Chemical and Metabolomic Analyses of Cuprizone-Induced

Home , Benmoxin, Octamoxin, Phenoxypropazine

CHEMICAL AND METABOLOMIC ANALYSES OF CUPRIZONE-INDUCED

DEMYELINATION AND REMYELINATION

A Thesis Presented to The Graduate Faculty of the University of Akron

In Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in Chemistry

Alexandra Taraboletti May 2017

CHEMICAL AND METABOLOMIC ANALYSES OF CUPRIZONE-INDUCED

DEMYELINATION AND REMYELINATION

Alexandra Taraboletti

Dissertation

Approved: Accepted:

______Advisor Department Chair Dr. Leah Shriver Dr. Christopher Ziegler

______Committee Member Dean of College Dr. Richard Londraville Dr. John Green

______Committee Member Dean of Graduate School Dr. Sailaja Paruchuri Dr. Chand Midha

______Committee Member Date Dr. Chrys Wesdemiotis

______Committee Member Dr. Christopher Ziegler

ABSTRACT

The cuprizone intoxication model of demyelination and remyelination has long been used to test myelin regenerative therapies for neurodegenerative diseases such as multiple sclerosis. Mice develop reversible and region-specific oligodendrocytosis and demyelination when fed this small molecule copper chelator. While the histopathology of the model has been well documented, to date, there is no consensus on cuprizone’s cellular mechanism of action, and whether it involves the chelation of copper in vivo.

Additionally, a variant model has been employed that combines rapamycin with cuprizone, causing more robust demyelination. In this model, rapamycin is hypothesized to suspend oligodendrocyte differentiation via the disruption of mTOR. In this work, the spatial and temporal effects of cuprizone were extensively studied in the in vivo model, and compared to a newly developed in vitro model. Global metabolomic and lipidomic profiling were utilized as analytic tools to determine biochemical pathways altered by cuprizone, rapamycin, or cuprizone and rapamycin treatments. Furthermore, oligodendrocyte differentiation in relation to mTOR was explored with transcriptomics.

Together this work utilizes OMICS technology paired with analytical techniques to investigate both the chemical nature of cuprizone and the biological function of the oligodendrocyte, gaining insight into the molecular interplay that occurs during the processes of demyelination and remyelination.

iii

DEDICATION

“Isn’t nice to know a lot? And a little bit not.”

-Stephen Sondheim (I Know Things Now, Into the Woods)

ACKNOWLEDGMENTS

The whole of this research has taken me on a long journey, starting in a seemingly empty lab as a one initial graduate student, and culminating in a group that has almost tripled. The lab has so much support and is filled with exciting energy, all in pursuit of many different scientific hypotheses. I was honored to start my academic path alongside my advisor and mentor Dr. Leah Shriver, who has instilled in me an extensive and diverse amount of knowledge. One day I too hope to have as detailed a grasp on human metabolism as her. I also have been shaped by my wonderful fellow students and lab- mates He Huang, Celina Cahalane, Hannah Baumann, and Rashmi Binjawadaji that I have had the honor to work alongside in our lab. I thank you especially for your endless support, and all of the coffee and snacks we have shared. I also want to thank my immediate family, Eva, Sabrina, and Matthew, as well as all of my extended family for their encouragement throughout my academic career; I hope I’ve made you proud. Thank you as well to all the new and wonderful friends I have made in my move to Ohio. I am so delighted to have a met a brilliant team of adventurers and fellow nerds in Mena

Klittich, Jake Hill, Katherine and Erik Willett, and Carolyn and Jake Scherger. Thanks,

Shadowmere’s Avengers for all the late nights, great food, and laughs. I was also very fortunate to befriend an excellent group of fellow graduate students, including Megan

Klufas, Shaun Christe, Louis Ray, Paul Mallory, Dan Morris, Joel Caporoso, Roger Shi, v

Farai Gombedza and many others who have all been so supportive during this expedition; my growth as a person, both emotionally and mentally, has been directly tied to these friendships. Finally, the work on my thesis would not have been completed without collaboration with the knowledgeable members of Renovo Neural, gracious funding from the Choose Ohio First Bioinformatics scholarship, and support from our NIH 1R15

GM119074-01 grant.

TABLE OF CONTENTS

Page

LIST OF FIGURES ...... xii

LIST OF TABLES ...... xiv

LIST OF SCHEMES ...... xv

ABBREVIATIONS ...... xvi

CHAPTER I ...... 1

1.1 Multiple Sclerosis ...... 1

1.1.1 Neuropathological Characteristics of Multiple Sclerosis ...... 3

1.1.2 Remyelination ...... 7

1.1.3 Animal models of Multiple Sclerosis ...... 10

1.2 The cuprizone animal model ...... 13

1.2.1 Pathology and clinical symptoms ...... 14

1.2.2 Proposed cuprizone mechanisms ...... 19

1.3 Global metabolomics ...... 21

1.3.1 Mass spectrometry ...... 23

1.3.1.1 Separation ...... 23

1.3.1.2 Ionization ...... 25

1.3.1.3 Mass analyzer...... 27

vii

1.3.2 Applications ...... 29

1.3.2.1 Targeted metabolomics ...... 29

1.3.2.2 Untargeted metabolomics ...... 30

1.3.2.2.1 Bioinformatics...... 32

1.3.2.3 Lipidomics ...... 37

CHAPTER II ...... 40

2.1 Introduction ...... 40

2.2 Methods...... 43

2.2.1 Absorbance Spectroscopy ...... 43

2.2.2 NMR Spectroscopy ...... 44

2.2.3 Mass Spectrometry...... 44

2.3 Results and Discussion ...... 45

2.3.1 Absorbance spectroscopy of CPZ, CuCPZ, and active site mimics ...... 45

2.3.2 NMR Spectroscopy of CPZ, and active site mimics...... 49

2.3.3 Mass Spectrometry of CPZ, CuCPZ, and active site mimics ...... 52

2.4 Conclusions ...... 56

CHAPTER III ...... 57

3.1 Introduction ...... 57

3.2 Methods...... 61

viii

3.2.1 Chemicals ...... 61

3.2.2 Cell Culture ...... 62

3.2.3 Preparation of CPZ Solution ...... 62

3.2.4 Live/Dead Cell Viability Assay and MTT assay ...... 62

3.2.5 Immunofluorescence ...... 63

3.2.6 Cuprizone and CuCPZ Absorbance Assay ...... 64

3.2.7 SRM-MS Analysis of CPZ Uptake in Cells ...... 64

3.2.8 Cuprizone Treatment of Mice ...... 65

3.2.9 Metabolomic analysis ...... 65

3.2.10 Shotgun Lipidomics ...... 67

3.2.11 Expression and Purification of rhTDO ...... 67

3.2.12 Characterization of TDO with SDS-PAGE gel and SEC ...... 68

3.2.13 Kinetic Assays of rhTDO...... 69

3.2.14 Mass Spectrometry and Absorbance Spectroscopy of P5P and CPZ ...... 70

3.2.15 Preparation of a Schiff base from P5P and Oxalydihydrazide ...... 71

3.2.16 Analysis of Transaminase Activity in Cells...... 72

3.2.17 NMR analysis of P5P and CPZ ...... 72

3.2.18 Q-PCR ...... 73

3.2.19 Data Processing ...... 73

3.3 Results and Discussion ...... 74

3.3.1 Cellular Uptake and Toxicity of CPZ ...... 74 ix

3.3.2 CPZ induces metabolic dysregulation in vitro and in vivo ...... 76

3.3.3 CPZ alters nicotinamide production in vitro ...... 83

3.3.4 CPZ binds pyridoxal 5’-phosphate perturbing aminotransferase activity ...... 87

3.4 Conclusions ...... 90

CHAPTER IV ...... 92

4.1 Introduction ...... 92

4.2 Methods...... 97

4.2.1 Chemicals ...... 97

4.2.2 Cell Culture and Differentiation ...... 97

4.2.3 Animals and Primary Cell Cultures ...... 98

4.2.4 Confocal Imaging...... 99

4.2.5 Cell Counting ...... 99

4.2.6 RAP and C/R Animal Model ...... 100

4.2.7 Metabolomic and Lipidomic Analysis ...... 101

4.2.8 Microarray for Transcriptomic Analysis ...... 101

4.2.9 Data Processing ...... 101

4.3 Results and Discussion ...... 102

4.3.1 Meta-analysis of RAP, CPZ, and C/R in vivo Models ...... 102

4.3.2 Lipidomic Analysis of RAP, and C/R in vivo Models ...... 108

4.3.3 Transcriptomic Analysis of OPC Differentiation ...... 110

4.3.4 BCAAs alters MO3.13 morphological state ...... 115

4.4 Conclusions ...... 117

CHAPTER V ...... 118

REFERENCES ...... 124

APPENDIX A: THESIS SUPPLEMENTARY ...... 173

APPENDIX B: FLUORESCENT FLAVONOIDS FOR ER CELL IMAGING ...... 193

APPENDIX C: LETTERS OF APPROVAL ...... 205

LIST OF FIGURES

Figure 1.1: Neuronal and glial cells of the CNS...... 4

Figure 1.2: Composition of myelin in the white matter of the CNS ...... 6

Figure 1.3: Process of oligodendrocyte maturation ...... 9

Figure 1.4: Three different models of murine demyelination (TMEV, EAE, and CPZ) .. 12

Figure 1.5: Structure of CPZ ...... 13

Figure 1.6: Visual representation of analyte interaction under differing solvent conditions in a representative RPLC and HILC column ...... 24

Figure 1.7: Example PCA data with divergent groups, and overlapping groups ...... 35

Figure 2.1: CuCPZ absorbance decays in solution in a manner dependent on the

Cu(II):CPZ ratio ...... 46

Figure 2.2: Copper active site mimics bind CPZ, shifting absorbance ...... 47

Figure 2.3: NMR reveals CPZ as partially hydrolyzed in solution...... 50

Figure 2.4: NMR reveals amidic CPZ hydrogens interact with mimics B and R ...... 52

Figure 2.5: Mass spectrometry confirms the presence of partially hydrolyzed CPZ and

CuCPZ formed at a 1:2 copper:CPZ ratio...... 53

Figure 2.6: Mass spectrometry reveals complex formed between B and CPZ ...... 55

Figure 3.1: Depiction of demyelination in the CNS ...... 59

xii

Figure 3.2: Cellular uptake and toxicity of CPZ in vitro ...... 75

Figure 3.3: CPZ perturbs the metabolism of MO3.13 cells ...... 78

Figure 3.4: CPZ induces region specific alterations in metabolism in the central nervous system...... 80

Figure 3.5: CPZ perturbs lipid homeostasis in the corpus callosum, but not the spinal cord...... 82

Figure 3.6: Purified recombinant human tryptophan 2,3-dioxygenase (rhTDO) is not inhibited by CPZ ...... 85

Figure 3.7: Pyridoxal 5’ Phosphate is a potential in vivo target for CPZ ...... 88

Figure 3.8: CPZ treatment of MO3.13 cells results in a decrease in transaminase gene expression after 16 hours ...... 89

Figure 4.1: mTORC1 and mTORC2 Complexes ...... 94

Figure 4.2: Meta-analysis of C/R, RAP, and CPZ reveals model specific dysregulation

...... 103

Figure 4.3: RAP and C/R effect lipid homeostasis differently in the corpus callosum (CC) versus the spinal cord (SC) ...... 109

Figure 4.4: PMA induced MO3.13 oligodendrocyte alters transcription of genes associated with cell cycle, cytoskeletal organization, and lipid synthesis ...... 113

Figure 4.5: BCAA and PMA modulate oligodendrocyte differentiation in vitro and are sensitive to RAP ...... 116

xiii

LIST OF TABLES

Table 1.1: Hydrazide neurological enzyme inhibitors…………………………………...20

Table 3.1: Resolving Gel Solution……………………………………………………….68

Table 3.2: Stacking Gel Solution.………………………………………………………..69

xiv

LIST OF SCHEMES

Scheme 2.1: Structure of intact CPZ, the proposed monohydrolyzed ligand, and three potential CuCPZ chelation modes……………………………………………………….42

Scheme 3.1: Preparation of Schiff base complex………………………………………..71

Scheme 3.2: Tryptophan/Nicotinamide pathway………………………………………...84

Scheme 4.1: Catabolism of branched chain amino acids……………………………….106

Scheme 4.2: Dopamine/Serotonin metabolism…………………………………………108

ABBREVIATIONS

1H-NMR Proton NMR CE Capillary electrophoresis 4E-BP1 Eukaryotic initiation factor CE Collision energy 4E binding protein 1 CES Collision energy spread AADC Aromatic amino acid CGT Ceramide galactosyltransferase decarboxylase CNPase 2',3'-Cyclic-nucleotide 3'- AAR Amino acid response pathway phosphodiesterase ABAT GABA aminotransferase CNS Central nervous system ANOVA Analysis of variance COX Cytochrome c oxidase AST Aspartate aminotransferase CPZ Cuprizone B Copper bound 1,4,7- CPZ-R CPZ with one cyclohexane triazacyclononane hydrolyzed BBB Blood-brain barrier CRM Consecutive reaction BCA Bicinchoninic Acid monitoring BCAA Branched chain amino acids CSF Cerebral Spinal Fluid BCAT Branched-chain amino acid CuCPZ Copper-bound CPZ aminotransferase CuZnSOD Copper zinc superoxide BCKD Branched-chain α-keto acid dismutase dehydrogenase DAG Diacylglycerol C/R Cuprizone with rapamycin DBH Dopamine β-hydroxylase CAT Peroxisome-bound catalase DC Direct current CC Corpus callosum DEPTOR DEP domain-containing CDK Cyclin-dependent kinase mTOR-interacting protein CDKI Cyclin-dependent kinase DFT Density functional theory inhibitor DIS Disseminated ins space

xvi

DIT Disseminated in time hEGFR Human epidermal growth DMEM Dulbecco’s Modification of factor receptor Eagle’s Medium HILIC Hydrophilic interaction liquid DP Declustering potential chromatography EAE Experimental autoimmune HLA Human leukocyte system encephalomyelitis HP Hippocampus EI Electron ionization IACUC Institutional Animal Care eIF-2α eukaryotic initiation factor 2 and Use Committee alpha IDA Information Dependent EPR Electron paramagnetic Acquisition resonance IDO Indoleamine-2,3-dioxygenase ER Endoplasmic reticulum IFN-γ interferon-γ ESI Electrospray ionization IGF Insulin-like growth factor ETC Electron transport chain IL-7Rα Interleukin-7 receptor alpha FBS Fetal bovine serum chain FDR False discovery rate IR Infrared FGF Fibroblast growth factor KIC α-ketoisocaproic acid FKBP12 FK-binding protein 12 KIV α-ketoisovaleric acid FTICR Fourier transform ion KMV keto-β-methylvaleric acid cyclotron resonance LC Liquid chromatography GC Gas chromatography LPS Lipopolysaccharide GPDH Glycerol phosphate LTQ-Orbitrap Linear quadrupole ion dehydrogenase trap-Orbitrap GPx Glutathione peroxidase MAG Myelin-associated GS Galactocerebroside glycoprotein GSH Reduced glutathione MALDI Matrix-assisted laser GWAS Genome-wide association desorption ionization studies MAO Monoamine oxidase HBSS Hank’s buffered saline MBP Myelin basic protein solution MHC Major histocompatibility complex

xvii

MLST8 Mammalian lethal with PC Phosphatidylcholine SEC13 protein 8 PCA Principal Component Analysis MnSOD Manganese superoxide PDGF Platelet-derived growth factor dismutase PKB Protein kinase B MOG Myelin oligodendrocyte PKC Protein kinase C protein PLD Phospholipase D MRM Multiple reaction monitoring PLP Proteolipid protein MS Multiple Sclerosis PLS-DA Partial Least Squares mSIN1 Mammalian stress-activated Discriminant Analysis protein kinase interacting protein 1 PMA 4-b-phorbol-12-myristate-13- MS/MS Tandem mass spectrometry acetate MT metallothioneins PNPO Pyridoxamine 5'-phosphate mTOR Mammalian/mechanistic oxidase target of rapamycin PPMS Primary progressive MS mTORC1 mTOR complex 1 PRAS40 Proline-rich AKT1 mTORC2 mTOR complex 2 substrate 1 NAWM Normal-appearing white PRM Parallel reaction monitoring matter PRMS Progressive-relapsing MS NCX Sodium/calcium exchangers Q1 First quadrupole NFκβ Nuclear factor kappa β Q2 Second quadrupole NMII Non-muscle myosin II Q3 Third quadrupole NMR Nuclear magnetic resonance Q-TOF quadrupole-TOF NOESY Nuclear Overhauser QTrap triple-quadrupole ion trap enhancement spectroscopy R Copper bound 1,8-dithia-4,11- NPLC Normal phase liquid diazacyclotetradecane chromatography RAP Rapamycin OPC Oligodendrocyte progenitor RAPTOR Regulatory-associated cell protein of mTOR P5P Pyridoxal 5’-phosphate Rb Rb tumor suppressor protein PA Phosphatidic acid RF Radio frequency PBS Phosphate-buffered saline rhTDO Recombinant human TDO

xviii

RICTOR Rapamycin-insensitive Sox SRY box containing companion of mammalian target of SPMS Secondary progressive MS rapamycin SRM Selective reaction monitoring RNS Reactive nitrogen species TDO Tryptophan-2,3-dioxygenase ROS Reactive oxygen species TEMED RPLC Reversed phase liquid Tetramethylethylenediamine chromatography THF Tetrahydrofolate s murine׳RRMS Relapsing-remitting MS TMEV Theiler RT Retention time encephalomyelitis virus S/N Signal to noise TNF Tumor necrosis factor S6K1 p70-S6 Kinase 1 TOF Time of Flight SAM S-adenosylmethionine TQ Triple quadrupole SEC Size exclusion chromatography XIC Extracted ion chromatogram SGK Serum-and glucocorticoid- XPS X-Ray photoelectron induced protein kinase spectroscopy SNPs Single-nucleotide polymorphisms

xix

CHAPTER I

INTRODUCTION

1.1 Multiple Sclerosis

Multiple sclerosis (MS) is a chronic autoimmune disease affecting the central nervous system (CNS).1 The pathology of this condition is characterized by immune cells that recognize and attack the myelin sheath. This tissue insulates nerves in the brain and spinal cord and loss of the sheath leads to the formation of lesions that are detectable by MRI .2,3 It is the most prevalent autoimmune disease of the CNS, affecting up to 2.3 million people per year, and is a leading cause of neurological disability in young individuals.4 The age of onset for the disease is between 20 and 50, and it is twice as common in women as in men.5,6 The cause of this disease is unknown; however, the risk of acquiring this complex disease currently is suggested to arise from a combination of genetic7 and environmental factors.8

MS presents as a broad spectrum of neurological symptoms, with the most common problems being visual, sensory, motor, autonomic and neuropsychiatric.3

Varying clinical outcomes with respect to disease progression have been described.9,10

Relapsing-remitting MS (RRMS) is the most common type. Patients experience reoccurring episodes of disability between periods of recovery. Among these patients, an initial course of relapses and remissions can be followed by a secondary-progressive stage (SPMS). During this phase of the disease, patients experience multiple episodes of disability that lead to a steady decline in function. A small percentage of patients initially and slowly accumulate neurological symptoms with no recovery phase (Primary

Progressive MS, PPMS) while some can exhibit acute relapses in between the progressive symptoms (Progressive-relapsing MS, PRMS). Though unpredictable, some relapses are proceeded by common triggers including viral infection, stress, and even seasonal deviations.3,11,12

MS is not regarded as a hereditary disease, though the probability of developing it is higher among relatives.3,13 The familial recurrence rate is about 15-20%, and the risk of disease is greater between those who are more closely related.14–16 Present information indicates that the disease results from combined genetic vulnerabilities, as opposed to

Mendelian inheritance.17 Some genetic variations are highly associated with increased risk, specifically, allele variations related to the human leukocyte system (HLA) on chromosome six, and its effects on the major histocompatibility complex (MHC).18–21

The HLA complex aids in immune system regulation via antigen presentation to T- cells.22 Between 20-60% of the genetic causes of MS can be attributed to changes in the

MHC.18 Single-nucleotide polymorphisms (SNPs) in the interleukin-7 receptor alpha chain (IL-7Rα) gene recently has been associated with a higher disease risk.23,24 IL-7Rα encodes the membrane-bound IL-7 receptor protein, aiding in the development and

1 maturation of lymphoid progenitor cells and as a survival factor for CD4 and CD8 T- cells. Variations associated with the disease cause the internalization of the IL-7 receptor protein causing a higher ratio of the soluble protein.24,25 A recent investigation confirmed higher levels of IL-7 mRNA in the cerebral spinal fluid (CSF) of patients with MS.26

Interestingly, IL-7Rα also is associated with other autoimmune diseases, including type I diabetes mellitus and rheumatoid arthritis.27 Other genes with modest relation have also been identified via genome-wide association studies (GWAS),18,28 including those expressing T-cell receptor β,29 cytotoxic T lymphocyte antigen 4,30 GPC5,28 TNFRSF1A,

IRF8, and CD6.31

Environmental factors, including several infectious agents, have been implicated as triggers to MS. Several virological studies implicate herpes, measles, mumps, rubella, and Epstein–Barr, though no conclusive evidence exists to tie them the disease.32–36 The strongest evidence for a viral connection is that approximately 90% of patients with MS have high concentrations of IgG and oligoclonal bands.37,38 When elevated levels of IgG in other diseases were studied, it was found to have high specificity for the disease of interest, providing a rationale that oligoclonal IgG in MS is specifically directed at the cause of the disease. Furthermore, it is postulated that the early exposure to these viral infections may be protective, in contrast to exposure as a young adult.35,39 Other indirect support for a viral association include the phenotypic similarities to demyelinating encephalomyelitis found in human and animal infections.40,41 Possible mechanisms include the exposure of myelin proteins during the viral assault to the CNS, resulting in mimicry between myelin and viral antigens, causing a direct immune response on the myelin sheath.42–44

Some evidence supports a geographical link to MS. Currently, the incidence of

MS increases with increasing distance from the equator.13,45 However, recent studies suggest that this latitude gradient is decreasing.5 The geographic distribution pattern may also simply result from the natural distribution of high-risk populations. Ultraviolet light, which is strongly connected to latitude, has been hypothesized to play a protective role, potentially mediated through vitamin D synthesis.46 Low vitamin D levels in patients with MS have been documented in several studies, though evidence for its effect on disease progression is lacking.47,48

1.1.1 Neuropathological Characteristics of Multiple Sclerosis

The characteristic neuropathological alteration of MS is the formation of lesions or plaques in the CNS disseminated in time (DIT) and space (DIS).49 These plaques are the result of rounds of inflammation, oligodendrocyte death, demyelination of nerves, disruption of the blood-brain barrier (BBB) and subsequent remyelination.50–53 Damage occurs to both neurons and glial cells. Glial cells include microglia, oligodendrocytes, and astrocytes. These cells provide distinct roles in the nervous system, including metabolic support, neurotransmitter uptake, structural support, and protection (Figure

1.1).54 In particular, astrogliosis in chronic lesions results in scar formation distinct to

MS.55,56

Disruption of the BBB is one characteristic of the pathology that occurs in MS.57

The BBB is formed by endothelial cells, bound by tight junctions, which line the vasculature in the CNS. The tight junctions of the BBB are formed by occludin and claudin, making the barrier highly selective and semipermeable, which restricts microscopic, large, and hydrophilic objects from entering the CSF.58,59 The migration of

Figure 1.1: Neuronal and glial cells of the CNS. Illustration depicts neuron (yellow) wrapped by myelin producing oligodendrocyte cells (blue) and interacting with the astrocyte (green) support cells, and immune functioning microglial cell (purple).

lymphocytes and monocytes into the surrounding tissue initiates the disruption of the

BBB. Activated adhesion molecules on the lymphocytes and monocytes interact with endothelial cells and cause the degradation of the BBB through matrix metalloproteases.60 It is still disputed whether the breakdown and dysfunction of the

BBB are causative or consequential of MS, though it is a distinct element involved in most cases of the disease.43

Loss of oligodendrocytes and subsequent demyelination are main hallmarks of the disease. Oligodendrocytes are glial cells responsible for neuronal insulation and support.61 These cells wrap around neuronal axons with an extended plasma membrane

4 called the myelin sheath, which is comprised of approximately 20% protein and 80% lipid.62 The myelin composition includes specific proteins such as myelin basic protein

(MBP), myelin oligodendrocyte glycoprotein (MOG), myelin-associated glycoprotein

(MAG), and proteolipid protein (PLP) and CNPase.62 Cholesterol, galactocerebroside, and sphingomyelin are essential lipid components of myelin (Figure 1.2).63 Myelinating oligodendrocytes encompass the white matter of the CNS, thus plaques are typically found in periventricular and juxtacortical white matter. Plaques can involve the cerebral cortex, cerebellar, brain stem, spinal cord, as well as optical nerve tracts.64

Detailed histopathological studies highlight the heterogeneity of the myelin destruction and lesion formation found in patients.65,66 Lesions can be segregated into four patterns of myelin destruction, possibly correlating to a different disease prognosis and course, all associated with an inflammatory process (composed mainly of T-cells and macrophages). Pattern I lesions are characterized by macrophage-associated demyelination. Scars are dominated by T-cells and macrophages, with intact oligodendrocytes. Pattern I lesions most closely resemble the strictly immune-mediated murine models of autoimmune encephalomyelitis in which activated macrophage products such as tumor necrosis factor-α (TNF-α) act to destroy myelin. Pattern II lesions resemble pattern I lesions, though are considered antibody mediated. These lesions additionally show signs of complement activation and the accumulation of immunoglobulin.67 Pattern II lesions most resemble experimental autoimmune encephalomyelitis (EAE) induced by the sensitization of antibodies with MOG.68–70 Both pattern I and II lesions are usually centered around veins or venules, have distinct boundaries, and show loss of myelin-specific proteins such as MAG, MOG,

Lipid Bilayer

PLP MOG

Myelin Cut-away MBP CNP

MAG

Figure 1.2: Grey matter is composed of glial cells, unmyelinated axons, and neuronal cell bodies, while white matter is composed primarily of long myelinated axons wrapped by oligodendrocyte cells. Myelin is composed of the wrapped oligodendrocyte plasma membrane surrounding the neuronal axon. Myelinating oligodendrocyte cells express specific proteins such as PLP, MBP, MOG, MAG, and CNPase.

PLP, and 2',3'-cyclic-nucleotide 3'-phosphodiesterase (CNPase). In contrast, pattern III lesions are not well demarcated and do not appear around vasculature. Instead, a perimeter of intact myelin usually surrounds inflamed blood vessels, all found within lesions. Pattern II lesions are dispersed with inflammatory elements such as macrophages, T-cells, and microglial cells. Within these areas, a gradient of oligodendrocyte loss can be found. Plaques contain total loss in the inactive center and progressive loss at the rim. The preferential loss of the myelin protein MAG is characteristic to these lesions. This type of distal oligodendrocyte-associated demyelination found in pattern III mimics lesions are often associated with virus-induced

6 human white matter diseases. Pattern IV lesions consist primarily of oligodendrocyte degeneration. These lesions are the least common and are usually isolated in patients with PPMS. The lesions have distinct borders with a boundary of normal-appearing white matter (NAWM) and have a complete absence of oligodendrocytes in the center of the plaques. Though speculative, oligodendrocytes found in pattern IV lesions could be susceptible to inflammatory mediators purely through metabolic disturbance.65–67

The distinct differences in these patterns make MS a heterogeneous disease, though some evidence proposes temporal heterogeneity as the root of distinct lesion types.71,72 Nonetheless, the four-pattern model is still largely recognized, and as such, different antibodies and biomarkers are being explored to identify the pathological subtypes. For example, mitochondrial damage to the respiratory chain complex IV

(cytochrome c oxidase, COX), and the subsequent lack of COX-I can distinguish pattern

II and III lesions.73 Another recent study used antigen microarrays to identify unique serum signatures of the pathological subtypes.74 Patients with pattern I lesions could be distinguished based on a heightened reactivity against oxidized derivatives of cholesterol, heat shock protein, and CNS antigens. Subjects with pattern II lesions instead displayed an increased reactivity to heat shock protein 60, oligodendrocyte-specific protein, MOG, and PLP peptide epitopes.

1.1.2 Remyelination

Remyelination occurs in the inactive plaques of MS. It is the phenomenon in which new myelin sheaths are produced around intact axons after a demyelinating event.

An acute inflammatory response simultaneously functions to promote debris clearance of damaged myelin and destroyed oligodendrocyte cells through phagocytosis.75

Remyelination immediately reestablishes saltatory conduction by the reallocation of ion channels at the nodes of Ranvier,76 lowers axonal mitochondrial content due to a partially resolved metabolic deficit,77 and can help recover behavioral and physical damage.78

Regenerated myelin is notably thinner than when formed during development, and after numerous cycles of demyelination and remyelination axons are left incapable of properly repairing the damage, lowering the conduction velocity, and potentially leading to axonal degradation.51,79–81 Axonal degradation is far more debilitating than demyelination, as axons cannot be regenerated, and thus most research focuses on therapeutically promoting remyelination.82 Mature oligodendrocytes spared during demyelination do not proliferate in the presence of demyelinated axons, nor do they participate in remyelination.83 Instead, oligodendrocyte progenitor cells (OPCs) are responsible for the myelin repair process. 84–87 Recruitment of OPCs is one of the first steps of remyelination, moving from adjacent white matter or from the subventricular zone.88–90

OPCs transition from a dormant to regenerative phenotype enhanced by the presence of astrocyte and microglial cells.91,92 OPCs progress through a series distinct morphological stages, starting as isolated bipolar early progenitors, differentiating into multipolar late progenitors, becoming postmitotic immature cells, and eventually developing and extending their membranes to change into mature oligodendrocytes expressing PLP,

MBP, and CNPase and other proteins (Figure 1.3).93–96 Though the distinct signaling pathway that controls these specific transitions is poorly understood, many growth factors, transcription factors, cytokines, chemokines and other signaling molecules impact OPC differentiation and remyelination.97 Human epidermal growth factor receptor

(hEGFR), insulin-like growth factor (IGF)-1, platelet derived growth

OPC Pre-oligodendrocyte Immature Myelinating PDGF-α PDGF-α Oligo 1/2 Oligo 1/2 MBP NG2 NG2 Sox10 Sox10 MAG Oligo 1/2 Oligo 1/2 O4 O4 MOG Sox10 Sox10 CNPase CNPase PLP O4 GalC GalC

Figure 1.3: Process of oligodendrocyte maturation, from OPC to mature myelinating cells. In each step

of differentiation cellular structure and morphology is unique.

factor (PDGF), and fibroblast growth factor (FGF) are all implicated in increasing OPC proliferation and accelerating remyelination.98–100 Growth factor stimulation of the

Ras/Raf/Mek/Erk and P13K/Akt/mTOR pathways is connected to the proliferation, migration, and survival of oligodendrocytes.101–110 Erk1/2 signaling, specifically, aids the transition of early progenitors to late progenitors, while mTOR signaling is required for the transition of immature oligodendrocyte cells to mature cells.111 Basic-helix-loop- helix (bHLH) transcription factors Olig1, Olig2, and Nkx2.2 all have designated roles in the development, maturation, and specification of oligodendrocyte cells.112 The disruption of Olig1 and Olig2 in progenitor cells halts the advancement of oligodendrocyte maturation, indicating a need for Olig-specific progenitors during development.113 Another transcription factor SOX10, is involved in terminal oligodendrocyte differentiation and can complex with Oligo1 to activate the transcription

9 of MBP.114 Additionally the cytokine TNF-α can play role in remyelination via TNFR2 signaling.115,116 TNFR2 binds membrane-bound TNF-α activating a survival and protective pathway mediated by nuclear factor kappa β (NFκβ).117 Many chemokines, expressed by reactive astrocytes, including CXCL12, and CXCL1 contribute to migration and differentiation during remyelination and repair of the CNS.118–120 CXCL12 mediates

OPC differentiation into mature oligodendrocytes in the CC through the GPCR CXCR4, and the neutralization CXCL12 prevents oligodendrocyte migration and proliferation.121–

123 The overexpression of CXCL1 in astrocytes, activating CXCR2, decreased EAE- induced demyelination and expedited remyelination in mice, however, in CXCR2−/− chimeric mice OPC proliferation occurred faster in lesions.124–126 Other molecules such as non-muscle myosin II (NMII)127,128 and retinoic acid129 have also been implicated in OPC differentiation and remyelination. Though it is clear that the CNS is normally capable inducing a response to myelin damage via many different actions, it is still unknown why patients with MS exhibit partial or failed remyelination, but the presence of OPCs and premyelinating oligodendrocytes is suggestive of an obstruction in OPC differentiation.130

1.1.3 Animal models of Multiple Sclerosis

As MS is a complex disease that involves two intricate biological systems: the immune system and CNS, animal models have been critical for comprehending pathogenesis and therapeutic development.131 However, the models of MS are innately and individually flawed, as each is unable to encapsulate the breadth and heterogeneity of the disease.132,133 Rodent immune systems are functionally distinct from humans, producing pathologies that can also manifest at a different rate than in humans.134,135

Though animal models have clear differences from human disease, several FDA- approved MS therapies have been developed by using them.136–138 Three distinct animal

s murine encephalomyelitis virus׳models are well characterized and employed: Theiler

(TMEV) infection, EAE (experimental autoimmune encephalomyelitis), and toxin- induced cuprizone (CPZ) demyelination (Figure 1.4).

TMEV is a neurotropic viral infection model for MS, induced by a non- enveloped, positive-sense, single-stranded RNA virus.139,140 The model produces an acute encephalomyelitis or chronic progressive inflammatory demyelinating disease state that is specific to mice. Two viral strains (GDVII, and DA) are commonly used and have discrete disease pathogenesis.141,142 GDVII is a neurovirulent strain with a high mortality rate (1-2 weeks). Neurons are predominantly affected in the absence of mononuclear cell recruitment, producing no demyelination.143–146 No viral persistence occurs after clearance in surviving mice. In contrast, the DA strain induces acute polioencephalomyelitis followed by chronic encephalomyelitis. During the acute phase of the DA infection, both CD4 and CD8 phenotypes infiltrate the gray matter of the brain, causing axonal damage, inflammation, and swelling.147–152 The chronic phase occurs approximately a month after infection, in which demyelination appears distributed in areas where the previous axonal damage occurred.143,147,149,153,154 This evidence suggests that an inflammatory response in the CNS triggered by initial axonal degeneration can precede the resulting loss of myelin.

EAE is widely used as an animal model of MS, and it is studied as a general example of T-cell-mediated autoimmune diseases. The model proceeds through the immunization of rodents (mice155, rats156 or guinea pigs157) or non-human primates158,159

TMEV GDVII/DA 1-2 weeks: Picornaviridae virus GDVII: Neuronal destruction, high mortality rate. DA: Axonal damage, inflammation, swelling, and demyelination.

EAE MOG/PLP/MBP 9-12 days: +(pertussis toxin) Nerve inflammation, demyelination, and paralysis. High rate of mortality or permanent paralysis.

5-6 weeks: CPZ 0.2%-0.5% CPZ Fed in chow Weight loss, lethargy, megalomitochondria non-immune demyelination of CNS white matter. Removal allows spontanious remyelination.

Figure 1.4: Three different models of murine demyelination (TMEV, EAE, and CPZ). The pathogenesis and pathology of each model differs, and are appropriately listed above.

with self-antigens injected subcutaneously, and derived from myelin. Purified myelin or whole myelin protein/peptides from MBP, PLP160, and MOG161 are commonly used as antigens, each presenting immunologically and pathologically as different models. T- cells already reactive towards the myelin antigens can also be introduced to induce

EAE.162 Animals can display acute monophasic, relapsing-remitting, or chronic acquired, inflammatory demyelinating autoimmune disease based on the antigen used.163,164

Activated myelin-antigen specific CD4+ T cells induce the breakdown of the BBB through the release adhesion molecules, cytokines, and chemokines, and migrate into the

CNS. Pertussis toxin can be used to help facilitate the initial BBB breakdown.165 Antigen presenting cells of the CNS are then recognized by the myelin antigen-specific T cells and begin demyelination and axonal loss.166,167 Disease onset usually occurs after 9-12

12 days, typically causing severe nerve inflammation, demyelination, and paralysis

(predominantly of the tail and hind limbs).168 Similar to MS, the disease symptoms are a reflection of the physical location of active lesions, which are primarily contained to the spinal cord in most models.169–171 Though EAE embodies many aspects of MS, the immune specific stimulation does not fully encapsulate the heterogeneity of the disease.172

1.2 The cuprizone animal model

Cuprizone (Bis(cyclohexanone)oxaldihydrazone), CPZ) is a small molecule made by condensing oxalylhydrazide and cyclohexanone, that was first illuminated in the

1950s by Nilsson for its analytical ability to quantify copper (II) ions through

chelation.173 The resulting copper-cuprizone

complex (CuCPZ) forms a distinct royal blue

solution with an absorption band centered at 595

nm, though to date the structure of this complex

Figure 1.5: Structure of CPZ.203 has yet to be fully elucidated. Renewed interest surfaced surrounding CPZ in 1966 when Carlton reported microscopic lesions, edema, hydrocephalus, demyelination, and astrogliosis found in the CNS after feeding mice a range of basic chow consisting of 0.2-0.5% CPZ.174 Originally Swiss mice were used to study CPZ demyelination.174–178 However, other mice strains (Albino, BALB/c, BSVS,

CD1, ICI and SJL) and rodent species (Albino and Wistar rats, guinea pigs and Syrian and Chinese hamsters)179–186 are known to exhibit a variable degree of demyelination.

Initial clinical observations made in the CPZ mouse model connected the dose and treatment length to the symptom severity and mortality rate.174,178,184,187–190 The age of the

13 mice also impacted the treatment; as juvenile mice are more susceptible to demyelination than older mice.174,184 Experimentations by Hiremath in 1998 worked to reduce model variability and clinical toxicity by feeding eight-week-old C57BL/6 mice a diet containing 0.2% CPZ consistently for six weeks.188 Two dominant setups based on

Hiremath’s work are now used often to induce either acute or chronic demyelination

(along with microgliosis and astrogliosis) by feeding mice 0.2% CPZ for six or twelve weeks, respectively. Unique to the CPZ model is spontaneous remyelination that can occur after removal of CPZ-containing chow. Remyelination is frequently studied after the six-week demyelination model, as the chronic twelve-week model causes inadequate and deficient remyelination.191

1.2.1 Pathology and clinical symptoms

Following CPZ intoxication, mice exhibit weight loss, reduced defecation, increased diuresis, lethargy, and altered sleep patterns, and though less common, some cases of ataxia, flaccid tail, and hind limb paralysis have also been noted.187,189,190,192–196

Ingestion of CPZ in mice induces very reproducible localized white matter tract demyelination of the CNS. Demyelination is found in the corpus callosum (CC), hippocampus (HP), cerebellum, and the caudate putamen.197–200 The grey matter of the cortex and hippocampus formation are also affected.196 As the CC becomes severely demyelinated, mice begin to exhibit impaired motor coordination.194

The model differs from TMEV and EAE in that it is not immune-mediated, instead, demyelination results from the degeneration of mature oligodendrocytes as opposed to a specific attack on myelin sheaths. Recent experiments treating rat primary glial cells with CPZ proved that toxicity was limited to mature oligodendrocytes

(decreased survival and lowered mitochondrial transmembrane potential) and not microglia, astrocytes, or OPCs.201,202 The death of myelinating oligodendrocytes in vivo is accompanied by the invasion of activated residential microglial cells and macrophages, astrogliosis, and an increased production of inflammatory cytokines.203,204

Histopathological features of the CPZ model closely resemble pattern III lesions found in

MS patients.66,73,202,205 Myelin loss in the model is preceded by the down-regulation of myelin-related proteins, such as MAG, and continues until CPZ treatment is halted.

Lesions have ill-defined borders and are not perivenous. In contrast to pattern III lesions from acute stages of MS, there are no signs of an adaptive immune response in the CPZ model.

Megamitochondria is a well-classified feature of the CPZ model found only in hepatic and oligodendrocyte cells, although neuronal, astrocyte, cardiac, kidney, Kupffer, and fat-storing cells have been examined.184,206–210 Hepatic megamitochondria have only been observed in chronic models using 0.5%, while oligodendrocyte cells quickly exhibit megamitochondria after three weeks of 0.2% CPZ. When CPZ is halted, mitochondria regain normal morphology, and enlargement ceases.188,211,212 In vitro mitochondrial enlargement usually occurs as a protective process to reduce oxidative stress, in response to elevated reactive oxygen species (ROS) and reactive nitrogen species (RNS),

213 •– eventually forming megamitochondria. The increase in O2 production after electron leakage from the electron transport chain (ETC) during CPZ treatment is reportedly due to the uncoupling of the oxidative phosphorylation process,210,214 and an inhibition of

Complex I–III, Complex II–III, and Complex IV.176,202,207,209,215–217 A recent paper

15 contests the inhibition of Complex I–III, and Complex II–III, claiming only Complex IV is inhibited by CPZ intoxication.218

Increased oxidative stress induced by CPZ causes numerous cellular changes contributing to the specific vulnerability of oligodendrocytes. Oligodendrocytes express only low amounts of metallothioneins (MT), which act as antioxidants and provide protection against metal toxicity, potentially increasing their susceptibility to oxidative stress.219 Astrocytes contrastingly show increased MT1 and MT2 expression following

CPZ treatment.212 Peroxisome-bound catalase (CAT) and glutathione peroxidase (GPx) are reduced following CPZ treatment.220,221 CAT and GPx are located in peroxisomes and aid in the reduction of the high amounts of hydrogen peroxide produced through β- oxidation and lipid synthesis.222,223 GPx specifically works through the use of the metabolite glutathione (GSH). Intrinsic levels of GSH are decreased specifically in oligodendrocytes, and a temporary decline in reduced GSH levels was reported after

CPZ-induced demyelination.212,224 The scavenging enzymes responsible for partitioning

•– of O2 , copper zinc superoxide dismutase (CuZnSOD), and manganese superoxide dismutase (MnSOD) correspondingly display decreased activity during CPZ treatment.215,218,221,225–227 Glycerol phosphate dehydrogenase (GPDH) levels are additionally altered in response to CPZ oxidative stress.212 GPDH is a contributor of electrons to the ETC, maintains cellular redox potential, and plays a major role in lipid biosynthesis.228,229 The wide array of enzymes presented solidifies the role of CPZ to produce oxidative stress distinctly in oligodendrocytes, arising in megamitochondria.

Oxidative stress, ROS production, and low levels of ATP after CPZ treatment eventually disrupts the endoplasmic reticulum (ER) of oligodendrocytes, causing

16 disturbed lipid and protein synthesis.230,231 A proper functioning ER and subsequent protein synthesis are dependent on amino acids levels. Reduced alanine, proline, and glycine were observed in the plasma after two weeks of CPZ intoxication, potentially activating the amino acid response pathway (AAR).232 The AAR activates eukaryotic initiation factor 2 (eIF-2α) resulting in decreased translation and successive transcription.

The mRNA levels of myelin protein genes and genes associated with lipid function drop as a consequence of ER stress. Both ceramide galactosyltransferase (CGT) for galactocerebroside synthesis and HMG-CoA reductase for cholesterol synthesis are examples of such enzymes found reduced in oligodendrocytes; the levels of CGT and

HMG-CoA also increase after CPZ is removed, during remyelination.187,233–235

Oligodendrocytes begin to display honeycomb vesiculation of the myelin, and eventual perikaryon disruption, causing oligodendrocyte degeneration only some days following

CPZ treatment.178,230,231,236,237

Early oligodendrocyte apoptosis in the CPZ model starts by means of a primary degeneration pathway from oxidative stress, ER dysfunction, and myelin degeneration, rather than an autoimmune-mediated attack. The degeneration pathway triggers additional damage from the immune system and the massive oligodendrocyte apoptosis found by week four of CPZ treatment.238 After one week neutrophils are observed infiltrating the CNS lesions,239 and astrocytic enzyme activity increases.175,240,241 The activation of astrocytes progresses to astrogliosis in the white matter by week four and can continue through remyelination.242 Microgliosis occurs as early as two weeks into

CPZ intoxication and peaks at week four when the largest number of proliferating microglia can be observed. The peak of microgliosis coincides with the initiation of

17 remyelination and the maximum demyelination.188 After the initial cascade of damage, microglia are the principal cause for oligodendrocyte apoptosis. However, microglia have been described as having both pro-inflammatory (M1) and regenerative (M2) phenotypes.243,244 Microglia along with astrocytes seem to attempt to restore homeostasis by the phagocytosis of myelin debris and dead cells, recruitment of OPCs, clearance of excess fluids, and the establishment of tissue restoration, and in this context, the occurrence of astrocyte and microglial processes are a pre-requisite for remyelination to occur.191,199,212,243,245–247

Remyelination begins after week three as OPCs are left unaffected, accumulate, and begin to repopulate mature cells. Remyelination becomes evident at week six when

185,191,248 these cells reach maturation. During continuous CPZ treatment, a second wave of oligodendrocyte apoptosis and demyelination occurs around week eight, followed by another incidence of remyelination between weeks ten and twelve.199,249–251 However, the second remyelination event is much smaller in comparison. OPCs found repopulating mature cells are either endogenous to the demyelinated tissue or originated in and migrated from peripheral tissues. Following the accumulation of OPCs to sites of damage, transcription factors and other proteins direct their differentiation into mature oligodendrocytes. In particular, the transcription factors Olig-2 and Olig-1 along with its downstream effector Zinc finger protein 488 are critical to elongation, branching, and maturation.252–254 Another zinc finger-containing transcription factor, Kruppel-like factor

9, is also key to effective differentiation.255 Maturation in the CPZ model is also dependent on signaling by golli-myelin basic protein and stimulation by BDNF.246,256,257

Additionally, immune-cell derived molecules such as galactin 3, GLU, and CXCL12

18 signaling via CXCR4; all can affect oligodendrocyte maturation.123,258,259 The development of oligodendrocyte cells and consequent myelination also appears intimately regulated by mTOR pathways. The addition of the mTOR inhibitor rapamycin

(RAP), as well as siRNA-mediated knockdown of mTOR,1 halts OPC differentiation in vitro.260–262 A variant model where RAP is simultaneously introduced with CPZ to the animal causes robust demyelination as it targets both OPCs and mature oligodendrocytes.263

1.2.2 Proposed cuprizone mechanisms

While the histology of CPZ demyelination is well documented, no real consensus exists concerning the mechanism of CPZ demyelination. To date, two main hypotheses exploring CPZ toxicity have been investigated. The first hypothesis purports pathological effects of CPZ to be the result of a disturbance in copper homeostasis via chelation.176,199,264 It is proposed that copper depletion in the brain, causes neurological signs paralleling inherited copper metabolism disorders such as Menke’s and Wilson’s disease. CPZ treatment decreased brain dry weight and copper content. In addition, CPZ could not be detected in the brain and liver by mass spectrometry while the gut showed the formation of Cu(II) oligomers.176,265 Based on this evidence, it has been proposed that

CPZ cannot enter the brain and instead copper specific enzymes (such as monoamine oxidase and cytochrome c oxidase) are altered via chelation. The copper deficiency model is contested by opposing claims of increased copper and zinc levels in the brain after dosing mice for nine months with 0.2% CPZ.241 Additionally, the supplementation of up to 260 ppm copper along with 0.2% CPZ did not ameliorate lesions of edema, spongy degeneration, astrogliosis, or megamitochondria.177,266 Interestingly, when the

Table 1.1: Hydrazide enzyme inhibitors

Activity Compound Chemical Structure

Inhibition of 3-hydroxy

GABA Aminotransferase benzylhydrazine

Irreversible Benmoxin

and nonselective MAO

inhibitors Iproniazid

Isocarboxazid

Mebanazine

Nialamide

Octamoxin

Phenelzine

Phenoxypropazine

CuCPZ complex was fed to mice, they exhibited no symptoms. As simple metal deficiencies do not encapsulate all of the physical characteristics produced by the CPZ model, the second hypothesis proposes CPZ’s neurotoxic effects are due to a metabolic interference and subsequent enzyme inhibition.199,203,241,267 The presence of CPZ in the brain and circulating in blood plasma as a free ligand have been documented, signifying its potential to directly affect various tissues. Furthermore, structural evidence suggests

CPZ hydrolysis freely occurs, liberating a cyclohexane ring and resulting in a monosubstituted hydrazide.203 Hydrazides are known to act as competitive neurological enzyme inhibitors.268–278 Much of the work involving the physicochemical nature of CPZ provides complex and variable information, proving further work is needed to elucidate a mechanism of action.

1.3 Global metabolomics

Metabolomics is the profiling of metabolites (molecules <1000 daltons) from biological systems, and is routinely used to identify biomarkers involved in disease or to identify altered system states.279,280 A major advantage of this application is the ability to detect hundreds to thousands of metabolites simultaneously, offering an efficient technique for monitoring biochemical alterations. Metabolomics is a sensitive method, and while genomics and proteomics provide substantial information regarding the genotype, metabolomics reveals phenotypic information.281 Perturbations or fluxes in metabolite concentrations happen quickly in response to stimuli, thus metabolomics offers dynamic experimental information (some methods demonstrating real-time information).282 Both untargeted metabolomics, which attempts to measure the complete set of metabolites in a biological system (the metabolome) using rapidly expanding data

21 acquisition and targeted metabolomics, which can quantitatively measure a class of metabolites, are techniques employed through numerous methods to assign the status of biological systems.

The two most common techniques utilized for metabolomic applications are nuclear magnetic resonance (NMR)283 and mass spectrometry.284 Mass spectrometry and

NMR methods complement one another, and various techniques for each method offer the detection and accurate identification of a variety of metabolites; along with the ability to correctly measure metabolite concentrations.285,286 NMR is a non-destructive technique that provides detailed structural information, does not depend on analyte polarity, but requires preparation using deuterated solvents.283,287,288 NMR also provides very reproducible quantitation, making it useful for metabolite concentration analysis.289

Simple proton (1H-NMR) and one-dimensional nuclear Overhauser enhancement spectroscopy (NOESY) are the most used NMR methods for metabolomic applications, though 13C NMR is often applied to isotopically-labeled flux experiments. While mass spectrometry is a destructive technique, it is intrinsically a highly sensitive detection method, with the lower limit of detection in the femtomole range, compared to the low nanomole range for NMR. In a single measurement, mass spectrometry allows for the structure elucidation, and quantification of several hundreds of metabolites, whereas

NMR typically detects ten to just over a hundred metabolites. Mass spectrometry, especially with liquid chromatography, is increasingly being used for metabolomic applications due to its larger metabolome coverage.290,291

1.3.1 Mass spectrometry

The analytical technique of MS can be coupled to various separation platforms, or used in a stand-alone mode. There are many variants of mass spectrometers, making MS- based metabolomic analysis highly dependent on the individual separation technique, source ionization method, and the means of detection. Each of these parts and their distinct merits are discussed below.

1.3.1.1 Separation

Metabolomic methods often require the use of a separation technique such as capillary electrophoresis (CE), gas chromatography (GC), and liquid chromatography

(LC) to due to the chemical complexity of the metabolome. LC and GC are the preferred methods in the field of metabolomics, though CE is gaining interest. All forms of chromatography utilize a stationary phase and a mobile phase. Compounds are carried with the mobile phase in a definite direction along the stationary phase and are retained by the stationary phase at different speeds depending on their affinity for the stationary and mobile phases. Differing affinities for the stationary and mobile phase allows for the separation of compounds from a mixture. The characteristic time it takes for a compound to elute is called its retention time. The retention time provides an additional data parameter to aid in metabolite identification.

GC achieves pronounced sample separation with a lowered risk of ion suppression, though it often requires chemical modification and can alter or destroy metabolites due to the heat associated with gas phase formation.292–294 GC uses a nonreactive carrier gas, such as helium, the mobile phase, which moves along the stationary phase comprised of a column with polymeric or liquid coated walls.295 GC

RPLC HILIC

Nonpolar Polar Analyte Analyte

Low High Low High Organic Organic

Figure 1.6: Visual representation of analyte interaction under differing solvent conditions in a representative RPLC and HILC column. RPLC: nonpolar analytes move quickly in the column under highly organic solvent ratios, but are retained by column hydrophobic chains when the organic solvent ratio drops. HILIC: polar analytes move quickly in the column under highly aqueous solvent ratios, but are retained by forming hydrogen bond interactions with head groups on chains residing in the aqueous layer when the aqueous solvent ratio drops. separation is useful for thermally stable and volatile metabolites, such as fruit and vegetable aromas, acids, esters, hydrocarbons, and short-chain alcohols.295–297 LC is often used for general metabolomic studies, in comparison to GC, as it can detect a large pool of intact metabolites without the use of chemical derivatization. LC works by separating injected samples on a stationary phase consisting of a column, packed with variable solids. Columns are commonly packed with derivatized silica or alumina.298,299 The mobile phase can include a single solvent or a mixture of solvents. Different elution methods enhance compound separation by controlling the solvent percent composition over time. An isocratic elution employs a single solvent or solvent mixture which remains constant through the analysis. A gradient elution is a separation where the mobile phase composition changes over time. Reversed phase liquid chromatography (RPLC) is traditionally used for the separation of nonpolar to slightly polar molecules for

24 metabolomics, though hydrophilic interaction liquid chromatography (HILIC) is increasingly becoming a technique of choice (Figure 1.5).300–302 RPLC methods utilize a nonpolar solid phase (often octadecyl carbon chain (C18)-bonded silica) as well a mobile phase consisting of water and a miscible organic solvent (acetonitrile, methanol, or tetrahydrofuran (THF)). The hydrophobic stationary phase has a stronger affinity for hydrophobic compounds. Gradients typically start with a polar mobile phase and gradually decrease in polarity, allowing hydrophilic compounds to elute quickly while retaining hydrophobic compounds longer.303,304 HILIC utilizes a polar chromatographic surface, making it a type of normal phase liquid chromatography (NPLC), employing

RPLC type eluents.305 The gradient oppositely starts with a nonpolar organic solvent and increases the mobile phase polarity. The aqueous phase interacting with the polar column creates a theoretical solvent-solvent system, in which compounds separate based on the degree of polarity and solvation. Polar compounds have a stronger interaction with the stationary aqueous phase and are retained longer than nonpolar compounds. HILIC is useful with regards to biomolecules and metabolites since most are semipolar or polar small molecules.300–302,306

1.3.1.2 Ionization

Ionization is a critical step that determines the ability to detect and quantify metabolites in MS-based metabolomics. Gas phase ions are formed by electron impact or chemical ionization. Several ionization techniques exist, however, the methods can broadly be categorized into hard and soft ionization. Hard ionization imparts a higher degree of energy into the target molecules through electron impact, breaking bonds to create numerous fragment molecules. The resulting information contains many low mass

25 to charge (m/z) analytes conveying abundant structural information. An example of a hard ionization technique is electron ionization (EI). EI uses the bombardment of electrons with solid or gas phase molecules to produce ions.307 It is often paired with GC, as it is suited for volatile and thermally stable compounds or atoms. In contrast, soft ionization imparts low amounts of energy to the sample, keeping molecules intact, producing little fragmentation, and resulting in larger m/z analytes. Soft ionization involves ion-molecule reactions between ionized reagent gas molecules and the molecule of interest. Ions can occur in positive or negative mode, imparting the respective charge on target molecules. Soft ionization techniques can be used universally for small molecules (metabolites) and large molecules (peptides and proteins), making them appropriate for metabolomic investigations. Matrix-assisted laser desorption ionization

(MALDI),308 and electrospray ionization (ESI)309 are both examples of soft ionization techniques often used for metabolomics. MALDI applies a pulsed laser to irradiate a sample that has been mixed with a matrix material and dried on a metal plate. The heat from the laser causes ablation, desorption, and ionization of the sample and matrix.

MALDI can provide quick analysis of metabolites in a small volume, however, it is difficult to hyphenate MALDI with LC, as the sample must be fractioned and dried. ESI can be paired with LC, and this combination is most often applied to the analysis of intracellular metabolites.310–312 LC-ESI-MS rapidly ionizes volatile liquid samples by applying a high voltage, creating a charged aerosol. The aerosolized droplets eventually evaporate into discrete charged analytes.309,313,314 Buffering agents used in LC paired with

ESI must also be volatile (usually ammonium acetate, formic acid, or ammonium formate), as salts produce deleterious effects.315 Ion suppression can occur in LC-ESI-MS

26 and is described as the coelution of matrix components with ions of interest competing for ionization in the source. Ion suppression can limit the validity of a result by affecting the precision, accuracy, or detection capability.316 Overall ESI offers the greatest extent of global metabolome coverage, stability, and ease of separation when directly paired with LC.279,291

1.3.1.3 Mass analyzer

The mass analyzer separates generated ions based on their m/z value. Mass analyzers use electric and magnetic fields to apply a force on ions. The applied force is reliant on the ionic charge of the ion (Lorentz force law), causing an acceleration that is mass dependent (Newton’s second law).317 Though it is desirable to detect metabolites in the mass analyzer with both high sensitivity and high resolution, generally, there exists a trade-off between the two properties, with different single mass analyzer configurations possessing distinct capabilities. Many mass spectrometers also use two or more mass analyzers for tandem mass spectrometry (MS/MS). Mass analyzers types include various forms of the ion trap, quadrupole mass filter, and time of flight (TOF). Tandem configurations often include the triple-quadrupole ion trap (QTrap), triple quadrupole

(TQ), quadrupole-TOF (Q-TOF), and linear quadrupole ion trap-Orbitrap (LTQ-

Orbitrap).

The Qtrap and QT analyzers are most often coupled to LC, due to their high sensitivity and selectivity, but Q-TOF and LTQ-Orbitrap are more suited to metabolomics due to their higher mass-resolving power. The front end of a Q-TOF contains one or more quadrupoles. Quadrupoles are formed from four parallel rods conducting oscillating electrical fields, used to stabilize or destabilize passing ions. The

27 combined direct current (DC) and radio frequency (RF) potentials can be used to create a mass filter, in which all other ions collide against the quadrupole rods. The first quadrupole in a TQ (Q1) transmit selected ions, acting as a mass filter. The second quadrupole (Q2) is an RF-only collision chamber, where the selected ion can be broken into fragments usually through collision-induced dissociation (CID) via a neutral gas. The third quadrupole (Q3) also acts as a mass filter for selected fragment ions.318 In a Q-TOF ions enter the TOF portion of the mass analyzer after passing through the three quadrupoles.319

A TOF uses a known electric field to accelerate ions and subsequently measures the time for ions to move from the source to the detector.320 This measurement requires ions to have the same starting time, which can be accomplished by electronic gating.

Other designs compensate for ions with different starting times by implementing a

“reflectron,” reversing the ion’s flight path.321 A curved-field or linear-field refectron can be used to ensure accurate TOF measurements and improve resolution.

An ion trap mass analyzer, such as the Qtrap, works by storing ions which are manipulated using DC and RF electric fields. The three-dimensional quadrupole ion trap is considered a “dynamic” trap, while the Orbitrap and Fourier transform ion cyclotron resonance (FTICR) mass spectrometers are regarded as a “static” trap.321 FT-based ion trap mass analyzers (Orbitrap and FTICR-MS) produce extremely high-resolution data, have non-destructive ion detection/ion remeasurement, and permit extended MS/MS experiments. Trapping ions for long time periods, however, can cause spontaneous ion decomposition, or ion-ion/ion-neutral molecule interactions.322 ICR mass analyzers are often used in tandem with MALDI.

1.3.2 Applications

Mass spectrometry-based applications for the field of metabolomics include targeted and untargeted approaches for the analysis of metabolite content, as well as separate methods developed for specifically analyzing lipids (lipidomics). The major differences between untargeted and targeted metabolomics are the accuracy of metabolite identification, the total number of detected metabolites, the sample preparation, and the level of quantification.

1.3.2.1 Targeted metabolomics

Targeted studies aim to analyze a specific and small number of metabolites using a hypothesis-driven method. Due to the focused quality, a targeted technique reduces background error and increases sensitivity, when compared to untargeted metabolomics.

Metabolites under investigation must already be chemically characterized, and often have an established biological importance. Through the use of chemical standards, investigations can be quantitative or semi-quantitative and often utilize a calibration curve. Quantitative metabolomics regularly uses chromatography along with MS/MS, combining retention time (RT), precursor m/z, and product ion spectra, to produce consistent and reproducible data.323–325

Sample preparation methods in targeted metabolomics often optimize the extraction protocols to reduce high-abundance molecules while preserving the metabolites of interest. Extractions are usually tailored to the physiochemical nature of the metabolite of interest. Consequently, many extraction parameters must be considered.

Extractions can be monophasic, biphasic, solid phase, or liquid-liquid, and employ differing ratios of organic and/or aqueous solvents.326–329 The pH and temperature during

29 the extraction are also of great importance. Though these are not all the possible extraction parameters, they are representative variables that can alter the selectivity and variability during sample extraction. Proper sample preparation can help clearly define analytes of interest, while reducing artifacts, overall, simplifying data analysis and the interpretation of biological associations.

Mass spectrometry-based targeted metabolomics often relies on selective reaction monitoring (SRM) MS/MS, during which a single or few ions can be quantified and fragmented for accurate identification.330 In general, ions are selected for in the first stage of the mass spectrometer, and a product fragment from the collision of the precursor is subsequently selected and detected in the second stage.331,332 The precursor and product ion set is called a transition. A serial application of dissociations, in which fragments are further fragmented in a third (or higher) stage is called consecutive reaction monitoring

(CRM). All transitions from the same precursor ion can also be sequentially analyzed

(multiple reaction monitoring, MRM),333 or analyzed in parallel (parallel reaction monitoring, PRM).334

1.3.2.2 Untargeted metabolomics

Untargeted metabolomics is a comprehensive analysis of all known and unknown metabolites (the metabolome). The method avoids the need for a hypothesis surrounding a specific set of metabolites. Instead, it evaluates a global profile. Samples are extracted, separated, analyzed, and statistically evaluated to acquire maximum information.

Untargeted approaches typically employ a comparative analysis of two or more data sets, usually altering one variable. Comparing two or more states, generally healthy vs. disease, generates profiles that most directly reflect the phenotype of a biological system.

Applications of untargeted metabolomics include personalized medicine, drug discovery, biomarkers, disease research, nutrition, environmental health, and systems biology.335–342

Profiling can be performed on a variety of whole extracts ranging from cells, tissue, or biofluids. While extraction protocols for untargeted analysis limit certain analytes, such as proteins, the main goal is to retain all intact aqueous metabolites from the sample. A broadened focus, however, does make single compounds more susceptible to ion suppression and quenched signal intensity. Samples are often prepared using a liquid-liquid extraction, or precipitation (biofluids).301,343–345 All conditions up to the MS analysis must be highly reproducible, as each complex sample is individually run. Many steps are introduced to normalize samples before analysis to ensure the analyte concentrations are true reflections of the source sample.346,347 Normalization can be performed before analysis using sample weight, volume, or protein concentration.

Samples may then be spiked with an isotopically labeled internal standard to further normalize peak intensities.348,349 Recently, quality control samples have been applied to control for signal drift, discard noisy background and reduce batch-to-batch variation.350–

355 Quality control samples are comprised aliquots from the study samples pooled together. After extraction, chromatography is optimized to maximize peak separation and definition over the shortest time. Small sample volumes (3-5 µL) of the highly- concentrated extracts are then separated using the gradient and column of choice.

Separation is imperative in untargeted metabolomics to reduce co-eluting species. The same sample is then run in duplicate, once using positive mode, and once using negative mode. The inclusion of both modes broadens the scope of metabolite identification. After

31 data collection, the analysis combines the chromatographic information with MS/MS transitions to deduce metabolite information using a metabolite database.

1.3.2.2.1 Bioinformatics

Metabolite data generated are often complex, and hence modern univariate and multivariate statistical methods have become essential to extract information from untargeted global metabolomic datasets. Prior to statistical analysis, raw LC-MS data is converted into an easily interpreted peak list through multiple preprocessing steps. A general workflow to pre-process most MS metabolomic data starts with feature detection.

Feature detection algorithms remove noise and convert the raw data into centroided discrete data, resulting in a single peak per ion. Feature detection is performed on the extracted ion chromatograms (XICs) by separating mass traces using data binning,

Kalman tracking,356,357 or, more recently, wavelet transforms358,359 to detect chromatographic peaks. After features are detected, peak alignment is performed, enabling a direct comparison between samples. Alignment algorithms are often used to compensate for RT-drift by using pattern recognition to accomplish peak finding and integration under set parameters, producing more reliable data.360 Peaks naturally drift between samples due to changes made over time to the column from the mobile phase, in sample-stationary phase interactions, and sample build-up. A common tactic is to incorporate internal standards to eliminate drift.348,349 This becomes less feasible for untargeted metabolomics, as the available internal standards represent a small fraction of possible metabolites.352,361,362 Features are then filtered using abundance and signal to noise ratio (S/N) information and then grouped based on similarity measurements. Lastly, features are annotated based on their m/z and RT information. Once alignment and

32 identification are complete, peaks can be matched across all samples creating a grouped list of features. If internal standards are used, relative ion abundance can be calculated based on the ratio of intensities.349

Data scaling is a technique occasionally used after data pre-processing to further standardize features. Popular scaling methods include mean centering, variance scaling, autoscaling, Pareto scaling, Log scaling, and power transformations. Different pretreatment methods highlight distinct aspects of the data to account for differences in metabolite orders of magnitude, normal secondary metabolite variation, and uninduced biological variation. While centering and scaling methods work to manipulate these variations, highlighting low abundant metabolites, they tend to inflate the measurement errors.363 Different scaling techniques can significantly affect the outcome of data analysis potentially correcting or skewing biologically relevant information and should be used with caution. Many software options (open source and proprietary) exist to complete data preprocessing and alignment for metabolomics, including XCMS,364 MZmine,365

MetAlign,366 Markerview (Sciex), AnalyzerPro (SpectralWorks), and Markerlynx

(Waters).

Processed metabolomic data is often analyzed first by using univariate techniques such as t-test or analysis of variance (ANOVA).367 Both are single variable tests used to compare group means (in this case mean intensity). A t-test is used when comparing two groups, while ANOVA is used for more than two groups. The calculated probability of statistically different mean peak intensities is represented by a p-value. A p-value is calculated using sample size, mean difference, and variance. The smaller the p-value, the greater the evidence against the null hypothesis, and generally the more statistically

33 significant the comparison. A cut-off of p-value 0.05 is typically used for biological data.

The p-value can be used in combination with other descriptive methods (such as intensity fold change between groups) to reduce large data sets and focus on statistically relevant features. Univariate test information greatly reduces redundancy but suffers from the multiple testing problem. In the case of a large sample, applying a t-test or ANOVA to each feature results in an increased chance of generating a false positive (not truly significant feature). Metabolomics innately produces large data sets, with a correspondingly large percent chance to generate false positives.368 To account for the accumulation of false positives, some researchers introduce more stringent p-value cutoffs or assign an adjusted p-value. The false discovery rate (FDR) is one method to adjust p-values by controlling the proportion of false discoveries, outputting a corresponding q-value.369,370 While the percentage of predicted false positives for a p- value is based on all features, the predicted false positives for a q-value takes into account only features below the set q-value threshold. Though each q-value may not result in less potential false positives when compared to the corresponding p-value, it gives a more accurate indication of significance and of false positives per threshold.

The complexity of metabolomic data requires the use of multivariate statistical methods to visualize and identify patterns and clusters of information. Multivariate methods reduce the dimensionality of data sets by observing and analyzing multiple variable interactions.371 Unsupervised multivariate methods are often first used to classify data without any knowledge of biological classifications. The unsupervised method used most often to visualize metabolomic data sets is Principal Component Analysis (PCA).

PCA transforms observed variable variance into a set of dimensionless principle

Figure 1.7: Example PCA data with (a) divergent groups, and (b) overlapping groups.

components, with the first principle component having the highest data variance and each successive principal component having the next highest variance while orthogonal to the preceding component.372,373 The principle components are used to construct a new set of axes to simplify the visualization of sample intragroup, and intergroup relations.374 A

PCA plot helps define natural clusters using sample-sample distance as a representative of sample association.373 In the context of metabolomics, the distinct clustering of each experimental group suggests metabolic similarity within the group, but divergent metabolism between groups (Figure 1.6a). In opposite, groups with a comparable metabolism are visualized as overlapping clusters (Figure 1.6b). Cluster analysis can also be performed using supervised methods. Supervised methods build and then apply a predictive model using a data set with a known classification and outcome.374,375 In particular, Partial Least Squares Discriminant analysis (PLS-DA) is a biased variant of

PCA in which orthogonal axes are constrained by the covariance between groups.376

Supervised methods are useful for sequential and temporal progression analysis, and for highlighting notable metabolites by forcing a group classification.

Individual metabolite identification currently remains the bottleneck in the field of metabolomics. Metabolites can first be putatively identified by comparing the precursor m/z to metabolites in a private or public library (HMDB,377 KEGG,378 METLIN379). Mass spectrometry-based metabolomics using soft ionization yields highly redundant data, as each compound can ionize differently under positive and negative mode, form a multimer, form a multiply charged ion, and/or associate with different adduct ions (i.e.

Na+, H+, NH4+, K+). Typical identification software allows the user to eliminate potential hits by restricting the ion mode, adduct type, and molecular weight tolerance. Still, automated directories for identifying metabolites using precursor information often list multiple potential hits per feature. Furthermore, search software does not consider RT except as a means to separately identify features, as the information is too variable. In general, accurate metabolite identification still relies heavily on using individual MS/MS spectra to confirm putative precursor identifications.380–382 Such confirmations are made by searching metabolite libraries by manually inputting MS/MS information. The fragment information found in libraries are submitted by research groups or can be generated via in silico prediction software.383 Machine and procedure differences, including the type of mass analyzer, type of CID inert gas, and collision energy level can cause different MS/MS fragment patterns. MS/MS spectra are also inconsistently dispersed between the multiple libraries, and sometimes not available. Often metabolites of interest must eventually be confirmed using a targeted approach. Such a labor- intensive procedure creates difficulty automating metabolite identification.

1.3.2.3 Lipidomics

Lipidomics is the study of biological networks and pathways associated with cellular lipids.384–386 As lipids are small molecules they fall under the general classification of metabolomics, though require a separate discipline due to their unique characteristics and specificity. Lipidomic investigations involve quantifying and identifying lipid structures, functions, and interactions with other macromolecules. Lipids are associated with numerous cellular functions including signaling, energy storage, and structural components of the cell membrane.384,387–389 Lipids are composed mainly of two structural units, a ketoacyl and isoprene group, variously combined depending on the lipid class.390 Variations of lipid groups are limited based on cell or tissue specificity, for example, cardiolipins are found almost exclusively in mitochondria.391

As most lipid classes are hydrophobic in nature, most extractions isolate lipids using organic solvents. Traditional procedures, such as the Folch,392 and Bligh and

Dyer393 protocol use mixtures of chloroform and methanol to generate a phase separation, with lipids subsiding in the bottom organic layer. Another type of lipid extraction, methyl-tert-butyl ether extraction,394 has the advantage of partitioning lipids into the upper layer of a two-phase extraction system. Extractions, useful for complex and low- abundance lipids have also be adapted from these protocols.395–398

Due to the diversity of lipid groups, no one procedure can accurately detect the whole lipidome. Each lipid group has separate fragmentation patterns and ionization efficiencies.399 Platforms are often combined to maximize the coverage of separate groups, and depending on the instrumental setup different levels of molecular and structural information can be discovered. Soft ionization techniques offer a

37 comprehensive detection range of lipids by preserving the intact molecule, so lipidomic platforms primarily use ESI ionization.309,400

Lipid extracts can be introduced to the mass spectrometer after LC separation.

LC-MS is a powerful tool for targeted analysis of specific lipid groups. Both NPLC and

RPLC are used to achieve lipid separation, and for untargeted lipidomic studies, it is common to analyze each sample using NPLC (or HILIC) along with RPLC to maximize coverage. NPLC effectively separates lipids by headgroup polarity whereas, RPLC separates on the basis of chain length and the degree of unsaturation. Lipid species must be fully identified using positive and negative ion MS/MS full scans. Quantitation is generally more difficult for untargeted lipidomics using LC-MS. However, due to reduced ion suppression, one can detect a higher number of lipids. Alternatively, crude lipid extracts can be introduced directly without prior chromatographic separation. This direct infusion intrasource separation method has been termed ‘shotgun lipidomics.'401,402

Shotgun lipidomics avoids complications due to chromatographic abnormalities and concentration alterations. Another advantage is that a longer scan time can be spent on each lipid class.

Lipids can be analyzed solely with accurate mass using FTMS, or FT-Orbitrap.

This method is referred to as top-down lipidomics,403,404 and the intention is to differentiate lipid patterns as opposed to quantifying individual lipid species. Most lipidomic experiments, however, are performed using an MS/MS MRM analysis. This bottom-up approach carries out a semi-targeted scan by the sequential acquisition of ions through windows of a specified mass range.404–406 The disadvantage of this approach is the deficiency in the detection of unanticipated lipids.

Lipid identification has been made easier through the growing available bioinformatic resources. Lipidomic tools such as LIPID MAPS and LipidBank facilitate annotation,407 and full spectra can be referenced from HMDB or XCMS. Other MS/MS software’s assist with the general identification of the lipid class, chain length, and degree of saturation; these include LipidXplorer, Lipid Search, SimLipid, and Lipidview.408

CHAPTER II

SPECTROSCOPIC STRUCTURAL ANALYSIS OF CPZ AND ITS COPPER

COMPLEXES

2.1 Introduction

CPZ, the small molecule hydrazine (Scheme 2.1a), has long been used as a sensitive quantitative determinate of copper(II) ions due to the intensely blue production of the resulting CuCPZ complex. The complex absorbs in the visible region centered at

595 nm (e/dm3 mol−1 cm−1 16 900).173,409 CPZ alone is a hydrophobic substance that when bound to copper the CuCPZ complex is aqueous. This poses an initial challenge, as using CPZ for later chelation requires the stock solution be dissolved in an organic solvent. Stabilization of the solution is normally achieved through phosphate buffering, or by gradual additions of sodium hydroxide to water.173,203,264,409 CPZ can also be dissolved in 50% ethanol/water with gentle heating. The method is cited in most papers involving treatment of cell cultures with CPZ.264,410 CuCPZ produces visible color that is stable for at least 15 minutes within the optimum pH range of 7 to 9.409 The ratio of

40 copper to CPZ also directly affects the complex stability.173,264,409 Benetti, reported that

CuCPZ solutions made with lower copper concentrations (Cu(II):CPZ ratios < 1:4) had a hyperbolic trend, peaking in stability at 17 days. Whilst at high copper concentrations

(Cu(II):CPZ ratios > 1:4) the trend observed is a linear decay due to product precipitation.264 Peterson and Bollier, however, originally reported a gradual fading of the color after only 60 minutes at a rate of 1% per hour, with solutions lasting a maximum of

3 days at room temperature. Other ions can produce a significant change in the absorbance of the CuCPZ complex. Lead, zinc, and nickel interfere at concentrations as low as 0.5 p.p.m. and cyanide prevents the development of color at concentrations less than 0.1 p.p.m.409 However, none of the tested ions alone formed any colored complex with CPZ.

Information regarding the isolated structures of both CPZ and CuCPZ remains limited and convoluted. Ab initio calculations on the isolated CPZ ligand, in a density functional theory (DFT) framework, combined with infrared (IR) and Ramen spectroscopy, and X-ray crystallography, all confirm CPZ packs as two centrosymmetric molecules in a monoclinic unit with a trans configuration.411 When subjected to an aqueous environment, CPZ has long been predicted to hydrolyze the hydrazone bond to cyclohexane (CPZ-R), though this has never been documented when examining solutions of the ligand alone (Scheme 2.1b). Structural information pertaining to the CuCPZ complex remain even more elusive, as a crystal of the complex has yet to be isolated.

Messori characterized the complex using ESI-TOF-MS, FTIR spectroscopy, EPR, and X-

Ray photoelectron spectrometry) and proposed a mononuclear complex, made of

Scheme 2.1: Structure of intact CPZ, the proposed monohydrolyzed ligand, and three potential CuCPZ chelation modes.203,413 two bidentate monohydrazones resulting from partial CPZ hydrolysis and bonded through the amidic nitrogens, (Scheme 2.1c). The MS and XPS data suggests that CuCPZ would consist of a d8 copper(III) center with a square planar arrangement.203 Other redox potential and DFT computational studies support the formation of the proposed d8 copper(III) center.412,413 A recent investigation isolated and crystallized (E)-1,2-diphenyl-

2-(2-(pyridine-2-yl)hydrozono)ethanone, an analogue of monohydrazone cuprizone, and performed XPS and DFT studies to predict the ground state energies of the various

CuCPZ chelation modes.413 Their work explored potential cis/trans-metal binding via both amidic nitrogens (CPZNN) (Scheme 2.1c), the keto imine nitrogen and oxygen

(CPZNO) (Scheme 2.1d), or both carbonyl oxygens (CPZOO) Scheme 2.1e). They concluded that both the cis- and trans-CPZNO configurations have the lowest ground state, while cis- and trans-CPZNN are approximately 3 × 103 kJ/mol higher in energy.

While the analogue (E)-1,2-diphenyl-2-(2-(pyridine-2-yl)hydrozono)ethanone formed a d9 complex made of two weakly coupled Cu(II) ions, modeling CuCPZ still left questions regarding copper’s oxidation state. Additionally, limited research has been

42 completed regarding the chelating properties of CPZ in relation to its biological activity.

Copper metabolism is tightly controlled with a series of chaperone and transporter proteins, as excess copper can initiate oxidative damage causing hepatic dysfunction, neurodegenerative changes in the CNS.414–419 Thus, the chelation of free copper by CPZ seems unlikely within a biological system. CPZ likely inhibits such copper-containing enzymes via copper extraction, or through complex formation and subsequent enzyme disruption. One study by Lindström and Pettersson examined the inhibition of pig plasma benzylamine oxidase with CPZ.267 Using electron paramagnetic resonance (EPR) and absorbance spectroscopy, they concluded that the inhibitory action of CPZ occurred from an interaction with the enzyme-bound coenzyme, pyridoxal phosphate, and was not attributed to its copper-chelating properties. Here we examine the structural properties of

CPZ and the CuCPZ complex, as well as the CPZ’s copper-chelating mechanism.

2.2 Methods

2.2.1 Absorbance Spectroscopy

For absorbance spectroscopy studies, CPZ solution was made fresh by dissolving

CPZ powder into ethanol (200 proof for molecular biology). CuCPZ was prepared by mixing aqueous CuSO4•5H2O with the CPZ solution in phosphate buffer (pH 7.4), and the mimics B and R were both dissolved in ultrapure water. All samples were measured in quartz cuvettes on a Spectramax M2 after 15 minutes of incubation. The kinetic study of CuCPZ was performed over a period of 60 hours, examining the following ratios: 1:1,

1:2, 1:4, 1:8, 1:16, 2:1, 4:1, 8:1 and 16:1 (Cu(II):CPZ). Samples were analyzed in triplicate on a 24-well plate using a Spectramax M2 plate reader.

The stoichiometric ligand binding ratio analysis of CuCPZ and CPZ with each protein mimic was performed by preparing solutions with differing mole fractions of

CPZ, ranging from 0.1-1 XCPZ in steps of 0.1 XCPZ, in phosphate buffer (pH 7.4) to keep pH consistent, at a combined molar concentration of 1mM. Triplicates of each solution were prepared and incubated for one hour. All samples were read on a 96-well plate at the respective complex λmax using a Spectramax M2 plate reader.

2.2.2 NMR Spectroscopy

For NMR spectroscopy studies, the CPZ solution was made at 100 µM in DMSO- d6, the mimics B and R were both dissolved in D2O at 100 µM. To study the CPZ-mimic complex, CPZ in DMSO-d6 was added to either B or R at a 1:1 ratio, with a total theoretical complex concentration of 100 µM. Samples were run on a Bruker 1H-NMR

300 MHz and the data was analyzed using ACDLabs 1D NMR processor. The CPZ theoretical spectra were generated using ChemDraw 1H-NMR predictor.

2.2.3 Mass Spectrometry

CPZ used for direct injection ESI-MS was prepared at 100 µM in methanol.

CuCPZ was prepared by mixing aqueous CuSO4•5H2O with the CPZ solution in phosphate buffer (pH 7.4). After the complex was formed it was diluted in methanol, at a final concentration of 100 µM, to promote solvent evaporation. The samples were analyzed by direct injection with the 5600+ TripleTOF Mass Spectrometer (SCIEX,

Framingham, MA, USA) in positive mode. Samples were injected at a flow rate of 10

µL/min. The ion source nebulizer gas was set at 18 psi, heater gas was 18 psi, and the curtain gas was 20 psi. The ionspray voltage was set to +5000 V, and the declustering

44 potential was set to +100. The TOF scan was performed over the mass range of 50-1,000

Da. Fragmentation data was subsequently collected for the CPZ precursor ion m/z =

279.18 over a range of 50-500 Da with a collision energy of +25 V.

For MS analysis of the CPZ and the mimic B complex, three solutions were prepared consisting of the following ratios: 1:3, 1:1, and 3:1 CPZ:B. The solutions were made in 35% acetonitrile at a total molar concentration of 1mM and left to incubate for one hour prior to analysis. Each solution was analyzed by direct injection in positive mode on a 5600+ TripleTOF Mass Spectrometer. Samples were injected at 10 µL/min over 10 minutes. The ion source nebulizer gas was set at 18 psi, heater gas was 18 psi, and the curtain gas was 20 psi. The ionspray voltage was set to +5000 V, and the declustering potential was set to +100. A TOF scan was performed over the mass range of 100-1,000 Da.

2.3 Results and Discussion

2.3.1 Absorbance spectroscopy of CPZ, CuCPZ, and active site mimics

Though already well documented, the absorbance spectroscopy for CPZ, CuSO4, and CuCPZ (Figure 2.1a) were verified as proof for the proper preparation of the complex. The absorbance maximum (λmax) was duplicated at 600 nm for CuCPZ prepared at pH 7.4. A kinetic study of the complex was performed over a period of 60 hours, examining the following ratios: 1:1, 1:2, 1:4, 1:8, 1:16, 2:1, 4:1, 8:1 and 16:1

(Cu(II):CPZ) (Figure 2.1b). As seen in the study by Benetti, ratios containing an excess of copper decayed rapidly, forming a precipitate in solution, while solutions with higher

CPZ content exhibited the aforementioned hyperbolic trend.264 However, in this study the

Figure 2.1:CuCPZ absorbance decays in solution in a manner dependent on the Cu(II):CPZ ratio. (a) Absorbance wavelength scan shows different peak shapes for CuCPZ (blue), CPZ (orange), and CuSO4 (green). The λmax of CuCPZ is confirmed at 600 nm. (b) The average OD600 of various solutions, ranging from a high to low Cu(II):CPZ ratio (light blue to dark blue), were observed over a period of 60 hours (done in triplicate).

Figure 2.2: Absorbance shifts induced by CPZ addition to copper active site mimics (a) Structures of both protein active site mimics, abbreviated as “R” and “B” and a colored photograph of both mimics (“R” and “B”) when dissolved in water at 50 mM, and the corresponding shifted color when 50mM CPZ is added (“R+CPZ” and “B+CPZ”). Spectra detail the shift of R from 548 nm to 583 nm with CPZ (b), and the shift of B from 651 nm to 627 nm (c) with CPZ. Job plots showing the stoichiometric ligand binding ratio of CPZ with (d) copper, (e) R mimic and (f) B mimic. Solutions were prepared with differing mole fractions of CPZ, ranging from 0.1-1 XCPZ in steps of 0.1 XCPZ to explore CPZ’s binding preference as a ligand in each scenario (repeated in triplicate).

47 rate of decay happened on a much faster timescale, similar to Peterson and Bollier’s observations.409 Solutions with a ratio of Cu(II):CPZ ≥ 1:1 decayed within 8 hours, while solutions with Cu(II):CPZ ratios ≤ 1:2, instead peaked at approximately 5 hours and decayed after 55 hours. The average tangential rate of color decay for solutions with a ratio of Cu(II):CPZ ≥ 1:1 was 14%/hr, and for solutions with a ratio of Cu(II):CPZ ≤ 1:2 it was 3%/hr.

To examine a simplified chelation scenario in which CPZ potentially interacts with copper proteins, CPZ was incubated with two different protein active site mimics.

We sought to test whether CPZ could efficiently extract copper confined in the mimic.

Copper bound 1,8-dithia-4,11-diazacyclotetradecane (R) is a type 1 copper center mimic, while copper bound 1,4,7-triazacyclononane (B) mimics a copper B center (Figure

2a).420,421 Upon incubation of each complex with CPZ, a noticeable shift in absorbance occurred that is unique from the characteristic 600 nm λmax of CuCPZ (Figure 2.2a).

Complex B was blueshifted from 651.00 nm to 627.00 nm, while complex R was redshifted from 548.00 nm to 583.00 nm, with the addition of CPZ at a 1:1 ratio (Figure

2.2b,c). The shift in color suggests the formation of a new complex as opposed to copper chelation, as the mimics are colorless without copper.

To characterize the binding stoichiometry of CPZ with the small molecule mimics, we utilized Job’s methods (method of continuous variation). In this method, the

λmax of each complex is measured against differing mole fractions of each reagent, while the total molar concentration is held constant. The maximum plotted absorbance is used to delineate the stoichiometry of the binding event between each reagent. While the stoichiometric ligand to metal ratio of CPZ with copper falls at 2:1 (OD600), when

48 combined with either mimic CPZ instead forms a 1:1 binding ratio at the respectively shifted absorbance (OD583 for R and OD627 for B) (Figure 2.2d-f). This provides additional evidence of a stable complex between CPZ and each mimic.

2.3.2 NMR Spectroscopy of CPZ, and active site mimics

Evidence collected from 1H-NMR studies of CPZ emphasizes the existence of monohydrazone CPZ (CPZ-R), in coexistence with the intact cuprizone under aqueous conditions. The collected spectra revealed a combined multiplet (M01 and M02) at 1.57 ppm and 1.62 ppm, a triplet at 2.28 ppm (M03), a singlet at 2.09 ppm (M04) and two down shifted singlets at 10.69 ppm (M05) and 10.89 ppm (M06) (Figure 2.3a). A predicted model (ChemDraw 1H-NMR prediction) of intact CPZ accounts for M01, M02,

M03, and M06, with reasonable deviation from the labile amidic hydrogen (Figure 2.3b).

The peaks from intact CPZ also match (with shifts due to differing solvents) a previously recorded spectrum of CPZ prepared in CDCl3 on a spectrometer operating at 399.65

MHz.422 The multiplets M01, M02, and M03 fit the predicted shape and shifts for the hydrogens found on the cyclohexane of CPZ, while M06 corresponds to the amidic proton (Figure 2.3c). The two unique singlets M04 and M05 match groups found only when one cyclohexane is cleaved from CPZ, corresponding to hydrogens from the cleaved amine, and the now asymmetric amidic hydrogen (Figure 2.3c). The signal from

M01, M02, M03, and M06 should overlap from both CPZ and CPZ-R. The lower signal from the singlets M04 and M05, which are attributed only to CPZ-R, reasonably suggests that a mixture exists of both intact and monohydrazone CPZ.

CPZ’s structure was next examined when exposed to the copper bound protein active site mimics B and R using 1H-NMR. Spectra were first collected for the individual

Figure 2.3: NMR reveals CPZ as partially hydrolyzed in solution. (a) 1H NMR of 100 µM CPZ (300 MHz, DMSO-d6). Peaks are labeled above, and each peak group is numbered and marked accordingly: s for singlet, t for triplet, and m for multiplet. The inlay shows closer detail of M01-M03. (b) The predicted spectra of intact CPZ was generated for comparison using ChemDraw 1H NMR prediction. (c) Peaks from the experimental spectrum are shown assigned to the two structures coexisting in solution.

50 mimics. Unfortunately, 1H-NMR provided poor resolution due to the paramagnetic nature of the bound Cu(II) center in each mimic. The spectrum of B reveals a singlet at 1.99 ppm (M01) (Figure 2.4a) corresponding to the symmetric methylene hydrogens. This peak is unchanged with the addition of CPZ (Figure 2.4b), suggesting that the copper interaction within mimic B is unaltered. The singlet at 3.25 ppm is a result DMSO contamination from the CPZ addition. Peaks also appear in Figure 2.4b that were previously characterized in Figure 2.3 as being unique to the structure of CPZ, though these are shifted due to solvent deviation. The peaks include the now separated multiplet at 1.80 ppm (M02), and 1.67 ppm (M03) and the triplet at 2.32 ppm (M04), all corresponding to hydrogens on the cyclohexane ring. There is, however, a lack of either of the down-shifted singlets associated with the amidic nitrogens and instead, a new singlet appears at 2.62 ppm (M05). The disappearance of the singlets near 10.69 ppm and

10.89 ppm suggests that the amidic nitrogens of CPZ participate in an interaction with the copper center of B. This provides additional evidence of a stable complex between CPZ and the mimic. The singlet (M04) is harder to elucidate due to deviations arising from the copper center, combined with the altered solvent conditions. The closest prediction fits it as the free amine of the CPZ-R ligand. No signal could be obtained with 1H-NMR for compound R under our operating conditions (Figure 2.4c). When CPZ was added to R the only peaks that appear are multiplets at 1.80 ppm (M01) and 1.67 ppm (M02), the triplet at 2.32 ppm (M03), and the singlet at 2.62 ppm (M04) (Figure 2.4d), confirming their association with CPZ.

Figure 2.4: NMR reveals amidic CPZ hydrogens interact with mimics B and R. (a) 1H NMR of 100 µM 1 B (300 MHz, D2O). (b) H NMR of 100 µM B with 100 µM CPZ (300 MHz, D2O with DMSO-d6). (c) 1 1 H NMR of 100 µM R (300 MHz, D2O). (d) H NMR of 100 µM R with 100 µM CPZ (300 MHz, D2O with DMSO-d6). Peaks are labeled above, and each peak group is numbered and marked accordingly: s for singlet, t for triplet, and m for multiplet.

2.3.3 Mass Spectrometry of CPZ, CuCPZ, and active site mimics

The structure of CPZ was subsequently analyzed in positive mode using direct injection ESI-MS. The collected spectrum revealed two main analytes with m/z = 279.18, and m/z = 199.22 corresponding to [CPZ+H]+ and to the predicted [CPZ-R+H]+, respectively (Figure 2.5a). A MS/MS scan was successively performed, selecting m/z =

Figure 2.5: Mass spectrometry confirms the presence of partially hydrolyzed CPZ and CuCPZ formed at a 1:2 copper:CPZ ratio. (a) Direct injection TOF-MS spectra of the CPZ. The solution was made in methanol at a total molar concentration of 1mM. (b) The MS/MS spectrum was collected for the precursor ion m/z = 279.18. (c) All structures for [CPZ+H]+, [CPZ-R+H]+, and the CPZ fragment ions are shown with the corresponding m/z. (d) TOF-MS of CuCPZ solution. CuCPZ was formed in phosphate buffer, and after 10 minutes mixed with in methanol (35%) to a final concentration of 100 µM.

279.18 as the precursor ion. The most abundant fragmented product is m/z = 139.09, followed by the products with m/z = 113.15, m/z = 96.08, and m/z = 184.12 (Figure

2.5b). The fragments for CPZ had not previously been cataloged. The fragments fit structural elements of CPZ and are assigned in Figure 2.5c along with the structures for

[CPZ+H]+ and [CPZ-R+H]+. Analysis of the CuCPZ complex with ESI-MS proved complex and convoluted. The main peaks reported by Messori were not discovered in our analysis. In positive mode, the spectra show the presence of dominant peaks m/z =

221.10 [CPZ+Na]+ and m/z = 301.16 [CPZ-R+Na]+, and of additional peaks at m/z =

+ + 419.21 [(CPZ-R)2+Na] , m/z = 499.39, m/z = 501.10, and m/z = 579.34 [(CPZ)2+Na]

(Figure 2.5d). The analyte with m/z = 499.39 corresponds to a species of formula

− + 203 ([Cu(CPZ-R)2] +2Na ), which is the species predicted by Messori, and the isotopic distribution, as noted by m/z = 501.10, is indicative of a copper ion. No evidence was found for the presence of the hypothetical copper chelate species formed from intact CPZ

− + ([Cu(CPZ-R)2] +2Na ), with a theoretical m/z = 661.23.

To confirm the presence of the proposed complex between CPZ and the B, we analyzed different molar ratios of CPZ with B using mass spectrometry. With excess

CPZ present, the mass spectrum shows that the intact [CPZ+H]+ peak dominates at m/z =

279.18 along with the presence of the [CPZ+Na]+ at m/z = 301.16 (Figure 2.6a).

Alternatively, when the ratio is shifted towards excess B, peaks that correspond to both

+ + the Cu(I) [B] (m/z = 192.05) and the Cu(II) [B+ClO4] (m/z = 291.00) forms of the mimics are observed (Figure 2.6b). These have been previously noted in the characterization of B.421 The spectra for the 1:1 ratio shows evidence of two different complex formations made of either the intact CPZ with B, [B+CPZ]+ at m/z = 470.23 or

Figure 2.6: Mass spectrometry reveals complex formed between B and CPZ. Direct injection TOF-MS spectrum of the CPZ and mimic B complex at (a) 3:1 CPZ:B, (b) 1:3 CPZ:B, and (c) 1:1 CPZ:B. The solutions were made in 35% acetonitrile at a total molar concentration of 1mM and left to incubate for one hour prior to analysis. Two different forms of the proposed complex between B and CPZ are denoted as [B+CPZ]+ and [B+CPZ-R]+, where CPZ-R represents a CPZ with one hydrolyzed cyclohexane ring.

CPZ-R with B, [B+CPZ-R]+ at m/z = 390.17 (Figure 2.6c). These complexes are also found when the stoichiometry is shifted to favor either CPZ or B, although at a lower intensity. Interestingly the [B+CPZ]+ complex is favored with excess CPZ, while the

[B+CPZ-R]+ complex is favored in the presence of excess B.

2.4 Conclusions

The spectroscopic examination of CPZ confirms the existence of monohydrolyzed

CPZ, stably existing in solution with intact CPZ, and that CPZ-R, not CPZ acts as the copper ligand (originally suggested by Nilsson).173 Our results obtained with small molecule protein active site mimics reveal the formation of a CPZ-mimic complex, as opposed to copper removal. These conclusions are similar to a previous study showing a weak binding event within the complicated structure of a folded copper protein.267 This data promotes the idea that CPZ-mediated toxicity may not arise from a functional depletion of copper from proteins, but perhaps by forming unstable or nonfunctional complexes.

CHAPTER III

GLOBAL METABOLOMIC ANALYSIS OF CPZ DEMYELINATION

3.1 Introduction

The myelination of neuronal axons in the CNS permits the rapid conduction of nerve impulses and maintains axonal integrity.423 This structure is formed by the spiral wrapping of the oligodendrocyte plasma membrane around axons and its destruction occurs during diseases such as multiple sclerosis.424 Cuprizone intoxication is used to study pathways involved in oligodendrocyte injury and test compounds that can promote remyelination.135 C57BL/6 mice are fed this copper chelator, leading to the development of demyelinating lesions in brain regions such as the CC, and HP, causing early and selective apoptosis of oligodendrocytes (Figure 3.1).425 Although most studies focus on the CC as the defined region of interest, it is acknowledged that demyelination is not restricted to this structure, nor even to white matter regions. Damage instead seems to be

57 site- and tract-specific.426,427 White matter tracts including the optic tract, the hippocampal fimbria, the mammillothalamic tract, and the columns of the fornix do not present explicit demyelination when observed with immunohistochemical staining after

CPZ exposure for 5 weeks.426 In contrast, overt demyelination can be found in grey matter areas such as the thalamus, cortex, and caudoputamen.426 Imaging shows reduced

(0.39%) grey matter-white matter contrast using magnetization transfer ratio after 6 weeks of CPZ exposure,428 and magnetic transfer imaging of deep grey matter demyelination after 6 weeks was also correlated with a loss of PLP via immunohistochemical staining.429 Strangely, subinterior regions such as areas of the caudoputamen, the hypothalamus, and the entirety of the spinal cord (SC) seem invulnerable to CPZ demyelination at the 5-week point,426,430 and it is unclear what causes this specificity.

As a result of early optimization of model parameters, the standard intoxication model of CPZ demyelination using C57BL/6 mice last from 5-6 weeks.188,235 At this point, maximal demyelination occurs, with almost complete loss of oligodendrocyte cells in the CC.188,199 Early oligodendrocyte dysfunction has, however, been described during initial model development, suggesting that mature oligodendrocyte dysfunction occurs within the first week of cuprizone treatment.242 Amino acid deprivation was observed occurring as early as 4 days into the model. Downregulation of myelin genes, growth factor disruption, and oligodendrocyte death have also been noted within and up to one week of the CPZ diet.234,235,431,432 The infiltration of neutrophils,239 astrogliosis,242,433 and microgliosis also develop between 2-4 weeks into the model.188 Work by Doan et al. showed that mice exposed to cuprizone for 2-4 week periods, with cessation, all exhibited

0 6 6+6 off Weeks with CPZ

Figure 3.1: Depiction of demyelination in the CNS. (Top) Regions of the CNS most affected by demyelination in CPZ model are highlighted, specifically the CC and HP. (Bottom) Myelin in CC after (panel 1) 0, or (panel 2) 6 weeks of CPZ intoxication. (Panel 3) remyelination of myelin after CPZ has been removed for 6 weeks following 6 weeks of demyelination. significant demyelination by the 4th week, and that mice treated for only 1 week (with removal for 3 weeks) developed negligible demyelination.434 Together, this work provides evidence of early oligodendrocyte dysfunction reliant on at least two weeks of exposure to cause subsequent demyelination. As made evident, fundamental work exploring the temporal nature of CPZ intoxication is still needed, and here we comparatively investigate both a 2- and 6-week model.

The brain pathology that occurs in this model is thought to result from both cell- specific toxicity and activation of innate immune effectors. As noted, the death of myelinating oligodendrocytes is accompanied by microglial and macrophage activation,

59 astrogliosis, and increased production of inflammatory cytokines.188,203,204,242,433 The accumulation of these cells produces a concentrated response producing proinflammatory molecules such as TNF-α, IL-1β, IFN-γ, and NO.115,188,435,436 Knockout experiments have demonstrated the importance of chemokine receptors such as CXCR2239 as well as inflammatory cytokines437,438 to the development of demyelination, however, this complicated inflammatory milieu in vivo has made it difficult to identify cell intrinsic pathways that lead to oligodendrocyte-specific injury. Attempts thus far to create a focused model of intoxication using neuron, astrocyte, and oligodendrocyte cell lines, or primary microglia, astrocytes, and OPCs 202,265,410 are limited and produced no decrease in viability with up to 200 µM CPZ. To this date, only one study confirmed selective intoxication using differentiated mature rat oligodendrocytes.201

Due to conflicting reports concerning the mechanisms of CPZ-mediated oligodendrocyte death, and based on our structural investigation, we sought to determine whether CPZ intoxication perturbs cellular metabolism, as opposed to disrupting copper homeostasis, by using a systems biology approach. Metabolomics is an emerging technology that detects changes in endogenous small molecules by using LC-MS. This technology allows for the global determination of dysregulation in biochemical pathways important for energy generation, myelin synthesis, and cell survival.339,439,440 We demonstrate that CPZ can accumulate within cells and leads to cell death in the oligodendroglial cell line, MO3.13. Metabolomic profiling of treated cells and tissue isolated from CPZ-fed mice shows that the compound induces widespread metabolic dysfunction that is region specific and consistent with the cellular pathology.

Additionally, we used mass spectrometry and NMR to probe the ability of CPZ to

60 interact with copper contained within small molecule mimics of protein copper sites as well as pyridoxal 5’-phosphate (P5P), a vitamin cofactor essential for amino acid metabolism. Our results indicate that CPZ toxicity in oligodendrocytes may be due to a disruption of the enzymes responsible for amino acid metabolism leading to increased susceptibility to reactive oxygen species and energy depletion.

3.2 Methods

3.2.1 Chemicals

HPLC grade (≥99.9%) acetonitrile, ethanol, isopropanol, methanol, and water used for extractions and LC-MS analysis were purchased from Fisher Scientific (Fair

Lawn, NJ, USA). Chloroform (HPLC grade, ≥99.5%) and NP-40 were purchased from

Alfa Asear (Ward Hill, MA, USA). Cuprizone (Biscyclohexanone oxaldihydrazone), copper (II) sulfate pentahydrate (BioReagent, ≥98%), thiazolyl blue tetrazolium bromide

(MTT) (BioReagent, ≥97.5%), P5P (Pyridoxal 5’-phosphate hydrate) (Bioreagent,

≥98%), DAPI, goat anti-mouse IgG (H+L) conjugated to FITC (F-2761), and ammonium acetate (HPLC grade), were purchased from Sigma-Aldrich (ST. Louis, MO, USA). Anti-

CNPase (ab6319) was purchased from Abcam (Cambridge, MA, USA). All cell culture materials including phosphate-buffered saline (PBS), fetal bovine serum (FBS)

Dulbecco’s Modification of Eagle’s Medium (DMEM), and penicillin/streptomycin were purchased from Corning (Manassas, VA, USA). D2O was purchased from Cambridge

Isotope Laboratories (Tewksbury, MA, USA).

3.2.2 Cell Culture

The MO3.13 human oligodendrocyte cell line was obtained from CELLutions

Biosystems Inc. (Burlington, Ontario, CA). MO3.13 cells were cultured in DMEM supplemented with 10% FBS, and 1% penicillin/streptomycin and grown at 37C in 5%

CO2 for all experiments.

For metabolomic analyses, cells were seeded in 6-well plates at a density of

1.0×106 cells/well for 24 hours, and left to attach. After incubation, the media was aspirated and replaced with medium containing 1 mM CPZ in a vehicle containing 0.3% ethanol. The cells were incubated with CPZ and metabolite extraction performed just prior to the loss of viability at 24 hours.

3.2.3 Preparation of CPZ Solution

CPZ solution was made fresh by dissolving CPZ powder in ethanol equal to 30% of the final solution volume. After the powder was partially dissolved, ddH2O was used to dilute the solution to the final volume. The solution was then gently heated with stirring until all of the CPZ dissolved. Afterwards the solution was filtered (22 µm) into its final container. Each batch was used within 2-3 days. The maximum concentration achieved, while fully dissolved, was 10 mM.

3.2.4 Live/Dead Cell Viability Assay and MTT assay

MO3.13 cells or the rat astrocyte cell line (DI TNC1) were seeded into three separate 96-well plates at a density of 73,000 cells/mL. The last row of each plate contained 100 µL of tissue culture medium with no cells to act as a blank control. The cells were left to attach for 24 hours followed by treatment with: 0.125, 0.25, 0.5, or 1

62 mM CPZ and 0.3% ethanol in DMEM, or a vehicle of DMEM with 0.3% ethanol. The plates were subsequently incubated for 6, 18, or 24 hours. For the Live/Dead Cell

Viability assay (ThermoFisher, Waltham, MA, USA) an additional row of cells was treated with 500 uM H2O2 for 10 minutes prior to any measurements, to act as a positive control for dead cells. The Live/Dead assay standard protocol for a fluorescence microplate was followed, using a final concentration of 2 µM calcein AM and 4 µM

EthD-1. The percentage of live cells was calculated as shown:

퐹(530)푠푎푚푝푙푒 − 퐹(530)푚푖푛 ( ⁄퐹(530)푚푎푥 − 퐹(530)푚푖푛) ×100%

Where F(530)sample is the fluorescence at 530 nm in the experimental cell sample, labeled with both calcein AM and EthD-1; F(530)min is a sample where all (or nearly all) cells are alive and labeled with EthD-1 only, and F(530)max is a sample where all (or nearly all) cells are alive and labeled with calcein AM only.

For the MTT assay, 20 µL of 5 mg/mL MTT was added to each well after the respective time point. Each plate was then incubated for 3.5 hours at 37°C, lysis buffer was added (4 mM HCl, 0.1% NP40 buffer, in isopropanol), and the absorbance was read at 590 nm with a reference filter of 620 nm by using a Spectramax M2 plate reader

(Molecular Devices, Sunnyvale, CA, USA). The percent metabolic activity was calculated using the vehicle control set as 100% metabolic activity.

3.2.5 Immunofluorescence

MO3.13 cells were seeded onto glass coverslips in a 6-well plate at a density of

1.0×106 cells/well and left to attach for 24 hours. After incubation, the media was aspirated and replaced with DMEM with 0.3% ethanol containing 1 mM CPZ or DMEM

63 with 0.3% ethanol vehicle. Following 24 hours of CPZ treatment, the media was removed and the cells were washed with PBS. Cells were fixed in 4% paraformaldehyde for 10 minutes and then permeabilized by using 1% Tween for 20 minutes. Cells were then left to block in 1% BSA/10% FBS for 2 hours followed by incubation with anti-CNPase at 10

µg/ml overnight at 4°C. Detection was performed with a goat anti-mouse IgG (H+L) conjugated to FITC (2 µg/ml). DAPI was used to stain the cell nuclei and cells were imaged by using a Nikon A1+ confocal microscope.

3.2.6 Cuprizone and CuCPZ Absorbance Assay

MO3.13 cells were first seeded in four 6-well plates at a density of 1.0×106 cells/well and left to attach for 24 hours. After incubation, the medium was aspirated and replaced with DMEM containing either: 10 µM CPZ with 0.003% ethanol, 10 µM CPZ and 5 µM CuSO4·5H2O with 0.003% ethanol, 5 µM CuSO4·5H2O, or 0.003% ethanol.

100 µL of the media was sampled from each plate at time points: 0, 12, 24, 36 and 48 hours. After collection of the medium, the absorbance of the complex (CuCPZ) was measured at 600 nm. To measure the levels of copper or CPZ in samples the opposite constituent was added in excess to first form the CuCPZ complex, followed by the absorbance measurement at 600 nm. After 48 hours, the cells were removed via scraping and extracted for SRM-MS analysis.

3.2.7 SRM-MS Analysis of CPZ Uptake in Cells

Standards and metabolite extracts from cells were resuspended in methanol and processed using the previously described LC gradient on the Micro200 LC coupled with the 5600+ TripleTOF Mass Spectrometer. The ion source nebulizer gas and heater gas were set at 18 psi, and the curtain gas was set at 20 psi. The ionspray voltage was set to

+5000 V with a declustering potential set at +100. A TOF scan was performed over the mass range of 100-500 Da. A product ion scan of 279.18 was collected over a range of

50-300 Da using a collision energy of +25 V. The XIC of transition m/z 279.18→ m/z

139.09 and the MS/MS fragmentation pattern were both use to verify the identify CPZ. A standard curve based on peak height versus concentration of the CPZ standards was constructed for quantification.

3.2.8 Cuprizone Treatment of Mice

All animal experiments were approved by the Institutional Animal Care and Use

Committee (IACUC) of the Cleveland Clinic. 6-week old C57BL/6 male mice were purchased from Jackson Laboratory (Bar Harbor, Maine) and used for all experiments.

Upon arrival, mice were placed on standard chow for 7-10 days. To induce demyelination, mice were fed a diet containing 0.3% cuprizone (bis-cyclohexanone oxaldihydrazone, Sigma-Aldrich), thoroughly mixed into standard chow and custom- made into pellets by Harlan Teklad (Madison, WI), for 2- or 6- weeks ad libitum.

Cuprizone chow was changed twice weekly and the weight of mice was monitored on a weekly basis. At the end of the 2- or 6-week demyelination, mice were perfused with

PBS and fresh tissue was harvested and placed on dry ice for subsequent analysis.

3.2.9 Metabolomic analysis

A modified form of the Bligh and Dyer Extraction was used for metabolite extraction.301,393 Following cuprizone treatment, MO3.13 cells were lysed by the addition of 180 µL HPLC grade water with 20 µL HPLC methanol and removed from the tissue culture dishes by gentle scraping. Tissue samples were normalized by weight and suspended in 200 µL of cold HPLC-grade methanol. The cell/tissue suspensions were

65 then subjected to three cycles of freezing in liquid nitrogen, thawing, and sonication. 750

µL of 1:2 (v:v) chloroform: methanol and 125 µL chloroform were added to each sample.

The samples were vortexed and an additional 250 µL of water was added. Cells were incubated at -20C for one hour and centrifuged at 1000 x g for 10 minutes at 4C to give a two-phase system: an aqueous layer on top, the organic layer below, with a protein disk interphase. The aqueous and organic phases were collected into 1.5 mL tubes separately.

Extracted metabolites were dried in a CentriVap Concentrator (LABCONCO, Kansas,

MO, USA) and then stored at -80oC until analysis. Protein pellets were used to normalize extracted metabolites quantities based on protein concentration with a Bicinchoninic Acid

(BCA) protein assay (G-Biosciences, St. Louis. MO, USA).441 For metabolomic analysis, the dried polar metabolites were first resuspended in 35% acetonitrile (volume normalized via BCA assay) and injected at a volume of 6 μL into a Micro200 LC

(Eksigent, Redwood, CA, USA) equipped with a HILIC column (Luna 3μ NH2 100Å,

150mm×1.0mm, Phenomenex, Torrance, CA, USA). The mobile phases for separation consisted of a water (A) and acetonitrile (B), both supplemented with 5 mM ammonium acetate and adjusted to pH 7.3 using ammonium hydroxide. The gradient proceeded at a flow rate of 30 μL/ min as follows: 0 min 98% B, 1 min 95% B, 5 min 80% B, 6 min

46% B, 13 min 14.7% B, 17 min 0% B, 17.1 min 100% B, 23 min 100% B. After separation, samples were analyzed on an AB SCIEX 5600+ TripleTOF mass spectrometer. The ion source nebulizer gas was set to 15 psi, heater gas 20 psi, and the curtain gas at 25 psi. The samples were collected in positive mode with an ionspray voltage of +5000 V and a declustering potential of +100 V. Samples were processed with

Information Dependent Acquisition (IDA), first utilizing a time of flight scan of 60-1,000

Da with a 250 ms accumulation time and a background threshold of 10 counts/second.

Fragmentation data was collected on all selected candidate ions using a collision energy spread (CES) of + (25-40) V.

3.2.10 Shotgun Lipidomics

Lipid samples were obtained from the dried organic phase after the modified

Bligh-Dyer extraction. The dried extracts were resuspended in a solvent of methanol:chloroform:water (v:v:v, 45:45:10, 5 mM ammonia acetate). Each sample was injected into the mass spectrometer at a flow rate of 10 µL/min. A MS/MSALL method was employed for the lipid analysis442 using an initial scan range of 200-1,200 Da with an accumulation time of 300 ms, proceeded by 1,000 individual MS/MS experiments with a

1.001 Da window width selected via a product ion IDA scan. The parameters were optimized as follows: ion source nebulizer gas was set at 14 psi, heater gas was 15 psi, and the curtain gas was set 25 psi. The method was repeated twice, once in positive and once in negative mode. The DP was set to ±80 V, CE was set to ±10.0 and CES set at

±30.0 V-50.0 V using positive or negative mode, respectively.

3.2.11 Expression and Purification of rhTDO

The expression and purification of rhTDO was performed as previously described.443 Transformed BL21(DE3) cells with the pEThTDO8 plasmid were grown in

10 mL of LB media supplemented with 30 μg/mLkanamycin overnight at 37°C. The overnight culture was then added to 0.5 L of 2×YT media containing 30

μg/mLkanamycin and grown in at 37 °C with shaking until the OD600 nm reached approximately 0.8. Isopropyl 1-thio-β-D-galactopyranoside was added at a concentration of 0.2 mM to induce expression, along with 0.5 mL of 3 mM hemin (in 10 mM NaOH),

67 and the temperature was lowered to 25.0°C. Cells were incubated overnight and then harvested by centrifugation (15 min, 10,000 x g, 4 °C). Cell pellets were stored at -80°C.

The collected pellets were resuspended in lysis buffer (50 mM NaH2PO4, 300 mM

NaCl, 10 mM imidazole, pH 8.0) with protease inhibitors. The solution was then lysed using a french pressure cell. The lysate was centrifuged at 8000 x g for 30 min, and the extract purified by immobilized metal affinity chromatography using a HisTrap FF (GE

Healthcare Bio-Sciences, PA) column on an AKTA HPLC/FPLC system (GE Healthcare

Bio-Sciences, PA). The buffers used for separation contained 50 mM NaH2PO4 and 300 mM NaCl (pH 8.0) with 20 or 250 mM imidazole for buffer A and B, respectively. rhTDO was separated using a linear gradient starting with 0% buffer B and increased to

100% buffer B over 5 minutes at a flow rate of 5 mL/min. rhTDO containing fractions were pooled and dialyzed against 50 mM Tris-HCl buffer. 100 μM aliquots of rhTDO were mixed with 150 μM hemin and incubated on ice in the dark for 1–2 hrs, followed by concentration using the CentriVap Concentrator.

3.2.12 Characterization of TDO with SDS-PAGE gel and SEC

The resolving gel, and stacking gel solutions were made according to the following tables:

Table 3.1

Resolving Gel Solution

Material 10% Gel 12% Gel 15% Gel Water 3.25 mL 2.7 mL 1.92 mL 4X resolving buffer 2 mL 2 mL 2 mL (1.5 M Tris-Cl at pH 8.8) 30% acrylamide 2.7 mL 3.2 mL 4 mL 10% ammonium persulfate 80 μL 80 μL 80 μL Tetramethylethylenediamine 3 μL 3 μL 3 μL (TEMED)

Table 3.2

Stacking Gel Solution

Material Volume Water 4.6 mL 4X stacking buffer 2 mL (0.5 M Tris pH 6.8) 30% acrylamide 1.3 mL 10% ammonium persulfate 48 μL Tetramethylethylenediamine (TEMED) 5 μL

The SDS-PAGE gel was made using 15% acrylamide. After the loading station was prepared the resolving gel was added into the chamber. A small layer of isopropanol was added and it was allowed to dry. After the gel polymerized, the isopropanol was removed, and the stacking gel was poured on top of the station. The comb was added, and the gel was allowed to solidify around the comb. After the gel was prepared, each sample was mixed with loading buffer and heated in boiling water for 5-10 min. The prepared samples were loaded into wells and along with a protein ladder. The gel was run for 1.5 hours at 120 V and stained with Coomassie Brilliant Blue for band visualization.

SEC was performed using a home packed column of Sephacryl S-300HR (GE Healthcare

Biosciences) and equilibrated and run with SEC buffer (50 mM NaH2PO4 and 100 mM

NaCl pH 8.0) on an AKTA HPLC/FPLC system (GE Healthcare Bio-Sciences, PA) at a flow rate of 1 mL/min. Protein elution was monitored with UV detection at 280 nm.

3.2.13 Kinetic Assays of rhTDO

Absorbance spectroscopy of rhTDO and rhTDO with L-tryptophan were collected on a Varian Cary 50 probe UV–visible spectrophotometer with a 1 cm light path. Kinetic reactions were performed at 25.0 °C in 50 mM Tris-HCl buffer (pH 8.0) containing 10

μM methylene blue, 100 μg of catalase, 20 mM L-ascorbate, and 2.5 µM rhTDO in a 96-

69 well plate and read on a Spectramax M2 plate reader. The reaction velocity was

-1 -1 calculated by monitoring the absorbance at 321 nm [ϵ321 ) 3750 M cm ] every 10 minutes for 60 minutes. Reactions were initiated with the addition of L-tryptophan, ranging in concentration from 10-1500 µM. Apparent Km values were determined by fitting the velocity data per concentration of L-tryptophan with either 0, 10, 50 or 100 µM

CPZ to a Michaelis–Menten and Lineweaver Burk plot.

3.2.14 Mass Spectrometry and Absorbance Spectroscopy of P5P and CPZ

All samples for LC-MS analysis were prepared fresh at a final concentration of

100 µM. P5P was prepared in water at a concentration of 100 µM and titrated with NaOH to a pH of 7 to dissolve. CPZ was prepared in ethanol at a concentration of 100 µM. To prepare the P5P+CPZ mixture, P5P was first dissolved in water and titrated with NaOH to pH 7. Stock CPZ in ethanol was then added, achieving a final concentration of 100 µM

CPZ and 100 µM P5P in 1% ethanol. Samples were injected at a volume of 6 μL into a

Micro200 LC with a HILIC column (Luna 3μ NH2 100Å, 150mm×1.0mm, Phenomenex,

Torrance, CA, USA). The mobile phases for separation consisted of a water (A) and acetonitrile (B), both containing 5 mM ammonium acetate and adjusted to pH 7.3 using ammonium hydroxide. The gradient (flow rate of 30 μL/ min) was as follows: 0 min

10% B, 3 min 10% B, 23 min 90% B, 30 min 90% B. The ion source nebulizer gas was set at 18 psi, heater gas was 18 psi, and the curtain gas was 20 psi. The ionspray voltage was set to +5500 V, and the declustering potential was set to +100. The TOF scan was performed over the mass range of 50-500 Da. Fragmentation data for m/z = 348.07 was subsequently collected over a range of 20-500 Da with a collision energy of +20 V.

For absorbance studies, all samples were prepared as described above and then added to a 96-well plate at a volume of at 150 µL with 12 replicates each. The plate was scanned from 300-500 nm at 0, 4, 30, 52, 72, and 86 hours at 27°C.

3.2.15 Preparation of a Schiff base from P5P and Oxalydihydrazide

Scheme 3.1: Preparation of Schiff base complex P5P 1 (0.2 mmol, 50 mg) was dissolved in distilled water (10 mL) at 80 ºC. The resulting solution was added dropwise with stirring to a five-fold excess of oxalydihydrazide 2 (1.0 mmol, 118 mg) in water (15 mL) at 80 ºC. After complete addition of pyridoxal phosphate, the reaction mixture was stirred for another 30 min at 80 °C. The light yellow precipitate was filtered without cooling and washed with hot water 2-3 times and dried

(0.1 mmol, 45 mg, 65% yield) and identified as Schiff-base 3 by NMR (SI).

1H NMR (500 MHz, DMSO-d6): δ12.88 (s, 1H), 12.07 (s, 1H), 10.48 (s, 1H), 9.01 (s,

1H), 8.00 (s, 1H), 5.01 (d, J = 7.73 Hz, 2H), 2.41 (s, 3H),

13C NMR (500 MHz, DMSO-d6): δ 157.15, 156.84, 151.05, 149.13, 149.00, 120.29,

19.34.

LCMS (ESI) m/z calculated for C35H45N5O8: 348.0 [M+H]+, found: 348.070.

3.2.16 Analysis of Transaminase Activity in Cells

MO3.13 cells were seeded in a 6-well plate at a density of 3.0×105 cells/well and left 24 hours at 37°C and 5% CO2. All cells were then treated with 1mM [15N]aspartate.

After adding 0.3% ethanol containing 1 mM CPZ or 0.3% ethanol as a vehicle; cells were incubated for 4 hours or 8 hours, respectively. The cell metabolites were extracted with the Bligh and Dyer method, and the concentration of [14N]glutamate and [15N]glutamate was analyzed by mass spectrometry. All the samples were normalized with the sample protein content measured with a BCA assay. The mobile phases for separation consisted of a water (A) and acetonitrile (B), both supplemented with 5 mM ammonium acetate and adjusted to pH 7.3 using ammonium hydroxide. A sample volume of 5 µL was eluted with a 45 min gradient: 0 min 85% B, 5 min 85% B, 35 min 20% B, 40 min 20% B, 41 min 85% B, 45 min 85% B. To detect glutamate and its isotope, m/z = 148.06 and m/z =

149.06 were selected as the parent ions for [14N]glutamate and [15N]glutamate respectively. A declustering potential (DP) of 70 V of and a collision energy (CE) of 10

V was used to monitor the transitions of m/z 148.061 → m/z 130.051 or m/z 149.061 → m/z 131.051.

3.2.17 NMR analysis of P5P and CPZ

All NMR data was collected at 298 K on an Agilent DD2 750 MHz spectrometer equipped with a HCN cryoprobe. To prepare the sample, P5P was first dissolved in water

(no NaOH) and then titrated with 100 mM CPZ in methanol until reaching a concentration of 1 mM of the predicted product (1:1 mol ratio). This was allowed to sit for 4 hours to achieve a reaction. The product solution had a final concentration of 1% methanol and was filtered through a 4 µm nylon syringe filter. Prior to data acquisition,

1 D2O was added to 8% as the lock solvent. H NMR was collected with 32 transients,

2048 points, and a recycle delay of 25 seconds.

3.2.18 Q-PCR

MO3.13 cells were plated in 6-well plates at 1×106 cells/well and left to adhere for 24 hours. The cultures were then treated with 1 mM CPZ (n=6), or vehicle (n=6) for

16 hours. Cells were collected in separate falcon tubes using trypsin and a Qiagen miRNeasy Mini kit was used to extract both the total RNA and microRNA. RNA samples were normalized to the lowest value using absorbance spectroscopy. Samples were then processed into cDNA using a cDNA Synthesis Kit (04897030001) Roche, and the Light

Cycler 480 SYBR green master mix (04887352001) Roche was used for Q-PCR analysis.

All primers were purchased from OriGene: qSTAR qPCR primers paired against Homo sapiens gene GPT (HP208469), Homo sapiens gene GOT1 (HP205825), and Homo sapiens gene ABAT (HP233728).

3.2.19 Data Processing

Untargeted metabolomics uses an unbiased approach, generating hundreds to thousands of “features” (peaks with a unique m/z ratio and retention time). Features do not necessarily correspond to a characterized metabolite entity. Relevant features must first be selected based on statistical criteria, limiting the baseline noise and aligning peaks.444 Metabolite identification can then be putatively assigned to a peak by using databases like METLIN and HMDB, followed by metabolite validation using the MS/MS data. Initial processing of HILIC-MS data was performed by using MarkerView (version

1.2.1.1).445 Isotopic ion peaks were excluded from analysis. Exploratory statistical analysis of the metabolites was performed using MetaboAnalyst 3.0 73

(http://www.metaboanalyst.ca),446 limited by P<0.05 and -log2 FC≥2. Features were identified by comparing accurate mass and fragmentation data to standards in the

METLIN (https://metlin.scripps.edu)379,447,448 and HMDB (http://www.hmdb.ca/) databases.449 Shotgun lipidomic data were processed using LipidView software (AB

Sciex) with a mass tolerance of 0.05 Da, min % intensity = 0.1% and S/N≥3. Lipids with different chain length were indicated by their fragments’ mass to charge ratio. Unpaired t- tests on all grouped data sets were performed using GraphPad Prism version 5.00 for

Windows, GraphPad Software, San Diego California USA (www.graphpad.com). Data for the 1H NMR was processed and analyzed using ACD/NMR Processor Academic

Edition, version 12.01. In silico MS/MS data was collected using CFM-ID 2.0

(cfmid.wishartlab.com/).

3.3 Results and Discussion

3.3.1 Cellular Uptake and Toxicity of CPZ

In order to identify cell-intrinsic mechanisms of CPZ toxicity, we first developed an in vitro assay of CPZ-mediated death by using the oligodendrocyte cell line, MO3.13.

CPZ is able to impact cell function in a cell-specific manner that is dependent on developmental stage.201 MO3.13 cells were treated with increasing concentrations of CPZ for 24 hours and viability was measured with a LIVE/DEAD fluorescence assay.

Following treatment, only cells treated for the longest time point and at 1 mM displayed any loss of viability or metabolic activity (Figure 3.2a-c, Figure A1a-c). A rat astrocyte cell line was treated with the same concentrations of CPZ but did not show reduced viability (Figure A1d). This indicates that CPZ can disrupt cellular function even in the

Figure 3.2: Cellular uptake and toxicity of CPZ in vitro. (a) A LIVE/DEAD fluorescence assay was used to measure the percentage of live MO3.13 cells treated with increasing concentrations of CPZ at 6, 18, and 24 hours. Concentrations ranged from 0 (vehicle) to 1 mM (black to white) (*P<0.001, n=24 cultures). Fluorescence microscopy of MO3.13 cell treated with vehicle (b) or 1mM CPZ (c) and stained with anti- CNPase antibody after 12 hours of treatment. (d) Absorbance measurements at 600 nm for MO3.13 cells treated with 10 µM CPZ (black, square), 10 µM CPZ and 5 µM CuSO4·5H2O (red, circle), and 5 µM CuSO4·5H2O (blue, triangle). Tissue culture medium was collected every 12 hours (n=6 cultures for each condition). (e) MS/MS data for unbound CPZ in cells matched to an authentic standard. The transition m/z 279.18→ m/z 139.09 was used to determine the concentration of unbound ligand in cells (* marks the parent ion). absence of innate immune cells. The chelation of copper by CPZ in solution results in the formation of a strong peak at 600 nm and we used this property to confirm the uptake of this compound into cells. We treated MO3.13 cells with CPZ, copper or the CuCPZ

75 complex and continuously sampled the tissue culture medium for 48 hours (Figure 3.2d).

The collected medium was then supplemented with the opposite complex component

(copper or CPZ) in excess to form CuCPZ, resulting in the characteristic λmax at 600 nm.173,203 Over time the presence of both CuCPZ and CPZ were diminished in the tissue culture medium suggesting that these compounds could penetrate the cell membrane. In contrast, the presence of copper stayed nearly constant, consistent with the tight control of its uptake into cells. We subsequently confirmed the presence of CPZ within cells by using mass spectrometry. Cellular extracts were collected after 48 hours and analyzed for the presence of unbound CPZ by monitoring the product ion transition m/z 279.18 → m/z

139.09 (Figure 3.2e). We detected unbound ligand in cells at a concentration of approximately 30 nM (Figure A2a,b) after treatment with 10 µM CPZ, indicating its ability to cross the membrane and accumulate in cells. Our data is consistent with a study in which mature rat oligodendrocytes had reduced viability when treated with CPZ;201 however, relatively high concentrations of CPZ are required to induce cell death in our system and this may explain the difficulty in establishing consistent in vitro models of demyelination with this compound.

3.3.2 CPZ induces metabolic dysregulation in vitro and in vivo

The ability of CPZ to induce cell death in cultured cells indicates that some component of its toxicity is due to the direct disruption of oligodendrocyte function.

Therefore, we sought to examine metabolic pathways that are altered by CPZ treatment using a global metabolomic approach. Cells were treated with 1 mM CPZ or vehicle followed by metabolite extraction and LC-MS analysis. We first identified alterations in cellular pathways that correlated with CPZ toxicity. Pathways that were significantly

76 dysregulated include NAD+ metabolism, cellular antioxidant capacity, ammonia homeostasis, and vitamin B6 metabolism (Figure 3.3a). Perturbations in energy generation have been suggested to play a role in CPZ toxicity and megamitochondria are seen after long-term treatment.207,214,450 In agreement with this, NAD+ metabolism was significantly dysregulated in CPZ-treated cells. NAD+ is formed in cells through two pathways, de novo synthesis using tryptophan and salvage from nicotinamide-containing compounds.451 Tryptophan and 5-hydroxytryptophan are both significantly increased in

MO3.13 cells after CPZ treatment, while kynurenine and nicotinamide levels are decreased (Figure 3.3b-e, Figure A3). Recently, it was shown that myelinating oligodendrocytes perform aerobic glycolysis as an energy source and this requires lactate dehydrogenase activity to regenerate NAD+.452 This metabolic state is associated with

Warburg metabolism and also occurs in rapidly proliferating cancer and stem cells.453,454

Myelination also requires an upregulation of NAD+-consuming enzymes, such as sirtuins,455 indicating that the preservation of NAD+ levels is critical for proper myelin maintenance.456 The disruption of the de novo synthesis and salvage pathways to produce

NAD+ may lead to a block in glycolysis and energy depletion that would preferentially affect these mature oligodendrocytes. Other metabolic changes include downregulation of reduced glutathione, an antioxidant tripeptide, as well as one of its precursor amino acids, cysteine (Figure 3.3f,g, Figure A4). Finally, members of the B vitamin family, choline and pyridoxine are also significantly downregulated by CPZ treatment (Figure

3.3h,i, Figure A4).

We sought to extend our in vitro studies by examining metabolites that are altered in tissue samples of the corpus callosum, hippocampus, and spinal cord taken from mice

Figure 3.3: CPZ perturbs the metabolism of MO3.13 cells. (a) Pathway analysis of metabolic changes induced by CPZ treatment of MO3.13 cells as determined by MetaboAnalyst 3.0. The overview shows all matched pathways according to p-values from pathway enrichment analysis and pathway impact values from pathway topology analysis (varying from yellow to red). The node color is based on its p value and the node radius is determined based on their pathway impact values. (b-i) Box and Whisker plots of significant features (P<0.05 t-test, fold change >2) identified both with parent ion and fragment ion information compared between the CPZ (n=10) and control (CON) (n=10) treated MO3.13 cells * P<0.001). 5-HT= 5-hydroxytryptophan and GSH=reduced glutathione.

78 fed a diet of 0.3% CPZ for a total of two or six weeks. In the mouse model of CPZ demyelination, the corpus callosum suffers a loss of oligodendrocytes by 6 weeks, causing lesion formation, along with demyelination in the hippocampus and cortex.196,457

Interestingly, mice do not form lesions in the spinal cord after CPZ treatment, but this regional variability remains unexplained.231,430 An examination of the total number of dysregulated features after the two treatment time points shows a pronounced difference in the hippocampus and corpus callosum when compared to the spinal cord, matching the extent of tissue demyelination in these regions (Figure 3.4a). Furthermore, a comparison of individual metabolites in the corpus callosum, hippocampus, and spinal cord shows that metabolic alterations in the spinal cord are opposite of those detected in corpus callosum and hippocampus (Figure 3.4c, Figure A5-9).

Similar to our cell data, perturbations in amino acid metabolism were particularly prominent in regions that undergo demyelination at both two and six weeks (Figure 3.4b).

Significant downregulation of amino acids or related metabolites such as glutamate, homocysteine, and cysteine were seen progressing from two to six weeks of CPZ treatment in both the corpus callosum and hippocampus, but did not change or were upregulated in the spinal cord (Figure 3.4c). Glutamine and tryptophan were upregulated in the CC and HP, while these metabolites did not change in the spinal cord. Amino acid perturbations have been documented in a shorter study (4 days of CPZ exposure) in the plasma of mice458 and our results are in line with this study. Products of amino acid metabolism also showed dysregulation after CPZ treatment. Reduced glutathione was downregulated in the brain but slightly upregulated in the spinal cord. Glutathione homeostasis works as an important antioxidant system to protect against reactive oxygen

Figure 3.4: CPZ induces region specific alterations in metabolism in the central nervous system. Global metabolic profiling of corpus callosum (CC), and spinal cord (SC) from C57Bl/6 mice after 2- or 6- weeks of treatment with CPZ and hippocampus after 6 weeks of CPZ treatment (n=10 mice per condition). (a) The percentage of dysregulated metabolic features (P<0.05, fold change >2) found in the CC, HP, and SC in CPZ-treated versus control mice. (b) Pathway analysis of metabolic changes induced by cuprizone in the brain. Pathways were constructed based on putatively identified metabolites found dysregulated in both the CC and HP based on information obtained with Metaboanalyst™. The node color is based on its p-value and the node radius is determined based on their pathway impact values. (c) Heatmap of metabolites changing in the CC, HP, and SC of CPZ-treated versus control mice (P<0.05,). Metabolites were identified with accurate mass and fragmentation information. The generated heatmap ranges from -log2 fold change 6 (red) to -6 (blue). Grey represents a metabolite with a matching accurate mass but no confirmative MS/MS data.

80 species, and the depletion of reduced glutathione has been implicated in several neurodegenerative diseases.459–461 The maintenance of reduced glutathione pools may facilitate oligodendrocyte protection in the spinal cord. Metabolites associated with 1C metabolism such as folate and tetrahydrofolate (THF) are increased in the brain; however, other metabolites that function in this pathway such as s-adenosylmethionine

(SAM) and 5-methyl THF are downregulated. Proteomic investigations previously have linked CPZ effects to the disruption of pathways associated with mitochondrial function and oxidative stress.462,463 Overall, the patterns detected from both cell and tissue investigations reflect a metabolic dysfunction of amino acid metabolism that may lead to a reduction in antioxidant molecules and perturbation of energy-generating pathways.

Interestingly, the region-specific vulnerability was recapitulated in the metabolic profiles of the different tissues, as spinal cord showed the lowest number of significant features changing upon CPZ treatments detected by LC-MS. Recently single cell transcriptomics has been used to identify region-specific heterogeneity in oligodendrocyte populations in the mature CNS.464 Our data suggests that metabolic heterogeneity may also occur in a localized manner and that these distinct populations may display differing functional responses after toxin administration.

CPZ demyelination has been examined extensively through histological analysis; however, global perturbations in lipid homeostasis may also occur in tandem with myelin loss. We also performed lipidomic analysis on tissue isolated from animals fed CPZ for two and six weeks. These time points represent a pathological continuum characterized by oligodendrocyte dysfunction, an overt loss (two weeks) to complete demyelination of

CC (six weeks).248 After two weeks of CPZ feeding, lipid levels in the corpus callosum

Figure 3.5: CPZ perturbs lipid homeostasis in the corpus callosum, but not the spinal cord. Shotgun lipidomics was performed on CNS tissue from CPZ-treated and control mice. Lipid profiles are expressed as the fold change ratio of the extracted corpus callosum (black) or the spinal cord (grey), using the peak intensity fold change of CPZ treated samples to vehicle treated samples. A ratio of one represents no change between the CPZ and the control sample group. Ten mice were analyzed for each condition at either 2-weeks (a) or 6-weeks (b). The relative intensity of (c) ceramide and (d) DAG lipids detected in the corpus callosum by shotgun lipidomics. The relative intensity for each lipid species for control tissue (diagonal fill) and CPZ-treated tissue (black) are the average of 10 mice per group. (All Cer and Dag lipid changes have a P<0.001).

82 showed slight changes compared to control, primarily downregulation of ceramides and diacylglycerols. In contrast, spinal cord lipid species, including phosphatidic acid and phosphatidylethanolamine, were upregulated at this early time point (Figure 3.5a). By six weeks, the CC lipidomic profile showed a reduction in most lipid classes, especially diacylglycerols and ceramides while the relative levels of spinal cord lipid species were preserved compared to controls (Figure 3.4b). The loss of both saturated and unsaturated ceramide species occurred after CPZ treatment when compared to controls (Figure 3.5c).

Glycosphingolipids, which consist of a ceramide core, are one of the major components of myelin465 and have been demonstrated to play a role in the functional regulation of oligodendrocyte differentiation.466 Depletion of these species would be expected to impact myelin structure and formation. Interestingly, neutral lipids, especially diacylglycerol (DAG) species, were depleted after CPZ treatment (Figure 3.5d). We found that choline was also significantly down-regulated in both the hippocampus and corpus callosum (Figure 3.4c). These changes (DAG and choline) might indicate impaired de novo synthesis of phosphatidylcholine (PC) species. This idea is supported by the ability of exogenous CDP-choline administration to promote remyelination after

CPZ feeding.246 Overall, the lipidomic results match well with histologic changes in myelin observed in affected brain regions.

3.3.3 CPZ alters nicotinamide production in vitro

One hypothesis of CPZ’s metabolic effect highlighted the deregulation of the tryptophan/nicotinamide pathway based on the documented low levels of nicotinamide paired with the high levels of tryptophan. The kynurenine pathway is responsible for the de novo synthesis of nicotinamide from tryptophan, and final production of NAD+

(Scheme 3.2). We hypothesized that the

toxic effect of CPZ could originate from the

inhibition of the first rate-limiting heme-

bound enzyme of the kynurenine pathway,

which can be indoleamine-2,3-dioxygenase

(IDO) or tryptophan-2,3-dioxygenase

(TDO),467,468 inducing energetic failure in

oligodendrocytes. IDO is ubiquitously

expressed, and its expression can be

elevated by proinflammatory stimuli such as

interferon-γ (IFN-γ), and lipopolysaccharide

(LPS) in proinflammatory cells.469,470 IDO

activity modulates T-cell response, causing Scheme 3.2: Tryptophan/Nicotinamide pathway. Intermediate metabolites are shown in black, and enzymes are shown in blue. Metabolite a suppression of T-cells through the abbreviations are as follows: nicotinic acid mononucleotide (NaMN), deamido-NAD (NaAD), depletion of tryptophan. IDO activity has nicotinamide mononucleotide (NMN).594 also been shown to enhance tumor progression via immunosuppression in animal models.471–473 In comparison to IDO, the role of TDO has been less explored, and its expression is limited mostly to the liver but has also been found in the testis, uterus, and in neurons.474–477 TDO-deficient (TDO−/−) mice are known to develop behavior issues and display alterations in neurogenesis.478

The malfunctioning of TDO has also been implicated in Alzheimer's disease,479 and was recently proposed to initiate neuroprotection when its expression was decreased in the

EAE model.480 We hypothesized that oligodendrocytes use TDO and that it could be a

Figure 3.6: Purified recombinant human tryptophan 2,3-dioxygenase (rhTDO) is not inhibited by CPZ. (a) SDS-PAGE 15% acrylamide gel stainded with Coomassie Brilliant Blue. From left to right the gel lanes were loaded with a ladder, the flow through from the Ni-histrap column, the fractions purified from the Ni- histrap column, and the fractions after dialysis. (b) Post-dialysis purified rhTDO run on a Sephacryl S- 300HR size exclusion column at 1 mL/min. Absorbance was measured at 280 nm over 165 minutes. Peak sizes were determined based on previously calibrated standard proteins, and each peak is labeled based on this calibration. (c) Absorbance scan of protein buffer (green), 2.5 µM rhTDO in buffer (red), and 2.5 µM TDO in buffer with 50 µM L-tryptophan (L-Trp) (blue). Spectral changes observed upon the reaction of rhTDO with L-Trp under aerobic conditions results in a peak at 321 nm from the product N- formylkynurenine, while the soret peak at 410 is common of porphyrin compounds (in this case heme). Both peaks prove rhTDO enzymatically active. (d) Kinetic absorbance scan of 2.5 µM rhTDO with 50 µM L-tryptophan over 60 minutes. Peak at 321 nm from N-formylkynurenine increases over the observed time. A scan was made every ten minutes (red to purple) (e) Michaelis–Menten and (f) Lineweaver burke plot of varying concentrations of tryptophan with 2.5 µM rhTDO in buffer with 0 (red), 10 (blue), 50 (green), or 100 (purple) µM CPZ. Plots both show no change in rhTDO’s enzymatic activity with CPZ additions.

85 target of CPZ in the brain and liver. Oligodendrocytes have high energetic demands in order to produce myelin, potentially leaving them susceptible to the dysfunction of this de novo pathway, while other cells may prefer the salvage pathway to produce NAD+.

The supplementation of nicotinamide to MO3.13 cells after CPZ treatment did rescue metabolic viability (Figure A10), seemingly negating the toxic effects of CPZ on oligodendrocytes. To further explore the possible direct inhibition of TDO via CPZ, we expressed recombinant human TDO (rhTDO) using a previously characterized vector and purification method (Figure A11).443 The homotetrameric heme-bound protein was confirmed using an SDS-Page gel, and size exclusion chromatography (SEC) (Figure

3.6a,b). The SEC spectra clearly distinguish four species: a monomer, dimer, tetramer, and multimer. The tetramer was isolated for further enzymatic studies. Using absorbance spectroscopy the characteristic heme Soret band was documented at 410 nm, as previously described for ferric TDO (Figure 3.6c).443 rhTDO activity was measured by monitoring the formation of the enzyme product, N-formylkynurenine, at 321 nm.443,481

Figure 3.6d shows a representative kinetic plot for the production of the N- formylkynurenine following initiation with L-tryptophan. The Km value of rhTDO was determined by measuring the linear velocity with varying concentrations of L-tryptophan using both a Michaelis-Menton curve and Lineweaver Burk plot (Figure 3.6 e,f). Our calculated Km of 169.16±15.46 µM fits within the range of previously calculated values.443,482 When CPZ was added at varying concentrations along with rhTDO and L- tryptophan, there was no significant change in the enzymatic activity. With this discovery, we turned to other enzymatic candidates involved in CPZ intoxication.

3.3.4 CPZ binds pyridoxal 5’-phosphate perturbing aminotransferase activity

Our metabolomic results indicate that amino acid metabolism is a major target of

CPZ in the CNS. Enzymes that utilize amino acids as substrates often use the active form of vitamin B6, P5P. P5P-containing enzymes act largely as aminotransferases, but also play roles in neurotransmitter synthesis, transsulfuration, glycogenolysis, sphingosine biosynthesis, and the conversion of tryptophan to nicotinic acid.483,484 Metabolites in these pathways were dysregulated early and late after CPZ feeding (Figure 3.4c) and a derivative of P5P, pyridoxine, was downregulated in CPZ-treated oligodendrocytes

(Figure 3.3i). The coenzyme P5P is also known to interact with hydrazine derivatives,485,486 thus we explored potential chemical interactions between CPZ and P5P as well as its possible downstream impact on aminotransferase activity in oligodendrocytes.

Addition of CPZ to P5P causes an immediate shift in the absorbance at 350 nm, as previously reported,267 which is stable for at least 86 hours (Figure 3.7a). A comparison of a solution of P5P+CPZ to CPZ, or P5P alone, by LC-MS, showed the appearance of a major peak with a retention time of 8.48 minutes at m/z = 348.070 (Figure 3.7b, Figure

A12). A secondary peak was also noted at m/z = 428.133. These peaks correspond to a predicted complex involving CPZ hydrolyzed to form oxalylhydrazide bound through a

Schiff base to P5P ([CPZ-2R+P5P]+), and a similar complex retaining one cyclohexane ring ([CPZ-1R+P5P]+), respectively. As the [CPZ-2R + P5P]+ species at m/z = 348.070 is the main analyte in solution, we sought to validate our proposed structure. We performed in silico MS/MS to predict a fragmentation pattern and subsequently synthesized a [CPZ-2R+P5P]+ standard to verify the structure (Figure 3.7c,

Figure 3.7: Pyridoxal 5’ Phosphate is a potential in vivo target for CPZ. (a) Absorbance spectroscopy of a CPZ- P5P solution (red), compared to P5P (blue). Both solutions were incubated from 0-86 hours to ensure stability (darker to lighter lines). (b) Solutions of P5P (red), CPZ (blue), and CPZ+P5P (black) were examined with LC-MS. A ToF scan was performed from 50-500 Da for each solution. The relative intensity for each analyte is shown. All peaks from the CPZ+P5P solution overlap with known peaks, from either CPZ or P5P, except m/z = 348.070 and m/z = 428.133 (labeled). Each novel peak has the predicted complex annotated above the peak, where CPZ-R represents a CPZ with one hydrolyzed cyclohexane ring, and CPZ-2R represents a CPZ with two hydrolyzed cyclohexane rings. (c) MS/MS data for m/z=348.070 (collision energy of 20V, * marks the parent ion) from the CPZ+P5P solution matched to a synthesized standard. (d) 1H-NMR data of the P5P-CPZ solution indicating the presence of the Schiff base via the down field shifted amino group. Large unlabeled peaks are attributed to sample solvent (methanol and water, respectively). (e) (Top) Cells were treated for 4 or 8 hours with CPZ after being pulsed with [15N]aspartate and the ratio of [15N]glutamate to unlabeled glutamate was measured for MO3.13 cells either treated with CPZ (grey) or vehicle (black). (*P<0.001, n=6 cultures, representative of duplicate independent experiments). (Bottom) A scheme showing the reaction catalyzed by AST with the stable isotope labeled nitrogen in red. 88

Figure 3.8: CPZ treatment of MO3.13 cells results in a decrease in transaminase gene expression after 16 hours. Q-PCR was used to analyze mRNA extracted from MO3.13 cells treated with 1 mM CPZ or vehicle for 16 hours with primers for the following transaminases: ABAT (GABA transaminase), AST (Aspartate transaminase) and ALT (Alanine transaminase). (Top) Melting curves indicate primer specificity. Curves shown in blue were in the range of normal specificity, while the curves in red (ALT) suggested non-specific binding of the primer. Fold change data for ALT was thus not included. (Bottom) Fold change of expression between control and CPZ-treated samples for ABAT and AST (*** P<0.001, n=6).

Figure A12a). The synthesized structure’s retention time (Figure A12b) and fragmentation data provide confirmation of the ability of CPZ to interact with P5P.

Nuclear magnetic resonance (NMR) data also supports a positively charged Schiff base linkage between P5P and CPZ (Figure 3.7d). The resonance of the proton on P5P coupled to the carbon of the Schiff base is shifted dramatically downfield (8.75 ppm) of

89 the aromatic proton (7.75 ppm) suggesting the presence of a positively charged amino group. The chemical shifts of the aromatic proton (7.75 ppm) and Schiff base proton

(8.75 ppm) also demonstrate the charged nitrogen in the heterocyclic ring. The same 1H

NMR peaks match the synthesized [CPZ-2R+P5P]+ (Figure A13).

Having support for the isolated reaction of P5P and CPZ, we chose to examine the effect of CPZ on aspartate aminotransferase (AST). AST is a prominent glial aminotransferase, which catalyzes the reversible transamination between aspartate and - ketoglutarate to form oxaloacetate and glutamate.487 We monitored AST activity in the presence of CPZ by feeding cells [15N]aspartate and examining the production of

[15N]glutamate (Figure 3.7e). The percentage of labeled [15N]glutamate was significantly decreased in CPZ treated cells at both four and six hours compared to vehicle controls, indicating that the transfer of the amino group catalyzed by AST is reduced in the presence of CPZ. Q-PCR was also used to confirm the downregulation of both AST and ABAT (GABA aminotransferase) after 16 hours of CPZ treatment (Figure

3.8). Overall, this data supports the idea P5P is an early target of cell intrinsic CPZ toxicity.

3.4 Conclusions

In this study, we have sought to identify oligodendrocyte-specific biochemical toxicity associated with CPZ treatment. We confirm that there is cellular uptake of CPZ into cells and unbound ligand is internalized in vitro. Global metabolomic profiling of both CPZ-treated cells and tissues, from 2-week and 6-week time points in the CPZ demyelination animal model, indicate a disruption of shared metabolic pathways directly

90 related to amino acid metabolism, NAD+ generation, glutathione metabolism, the 1C cycle, and lipid synthesis. These changes are region specific and correlate to areas that develop demyelinating lesions. While exploring enzymatic targets, we found that TDO’s activity was not directly altered by CPZ. However, our spectroscopic data indicates that

CPZ reacts with P5P, revealing a capacity for CPZ to disrupt metabolism through the perturbation of enzymes, such as AST, requiring the P5P coenzyme. Our results point to region specific perturbations in amino acid metabolism that potentially represent a unique metabolic vulnerability by oligodendrocytes.

CHAPTER IV

DEMYELINATION AND DIFFERENTIATION IN THE CPZ/RAPAMYCIN MODEL

4.1 Introduction

Myelination of the CNS provides critical support for neuronal axons and enables rapid nerve signal transduction. Oligodendrocyte cells are responsible for the formation and upkeep of the protective myelin sheath, which is composed of a spirally wrapped plasma membrane covering neuronal axons. Myelin is produced, specifically by mature oligodendrocytes, which differentiate from OPCs. Loss of the myelin sheath

(demyelination) causes serious neurological impairment and damage associated with many neurodegenerative diseases, including multiple sclerosis. One model used to better understand the mechanisms of demyelination and remyelination is the CPZ murine model. In the model, C57BL/6 mice are orally delivered the small molecule toxin CPZ,

92 resulting in the cell-specific death of oligodendrocytes.201,457 Demyelination occurs in oligodendrocyte rich regions, prominently the hippocampus, corpus callosum, and other sections of white matter. Alone, CPZ causes loss of mature oligodendrocyte populations by 6 weeks and a 60% decrease in myelin, nevertheless, OPCs migrate into lesions and begin to repair the myelin sheath. The variability introduced by actively proliferating and differentiating OPCs can lead to difficulties interpreting molecular changes that are due to either the demyelinating process versus remyelination.

A new CPZ model been developed that eliminates the OPC population by targeting the PI3K/Akt/mTOR pathway.263 mTOR (mammalian/mechanistic target of rapamycin) is a serine/threonine kinase involved in cellular growth, proliferation, survival, transcription, translation, and motility and is present as two complexes: mTOR complex 1 (mTORC1) and mTOR complex 2 (mTORC2) (Figure 4.1).488,489 mTORC1 consists of mTOR complexed with proteins; the mammalian lethal with SEC13 protein 8

(MLST8) and the regulatory-associated protein of mTOR (RAPTOR).490–492 Additional protein factor inhibitors include Proline-rich AKT1 substrate 1 (PRAS40) and DEP domain-containing mTOR-interacting protein (DEPTOR).493,494 Numerous stimuli regulate mTORC1 including: phosphatidic acid, amino acids levels, insulin, growth factors, and oxidative stress.495–497 mTORC1 can be activated upstream by different signaling pathways, including Rheb-GTP, Akt/PKB, MAPK/ERK, and Wnt.498–500 After activation mTORC1 triggers transcription and translation via p70-S6 Kinase 1 (S6K1) and the eukaryotic initiation factor 4E binding protein 1 (4E-BP1), increasing protein production.498 mTOR, rapamycin-insensitive companion of mammalian target of rapamycin (RICTOR), PROTOR, and mammalian stress-activated protein kinase

Figure 4.1: mTORC1 and mTORC2 Complexes. The mTOR kinase is occurs via two distinct protein complexes: mTORC1 and mTORC2. Amino acids, oxygen, phosphatidic acid, insulin and growth factors activate mTORC1. mTORC1sensitive to stress and is inhibited by rapamycin. mTORC1 promotes cell growth, drives cell-cycle progression, induces anabolic metabolism. mTORC2 is activated mainly by growth factors and regulates the cytoskeletal structure, survival, and cell metabolism. Only chronic rapamycin exposure can directly inhibit mTORC2. The circles represent the protein components that make up each mTOR complex. Purple small circles are normal scaffolding/required components, while green small circles are inhibitors.595 interacting protein 1 (mSIN1) make up mTORC2.501–504 mTORC2 activation regulates

proteins associated with cytoskeletal rearrangement including protein kinase c alpha

(PKCα), F-actin stress fibers, and paxillin.490 mTORC2 additionally regulates cellular

metabolism and proliferation through serum and glucocorticoid-induced protein kinase

(SGK), and Akt/PKB.505

The mTOR-specific inhibitor rapamycin (RAP), is an approved

immunosuppressant originally developed as an antifungal and currently used to prevent

organ transplant rejection.506,507 RAP has also been associated with promoting tumor

regression, treating tuberous sclerosis complex, preventing cognitive deficits in the

94 mouse model of Alzheimer’s, and lifespan extension in yeast and mice.508–514 When introduced, RAP binds the FK-binding protein 12 (FKBP12), forming a complex. RAP-

FKBP12 binds the FKBP12 site on free mTOR, inhibiting mTOR autophosphorylation and allosterically blocking the binding of RAPTOR preventing the formation of mTORC1.498,515,516 RAP-FKBP12 does not interact with mTORC2, and so it was originally thought to be RAP insensitive.490 However, chronic treatment, or high doses of

RAP have a cell-dependent effect, eventually confining enough free mTOR so that it is unable to regenerate mTORC2. Once mTORC2 levels drop, the phosphorylation and activation of Akt/PKB can no longer be sustained.517

Understanding the signaling mechanism and the development of oligodendrocyte cells is key to developing advances in myelin repair, as inefficient remyelination due to improper oligodendrocyte differentiation occurs in human demyelinating disorders.518,519

Oligodendrocytes progress through a series of morphologically distinct stages characterized in vitro and in vivo.93–95 Several investigations have connected myelin formation and integrity with the PI3K/Akt pathway,520–522 and the proliferation, migration or survival of oligodendrocytes with the growth factor-induced P13K/Akt/mTOR pathway.100,101,103–108,110 The oligodendrocyte-specific ablation of RAPTOR in vitro produced dysmyelination in the CNS.523 Differentiation appears intimately regulated by mTOR pathways, as addition of RAP,111,260,263,524–526 as well as siRNA-mediated knockdown of mTOR,527 halts OPC differentiation in vitro. Primary OPCs cultured with

15 nM RAP displayed decreased morphological complexity, where GalC negative, and had reduced expression of CNPase, PLP, and MBP.260 When rapamycin treatment was delayed, the rate of differentiation of late progenitors to immature oligodendrocytes was

95 not affected, when assessed by stage-specific antigens (O1), however, cells were morphologically affected and displayed simpler multiprocess.111

In the investigated CPZ with RAP (C/R) model, mice are administered 10 mg/kg

RAP along with 0.3% CPZ over the course of six weeks. The addition of RAP with CPZ leads to, a greater loss of myelin, slower remyelination, delayed maturation in the CC and loss of mature cells in the CC.263 It should also be noted that astrocyte and microglial immunoreactivity increased slightly in the C/R model (no activity was noted in the RAP control) when compared the CPZ model.263 The C/R model causes demyelination of mouse white matter and the cerebral cortex, with faster remyelination occurring in the cerebral cortex. These properties more accurately replicated phenotypes found in postmortem MS brains and with the introduction of thyroid hormone triiodothyronine to promote remyelination, a new model was developed for testing remyelinating therapies.528 The C/R model has significant advantages for exploring the cellular and biochemical requirements for demyelination separate from remyelination by creating a distinct timeline of demyelination removed of unconstrained OPC maturation. The model can also inform on the involvement of mTOR in myelin production, providing targets that could enhance remyelination in diseases such as MS.

The mTOR signaling pathway is sensitive to many environmental and hormonal cues, allowing it to uniquely modulate cellular metabolism.529,530 To better understand the collective demyelinating effects of CPZ and RAP on the CNS, we integrated untargeted metabolomics with transcriptomics to correlate the activities of RAP-induced metabolic pathways with specific genetic changes to oligodendrocyte differentiation. We used LC-

MS to elucidate biochemical pathways important for cell survival, growth, and

96 maintenance by monitoring changes in endogenous small molecules. The expression of all mRNA was also monitored using a microarray. Microarrays measure gene expression levels by the hybridization of genetic probes to cDNA or cRNA from samples.531,532 The hybridized product is quantified using a fluorescent or chemiluminescent label. We have used these two OMICs approaches to identify biochemical signals necessary for OPC differentiation. We found that branched chain amino acids (BCAA) may play an important role in driving OPC maturation and myelin production.

4.2 Methods

4.2.1 Chemicals

Phorbol 12-myristate 13-acetate (≥99%), L-Valine (Bioultra, ≥99.5%), L-Leucine

(Bioultra, ≥99.5%), Poly-D-lysine hydrobromide (average mol wt 30,000-70,000, lyophilized powder, γ-irradiated, BioReagent, suitable for cell culture), Rapamycin, and

Anti-Galactocerebroside Antibody (clone mGalC, Alexa Fluor®488) were all purchased from Sigma Aldrich (MilliporeSigma, St. Louis, MO, USA). DNase I, and the oligodendrocyte marker O1 monoclonal antibody (O1) conjugated toeFluor 660, were purchased from ThermoFisher (part of eBioscience, Fair Lawn, NJ, USA). All other reagents have been previously detailed in the preceding chapter methods.

4.2.2 Cell Culture and Differentiation

MO3.13 cells were cultured in DMEM with 10% FBS, and 1% penicillin/streptomycin. For all experiments, the cells were maintained in an incubator set at 37C with 5% CO2. MO3.13 cells monitored with brightfield imaging were plated in

24-well plates at a density of 5×104 cells/well. The cells were immediately treated with either 100 nM PMA (n=3), 100 nM PMA and 1 mM valine (n=3), 100 nM PMA and 10 mM valine (n=3), 100 nM rapamycin (n=3), 1 mM leucine (n=3), or 100 nM rapamycin and 1 mM leucine (n=3) and were grown for 3 days. Images were taken using an inverted tissue culture microscope (20X) and Amscope software.

4.2.3 Animals and Primary Cell Cultures

The mice used in this work were cared for according to the IACUC of the

University of Akron. Ethical approval for experiments conducted was obtained protocol number 17-02-01-SMRD. Breeding was done in-house at animal facilities of the

University of Akron. Stock CD1 mice were purchased from Charles River Laboratories,

Ashland, OH.

Primary glial cultures were isolated from the brain tissue of 1-2 day postnatal

CD1 mice. Mice were decapitated and heads were stored in a sterile Falcon tube with cold Hank’s buffered saline solution (HBSS). Cortices were dissected and placed in a clean Petri dish containing cold HBSS. After removal, the tissues were placed in a clean falcon tube along with approximately 5 mL of the HBSS and a single cell suspension was made by mechanical trituration with a glass Pasteur pipette. 0.1 mL DNase I stock solution (0.2 mg/mL) and 0.1 mL trypsin (0.25%) were subsequently added to the cell suspension, and the solution was left to incubate for 15 min at 37 °C. Cells were pelleted with swinging bucket centrifugation at 1000 rpm for 10 min. The supernatant was removed and replaced with 5 mL DMEM supplemented with 10% FBS, and 1% penicillin/streptomycin. The cell pellet was then further homogenized in solution, and passed through a 40 µm cell strainer and then a 20 µm cell strainer. The cells were

98 counted and plated in T75 culture flasks at a density of 1.0×105 cells/mL. Cells were grown for 10 days, or until confluence, at 5% CO2 and 37 °C.

For the lipidomic analysis, mixed glial cells were seeded in 6-well plates, at a density of 1.0×106 cells/well, and left to grow for 3 days in medium containing 100 nM rapamycin (n=6), or the vehicle control (n=6). After incubation, a metabolite extraction was performed, and shotgun lipidomics was used for analysis.

4.2.4 Confocal Imaging

For imaging experiments, MO3.13 cells were seeded onto glass coverslips coated with Poly-D-lysine in a 6-well plate at a density of 5×104 cells/well. The media was supplemented with either 100 nM PMA, 100 nM PMA and 10 mM valine, or a vehicle control. Following 3 days of treatment, the media was removed and the cells were washed with PBS. Cells were fixed in 4% paraformaldehyde for 10 minutes and then permeabilized by using 1% Tween for 20 minutes. Cells were then left to block in 1%

BSA/10% FBS for 2 hours followed by an overnight at 4°C incubation with the Alexa

488 anti galactocerebroside marker (10 µg/mL) or the oligodendrocyte marker O1 eFluor

660 antibody (20 µg/mL). Following incubation, the cells were washed twice with PBS and DAPI was used to stain the cell nuclei. Cells were then mounted and imaged (600x) using a Nikon A1+ confocal microscope.

4.2.5 Cell Counting

MO3.13 cells were seeded 24-well plates at a density of 5×104 cells/well and treated with 100 nM PMA (n=6), 100 nM PMA and 1mM valine (n=6), 100 nM PMA and

10 mM valine (n=6) or the vehicle control (n=6) all in DMEM. The cells were left to incubate for 3 days to allow for differentiation. Following differentiation, cells were 99 collected after the addition of trypsin and spun down at 1000 rpm for 10 minutes and the supernatant was aspirated. Each sample was resuspended in 100 µL of DMEM. Cell counts were performed on each sample by mixing with 80 µL Trypan blue with 20 µL of homogeneous cell solution. The dyed mixture was then introduced onto a hemocytometer and an average live cell count from all four grids was obtained. Total solution cell counts were calculated as follows:

퐴푣푒푟푎𝑔푒 푐푒푙푙 푐표푢푛푡 푚푚2 𝑔푟𝑖푑 ×푑푖푙푢푡푖표푛 푓푎푐푡표푟 (5)× ×푚퐿 푐푒푙푙 푠표푙푢푡푖표푛 (0.1 푚퐿) = 푡표푡푎푙 푐푒푙푙 푐표푢푛푡 푚푚2 𝑔푟𝑖푑 0.0001 푚퐿

4.2.6 RAP and C/R Animal Model

All animal experiments were approved by the IACUC of the Cleveland Clinic. 6- week old C57BL/6 male mice were purchased from Jackson Laboratory (Bar Harbor,

Maine) and used for these experiments. Upon arrival, mice were placed on standard chow for 7-10 days. Food and water were available ad libitum, and mice were weighed weekly. To induce demyelination for the C/R model, mice were fed a diet containing

0.3% CPZ (biscyclohexanone oxaldihydrazone, Sigma-Aldrich), thoroughly mixed into standard chow and custom-made into pellets by Harlan Teklad (Madison, WI), for 6- weeks ad libitum. In addition, the mice received injections of rapamycin (10 mg/kg) 5 days a week dissolved in the 5% polyethylene glycol 400 (PEG 400), and 5% Tween 80.

Cuprizone chow was changed twice weekly and the weight of mice was monitored on a weekly basis. For the RAP model, mice were fed control chow, containing no CPZ and while receiving the aforementioned RAP injections. A third group of control mice were administered injections of the vehicle, while on normal chow. At the end of 6 weeks, mice were perfused with PBS and fresh tissue was harvest and stored at -80°C until subsequent analysis.

100

4.2.7 Metabolomic and Lipidomic Analysis

All extraction and analysis parameters for both polar metabolites and lipids were performed as previously discussed in the preceding chapter methods. Tissue was extracted using a modified Bligh and Dyer procedure, and the aqueous (polar metabolites) and organic (lipids) layers were separated and analyzed separately. Polar metabolites were separated and using HILIC-LC and analyzed using an ESI-MS/MS protocol with IDA on a TripleTOF 5600+. Lipids were directly injected and analyzed using an MS/MSALL method with IDA.

4.2.8 Microarray for Transcriptomic Analysis

MO3.13 cells were plated in T25 flasks at 1×107 cells/flask the cultures were treated with 100 nM PMA (n=3), or vehicle (n=3) and grown to confluence over 3 days.

Cells were collected in separate falcon tubes using trypsin and a Qiagen miRNeasy Mini kit was used to extract both the total RNA and microRNA. RNA samples were normalized to 1000ng/ml using absorbance spectroscopy. Samples were processed by the

Case Western Reserve Genomics core using an Affymetrix® Human Genome U219 microarray.

4.2.9 Data Processing

MarkerView (version 1.2.1.1) was used for HILIC-MS data alignment and peak picking.445 Features were limited by statistical significance using MetaboAnalyst 3.0

(http://www.metaboanalyst.ca),446 P<0.05 and -log2 FC≥2. Identified was completed with the use of the METLIN (https://metlin.scripps.edu)379,447,448 and HMDB

(http://www.hmdb.ca/)449 databases, by searching based on accurate mass and confirming

101 with fragmentation data. Meta-analysis was performed by feeding the aligned, statistically significant features into metaXCMS (https://xcmsonline.scripps.edu).533,534

Shotgun lipidomic data was processed using LipidView software (AB Sciex) with a mass tolerance of 0.05 Da, min % intensity = 0.1% and S/N≥3. Unpaired t-tests on all metabolomic/lipidomic grouped data sets were performed using GraphPad Prism version

5.00 for Windows, GraphPad Software, San Diego California USA (www.graphpad.com).

The Affymetrix GeneChip Command Console (AGCC) Software, and Transcriptome

Analysis Console (TAC) Software were used to analyze the microarray data.

4.3 Results and Discussion

4.3.1 Meta-analysis of RAP, CPZ, and C/R in vivo Models

The CPZ intoxication model is used extensively as a model for oligodendropathy; however, the chemical characterization of myelin loss is complicated by the presence of active remyelination driven by OPC that is present during the demyelinating lesion formation. We first sought to compare metabolites dysregulated in each of these models to learn more about the how the models are both metabolically unique as well as any interrelated pathways. C57BL/6 male mice were fed a diet containing 0.3% cuprizone

(CPZ model), fed CPZ while additionally receiving rapamycin injections (10 mg/kg) 5 days a week (C/R model), or mice just receiving rapamycin injections while consuming normal chow (RAP model). The mice were sacrificed after 6 weeks of the respective model and the CC, HP, and SC were removed.

102

Figure 4.2: Meta-analysis of C/R, RAP,and CPZ reveals model specific dysregulation. (a) Total dysregulated features (P<0.05, fold change >2) found in C/R (green), RAP (blue), and CPZ (purple) model detected in the hippocampus (HP) and corpus callosum (CC) (n=10 mice per condition). (b) Meta-analysis performed through XCMS. Significant features shared between C/R (green), RAP (blue), and CPZ (purple) from the HP and CC are numerically represented in the corresponding overlap. (c-k) Box and Whisker plots of significant features (P<0.05 1-way ANOVA, fold change >2) identified both with parent ion and fragment ion information. Each feature is collectively and significantly altered across the respective models when compared to the control.

103

The tissue was analyzed via our global metabolomics platform which combines the separate analysis of polar metabolites followed by shotgun lipidomics.

As discovered previously, metabolic dysregulation of the SC is minimal, so we focused on a meta-analysis comparing dysregulated features in the HP and CC. This analysis highlighted characteristics unique to each model, as well as features that are shared (Figure 4.2). In both tissues, the combined C/R model had the most metabolic differences when compared to the CPZ or RAP model alone, followed by the RAP then

CPZ model (Figure 4.2a). As previously characterized, the CPZ model displays more dysregulated features in the CC than the HP. RAP and C/R seemed to equally affect the

CC and HP, though CPZ causes dramatic histological changes in both the CC and HP.

Based on the number of dysregulated features, RAP was a significant driver of metabolic alterations when compared to CPZ alone.

The overlap of the CPZ and C/R model contains metabolites that are specific to the effects of CPZ, and correlate to the phenotypic changes CPZ induces within the CNS

(Figure 2.2b). Many of the metabolites uniquely disrupted in the CPZ and C/R model have been characterized in our previous work535 and include changes in metabolites associated with the P5P-bound enzymes. P5P is the active form of vitamin B6 and mainly participates in the transamination of amino acids. Specifically, metabolites associated with tryptophan metabolism including tryptophan, and 5-hydroxytrptophan were found disrupted in the CPZ and C/R model.

The RAP and C/R models contain the most overlapping dysregulated features and are associated with the direct effects of rapamycin on the CNS. This includes glutathione and homocysteine, which are associated with oxidative damage and were upregulated

104

(Figure 4.2c-f), the opposite FC as with CPZ intoxication alone (Figure 3.4c). Recent work has noted RAP to increase the expression of oxidative stress response genes in stem cells, ultimately producing more reduced glutathione.536 Between RAP and C/R BCAA were also downregulated, specifically leucine and valine (Figure 4.2g,h). Leucine is a known activator of the mTORC1 pathway537 and previous work genetically profiling human BJAB B-lymphoma cells and murine CTLL-2 T lymphocytes found overlaps between the rapamycin treatment group and those under leucine or general amino acid deprivation.538 Genes classified in this study (and supported in others) involve the upregulated oxidation of fatty acids (ACADVL,539 CRAT), catabolism of amino acids

(BCKADHA,540 GCDH, HMGCS2541), and the downregulated synthesis of fatty acids

(SREBF1,542 SCD,543 FABP5, FASN543), cholesterol (IDI1, SQLE), and proteins (TARS,544

EIF2S1544), and synthesis of NADPH (IDH1).538 BCAAs have been connected to the activation of mTORC1, and their downregulation noted in our samples may be a marker of RAP’s effect in the CNS.529 The first transaminase step of BCAA catabolism (Scheme

4.1) is catalyzed by either the mitochondrial (BCATm) or cystolic (BCATc) isoform of the PLP-dependent branched-chain amino acid aminotransferase (BCAT). BCAT converts the BCAAs into their deaminated derivatives α-ketoisocaproic acid (KIC), α- keto-β-methylvaleric acid (KMV), and α-ketoisovaleric acid (KIV) (Scheme 4.1).

The isoform BCATc is predominantly expressed in the CNS and is cell specific. In adult rat brains BCATc was found to be solely in neurons, with astrocytes uniformly expressing BCATm.545,546 In humans, BCATm was discovered in endothelial cells, with

BCATc similarly restricted to neurons.547 However, in an examination of primary cell cultures, BCATm was detected in both astrocytes and microglia, and BCATc in

105

Scheme 4.1: Catabolism of branched chain amino acids. Intermediate metabolites are shown in black, and enzymes are shown in blue. Metabolite abbreviations are as follows: 2-ketoisocaproate (KIC), 2- ketoisovalerate (KIV), and α-keto-β-methyl-n-valerate (KMV).596 oligodendrocytes and neurons.548,549 Branched-chain α-keto acid dehydrogenase (BCKD) acts as the regulatory step in BCAA catabolism, eventually participating in the synthesis of fatty acids. BCKD is a mitochondrial enzyme complex made up of three different catalytic subunits performing oxidative decarboxylation on branched-chain α-ketoacids, producing isovaleryl-CoA from KIC, isobutyrylCoA from KIV, and alpha-methylbutyryl-

CoA from KMV. The first subunit (alpha-ketoacid dehydrogenase, E1) decarboxylates the α-ketoacid using the cofactor thiamine pyrophosphate, and subsequently oxidizes the substrate, transferring it to a swinging lipoyl moiety, and regenerating thiamine pyrophosphate. The second subunit (dihydrolipoyl transacylase, E2) generates acetyl-

CoA through the transfer of the acetyl group off of the lipoyl arm and onto CoA. The reduced lipoyl arm then swings to the third subunit (dihydrolipoamide dehydrogenase,

E3) where it is regenerated through the use of FAD and NAD+. Downstream BCAA catabolism utilizes diverging enzymes, first employing dehydrogenases to form 3- methylcrotonyl CoA, tigloyl-CoA and methylacryl-CoA and eventually forming

106 energetic products such as acetyl-CoA, and propionyl-CoA.550 BCAA metabolism is an important mechanism that has also been implicated in cell differentiation and the suppression of stem-like cancer cells.551–553 RAPs observed ability to halt OPC differentiation may be due to the intimate connection of BCAA catabolism and mTOR signaling.554

Alterations in amino acid metabolism among the three models also included other amino acids. Arginine, glutamine, and serine were all downregulated (Figure 4.2i-k), and have all been linked to signaling through mTORC1.555–558. Metabolites associated with tyrosine metabolism (Scheme 4.2) including homovanillic acid, 5‐hydroxyindoleacetic acid, and norepinephrine were also reduced in the CPZ, RAP, and C/R models when compared to the control (Figure 4.2l-n). These metabolites are involved in the formation of dopamine and serotonin (Scheme 4.2) and are characteristic biomarkers associated with monoamine oxidase (MAO) inhibition.559 Mice under a chronic (8-30 weeks) RAP treatment have exhibited augmented levels of the matching metabolites in addition to 5- hydroxytryptamine, 3,4-dihydroxyphenylacetic acid, and dopamine associated with alterations in MAO.514 The reasoning for alterations in this pathway from CPZ alone are less clear. Lowered activities of MAO and dopamine β-hydroxylase (DBH) have been noted as early as 3 weeks into CPZ exposure, along with decreased concentrations of norepinephrine.175,560 A deficiency in the P5P-enzyme aromatic amino acid decarboxylase (AADC) could also cause a disruption in the dopamine/serotonin pathway, producing similar metabolic effects.561–563 AADC deficiency can occur through direct inborn error or due to complications with P5P synthesis, via defections in the pyridoxamine 5'-phosphate oxidase (PNPO) gene.561,564 PNPO intracellularly converts

107

Scheme 4.2: Dopamine/serotonin metabolism. Intermediate metabolites are shown in black, and enzymes are shown in blue. Abbreviations are as follows: tryptophan hydroxylase (TPH), 5- hydroxytryptophan (5HT), amino acid decarboxylase (AADC), catechol-O-methyl-transferase (COMT), aldehyde dehydrogenase (ADH), monoamine oxidase (MAO), 3-methoxytyramine (3-MT), 3,4-dihydroxyphenylacetaldehyde (DHPA), 3-methoxy-4-hydroxyphenylacetaldehyde (MHPA), 3,4- didroxyphenylacetic acid (DOPAC), and homovanillic acid (HVA).597 pyridoxamine phosphate and pyridoxine phosphate from the CSF into the active cofactor

P5P. A PNPO deficiency can cause seizures, brain atrophy, and hypomyelination.565 We have hypothesized CPZ to bind P5P and disrupt its function as a coenzyme.535 Thus, CPZ may also act to decrease AADC activity, further altering the dopamine/serotonin pathway. Though RAP and CPZ may act through different mechanisms, their shared disruption of tyrosine metabolism may be connected to the collective oligodendrocyte dysfunction.

4.3.2 Lipidomic Analysis of RAP, and C/R in vivo Models

We have previously characterized the damaging effects of CPZ on the general synthesis of lipids localized in the CC and HP. Here we explore the global changes of glial lipid composition under the effects of RAP or C/R. After 6 weeks, lipid production was lowered in nearly all classes in the corpus callosum and the spinal cord, for both the

108

Figure 4.3: RAP and C/R effect lipid homeostasis differently in the corpus callosum (CC) versus the spinal cord (SC). Shotgun lipidomics was performed on CNS tissue from RAP, C/R and control mice (n=10, for each model). Lipid profiles are expressed as the fold change ratio of the extracted corpus callosum (black) or the spinal cord (grey), using the peak intensity fold change (FC) of RAP (a) or C/R (b) treated samples to vehicle treated samples. A ratio of one represents no change between the respective model and the control sample group

RAP and C/R model. The RAP administration did not affect the overall lipid composition in the corpus callosum or spinal cord to nearly the same extent as CPZ alone. The RAP model also did not preferentially affect the corpus callosum over the spinal cord.

Sphingomyelin and ceramide were considerably downregulated in the RAP model. Both of which are important for the formation of myelin during cell differentiation. In Yeast cells deficient in TORC2 ceramide species are impaired due to the inactivation of ceramide synthase,566 and Drosophila treated with rapamycin resulted in the clearance of

BODIPY-labeled ceramide in neurons.567 Additionally, the de novo synthesis of sphingolipid species was found decreased in rapamycin-treated diabetic rats, determined

109 by MALDI-TOF MS.568 Thus, it is reasonable to believe that rapamycin negatively impacts ceramide synthesis, arresting myelin production as OPC mature into oligodendrocytes.

Like the CPZ model, lipid classes the C/R model had changes in lipid composition in the corpus callosum, though sphingomyelin, ceramide, and PC were found reduced in both the corpus callosum and spinal cord. Neutral lipids like DAG,

MADAG, and TAG are distinctively reduced in the corpus callosum of C/R-treated mice.

Neutral lipid metabolism was also altered in the CPZ model, implying that these effects are driven by CPZ intoxication. Inactivation mTORC2 through a concerted effort in the

C/R model could also impair the PKCα signaling of phospholipase D (PLD), downregulating the production of sphingomyelin. PLD catalyzes the hydrolysis of phosphatidylcholine (PC) into phosphatidic acid (PA) and choline, and PKCα has been specifically implicated in its activation.569,570 The low levels of choline and neutral lipids in the CPZ model also implied some deficit connected to the de novo synthesis of PC via the Kennedy Pathway.571

4.3.3 Transcriptomic Analysis of OPC Differentiation

The MO3.13 cell line is an immortalized oligodendrocyte hybrid line expressing phenotypic characteristics of primary oligodendrocytes while retaining reactivity for

MBP, PLP, and galactocerebroside (GS). Calcium levels, mediated by sodium/calcium exchangers (NCX), have been shown to influence oligodendrocyte maturation through the activation of protein kinase C. Studies have shown that when MO3.13 cells are cultured in 100nM 4-b-phorbol-12-myristate-13-acetate (PMA) (a structurally mimic of

DAG) and serum starved, signal transduction of PKC occurs, releasing calcium, and

110 causing cell maturation, branching, and increased levels of MAG, GS, and CNPase.572–574

PKCα is a downstream target of mTORC2 that is calcium-activated and dependent on the binding of a phospholipid and DAG, which is activated in part by PMA. PKCα activity in glioma cells is known to create a positive feedback loop with mTOR via an Akt- independent EGFG signaling pathway.575 Additionally, PMA-stimulation of U-251 glioblastoma cells resulted in the stimulation of PKCη activating both Akt and mTOR.576

To gain more information surrounding oligodendrocyte differentiation and the role of mTOR, we performed a transcriptomic investigation of PMA treated MO3.13 cells versus control cells. In agreement with previous reports, 3 days of 100 nM PMA treatment with

1% FBS altered the cell structure, indicative of maturation (Figure 4.3 a,b).574 PMA- treated cells also expressed galactocerebroside, a lipid generated by myelinating oligodendrocytes (Figure 4.3 a,b). After PMA treatment, we extracted total and RNA and performed a comparative analysis of PMA treated MO3.13 cells versus vehicle controls by using a comprehensive expression microarray (U219 array). Of the 49372 total genes identified, 2177 genes were significantly upregulated, and 2084 genes were significantly down-regulated in the PMA versus control samples (Figure 4.3c). Gene pathways most affected by PMA induced differentiation involved mitotic arrest, adhesion, cytoskeletal rearrangement, mTOR signal transduction, PKC signaling, lipid regulation, and amino acid metabolism (Figure 4.3d, Table A1).

Activation of the PI3K/Akt/mTOR signaling pathway associated with the differentiation of the MO3.13 cells was made evident by the upregulation of genetic regulators immediately downstream of mTORC1 and mTORC2. This included genes responsible for lipid biosynthesis (LPIN1 and SGK1), the cell fate determining SRY box

111 containing (Sox) family members (SOX4, SOX6, SOX11) and cytoskeletal arrangement

(PRKCA). PMA treatment also directly affects genes related to the PKC signal transduction, including PRKCA which regulates PKCα. PKCα mediates the activation of

MAPK1/3(ERK1/2), RAP1GAP, EGFG, and VEGFA-induced cell proliferation. Genes associated with the PKCα secondary messenger, calcium, are also upregulated in PMA- treated cells. This includes CALM1, and SLC8A3, which encode calmodulin and NCX, respectively. NCX expression has shown to be reliant on MAPK ERK1/2 activity577 in the brain and when silenced or knocked out impairs oligodendrocyte differentiation.574

As oligodendrocytes mature, they become postmitotic, exiting the cell cycle. The

E2F/Rb pathway is an important regulator of the cellular transition into quiescence and has been associated with the differentiation of OPCs.578,579 PMA is also known to mediate the E2F/Rb pathway through the PKCα stimulated MAPK/ERK signaling cascade.580,581

The pathway begins with the transcription factor E2F1 complexing with DP1, binding

DNA, and results in the promotion of genes mediating G1/S transition and S-phase initiation. The Rb tumor suppressor protein (Rb) sequesters E2F1, until phosphorylated, which occurs through the activity of cyclin-dependent kinase (CDK) complexes. CDK inhibitors (CKDIs) negatively control Rb phosphorylation, halting cell G1/S transitions and DNA replication, and genes encoding CDKIs such as p21Cip1, p18 Ink, and p27Kip1, have been reported to impair OPC differentiation. Many gene targets associated with the E2F/Rb pathway and downstream cell cycle regulation were found downregulated in PMA-treated MO3.13 cells, including E2F1 which encodes for the E2F transcription factor. The CDKs CDK1, CDK2, CCNA2, CCNB2, and CCNE2 are downregulated and CDKIs CDKN1A and CDKN1B are upregulated. The gene MCM7

112

Figure 4.4: PMA induced MO3.13 oligodendrocyte alters transcription of genes associated with cell cycle, cytoskeletal organization, and lipid synthesis. Fluorescence microscopy of MO3.13 cells in (a) DMEM with 1% FBS or (b) DMEM with 1% FBS and 100 nM PMA. Cells are stained with anti-galactocerebroside antibody (green) and DAPI (blue). After 3 days of treatment differentiated cells are notable smaller with many more processes. (c) Volcano plot depicting all identified transcripts found in both PMA-treated MO3.13 cells and control cells (n=3 for each condition). Transcripts are limited by significance (P<0.05 t- test, fold change >2). Transcripts are either non-significant (grey), significant and upregulated in PMA- treated cells (blue), or significant and down regulated in PMA treated cells (red). (d) Gene pathway enrichment of all significant transcripts (n =3), analyzed using Reactome for interpretation. Each slice represents the number of transcripts dysregulated in the labeled pathway.

113 binds Rb controlling the S phase checkpoint, and inhibiting DNA replication.582,583

MCM7 is found downregulated in PMA-treated cells, indicating that Rb is bound with

E2F. A collection of downstream gene expression was found significantly inhibited by the E2F/Rb complex. These include genes important in cell cycle checkpoints (CHEK1,

RFC3, RFC4, RFC5, WEE1, ORC1, POLE, and CDC45), mitotic cell cycle and spindle organization (RRM2, DHFR, TYMS, TTK, STMN1, PRIM1, POLA1, POLE2, and PCNA), and DNA repair and replication (SMC2, BARD1, RRM1, TOP2A, CDT1, ANLN, HMGB1,

H2AFZ, POLD3, DCK, and KIF4A).

The activation of PI3K/Akt/mTOR signaling and lipid biosynthesis has been abundantly characterized during the differentiation of adipocytes and hepatocytes and much of the regulation has been hypothesized to act during oligodendrocyte differentiation.584 Lipid biosynthesis in these cells is coordinated mainly through the transcription factors: C/EBP, PPARγ, and SREBP-1. PPARγ activates genes managing the synthesis of fatty acids, triacylglycerols, cholesterol, and phospholipids. Inhibiting mTORC1/mTORC2 with rapamycin has been shown to block PPARγ activity in adipocytes by disrupting the C/EBP and PPARγ positive feedback loop, blocking cell differentiation and lipid accumulation.585 Recent work also indicates that mTORC1 activity causes the phosphorylation and sequestration of Lipin-1, an inhibitor of SREBP-

1. SREBP-1 regulates genes guiding the synthesis of cholesterol and fatty acid.586

PPARG, LPIN1, and SREBP-1 were found upregulated in the differentiated MO3.13 cells, though below the significance threshold. SGK-1, which is responsible for ceramide synthesis, has also been implicated as a target of mTORC2 and was upregulated in PMA- treated cells. Many genes downstream of these transcription factors that are involved in

114 the synthesis of myelin lipids, particularly sphingomyelin (SGMS1, SMPD1, DEGS1,

UGCG, ACER3) and cholesterol (HMGCR, MVK, MVD, LSS, DHCR7) were also found upregulated with PMA.

In addition to genes associated with cell organization, cell cycle, and lipid synthesis, we investigated the disruption of BCAA in relation to PMA-induced differentiation. mTOR integrates signals from different growth factors, insulin, and amino acids. Leucine is known to stimulate protein synthesis by activation of the mTORC1 signaling pathway leading to the phosphorylation of target proteins. Amino acid starvation also causes the deactivation of mTORC1.587 As PMA has been shown to indirectly activate mTOR and the subsequent proliferation of glioma, the differentiation of oligodendrocytes may act through a similar mTOR signaling pathway. Stimulation of

PKCs in oligodendrocytes via PMA caused an increase in the acetyl-CoA transporter transcript ACAA1 and a decrease in BCAT1 and other transcripts involved in BCAA catabolism including IVD and HADH. BCAAs have been shown to play a role in adipocyte,551 and hepatocyte differentiation,552,553 thus the presence of available BCAAs may induce mTOR activation.

4.3.4 BCAAs alters MO3.13 morphological state

As BCAA metabolism was implicated in oligodendrocyte differentiation, we chose to visualize the effects of BCAA supplementation on cell morphology. To gain an enhanced understanding of the collective functions of mTORC1 and mTORC2 on

MO3.13 differentiation we combined the effects of PMA and amino acid supplementation. When up to 10 mM valine is added to MO3.13 cells along with PMA, density and growth dissipates, morphology changes, and cells stain positive for O1

115

Figure 4.5: BCAA and PMA modulate oligodendrocyte differentiation in vitro and are sensitive to RAP. (a) The number of live MO3.13 cells were counted using Trypan blue after 3 days of treatment with 100 nM PMA (white), 100 nM PMA and 1mM valine (light grey), or 100 nM PMA and 10 mM valine (dark grey) to discern the effects of BCAA on further maturation. Immunofluorescence was also used to visualize the O1 (magenta) and DAPI (blue) staining of MO3.13 cells treated with (b) 100 nM PMA and (c) 100 nM PMA with 10 mM valine (60X). MO3.13 cells treated with (d) vehicle, (e) 100 nM PMA, (f) 100 nM PMA and 10 mM leucine, (g) 10 mM leucine, or (h) 100 nM PMA and 10 mM leucine, and 100 µM rapamycin were imaged using brightfield microscopy (200X)

(Figure 4.5a-c). The O1 monoclonal antibody serves as a marker of oligodendrocyte differentiation, reacting with lipids, like GS, that are produced by late progenitor and mature cells. MO3.13 morphology also changed with additions of 100 nM PMA or 10 mM leucine with 100 nM PMA (Figure 4.5d-f). Additions of 10 mM leucine alone caused some morphological effects on the cells, though different from the PMA-induced

PKCα cytoskeletal restructuring (Figure 4.5g). To confirm these changes acted through

116 mTOR, RAP was supplemented in excess prior to additions of PMA and leucine. With the addition of RAP MO3.13, cells remained unchanged and the morphology reflects that of the controls (Figure 4.5h). In summation, MO3.13 cells undergo diverse changes in growth and morphology, caused by both PMA and BCAA, which are sensitive to mTOR deactivation via RAP.

4.4 Conclusions

Our investigation of the C/R in vivo model of demyelination using metabolomics, and the in vitro model of PMA-induced oligodendrocyte differentiation using transcriptomics, revealed overlapping pathway information pertaining to the maturation and intoxication of oligodendrocytes. The combined effects of CPZ and RAP caused an increase in cellular metabolic effects in both the CC and HP connected to the dysregulation of lipids synthesis, amino acid metabolism, and the production of homovanillic acid. mTOR signaling via the activation of PKCα proved important for oligodendrocyte differentiation, revealing numerous transcripts involved in the process, including BCAT1 (encoding BCATc). Together, metabolomic, transcriptomic, and microscopy data connected BCAA to mTOR activity, suggesting that like adipocytes and hepatocytes, OPCs may rely on pools of BCAAs for differentiation through mTOR. To further explore these theories, work should be done with primary oligodendrocytes or animal models to confirm BCAA acid levels post RAP treatment using targeted metabolomics, validate transcript changes using RT-PCR, and reinforce BCAA driven differentiation with additional immunofluorescent markers.

117

CHAPTER V

SUMMARY

Despite its cellular mechanisms and etiology remaining unclear, the CPZ murine model has been used since its conception as a model of demyelination and remyelination of both white and grey matter tracts in the CNS. Feeding C57BL ⁄6 mice a diet supplemented with 0.2% CPZ for 6 weeks results in reproducible demyelination via the selective death of oligodendrocytes. Since its original report, the described model variation has been mainly utilized due to minimal clinical toxicity and well established histological effects. Oligodendrocyte apoptosis together with microglial activation represent two major pathological features of the CPZ model. The features resemble patterns characteristic of MS lesions, particularly patterns III and IV, which are described as primarily oligodendrocyte-mediated. Consequently, the CPZ model is widely used to test the practicality of therapeutics serving to regulate myelin, particularly for patients with MS. As the origins of MS are themselves poorly understood, the mechanistic nature through which CPZ induces demyelination can illuminate interactions specific to oligodendrocyte maintenance, maturation, and myelin formation. Additionally, the

118 validity of MS therapeutics remains tied to fully understanding the CPZ-induced pathology. Two, seemingly opposed hypotheses, have been suggested describing potential modes of action triggered by CPZ administration. The first hypothesis suggests

CPZ’s copper chelating ability results in a disturbance of internal copper homeostasis, producing toxicity, and the second hypothesis proposes CPZ to instead induce cell- specific metabolic interference and subsequent enzyme inhibition. Here we sought to uncover the pathological processes by which CPZ acts by investigating the complex chemical nature of CPZ and by understanding the metabolic demands, and pathway interactions important for oligodendrocyte development and differentiation.

The physio-chemical nature of CPZ, alone, had been weakly explored, largely due to its poor solution stability. Using a combination of analytical methods, we sought to examine and reproduce experiments, while also expanding on basic knowledge regarding the structure and chelating abilities of CPZ. The hydrolysis of CPZ to CPZ-R proved to occur spontaneously in solution and we collected the first reported MS/MS data for the intact compound. Additionally, we developed a method to produce 10 mM CPZ in 30% ethanol (the lowest ethanol content previously reported was 50%), which has merit for developing a nontoxic cell culture treatment. The structure of the CuCPZ complex remains elusive, though our work has expanded the current knowledge base, presenting evidence that CPZ-R acts as the ligand to bind copper. As hypothesized in the past,

CuCPZ binds copper at a 2:1 (CPZ:Cu) ratio and its solution stability is affected by the present copper concentration. Chelation studies were additionally performed, applying a novel method of competition between CPZ and synthesize copper protein active site mimics. CPZ was unable to chelate copper away from the center of the protein active site

119 mimics and, instead, evidence of a CPZ-mimic complex was documented. Our work further supports the latter hypothesis of metabolically driven intoxication by providing more evidence of CPZ’s inability to chelate tightly regulated pools of copper in a biological system.

The development of an oligodendrocyte cell model to explore metabolic alterations occurring during CPZ intoxication was key to understanding in vivo cell- specific impairment. Treatment of 1 mM CPZ for no less than 24 hours resulted in the death of MO3.13 human oligodendrocyte cells. Using the developed cell model, we identified metabolites with a global metabolomics platform that were uniquely dysregulated in response to CPZ treatment. Amino acids connected with energy generation and members of the B vitamin family were specifically affected. To complement the cell model, we analyzed samples from C57BL/6 mice after a 0.2% CPZ diet. Samples were collected either 2- or 6-weeks after the introduction of CPZ feed and from differing regions of the CNS. This method allowed us both temporal and spatial information regarding the in vivo effects of CPZ. Metabolic discrepancies were noted between white matter tracts (CC and HP), and the SC, with the number of dysregulated features mirroring documented lesion pattern pathology. The loss of lipids, characterized by our shotgun lipidomics method, also reproduced the prominent demyelination in the

CC and HP, with neutral lipids being especially affected. Metabolic changes increased in significance from the 2- to 6-week time point paralleling the apoptosis of oligodendrocyte cells. Like the cell model, the CC and HP presented altered amino acid metabolism potentiating reduced antioxidant molecules and the perturbation of energy-generating

120 pathways. Collectively, the in vitro and in vivo CPZ models provide evidence of metabolic disruption tied to the events of demyelination and oligodenrogliosis.

Following the collected metabolic evidence, enzymatic targets for CPZ demyelination were investigated. Our first hypothesized target, TDO, is a heme containing, rate-limiting enzyme catalyzing the degradation of tryptophan which funnels into the de novo synthesis of NAD+. Though a promising target, purified rhTDO was not inhibited in the presence of up to 100 µM CPZ. Instead, we turned to P5P-bound transaminase enzymes.

Numerous amino acids regulated by transaminases were found dysregulated by CPZ, as well as B vitamins associated with the shared enzymatic cofactor P5P. Evidence in previous literature described P5P interacting with other structurally similar hydrazides, and we have demonstrated with mass spectrometry and NMR that CPZ too can react with

P5P via a Schiff base link. This inhibitory mechanism of action may be at play biologically, as transaminase activity and expression are both altered by the presence of

CPZ in vitro.

To expand on the demyelination and remyelination potential of the CPZ model, a new model has been developed, combining RAP and CPZ over a 6-week period. RAP is hypothesized to stop OPC differentiation through the inhibition of mTOR, ceasing the background supply of maturing oligodendrocytes present during treatments of CPZ alone.

Using metabolomics and transcriptomics, we explored the interplay between CPZ and

RAP while gaining further knowledge pertaining to the process of mTOR-driven oligodendrocyte differentiation. Lipogenesis decreases with both CPZ and RAP, although

RAP in not spatially discriminatory. The CPZ, RAP, and C/R models also all altered tyrosine metabolism in the CNS, a possible consequence of MAO inhibition. Models

121 employing RAP, however, produced metabolic alterations involving BCAA, providing a target of oligodendrocyte maturation associated with mTOR. Additionally, a transcriptomic analysis of MO3.13 cells treated with differentiation-inducing PMA revealed numerous genes related to the maturation process including ones associated with the catabolism of BCAA. Expanding on the exploratory evidence, oligodendrocyte cells treated with a combination of PMA and the BCAA leucine, or leucine alone decreased their rate of growth and presented morphological differences indicative of maturation.

Though indefinite, this data provides a promising lead, as BCAAs may serve a principal function in oligodendrocyte differentiation and subsequent myelination.

Bringing together information regarding the chemical nature of CPZ, and the biological function of its target, the oligodendrocyte, has revealed tandem information important to further understanding both the processes of demyelination and remyelination and the modus operandi of the molecule itself. Yet, after highlighting the important discoveries gained from this research, it leads only to further questions and a desire for further studies. Regarding the chemical properties of CPZ, many of the proposed complexes (CuCPZ, PLP+CPZ, and both mimics R and B with CPZ) could be further confirmed by X-ray crystallography and the oxidation state of bound copper with electron paramagnetic resonance. The actual presence of CPZ in the CNS has also been debated, with some questioning CPZs ability to cross the BBB. In our studies CPZ in its intact form was not found in any analyzed tissue but was found in small quantities soon after its addition to cells. CPZ studied in vivo may take on a different form, hydrolyzing and interacting with enzymes, and thus may be imperceptible in its intact formula. To track the biodiversity of CPZ, an isotopically labeled form of the molecule can be introduced

122 and the byproducts traced using mass spectrometry. Many metabolite and enzyme targets for both CPZ intoxication and oligodendrocyte function and differentiation have been proposed in this document. Metabolite concentrations can be verified using the described

MRM-MS procedure, and enzyme expression using RT-PCR. Specifically, a fuller investigation of the concentration of BCAAs and their byproducts in response to RAP treatment in vitro and in vivo. Furthermore, an examination of P5P-bound protein inhibition can be performed using expressed enzymes of interest. Following positive inhibition, P5P and CPZ could also be co-crystalized and structurally examined via X-ray crystallography. These are but a few suggestive directions when taking into account all of the studies made, and questions raised, involving both the structural and biological function of CPZ and the oligodendrocyte, however, these serve as direct inquiries based on information gained in this study. Considerable progress was made in studying this model, and would not have been possible without the integration of so many techniques.

These results have resolved many (even unexpected) questions, but of course, leave many more unanswered.

“It isn't all over; everything has not been invented; the human adventure is just

beginning.”

- Gene Roddenberry

123

REFERENCES

(1) Berer, K.; Krishnamoorthy, G. FEBS Lett. 2014, 588 (22), 4207–4213.

(2) Multiple Sclerosis: National Institute of Neurological Disorders and Stroke

https://www.ninds.nih.gov/Disorders/All-Disorders/Multiple-Sclerosis-

Information-Page (accessed Feb 14, 2017).

(3) Compston, A.; Coles, A. Lancet 2008, 372 (9648), 1502–1517.

(4) Vos, T.; Barber, Ryan, Brad, Bell, et. all. Lancet 2015, 386 (9995), 743–800.

(5) Alonso, A.; Hernan, M. A. Neurology 2008, 71 (2), 129–135.

(6) O’Connor, P.; O’Connor, P.; Group, on behalf of T. C. M. S. W. Neurology 2002,

59 (Issue 6, Supplement 3), S1–S33.

(7) International Multiple Sclerosis Genetics Consortium (IMSGC). Lancet Neurol.

2008, 7 (7), 567–569.

(8) Marrie, R. A. Lancet Neurol. 2004, 3 (12), 709–718.

(9) Lublin, F. D.; Reingold, S. C.; Sclerosis*, N. M. S. S. (USA) A. C. on C. T. of N.

A. in M. Neurology 1996, 46 (4), 907–911.

(10) Lublin, F. D.; Reingold, S. C.; Cohen, J. A.; Cutter, G. R.; Sørensen, P. S.;

Thompson, A. J.; Wolinsky, J. S.; Balcer, L. J.; Banwell, B.; Barkhof, F.; Bebo,

B.; Calabresi, P. A.; Clanet, M.; Comi, G.; Fox, R. J.; Freedman, M. S.; Goodman,

A. D.; Inglese, M.; Kappos, L.; Kieseier, B. C.; Lincoln, J. A.; Lubetzki, C.;

124

Miller, A. E.; Montalban, X.; O’Connor, P. W.; Petkau, J.; Pozzilli, C.; Rudick, R.

A.; Sormani, M. P.; Stüve, O.; Waubant, E.; Polman, C. H. Neurology 2014, 83

(3), 278–286.

(11) Tataru, N.; Vidal, C.; Decavel, P.; Berger, E.; Rumbach, L. Neuroepidemiology

2006, 27 (1), 28–32.

(12) Heesen, C.; Mohr, D.; Huitinga, I.; Bergh, F. T.; Gaab, J.; Otte, C.; Gold, S. Mult.

Scler. J. 2007, 13 (2), 143–148.

(13) Milo, R.; Kahana, E. Autoimmun. Rev. 2010, 9 (5), A387–A394.

(14) Nielsen, N. M.; Westergaard, T.; Rostgaard, K.; Frisch, M.; Hjalgrim, H.;

Wohlfahrt, J.; Koch-Henriksen, N.; Melbye, M. Am. J. Epidemiol. 2005, 162 (8),

774–778.

(15) Lindsey, J. W. Am. J. Med. Genet. Part A 2005, 135A (1), 53–58.

(16) Robertson, N. P.; Clayton, D.; Fraser, M.; Deans, J.; Compston, D. A. Neurology

1996, 47 (2), 347–352.

(17) Dyment, D. A.; Ebers, G. C.; Dessa Sadovnick, A. Lancet Neurol. 2004, 3 (2),

104–110.

(18) Baranzini, S. E. Curr. Opin. Genet. Dev. 2011, 21 (3), 317–324.

(19) Haines, J. L.; Terwedow, H. A.; Burgess, K.; Pericak-Vance, M. A.; Rimmler, J.

B.; Martin, E. R.; Oksenberg, J. R.; Lincoln, R.; Zhang, D. Y.; Banatao, D. R.;

Gatto, N.; Goodkin, D. E.; Hauser, S. L. Hum. Mol. Genet. 1998, 7 (8), 1229–

1234.

(20) Olerup, O.; Hillert, J. Tissue Antigens 1991, 38 (2), 1–15.

(21) Sotgiu, S.; Pugliatti, M.; Sanna, A.; Sotgiu, A.; Castiglia, P.; Solinas, G.; Dolei,

125

A.; Serra, C.; Bonetti, B.; Rosati, G. Eur. J. Neurol. 2002, 9 (4), 329–341.

(22) CA Jr, J.; P, T.; M, W.; et al. In Immunobiology: The Immune System in Health

and Disease; New York: Garland Science, 2001.

(23) Zuvich, R. L.; McCauley, J. L.; Oksenberg, J. R.; Sawcer, S. J.; De Jager, P. L.;

International Multiple Sclerosis Genetics Consortium; Aubin, C.; Cross, A. H.;

Piccio, L.; Aggarwal, N. T.; Evans, D.; Hafler, D. A.; Compston, A.; Hauser, S. L.;

Pericak-Vance, M. A.; Haines, J. L. Hum. Genet. 2010, 127 (5), 525–535.

(24) Kreft, K. L.; Verbraak, E.; Wierenga-Wolf, A. F.; van Meurs, M.; Oostra, B. A.;

Laman, J. D.; Hintzen, R. Q. Genes Immun. 2012, 13 (7), 587–592.

(25) Lee, L.-F.; Axtell, R.; Tu, G. H.; Logronio, K.; Dilley, J.; Yu, J.; Rickert, M.; Han,

B.; Evering, W.; Walker, M. G.; Shi, J.; de Jong, B. A.; Killestein, J.; Polman, C.

H.; Steinman, L.; Lin, J. C. Sci. Transl. Med. 2011, 3 (93), 93ra68.

(26) Lundmark, F.; Duvefelt, K.; Iacobaeus, E.; Kockum, I.; Wallström, E.; Khademi,

M.; Oturai, A.; Ryder, L. P.; Saarela, J.; Harbo, H. F.; Celius, E. G.; Salter, H.;

Olsson, T.; Hillert, J. Nat. Genet. 2007, 39 (9), 1108–1113.

(27) O’Doherty, C.; Alloza, I.; Rooney, M.; Vandenbroeck, K. Tissue Antigens 2009,

74 (5), 429–431.

(28) Baranzini, S. E.; Wang, J.; Gibson, R. A.; Galwey, N.; Naegelin, Y.; Barkhof, F.;

Radue, E.-W.; Lindberg, R. L. P.; Uitdehaag, B. M. G.; Johnson, M. R.;

Angelakopoulou, A.; Hall, L.; Richardson, J. C.; Prinjha, R. K.; Gass, A.; Geurts,

J. J. G.; Kragt, J.; Sombekke, M.; Vrenken, H.; Qualley, P.; Lincoln, R. R.;

Gomez, R.; Caillier, S. J.; George, M. F.; Mousavi, H.; Guerrero, R.; Okuda, D. T.;

Cree, B. A. C.; Green, A. J.; Waubant, E.; Goodin, D. S.; Pelletier, D.; Matthews,

126

P. M.; Hauser, S. L.; Kappos, L.; Polman, C. H.; Oksenberg, J. R. Hum. Mol.

Genet. 2009, 18 (4), 767–778.

(29) Seboun, E.; Robinson, M. A.; Doolittle, T. H.; Ciulla, T. A.; Kindt, T. J.; Hauser,

S. L. Cell 1989, 57 (7), 1095–1100.

(30) Ligers, A.; Xu, C.; Saarinen, S.; Hillert, J.; Olle Olerup. J. Neuroimmunol. 1999,

97 (1), 182–190.

(31) De Jager, P. L.; Jia, X.; Wang, J.; de Bakker, P. I. W.; Ottoboni, L.; Aggarwal, N.

T.; Piccio, L.; Raychaudhuri, S.; Tran, D.; Aubin, C.; Briskin, R.; Romano, S.;

Baranzini, S. E.; McCauley, J. L.; Pericak-Vance, M. A.; Haines, J. L.; Gibson, R.

A.; Naeglin, Y.; Uitdehaag, B.; Matthews, P. M.; Kappos, L.; Polman, C.;

McArdle, W. L.; Strachan, D. P.; Evans, D.; Cross, A. H.; Daly, M. J.; Compston,

A.; Sawcer, S. J.; Weiner, H. L.; Hauser, S. L.; Hafler, D. A.; Oksenberg, J. R.;

Oksenberg, J. R. Nat. Genet. 2009, 41 (7), 776–782.

(32) Ascherio, A.; Munch, M. Epidemiology 2000, 11 (2), 220–224.

(33) Haahr, S.; Munch, M. J. Neurovirol. 2000, 6 Suppl 2, S76-9.

(34) Casetta, I.; Granieri, E. J. Neurovirol. 2000, 6 Suppl 2, S147-51.

(35) Levin, L. I.; Munger, K. L.; Rubertone, M. V.; Peck, C. A.; Lennette, E. T.;

Spiegelman, D.; Ascherio, A. JAMA 2003, 289 (12), 1533.

(36) Granieri, E.; Casetta, I. Neurology 1997, 49 (2 Suppl 2), S42-54.

(37) Gilden, D. H. Lancet Neurol. 2005, 4 (3), 195–202.

(38) Correale, J.; de los Milagros Bassani Molinas, M. J. Neurol. 2002, 249 (4), 375–

389.

(39) Okada, H.; Kuhn, C.; Feillet, H.; Bach, J.-F. Clin. Exp. Immunol. 2010, 160 (1), 1–

127

(40) Fazakerley, J. K.; Walker, R. J. Neurovirol. 2003, 9 (2), 148–164.

(41) Stohlman, S. A.; Hinton, D. R. Brain Pathol. 2006, 11 (1), 92–106.

(42) Olson, J. K.; Croxford, J. L.; Calenoff, M. A.; Dal Canto, M. C.; Miller, S. D. J.

Clin. Invest. 2001, 108 (2), 311–318.

(43) Tsunoda, I.; Fujinami, R. S. Springer Semin. Immunopathol. 2002, 24 (2), 105–

125.

(44) Oldstone, M. B. Cell 1987, 50 (6), 819–820.

(45) Kurtzke, J. F. Acta Neurol. Scand. 1975, 51 (2), 110–136.

(46) Freedman, D. M.; Dosemeci, M.; Alavanja, M. C. Occup. Environ. Med. 2000, 57

(6), 418–421.

(47) Alharbi, F. M. Neurosciences (Riyadh). 2015, 20 (4), 329–335.

(48) Ascherio, A.; Munger, K. L.; White, R.; Köchert, K.; Simon, K. C.; Polman, C. H.;

Freedman, M. S.; Hartung, H.-P.; Miller, D. H.; Montalbán, X.; Edan, G.; Barkhof,

F.; Pleimes, D.; Radü, E.-W.; Sandbrink, R.; Kappos, L.; Pohl, C. JAMA Neurol.

2014, 71 (3), 306.

(49) Dutta, R.; Trapp, B. D. Rev. Prat. 2006, 56 (12), 1293–1298.

(50) Goodkin, D. E.; Rooney, W. D.; Sloan, R.; Bacchetti, P.; Gee, L.; Vermathen, M.;

Waubant, E.; Abundo, M.; Majumdar, S.; Nelson, S.; Weiner, M. W. Neurology

1998, 51 (6), 1689–1697.

(51) Tallantyre, E. C.; Bo, L.; Al-Rawashdeh, O.; Owens, T.; Polman, C. H.; Lowe, J.

S.; Evangelou, N. Mult. Scler. 2010, 16 (4), 406–411.

(52) Ontaneda, D.; Sakaie, K.; Lin, J.; Wang, X.; Lowe, M. J.; Phillips, M. D.; Fox, R.

128

J. J. Neuroimaging 2014, 24 (6), 569–576.

(53) Cudrici, C.; Niculescu, T.; Niculescu, F.; Shin, M. L.; Rus, H. J Rehabil Res Dev

2006, 43 (1), 123–132.

(54) Purves, D.; Williams, S. M. (Stephen M. Sinauer Associates, 2001.

(55) Clanet, M. Int. MS J. 2008, 15 (2), 59–61.

(56) Brosnan, C. F.; Raine, C. S. Glia 2013, 61 (4), 453–465.

(57) Minagar, A.; Alexander, J. S. Mult. Scler. 2003, 9, 540–549.

(58) Ballabh, P.; Braun, A.; Nedergaard, M. Neurobiol. Dis. 2004, 16 (1), 1–13.

(59) Bradbury, M. W. Circ. Res. 1985, 57 (2).

(60) Correale, J.; Villa, A. Autoimmunity 2007, 40 (2), 148–160.

(61) Bunge, R. P. Physiol. Rev. 1968, 48 (1), 197–251.

(62) Morell, P.; Quarles, R. H. In Basic Neurochemistry: Molecular, Cellular and

Medical Aspects; Lippincott-Raven: Philadelphia, 1999.

(63) Saher, G.; Brügger, B.; Lappe-Siefke, C.; Möbius, W.; Tozawa, R.; Wehr, M. C.;

Wieland, F.; Ishibashi, S.; Nave, K.-A. Nat. Neurosci. 2005, 8 (4), 468.

(64) Frischer, J. M.; Weigand, S. D.; Guo, Y.; Kale, N.; Parisi, J. E.; Pirko, I.;

Mandrekar, J.; Bramow, S.; Metz, I.; Br?ck, W.; Lassmann, H.; Lucchinetti, C. F.

Ann. Neurol. 2015, 78 (5), 710–721.

(65) Lucchinetti, C. F.; Brück, W.; Rodriguez, M.; Lassmann, H. Brain Pathol. 1996, 6

(3), 259–274.

(66) Lucchinetti, C.; Brück, W.; Parisi, J.; Scheithauer, B.; Rodriguez, M.; Lassmann,

H. Ann. Neurol. 2000, 47 (6), 707–717.

(67) Lucchinetti, C.; Brück, W.; Parisi, J.; Scheithauer, B.; Rodriguez, M.; Lassmann,

129

H. Brain 1999, 122 (12), 2279–2295.

(68) Spadaro, M.; Gerdes, L. A.; Mayer, M. C.; Ertl-Wagner, B.; Laurent, S.;

Krumbholz, M.; Breithaupt, C.; Högen, T.; Straube, A.; Giese, A.; Hohlfeld, R.;

Lassmann, H.; Meinl, E.; Kümpfel, T. Ann. Clin. Transl. Neurol. 2015, 2 (3), 295–

301.

(69) Lindert, R. B.; Haase, C. G.; Brehm, U.; Linington, C.; Wekerle, H.; Hohlfeld, R.

Brain 1999, 2089–2100.

(70) Reindl, M.; Linington, C.; Brehm, U.; Egg, R.; Dilitz, E.; Deisenhammer, F.;

Poewe, W.; Berger, T. Brain 1999, 2047–2056.

(71) Barnett, M. H.; Prineas, J. W. Ann. Neurol. 2004, 55 (4), 458–468.

(72) Brück, W.; Popescu, B.; Lucchinetti, C. F.; Markovic-Plese, S.; Gold, R.; Thal, D.

R.; Metz, I. Ann. Neurol. 2012, 72 (3), 385–394.

(73) Mahad, D.; Ziabreva, I.; Lassmann, H.; Turnbull, D. Brain 2008, 131 (7), 1722–

1735.

(74) Quintana, F. J.; Farez, M. F.; Viglietta, V.; Iglesias, A. H.; Merbl, Y.; Izquierdo,

G.; Lucas, M.; Basso, A. S.; Khoury, S. J.; Lucchinetti, C. F.; Cohen, I. R.;

Weiner, H. L. Proc. Natl. Acad. Sci. U. S. A. 2008, 105 (48), 18889–18894.

(75) Patrikios, P.; Stadelmann, C.; Kutzelnigg, A.; Rauschka, H.; Schmidbauer, M.;

Laursen, H.; Sorensen, P. S.; Bruck, W.; Lucchinetti, C.; Lassmann, H. Brain

2006, 129 (12), 3165–3172.

(76) Smith, K. J.; Blakemore, W. F.; McDonald, W. I. Nature 1979, 280 (5721), 395–

396.

(77) Zambonin, J. L.; Zhao, C.; Ohno, N.; Campbell, G. R.; Engeham, S.; Ziabreva, I.;

130

Schwarz, N.; Lee, S. E.; Frischer, J. M.; Turnbull, D. M.; Trapp, B. D.; Lassmann,

H.; Franklin, R. J. M.; Mahad, D. J. Brain 2011, 134 (7), 1901–1913.

(78) Jeffery, N. D.; Blakemore, W. F. Brain 1997, 27–37.

(79) Haines, J. D.; Inglese, M.; Casaccia, P. Mt. Sinai J. Med. 2011, 78 (2), 231–243.

(80) Evangelou, N.; Konz, D.; Esiri, M. M.; Smith, S.; Palace, J.; Matthews, P. M.

Brain 2000, 1845–1849.

(81) Kornek, B.; Storch, M. K.; Weissert, R.; Wallstroem, E.; Stefferl, A.; Olsson, T.;

Linington, C.; Schmidbauer, M.; Lassmann, H. Am. J. Pathol. 2000, 157 (1), 267–

276.

(82) Irvine, K. A.; Blakemore, W. F. Brain 2008, 131 (6), 1464–1477.

(83) Keirstead, H. S.; Blakemore, W. F. J. Neuropathol. Exp. Neurol. 1997, 56 (11),

1191–1201.

(84) Richardson, W. D.; Young, K. M.; Tripathi, R. B.; McKenzie, I. Neuron 2011, 70

(4), 661–673.

(85) Gensert, J. M.; Goldman, J. E. Neuron 1997, 19 (1), 197–203.

(86) Zhang, S. C.; Ge, B.; Duncan, I. D. Proc. Natl. Acad. Sci. U. S. A. 1999, 96 (7),

4089–4094.

(87) Zawadzka, M.; Rivers, L. E.; Fancy, S. P. J.; Zhao, C.; Tripathi, R.; Jamen, F.;

Young, K.; Goncharevich, A.; Pohl, H.; Rizzi, M.; Rowitch, D. H.; Kessaris, N.;

Suter, U.; Richardson, W. D.; Franklin, R. J. M. Cell Stem Cell 2010, 6 (6), 578–

590.

(88) Franklin, R. J. M.; Gilson, J. M.; Blakemore, W. F. J. Neurosci. Res. 1997, 50 (2),

337–344.

131

(89) Menn, B.; Garcia-Verdugo, J. M.; Yaschine, C.; Gonzalez-Perez, O.; Rowitch, D.;

Alvarez-Buylla, A. J. Neurosci. 2006, 26 (30).

(90) Nait-Oumesmar, B.; Decker, L.; Lachapelle, F.; Avellana-Adalid, V.; Bachelin, C.;

Baron-Van Evercooren, A. Eur. J. Neurosci. 1999, 11 (12), 4357–4366.

(91) Glezer, I.; Lapointe, A.; Rivest, S. FASEB J. 2006, 20 (6), 750–752.

(92) Rhodes, K. E.; Raivich, G.; Fawcett, J. W. Neuroscience 2006, 140 (1), 87–100.

(93) Pfeiffer, S. E.; Warrington, A. E.; Bansal, R. Trends Cell Biol. 1993, 3 (6), 191–

197.

(94) Warrington, A. E.; Barbarese, E.; Pfeiffer, S. E. J. Neurosci. Res. 1993, 34 (1), 1–

13.

(95) Miller, R. H. Prog. Neurobiol. 2002, 67 (6), 451–467.

(96) Lindner, M.; Heine, S.; Haastert, K.; Garde, N.; Fokuhl, J.; Linsmeier, F.; Grothe,

C.; Baumgärtner, W.; Stangel, M. Neuropathol. Appl. Neurobiol. 2007, 0 (0),

071026235340007–???

(97) Patel, J. R.; Klein, R. S. FEBS Lett. 2011, 585 (23), 3730–3737.

(98) Aguirre, A.; Dupree, J. L.; Mangin, J. M.; Gallo, V. Nat. Neurosci. 2007, 10 (8),

990–1002.

(99) Mason, J. L.; Xuan, S.; Dragatsis, I.; Efstratiadis, A.; Goldman, J. E. J. Neurosci.

2003, 23 (20), 7710–7718.

(100) Frost, E. E.; Nielsen, J. A.; Le, T. Q.; Armstrong, R. C. J. Neurobiol. 2003, 54 (3),

457–472.

(101) Bhat, N. R. Dev. Neurosci. 1995, 17 (5–6), 267–284.

(102) Baron, W.; Metz, B.; Bansal, R.; Hoekstra, D.; de Vries, H. Mol. Cell. Neurosci.

132

2000, 15 (3), 314–329.

(103) Flores, A. I.; Mallon, B. S.; Matsui, T.; Ogawa, W.; Rosenzweig, A.; Okamoto, T.;

Macklin, W. B. J. Neurosci. 2000, 20 (20), 7622–7630.

(104) Ebner, S.; Dunbar, M.; McKinnon, R. D. J. Neurosci. Res. 2000, 62 (3), 336–345.

(105) Yim, S. H.; Hammer, J. A.; Quarles, R. H. J. Neurochem. 2001, 76 (6), 1925–

1934.

(106) Bansal, R.; Magge, S.; Winkler, S. J. Neurosci. Res. 2003, 74 (4), 486–493.

(107) Cui, Q.-L.; Almazan, G. J. Neurochem. 2007, 0 (0), 070210024758017–???

(108) Frederick, T. J.; Min, J.; Altieri, S. C.; Mitchell, N. E.; Wood, T. L. Glia 2007, 55

(10), 1011–1022.

(109) Frost, E. E.; Zhou, Z.; Krasnesky, K.; Armstrong, R. C. Neurochem. Res. 2009, 34

(1), 169–181.

(110) Van’t Veer, A.; Du, Y.; Fischer, T. Z.; Boetig, D. R.; Wood, M. R.; Dreyfus, C. F.

J. Neurosci. Res. 2009, 87 (1), 69–78.

(111) Guardiola-Diaz, H. M.; Ishii, A.; Bansal, R. Glia 2012, 60 (3), 476–486.

(112) Arnett, H. A.; Fancy, S. P. J.; Alberta, J. A.; Zhao, C.; Plant, S. R.; Kaing, S.;

Raine, C. S.; Rowitch, D. H.; Franklin, R. J. M.; Stiles, C. D. Science (80-. ). 2004,

306 (5704), 2111–2115.

(113) Lu, Q. R.; Sun, T.; Zhu, Z.; Ma, N.; Garcia, M.; Stiles, C. D.; Rowitch, D. H.;

Richardson, W. D.; Zalc, B.; Qiu, M.; Black, P. M. Cell 2002, 109 (1), 75–86.

(114) Li, H.; Lu, Y.; Smith, H. K.; Richardson, W. D. J. Neurosci. 2007, 27 (52), 14375–

14382.

(115) Arnett, H. A.; Mason, J.; Marino, M.; Suzuki, K.; Matsushima, G. K.; Ting, J. P.-

133

Y. Nat. Neurosci. 2001, 4 (11), 1116–1122.

(116) Kircik, L. H.; Del Rosso, J. Q. J. Drugs Dermatol. 2009, 8 (6), 546–559.

(117) Barnhart, B. C.; Peter, M. E. Cell 2003, 114 (2), 148–150.

(118) Calderon, T. M.; Eugenin, E. A.; Lopez, L.; Kumar, S. S.; Hesselgesser, J.; Raine,

C. S.; Berman, J. W. J. Neuroimmunol. 2006, 177 (1–2), 27–39.

(119) Omari, K. M.; John, G. R.; Sealfon, S. C.; Raine, C. S. Brain 2005, 128 (5), 1003–

1015.

(120) Omari, K. M.; John, G.; Lango, R.; Raine, C. S. Glia 2006, 53 (1), 24–31.

(121) Nagasawa, T. Int. J. Hematol. 2000, 72 (4), 408–411.

(122) Carbajal, K. S.; Schaumburg, C.; Strieter, R.; Kane, J.; Lane, T. E. Proc. Natl.

Acad. Sci. 2010, 107 (24), 11068–11073.

(123) Patel, J. R.; McCandless, E. E.; Dorsey, D.; Klein, R. S. Proc. Natl. Acad. Sci.

2010, 107 (24), 11062–11067.

(124) Omari, K. M.; Lutz, S. E.; Santambrogio, L.; Lira, S. A.; Raine, C. S. Am. J.

Pathol. 2009, 174 (1), 164–176.

(125) Liu, L.; Belkadi, A.; Darnall, L.; Hu, T.; Drescher, C.; Cotleur, A. C.; Padovani-

Claudio, D.; He, T.; Choi, K.; Lane, T. E.; Miller, R. H.; Ransohoff, R. M. Nat.

Neurosci. 2010, 13 (3), 319–326.

(126) Liu, L.; Darnall, L.; Hu, T.; Choi, K.; Lane, T. E.; Ransohoff, R. M. J. Neurosci.

2010, 30 (27), 9074–9083.

(127) Urbanski, M. M.; Kingsbury, L.; Moussouros, D.; Kassim, I.; Mehjabeen, S.;

Paknejad, N.; Melendez-Vasquez, C. V. Sci. Rep. 2016, 6, 33751.

(128) Wang, H.; Tewari, A.; Einheber, S.; Salzer, J. L.; Melendez-Vasquez, C. V. J. Cell

134

Biol. 2008, 182 (6), 1171–1184.

(129) Noll, E.; Miller, R. H. Development 1994, 120 (3), 649–660.

(130) Snethen, H.; Love, S.; Scolding, N. Regen. Med. 2008, 3 (6), 835–847.

(131) Procaccini, C.; De Rosa, V.; Pucino, V.; Formisano, L.; Matarese, G. Eur. J.

Pharmacol. 2015, 759, 182–191.

(132) Ransohoff, R. M. Nat. Neurosci. 2012, 15 (8), 1074–1077.

(133) Denic, A.; Johnson, A. J.; Bieber, A. J.; Warrington, A. E.; Rodriguez, M.; Pirko,

I. Pathophysiol. Off. J. Int. Soc. Pathophysiol. 2011, 18 (1), 21–29.

(134) Baker, D.; Amor, S. Curr. Pharm. Des. 2015, 21 (18), 2440–2452.

(135) Denic, A.; Johnson, A. J.; Bieber, A. J.; Warrington, A. E.; Rodriguez, M.; Pirko,

I. Pathophysiology 2011, 18 (1), 21–29.

(136) MD Biosciences. EAE Preclinical Model Associated With FDA Approved MS

Drugs http://www.mdbiosciences.com/blog/bid/39317/the-eae-preclinical-model-

is-associated-with-fda-approved-ms-drugs (accessed Mar 30, 2017).

(137) Li, Z.; He, Y.; Fan, S.; Sun, B. Neurosci. Bull. 2015, 31 (5), 617–625.

(138) Najm, F. J.; Madhavan, M.; Zaremba, A.; Shick, E.; Karl, R. T.; Factor, D. C.;

Miller, T. E.; Nevin, Z. S.; Kantor, C.; Sargent, A.; Quick, K. L.; Schlatzer, D. M.;

Tang, H.; Papoian, R.; Brimacombe, K. R.; Shen, M.; Boxer, M. B.; Jadhav, A.;

Robinson, A. P.; Podojil, J. R.; Miller, S. D.; Miller, R. H.; Tesar, P. J. Nature

2015, 522 (7555), 216–220.

(139) Theiler, M. J. Exp. Med. 1937, 65 (5), 705–719.

(140) Oleszak, E. L.; Chang, J. R.; Friedman, H.; Katsetos, C. D.; Platsoucas, C. D. Clin.

Microbiol. Rev. 2004, 17 (1), 174–207.

135

(141) Stavrou, S.; Feng, Z.; Lemon, S. M.; Roos, R. P. J. Virol. 2010, 84 (18), 9181–

9189.

(142) Zoecklein, L. J.; Pavelko, K. D.; Gamez, J.; Papke, L.; McGavern, D. B.; Ure, D.

R.; Njenga, M. K.; Johnson, A. J.; Nakane, S.; Rodriguez, M. Brain Pathol. 2003,

13 (3), 291–308.

(143) Tsunoda, I.; Iwasaki, Y.; Terunuma, H.; Sako, K.; Ohara, Y. Acta Neuropathol.

1996, 91 (6), 595–602.

(144) Yamada, M.; Zurbriggen, A.; Fujinami, R. S. 1991; pp 291–320.

(145) Tsunoda, I.; Fujinami, R. S. J. Neuropathol. Exp. Neurol. 1996, 55 (6), 673–686.

(146) Tsunoda, I.; Wada, Y.; Libbey, J. E.; Cannon, T. S.; Whitby, F. G.; Fujinami, R. S.

J. Virol. 2001, 75 (16), 7494–7505.

(147) Dal Canto, M. C.; Lipton, H. L. Am. J. Pathol. 1982, 106 (1), 20–29.

(148) Dal Canto, M. C.; Lipton, H. L. Lab. Invest. 1975, 33 (6), 626–637.

(149) Rodriguez, M.; Lucchinetti, C. F.; Clark, R. J.; Yakash, T. L.; Markowitz, H.;

Lennon, V. A. J. Immunol. 1988, 140 (3), 800–806.

(150) Oleszak, E. L.; Kuzmak, J.; Good, R. A.; Platsoucas, C. D. Immunol. Res. 1995, 14

(1), 13–33.

(151) Begolka, W. S.; Vanderlugt, C. L.; Rahbe, S. M.; Miller, S. D. J. Immunol. 1998,

161 (8), 4437–4446.

(152) Murray, P. D.; Pavelko, K. D.; Leibowitz, J.; Lin, X.; Rodriguez, M. J. Virol.

1998, 72 (9), 7320–7329.

(153) Zhang, J.; Lipton, H. L.; Perelson, A. S.; Dahari, H. J. Virol. 2013, 87 (7), 4052–

4059.

136

(154) Dal Canto MC; Kim; Miller; Melvold. Methods 1996, 10 (3), 453–461.

(155) OLITSKY, P. K.; YAGER, R. H. J. Exp. Med. 1949, 90 (3), 213–224.

(156) Lipton, M. M.; Freund, J. Exp. Biol. Med. 1952, 81 (1), 260–261.

(157) Freund, J. J. Immunol. 1947, 57 (2).

(158) Rivers, T. M.; Sprunt, D. H.; Berry, G. P. J. Exp. Med. 1933, 58 (1).

(159) Rivers, T. M.; Schwentker, F. F. J. Exp. Med. 1935, 61 (5).

(160) Tuohy, V. K.; Lu, Z.; Sobel, R. A.; Laursen, R. A.; Lees, M. B. J. Immunol. 1989,

142 (5), 1523–1527.

(161) Tompkins, S. M.; Padilla, J.; Dal Canto, M. C.; Ting, J. P.-Y.; Van Kaer, L.;

Miller, S. D. J. Immunol. 2002, 168 (8), 4173–4183.

(162) Stromnes, I. M.; Goverman, J. M. Nat. Protoc. 2006, 1 (4), 1952–1960.

(163) Miller, S. D.; Karpus, W. J.; Davidson, T. S. Curr. Protoc. Immunol. 2007,

Chapter 15, Unit 15.1.

(164) Mannie, M.; Swanborg, R. H.; Stepaniak, J. A. In Current Protocols in

Immunology; John Wiley & Sons, Inc.: Hoboken, NJ, USA, 2009; Vol. Chapter

15, p Unit 15.2.

(165) Munoz, J. J.; Bernard, C. C. A.; Mackay, I. R. Cell. Immunol. 1984, 83, 92–100.

(166) Constantinescu, C. S.; Farooqi, N.; O’Brien, K.; Gran, B. Br. J. Pharmacol. 2011,

164 (4), 1079–1106.

(167) Hemmer, B.; Cepok, S.; Nessler, S.; Sommer, N. Curr. Opin. Neurol. 2002, 15 (3),

227–231.

(168) BATOULIS, H.; RECKS, M. S.; ADDICKS, K.; KUERTEN, S. APMIS 2011, 119

(12), 819–830.

137

(169) Tanuma, N.; Shin, T.; Matsumoto, Y. J. Neuroimmunol. 2000, 108 (1–2), 171–

180.

(170) Day, M. J. In Experimental Models of Multiple Sclerosis; Springer US: Boston,

MA, 2005; pp 25–43.

(171) Constantinescu, C. S.; Farooqi, N.; O’Brien, K.; Gran, B. Br. J. Pharmacol. 2011,

164 (4), 1079–1106.

(172) Sriram, S.; Steiner, I. Ann. Neurol. 2005, 58 (6), 939–945.

(173) Nilsson, G.; Erdtman, H.; Lindstedt, G.; Kinell, P.-O. Acta Chem. Scand. 1950, 4,

205–205.

(174) Carlton, W. Toxicol. Appl. Pharmacol. 1966, 8, 512–521.

(175) Kesterson, J. W.; Carlton, W. W. Toxicol. Appl. Pharmacol. 1971, 20 (3), 386–

395.

(176) Venturini, G. J. Neurochem. 1973, 21 (5), 1147–1151.

(177) Carlton, W. W. Life Sci 1967, 6 (1), 11–19.

(178) Blakemore, W. F. J. Neurol. Sci. 1973, 20 (1), 63–72.

(179) Adamo, A. M.; Paez, P. M.; Escobar Cabrera, O. E.; Wolfson, M.; Franco, P. G.;

Pasquini, J. M.; Soto, E. F. Exp. Neurol. 2006, 198 (2), 519–529.

(180) Basoglu, H.; Boylu, N. T.; Kose, H. Eur. Rev. Med. Pharmacol. Sci. 2013, 17 (20),

2711–2717.

(181) Elsworth, S.; Howell, J. M. Res. Vet. Sci. 1973, 14 (3), 385–387.

(182) Franco, P. G.; Silvestroff, L.; Soto, E. F.; Pasquini, J. M. Exp. Neurol. 2008, 212

(2), 458–467.

(183) KANNO, T.; SASAKI, S.; YAMADA, N.; KAWASAKO, K.; TSUCHITANI, M.

138

J. Vet. Med. Sci. 2012, 74 (7), 837–843.

(184) Ludwin, S. K. Lab. Invest. 1978, 39 (6), 597–612.

(185) Skripuletz, T.; Gudi, V.; Hackstette, D.; Stangel, M. Histol. Histopathol. 2011, 26

(12), 1585–1597.

(186) Wakabayashi, T.; Asano, M.; Ishikawa, K.; Kishimoto, H. J. Electron Microsc.

(Tokyo). 1977, 26 (2), 137–140.

(187) Carey, E. M.; Freeman, N. M. Neurochem. Res. 1983, 8 (8), 1029–1044.

(188) Hiremath, M. M.; Saito, Y.; Knapp, G. W.; Ting, J. P.; Suzuki, K.; Matsushima, G.

K. J. Neuroimmunol. 1998, 92 (1–2), 38–49.

(189) Kesterson, J. W.; Carlton, W. W. Exp. Mol. Pathol. 1970, 13 (3), 281–294.

(190) Stidworthy, M. F.; Genoud, S.; Suter, U.; Mantei, N.; Franklin, R. J. M. Brain

Pathol. 2003, 13 (3), 329–339.

(191) Matsushima, G. K.; Morell, P. Brain Pathol. 2001, 11 (1), 107–116.

(192) Pattison, I. H.; Jebbett, J. N. Res. Vet. Sci. 1971, 12 (4), 378–380.

(193) Petronilli, V.; Zoratti, M. J. Bioenerg. Biomembr. 1990, 22 (5), 663–677.

(194) Franco-Pons, N.; Torrente, M.; Colomina, M. T.; Vilella, E. Toxicol. Lett. 2007,

169 (3), 205–213.

(195) Kesterson, J. W.; Carlton, W. W. Toxicol. Appl. Pharmacol. 1983, 22 (1), 6–13.

(196) Kipp, M.; Clarner, T.; Dang, J.; Copray, S.; Beyer, C. Acta Neuropathol 2009, 118

(6), 723–736.

(197) Norkute, A.; Hieble, A.; Braun, A.; Johann, S.; Clarner, T.; Baumgartner, W.;

Beyer, C.; Kipp, M. J. Neurosci. Res. 2009, 87 (6), 1343–1355.

(198) Steelman, A. J.; Thompson, J. P.; Li, J. Neurosci. Res. 2012, 72 (1), 32–42.

139

(199) Praet, J.; Guglielmetti, C.; Berneman, Z.; Van der Linden, A.; Ponsaerts, P.

Neurosci. Biobehav. Rev. 2014, 47, 485–505.

(200) Tagge, I.; O’Connor, A.; Chaudhary, P.; Pollaro, J.; Berlow, Y.; Chalupsky, M.;

Bourdette, D.; Woltjer, R.; Johnson, M.; Rooney, W. PLoS One 2016, 11 (4),

e0152480.

(201) Benardais, K.; Kotsiari, A.; Skuljec, J.; Koutsoudaki, P. N.; Gudi, V.; Singh, V.;

Vulinovic, F.; Skripuletz, T.; Stangel, M. Neurotox Res 2013, 24 (2), 244–250.

(202) Pasquini, L. A.; Calatayud, C. A.; Bertone Uña, A. L.; Millet, V.; Pasquini, J. M.;

Soto, E. F. Neurochem Res 2007, 32 (2), 279–292.

(203) Messori, L.; Casini, A.; Gabbiani, C.; Sorace, L.; Muniz-Miranda, M.; Zatta, P.

Dalt. Trans. 2007, No. 21, 2112–2114.

(204) Buschmann, J. P.; Berger, K.; Awad, H.; Clarner, T.; Beyer, C.; Kipp, M. J. Mol.

Neurosci. 2012, 48 (1), 66–76.

(205) Aboul-Enein, F.; Rauschka, H.; Kornek, B.; Stadelmann, C.; Stefferl, A.; Brück,

W.; Lucchinetti, C.; Schmidbauer, M.; Jellinger, K.; Lassmann, H. J. Neuropathol.

Exp. Neurol. 2003, 62 (1), 25–33.

(206) Suzuki, K. Science 1969, 163 (3862), 81–82.

(207) Petronilli, V.; Zoratti, M. J. Bioenerg. Biomembr. 1990, 22 (5), 663–677.

(208) Wagner, T.; Rafael, J. Exp. Cell Res. 1977, 107 (1), 1–13.

(209) Wakabayashi, T.; Asano, M.; Ishikawa, K.; Kishimoto, H. Acta Pathol. Jpn. 1978,

28 (2), 215–223.

(210) Tandler, B.; Hoppel, C. L. J. Cell Biol. 1973, 56 (1), 266–272.

(211) Acs, P.; Komoly, S. Ideggyogy. Sz. 2012, 65 (7–8), 266–270.

140

(212) Biancotti, J. C.; Kumar, S.; de Vellis, J. Neurochem. Res. 2008, 33 (12), 2615–

2628.

(213) Wakabayashi, T. J. Cell. Mol. Med. 6 (4), 497–538.

(214) Wakabayashi, T.; Asano, M.; Kurono, C. Acta Pathol. Jpn. 1975, 25 (1), 39–49.

(215) De, A. K.; Subramanian, M. Experientia 1982, 38 (7), 784–785.

(216) Kimberlin, R. H.; Collis, S. C.; Walker, C. A. J. Comp. Pathol. 1976, 86 (1), 135–

142.

(217) Millet, V.; Moiola, C. P.; Pasquini, J. M.; Soto, E. F.; Pasquini, L. A. Exp. Neurol.

2009, 217 (2), 282–296.

(218) Acs, P.; Selak, M. A.; Komoly, S.; Kalman, B. J. Neuroimmunol. 2013, 262 (1–2),

128–131.

(219) Nakajima, K.; Suzuki, K. Neurochem. Int. 1995, 27 (1), 73–87.

(220) Xu, H.; Yang, H.-J.; Li, X.-M. Eur. Arch. Psychiatry Clin. Neurosci. 2014, 264

(2), 121–129.

(221) Witherick, J.; Wilkins, A.; Scolding, N.; Kemp, K. Autoimmune Dis. 2010, 2011,

164608.

(222) Ohdate, T.; Inoue, Y. Biochim. Biophys. Acta - Mol. Cell Biol. Lipids 2012, 1821

(9), 1295–1305.

(223) Horiguchi, H.; Yurimoto, H.; Kato, N.; Sakai, Y. J. Biol. Chem. 2001, 276 (17),

14279–14288.

(224) Juurlink, B. H.; Thorburne, S. K.; Hertz, L. Glia 1998, 22 (4), 371–378.

(225) Bernardo, A.; Greco, A.; Levi, G.; Minghetti, L. J. Neuropathol. Exp. Neurol.

2003, 62 (5), 509–519.

141

(226) Russanov, E. M.; Ljutakova, S. G. Gen. Pharmacol. 1980, 11 (6), 535–538.

(227) Ljutakova, S. G.; Russanov, E. M. Acta Physiol. Pharmacol. Bulg. 1985, 11 (2),

56–61.

(228) Harding, J. W.; Pyeritz, E. A.; Copeland, E. S.; White, H. B. Biochem. J. 1975,

146 (1).

(229) Mráček, T.; Drahota, Z.; Houštěk, J. Biochim. Biophys. Acta - Bioenerg. 2013,

1827 (3), 401–410.

(230) Hemm, R. D.; Carlton, W. W.; Welser, J. R. Toxicol. Appl. Pharmacol. 1971, 18

(4), 869–882.

(231) Love, S. J Neurol Sci 1988, 84 (2–3), 223–237.

(232) Goldberg, J.; Daniel, M.; van Heuvel, Y.; Victor, M.; Beyer, C.; Clarner, T.; Kipp,

M. Cell. Mol. Neurobiol. 2013, 33 (8), 1087–1098.

(233) Jurevics, H.; Hostettler, J.; Muse, E. D.; Sammond, D. W.; Matsushima, G. K.;

Toews, A. D.; Morell, P. J. Neurochem. 2001, 77 (4), 1067–1076.

(234) Jurevics, H.; Largent, C.; Hostettler, J.; Sammond, D. W.; K. Matsushima, G.;

Amber, K.; Toews, A. D.; Morell, P. J. Neurochem. 2002, 82 (1), 126–136.

(235) Morell, P.; Barrett, C. V.; Mason, J. L.; Toews, A. D.; Hostettler, J. D.; Knapp, G.

W.; Matsushima, G. K. Mol. Cell. Neurosci. 1998, 12 (4–5), 220–227.

(236) Blakemore, W. F. J. Neurocytol. 1972, 1 (4), 413–426.

(237) Ludwin, S. K.; Sternberger, N. H. Acta Neuropathol. 1984, 63 (3), 240–248.

(238) Doan, V.; Kleindienst, A. M.; McMahon, E. J.; Long, B. R.; Matsushima, G. K.;

Taylor, L. C. J Neurosci Res 2013, 91 (3), 363–373.

(239) Liu, L.; Belkadi, A.; Darnall, L.; Hu, T.; Drescher, C.; Cotleur, A. C.; Padovani-

142

Claudio, D.; He, T.; Choi, K.; Lane, T. E.; Miller, R. H.; Ransohoff, R. M. Nat.

Neurosci. 2010, 13 (3), 319–326.

(240) Cammer, W.; Zhang, H. J. Neurosci. Res. 1993, 36 (2), 183–190.

(241) Zatta, P.; Raso, M.; Zambenedetti, P.; Wittkowski, W.; Messori, L.; Piccioli, F.;

Mauri, P. L.; Beltramini, M. Cell. Mol. Life Sci. 2005, 62 (13), 1502–1513.

(242) Kesterson, J. W.; Carlton, W. W. Exp. Mol. Pathol. 1971, 15 (1), 82–96.

(243) Olah, M.; Amor, S.; Brouwer, N.; Vinet, J.; Eggen, B.; Biber, K.; Boddeke, H. W.

G. M. Glia 2012, 60 (2), 306–321.

(244) Arnett, H. A.; Wang, Y.; Matsushima, G. K.; Suzuki, K.; Ting, J. P.-Y. J.

Neurosci. 2003, 23 (30), 9824–9832.

(245) Liberto, C. M.; Albrecht, P. J.; Herx, L. M.; Yong, V. W.; Levison, S. W. J.

Neurochem. 2004, 89 (5), 1092–1100.

(246) Skripuletz, T.; Manzel, A.; Gropengießer, K.; Schäfer, N.; Gudi, V.; Singh, V.;

Salinas Tejedor, L.; Jörg, S.; Hammer, A.; Voss, E.; Vulinovic, F.; Degen, D.;

Wolf, R.; Lee, D.-H.; Pul, R.; Moharregh-Khiabani, D.; Baumgärtner, W.; Gold,

R.; Linker, R. A.; Stangel, M. Brain 2015, 138 (Pt 2), 398–413.

(247) McMahon, E. J.; Suzuki, K.; Matsushima, G. K. J. Neuroimmunol. 2002, 130 (1–

2), 32–45.

(248) Mason, J. L.; Jones, J. J.; Taniike, M.; Morell, P.; Suzuki, K.; Matsushima, G. K.

J. Neurosci. Res. 2000, 61 (3), 251–262.

(249) Ludwin, S. K. Lab. Invest. 1980, 43 (4), 382–387.

(250) Lindner, M.; Fokuhl, J.; Linsmeier, F.; Trebst, C.; Stangel, M. Neurosci. Lett.

2009, 453 (2), 120–125.

143

(251) Gudi, V.; Gingele, S.; Skripuletz, T.; Stangel, M. Front. Cell. Neurosci. 2014, 8,

73.

(252) Soundarapandian, M. M.; Selvaraj, V.; Lo, U.-G.; Golub, M. S.; Feldman, D. H.;

Pleasure, D. E.; Deng, W. Sci. Rep. 2011, 1, 2.

(253) Othman, A.; Frim, D. M.; Polak, P.; Vujicic, S.; Arnason, B. G. W.; Boullerne, A.

I. Glia 2011, 59 (6), 914–926.

(254) Chen, L. P.; Li, Z. F.; Ping, M.; Li, R.; Liu, J.; Xie, X. H.; Song, X. J.; Guo, L.

Neuroscience 2012, 221, 96–107.

(255) Dugas, J. C.; Ibrahim, A.; Barres, B. A. Mol. Cell. Neurosci. 2012, 50 (1), 45–57.

(256) Paez, P. M.; Cheli, V. T.; Ghiani, C. A.; Spreuer, V.; Handley, V. W.;

Campagnoni, A. T. Glia 2012, 60 (7), 1078–1093.

(257) VonDran, M. W.; Singh, H.; Honeywell, J. Z.; Dreyfus, C. F. J. Neurosci. 2011, 31

(40).

(258) Hoyos, H. C.; Rinaldi, M.; Mendez-Huergo, S. P.; Marder, M.; Rabinovich, G. A.;

Pasquini, J. M.; Pasquini, L. A. Neurobiol. Dis. 2014, 62, 441–455.

(259) Li, C.; Xiao, L.; Liu, X.; Yang, W.; Shen, W.; Hu, C.; Yang, G.; He, C. Glia 2013,

61 (5), 732–749.

(260) Tyler, W. A.; Gangoli, N.; Gokina, P.; Kim, H. A.; Covey, M.; Levison, S. W.;

Wood, T. L. J. Neurosci. 2009, 29 (19), 6367–6378.

(261) Norrmén, C.; Suter, U. Biochem. Soc. Trans. 2013, 41 (4), 944–950.

(262) Wahl, S. E.; McLane, L. E.; Bercury, K. K.; Macklin, W. B.; Wood, T. L. J.

Neurosci. 2014, 34 (13), 4453–4465.

(263) Sachs, H. H.; Bercury, K. K.; Popescu, D. C.; Narayanan, S. P.; Macklin, W. B.

144

ASN Neuro 2014, 6 (5).

(264) Benetti, F.; Ventura, M.; Salmini, B.; Ceola, S.; Carbonera, D.; Mammi, S.; Zitolo,

A.; D’Angelo, P.; Urso, E.; Maffia, M.; Salvato, B.; Spisni, E. Neurotoxicology

2010, 31 (5), 509–517.

(265) Benetti, F.; Ventura, M.; Salmini, B.; Ceola, S.; Carbonera, D.; Mammi, S.; Zitolo,

A.; D’Angelo, P.; Urso, E.; Maffia, M.; Salvato, B.; Spisni, E. Neurotoxicology

2010, 31 (5), 509–517.

(266) Tandler, B.; Hoppel, C. L. Beitr. Pathol. 1975, 156 (1), 56–64.

(267) LindstrÖM, A.; Pettersson, G. Eur. J. Biochem. 1974, 48 (1), 229–236.

(268) Finberg, J. P. M.; Rabey, J. M. Front. Pharmacol. 2016, 7, 340.

(269) Lightcap, E. S.; Silverman, R. B. J. Med. Chem. 1996, 39 (3), 686–694.

(270) Patek, D. R.; Hellerman, L. J. Biol. Chem. 1974, 249 (8), 2373–2380.

(271) Roth, J. A. Biochem. Pharmacol. 1979, 28 (6), 729–732.

(272) Levy, J.; Michel-Ber, E. Therapie 21 (4), 929–945.

(273) Fagervall, I.; Ross, S. B. Biochem. Pharmacol. 1986, 35 (8), 1381–1387.

(274) Gilmour, S. J. Br. J. Psychiatry 1965, 111 (478), 899–902.

(275) Fiedorowicz, J. G.; Swartz, K. L. J. Psychiatr. Pract. 2004, 10 (4), 239–248.

(276) Herman, Z. S.; Sokola, A.; Lenartowicz, H.; Zieliński, M.; Depta, L. Pol. J.

Pharmacol. Pharm. 28 (4), 313–321.

(277) LEAHY, M. R.; ROSE, J. T.; PLOWMAN, R. Am. J. Psychiatry 1963, 119 (10),

986–987.

(278) Dictionary of pharmacological agents.; Chapman & Hall, 1997.

(279) Wishart, D. S. Nat. Rev. Drug Discov. 2016, 15 (7), 473–484.

145

(280) Johnson, C. H.; Ivanisevic, J.; Siuzdak, G. Nat. Rev. Mol. Cell Biol. 2016, 17 (7),

451–459.

(281) Fukusaki, E. Mass Spectrom. (Tokyo, Japan) 2014, 3 (Spec Iss 3), S0045.

(282) Link, H.; Fuhrer, T.; Gerosa, L.; Zamboni, N.; Sauer, U. Nat. Methods 2015, 12

(11), 1091–1097.

(283) Larive, C. K.; Barding, G. A.; Dinges, M. M. Anal. Chem. 2015, 87 (1), 133–146.

(284) Junot, C.; Fenaille, F.; Colsch, B.; Bécher, F. Mass Spectrom. Rev. 2014, 33 (6),

471–500.

(285) Marshall, D. D.; Lei, S.; Worley, B.; Huang, Y.; Aracely, @bullet; @bullet, G.-G.;

Franco, R.; Dodds, E. D.; Powers, R. .

(286) Bingol, K.; Brüschweiler, R. Curr. Opin. Clin. Nutr. Metab. Care 2015, 18 (5),

471–477.

(287) Teng, Q. In Structural Biology; Springer US: Boston, MA, 2013; pp 103–116.

(288) Beltran, A.; Suarez, M.; Rodríguez, M. A.; Vinaixa, M.; Samino, S.; Arola, L.;

Correig, X.; Yanes, O. Anal. Chem. 2012, 84 (14), 5838–5844.

(289) Ian A. Lewis; Seth C. Schommer; Brendan Hodis; Kate A. Robb; Marco Tonelli;

William M. Westler; Michael R. Sussman, and; Markley*, J. L. 2007.

(290) Yin, P.; Xu, G. J. Chromatogr. A 2014, 1374, 1–13.

(291) Rainville, P. D.; Theodoridis, G.; Plumb, R. S.; Wilson, I. D. TrAC Trends Anal.

Chem. 2014, 61, 181–191.

(292) Fang, M.; Ivanisevic, J.; Benton, H. P.; Johnson, C. H.; Patti, G. J.; Hoang, L. T.;

Uritboonthai, W.; Kurczy, M. E.; Siuzdak, G. Anal. Chem. 2015, 87 (21), 10935–

10941.

146

(293) Fiehn, O.; Kopka, J.; Dörmann, P.; Altmann, T.; Trethewey, R. N.; Willmitzer, L.

Nat. Biotechnol. 2000, 18 (11), 1157–1161.

(294) Dunn, W. B.; Broadhurst, D.; Ellis, D. I.; Brown, M.; Halsall, A.; O’Hagan, S.;

Spasic, I.; Tseng, A.; Kell, D. B. Int. J. Epidemiol. 2008, 37 (Supplement 1), i23–

i30.

(295) Heftmann, E. Chromatography : fundamentals and applications of

chromatography and related differential migration methods. Part A, Fundamentals

and techniques; Elsevier, 2004.

(296) GC-MS Metabolomics Analysis

https://www.thermofisher.com/us/en/home/industrial/mass-spectrometry/mass-

spectrometry-learning-center/mass-spectrometry-applications-area/metabolomics-

mass-spectrometry/practical-guide-metabolomics/gc-ms-metabolomics-

analysis.html (accessed Feb 16, 2017).

(297) Dettmer, K.; Aronov, P. A.; Hammock, B. D. Mass Spectrom. Rev. 2007, 26 (1),

51–78.

(298) Pinkerton, T. C. J. Chromatogr. A 1991, 544, 13–23.

(299) Chromatographic Columns

https://chem.libretexts.org/Core/Analytical_Chemistry/Instrumental_Analysis/Chr

omatography/Chromatographic_Columns (accessed May 3, 2017).

(300) Cubbon, S.; Antonio, C.; Wilson, J.; Thomas-Oates, J. Mass Spectrom Rev 2010,

29 (5), 671–684.

(301) Fei, F.; Bowdish, D. M.; McCarry, B. E. Anal Bioanal Chem 2014, 406 (15),

3723–3733.

147

(302) Jian, W.; Edom, R. W.; Xu, Y.; Weng, N. J Sep Sci 2010, 33 (6–7), 681–697.

(303) HPLC Separation Modes http://www.waters.com/waters/en_US/HPLC-

Separation-Modes/nav.htm?cid=10049076&locale=en_US (accessed Feb 16,

2017).

(304) Aguilar, M.-I. In HPLC of Peptides and Proteins; Humana Press: New Jersey; pp

9–22.

(305) Cooper, W. T.; Cooper; T., W. In Encyclopedia of Analytical Chemistry; John

Wiley & Sons, Ltd: Chichester, UK, 2006.

(306) Bajad, S. U.; Lu, W.; Kimball, E. H.; Yuan, J.; Peterson, C.; Rabinowitz, J. D. J.

Chromatogr. A 2006, 1125 (1), 76–88.

(307) Mä rk, T. D. (Tilmann D. .; Dunn, G. H. (Gordon H. . Electron impact ionization;

Springer-Verlag, 1985.

(308) Karas, M.; Bachmann, D.; Bahr, U.; Hillenkamp, F. Int. J. Mass Spectrom. Ion

Process. 1987, 78, 53–68.

(309) Fenn, J.; Mann, M.; Meng, C.; Wong, S.; Whitehouse, C. Science (80-. ). 1989,

246 (4926).

(310) Zhou, B.; Xiao, J. F.; Tuli, L.; Ressom, H. W. Mol. Biosyst. 2012, 8 (2), 470–481.

(311) Xiao, J. F.; Zhou, B.; Ressom, H. W. TrAC Trends Anal. Chem. 2012, 32, 1–14.

(312) Lu, W.; Bennett, B. D.; Rabinowitz, J. D. J. Chromatogr. B 2008, 871 (2), 236–

242.

(313) Wilm, M. Mol. Cell. Proteomics 2011, 10 (7), M111.009407.

(314) Ho, C. S.; Lam, C. W. K.; Chan, M. H. M.; Cheung, R. C. K.; Law, L. K.; Lit, L.

C. W.; Ng, K. F.; Suen, M. W. M.; Tai, H. L. Clin. Biochem. Rev. 2003, 24 (1), 3–

148

12.

(315) García, M. C. J. Chromatogr. B 2005, 825 (2), 111–123.

(316) Mei, H.; Hsieh, Y.; Nardo, C.; Xu, X.; Wang, S.; Ng, K.; Korfmacher, W. A.

Rapid Commun. Mass Spectrom. 2003, 17 (1), 97–103.

(317) Hoffmann, E. de.; Stroobant, V. Mass spectrometry : principles and applications;

J. Wiley, 2007.

(318) Yost, R. A.; Enke, C. G. J. Am. Chem. Soc. 1978, 100 (7), 2274–2275.

(319) Chernushevich, I. V.; Loboda, A. V.; Thomson, B. A. J. Mass Spectrom. 2001, 36

(8), 849–865.

(320) Wollnik, H. Mass Spectrom. Rev. 1993, 12 (2), 89–114.

(321)

(322) Tolmachev, A. V.; Vilkov, A. N.; Bogdanov, B.; Pǎsa-Tolić, L.; Masselon, C. D.;

Smith, R. D. J. Am. Soc. Mass Spectrom. 2004, 15 (11), 1616–1628.

(323) Roberts, L. D.; Souza, A. L.; Gerszten, R. E.; Clish, C. B. In Current Protocols in

Molecular Biology; John Wiley & Sons, Inc.: Hoboken, NJ, USA, 2012; Vol.

Chapter 30, p Unit 30.2.1-24.

(324) Guo, B.; Chen, B.; Liu, A.; Zhu, W.; Yao, S. Curr. Drug Metab. 2012, 13 (9),

1226–1243.

(325) Chen, J.; Hou, W.; Han, B.; Liu, G.; Gong, J.; Li, Y.; Zhong, D.; Liao, Q.; Xie, Z.

Anal. Bioanal. Chem. 2016, 408 (10), 2527–2542.

(326) Dietmair, S.; Timmins, N. E.; Gray, P. P.; Nielsen, L. K.; Kromer, J. O. Anal

Biochem 2010, 404 (2), 155–164.

(327) Dettmer, K.; Nurnberger, N.; Kaspar, H.; Gruber, M. A.; Almstetter, M. F.;

149

Oefner, P. J. Anal Bioanal Chem 2011, 399 (3), 1127–1139.

(328) Sellick, C. A.; Knight, D.; Croxford, A. S.; Maqsood, A. R.; Stephens, G. M.;

Goodacre, R.; Dickson, A. J. Metabolomics 2010, 6 (3), 427–438.

(329) Lin, C.; Wu, H.; Tjeerdema, R.; Viant, M. Metabolomics 2007, 3 (1), 55–67.

(330) Wei, R.; Li, G.; Seymour, A. B. Anal. Chem. 2010, 82 (13), 5527–5533.

(331) de Hoffmann, E. J. Mass Spectrom. 1996, 31 (2), 129–137.

(332) Lange, V.; Picotti, P.; Domon, B.; Aebersold, R. Mol. Syst. Biol. 2008, 4 (1), 222.

(333) Kondrat, R. W.; McClusky, G. A.; Cooks, R. G. Anal. Chem. 1978, 50 (14), 2017–

2021.

(334) Peterson, A. C.; Russell, J. D.; Bailey, D. J.; Westphall, M. S.; Coon, J. J. Mol.

Cell. Proteomics 2012, 11 (11), 1475–1488.

(335) Yanes, O.; Clark, J.; Wong, D. M.; Patti, G. J.; Sánchez-Ruiz, A.; Benton, H. P.;

Trauger, S. A.; Desponts, C.; Ding, S.; Siuzdak, G. Nat. Chem. Biol. 2010, 6 (6),

411–417.

(336) Patti, G. J.; Yanes, O.; Shriver, L. P.; Courade, J.-P.; Tautenhahn, R.; Manchester,

M.; Siuzdak, G. Nat. Chem. Biol. 2012, 8 (3), 232–234.

(337) Jamin, E. L.; Bonvallot, N.; Tremblay-Franco, M.; Cravedi, J.-P.; Chevrier, C.;

Cordier, S.; Debrauwer, L. Anal. Bioanal. Chem. 2014, 406 (4), 1149–1161.

(338) Astarita, G.; Langridge, J. J. Nutrigenet. Nutrigenomics 2013, 6 (4–5), 181–200.

(339) Huang, H.; Taraboletti, A.; Shriver, L. P. Redox Biol. 2015, 5 (0), 169–175.

(340) Scalbert, A.; Brennan, L.; Fiehn, O.; Hankemeier, T.; Kristal, B. S.; van Ommen,

B.; Pujos-Guillot, E.; Verheij, E.; Wishart, D.; Wopereis, S. Metabolomics 2009, 5

(4), 435–458.

150

(341) Robertson, D. G.; Watkins, P. B.; Reily, M. D. Toxicol. Sci. 2011, 120

(Supplement 1), S146–S170.

(342) Trock, B. J. Urol. Oncol. Semin. Orig. Investig. 2011, 29 (5), 572–581.

(343) Yanes, O.; Tautenhahn, R.; Patti, G. J.; Siuzdak, G. Anal Chem 2011, 83 (6),

2152–2161.

(344) Sapcariu, S. C.; Kanashova, T.; Weindl, D.; Ghelfi, J.; Dittmar, G.; Hiller, K.

MethodsX 2014, 1, 74–80.

(345) Patti, G. J. J Sep Sci 2011, 34 (24), 3460–3469.

(346) Wu, Y.; Li, L. J. Chromatogr. A 2016, 1430, 80–95.

(347) Alonso, A.; Marsal, S.; Julià, A. Front. Bioeng. Biotechnol. 2015, 3, 23.

(348) Bijlsma, S.; Bobeldijk, I.; Verheij, E. R.; Ramaker, R.; Kochhar, S.; Macdonald, I.

A.; van Ommen, B.; Smilde, A. K. Anal. Chem. 2006, 78 (2), 567–574.

(349) Sysi-Aho, M.; Katajamaa, M.; Yetukuri, L.; Orešič, M. BMC Bioinformatics 2007,

8 (1), 93.

(350) Kirwan, J. A.; Broadhurst, D. I.; Davidson, R. L.; Viant, M. R. Anal. Bioanal.

Chem. 2013, 405 (15), 5147–5157.

(351) Dunn, W. B.; Broadhurst, D.; Begley, P.; Zelena, E.; Francis-McIntyre, S.;

Anderson, N.; Brown, M.; Knowles, J. D.; Halsall, A.; Haselden, J. N.; Nicholls,

A. W.; Wilson, I. D.; Kell, D. B.; Goodacre, R.; Human Serum Metabolome

(HUSERMET) Consortium. Nat. Protoc. 2011, 6 (7), 1060–1083.

(352) Dunn, W. B.; Wilson, I. D.; Nicholls, A. W.; Broadhurst, D. Bioanalysis 2012, 4

(18), 2249–2264.

(353) Kamleh, M. A.; Ebbels, T. M. D.; Spagou, K.; Masson, P.; Want, E. J. Anal.

151

Chem. 2012, 84 (6), 2670–2677.

(354) Fernandez-Albert, F.; Llorach, R.; Garcia-Aloy, M.; Ziyatdinov, A.; Andres-

Lacueva, C.; Perera, A. Bioinformatics 2014, 30 (20), 2899–2905.

(355) Nezami Ranjbar, M. R.; Zhao, Y.; Tadesse, M. G.; Wang, Y.; Ressom, H. W.

Proteome Sci. 2013, 11 (Suppl 1), S13.

(356) Stolt, R.; Torgrip, R. J. O.; Lindberg, J.; Csenki, L.; Kolmert, J.; Schuppe-

Koistinen, I.; Jacobsson, S. P. Anal. Chem. 2006, 78 (4), 975–983.

(357) Åberg, K. M.; Torgrip, R. J. O.; Kolmert, J.; Schuppe-Koistinen, I.; Lindberg, J. J.

Chromatogr. A 2008, 1192 (1), 139–146.

(358) Xueguang Shao, *,†; Wensheng Cai, ‡; Peiyan Sun, ‡; Maosen Zhang, ‡ and;

Zhao†, G. 1997.

(359) Tautenhahn, R.; Bottcher, C.; Neumann, S. BMC Bioinformatics 2008, 9 (1), 504.

(360) Smith, R.; Ventura, D.; Prince, J. T. Brief. Bioinform. 2015, 16 (1), 104–117.

(361) Ejigu, B. A.; Valkenborg, D.; Baggerman, G.; Vanaerschot, M.; Witters, E.;

Dujardin, J.-C.; Burzykowski, T.; Berg, M. Omi. A J. Integr. Biol. 2013, 17 (9),

473–485.

(362) Vinayavekhin, N.; Saghatelian, A. In Current Protocols in Molecular Biology;

John Wiley & Sons, Inc.: Hoboken, NJ, USA, 2010; Vol. Chapter 30, p Unit

30.1.1-24.

(363) van den Berg, R. A.; Hoefsloot, H. C. J.; Westerhuis, J. A.; Smilde, A. K.; van der

Werf, M. J. BMC Genomics 2006, 7, 142.

(364) Smith, C. A.; Want, E. J.; O’Maille, G.; Abagyan, R.; Siuzdak, G. Anal. Chem.

2006, 78 (3), 779–787.

152

(365) Katajamaa, M.; Orešič, M. BMC Bioinformatics 2005, 6 (1), 179.

(366) Lommen, A. Anal. Chem. 2009, 81 (8), 3079–3086.

(367) Vinaixa, M.; Samino, S.; Saez, I.; Duran, J.; Guinovart, J. J.; Yanes, O.

Metabolites 2012, 2 (4), 775–795.

(368) Ioannidis, J. P. A. PLoS Med. 2005, 2 (8), e124.

(369) Storey, J. D.; Tibshirani, R. Proc. Natl. Acad. Sci. 2003, 100 (16), 9440–9445.

(370) Broadhurst, D. I.; Kell, D. B. Metabolomics 2007, 2 (4), 171–196.

(371) Worley, B.; Powers, R. Curr. Metabolomics 2013, 1 (1), 92–107.

(372) Shlens, J. 2005.

(373) Abdi, H.; Williams, L. J. Wiley Interdiscip. Rev. Comput. Stat. 2010, 2 (4), 433–

459.

(374) Bartel, J.; Krumsiek, J.; Theis, F. J. Comput. Struct. Biotechnol. J. 2013, 4 (5).

(375) Orlitsky, A. .

(376) Wold, S.; Sjöström, M.; Eriksson, L. Chemom. Intell. Lab. Syst. 2001, 58 (2), 109–

130.

(377) Wishart, D. S.; Tzur, D.; Knox, C.; Eisner, R.; Guo, A. C.; Young, N.; Cheng, D.;

Jewell, K.; Arndt, D.; Sawhney, S.; Fung, C.; Nikolai, L.; Lewis, M.; Coutouly, M.

A.; Forsythe, I.; Tang, P.; Shrivastava, S.; Jeroncic, K.; Stothard, P.; Amegbey, G.;

Block, D.; Hau, D. D.; Wagner, J.; Miniaci, J.; Clements, M.; Gebremedhin, M.;

Guo, N.; Zhang, Y.; Duggan, G. E.; Macinnis, G. D.; Weljie, A. M.; Dowlatabadi,

R.; Bamforth, F.; Clive, D.; Greiner, R.; Li, L.; Marrie, T.; Sykes, B. D.; Vogel, H.

J.; Querengesser, L. Nucleic Acids Res 2007, 35 (Database issue), D521-6.

(378) Kanehisa, M.; Goto, S. Nucleic Acids Res. 2000, 28 (1), 27–30.

153

(379) Smith, C. A.; O’Maille, G.; Want, E. J.; Qin, C.; Trauger, S. A.; Brandon, T. R.;

Custodio, D. E.; Abagyan, R.; Siuzdak, G. Ther Drug Monit 2005, 27 (6), 747–

751.

(380) Weber, R. J. M.; Lawson, T. N.; Salek, R. M.; Ebbels, T. M. D.; Glen, R. C.;

Goodacre, R.; Griffin, J. L.; Haug, K.; Koulman, A.; Moreno, P.; Ralser, M.;

Steinbeck, C.; Dunn, W. B.; Viant, M. R. Metabolomics 2017, 13 (2), 12.

(381) Want, E. J.; Wilson, I. D.; Gika, H.; Theodoridis, G.; Plumb, R. S.; Shockcor, J.;

Holmes, E.; Nicholson, J. K. Nat. Protoc. 2010, 5 (6), 1005–1018.

(382) Lynn, K.-S.; Cheng, M.-L.; Chen, Y.-R.; Hsu, C.; Chen, A.; Lih, T. M.; Chang, H.-

Y.; Huang, C.; Shiao, M.-S.; Pan, W.-H.; Sung, T.-Y.; Hsu, W.-L. Anal. Chem.

2015, 87 (4), 2143–2151.

(383) Allen, F.; Greiner, R.; Wishart, D. Metabolomics 2015, 11 (1), 98–110.

(384) Wenk, M. R. Nat. Rev. Drug Discov. 2005, 4 (7), 594–610.

(385) Loizides-Mangold, U. FEBS J. 2013, 280 (12), 2817–2829.

(386) Köfeler, H. C.; Fauland, A.; Rechberger, G. N.; Trötzmüller, M. Metabolites 2012,

2 (4), 19–38.

(387) Orešič, M.; Hänninen, V. A.; Vidal-Puig, A. Trends Biotechnol. 2008, 26 (12),

647–652.

(388) Hannun, Y. A.; Obeid, L. M. Nat. Rev. Mol. Cell Biol. 2008, 9 (2), 139–150.

(389) Wymann, M. P.; Schneiter, R. Nat. Rev. Mol. Cell Biol. 2008, 9 (2), 162–176.

(390) Fahy, E.; Subramaniam, S.; Brown, H. A.; Glass, C. K.; Merrill, A. H.; Murphy, R.

C.; Raetz, C. R. H.; Russell, D. W.; Seyama, Y.; Shaw, W.; Shimizu, T.; Spener,

F.; van Meer, G.; VanNieuwenhze, M. S.; White, S. H.; Witztum, J. L.; Dennis, E.

154

A. J. Lipid Res. 2005, 46 (5), 839–862.

(391) Paradies, G.; Paradies, V.; De Benedictis, V.; Ruggiero, F. M.; Petrosillo, G.

Biochim. Biophys. Acta - Bioenerg. 2014, 1837 (4), 408–417.

(392) FOLCH, J.; LEES, M.; SLOANE STANLEY, G. H. J. Biol. Chem. 1957, 226 (1),

497–509.

(393) Bligh, E. G.; Dyer, W. J. Can. J. Biochem. Physiol. 1959, 37 (8), 911–917.

(394) Matyash, V.; Liebisch, G.; Kurzchalia, T. V; Shevchenko, A.; Schwudke, D. J.

Lipid Res. 2008, 49 (5), 1137–1146.

(395) Deems, R.; Buczynski, M. W.; Bowers‐Gentry, R.; Harkewicz, R.; Dennis, E. A.

Methods Enzymol. 2007, 432, 59–82.

(396) Garrett, T. A.; Guan, Z.; Raetz, C. R. H. Methods Enzymol. 2007, 432, 117–143.

(397) Sullards, M. C.; Allegood, J. C.; Kelly, S.; Wang, E.; Haynes, C. A.; Park, H.;

Chen, Y.; Merrill, A. H. Methods Enzymol. 2007, 432, 83–115.

(398) Schmelzer, K.; Fahy, E.; Subramaniam, S.; Dennis, E. A. Methods Enzymol. 2007,

432, 171–183.

(399) Hsu, F.-F.; Turk, J. J. Chromatogr. B 2009, 877 (26), 2673–2695.

(400) Han, X.; Gross, R. W. J. Lipid Res. 2003, 44 (6), 1071–1079.

(401) Han, X.; Gross, R. W. Mass Spectrom Rev 2005, 24 (3), 367–412.

(402) Simons, B.; Kauhanen, D.; Sylvänne, T.; Tarasov, K.; Duchoslav, E.; Ekroos, K.

Metabolites 2012, 2 (1), 195–213.

(403) Schwudke, D.; Hannich, J. T.; Surendranath, V.; Grimard, V.; Moehring, T.;

Burton, L.; Kurzchalia, T.; Shevchenko, A. Anal. Chem. 2007, 79 (11), 4083–

4093.

155

(404) Shevchenko, A.; Simons, K. Nat. Rev. Mol. Cell Biol. 2010, 11 (8), 593–598.

(405) Schuhmann, K.; Herzog, R.; Schwudke, D.; Metelmann-Strupat, W.; Bornstein, S.

R.; Shevchenko, A. Anal. Chem. 2011, 83 (14), 5480–5487.

(406) Zhu, X.; Chen, Y.; Subramanian, R. Anal. Chem. 2014, 86 (2), 1202–1209.

(407) Fahy, E.; Subramaniam, S.; Murphy, R. C.; Nishijima, M.; Raetz, C. R. H.;

Shimizu, T.; Spener, F.; van Meer, G.; Wakelam, M. J. O.; Dennis, E. A. J. Lipid

Res. 2008, 50 (Supplement), S9–S14.

(408) Herzog, R.; Schuhmann, K.; Schwudke, D.; Sampaio, J. L.; Bornstein, S. R.;

Schroeder, M.; Shevchenko, A. PLoS One 2012, 7 (1), e29851.

(409) Peterson, R. E.; Bollier, M. E. Anal. Chem. 1955, 27 (7), 1195–1197.

(410) Bénardais, K.; Kotsiari, A.; Škuljec, J.; Koutsoudaki, P. N.; Gudi, V.; Singh, V.;

Vulinović, F.; Skripuletz, T.; Stangel, M. Neurotox. Res. 2013, 24 (2), 244–250.

(411) Muniz-Miranda, M.; Pagliai, M.; Cardini, G.; Messori, L.; Bruni, B.; Casini, A.; Di

Vaira, M.; Schettino, V. CrystEngComm 2008, 10 (4), 416–422.

(412) Yamamoto, N.; Kuwata, K. J. Mol. Struct. THEOCHEM 2009, 895 (1–3), 52–56.

(413) Kundu, S.; Mondal, A.; Weyhermüller, T.; Sproules, S.; Ghosh, P. Inorganica

Chim. Acta 2016, 451, 23–30.

(414) Rosenzweig, A. C. Acc. Chem. Res. 2001, 34 (2), 119–128.

(415) Prohaska, J. R. Am. J. Clin. Nutr. 2008, 88 (3), 826S–9S.

(416) Boal, A. K.; Rosenzweig, A. C. Chem. Rev. 2009, 109 (10), 4760–4779.

(417) Burkitt, M. J. Arch. Biochem. Biophys. 2001, 394 (1), 117–135.

(418) Evans, P. J.; Halliwell, B. Methods Enzymol. 1994, 233, 82–92.

(419) Gaetke, L. M.; Chow-Johnson, H. S.; Chow, C. K. Arch. Toxicol. 2014, 88 (11),

156

1929–1938.

(420) Walker, T. L.; Malasi, W.; Bhide, S.; Parker, T.; Zhang, D.; Freedman, A.;

Modarelli, J. M.; Engle, J. T.; Ziegler, C. J.; Custer, P.; Youngs, W. J.; Taschner,

M. J. Tetrahedron Lett. 2012, 53 (48), 6548–6551.

(421) Walker, T. L.; Mula, S.; Malasi, W.; Engle, J. T.; Ziegler, C. J.; van der Est, A.;

Modarelli, J.; Taschner, M. J. Dalt. Trans. 2015, 44 (46), 20200–20206.

(422) The NMR of CUPRIZONE

http://www.hanhonggroup.com/nmr/nmr_en/RE04000064.html (accessed Feb 25,

2017).

(423) Aggarwal, S.; Yurlova, L.; Simons, M. Trends Cell Biol. 2011, 21 (10), 585–593.

(424) Deshmukh, V. A.; Tardif, V.; Lyssiotis, C. A.; Green, C. C.; Kerman, B.; Kim, H.

J.; Padmanabhan, K.; Swoboda, J. G.; Ahmad, I.; Kondo, T.; Gage, F. H.;

Theofilopoulos, A. N.; Lawson, B. R.; Schultz, P. G.; Lairson, L. L. Nature 2013,

502 (7471), 327–332.

(425) Schmidt, T.; Awad, H.; Slowik, A.; Beyer, C.; Kipp, M.; Clarner, T. J. Mol.

Neurosci. 2013, 49 (1), 80–88.

(426) Goldberg, J.; Clarner, T.; Beyer, C.; Kipp, M. J. Mol. Neurosci. 2015, 57 (2), 166–

175.

(427) Wergeland, S.; Torkildsen, O.; Myhr, K. M.; Mork, S. J.; Bo, L. APMIS 2012, 120

(8), 648–657.

(428) Nathoo, N.; Yong, V. W.; Dunn, J. F. NeuroImage. Clin. 2014, 4, 743–756.

(429) Fjær, S.; Bø, L.; Lundervold, A.; Myhr, K.-M.; Pavlin, T.; Torkildsen, Ø.;

Wergeland, S. PLoS One 2013, 8 (12), e84162.

157

(430) Herder, V.; Hansmann, F.; Stangel, M.; Skripuletz, T.; Baumgartner, W.; Beineke,

A. Neuropathol Appl Neurobiol 2011, 37 (6), 676–684.

(431) Hesse, A.; Wagner, M.; Held, J.; Brück, W.; Salinas-Riester, G.; Hao, Z.;

Waisman, A.; Kuhlmann, T. Neurobiol. Dis. 2010, 37 (2), 362–369.

(432) Gudi, V.; Škuljec, J.; Yildiz, Ö.; Frichert, K.; Skripuletz, T.; Moharregh-Khiabani,

D.; Voß, E.; Wissel, K.; Wolter, S.; Stangel, M. PLoS One 2011, 6 (7), e22623.

(433) Hibbits, N.; Yoshino, J.; Le, T. Q.; Armstrong, R. C. ASN Neuro 2012, 4 (6), 393–

408.

(434) Doan, V.; Kleindienst, A. M.; McMahon, E. J.; Long, B. R.; Matsushima, G. K.;

Taylor, L. C. J. Neurosci. Res. 2013, 91 (3), 363–373.

(435) Voß, E. V.; Škuljec, J.; Gudi, V.; Skripuletz, T.; Pul, R.; Trebst, C.; Stangel, M.

Neurobiol. Dis. 2012, 45 (1), 519–528.

(436) Mason, J. L.; Suzuki, K.; Chaplin, D. D.; Matsushima, G. K. J. Neurosci. 2001, 21

(18), 7046–7052.

(437) Maña, P.; Liñares, D.; Fordham, S.; Staykova, M.; Willenborg, D. Am. J. Pathol.

2006, 168 (5), 1464–1473.

(438) Kang, Z.; Liu, L.; Spangler, R.; Spear, C.; Wang, C.; Gulen, M. F.; Veenstra, M.;

Ouyang, W.; Ransohoff, R. M.; Li, X. J. Neurosci. 2012, 32 (24), 8284–8292.

(439) Patti, G. J.; Yanes, O.; Siuzdak, G. Nat Rev Mol Cell Biol 2012, 13 (4), 263–269.

(440) Blasco, H.; Corcia, P.; Pradat, P.-F.; Bocca, C.; Gordon, P. H.; Veyrat-Durebex,

C.; Mavel, S.; Nadal-Desbarats, L.; Moreau, C.; Devos, D.; Andres, C. R.; Emond,

P. J. Proteome Res. 2013, 12 (8), 3746–3754.

(441) Smith, P. K.; Krohn, R. I.; Hermanson, G. T.; Mallia, A. K.; Gartner, F. H.;

158

Provenzano, M. D.; Fujimoto, E. K.; Goeke, N. M.; Olson, B. J.; Klenk, D. C. Anal

Biochem 1985, 150 (1), 76–85.

(442) Simons, B.; Kauhanen, D.; Sylvanne, T.; Tarasov, K.; Duchoslav, E.; Ekroos, K.

Metabolites 2012, 2 (1), 195–213.

(443) Basran, J.; Rafice, S. A.; Chauhan, N.; Efimov, I.; Cheesman, M. R.; Ghamsari, L.;

Raven, E. L. Biochemistry 2008, 47 (16), 4752–4760.

(444) Vinaixa, M.; Samino, S.; Saez, I.; Duran, J.; Guinovart, J. J.; Yanes, O.

Metabolites 2012, 2 (4), 775–795.

(445)

(446) Xia, J.; Sinelnikov, I. V; Han, B.; Wishart, D. S. Nucleic Acids Res 2015.

(447) Tautenhahn, R.; Cho, K.; Uritboonthai, W.; Zhu, Z.; Patti, G. J.; Siuzdak, G. Nat

Biotech 2012, 30 (9), 826–828.

(448) Ivanisevic, J.; Zhu, Z.-J.; Plate, L.; Tautenhahn, R.; Chen, S.; O’Brien, P. J.;

Johnson, C. H.; Marletta, M. A.; Patti, G. J.; Siuzdak, G. Anal. Chem. 2013, 85

(14), 6876–6884.

(449) Wishart, D. S.; Jewison, T.; Guo, A. C.; Wilson, M.; Knox, C.; Liu, Y.;

Djoumbou, Y.; Mandal, R.; Aziat, F.; Dong, E.; Bouatra, S.; Sinelnikov, I.; Arndt,

D.; Xia, J.; Liu, P.; Yallou, F.; Bjorndahl, T.; Perez-Pineiro, R.; Eisner, R.; Allen,

F.; Neveu, V.; Greiner, R.; Scalbert, A. Nucleic Acids Res 2013, 41 (Database

issue), D801–D807.

(450) Flatmark, T.; Kryvi, H.; Tangerås, A. Eur. J. Cell Biol. 1980, 23 (1), 141–148.

(451) Belenky, P.; Bogan, K. L.; Brenner, C. Trends Biochem Sci 2007, 32 (1), 12–19.

(452) Funfschilling, U.; Supplie, L. M.; Mahad, D.; Boretius, S.; Saab, A. S.; Edgar, J.;

159

Brinkmann, B. G.; Kassmann, C. M.; Tzvetanova, I. D.; Mobius, W.; Diaz, F.;

Meijer, D.; Suter, U.; Hamprecht, B.; Sereda, M. W.; Moraes, C. T.; Frahm, J.;

Goebbels, S.; Nave, K.-A. Nature 2012, 485 (7399), 517–521.

(453) Warburg, O. Science 1956, 124 (3215), 269–270.

(454) Varum, S.; Rodrigues, A. S.; Moura, M. B.; Momcilovic, O.; Easley, C. A.;

Ramalho-Santos, J.; Van Houten, B.; Schatten, G. PLoS One 2011, 6 (6), e20914.

(455) Li, W.; Zhang, B.; Tang, J.; Cao, Q.; Wu, Y.; Wu, C.; Guo, J.; Ling, E. A.; Liang,

F. J Neurosci 2007, 27 (10), 2606–2616.

(456) Bogan, K. L.; Brenner, C. Annu Rev Nutr 2008, 28, 115–130.

(457) Torkildsen, Ø.; Brunborg, L. A.; Myhr, K. M.; Bø, L. Acta Neurol. Scand. 2008,

117, 72–76.

(458) Goldberg, J.; Daniel, M.; van Heuvel, Y.; Victor, M.; Beyer, C.; Clarner, T.; Kipp,

M. Cell. Mol. Neurobiol. 2013, 33 (8), 1087–1098.

(459) Carvalho, A. N.; Lim, J. L.; Nijland, P. G.; Witte, M. E.; Van Horssen, J. Mult.

Scler. 2014, 20 (11), 1425–1431.

(460) Johnson, W. M.; Wilson-Delfosse, A. L.; Mieyal, J. J. Nutrients 2012, 4 (10),

1399–1440.

(461) Pompella, A.; Visvikis, A.; Paolicchi, A.; Tata, V. De; Casini, A. F. Biochem.

Pharmacol. 2003, 66 (8), 1499–1503.

(462) Werner, S. R.; Saha, J. K.; Broderick, C. L.; Zhen, E. Y.; Higgs, R. E.; Duffin, K.

L.; Smith, R. C. J. Mol. Neurosci. 2010, 42 (2), 210–225.

(463) Partridge, M. A.; Gopinath, S.; Myers, S. J.; Coorssen, J. R. J. Chem. Biol. 2016, 9

(1), 9–18.

160

(464) Marques, S.; Zeisel, A.; Codeluppi, S.; van Bruggen, D.; Mendanha Falcão, A.;

Xiao, L.; Li, H.; Häring, M.; Hochgerner, H.; Romanov, R. A.; Gyllborg, D.;

Muñoz-Manchado, A. B.; La Manno, G.; Lönnerberg, P.; Floriddia, E. M.;

Rezayee, F.; Ernfors, P.; Arenas, E.; Hjerling-Leffler, J.; Harkany, T.; Richardson,

W. D.; Linnarsson, S.; Castelo-Branco, G. Science 2016, 352 (6291), 1326–1329.

(465) Morell, P.; Quarles, R. H. Lippincott-Raven 1999.

(466) Bansal, R.; Winkler, S.; Bheddah, S. J. Neurosci. 1999, 19 (18), 7913–7924.

(467) Sugimoto, H.; Oda, S. -i.; Otsuki, T.; Hino, T.; Yoshida, T.; Shiro, Y. Proc. Natl.

Acad. Sci. 2006, 103 (8), 2611–2616.

(468) Zhang, Y.; Kang, S. A.; Mukherjee, T.; Bale, S.; Crane, B. R.; Begley, T. P.;

Ealick, S. E. Biochemistry 2006, 46 (1), 145–155.

(469) Fujigaki, S.; Saito, K.; Sekikawa, K.; Tone, S.; Takikawa, O.; Fujii, H.; Wada, H.;

Noma, A.; Seishima, M. Eur. J. Immunol. 2001, 31 (8), 2313–2318.

(470) Lanz, T. V.; Opitz, C. A.; Ho, P. P.; Agrawal, A.; Lutz, C.; Weller, M.; Mellor, A.

L.; Steinman, L.; Wick, W.; Platten, M. Stem Cells Dev. 2010, 19 (5), 657–668.

(471) Pilotte, L.; Larrieu, P.; Stroobant, V.; Colau, D.; Dolusic, E.; Frederick, R.; De

Plaen, E.; Uyttenhove, C.; Wouters, J.; Masereel, B.; Van den Eynde, B. J. Proc.

Natl. Acad. Sci. 2012, 109 (7), 2497–2502.

(472) Muller, A. J.; Sharma, M. D.; Chandler, P. R.; DuHadaway, J. B.; Everhart, M. E.;

Johnson, B. A.; Kahler, D. J.; Pihkala, J.; Soler, A. P.; Munn, D. H.; Prendergast,

G. C.; Mellor, A. L. Proc. Natl. Acad. Sci. 2008, 105 (44), 17073–17078.

(473) Platten, M.; Wick, W.; Van den Eynde, B. J. Cancer Res. 2012, 72 (21), 5435–

5440.

161

(474) KNOX, W. E.; MEHLER, A. H. J. Biol. Chem. 1950, 187 (1), 419–430.

(475) Tatsumi, K.; Higuchi, T.; Fujiwara, H.; Nakayama, T.; Egawa, H.; Itoh, K.; Fujii,

S.; Fujita, J. Biochem. Biophys. Res. Commun. 2000, 274 (1), 166–170.

(476) Britan, A.; Maffre, V.; Tone, S.; Drevet, J. R. Cell Tissue Res. 2006, 324 (2), 301–

310.

(477) Haber, R.; Bessette, D.; Hulihan-Giblin, B.; Durcan, M. J.; Goldman, D. J.

Neurochem. 1993, 60 (3), 1159–1162.

(478) Kanai, M.; Funakoshi, H.; Takahashi, H.; Hayakawa, T.; Mizuno, S.; Matsumoto,

K.; Nakamura, T. Mol. Brain 2009, 2 (1), 8.

(479) Wu, W.; Nicolazzo, J. A.; Wen, L.; Chung, R.; Stankovic, R.; Bao, S. S.; Lim, C.

K.; Brew, B. J.; Cullen, K. M.; Guillemin, G. J. PLoS One 2013, 8 (4), e59749.

(480) Lanz, T. V; Williams, S. K.; Stojic, A.; Iwantscheff, S.; Sonner, J. K.; Grabitz, C.;

Becker, S.; Böhler, L.-I.; Mohapatra, S. R.; Sahm, F.; Küblbeck, G.; Nakamura, T.;

Funakoshi, H.; Opitz, C. A.; Wick, W.; Diem, R.; Platten, M. Sci. Rep. 2017, 7,

41271.

(481) Ishimura, Y.; Nozaki, M.; Hayaishi, O. J. Biol. Chem. 1970, 245 (14), 3593–3602.

(482) Batabyal, D.; Yeh, S.-R. J. Am. Chem. Soc. 2007, 129 (50), 15690–15701.

(483) Percudani, R.; Peracchi, A. EMBO Rep. 2003, 4 (9), 850–854.

(484) Ciorba, M. A. Am. J. Clin. Nutr. 2013, 98 (4), 863–864.

(485) Echevarrı́a Gorostidi, G. R.; Basagoitıa,́ A.; Santos Blanco, J. G.; Garcı́a Blanco,

F. J. Mol. Catal. A Chem. 1998, 129 (2), 173–177.

(486) Echevarrı́a, G. R.; Basagoitia, A.; Santos, J. G.; Garcı́a Blanco, F. J. Mol. Catal. A

Chem. 2000, 160 (2), 209–215.

162

(487) Kirsch, J. F.; Eichele, G.; Ford, G. C.; Vincent, M. G.; Jansonius, J. N.; Gehring,

H.; Christen, P. J. Mol. Biol. 1984, 174 (3), 497–525.

(488) Hay, N.; Sonenberg, N. Genes Dev. 2004, 18 (16), 1926–1945.

(489) Lipton, J. O.; Sahin, M. Neuron 2014, 84 (2), 275–291.

(490) Sarbassov, D. D.; Ali, S. M.; Kim, D.-H.; Guertin, D. A.; Latek, R. R.; Erdjument-

Bromage, H.; Tempst, P.; Sabatini, D. M.; Hahn, W. .; Sharp, P. .; al., et. Curr.

Biol. 2004, 14 (14), 1296–1302.

(491) Frias, M. A.; Thoreen, C. C.; Jaffe, J. D.; Schroder, W.; Sculley, T.; Carr, S. A.;

Sabatini, D. M.; Sabatini, D. M.; Firtel, R. A. Curr. Biol. 2006, 16 (18), 1865–

1870.

(492) Kim, D.-H.; Sarbassov, D. D.; Ali, S. M.; King, J. E.; Latek, R. R.; Erdjument-

Bromage, H.; Tempst, P.; Sabatini, D. M.; Boeke, J. D.; Bussey, H.; al., et. Cell

2002, 110 (2), 163–175.

(493) Peterson, T. R.; Laplante, M.; Thoreen, C. C.; Sancak, Y.; Kang, S. A.; Kuehl, W.

M.; Gray, N. S.; Sabatini, D. M.; Sabatini, D. M.; Gray, N. S.; al., et. Cell 2009,

137 (5), 873–886.

(494) Kovacina, K. S.; Park, G. Y.; Bae, S. S.; Guzzetta, A. W.; Schaefer, E.; Birnbaum,

M. J.; Roth, R. A. J. Biol. Chem. 2003, 278 (12), 10189–10194.

(495) Bar-Peled, L.; Sabatini, D. M.; Fox, C.; Healy, J.; Ferguson, B.; Lowery, R.;

McNally, R.; Lockwood, C.; Stout, J.; Kavazis, A. Trends Cell Biol. 2014, 24 (7),

400–406.

(496) Fang, Y.; Vilella-Bach, M.; Bachmann, R.; Flanigan, A.; Chen, J. Science (80-. ).

2001, 294 (5548).

163

(497) Bond, P. J. Int. Soc. Sports Nutr. 2016, 13 (1), 8.

(498) Hay, N.; Sonenberg, N. Genes Dev. 2004, 18 (16), 1926–1945.

(499) Mendoza, M. C.; Er, E. E.; Blenis, J.; al., et. Trends Biochem. Sci. 2011, 36 (6),

320–328.

(500) Inoki, K.; Ouyang, H.; Zhu, T.; Lindvall, C.; Wang, Y.; Zhang, X.; Yang, Q.;

Bennett, C.; Harada, Y.; Stankunas, K.; Wang, C.; He, X.; MacDougald, O. A.;

You, M.; Williams, B. O.; Guan, K.-L. Cell 2006, 126 (5), 955–968.

(501) Pearce, L. R.; Sommer, E. M.; Sakamoto, K.; Wullschleger, S.; Alessi, D. R.

Biochem. J. 2011, 436 (1), 169–179.

(502) Laplante, M.; Sabatini, D. M.; Efeyan, A.; Wang, S.; Sancak, Y.; Sabatini, D. M.;

Benzer, S.; Kapahi, P.; Cantley, L. C.; Gygi, S. P.; Blenis, J. Cell 2012, 149 (2),

274–293.

(503) Frias, M. A.; Thoreen, C. C.; Jaffe, J. D.; Schroder, W.; Sculley, T.; Carr, S. A.;

Sabatini, D. M.; Sabatini, D. M.; Firtel, R. A. Curr. Biol. 2006, 16 (18), 1865–

1870.

(504) Sarbassov, D. D.; Ali, S. M.; Kim, D.-H.; Guertin, D. A.; Latek, R. R.; Erdjument-

Bromage, H.; Tempst, P.; Sabatini, D. M.; Hahn, W. .; Sharp, P. .; al., et. Curr.

Biol. 2004, 14 (14), 1296–1302.

(505) Sarbassov, D. D.; Guertin, D. A.; Ali, S. M.; Sabatini, D. M. Science (80-. ). 2005,

307 (5712).

(506) VÉZINA, C.; KUDELSKI, A.; SEHGAL, S. N. J. Antibiot. (Tokyo). 1975, 28 (10),

721–726.

(507) Camardo, J. Transplant. Proc. 2003, 35 (3 Suppl), 18S–24S.

164

(508) Easton, J. B.; Houghton, P. J. Oncogene 2006, 25 (48), 6436–6446.

(509) Balestri, R.; Neri, I.; Patrizi, A.; Angileri, L.; Ricci, L.; Magnano, M. J. Eur. Acad.

Dermatology Venereol. 2015, 29 (1), 14–20.

(510) Spilman, P.; Podlutskaya, N.; Hart, M. J.; Debnath, J.; Gorostiza, O.; Bredesen, D.;

Richardson, A.; Strong, R.; Galvan, V. PLoS One 2010, 5 (4), e9979.

(511) Powers, R. W.; Kaeberlein, M.; Caldwell, S. D.; Kennedy, B. K.; Fields, S. Genes

Dev. 2006, 20 (2), 174–184.

(512) Harrison, D. E.; Strong, R.; Sharp, Z. D.; Nelson, J. F.; Astle, C. M.; Flurkey, K.;

Nadon, N. L.; Wilkinson, J. E.; Frenkel, K.; Carter, C. S.; Pahor, M.; Javors, M.

A.; Fernandez, E.; Miller, R. A. Nature 2009, 460 (7253), 392–395.

(513) Miller, R. A.; Harrison, D. E.; Astle, C. M.; Baur, J. A.; Boyd, A. R.; de Cabo, R.;

Fernandez, E.; Flurkey, K.; Javors, M. A.; Nelson, J. F.; Orihuela, C. J.; Pletcher,

S.; Sharp, Z. D.; Sinclair, D.; Starnes, J. W.; Wilkinson, J. E.; Nadon, N. L.;

Strong, R. Journals Gerontol. Ser. A Biol. Sci. Med. Sci. 2011, 66A (2), 191–201.

(514) Halloran, J.; Hussong, S. A.; Burbank, R.; Podlutskaya, N.; Fischer, K. E.; Sloane,

L. B.; Austad, S. N.; Strong, R.; Richardson, A.; Hart, M. J.; Galvan, V.

Neuroscience 2012, 223, 102–113.

(515) Sabatini, D. M.; Erdjument-Bromage, H.; Lui, M.; Tempst, P.; Snyder, S. H. Cell

1994, 78 (1), 35–43.

(516) Sabers, C. J.; Martin, M. M.; Brunn, G. J.; Williams, J. M.; Dumont, F. J.;

Wiederrecht, G.; Abraham, R. T. J. Biol. Chem. 1995, 270 (2), 815–822.

(517) Sarbassov, D. D.; Ali, S. M.; Sengupta, S.; Sheen, J.-H.; Hsu, P. P.; Bagley, A. F.;

Markhard, A. L.; Sabatini, D. M. Mol. Cell 2006, 22 (2), 159–168.

165

(518) Franklin, R. J. M. Nat. Rev. Neurosci. 2002, 3 (9), 705–714.

(519) Fancy, S. P. J.; Chan, J. R.; Baranzini, S. E.; Franklin, R. J. M.; Rowitch, D. H.

Annu. Rev. Neurosci. 2011, 34 (1), 21–43.

(520) Flores, A. I.; Narayanan, S. P.; Morse, E. N.; Shick, H. E.; Yin, X.; Kidd, G.;

Avila, R. L.; Kirschner, D. A.; Macklin, W. B. J. Neurosci. 2008, 28 (28), 7174–

7183.

(521) Goebbels, S.; Oltrogge, J. H.; Kemper, R.; Heilmann, I.; Bormuth, I.; Wolfer, S.;

Wichert, S. P.; Mobius, W.; Liu, X.; Lappe-Siefke, C.; Rossner, M. J.; Groszer,

M.; Suter, U.; Frahm, J.; Boretius, S.; Nave, K.-A. J. Neurosci. 2010, 30 (26),

8953–8964.

(522) Harrington, E. P.; Zhao, C.; Fancy, S. P. J.; Kaing, S.; Franklin, R. J. M.; Rowitch,

D. H. Ann. Neurol. 2010, 68 (5), 703–716.

(523) Bercury, K. K.; Dai, J.; Sachs, H. H.; Ahrendsen, J. T.; Wood, T. L.; Macklin, W.

B. J. Neurosci. 2014, 34 (13), 4466–4480.

(524) Wahl, S. E.; McLane, L. E.; Bercury, K. K.; Macklin, W. B.; Wood, T. L. J.

Neurosci. 2014, 34 (13), 4453–4465.

(525) Narayanan, S. P.; Flores, A. I.; Wang, F.; Macklin, W. B. J. Neurosci. 2009, 29

(21), 6860–6870.

(526) Tyler, W. A.; Jain, M. R.; Cifelli, S. E.; Li, Q.; Ku, L.; Feng, Y.; Li, H.; Wood, T.

L. Glia 2011, 59 (11), 1754–1769.

(527) Tyler, W. A.; Gangoli, N.; Gokina, P.; Kim, H. A.; Covey, M.; Levison, S. W.;

Wood, T. L. J. Neurosci. 2009, 29 (19), 6367–6378.

(528) Bai, C. B.; Sun, S.; Roholt, A.; Benson, E.; Edberg, D.; Medicetty, S.; Dutta, R.;

166

Kidd, G.; Macklin, W. B.; Trapp, B. Exp. Neurol. 2016, 283, 330–340.

(529) Kennedy, B. K.; Lamming, D. W. Cell Metab. 2016, 23 (6), 990–1003.

(530) Hay, N.; Sonenberg, N. Genes Dev. 2004, 18 (16), 1926–1945.

(531) Baldi, P.; Hatfield, G. W. 2002.

(532) Kerr, G.; Ruskin, H. J.; Crane, M.; Doolan, P. Comput. Biol. Med. 2008, 38 (3),

283–293.

(533) Tautenhahn, R.; Patti, G. J.; Kalisiak, E.; Miyamoto, T.; Schmidt, M.; Lo, F. Y.;

McBee, J.; Baliga, N. S.; Siuzdak, G. Anal. Chem. 2011, 83 (3), 696–700.

(534) Patti, G. J.; Tautenhahn, R.; Siuzdak, G. Nat. Protoc. 2012, 7 (3), 508–516.

(535) Taraboletti, A.; Walker, T.; Avila, R.; Huang, H.; Caporoso, J.; Manandhar, E.;

Leeper, T. C.; Modarelli, D. A.; Medicetty, S.; Shriver, L. P. Biochemistry 2017,

56 (10), 1518–1528.

(536) Kofman, A. E.; McGraw, M. R.; Payne, C. J. Aging (Albany. NY). 2012, 4 (4),

279–289.

(537) Gran, P.; Cameron-Smith, D. BMC Physiol. 2011, 11 (1), 10.

(538) Peng, T.; Golub, T. R.; Sabatini, D. M. Mol. Cell. Biol. 2002, 22 (15), 5575–5584.

(539) Zhang, J.; Underwood, L. E.; D’Ercole, A. J. FASEB J. 2001, 15 (7), 1261–1263.

(540) Costeas, P. A.; Chinsky, J. M. Biochem. J. 1996, 85–92.

(541) Casals, N.; Roca, N.; Guerrero, M.; Gil-Gómez, G.; Ayté, J.; Ciudad, C. J.;

Hegardt, F. G. Biochem. J. 1992, 261–264.

(542) Horton, J. D.; Bashmakov, Y.; Shimomura, I.; Shimano, H. Proc. Natl. Acad. Sci.

U. S. A. 1998, 95 (11), 5987–5992.

(543) Soukas, A.; Cohen, P.; Socci, N. D.; Friedman, J. M. Genes Dev. 2000, 14 (8),

167

963–980.

(544) Weindruch, R.; Prolla, T. A.; Lee, C.-K. Nat. Genet. 2000, 25 (3), 294–297.

(545) Sweatt, A. J.; Garcia-Espinosa, M. A.; Wallin, R.; Hutson, S. M. J. Comp. Neurol.

2004, 477 (4), 360–370.

(546) Cole, J. T.; Sweatt, A. J.; Hutson, S. M. Front. Neuroanat. 2012, 6, 18.

(547) Hull, J.; Hindy, M. El; Kehoe, P. G.; Chalmers, K.; Love, S.; Conway, M. E. J.

Neurochem. 2012, 123 (6), 997–1009.

(548) Bixel, M. G.; Hutson, S. M.; Hamprecht, B. J. Histochem. Cytochem. 1997, 45 (5),

685–694.

(549) Bixel, M. G.; Shimomura, Y.; Hutson, S. M.; Hamprecht, B. J. Histochem.

Cytochem. 2001, 49 (3), 407–418.

(550) Cole, J. T. In Branched Chain Amino Acids in Clinical Nutrition; Springer New

York: New York, NY, 2015; pp 13–24.

(551) Green, C. R.; Wallace, M.; Divakaruni, A. S.; Phillips, S. A.; Murphy, A. N.;

Ciaraldi, T. P.; Metallo, C. M. Nat Chem Biol 2016, 12 (1), 15–21.

(552) TOMIYA, T.; OMATA, M.; FUJIWARA, K. Hepatol. Res. 2004, 30, 14–18.

(553) Nishitani, S.; Horie, M.; Ishizaki, S.; Yano, H.; Noguchi, N.; Munsell, M. PLoS

One 2013, 8 (11), e82346.

(554) Zhen, H.; Kitaura, Y.; Kadota, Y.; Ishikawa, T.; Kondo, Y.; Xu, M.; Morishita, Y.;

Ota, M.; Ito, T.; Shimomura, Y. FEBS Open Bio 2016, 6 (1), 43–49.

(555) Carroll, B.; Maetzel, D.; Maddocks, O. D.; Otten, G.; Ratcliff, M.; Smith, G. R.;

Dunlop, E. A.; Passos, J. F.; Davies, O. R.; Jaenisch, R.; Tee, A. R.; Sarkar, S.;

Korolchuk, V. I. Elife 2016, 5.

168

(556) Jewell, J. L.; Kim, Y. C.; Russell, R. C.; Yu, F.-X.; Park, H. W.; Plouffe, S. W.;

Tagliabracci, V. S.; Guan, K.-L. Science (80-. ). 2015, 347 (6218), 194–198.

(557) Fan, S.-J.; Snell, C.; Turley, H.; Li, J.-L.; McCormick, R.; Perera, S. M. W.;

Heublein, S.; Kazi, S.; Azad, A.; Wilson, C.; Harris, A. L.; Goberdhan, D. C. I.

Oncogene 2016, 35 (23), 3004–3015.

(558) Wang, S.; Tsun, Z.-Y.; Wolfson, R. L.; Shen, K.; Wyant, G. A.; Plovanich, M. E.;

Yuan, E. D.; Jones, T. D.; Chantranupong, L.; Comb, W.; Wang, T.; Bar-Peled, L.;

Zoncu, R.; Straub, C.; Kim, C.; Park, J.; Sabatini, B. L.; Sabatini, D. M. Science

(80-. ). 2015, 347 (6218), 188–194.

(559) Hovevey-Sion, D.; Kopin, I. J.; Stull, R. W.; Goldstein, D. S. Neuropharmacology

1989, 28 (8), 791–797.

(560) Xu, H.; Yang, H.-J.; Zhang, Y.; Clough, R.; Browning, R.; Li, X.-M. Behav.

Neurosci. 2009, 123 (2), 418–429.

(561) Manegold, C.; Hoffmann, G. F.; Degen, I.; Ikonomidou, H.; Knust, A.; Laaß, M.

W.; Pritsch, M.; Wilichowski, E.; Hörster, F. J. Inherit. Metab. Dis. 2009, 32 (3),

371–380.

(562) Hyland, K.; Surtees, R. A.; Rodeck, C.; Clayton, P. T. Neurology 1992, 42 (10),

1980–1988.

(563) Hyland, K.; Clayton, P. T. Clin. Chem. 1992, 38 (12), 2405–2410.

(564) Mills, P. B.; Surtees, R. A. H.; Champion, M. P.; Beesley, C. E.; Dalton, N.;

Scambler, P. J.; Heales, S. J. R.; Briddon, A.; Scheimberg, I.; Hoffmann, G. F.;

Zschocke, J.; Clayton, P. T. Hum. Mol. Genet. 2005, 14 (8), 1077–1086.

(565) Goyal, M.; Fequiere, P. R.; McGrath, T. M.; Hyland, K. Pediatr. Neurol. 2013, 48

169

(3), 227–231.

(566) Aronova, S.; Wedaman, K.; Aronov, P. A.; Fontes, K.; Ramos, K.; Hammock, B.

D.; Powers, T. Cell Metab. 2008, 7 (2), 148–158.

(567) Hebbar, S.; Sahoo, I.; Matysik, A.; Argudo Garcia, I.; Osborne, K. A.; Papan, C.;

Torta, F.; Narayanaswamy, P.; Fun, X. H.; Wenk, M. R.; Shevchenko, A.;

Schwudke, D.; Kraut, R. Sci. Rep. 2015, 5 (1), 15926.

(568) Liu, G.; Han, F.; Yang, Y.; Xie, Y.; Jiang, H.; Mao, Y.; Wang, H.; Wang, M.;

Chen, R.; Yang, J.; Chen, J. Nephrol. Dial. Transplant. 2011, 26 (5), 1493–1502.

(569) Balboa, M. A.; Firestein, B. L.; Godson, C.; Bell, K. S.; Insel, P. A. 1994, 269

(14), 10511–10516.

(570) Hu, T.; Exton, J. H. Biochem. Biophys. Res. Commun. 2005, 333 (3), 750–753.

(571) Gibellini, F.; Smith, T. K. IUBMB Life 2010, 62 (6), n/a-n/a.

(572) McLaurin, J.; Trudel, G. C.; Shaw, I. T.; Antel, J. P.; Cashman, N. R. J. Neurobiol.

1995, 26 (2), 283–293.

(573) Buntinx, M.; Vanderlocht, J.; Hellings, N.; Vandenabeele, F.; Lambrichts, I.; Raus,

J.; Ameloot, M.; Stinissen, P.; Steels, P. J. Neurocytol. 2003, 32 (1), 25–38.

(574) Boscia, F.; D’Avanzo, C.; Pannaccione, A.; Secondo, A.; Casamassa, A.;

Formisano, L.; Guida, N.; Annunziato, L.; Herchuelz, A.; Annunziato, L. Cell

Death Differ. 2012, 19 (4), 562–572.

(575) Fan, Q.-W.; Cheng, C.; Knight, Z. A.; Haas-Kogan, D.; Stokoe, D.; James, C. D.;

McCormick, F.; Shokat, K. M.; Weiss, W. A. Sci. Signal. 2009, 2 (55), ra4-ra4.

(576) Aeder, S. E.; Martin, P. M.; Soh, J.-W.; Hussaini, I. M. Oncogene 2004, 23 (56),

9062–9069.

170

(577) Sirabella, R.; Secondo, A.; Pannaccione, A.; Molinaro, P.; Formisano, L.; Guida,

N.; Di Renzo, G.; Annunziato, L.; Cataldi, M. J. Neurochem. 2012, 122 (5), 911–

922.

(578) Nygård, M.; Wahlström, G. M.; Gustafsson, M. V.; Tokumoto, Y. M.; Bondesson,

M. Mol. Endocrinol. 2003, 17 (1), 79–92.

(579) Huang, Z.; Tang, X.-M.; Cambi, F. Mol. Cell. Neurosci. 2002, 19 (2), 250–262.

(580) Ikeda, M.-A.; Jakoi, L.; Nevins, J. R. Biochemistry 1996, 93, 3215–3220.

(581) ZHANG, W.; LIU, H. T. Cell Res. 2002, 12 (1), 9–18.

(582) Mukherjee, P.; Winter, S. L.; Alexandrow, M. G. Mol. Cell. Biol. 2010, 30 (3),

845–856.

(583) Luo, J.-H. World J. Clin. Oncol. 2011, 2 (2), 120–124.

(584) Wood, T. L.; Bercury, K. K.; Cifelli, S. E.; Mursch, L. E.; Min, J.; Dai, J.;

Macklin, W. B. ASN Neuro 2013, 5 (1), AN20120092.

(585) Kim, J. E.; Chen, J. Diabetes 2004, 53 (11), 2748–2756.

(586) Peterson, T. R.; Sengupta, S. S.; Harris, T. E.; Carmack, A. E.; Kang, S. A.;

Balderas, E.; Guertin, D. A.; Madden, K. L.; Carpenter, A. E.; Finck, B. N.;

Sabatini, D. M. Cell 2011, 146 (3), 408–420.

(587) Demetriades, C.; Doumpas, N.; Teleman, A. A. Cell 2014, 156 (4), 786–799.

(588) Lin, W.; Popko, B. Nat. Neurosci. 2009, 12 (4), 379–385.

(589) Schröder, M. Cell. Mol. Life Sci. 2008, 65 (6), 862–894.

(590) Ma, Y.; Hendershot, L. M. Cell 2001, 107 (7), 827–830.

(591) Ron, D.; Walter, P. Nat. Rev. Mol. Cell Biol. 2007, 8 (7), 519–529.

(592) Mháille, A. N.; Mcquaid, S.; Windebank, A.; Cunnea, P.; Mcmahon, J.; Samali,

171

A.; Fitzgerald, U. .

(593) McDonald, L.; Liu, B.; Taraboletti, A.; Whiddon, K.; Shriver, L. P.; Konopka, M.;

Liu, Q.; Pang, Y. J. Mater. Chem. B 2016, 4 (48), 7902–7908.

(594) Fukuwatari, T.; Shibata, K. Int. J. Tryptophan Res. 2013, 3.

(595) Frost, R. A.; Lang, C. H. Physiology 2011, 26 (2).

(596) Kochevenko, A.; Fernie, A. R. J. Exp. Bot. 2011, 62 (11), 3895–3906.

(597) Tunbridge, E. M.; Harrison, P. J.; Weinberger, D. R. Biol. Psychiatry 2006, 60 (2),

141–151.

172

APPENDIX A: THESIS SUPPLEMENTARY

A1. CPZ is taken up into cells causing cell death

Additional information is provided documenting the cytotoxic, and oligodendrocyte-specific nature of CPZ. An MTT metabolic activity assay and a

LIVE/DEAD viability assay were used in tandem to show reduced cellular activity and viability after at least 24 hours with 1 mM CPZ. To prove CPZ was indeed taken up into cells, we performed and SRM-MS experiment. The supplemental calibration curve for the SRM is provided.

A1.1 Methods

A1.1.1 MTT Assay

MTT assay, 20 µL of 5 mg/mL MTT was added to each well after the respective time point. Each plate was then incubated for 3.5 hours at 37°C, lysis buffer was added (4 mM HCl, 0.1% NP40 buffer, in isopropanol), and the absorbance was read at 590 nm with a reference filter of 620 nm by using a Spectramax M2 plate reader (Molecular

173

Devices, Sunnyvale, CA, USA). The percent metabolic activity was calculated using the vehicle control set as 100% metabolic activity.

A1.1.2 Cell Imaging

MO3.13 cells were seeded onto were plated on microscopy dishes at a density of 5 × 105 cells per mL in DMEM media supplemented with 10% FBS and 1% penicillin/streptomycin and left to attach for 24 hours. After incubation, the media was aspirated and replaced with DMEM with 0.3% ethanol containing 1 mM CPZ or DMEM with 0.3% ethanol vehicle. Following 12 hours of CPZ treatment, the media was removed and the cells were washed and then plated with Live Cell Imaging Solution

(ThermoFisher A14291DJ), along with 2 µM calcein AM and 4 µM EthD-1. Cells were then left to incubate at room temperature for 30 minutes before imaging.

Imaging was performed on a Nikon A1 confocal system with a 100× Plan Apo λ,

NA = 1.45 oil objective. Calcein AM is excited using 495 nm with an emission at 515, and EthD-1 was excited at 528 nm with an emission at 617. All imaging was done in an

Okolab Bold Cage Incubator at 37 °C, and images were processed using NIS Elements or

ImageJ Pro imaging software.

174

Figure A1: Oligodendrocyte cells show changes in metabolic activity after cuprizone treatment while a rat astrocyte cell line does not. (a) An MTT assay was used to measure the metabolic activity of the MO3.13 cells treated with increasing concentrations of CPZ at 6, 18, and 24 hours. Concentrations ranged from 0.125 µM to 1 mM (black to white) (n=12 cultures per concentration per time point). Percent activity was compared to vehicle-treated cells. Live cell fluorescence microscopy of MO3.13 cells (100X) treated with vehicle (b) or 1mM CPZ (c) and stained with 2 µM calcein AM (green) and 4 µM EthD-1 (red) after 12 hours of treatment. (d) Metabolic Activity measured by MTT assay in a rat astrocyte cell line (DI TNC1) treated with increasing concentrations of CPZ at 6, 18, and 24 hours. Concentrations ranged from 0.125 µM to 1 mM (black to white) (n=12 cultures per concentration per time point).

175

Figure A2: Determination of CPZ uptake in cells by mass spectrometry. (a) MS/MS data for unbound CPZ (* parent ion) (b) The standard curve for CPZ concentrations from 1 to 100 nM. Levels in cells are shown in red. The transition m/z 279.18→ m/z 139.09 was used to determine the concentration in cells.

176

A2. MS/MS proves metabolic disruption in vitro and in vivo from CPZ

The MS/MS data is provided for all metabolites identified as dysregulated in either MO3.13 cells or tissue after CPZ intoxication. MS/MS fragmentation was used to positively identify putatively identified features. All metabolites are compared to a standard provided by either the HMDB or Metlin metabolite library.

177

Figure A3: MS/MS spectra of identified metabolites from MO3.13 cells. The best IDA product spectrum is chosen from all aligned cell data. Each metabolite spectrum is shown in red with the corresponding standard spectra shown above in black. All standard spectra are referenced at CE 20-30 (v) for positive mode. 178

Figure A4: MS/MS spectra of identified metabolites from MO3.13 cells. The best IDA product spectrum is chosen from all aligned cell data. Each metabolite spectrum is shown in red with the corresponding standard spectra shown above in black. All standard spectra are referenced at CE 20-30 (v) for positive mode.

179

Figure A5: MS/MS spectra of identified metabolites from mouse tissue. The best IDA product spectrum is chosen from all aligned tissue data. Each metabolite spectrum is shown in red with the corresponding standard spectra shown above in black. All standard spectra are referenced at CE 20-30 (v) for positive mode.

180

Figure A6: MS/MS spectra of identified metabolites from mouse tissue. The best IDA product spectrum is chosen from all aligned tissue data. Each metabolite spectrum is shown in red with the corresponding standard spectra shown above in black. All standard spectra are referenced at CE 20-30 (v) for positive mode.

181

Figure A7: MS/MS spectra of identified metabolites from mouse tissue. The best IDA product spectrum is chosen from all aligned tissue data. Each metabolite spectrum is shown in red with the corresponding standard spectra shown above in black. All standard spectra are referenced at CE 20-30 (v) for positive mode.

182

Figure A8: MS/MS spectra of identified metabolites from mouse tissue. The best IDA product spectrum is chosen from all aligned tissue data. Each metabolite spectrum is shown in red with the corresponding standard spectra shown above in black. All standard spectra are referenced at CE 20-30 (v) for positive mode.

183

Figure A9: MS/MS spectra of identified metabolites from mouse tissue. The best IDA product spectrum is chosen from all aligned tissue data. Each metabolite spectrum is shown in red with the corresponding standard spectra shown above in black. All standard spectra are referenced at CE 20-30 (v) for positive mode.

184

A3. TDO and P5P enzymes are targets explored for CPZ toxicity

Supplemental data is provided detailing our initial investigation of TDO inhibition by CPZ, and the latter investigation of CPZ binding P5P. As tryptophan metabolism is affected by CPZ, nicotinamide was first supplemented into cell cultures with CPZ in an attempt to rescue viability. With promising initial results, rHTDO was expressed and purified using information from the provided ProtPram. P5P was found to bind hydrolyzed CPZ and potentially affect enzymes utilizing it as a cofactor. The actual binding event further was characterized using LC separation and in silico prediction patterns. We proposed CPZ to bind P5P via a Schiff base bond and synthesized a standard to compare to our sample for verification. The NMR information from the standard is provided.

A3.1 Methods

A3.1.1 MTT assay

MO3.13 cells were seeded in 96-well plates at a density of 73,000 cells/mL. The last row of each plate contained 100 µL of tissue culture medium with no cells to act as a blank control. The cells were left to attach for 24 hours followed by treatment with: a vehicle of DMEM with 0.3% ethanol, 0.3% ethanol containing 1 mM CPZ, or 0.3% ethanol containing 1 mM CPZ with 0.25-5 mM nicotinamide. The assay was performed as previously described, and the percent metabolic activity was calculated using the vehicle control set as 100% metabolic activity.

185

Figure A10: An MTT assay was used to measure the metabolic activity of the MO3.13 cells after of 24 hours of treatment with increasing concentrations of nicotinamide and CPZ. Concentrations ranged from 0 mM to 5 mM (black to light grey) (n=12 cultures per concentration). Percent activity was compared to vehicle-treated cells (white).

186

Figure A11: TDO ProtParam User-provided sequence: 10 20 30 40 50 60 HHHHHHMSGC PFLGNNFGYT FKKLPVEGSE EDKSQTGVNR ASKGGLIYGN YLHLEKVLNA

70 80 90 100 110 120 QELQSETKGN KIHDEHLFII THQAYELWFK QILWELDSVR EIFQNGHVRD ERNMLKVVSR

130 140 150 160 170 180 MHRVSVILKL LVQQFSILET MTALDFNDFR EYLSPASGFQ SLQFRLLENK IGVLQNMRVP

190 200 210 220 230 240 YNRRHYRDNF KGEENELLLK SEQEKTLLEL VEAWLERTPG LEPHGFNFWG KLEKNITRGL

250 260 270 280 290 300 EEEFIRIQAK EESEEKEEQV AEFQKQKEVL LSLFDEKRHE HLLSKGERRL SYRALQGALM

310 320 330 340 350 360 IYFYREEPRF QVPFQLLTSL MDIDSLMTKW RYNHVCMVHR MLGSKAGTGG SSGYHYLRST

370 380 390 400 410 VSDRYKVFVD LFNLSTYLIP RHWIPKMNPT IHKFLYTAEY CDSSYFSSDE SD

Number of amino acids: 412 Gly (G) 24 5.8% His (H) 21 5.1% Molecular weight: 48694.3 Ile (I) 17 4.1% Leu (L) 50 12.1% Theoretical pI: 6.63 Lys (K) 27 6.6% Met (M) 11 2.7% Phe (F) 24 5.8% Amino acid composition: Pro (P) 11 2.7% Ala (A) 12 2.9% Ser (S) 30 7.3% Arg (R) 26 6.3% Thr (T) 16 3.9% Asn (N) 19 4.6% Trp (W) 6 1.5% Asp (D) 15 3.6% Tyr (Y) 18 4.4% Cys (C) 3 0.7% Val (V) 21 5.1% Gln (Q) 19 4.6% Pyl (O) 0 0.0% Glu (E) 42 10.2% Sec (U) 0 0.0%

Total number of negatively charged residues (Asp + Glu): 57 Total number of positively charged residues (Arg + Lys): 53

Formula: C2193H3369N603O629S14 Total number of atoms: 6808

Extinction coefficients:

Extinction coefficients are in units of M-1cm-1, at 280 nm measured in water. Ext. coefficient 59945 Abs 0.1% (=1 g/l) 1.231, assuming all pairs of Cys residues form cystines Ext. coefficient 59820 Abs 0.1% (=1 g/l) 1.228, assuming all Cys residues are reduced

187

Figure A12: Chemical interaction of Pyridoxal 5’ Phosphate and CPZ. a) In silico fragmentation data for the proposed Schiff-base complex of P5P and CPZ. Note that many more peaks are predicted to occur using fragment-predicting algorithms than are seen with the synthesized standard. b) Total Ion Chromatograms from solutions containing CPZ (black), P5P (green), P5P with the addition of CPZ (blue), and the synthesized standard for the putative Schiff base (red).

188

Figure A13: Preparation of Schiff base from Pyridoxal-5-phosphate and oxalydihydrazide. 1H NMR (500 MHz, D2O with sodium bicarbonate).

189

A4. Transcriptomics data after PMA-induced MO3.13 differentiation

Full information including the significance and fold change values have been included for all dysregulated genes mentioned in Chapter IV. Genes are organized by their respective pathway.

Table A1:

Pathway Gene Fold Change ANOVA p-value FDR p-value Description Symbol PMA to Control Cell Fate SOX6 -5.19 0.000076 0.001011 SRY box 6 SOX11 3.34 0.000081 0.001042 SRY box 11 SOX6 -3.08 0.000093 0.001134 SRY box 6 SOX4 -4.87 1.83E-08 0.000053 SRY box 4 SOX4 -4.44 0.000001 0.000145 SRY box 4 SOX11 4.75 0.000001 0.000145 SRY box 11 SOX6 -3.16 0.000001 0.000147 SRY box 6 PKCα PRKCA -2.14 0.000346 0.002602 protein kinase C, alpha Calcium CALM1; 4.96 1.67E-07 0.000075 calmodulin 1 Signaling CALM2 (phosphorylase kinase, delta); calmodulin 2 (phosphorylase kinase, delta) SLC8A3 2.31 0.000012 0.000378 solute carrier family 8 (sodium/calcium exchanger), member 3 E2F/Rb E2F1 -2.29 0.000096 0.001163 E2F transcription factor 1 CDK1 -32.44 0.000002 0.000154 cyclin-dependent kinase 1 CDK2 -2.63 0.00004 0.000707 cyclin-dependent kinase 2 CCNA2 -28.83 0.000001 0.000145 cyclin A2 CCNB2 -14.62 7.29E-07 0.000118 cyclin B2 CCNE2 -2.9 0.00008 0.00103 cyclin E2 CDKN1B 4.03 0.000574 0.003658 cyclin-dependent kinase inhibitor 1B (p27, Kip1) MCM7 -4.63 6.58E-07 0.000113 minichromosome maintenance complex component 7

190

Cell CHEK1 -5.61 0.000001 0.000146 checkpoint kinase 1 Checkpoints RFC3 -4.04 0.000008 0.000309 replication factor C subunit 3 RFC4 -4.59 0.000007 0.000284 replication factor C subunit 4 RFC5 -4.24 0.000017 0.000458 replication factor C subunit 5 WEE1 -4.79 0.000005 0.00026 WEE1 G2 checkpoint kinase ORC1 -4.11 0.000073 0.00099 origin recognition complex subunit 1 POLE2 -15.46 0.000017 0.000449 polymerase (DNA directed), epsilon 2, accessory subunit CDC45 -7.77 0.000001 0.000146 cell division cycle 45 Mitotic cell RRM2 -18.95 0.000008 0.000319 ribonucleotide cycle and reductase M2 spindle DHFR -8.42 0.000051 0.000801 dihydrofolate organization reductase TYMS -11.12 2.34E-07 0.000082 thymidylate synthetase TTK -17.91 0.000003 0.000199 TTK protein kinase STMN1 -2.89 0.000006 0.000269 stathmin 1 STMN1 -3.16 0.000009 0.00033 stathmin 1 PRIM1 -14.88 7.86E-08 0.000066 primase, DNA, polypeptide 1 (49kDa) POLA1 -2.92 0.000006 0.000272 polymerase (DNA directed), alpha 1, catalytic subunit PCNA -3.45 0.000004 0.000222 proliferating cell nuclear antigen DNA repair SMC2 -8.61 0.000005 0.000261 structural and maintenance of replication chromosomes 2 BARD1 -4.22 0.000003 0.000199 BRCA1 associated RING domain 1 RRM1 -4.41 0.000006 0.000279 ribonucleotide reductase M1 TOP2A -46.67 0.00001 0.00034 topoisomerase (DNA) II alpha TOP2A -50.28 6.67E-07 0.000113 topoisomerase (DNA) II alpha CDT1 -3.14 0.000004 0.000226 chromatin licensing and DNA replication factor 1 ANLN -6.41 0.000009 0.000333 anillin actin binding protein HMGB1 -4.55 0.000057 0.000854 high mobility group box 1

191

H2AFZ -2.38 0.000001 0.000137 H2A histone family, member Z POLD3 -2.28 0.000281 0.002274 polymerase (DNA- directed), delta 3, accessory subunit DCK -3.42 0.000487 0.003266 deoxycytidine kinase KIF4A; -36.52 9.06E-07 0.000127 kinesin family KIF4B member 4A; kinesin family member 4B KIF4A -21.43 0.000003 0.0002 kinesin family member 4A Lipid LPIN1 2.36 0.000035 0.000654 lipin 1 Synthesis HMGCR 2.02 0.001352 0.006622 3-hydroxy-3- methylglutaryl-CoA reductase SGMS1 2.16 0.004161 0.015221 sphingomyelin synthase 1 SMPD1 2.25 0.000036 0.000662 sphingomyelin phosphodiesterase 1, acid lysosomal

DEGS1 2.12 0.000252 0.002118 delta(4)-desaturase, sphingolipid 1 UGCG 2.32 7.16E-08 0.000065 UDP-glucose ceramide glucosyltransferase ACER3 2.15 0.000199 0.001817 alkaline ceramidase 3 MVK 2.3 0.000179 0.001701 mevalonate kinase MVD 2.02 0.004129 0.015134 mevalonate (diphospho) decarboxylase LSS 2.23 2.23E-07 0.000081 lanosterol synthase (2,3-oxidosqualene- lanosterol cyclase)

DHCR7 2.46 4.63E-07 0.000099 7-dehydrocholesterol reductase SGK1 2.65 0.000028 0.000574 serum/glucocorticoid regulated kinase 1 BCAA ACAA1 2.02 0.000022 0.000506 acetyl-CoA Metabolism acyltransferase 1 BCAT1 -2.01 0.000322 0.00248 branched chain amino-acid transaminase 1, cytosolic IVD -2.05 0.000016 0.000442 isovaleryl-CoA dehydrogenase HADH -4.29 0.000001 0.000148 hydroxyacyl-CoA dehydrogenase

192

APPENDIX B: FLUORESCENT FLAVONOIDS FOR ER CELL IMAGING

B1. Introduction

The endoplasmic reticulum (ER) serves as the site for the synthesis and exportation of proteins and membrane lipids. Protein synthesis is initiated on the cytosolic surface of the ER, translocating through a pore into the ER lumen. Tight control of protein maturation and folding ensures only correctly folded proteins are transported out of the ER, safeguarding cellular homeostasis. Perturbations of the cellular state, such as alterations to the ER oxidizing environment, protein folding capacity, glycosylation, or calcium stores, can result in a loss of ER homeostasis causing protein retention and ultimately initiating a cascade of pathways known as the unfolded protein response or ER stress.588–591 The ER stress response has been associated with oligodendrocyte perikaryon and myelin sheath degeneration found in MS and following the CPZ intoxication model.199,592 Herein we show work resulting in the development of a non-toxic flavonoid- based fluorophore useful for ER localization in live eukaryotic cells.593 As the effects of

ER stress in oligodendrocytes is of interest in relation to our work on the CPZ model of demyelination, we evaluated the novel fluorophore in MO3.13 cells. Additional glial cells such as GL261 cells (astrocyte-derived glioblastoma), and SIM-A9 cells (semi-adherent

193 microglial cells), were also monitored to confirm reproducibility. Text and figures were reproduced593 with permission from the Royal Society of Chemistry.

B2. Methods

B2.1. Cell Viability and IC50

MO3.13 cells were seeded into three separate 96-well plates at a density of 73,000 cells/mL. The last row of each plate contained 100 µL of tissue culture medium with no cells to act as a blank control. The cells were left to attach for 24 hours followed by treatment with: 0.0005 to 1 µM 2a or 2b in DMSO, or a vehicle of DMSO. The plates were subsequently incubated for 24, 48, or 72 hours. The MTT assay was performed and the absorbance was read at 590 nm with a reference filter of 620 nm by using a

Spectramax M2 plate reader. MTT calculations were performed using Graphpad Prism 5 software. All absorbance values were translated to percent viability based on the mean

(µ) vehicle treated absorbance:

퐴퐵푆퐸푥푝푒푟𝑖푚푒푛푡 ×100% µ(퐴퐵푆푉푒ℎ𝑖푐푙푒)

Using the normalized cell percent viability data, the IC50 was then calculated by fitting the log of the concentration (µM) to the percent viability using a dose-response curve.

Specifically, a nonlinear regression was used to fit the data to a log(inhibitor) vs. response (variable slope) curve:

(푇표푝 − 퐵표푡푡표푚) 푦 = 퐵표푡푡표푚 + (1 + 10(푙표𝑔퐼퐶50−푥)∗퐻𝑖푙푙푠푙표푝푒))

HillSlope: describes the steepness of the family of curves.

Top and Bottom: are plateaus in the units of the Y axis.

194

IC50: is the concentration of agonist that gives a response halfway between Bottom and

Top.

B2.2. Cell culturing and staining

MO3.13 cells (oligodendrocytes), GL261 cells (astrocyte-derived glioblastoma) or SIM A9 cells (semi-adherent microglia) were plated on microscopy dishes at a density of 5 × 105 cells per mL in DMEM media supplemented with 10% FBS and 1% penicillin/streptomycin. Cells were allowed to attach overnight at 37 °C and in 5% CO2.

After cells attached, ER-Tracker™ Red (Sigma E34250) was added to the cell media at a final concentration of 1 μM along with a final concentration at 500 nM of the flavonoid dye to be tested. Cells were left to incubate for 2 hours and then washed 5 times with

PBS. Live Cell Imaging Solution (ThermoFisher A14291DJ) was then added, along with

10 μM of DRAQ5 (Sigma 62251) fluorescent probe, to stain cell nuclei. Cells were then left to incubate at room temperature for 30 minutes before imaging.

B2.3. Live imaging

Imaging was performed on a Nikon A1 confocal system with a 100× Plan Apo λ,

NA = 1.45 oil objective with both GaAsP detectors and high sensitivity low noise PMTs for detection. The excitation used for our dyes was 405 nm with standard DAPI, FITC, and Texas Red filters. The ER-Tracker™ was excited using 560 nm and with a 600/50 nm bandpass filter used for emission. The nuclear dye, DRAQ5, was excited at 620 nm and 680/75 nm bandpass filter was used for emission. All imaging was done in an Okolab

Bold Cage Incubator at 37 °C, and images were processed using NIS Elements or ImageJ

Pro imaging software.

195

B3 Results

B3.1 Synthesis of Flavonoids

The new flavonoids 2a–2b include a donor–acceptor (D–A) interaction, which retains the desirable fluorescent solvatochromic property. Flavonoids 2 were synthesized by reaction of N-(3-acetyl-4-hydroxyphenyl)butyramide 3 with 4-

(dialkylamino)benzaldehyde 4 in two steps (Scheme B1), i.e. by Claisen–Schmidt condensation followed by Algar–Flynn–Oyamada reaction.23,24 The crude product was purified by recrystallization from hexane/ethanol mixture in good yield.593

Scheme B1: Chemical structures of flavonoids 1–2 and synthesis of 2. Fluorophores’ 2a and 2b were further investigated due to low toxicity.

B3.2 Cytotoxicity of flavonoids

Cytotoxicity of flavonoids was examined by using an MTT assay, showing that the half maximal inhibitory concentration (IC50) for 2a–2b was 18.59 and 33.54 µM, respectively (Figure B1). A kinetic assay additionally showed no further acquired cytotoxic 196 effects when observed from 24-72 hours (Figure B2). Low cytotoxicity of 2 encouraged us to further examine the staining of the flavonoids in MO3.13 cells. When the cells were incubated with 2 and imaged by confocal microscopy, the dye was found to quickly penetrate cells, giving strong fluorescence (Figure B3). Weak fluorescence could be observed immediately after adding dyes into cells and mounting the sample onto microscope (less than 1 minute).

Since the dye was relatively non-fluorescent in an aqueous environment, the observed fluorescence was assumed to arise from its interaction with cellular lipids or hydrophobic pockets of proteins.

B3.3 Colocalization of 2 and ER-Tracker in oligodendrocytes

ER-Tracker™ Red (concentration 1 μM) along with flavonoid 2 (concentration

500 nM) were added to the MO3.13 cell media. Fluorescent probe DRAQ5 (10 μM) was also added to the tissue culture media for staining cell nuclei. The fluorescence confocal imaging (Figure B4) revealed that flavonoid 2 stained the cell components surrounding the nuclei. The sharp circles in the fluorescence image of 2 (Figure B3) were the outer contour of the nuclei since they are not stained and non-fluorescent. The non-uniform stain of 2 on the cells suggests that the dyes might be selectively binding to intracellular organelles.

In order to confirm the intracellular location and distribution of 2, we examined the colocalization of our dye with the commercial ER probe, ER-Tracker™ Red. Incubation of

MO3.13 cells with ER–Tracker™ Red showed robust fluorescence in structures surrounding the nucleus, a pattern consistent with ER staining (Figure B4 A,D). Staining with both compounds 2a and 2b showed similar localization (Figure B4 E,B) and colocalization was observed when cells were incubated with both the commercial ER dye and our probes (Figure

197

B4 C,F). The stained cells were scanned repeated at 5-minute interval for 3 hours which revealed good stability in comparison with commercial ER-Tracker (Figure B5).

It is well known that ER forms an interconnected network of membrane-enclosed sacs which are embedded with transmembrane proteins that contain hydrophobic structural motifs.

It is attractive to speculate that the flavonoid dyes were binding to these types of hydrophobic structures. Such ER binding-induced fluorescence turn-on would be valuable for dynamic tracking since 2 was nearly non-fluorescent in the aqueous environment. The assumption was confirmed by the observation of clear cell contours without post-staining washing (Figure B3), as the dyes outside the cells did not give observable fluorescence. A direct comparison also revealed that flavonoid 2b gave stronger green fluorescence than 2a under the same experimental conditions (Figure B4 B ,E), illustrating that the ER binding-induced fluorescence turn on was also sensitive to substituents on the flavonoid.

In order to verify the ER selectivity, flavonoid 2b was further examined by staining in

GL261 cells, its growth is known to respond to ER stress.28 Colocalization of 2b with the commercial ER-Tracker™ Red revealed the identical patterns (Figure B6 A,B), showing that the flavonoid was indeed selective to ER. The ER selectivity was also examined on SIM-A9 cells.29 As a type of glial cell located throughout the brain, microglial cells account for 10–

15% of all cells found within the brain,30 and ER stress plays a significant role in the microglial cell death.31 Although the ER distribution in GL261 cells is quite different from that in SIM-A9 cells (Fig. 5A and D) due to their differing morphology, colocalization of 2b with the commercial ER-Tracker™ Red revealed the same patterns.

198

150 2a 2b

100

50 Cell Viability (%) Viability Cell

0 -3 -2 -1 0 log concentration (uM)

Figure B1: IC50 plots for 2a and 2b with MO3.13 cells treated for 24 hours, with mean IC50 values being 18.59 and 33.54 µM, respectively. The 95% confidence intervals are 11.99 to 28.80 µM for 2a and 25.45 to 44.22 µM for 2b. Cells were tested on two separate occasions with all data combined for a total of n=6 for each concentration.

199

MTT: MO3.13 2a

24 hours 48 hours 100 72 hours

50 Cell Viability (%) Viability Cell

0 2 62 250 Concentration (uM) b MTT: MO3.13 2b

24 hours 48 hours 100 72 hours

50 Cell Viability (%) Viability Cell

0 2 62 250 Concentration (uM)

Figure B2: Kinetic cell viability plots for 2a (a) and 2b (b) with MO3.13 cells. Cells were treated with a range of concentrations for 24, 48, or 72 hours and the viability was measured via an MTT assay. The cells were tested in duplicate for each concentration at each time point.

200

Figure B3: Confocal imaging of oligodendrocyte cells incubated with 2b (10 μM) at <1, 15, 30, and 60 minutes after probe addition. Magnification is 100×.

201

Figure B4: Confocal imaging of MO3.13 cells incubated with ER-Tracker (1 μM) (panels A and D), flavonoid compounds 2b (panel B) or 2a (panel E) (500 nM, for each dye respectively). Panels (C) and (F) are the merged images of ER-Tracker Red and flavonoid probes. The nuclei of cells were stained with DRAQ5 (shown in purple).

202

Figure B5: Confocal imaging of MO3.13 cells incubated with (Top) 2b (1 µM) and (Bottom) ER-Tracker (1 µM). The images were acquired at 5-minute interval for 3 hours.

203

Figure B6: Confocal imaging of GL261 cells (panels A–C) and SIM A9 cells (panels D–F), which were incubated with ER-Tracker (1 μM, panels A and D), flavonoid 2b (500 nM, panels B and E) for 30 minutes. Panels (C) and (F) are the merged images of ER-Tracker Red and flavonoid probes. The nuclei of cells were stained with DRAQ5 (shown in purple).

204

APPENDIX C: LETTERS OF APPROVAL

Title: Cuprizone Intoxication Induces Cell Intrinsic Alterations in Oligodendrocyte Metabolism Independent of Copper Chelation Author: Alexandra Taraboletti, Tia Walker, Robin Avila, et al Publication: Biochemistry Publisher: American Chemical Society Date: Mar 1, 2017 Copyright © 2017, American Chemical Society

PERMISSION/LICENSE IS GRANTED FOR YOUR ORDER AT NO CHARGE

This type of permission/license, instead of the standard Terms & Conditions, is sent to you because no fee is being charged for your order. Please note the following:

▪ Permission is granted for your request in both print and electronic formats, and translations. ▪ If figures and/or tables were requested, they may be adapted or used in part. ▪ Please print this page for your records and send a copy of it to your publisher/graduate school. ▪ Appropriate credit for the requested material should be given as follows: "Reprinted (adapted) with permission from (COMPLETE REFERENCE CITATION). Copyright (YEAR) American Chemical Society." Insert appropriate information in place of the capitalized words. ▪ One-time permission is granted only for the use specified in your request. No additional uses are granted (such as derivative works or other editions). For any other uses, please submit a new request.

205

Fluorescent flavonoids for endoplasmic reticulum cell imaging

You do not need to request permission to reuse your own figures, diagrams, etc, that were originally published in a Royal Society of Chemistry publication. However, permission should be requested for use of the whole article or chapter except if reusing it in a thesis. If you are including an article or book chapter published by us in your thesis please ensure that your co-authors are aware of this. Reuse of material that was published originally by the Royal Society of Chemistry must be accompanied by the appropriate acknowledgement of the publication. The form of the acknowledgement is dependent on the journal in which it was published originally, as detailed in 'Acknowledgements'.

206

207