SELECTIVE NEURODEGENERATION IN ALZHEIMER’S DISEASE AND PARKINSON’S DISEASE by Juelu Wang B. Medicine, Hunan Normal University, 2008 M. Medicine, Central South University, 2011 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Neuroscience) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) October 2017 © Juelu Wang, 2017 ii Abstract Alzheimer’s disease (AD) and Parkinson’s disease (PD) are featured by cholinergic and dopaminergic neuron loss, respectively. As a unique pathological hallmark of AD, neuritic plaques contain aggregated amyloid β protein (Aβ), generated from amyloid β precursor protein (APP). APP mutations cause familial AD; mutations in the alpha-synuclein (SNCA) and leucine-rich repeat kinase 2 (LRRK2) genes are associated with PD. Recent studies suggest that the level of LRRK2 affects its toxicity in neurons. Therefore, understanding the mechanisms underlying LRRK2 expression would help to examine its pathogenic effects on PD. However, the features of the LRRK2 promoter remain elusive. In the first project, we cloned and characterized the LRRK2 promoter. There were two functional cis-acting specificity protein 1(Sp1)-responsive elements in its promoter. Our study demonstrates that LRRK2 transcription and translation were facilitated by Sp1 overexpression and blocked by an Sp1 inhibitor in vitro. The Lewy bodies primarily consist of α-synuclein protein, encoded by SNCA, and SNCAA53T mutation promotes α-synuclein aggregation. The Swedish APP mutation (APPSWE) promotes Aβ generation and AD pathogenesis. However, the mechanisms underlying selective neurodegeneration in AD and PD are still unknown. In the second project, we stably overexpressed wildtype and mutated APP and SNCA genes in cholinergic SN56 and dopaminergic MN9D cells. APPSWE and SNCAA53T mutations enhanced Aβ generation and α-synuclein inclusion formation in SN56 and MN9D cells, respectively. Aβ42 and mutant α-synuclein oligomers caused severe cell death in SN56-APPSWE and MN9D-SNCAA53T cells, iii respectively. Furthermore, syndecan 3 (SDC3) and fibroblast growth factor receptor like 1 (FGFRL1) genes were identified as two of the differentially expressed genes in APP- and SNCA- related stable cells by microarrays. SDC3 was increased in the cholinergic nucleus of APPSWE knock-in mouse brains, whereas FGFRL1 was elevated in dopaminergic neurons in SNCAA53T transgenic mice. Finally, knockdown of SDC3 and FGFRL1 attenuated oxidative stress-induced cell death in SN56-APPSWE and MN9D-SNCAA53T cells. Overall, these demonstrate that SDC3 and FGFRL1 mediated the specific effects of APPSWE and SNCAA53T on cholinergic and dopaminergic neurodegeneration in AD and PD, respectively. Our study suggests that SDC3 and FGFRL1 could be potential targets to alleviate the selective neurodegeneration in AD and PD. iv Lay summary Alzheimer’s disease (AD) and Parkinson’s disease (PD) are prevalent in the aging population. Memory loss in AD patients and motoric dysfunction in PD patients are caused by cell death within two different neuronal groups in distinct brain regions. To examine the pathogenesis of AD and PD, we first studied a PD-related gene, LRRK2. We demonstrated that LRRK2 transcription and translation were facilitated by Sp1 overexpression and blocked by a Sp1 inhibitor. In the second project, we aimed to examine why AD- and PD-associated gene mutations only kill certain neuronal types without affecting the others. We introduced these mutated genes into cholinergic and dopaminergic cells. SDC3 and FGFRL1 genes were found to mediate specific effects of these mutations on the selective neurodegeneration in AD and PD. Our study suggests that SDC3 and FGFRL1 could be potential targets to alleviate the selective neurodegeneration in AD and PD. v Preface After obtaining the Master’s Degree in Clinical Medicine specialized in Psychiatry, I joined Dr. Weihong Song’s lab to investigate the mechanisms underlying neurodegeneration of Alzheimer’s disease and Parkinson’s disease. As a training for the first-year PhD student, Dr. Song assigned me a project focusing on the transcriptional regulation of LRRK2 gene promoter. Chapter 2 is based on the findings from this project, which was completed by myself from beginning. Two of the promoter deletion plasmids were constructed by Dr. Michelle, a previous PhD student working in our lab. All the experiments and data analyses in this project were performed by myself. I am the first author on a paper published in Molecular Brain in March 2015 (Wang & Song, 2016). The journal granted permission for the author to include the published materials in this thesis. Chapter 3 and Chapter 4 addressed the second project, which was completed during my PhD training. It aims to reveal possible mechanisms underlying the selective neurodegeneration in AD and PD. More specifically, Chapter 3 discussed cell type-specific effects of the APPSWE and SNCAA53T mutations contributing to selective neurodegeneration in AD and PD. Chapter 4 described the effects of SDC3 and FGFRL1 on neurodegeneration in APPSWE-associated AD and SNCAA53T-associated PD. I designed and carried out all the experiments. Dr. Fang Cai, Dr. Yun Zhang, and Ms. Xinxin Liao helped me to breed and maintain APPSWE knock-in mouse line (APPNL/NL) and C57 mice. Dr. Qin Xu prepared Aβ oligomers for my experiments. APPNL/NL mice were originally generated by Dr. Takashi Saito’ group (Saito et al., 2014) and obtained from Bioken Bioresource Center. SNCAA53T transgenic mice (Prnp-SNCAA53T) was generated by vi Dr. Michael K. Lee et al. (M. K. Lee et al., 2002) and ordered from The Jackson Laboratory. All animal studies were approved by the University of British Columbia Animal Care Committee (protocol number: A14-0191). vii Table of contents Abstract .......................................................................................................................................... ii Lay summary ................................................................................................................................ iv Preface .............................................................................................................................................v Table of contents ......................................................................................................................... vii List of tables................................................................................................................................ xiii List of figures .............................................................................................................................. xiv List of abbreviations .................................................................................................................. xvi Acknowledgements .................................................................................................................. xxiii Dedication ...................................................................................................................................xxv Chapter 1: Introduction ................................................................................................................1 1.1 Alzheimer’s Disease (AD) .............................................................................................. 1 1.1.1 Clinical features of AD ............................................................................................... 1 1.1.2 Pathological features of AD ........................................................................................ 2 1.1.3 Genetics of AD ........................................................................................................... 4 1.2 Parkinson’s Disease (PD) ............................................................................................... 5 1.2.1 Clinical presentations of PD ....................................................................................... 5 1.2.2 Pathological features of PD ........................................................................................ 7 1.2.3 Genetics of PD ............................................................................................................ 8 1.3 APP processing and the amyloid hypothesis .................................................................. 9 1.3.1 The amyloidogenic pathway and β-secretase ........................................................... 10 1.3.2 The non-amyloidogenic pathway and α-secretase .................................................... 11 viii 1.3.3 Presenilins and γ-secretase ........................................................................................ 14 1.3.4 The amyloid hypothesis of AD ................................................................................. 16 1.3.5 The challenges of amyloid hypothesis and its revision ............................................ 18 1.3.6 The effects of APPSWE mutation on AD pathogenesis .............................................. 20 1.3.7 The toxicity of Aβ oligomer ..................................................................................... 23 1.4 The role of α-synuclein (αSyn) in PD ........................................................................... 24 1.4.1 The structure of αSyn protein ................................................................................... 25 1.4.2 Physiological functions of αSyn ............................................................................... 26 1.4.3 SNCAA53T mutation and its effects on PD pathogenesis .......................................... 27 1.4.4 αSyn oligomer and its toxicity .................................................................................. 30 1.4.5 The secretion and transmission of αSyn ................................................................... 31 1.5 Selective neurodegeneration in AD and PD ................................................................. 33 1.5.1 Selective neurodegeneration in AD .......................................................................... 34 1.5.2 Cholinergic hypothesis of AD .................................................................................. 36 1.5.3 Interaction between Aβ and cholinergic neurons...................................................... 38 1.5.4 Selective neurodegeneration in PD ........................................................................... 41 1.5.5 Interaction between αSyn and dopaminergic neurons .............................................. 43 1.6 LRRK2 and PD ............................................................................................................. 44 1.6.1 The expression and cellular functions of LRRK2..................................................... 45 1.6.2 LRRK2 protein domains ........................................................................................... 47 1.6.3 LRRK2 mutations and their effects .......................................................................... 49 1.6.4 The effects of LRRK2 protein level on cellular toxicity .......................................... 51 1.7 Overall goal of this research ......................................................................................... 53 ix 1.7.1 Sp1 enhances LRRK2 promoter activity and gene expression ................................. 53 1.7.2 Cell type- specific effects of the APPSWE and SNCAA53T mutations contributing to the selective neurodegeneration in AD and PD .................................................................... 55 1.7.3 The effects of syndecan 3 (SDC3) and fibroblast growth factor like 1 (FGFRL1) on the neurodegeneration in APPSWE-associated AD and SNCAA53T-associated PD ................ 57 Chapter 2: Sp1 enhances LRRK2 promoter activity and gene expression ............................59 2.1 Introduction ................................................................................................................... 59 2.2 Methods......................................................................................................................... 61 2.2.1 Primers and plasmids ................................................................................................ 61 2.2.2 Cell culture and transfection ..................................................................................... 62 2.2.3 Dual-luciferase reporter assay................................................................................... 63 2.2.4 5’-Rapid amplification of cDNA end (5’-RACE) assay ........................................... 63 2.2.5 Electrophoretic mobility shift assay (EMSA) ........................................................... 64 2.2.6 Sp1 knockdown ......................................................................................................... 65 2.2.7 RT-PCR..................................................................................................................... 65 2.2.8 Immunoblotting......................................................................................................... 66 2.2.9 Mithramycin A (MTM) treatment ............................................................................ 67 2.2.10 Statistical analyses .................................................................................................... 67 2.3 Results ........................................................................................................................... 67 2.3.1 Characterization of the human LRRK2 gene promoter ............................................ 67 2.3.2 Functional analyses of the human LRRK2 gene promoter ....................................... 69 2.3.3 Upregulation of the LRRK2 promoter activity by Sp1 ............................................. 71 2.3.4 Upregulation of the LRRK2 gene expression by Sp1 ............................................... 75 x 2.3.5 Inhibition of the LRRK2 promoter activity and gene expression by MTM ............. 77 2.4 Discussion ..................................................................................................................... 79 2.5 Conclusion .................................................................................................................... 82 Chapter 3: Cell type- specific effects of the APPSWE and SNCAA53T mutations in cholinergic and dopaminergic cells ................................................................................................................83 3.1 Introduction ................................................................................................................... 83 3.2 Methods......................................................................................................................... 86 3.2.1 Construction of plasmids .......................................................................................... 86 3.2.2 Cell culture and generation of stable cell lines ......................................................... 86 3.2.3 Immunoblotting......................................................................................................... 87 3.2.4 Aβ40/42 enzyme-linked immunosorbent assay (ELISA) ............................................ 88 3.2.5 Immunocytochemistry .............................................................................................. 88 3.2.6 Lactase dehydrogenase (LDH) assay and MTS assay .............................................. 89 3.2.7 Preparation of Aβ oligomers and αSyn oligomers .................................................... 89 3.2.8 Caspase-3/7 activity assay ........................................................................................ 90 3.2.9 Statistical analyses .................................................................................................... 90 3.3 Results ........................................................................................................................... 91 3.3.1 APPSWE mutation enhances the amyloidogenesis in cholinergic SN56 cells ........... 91 3.3.2 SNCAA53T mutation promotes the formation of cytoplasmic aggregates in dopaminergic MN9D cells .................................................................................................... 93 3.3.3 APPSWE mutation elevates oxidative stress-induced cell death in cholinergic SN56 cells through the apoptotic pathway ..................................................................................... 97 xi 3.3.4 SNCAA53T mutation sensitizes dopaminergic MN9D cells to oxidative stress through caspase-3 independent pathway .......................................................................................... 101 3.3.5 APPSWE mutation makes SN56 cells more vulnerable to Aβ42 oligomer treatment 104 3.3.6 Synergic effects of extracellular and intercellular αSyn species on cell death in MN9D cells ......................................................................................................................... 106 3.4 Discussion ................................................................................................................... 110 3.5 Conclusion .................................................................................................................. 114 Chapter 4: The effects of SDC3 and FGFRL1 on neurodegeneration in AD and PD .........115 4.1 Introduction ................................................................................................................. 115 4.2 Methods....................................................................................................................... 118 4.2.1 Cell culture, transfection, and knockdown of SDC3 and FGFRL1 ........................ 118 4.2.2 Animal and genotyping ........................................................................................... 118 4.2.3 Whole-genome expression profiling ....................................................................... 119 4.2.4 Analysis of DEGs ................................................................................................... 120 4.2.5 Gene Ontology (GO) enrichment analysis and Ingenuity pathway analysis (IPA) analysis ................................................................................................................................ 120 4.2.6 Quantitative reverse transcription PCR (qRT-PCR) ............................................... 121 4.2.7 Dissection of mouse brain and preparation of mouse brain tissue homogenate ..... 122 4.2.8 Immunofluorescent staining.................................................................................... 123 4.2.9 LDH assay ............................................................................................................... 124 4.2.10 RT-PCR................................................................................................................... 124 4.2.11 Immunoblotting....................................................................................................... 125 4.2.12 Statistical analysis ................................................................................................... 125 xii 4.3 Results ......................................................................................................................... 126 4.3.1 Data processing for DNA microarray and comparison of gene expression in SN56 and MN9D cells .................................................................................................................. 126 4.3.2 APPSWE and SNCAA53T mutations have differential effects on gene expression in cholinergic and dopaminergic neuronal cells ..................................................................... 130 4.3.3 Validation of selected genes differentially affected by APPSWE and SNCAA53T mutations in cholinergic and dopaminergic neuronal cells by qRT-PCR........................... 134 4.3.4 Differential expression of SDC3 and FGFRL1 proteins in cholinergic and dopaminergic cells carrying APPSWE or SNCAA53T mutations ........................................... 138 4.3.5 Differential expression of endogenous SDC3 and FGFRL1 protein in AD and PD mouse models...................................................................................................................... 141 4.3.6 The effects of SDC3 and FGFRL1 on oxidative stress-induced cell death ............ 146 4.4 Discussion ................................................................................................................... 148 4.5 Conclusion .................................................................................................................. 152 Chapter 5: Conclusions and discussions ..................................................................................154 5.1 General discussion ...................................................................................................... 154 5.2 Novelty and significance of research .......................................................................... 157 5.3 Limitation and future directions .................................................................................. 159 5.4 Further discussion ....................................................................................................... 161 5.4.1 The application of MTM in treating neurodegenerative diseases ........................... 161 5.4.2 Oligomer-based treatment in AD and PD ............................................................... 162 Bibliography ...............................................................................................................................165 xiii List of tables Table 4.1 Oligonucleotides for the qRT-PCR............................................................................. 122 Table 4.2 Go analysis for DEGs in SN56 and MN9D cells........................................................ 128 Table 4.3 GO analysis for DEGs in APP-related stable cells ..................................................... 133 Table 4.4 GO analysis for DEGs in SNCA-related stable cells .................................................. 133 xiv List of figures Figure 1.1 The amyloidogenic and non-amyloidogenic pathways of APP processing. ............... 14 Figure 1.2 Major pathogenic events leading to AD proposed by the ‘revised amyloid hypothesis’........................................................................................................................................................ 20 Figure 1.3 Protein domains and pathogenic mutations of LRRK2 ............................................... 47 Figure 2.1 Identification of TSS and sequence features of the human LRRK2 gene promoter. .. 69 Figure 2.2 Deletion analyses of the human LRRK2 promoter activity. ....................................... 71 Figure 2.3 The binding between Sp1 and cis-acting elements on the LRRK2 gene promoter promotes LRRK2 promoter activity. ............................................................................................ 73 Figure 2.4 Sp1 upregulates LRRK2 gene expression at both mRNA and protein levels. ............ 76 Figure 2.5 MTM treatment inhibits LRRK2 promoter activity and gene expression. .................. 78 Figure 3.1 APPSWE mutation increases C99/ C83 ratio and promotes Aβ generation in cholinergic SN56 cells. .................................................................................................................................... 93 Figure 3.2 SNCAA53T mutation increases the size of LBs-like inclusions in dopaminergic MN9D cells. .............................................................................................................................................. 96 Figure 3.3 APPSWE mutation enhances oxidative stress-induced cytotoxicity in cholinergic SN56 cells through the apoptotic pathway. .......................................................................................... 100 Figure 3.4 SNCAA53T mutation promotes oxidative stress-induced cytotoxicity in dopaminergic MN9D cells through caspase-3 independent pathway. ............................................................... 103 Figure 3.5 Aβ42 oligomers increase cell death and caspase-3/7 activation in SN56-APPSWE cells but not MN9D-APPSWE cells. ..................................................................................................... 106 xv Figure 3.6 αSyn WT and A53T oligomers elevate cell death in dopaminergic MN9D cells overexpressing SNCA with the same genotype through caspase-3/7 independent pathway. ..... 109 Figure 4.1 Data processing and exploratory analysis for 20 datasets. ........................................ 129 Figure 4.2 Analysis of DEGs in APP-related and SNCA-related stable cells. ........................... 132 Figure 4.3 Cellular functions and top canonical pathways overrepresented in the DEGs for analyzing APP- and SNCA-related arrays by IPA. .................................................................... 134 Figure 4.4 Validation of gene expression by qRT-PCR for selected DEGs from analyzing APP-related stable cells. ...................................................................................................................... 136 Figure 4.5 Validation of gene expression by qRT-PCR for selected DEGs from analyzing SNCA-related stable cells. ...................................................................................................................... 138 Figure 4.6 Differential expression of endogenous SDC3 and FGFRL1 protein in APPSWE- and SNCAA53T- overexpressing SN56 and MN9D cells.................................................................... 140 Figure 4.7 Differential expression of the endogenous SDC3 protein in the MS and SN of APPNL/NL mice. ............................................................................................................................ 143 Figure 4.8 Differential expression of the endogenous FGFRL1 protein in the cholinergic and dopaminergic neurons of Prnp-SNCAA53T mice. ........................................................................ 145 Figure 4.9 . Knockdown of SDC3 and FGFRL1 alleviates oxidative stress- induced cell death in SN56-APPSWE and MN9D-SNCAA53T cells, respectively. ......................................................... 147 xvi List of abbreviations 4-HNE 4-hydroxynonenal 4E-BP Eukaryotic initiation factor 4E-binding protein 5’-RACE 5’-rapid amplification of cDNA ends αSyn α-synuclein AADC L-aromatic amino acid decarboxylase Aβ Amyloid-β Aβ*56 56-kDa soluble Aβ species Ach Acetylcholine AChE Acetylcholinesterase AchRs Acetylcholine receptors ADCY4 Adenylate cyclase 4 AD Alzheimer’s disease ADAM A disintegrin and metalloproteinase ADDLs Aβ-derived diffusible ligands AICD APP intercellular domain AMPA α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid ANOVA Analysis of variance APOE Apolipoprotein E APP Amyloid β precursor protein APPSWE Swedish APP mutation BACE1 β-amyloid cleaving enzyme-1 xvii BACE2 β-amyloid cleaving enzyme-2 Bp Base pair C83 Carboxy-terminal fragment 83 C89 Carboxy-terminal fragment 89 C99 Carboxy-terminal fragment 99 CBF Cholinergic basal forebrain cDNA complementary DNA CHOP C/EBP homologous protein ChAT Choline acetyltransferase CHO Chinese hamster ovary CHIP Carboxyl terminus of HSP70-interacting protein CMA Chaperone-mediated autophagy COR C-terminal of ROC cRNA complementary RNA CSF Cerebrospinal fluid CXCL1 C-X-C motif chemokine ligand 1 CXCL2 C-X-C motif chemokine ligand 2 DBB Diagonal band of Broca DEGs Differentially expressed genes DLB Dementia with Lewy bodies DMEM Dulbecco’s modified Eagle’s medium DOPAL 3,4-dihydroxyphenylacetaldehyde xviii DS Down syndrome eIF4E Eukaryotic initiation factor 4E EOAD Early-onset Alzheimer’s disease EOPD Early-onset Parkinson’s disease ER Endoplasmic reticulum FAD Familial Alzheimer’s disease FBS Fetal bovine serum FC Fold change FDR False discovery rate EGF Epidermal growth factor receptor FGFRL1 Fibroblast growth factor receptor like 1 EMSA Electrophoretic mobility shift assay FN1 Fibronectin 1 FTDP-17 Frontotemporal dementia with parkinsonism on chromosome 17 GAGs Glycosaminoglycans GAP GTPase activating protein gDNA genomic DNA GDP Guanosine diphosphate GEF Guanine nucleotide exchange factor GO Gene ontology GTP Guanosine-5’-triphosphate HA Hemagglutinin xix HCHWA-Dutch Hereditary cerebral hemorrhage with amyloidosis-Dutch type H2O2 Hydrogen peroxide HD Huntington's disease HEK293 Human embryonic kidney 293 hiPSC Human induced pluripotent stem cell Hsp90 Heat shock protein 90 iPSC Induced pluripotent stem cell KPI Kunitz protease inhibitor LBs Lewy bodies LBVAD Lewy body variant of Alzheimer’s disease LCM Laser capture microdissection Limma Linear models for microarray data LNs Lewy neurites LOAD Last-onset Alzheimer’s disease LRRK2 Leucine-rich repeat kinase 2 LTA Lymphotoxin alpha LTD Long-term depression LTP Long-term potentiation mAchRs muscarinic Acetylcholine receptors MBP Myelin basic protein MEF Mouse embryonic fibroblasts METH Methamphetamine xx MRAS Muscle RAS oncogene homolog MS Medial septum MSA Multiple system atrophy MYD88 Myeloid differentiation primary response gene 88 MTM Mithramycin A MTS 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium MTT 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide nAchRs nicotinic Acetylcholine receptors NAC Non-Aβ component of Alzheimer’s disease amyloid plaques NBM Nucleus basalis of Meynert NFTs Neurofibrillary tangles NTRK3 Neurotrophic receptor tyrosine kinase 3 p3 3-kDa peptide PAGE Polyacrylamide gel electrophoresis PBS Phosphate-buffered saline PCR Polymerase chain reaction PDGFRA Platelet derived growth factor receptor alpha PDL Poly-D-lysine PET Positron emission tomography PFs Protofibrils PHFs Paired helical filaments xxi PKC Protein kinase C PPP1R3C Protein phosphatase 1 regulatory subunit 3C PS1 Presenilin 1 PS2 Presenilin 2 PSP Progressive supranuclear palsy PVDF-FL Polyvinylidene fluoride PTS 6-pyruvoyl-tetrahydropterin synthase qRT-PCR Quantitative reverse transcription polymerase chain reaction RLU Relative luciferase unit ROC Ras of complex proteins RT-PCR Reverse transcription polymerase chain reaction sAPPα Secretory APPα sAPPβ Secretory APPβ SDC3 Syndecan 3 SDS Sodium dodecyl sulfate SN Substantia nigra SNCA Alpha-synuclein SNpc pars compacta of substantial nigra SNARE Soluble N-ethylmaleimide-sensitive factor attachment protein receptor SP Sodium pyruvate Sp1 Specificity protein 1 tAβ truncated Aβ xxii TKL Tyrosine kinase-like TBS Tris-buffered saline TSS Transcription start site TH Tyrosine hydroxylase TLR2 Toll-like receptor 2 xxiii Acknowledgements I would like to express my sincere thanks to so many wonderful people I met during my 26 years of experience in studying, especially the challenging time in my Ph.D. training. First and foremost, I would like to thank my supervisor, Dr. Weihong Song. I am thankful for your guidance and suggestion for my research and projects along the way. You always offered the supports with patience when I got stuck in my experiments, and you always helped me find the potential in myself when I felt lost in my career development. Thank you for helping me establish the ability of critical thinking, rationalize the experiments I performed, and develop my self-confidence. Your opinions about science and academia broadened my scope, your passion for life motivated me, and your financial supports made me feel independent and secured to take every effort in studying and finishing my degree. I would like to thank my supervisory committee members, Dr. Yutian Wang, Dr. Ging-Yuek Robin Hsiung, and Dr. Shernaz Bamji. They provided me with invaluable advices and great opinions on my projects through these years, and kindly offered any help when I asked for. I would like to thank Dr. Yutian Wang for asking me the enlightening questions in the committee meetings, letting me use the lyophilization equipment in his lab, and helping me a lot when I prepared oligomer samples. I would also like to thank Dr. Shernaz Bamji to discuss with me about the experimental design to improve my project’s overall impact and significance. I would like to thank Dr. Ging-Yuek Robin Hsiung for attending my every committee meeting and giving me excellent feedbacks on my project. I would like to express my gratitude to Dr. Hongling Luo, who was the external examiner of my comprehensive examination. xxiv To my current and past colleagues in the Song lab, I have been appreciated your mental supports and accompanies in the journey of pursuing my PhD degree. Special thanks are given to Dr. Yili Wu, Dr. Qin Xu, and Dr. Fang Cai, who helped me acquire relevant techniques, and gave me valuable advices on my projects. Many thanks to Dr. Mingming Zhang and Dr. Si Zhang, who gave me strategic advices on drafting my project proposal. I would also like to thank Dr. Yun Zhang and Dr. Xinxin Liao, who taught me techniques to process mouse samples. Special thanks to Ms. Susan Lin and Ms. Beibei Song for proofreading my thesis. Many thanks to other colleagues, Dr. Yuhang Liu, Mr. Bruno Herculano, and Dr. Zhe Wang. I would also like to acknowledge the financial support during my doctoral training, the international student scholarship and College for Interdisciplinary Studies Graduate Award from UBC. Last but not least, I cannot make these without the supports from my beloved family. I thank my parents from the deepest of my heart for their endless love and unconditioned trust in me. They create a free environment for me, letting me choose the things I love and be the person I want to be. Their words cheered me up during my tough days, and their optimism towards life makes me a positive person. Thanks for having my little one (Xiao Qi) two years ago. His smile brings me the sunshine everyday, and makes me become a brave and fearless mother. Thanks to my husband for sharing the happiness with me, taking me to feel the world around, and comforting me whenever I feel frustrated. Many thanks to my four best friends, who have been supportive for 20 years and will always be there with me in the future. Thanks to Ms. Guan Ling, Ms. Qian Zhang, and all my good friends met in Vancouver. xxv Dedication To my beloved family 献给我最爱的家人 1 Chapter 1: Introduction 1.1 Alzheimer’s Disease (AD) Alzheimer’s Disease (AD) was first described by Alois Alzheimer in a meeting presentation in 1906. Nowadays, the clinical and pathological features have become well-known due to the extensive work done in the field. As the most common neurodegenerative disorder, AD is the leading cause of dementia - a gradual deterioration of memory and other cognitive functions that is severe enough to interfere daily life. The prevalence of AD increases with aging. It develops in about 11% of people aged over 65-year-old and one-third of individuals who are over 85 years old (Hebert, Weuve, Scherr, & Evans, 2013). In the United States, over 5.2 million people aged over 65 have AD, and it costs over 200 billion dollars for taking care of AD patients according to the annual report of Alzheimer’s Association in 2016. As mentioned in the World Alzheimer Report 2015, 46.8 million people were living with dementia in 2015, and this number will double every 20 years. The increasing prevalence due to extended lifespan creates an urgency in the field to explore the molecular mechanisms underlying the disease progression and to search for disease-modifying strategies. 1.1.1 Clinical features of AD The most prominent symptoms observed in AD patient is dementia, which is a gradual deterioration in memory and other cognitive functions. Based on the severity of dementia, the course of AD can be roughly divided into four stages: pre-dementia stage, mild dementia stage, moderate dementia stage, and severe dementia stage (Forstl & Kurz, 1999). It has been reported that AD-related pathological alterations probably occur 10 to 20 years before the manifestation of the first clinical syndrome. During the pre-dementia stage, patients may have mild cognitive 2 impairment with no obvious malfunction in daily living, and the diagnostic criteria for dementia are not met. In mild dementia stage lasting from 2 to 5 years, learning and memory functions deteriorate, but patients are still able to live independently. When AD patients enter the moderate dementia stage, recent memory is severely impaired, and the deficits in cognitive function become more generalized, making it hard for patients to live by themselves. During the severe dementia stage, almost all the cognitive functions are affected and patients are dependent on caregivers for daily life (Holtzman, Morris, & Goate, 2011). 1.1.2 Pathological features of AD AD is pathologically characterized by extracellular neuritic plaques, intracellular neurofibrillary tangles (NFTs), extensive neuron loss, and brain atrophy. Neuritic plaques are extracellular proteinaceous deposits, associated with dystrophic neurites and surrounded by activated microglia and astrocytes (Hyman et al., 2012). The essential components of the neuritic plaques are amyloid-β (Aβ) protein in its aggregated form (Glenner & Wong, 1984a, 1984b; Masters et al., 1985). Aβ is generated from its precursor protein, amyloid β precursor protein (APP) (Kang et al., 1987; Robakis, Ramakrishna, Wolfe, & Wisniewski, 1987; Tanzi et al., 1987), with varying length from 37 to 43 amino acids (Benilova, Karran, & De Strooper, 2012). Dystrophic neurites are degenerated neuronal processes within or around plaques, and they contain a number of mitochondria, dilated lysosomes, and paired helical filaments (PHFs) (Dickson, 1997). Besides neuritic plaques, diffuse plaques (another type of plaque), appear as amorphous and granular deposits without a clearly condensed core. They are often devoid of dystrophic neurites (Joachim, Morris, & Selkoe, 1989; Tagliavini, Giaccone, Frangione, & Bugiani, 1988; Yamaguchi, Hirai, Morimatsu, Shoji, & Harigaya, 1988). These diffuse plaques coexist with 3 typical neuritic plaques, but they differ in the Aβ species they are made of. The majority of the Aβ species in the diffuse plaques ends at amino acid 42, whereas both of Aβ40 and Aβ42 are present in the neuritic plaques (D. Selkoe, 2001). NFTs are intraneuronal filamentous inclusions, occupying a large part of the cytoplasm near the nucleus. The abnormal inclusions are made of β-sheet rich paired helical filaments (PHFs), two ~10nm filaments wind with each other to form helices (D. Selkoe, 2001). By performing immunohistochemical and molecular analyses, NFTs are found to be composed of hyperphosphorylated microtubule-associated protein tau (Brion, Couck, Passareiro, & Flament-Durand, 1985; Grundke-Iqbal et al., 1986). Phosphorylation of various residues in the tau protein makes it dissociate with microtubules and become insoluble in the cytoplasm. Both the neuritic plaques and NFTs are considered as the neuropathological criteria for diagnosing AD, but only neuritic plaques are exclusive to AD (Dubois et al., 2007). NFTs are also found in other neurodegenerative diseases, such as frontotemporal dementia with parkinsonism on chromosome 17 (FTDP-17), Pick’s disease, and progressive supranuclear palsy (PSP) (Dubois et al., 2007; Hyman et al., 2012). In addition to extracellular neuritic plaques and intracellular neurofibrillary tangles, there is extensive neuron loss within the basal forebrain, and cholinergic denervation in the cerebral cortex and associated limbic areas. Neurodegeneration in those areas are closely associated with the cognitive symptoms and memory deficits seen in AD (Bartus, Dean, Pontecorvo, & Flicker, 1985; P. Whitehouse, Price, Clark, Coyle, & DeLong, 1981). The discoveries of cholinergic deficiency in AD will be discussed further in Section 1.5.1. 4 1.1.3 Genetics of AD From a genetic perspective, AD can be divided into two categories, familial and sporadic. Familial AD (FAD) accounts for less than 1% of all AD cases. The onset age for most FAD is before 65 years of age, known as early-onset AD (EOAD) (Selkoe & Hardy, 2016). So far, there are three identified genes containing pathogenic mutations involved in early-onset FAD, including APP, presenilin 1(PS1), and presenilin 2 (PS2) (Bekris, Yu, Bird, & Tsuang, 2010). In sporadic AD, apolipoprotein E (APOE) genotype is the only confirmed genetic risk factor across different studies, with the ε4 allele being a major risk factor and the ε2 allele being protective (Corder et al., 1993; Rebeck, Reiter, Strickland, & Hyman, 1993; Saunders et al., 1993; Strittmatter et al., 1993). APP gene was cloned in 1987 and the first pathogenic mutation in the APP gene was discovered in hereditary cerebral hemorrhage with amyloidosis-Dutch type (HCHWA-Dutch) in 1990 (Goldgaber, Lerman, McBride, Saffiotti, & Gajdusek, 1987); Kang et al., 1987; Levy et al., 1990; Robakis, Ramakrishna, Wolfe, & Wisniewski, 1987; Tanzi et al., 1987). Afterward, the first FAD-associated mutation was identified in APP gene - APP London mutation, causing an amino acid substitution from valine to isoleucine at the position of 717 (APPV717I) (Chartier-Harlin et al., 1991; Goate et al., 1991; Murrell, Farlow, Ghetti, & Benson, 1991). With more studies to screen APP mutations, one of the most well-known disease-causing Swedish APP (APPSWE) mutation was reported in 1992. APPSWE is a double mutation at the site of amino acids 670 and 671 resulting in a lysine and methionine to asparagine and leucine (K670N/M671L), respectively (Citron et al., 1992). Until now, there are over 25 APP mutations have been linked 5 to the pathogenesis of AD, most of which are located within or adjacent to the Aβ sequence (http://www.alzforum.org/mutations/). 1.2 Parkinson’s Disease (PD) PD was first described by Dr. James Parkin in 1817 as the ‘shaking palsy’. It is the leading cause of parkinsonism (Dickson, 2012) and second most prevalent neurodegenerative disorder after AD (Jankovic, 2008). The prevalence of PD is age-dependent, with ~1% of the population at the age over 65 years old and increasing to 4-5% in the individuals older than 85 years (de Lau & Breteler, 2006). According to the reports from Parkinson Society Canada, more than 25 people diagnosed with PD every day in Canada. The individuals affected by PD are more than 100,000 in 2003, and this number is estimated to increase to more than 163,700 between 2011 and 2031. Additionally, PD causes enormous social and economic burden with the annual cost of over $120 million nationwide, ranking as the third of highest direct healthcare costs. Although many studies have deepened our understanding of the pathological mechanisms underlying PD, more data is required to reveal the culprits contributive to the etiology of PD in order to formulate effective therapeutic strategies. 1.2.1 Clinical presentations of PD PD is a progressive neurodegenerative disorder with significant motor symptoms, including bradykinesia, resting tremor, postural instability, and rigidity. During disease progression, patients may display some but not necessarily all of the symptoms. (Lang & Lozano, 1998a, 1998b; Massano & Bhatia, 2012). 6 Bradykinesia is characterized by slowness of movement with gradual loss of speed and amplitude during repetitive alternating movement. It is the most recognizable clinical feature seen in PD patients (Berardelli, Rothwell, Thompson, & Hallett, 2001; Jankovic, 2008). In a clinical study, the subscale of bradykinesia in the modified Columbia score correlates best with the severity of dopaminergic deficiency determined by fluorodopa positron emission tomography (PET) (Vingerhoets, Schulzer, Calne, & Snow, 1997). This finding is in accord with the observation that the degree of bradykinesia is significantly correlated with the decreased fluorodopa uptake in putamen and globus pallidum (Lozza, Marie, & Baron, 2002). Resting tremor usually occurs at the frequency of 4 to 6 Hz affecting one side of the body, and it is most evident when the affected segment is relaxed (Jankovic, 2008). However, the appearance of resting tremor in the patients is varied. Some studies report its occurrence is 69% in PD patients at the onset of disease (Hughes, Daniel, Blankson, & Lees, 1993), and others point out the absence of resting tremor in around 11% of PD patients (Martin, Loewenson, Resch, & Baker, 1973). In addition, rigidity is used to describe the increased resistance of muscle during passive movement, and ‘cogwheel’ phenomenon, periodic interruption in the passive movement of affected limbs, is the most identifiable one in PD (Rodriguez-Oroz et al., 2009). Due to the loss of postural reflexes, postural instability appears at the later stages of PD progression (Coelho & Ferreira, 2012). Besides motor symptoms, non-motor symptoms are common to PD patients, including autonomous symptoms (postprandial hypotension, constipation, sexual dysfunction, sweating dysfunction), sleep disorders (rapid eye movement disorder), neuropsychiatric symptoms (depression), and hyposmia. Although motor symptoms are prominent and often used for 7 diagnosis, the non-motor symptoms cannot be ignored. They are valuable for early diagnosis since they precede the classical motor signs (Jankovic, 2008). 1.2.2 Pathological features of PD PD is characterized by two major neuropathological features. One is the degeneration of dopaminergic neurons in the pars compacta of substantial nigra (SNpc), leading to dopaminergic deficiency in the striatum (Fearnley & Lees, 1991;Lang & Lozano, 1998a). The discovery of dopamine deficiency in PD was initiated with the findings in the 1960s, when significant dopaminergic neuron loss was reported in the caudate nucleus and putamen in PD patients by examining autopsy samples (Ehringer & Hornykiewicz, 1960). Since then, multiple aspects of the biochemical alterations within the dopaminergic system has been thoroughly studied, which will be further discussed in Section 1.5.4. The other neuropathological feature is the presence of Lewy bodies (LBs) and Lewy neurites (LNs) in the surviving neurons (Spillantini et al., 1997). The classic type of LBs, mostly found in the brain stem, appears as spherical inclusions with a hyaline eosinophilic core and a pale peripheral halo in the neuronal perikarya (Forno, 1996). Another type of LB - ‘cortical Lewy bodies’- was first found in the cerebral cortex, and has pale staining and less compact than classic LBs (Ikeda, Ikeda, Yoshimura, Kato, & Namba, 1978). As a milestone in PD, LBs were found to mainly consist of α-synuclein (αSyn) in its aggregated form (Spillantini et al., 1997). Additionally, antibodies to neurofilament (Galvin et al., 1997), ubiquitin (Kuzuhara, Mori, Izumiyama, Yoshimura, & Ihara, 1988), and the ubiquitin binding protein p62 (Kuusisto, Parkkinen, & Alafuzoff, 2003) are consistently positive for detecting LBs with varying sensitivities, suggesting their existence in LBs. αSyn immunoactivity is also found 8 in the degenerated neuronal processes, which are termed as LNs (Braak, Sandmann-Keil, Gai, & Braak, 1999). Although LBs are essential for diagnosing PD from a pathological perspective, they are not exclusive to nor required for all PD cases. They are also present in multiple system atrophy (MSA) and dementia with Lewy body (DLB) (Dickson et al., 2009), while absent in Parkin mutation carriers and some leucine-rich repeat kinase 2 (LRRK2) mutations (Farrer et al., 2001; Zimprich et al., 2004). To better evaluate the progression of PD, Braak and his co-works have developed a staging criterion based on the distribution of αSyn-related pathology. It usually starts from the olfactory bulb and dorsal motor vagal nucleus, then evolve to the brainstem, and finally reach to the cortical motor and sensory areas (Braak et al., 2003). 1.2.3 Genetics of PD The average age of onset for PD is 60-65 years old, although rare cases develop PD younger than 50 years old, which are defined as early-onset PD (EOPD) (Farrer, 2006). The notion of genetic factors contributing to the etiology of PD started in 1997, when the mutation in the alpha synculein (SNCA) gene was identified in familial PD (Polymeropoulos, 1997). Since then, studies of PD-linked mutations and PD-associated risk loci have flourished. Similar to AD, PD is categorized into sporadic and familial PD, with the latter accounting for 10%-15% of PD cases (Gasser, 2009). However, the genetics of inheritable PD is more complicated than AD since it involves more genes (Singleton, Farrer, & Bonifati, 2013). Linkage analyses have discovered mutations in the SNCA gene related to PD including both missense point mutations (Appel-Cresswell et al., 2013; Kiely et al., 2013; Kruger et al., 1998; Lesage et al., 2013; Polymeropoulos, 1997; Proukakis et al., 2013; Zarranz et al., 2004) and gene multiplication 9 (Chartier-Harlin et al., 2004; Singleton et al., 2003), which will be further discussed in Section 1.4.3. There are several other genes linked to familial PD, including LRRK2 (Nichols et al., 2005; West et al., 2005), Parkin (PARK2) (Kitada et al., 1998), PINK (PARK6) (Valente et al., 2004), DJ-1(PARK7) (Bonifati et al., 2003), ATP13A2 (Ramirez et al., 2006), PLA2G6 (Paisan-Ruiz et al., 2009), and VSP35 (Vilarino-Guell et al., 2011; Zimprich et al., 2011). As the most common genetic cause of PD, LRRK2 accounts for 2-6% of hereditary PD and 1-2% of sporadic PD (Brice, 2005; Healy et al., 2008; Lesage et al., 2006). Seven pathogenic LRRK2 mutations have been confirmed across different studies, including G2019S, I2020T, N1437H, R1441G/C/H, and Y1699C (Trinh & Farrer, 2013). 1.3 APP processing and the amyloid hypothesis In 1984, an identical short peptide, Aβ protein, was purified from cerebrovascular amyloid fibrils in both AD and Down syndrome (DS) patients. At the time, it was already known that DS patients over the age of 40 inevitably develop AD neuropathology, since DS patients carry an extra copy of chromosome 21(Olson & Shaw, 1969). Glenner and Wong hypothesized that the gene encoding Aβ should be located on chromosome 21 (Glenner & Wong, 1984a, 1984b). Shortly after, another group purified Aβ from cerebral plaque cores in brain samples from both AD and aged DS cases (Masters et al., 1985). Following these exciting findings, the coding sequence of Aβ was shown to be a part of a larger precursor protein, APP, and its full-length complementary DNA (cDNA) was successfully cloned by four different groups at the same time (Goldgaber, Lerman, McBride, Saffiotti, & Gajdusek, 1987; Kang et al., 1987; Robakis, 10 Ramakrishna, Wolfe, & Wisniewski, 1987; Tanzi et al., 1987). APP is a type I integral transmembrane protein and it has three isoforms of varying lengths: APP695, APP751, and APP770. The two longer isoforms bear a Kunitz protease inhibitor (KPI) domain and are mostly expressed in peripheral tissues, but the APP695 isoform is most abundant in the brain (Kang et al., 1987; Kitaguchi, Takahashi, Tokushima, Shiojiri, & Ito, 1988; Ponte et al., 1988). 1.3.1 The amyloidogenic pathway and β-secretase APP undergoes proteolysis in two distinctive pathways, the amyloidogenic pathway and the non-amyloidogenic pathway, as presented in Figure 1.1. The amyloidogenic pathway is initiated by the cleavage of APP at its β-site by an aspartyl protease, β-secretase, at the first amino acid of the Aβ domain. This cleavage releases the secretory APPβ (sAPPβ) and leaves a carboxy-terminal fragment 99 (C99) anchored to the membrane. Membrane-bound C99 is further cleaved by γ-secretase to generate the Aβ peptide and the APP intercellular domain (AICD). Beta-amyloid cleaving enzyme-1 (BACE1) was identified as the active β-secretase in vivo (Hussain et al., 1999; Sinha et al., 1999; Vassar et al., 1999; Yan et al., 1999). Data from our lab clearly demonstrated that the preferential cleavage sites by BACE1 were different between wildtype APP (APPWT) and APPSWE mutant. BACE1 cleaves APPWT at the Glu-11 site (β’-site) within the Aβ domain to produce a carboxy-terminal fragment 89 (C89), thereby preventing Aβ generation. However, when BACE1 was overexpressed together with APPSWE, APPSWE shifted the BACE1 cleavage site from Glu-11 to Asp-1 and substantially increased Aβ production (Deng et al., 2013). The details of APP processing at the β’-site will be discussed in Section 1.3.2. 11 BACE1 is a 501-amino acid type I transmembrane aspartic protease that is related to the pepsin-like family of aspartic peptidases. It has two catalytic sites and is located in the lumen of acidic intercellular compartments (Hussain et al., 1999; Sinha et al., 1999; Vassar et al., 1999; Yan et al., 1999). There is very low expression of BACE1 in peripheral tissues; its expression is confined to the brain and is especially enriched in neurons (Vassar et al., 1999). Although BACE1 serves an essential role in generating Aβ in the neuron, BACE1 knockout mice were initially reported to be viable and fertile (H. Cai et al., 2001; Luo et al., 2001; Roberds et al., 2001). However, following studies revealed that BACE1 knockout mice actually displayed extensive deficits in the nervous system, such as hypomyelination (Willem et al., 2006), abnormal neurogenesis and astrogenesis (Hu, He, Luo, Tsubota, & Yan, 2013), neurodegeneration, and alteration of neuronal activities (Hu et al., 2010). Most importantly, genetic ablation of BACE1 ameliorates the Aβ burden, rescues Aβ-dependent memory deficits, and improves cholinergic dysfunction in various APP transgenic mouse models (Laird et al., 2005; Luo et al., 2003; McConlogue et al., 2007; Ohno et al., 2004). 1.3.2 The non-amyloidogenic pathway and α-secretase Under physiological conditions, the majority of APP is first cleaved at the α-cleavage site located between Lys-16 and Leu-17 within the Aβ domain. This α-cleavage releases the ectodomain secretory APPα (sAPPα) and leaves an 83-amino acid membrane-bound carboxy-terminal fragment (C83) (Buxbaum et al., 1998; Koike et al., 1999; Lammich et al., 1999). Following α-cleavage, C83 is further cleaved by γ-secretase to generate a small 3 kDa peptide (p3) and the AICD (Haass, Hung, Schlossmacher, Teplow, & Selkoe, 1993; Haass et al., 1992). As mentioned 12 before, APP can also be cleaved at the β’-site between Try-10 and Glu-11 within Aβ domain by BACE1 (Vassar et al., 1999). Its cleavage generates secreted APPβ (sAPPβ’) and membrane-bound C89, and C89 is further cleaved by γ-secretase to produce truncated Aβ (tAβ) and AICD. Additionally, the non-amyloidogenic processing of APP can be conferred by β-amyloid cleaving enzyme-2 (BACE2) at the θ-site between Phe-19 and Phe-20 within the Aβ domain to generate secretory APPθ (sAPPθ) and membrane-bound C80 (Sun, He, & Song, 2006), and C80 is further cleaved by γ-secretase (Figure 1.1). Collectively, the proteolytic cleavage of APP at the α-site, β’-site, and θ-site prevents Aβ generation and is involved in the non-amyloidogenic pathways. Several α-secretase candidates belonging to the a disintegrin and metalloproteinase (ADAM) protein family have been identified by previous studies, including ADAM9, 10, 17 and 19 (Asai et al., 2003; Koike et al., 1999; Lammich et al., 1999; Tanabe et al., 2007). Overexpression of ADAM10 in human embryonic kidney 293 (HEK293) cells results in a substantial increase of sAPPα, while the inhibition of endogenous α-secretase activity by a dominant negative mutant of ADAM10 significantly decreases sAPPα (Lammich et al., 1999). The increase of sAPPα through protein kinase C (PKC) activation is abolished in cells derived from ADAM17 knockout mice, suggesting the role of ADAM17 in mediating α-cleavage of APP by activating PKC (Buxbaum et al., 1998). ADAM9-overexpressing cells exclusively cleave APP at the α-cleavage site after being stimulated by PKC (Koike et al., 1999). More recently, a study has reported that overexpression of ADAM19 elevates sAPPα level in HEK293 cells, and sAPPα is down-regulated by ADAM19 siRNA treatment (Tanabe et al., 2007). 13 As a homolog of BACE1, BACE2 shares 64% similarity of amino acid sequence between them (Bennett et al., 2000). BACE2 is weakly expressed in neurons, but has a relatively higher expression level in glial cells (Laird et al., 2005). Because of the sequence similarity between BACE1 and BACE2, BACE2 was originally thought as another β-secretase. However, it turned out to be a novel θ-site secretase (Sun et al., 2006). Interestingly, BACE2 is located on chromosome 21, but its protein levels do not change in DS and the θ- cleavage site does not contribute to the AD-related neuropathology seen in DS (Sun, He, & Song, 2006). Moreover, our lab also identified different transcription factor binding sites in the promoter region of BACE1 and BACE2, providing a possible explanation for their distinctive expression pattern (Sun et al., 2005). Taken together, BACE2 can be distinguished from BACE1 in terms of transcriptional regulation, expression pattern, the specificity of APP processing, and roles in Aβ-related pathogenesis. 14 Figure 1.1 The amyloidogenic and non-amyloidogenic pathways of APP processing. The amyloidogenic pathway is initiated by cleaving APP at its β-site by β-secretase to release sAPPβ and leave a C99 anchored to the membrane. Membrane-bound C99 is further cleaved by γ-secretase to generate Aβ peptide and AICD. Under physiological conditions, the majority of APP is first cleaved at α-cleavage site to release the ectodomain sAPPα and leave a membrane-bound C83. Followed by α-cleavage, C83 is further cleaved by γ-secretase to generate p3 and AICD. BACE1 can cleave APP at β’ site within Aβ domain to generate sAPPβ’ and membrane-bound C89, and C89 is further cleaved by γ-secretase to generate truncated Aβ and AICD. A homologue of BACE1, BACE2, cleaves APP at θ site and prevent Aβ production. 1.3.3 Presenilins and γ-secretase The presenilin story started with the discovery of several missense mutations in two newly cloned genes, presenilin 1 (PS1) and presenilin 2 (PS2), from the families of inherited forms of AD. PS1 and PS2 are located on chromosomes 14 and 1, respectively (Alzheimer's Disease Collaborative, 1995; Levy-Lahad et al., 1995; Rogaev et al., 1995; Sherrington et al., 1995). Following these discoveries, De Strooper and his colleagues later demonstrated that γ-cleavage of APP is prevented in PS1-deficient neuronal culture without affecting α- and β- cleavage, resulting in the accumulation of γ-secretase substrate, C83 and C99, and a corresponding 15 decrease of Aβ. This result suggested that PS1 is the major component of γ-secretase complex (De Strooper et al., 1998). Afterward, PS2, a PS1 homolog, was found to be another constituent of the γ-secretase complex based on the observation that Aβ generation is only completely abolished in PS1/ PS2 double knockout cells, and not in single knockouts (Herreman et al., 2000; Z. Zhang et al., 2000). Presenilin is an intramembranous aspartyl protease, and it undergoes autocatalytic cleavage between transmembrane domain 6 and 7 to generate ~27-28 kDa N-terminal and ~16-17 kDa C-terminal derivatives when overexpressed in cell cultures and transgenic mouse brains (Thinakaran et al., 1996). However, its full function still requires three other cofactors, including nicastrin, Aph 1, and Pen-2 (De Strooper, 2003; Edbauer et al., 2003; Kimberly et al., 2003; Takasugi et al., 2003). Coexpression of the four complex components leads to a marked increase of γ-cleavage in mammalian cells (Kimberly et al., 2003) and can reconstitute γ-secretase activity in yeast, which lacks an endogenous γ-secretase (Edbauer et al., 2003). Apart from the essential role of PS in APP processing, it has other substrates, such as Notch. Cultured neuronal cells lacking PS1 display decreased cleavage of a truncated form of Notch which only contains a transmembrane domain and its intercellular domain. Application of an γ-secretase inhibitor also blocks this process of Notch cleavage, which is similar to its effects on APP processing (De Strooper et al., 1999). Song et al. also demonstrated that endoproteolysis of the Notch 1 intracellular domain is substantially inhibited in fibroblasts derived from PS1 knockout mice, whereas overexpression of PS1 rescues this defect (Song et al., 1999). Furthermore, null mutations of presenilin in Drosophilia prevent normal Notch signaling and block its intracellular domain from translocating into the nucleus (Struhl & Greenwald, 1999; Y. 16 Ye, Lukinova, & Fortini, 1999). Collectively, presenilins are not only essential for cleaving APP at the γ-site, but also play an indispensable role in processing other substrates. 1.3.4 The amyloid hypothesis of AD The origins of the amyloid hypothesis can be traced back to when the amino acid sequence of Aβ protein was determined, and the gene encoding Aβ was predicted to be located on chromosome 21 (Glenner & Wong, 1984a, 1984b). Following the successful cloning of the APP gene and the identification of several APP mutations, the amyloid hypothesis was officially proposed by Hardy and Higgins in 1992. It stated that the deposition of Aβ is the principal cause of AD pathogenesis, leading to NFT formation, neuron loss, and memory deficits seen in AD patients (Hardy & Higgins, 1992). Since this hypothesis became the topic of intense investigation, extensive studies have been carried out that provide strong support for several aspects of the amyloid hypothesis. First, various APP mutations found in familial AD cases affect β-cleavage or γ-cleavage to generate more Aβ species or change the ratio of Aβ42/40 (Citron et al., 1992; De Jonghe et al., 2001; Hendriks et al., 1992; Mullan et al., 1992; Suzuki et al., 1994). The pathogenic AD-associated mutations in the Aβ domain were discovered to change the kinetics of Aβ oligomerization as their underlying mechanisms (Nilsberth et al., 2001; Tomiyama et al., 2008). Second, it has long been recognized that DS patients inevitably develop AD-associated neuropathological features, as they carry an extra copy of the APP gene due to the trisomy of chromosome 21 (Mann, Yates, & Marcyniuk, 1984; Motte & Williams, 1989; Olson & Shaw, 1969). Although it is arguable that many genes other than APP located on chromosome 21 could 17 contribute to the development of these pathological events, a rare case of DS with partial trisomy 21 containing only two copies of the APP gene was negative for AD-related neuropathological changes (Prasher et al., 1998). This further emphasizes the central role of APP in the development of AD-associated pathology. Third, converging lines of evidence suggest that presenilin mutations, the most common genetic cause of familial AD, exert their pathogenic effects by increasing the ratio of Aβ42/40 (Bentahir et al., 2006; Kumar-Singh et al., 2006; Scheuner et al., 1996). Fourth, APOE4 carriers, the only confirmed genetic factor in sporadic AD, have a higher risk of developing AD by either decreasing Aβ clearance or promoting Aβ aggregation (Bales et al., 1997; Castellano et al., 2011; Cerf, Gustot, Goormaghtigh, Ruysschaert, & Raussens, 2011; Fagan et al., 2002; Rebeck et al., 1993). Fifth, as another pathological feature of AD, tau is always under debate on whether it has a more dominant role in the progression of AD compared with Aβ. However, the amyloid hypothesis is upheld by the observation that mutations in the gene coding tau protein lead to frontotemporal dementia with parkinsonism instead of AD (Hutton et al., 1998; Poorkaj et al., 1998; Spillantini, Bird, & Ghetti, 1998). Additionally, injection of Aβ fibrils into the brains of P301l tau transgenic mice caused a fivefold increase of NFTs formation, and crossing Aβ-promoting APP transgenic mice with mutant tau transgenic mice markedly elevated NFTs pathologies compared with its parental tau transgenic mice (Gotz, Chen, van Dorpe, & Nitsch, 2001; Lewis et al., 2001). Conversely, application of anti-Aβ antibody to 3xTg-AD mice (containing three AD-associated mutations) not only lowers Aβ load but also substantially reduces tau lesions, strongly suggesting that tau is a downstream effector in the amyloid hypothesis (Oddo, Billings, Kesslak, Cribbs, & LaFerla, 2004). Lastly, the toxicities of various Aβ species were demonstrated both in vitro and in vivo. Exogenous application of Aβ peptides causes free radical damage (Behl, Davis, Lesley, & 18 Schubert, 1994), synaptic degeneration (Lorenzo & Yankner, 1994), and decreased cell viability (Pike, Burdick, Walencewicz, Glabe, & Cotman, 1993; Yankner, Duffy, & Kirschner, 1990) in cell culture. Microinjection of synthetic Aβ or cell medium-derived Aβ into the animal brain induces a decrease of endogenous acetylcholine (Ach) release, inhibition of hippocampal long-term potentiation (LTP), neuronal death, and abnormalities in learning behaviors (Cleary et al., 2005; Geula et al., 1998; Vaucher et al., 2001; Walsh et al., 2002). Collectively, converging lines of evidence support Aβ as the principal driver in the pathogenic processes of AD and as the upstream regulator of tau pathologies, leading to further deteriorating events in the later stages of disease progression. 1.3.5 The challenges of amyloid hypothesis and its revision Although the development of the amyloid hypothesis is a milestone in AD history, it has received long-lasting debate from skeptics and has been challenged by the following observations. First, neuritic plaque load correlates poorly with cognitive decline and severity of dementia (Katzman, 1986; Nelson et al., 2012; Terry et al., 1991). Second, APP transgenic mice showing phenotypes of synaptic disruption and behavioral and memory deficits are devoid of plaque formation (Hsia et al., 1999; Moechars et al., 1999; Mucke et al., 2000). In addition to fibrillary Aβ, other forms of Aβ, such as oligomeric Aβ species, have been consistently detected in various conditions, from cell culture systems to brain samples and human cerebrospinal fluid (Gong et al., 2003; Podlisny et al., 1995; Vigo-Pelfrey, Lee, Keim, Lieberburg, & Schenk, 1993). Aβ oligomers also better correlate with memory decline than plaque numbers (Lue et al., 1999; C. McLean et al., 1999; J. Wang, Dickson, Trojanowski, & 19 Lee, 1999). Due to these observations, the amyloid hypothesis has shifted its focus from Aβ plaques to Aβ oligomers, and is referred to the 'revised amyloid hypothesis' as stated by Hardy and Selkoe (Figure 1.2). Either genetic mutations in familial AD or APOE4-associated mechanisms in sporadic and familial cases contribute to the accumulation of Aβ42 oligomers, which are deemed as pathological initiators triggering a sequential of pathological events and leading to dementia (Hardy & Selkoe, 2002; Selkoe & Hardy, 2016). It has been suggested that plaques are more like the ‘inert pool’ of Aβ species, and are even deemed as protective for cells to circumvent toxic oligomers, which is opposite to the earlier findings that fibrillary Aβ induced cellular toxicity (Lorenzo & Yankner, 1994; Pike et al., 1993; Yankner et al., 1990). These conflicting reports cannot be resolved with certainty for now, but accumulating evidence supports the idea that Aβ oligomers manifest detrimental effects both in vitro and in vivo, as discussed in Section 1.3.7. 20 Figure 1.2 Major pathogenic events leading to AD proposed by the ‘revised amyloid hypothesis’. The ‘revised amyloid hypothesis’ was brought about by Dr. Selkoe and Dr. Hardy in 2016 (Selkoe & Hardy, 2016). It incorporates the recent findings of Aβ oligomers. Either genetic mutations in familial AD or APOE4-associated mechanisms in sporadic and familial cases contribute to the accumulation of Aβ42 oligomers, which are deemed as pathological initiators triggering a sequential of pathological events and leading to dementia. Aβ42 oligomers may directly result in tangle formation in addition to microglia and astrocytes activation, as shown by the dash arrow. 1.3.6 The effects of APPSWE mutation on AD pathogenesis In 1992, a double mutation at codons 670 and 671 in exon 16 of APP (KM670/671NL) was discovered in two Swedish AD families with an average age onset of 55; it is now referred to as the APPSWE mutation (Mullan et al., 1992). The clinical manifestations of the Swedish familial AD cases resemble that of the sporadic cases, suggesting that they share common molecular pathways leading to disease (Schellenberg & Montine, 2012). Since then, the pathogenic 21 mechanisms of APPSWE mutation have become a topic of intense investigation in hopes that the results can lead to therapeutic strategies to treat not only familial AD but also sporadic AD. Since the Swedish double mutation is located outside of the N-terminal of Aβ sequence, it was originally proposed that this mutation could affect APP cleavage and Aβ generation. Indeed, in vitro studies unequivocally indicate that the APPSWE mutation promotes the proteolytic processing of APP through the amyloidogenic pathway. It has been demonstrated by various groups that cultured cells bearing APPSWE mutation produce 6-8 folds Aβ and decrease the release of p3 in the cell culture medium compared with cells with wildtype APP (Cai, Golde, & Younkin, 1993; Mullan et al., 1992). Consistent with these results, primary skin fibroblasts from symptomatic and presymptomatic AD patients generate 3-fold more extracellular Aβ than fibroblasts from non-affected subjects (Citron et al., 1994). Moreover, our lab has demonstrated that while wildtype APP is primarily cleaved by BACE1 at Glu 11, the APPSWE mutation shifts the BACE1 cleavage site from Glu11 to Asp1, thereby promoting C99 fragment and Aβ generation (Deng et al., 2013). Regarding cytotoxicity, APPSWE-overexpressing PC12 cells undergo more cell death in response to hydrogen peroxide treatment through apoptotic mechanisms as indicated by elevated levels of caspase-3 cleavage and activation of JNK signaling pathway (Eckert, Steiner, Marques, Leutz, Romig, Haass, & Muller, 2001; Marques et al., 2003). To better explore the pathogenic mechanisms caused by APPSWE mutation, several lines of transgenic mice have been developed to overexpress Swedish mutation-bearing APP protein (Folkesson et al., 2007; Hsiao et al., 1996; Sturchler-Pierrat, Abramowski, Duke, Wiederhold, 22 Mistl, Rothacher, Ledermann, Bürki, et al., 1997). It was reported that Tg2576 mice, which overexpress human APPSWE mutation driven by a hamster prion protein promoter, manifest memory impairment by 9-10 months with a concurrent substantial increase of both Aβ40 and Aβ42. An abundance of extracellular neuritic plaques positive for Congo red staining is detected at around 12 months (Hsiao et al., 1996). Inconsistent results have been reported for the deficits in the cholinergic neurotransmitter system without overt neuron loss (Apelt, Kumar, & Schliebs, 2002; Gau et al., 2002; Lüth, Apelt, Ihunwo, Arendt, & Schliebs, 2003). APP23, a transgenic mouse model overexpressing 7-fold of APP protein harboring Swedish mutation under the murine Thy1 promoter, is reported to have an age-dependent increase of Aβ and neuritic plaques starting at 6 months (Sturchler-Pierrat, Abramowski, Duke, Wiederhold, Mistl, Rothacher, Ledermann, Bürki, et al., 1997). A subsequent study confirmed that neuron loss is only detected in the CA1 hippocampal region of APP23 mice aged 14-18 months (Calhoun et al., 1998). Additionally, spatial learning memory is impaired in 24-month-old APP23 mice (Dumont, Strazielle, Staufenbiel, & Lalonde, 2004). However, the phenotypes observed in the APPSWE transgenic mice are challenged by a recent novel APPSWE knock-in mouse line (APPNL/NL). In this line, both Aβ40 and Aβ42 were modestly increased with no detection of neuritic plaques even at 2 years, suggesting the knock-in mouse model displays more subtle phenotypes (Masuda et al., 2016; Saito et al., 2014). Taken together, the role of APPSWE in enhancing Aβ generation is consistent in findings from in vivo and in vitro, while its cytotoxic effects observed in cell culture has scarcely been detected in transgenic mouse models. 23 1.3.7 The toxicity of Aβ oligomer Although Aβ oligomer is a well-accepted concept in the AD field, the term is ill-defined and refers to heterogeneous Aβ species based on different preparation methodology. There are three methods to produce Aβ oligomers: synthetic Aβ peptide-derived, cell culture-derived, and brain-derived oligomeric assemblies (Benilova, Karran, & De Strooper, 2012). During the process of Aβ fibrillogenesis using synthetic Aβ peptide, an intermediate Aβ species with distinct properties from fibrils was identified and referred to as protofibrils (PFs) (Harper, Lieber, & Lansbury, 1997; Harper, Wong, Lieber, & Lansbury, 1997; Walsh, Lomakin, Benedek, Condron, & Teplow, 1997). Such PFs increase cell death in cortical culture as determined by a cytotoxicity assay, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (Walsh et al., 1999); this is further validated by the findings that PFs change the electrical activities of cortical neurons and decrease cell viability in a time- and concentration- dependent manner (Hartley et al., 1999; Isaacs, Senn, Yuan, Shine, & Yankner, 2006; Ye, Selkoe, & Hartley, 2003). Similarly, small diffusible Aβ oligomers, called Aβ-derived diffusible ligands (ADDLs), are recognized for their ability to directly kill neurons in organotypic cultures and inhibit LTP in hippocampal slices (Lambert et al., 1998; Wang et al., 2002). In cell culture, low molecular weight Aβ oligomers are detected in 7PA2 cells (Chinese hamster ovary (CHO) cells that stably overexpress APPV717F mutant), and these oligomers are stable in sodium dodecyl sulfate (SDS) (Morishima-Kawashima & Ihara, 1998; Podlisny et al., 1995; Walsh, Tseng, Rydel, Podlisny, & Selkoe, 2000; Xia et al., 1997). Cerebral injection of 7PA2 24 cell medium containing SDS-stable Aβ oligomers into rats causes LTP inhibition, cognitive dysfunction, and reversible synaptic degeneration without apparent cytotoxicity (Cleary et al., 2005; Shankar et al., 2007; Townsend, Shankar, Mehta, Walsh, & Selkoe, 2006; Walsh et al., 2005; Walsh et al., 2002). In the case of brain-derived Aβ oligomers, AD brain-derived dimers introduce cytoskeletal disruption and neurodegeneration in hippocampal neurons (Jin et al., 2011), and result in reduced LTP, enhanced long-term depression (LTD), and decreased dendritic spine density in rat hippocampus (Shankar et al., 2008). Another study claims that a 56-kDa soluble Aβ species (Aβ*56) was purified from APPSWE transgenic mice. Aβ*56 is responsible for memory impairment in middle-aged transgenic mice and introduces memory deficits when infused into wildtype rats (Lesne et al., 2006). The existence of Aβ*56 is also confirmed in other transgenic mouse lines either expressing wildtype or Arctic mutant APP (Cheng et al., 2007). Overall, Aβ oligomers render deleterious effects on cell viability, synaptic transmission, and memory functions in vitro and in vivo. However, the toxic profiles of Aβ oligomers are not consistently reported by different investigators due to the methodological differences to prepare them. 1.4 The role of α-synuclein (αSyn) in PD αSyn, the major component in the LBs and LNs, is a presynaptic protein with a molecular weight of 14.4 kDa and accounts for almost 1% of total protein in the brain (Jakes, Spillantini, & Goedert, 1994; Uéda et al., 1993; Uéda, Saitoh, & Mori, 1994). The synuclein family also 25 consists of the other two members, β-synuclein and γ-synuclein, with the sequences of 134 and 127 amino acids, respectively (Ji et al., 1997; Nakajo, Tsukada, Omata, Nakamura, & Nakaya, 1993). The central region of αSyn protein was first identified in the amyloid plaques of AD patients as a non-Aβ component (NAC), and its full-length sequence of 140 amino acids was discovered in the same study (Uéda et al., 1993). Further details of αSyn's structural features have been revealed by the following studies. 1.4.1 The structure of αSyn protein αSyn protein has an amphipathic amino-terminal, a central hydrophobic region (comprising NAC), and an acidic carboxy-terminal (Lashuel, Overk, Oueslati, & Masliah, 2013). Purified NAC self-aggregates in a time-, temperature-, and concentration-dependent manner in aqueous solutions and forms fibril-like structures with dye-binding properties (Iwai, Yoshimoto, Masliah, & Saitoh, 1995). A 12-amino acid sequence within NAC is critical for fibrillogenesis, as the deletion of this region abolishes filament assembly and is resistant to protease digestion, suggesting NAC is required and sufficient for its aggregation (Giasson, Murray, Trojanowski, & Lee, 2001). Concerning its N-terminal, there is a consensus KTKEGV motif similar to the lipid-binding domain of apolipoproteins and it is capable of binding to phospholipid vesicles (Davidson, Jonas, Clayton, & George, 1998; Eliezer, Kutluay, Bussell, & Browne, 2001). The acidic C-terminal is largely disordered without secondary structure (Eliezer et al., 2001; Ulmer, Bax, Cole, & Nussbaum, 2005), but it appears to affect the propensity of αSyn to aggregate (Crowther, Jakes, Spillantini, & Goedert, 1998; Murray et al., 2003) 26 1.4.2 Physiological functions of αSyn With respect to the cellular functions of αSyn, early studies in songbirds found that this protein was upregulated during a critical period for song learning and it was particularly enriched in synaptic vesicles, suggesting a possible role in modulating synaptic plasticity and vesicular functions (George, Jin, Woods, & Clayton, 1995; Iwai, Masliah, et al., 1995). Consistent with this idea, converging evidence supports αSyn as a key molecule in presynaptic sites. Purified αSyn from bacteria is a potent inhibitor of phospholipase D2 activity in vitro in a concentration-dependent manner, and PLD2 is implicated in signal transduction, vesicular transportation, and cytoskeleton dynamics. (Jenco, Rawlingson, Daniels, & Morris, 1998). Additionally, accumulating data suggest that αSyn interacts with other synaptic proteins, such as Ras-related proteins in brain (Rab) small GTPase Rab3a (Dalfo & Ferrer, 2005) and prenylated rab acceptor protein 1 (PRA1) (Lee, Kang, Lee, & Im, 2011). Furthermore, αSyn can bind to synaptobrevin-2 and promote the assembly of the soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complex, which is implicated in neurotransmitter release (Burre et al., 2010). In addition to in vitro experiments, studies focusing on αSyn animal models provide further knowledge. Genetic deletion of αSyn in mice was shown to increase the recovery rate of dopamine release in paired pulse induced-stimulation at nigrostriatal terminals, suggesting αSyn works as a negative regulator of the readily releasable pool of dopaminergic vesicles (Abeliovich et al., 2000). Suppressing the expression of αSyn by antisense oligonucleotides resulted in a significant decrease of the distal reserve pool of vesicles in cultured hippocampal neurons (Murphy, Rueter, Trojanowski, & Lee, 2000). Consistent with this observation, αSyn knock-out mice display impaired synaptic responses to repeated stimulation in the hippocampal slices due 27 to depletion of the vesicles in both docking and reserve pools, supporting the role of αSyn in the generation and maintenance of the reserve pool of vesicles (Cabin et al., 2002). In contrast, transgenic mice overexpressing human αSyn have enlarged synaptic vesicles at transgenic boutons and significant deficits in neurotransmitter release. The expression of critical endogenous presynaptic proteins involved in endocytosis and exocytosis is severely impaired in αSyn-overexpressing neurons, demonstrating that elevated αSyn causes synaptic dysfunction (Scott et al., 2010). Another study also found that increased αSyn expression results in decreased neurotransmitter release by reducing the size of the synaptic vesicle recycling pool (Nemani et al., 2010). A large wealth of data unequivocally points to the essential role of αSyn in vesicle transportation, neurotransmitter release, and synaptic plasticity. αSyn also interacts extensively with various components of the dopaminergic neurotransmitter system and this will be discussed in Section 1.5.5. 1.4.3 SNCAA53T mutation and its effects on PD pathogenesis The discovery of SNCA as the first gene linked to familial PD caused a seismic shift in the field, as PD was originally thought to be a sporadic disease. In 1997, a missense mutation was discovered in the coding sequence of the SNCA gene in familial PD patients of Italian origin where guanine in position 209 was replaced by adenine resulting in a substitution of Ala to Thr at position 53 (A53T) in αSyn protein (Polymeropoulos et al., 1997). Since the discovery of the SNCAA53T mutation, scientists have delved into elucidating the mechanisms underlying its pathogenic effects on PD. αSyn protein has long been recognized as a ‘natively unfolded’ monomer, unless it is bound to lipids to acquire a α-helical secondary 28 structure (Davidson, Jonas, Clayton, & George, 1998; Weinreb, Zhen, Poon, Conway, & Lansbury, 1996). However, this view has been challenged by several recent findings that physiological αSyn exists as a helically folded tetramer and this conformation is resistant to aggregation (Bartels, Choi, & Selkoe, 2011; Dettmer, Newman, Luth, Bartels, & Selkoe, 2013; Wang et al., 2011). Although the conformation of naturally existing αSyn is still under debate, SNCAA53T mutation was consistently reported to accelerate the processes of oligomerization and fibrillization. Recombinant αSyn A53T mutant forms mature fibrils containing β-sheet structure more quickly than both wildtype and A30P mutant in in vitro fibrillization assay (Conway, Harper, & Lansbury, 1998; El-Agnaf, Jakes, Curran, & Wallace, 1998; Giasson, Uryu, Trojanowski, & Lee, 1999; Narhi et al., 1999). Another in vitro study has shown that αSyn A53T mutant and an equimolar mixture of A53T and wildtype (WT) fibrillize faster than WT, while A30P and an equimolar mixture of A30P and WT fibrillize more slowly than WT. However, both A53T and A30P monomers are consumed more rapidly than WT. This suggests that acceleration of oligomerization, rather than fibrillization, is a shared property of αSyn mutants (Conway et al., 2000). A recent study revealed that all PD-linked SNCA mutants tested in the study, including A53T, A30P, E45K, G51D, and H50Q, significantly decreased the ratio of monomer to tetramer αSyn as detected by immunoblotting followed by intact cell crosslinking. This indicates that shifting from tetramer to monomer is the common pathogenic effect of αSyn mutations for the initiation of PD pathogenesis, but it is a tentative conclusion until further validation (Dettmer et al., 2015). In contrast to the consistent findings of SNCAA53T mutation’s effect on oligomerization, whether the mutation is sufficient to cause neuronal death is still under debate, especially when taking in 29 vivo data into consideration. When SNCAA53T mutation was overexpressed in the neuroblastoma BE-M17 cell line, it only increased cell vulnerability to iron but not to hydrogen peroxide compared with wildtype and A30P αSyn, suggesting that cells with SNCAA53T are more sensitive to iron-mediated toxicity (Ostrerova-Golts et al., 2000). However, overexpression of SNCAA53T in NT-2/D1 and SK-N-MC cells resulted in elevated cell death in response to all tested insults (including hydrogen peroxide, serum withdrawal, and protease inhibitors) compared with wildtype αSyn, but this increase was not apparent when comparing with A30P mutation (Lee, Hyun, Halliwell, & Jenner, 2001). Additionally, another study demonstrated that PC12 cells stably overexpressing SNCAA53T caused more cell death than overexpressing wildtype through nonapoptotic mechanisms (Stefanis, Larsen, Rideout, Sulzer, & Greene, 2001). The impact of SNCAA53T mutation on cytotoxicity has been examined in vivo by generating SNCAA53T transgenic animal models. When human SNCAA53T is expressed in Drosophila, it produces the essential phenotypes of PD: adult-onset dopaminergic neuron loss, intracellular inclusions, and motoric dysfunction (Feany & Bender, 2000). Additionally, in the Caenorhabditis elegans transgenic model, even though both wildtype and SNCAA53T mutation manifest neurotoxicity in the dopaminergic system, there is no difference in neuron loss in the SNCAA53T transgenic Caenorhabditis elegans compared with its wildtype (Kuwahara et al., 2006; Lakso et al., 2003). Regarding the different lines of transgenic mice, only one line carrying SNCAA53T mutation under the control of murine prion promoter (line G2-3) exhibits adult-onset neurodegeneration, but not for wildtype and SNCAA30P mutation, suggesting that the A53T mutant has the most severe in vivo toxicity (Lee et al., 2002). However, other transgenic mice (PrPmtB and M83) expressing human SNCAA53T driven by the murine prion promoter 30 consistently show motor deficits with varying effects on neurodegeneration and LBs-like inclusion formation (Giasson et al., 2002; Gispert et al., 2003). SNCAA53T transgenic mice driven by the tyrosine hydroxylase (TH) promoter lack αSyn-positive intracellular inclusions and neuropathology in the nigrostriatal pathway, and do not show overt neuron loss (Matsuoka et al., 2001). However, SNCAA53T transgenic mice controlled by the Thy1 promoter develop α-synucleinopathy, neuronal degeneration, and motor deficits (van der Putten et al., 2000). Taken together, the role of SNCAA53T in facilitating oligomerization is consistent from in vitro studies, but different lines of SNCAA53T transgenic mice display motor symptoms with discrepancies in inclusion formation and neurodegeneration. 1.4.4 αSyn oligomer and its toxicity Although the presence of LBs in biopsy samples is considered to be the neuropathological hallmark of PD diagnosis, LBs are not exclusive to PD nor present in all PD patients (Houlden & Singleton, 2012). Moreover, neuronal death in PD animal models does not consistently correlate with LB-pathology (Kirik et al., 2003; Lo Bianco, Ridet, Schneider, Deglon, & Aebischer, 2002; Maries, Dass, Collier, Kordower, & Steece-Collier, 2003). Finally, it has been suggested that LBs could be protective entities, as an increasing number of LBs correlates with lower neurotoxicity and LBs-containing neurons are healthier (Chen & Feany, 2005; Tompkins & Hill, 1997). To better resolve this controversy, the concept of αSyn oligomers has emerged, and its importance in the progression of PD has been increasingly appreciated, sparking great interest in exploring potentially valuable therapeutics. 31 By employing different protocols for producing recombinant αSyn oligomers in vitro, it has been demonstrated that αSyn oligomers are heterogeneous species. Oligomers with annular structures exert cytotoxicity by disrupting calcium homeostasis via pore-forming mechanisms (Danzer et al., 2007), which is consistent with an earlier study that suggested αSyn oligomers carrying A30P or A53T mutations have greater permeabilizing activities than the wildtype protein (Volles & Lansbury, 2002). A more recent study that precisely monitored αSyn assembly process revealed that the formation of condensed proteinase K-resistant oligomers from initial oligomers is the critical step and produces more oxidative stress than the initial oligomers (Cremades et al., 2012). Additionally, αSyn oligomers prepared in vitro are suggested to impair α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA)-receptor neuronal transmission and LTP (Diogenes et al., 2012; Huls et al., 2011). With the advancement of techniques to visualize oligomerization via bimolecular fluorescence complementation assay, the toxicity of αSyn oligomers formed in cells is supported by the increased release of adenylate kinase from damaged cells (Outeiro et al., 2008; Tetzlaff et al., 2008). More recent evidence from in vivo models showed that overexpression of different αSyn variants, either oligomer-forming mutations or fibril-promoting mutations, in the substantial nigra (SN) of rats by viral injection induces either severe dopaminergic neuronal demise or no significant toxic effects (Winner et al., 2011). Although there is ample evidence unequivocally indicating the toxicity of αSyn oligomers, the mechanisms underlying these deleterious effects are still vague. 1.4.5 The secretion and transmission of αSyn The longstanding view of αSyn as a cytoplasmic protein has been challenged with the finding that it can also exist in the extracellular space. αSyn has been detected in the conditioned 32 medium of neuroblastoma cells either with or without SNCA overexpression, in human cerebrospinal fluid (CSF), and in the plasma of PD patients and healthy subjects (Borghi et al., 2000; El-Agnaf et al., 2003). In the cellular models, release of αSyn is thought to be mediated by exocytosis; however, this process is different from the classic exocytotic pathways as it is not sensitive to endoplasmic reticulum (ER)-Golgi transport blocker (Jang et al., 2010; Lee, Patel, & Lee, 2005). In addition to these discoveries, case reports from PD patients who received mesencephalic tissue transplantation also provide evidence for supporting the transmission of αSyn. In the postmortem brains of these patients, the grafted nigral neurons were positive for LBs-like inclusions, staining for αSyn, ubiquitin, and thioflavin-S after one or two decades of transplantation; this strongly implies host-to-graft transmission of αSyn pathology (Kordower, Chu, Hauser, Freeman, & Olanow, 2008; Kordower, Chu, Hauser, Olanow, & Freeman, 2008; Kurowska et al., 2011; Li et al., 2008). The host-to-graft transmission phenomenon is also observed in αSyn transgenic mice (Desplats et al., 2009; Hansen et al., 2011). Based on these observations, scientists started to speculate on the route of interneuronal transmission of αSyn after it has been secreted extracellularly. Endocytosis is a proposed mechanism for oligomeric and fibrillar αSyn species, while monomeric αSyn is more likely to translocate across the plasma membrane directly (Ahn, Paik, Chung, & Kim, 2006; Lee, Suk, Bae, Lee, et al., 2008). Another important component of αSyn transmission is to define the seeding properties of αSyn, a well-known concept in the study of prion disease. It has been reported that extracellular application of αSyn oligomers prepared in vitro caused transmembrane seeding of αSyn 33 aggregates in neuronal cells or primary neurons in a dose- and time- dependent manner (Danzer, Krebs, Wolff, Birk, & Hengerer, 2009; Volpicelli-Daley et al., 2011). Another study reported that intrastriatal inoculation of fibrillar αSyn results in spreading of αSyn-related pathology in non-transgenic wildtype mice (Luk, Kehm, Carroll, et al., 2012). More recent studies have revealed that inoculation of brain homogenates from old symptomatic αSyn transgenic mice induces the formation of LBs/LNs-like inclusion in young healthy αSyn transgenic mice and accelerates neurological dysfunction (Luk, Kehm, Zhang, et al., 2012; Mougenot et al., 2012). Collectively, the secretion and transmission of αSyn is a consistent finding from these enlightening seminal studies, although the transmission routes and the definition of αSyn ‘seeds’ mediating its propagation are far from being well understood. 1.5 Selective neurodegeneration in AD and PD Both AD and PD are neurodegenerative disorders, defined by progressive neuron loss and degeneration within specific neurotransmitter systems and thereby manifesting related clinical symptoms due to deficiency of normal neurotransmission (Bartus et al., 1985; Greffard et al., 2006). Additionally, certain misfolded proteins, namely Aβ in AD and αSyn in PD, exist as the prominent neuropathological features (Glenner & Wong, 1984a, 1984b; Spillantini et al., 1997). Although such selectivity of neuronal vulnerability is poorly understood, extensive evidence has deepened our insights and broadened our views on how such potentially pathogenic proteins interact with vulnerable neurons. 34 1.5.1 Selective neurodegeneration in AD AD is a well-known neurodegenerative disorder, characterized by progressive and prominent neurodegeneration within cholinergic neurons in the basal forebrain and dysfunction of cholinergic transmission in the hippocampus and cerebral cortex (Bartus, Dean, Pontecorvo, & Flicker, 1985). Although the deficits are not limited to the cholinergic system, its functional disturbance is closely related to memory impairment, a major symptom seen in AD patients (Aletrino, Vogels, Van Domburg, & Ten Donkelaar, 1992; Francis et al., 1985; Halliday et al., 1992; Marcyniuk, Mann, & Yates, 1986; Perry et al., 1981). Since the human brain samples from AD patients have been available for examination, profound cell loss has been consistently identified in the basal forebrain cholinergic system and severe loss of cortical cholinergic innervation has been detected in an advanced stage of AD (Geula & Mesulam, 1989; Mesulam, 2004). From an anatomical perspective, the basal forebrain cholinergic system is composed of four cholinergic nuclei, including the medial septum (MS) nucleus, the horizontal and vertical limb of the diagonal band of Broca (DBB), and the nucleus basalis of Meynert (NBM). Specifically, the cerebral cortex mainly receives cholinergic innervation from the NBM, whereas hippocampus receives innervation from MS and vertical limb of DBB (Mesulam & Geula, 1988; Mesulam, Mufson, Levey, & Wainer, 1983). The selectivity of cholinergic neurodegeneration within the cholinergic basal forebrain (CBF) complex in AD is supported by several observations. The severity of the cholinergic denervation in the cerebral cortex is not universal, with the most severe in temporal lobes, including the entorhinal cortex (Mesulam, 2004). It has been reported that there is an average of 55% loss of 35 cortical cholinergic fibers in AD brain samples, while the temporal lobe is almost depleted of any fibers (Geula & Mesulam, 1996). Within the hippocampus, the most drastic neuron loss is situated in the CA1 region of AD patients (Padurariu, Ciobica, Mavroudis, Fotiou, & Baloyannis, 2012; West, Coleman, Flood, & Troncoso, 1994). Within the basal forebrain, the most significant cholinergic loss is detected in the NBM (Candy et al., 1983; Whitehouse et al., 1981). Besides CBF, other cholinergic nuclear groups located in the brain stem and striatum are relatively spared in AD patients (Woolf, Jacobs, & Butcher, 1989). As the initial fundamental findings, three independent British groups discovered that the activity of choline acetyltransferase (ChAT), the enzyme for synthesizing Ach, was impaired in AD cortex and hippocampus compared to the age-matched controls (Bowen, Smith, White, & Davison, 1976; Davies & Maloney, 1976; Perry, Perry, Blessed, & Tomlinson, 1977). Due to great advancements in basic neuroscience research, the limbic cholinergic pathway has since been revealed and it is now known that the pathological changes mentioned above are more confined to the postsynaptic sites innervated by cholinergic neurons. Ongoing studies reveal further alterations of cholinergic neurons at the presynaptic sites in AD, including marked neuron loss in the CBF (Candy et al., 1983; Whitehouse et al., 1981), reduced Ach synthesis (Sims et al., 1983; Sims et al., 1980), and decreased high affinity choline uptake (Rylett, Ball, & Colhoun, 1983). Furthermore, additional mechanisms of neurodegeneration have been discovered, including cholinergic receptor dysfunction, impaired postsynaptic cholinergic responses (Lang & Henke, 1983; Rinne, Rinne, Laakso, Paljarvi, & Rinne, 1984), and lowered density of muscarinic and nicotinic acetylcholine receptors (AchRs) of varying subtypes (London, Ball, & Waller, 1989; Mash, Flynn, & Potter, 1985). Taken together, cholinergic neurodegeneration involves 36 both presynaptic and postsynaptic mechanisms, but the findings on AchRs deficits are relatively controversial. 1.5.2 Cholinergic hypothesis of AD The development of the cholinergic hypothesis was based on more understanding of the neural circuits in basic neuroscience, early pharmacological studies, and seminal observations in aged and AD patients. It proposes that the disturbance of normal cholinergic function in the brain predominantly contributes to the memory loss and cognitive impairments seen in AD patients and the elderly, and enhancing cholinergic function can significantly alleviate these cognitive symptoms (Bartus, Dean, Beer, & Lippa, 1982; Bartus et al., 1985). Although this was greatly challenged after being postulated, nevertheless it propelled the application of acetylcholinesterase (AChE) inhibitors as a treatment for AD in the clinic and it is still the most validated treatment available in the market. During the 1960s to 1970s, researchers identified the cholinergic projection from basal forebrain to the cortex and related limbic system (Lewis & Shute, 1967; Shute & Lewis, 1967). Concurrent with these fundamental findings, it was reported that when young subjects were treated with an anticholinergic drug, namely scopolamine, it produced dementia symptoms seen in the elderly (Drachman, 1977; Drachman & Leavitt, 1974). Likewise, other studies observed a similar phenomenon in treated monkeys (Bartus & Johnson, 1976). More importantly, with the help of cholinomimetic therapy, memory impairment in aged monkeys and patients were partially improved after receiving physostigmine, an anticholinesterase drug, although the responsive dosage varied to a great extent (Bartus, 1978, 1979; Davis, Mohs, & Tinklenberg, 37 1979). Studies in post-mortem brains dating back to as early as 1964 have found that the activity of AChE was decreased in AD cases (Pope et al., 1964). Three successive studies from British groups reported a 50% ~ 80% reduction of ChAT in the hippocampus and cerebral cortex of AD patients, marking the beginning of a cholinergic era in the AD field (Bowen et al., 1976; Davies & Maloney, 1976; Perry et al., 1977). Subsequent research focusing on the cholinergic system revealed further details on the cholinergic dysfunction in AD, as discussed in the previous section. Although the cholinergic hypothesis received lots of credit in the early years of AD research, a remarkable convergence of studies has casted doubts on its validity, especially after the advent of the amyloid hypothesis. The clinical application of the cholinesterase inhibitor to treat AD is based on the cholinergic hypothesis, but these remedies are only modestly effective (Courtney et al., 2004; Schneider, Insel, Weiner, & Alzheimer's Disease Neuroimaging, 2011; Trinh, Hoblyn, Mohanty, & Yaffe, 2003). Moreover, results from patients in the prodromal stage of AD or patients with mild cognitive impairment (MCI) support the notion that cholinergic lesion is questionable in early AD. There was no significant loss of ChAT activity in the brain samples from these subjects (Davis et al., 1999; Tiraboschi et al., 2000) or even increased activities of ChAT in their brains (DeKosky et al., 2002). Furthermore, no noticeable or only mild decrement of AChE activity was detected in patients with MCI (Davis et al., 1999; Rinne et al., 2003). Although the importance of the cholinergic hypothesis has recently been taken over by the amyloid hypothesis, nevertheless it laid the foundation for treating AD. It is more likely that cholinergic deficits are secondary to Aβ-related pathology, but further illumination of molecular events from cholinergic 38 neurodegeneration to memory dysfunction will give us more hints to search for more effective symptom-relieving treatments. 1.5.3 Interaction between Aβ and cholinergic neurons As mentioned before, the cholinergic hypothesis and amyloid hypothesis are the two successive major hypotheses in the journey of AD research. Extracellular neuritic plaques composed of Aβ and severe neuron loss within the cholinergic system are two prominent neuropathological features in AD (Hyman et al., 2012). It is therefore logical to explore the interaction between Aβ and cholinergic neurons in order to explain the mechanisms of AD development. Not only can cholinergic neurons modulate APP processing and therefore Aβ generation, but Aβ can also interact with various aspects of cholinergic machinery (Govoni et al., 2014; Kar, Slowikowski, Westaway, & Mount, 2004; Pakaski & Kalman, 2008). With respect to the role of cholinergic receptors in regulating APP processing, Nitsch et al. first demonstrated that HEK293 cells overexpressing the m1/m3 acetylcholine receptor had increased release of sAPPα and a corresponding decrease of Aβ generation after being stimulated by carbachol, a non-selective muscarinic acetylcholine receptor (mAchR) activator. These alterations were blocked by staurosporine, indicating the involvement of PKC (Nitsch, Slack, Wurtman, & Growdon, 1992). Carbachol’s effect on APP processing is consistent, but other signaling transduction pathways also have been suggested, such as the phospholipase C signaling pathway (Wolf et al., 1995). Moreover, it has been reported that in the 3xTg AD animal model, an m1AchR agonist, AF267B, decreases BACE1 level, increases ADAM 10 and 17, reduces Aβ burden, and ameliorates tau pathology, whereas dicyclomine, an m1AchR antagonist, produces 39 the opposite effects. These results suggest that m1AchR stimulation shifts the amyloidogenic processing of APP to the non-amyloidogenic pathway (Caccamo et al., 2006). Besides muscarinic AchRs, it has been consistently found that nicotine treatment elevates sAPPα and decreases Aβ in the conditioned cell medium (Kim et al., 1997; Lahiri et al., 2002; Mousavi & Hellström-Lindahl, 2009), and reduces Aβ42-positive plaques by more than 80% in APPSWE transgenic mice (Nordberg et al., 2002). Several studies further demonstrated that α7 nicotinic acetylcholine receptors (nAchRs) activation enhanced α-cleavage of APP processing and diminished Aβ production. Collectively, activation of cholinergic transmission by various receptors is more likely to drive APP processing through the non-amyloidogenic pathway. Aβ modulates the cholinergic system at multiple sites, including Ach synthesis, release, and cholinergic neurotransmission at the postsynaptic sites through nAchRs and mAchRs. In cell culture or primary neuron culture, application of exogenous Aβ peptides reduces ChAT activity (Pedersen & Blusztajn, 1997; Pedersen, Kloczewiak, & Blusztajn, 1996), decreases high-affinity choline uptake (Kar et al., 1998; Kristofikova, Tejkalova, & Klaschka, 2001), and diminishes intracellular Ach concentration (Hoshi et al., 1997; Pedersen & Blusztajn, 1997; Pedersen et al., 1996). A relatively low concentration of Aβ peptides, ranging from nM to μM, is sufficient to inhibit Ach release in hippocampal slice culture and synaptosome (Kar et al., 1998; Kar, Seto, Gaudreau, & Quirion, 1996; Satoh et al., 2001; Vaucher et al., 2001). Injecting Aβ peptides into the CBF of rat brains only induces cholinergic neuron loss with reduced activity of AChE, but has no effect on the parvalbumin-containing neurons located in the same brain region, suggesting neuronal-type specific effects. In line with these findings, local injection of Aβ peptides in rat brains results in a marked decrease of AChE and ChAT activity and the number of mAchRs. 40 However, Aβ1-40 or Aβ25-35 peptides have been shown to stimulate AChE activity in P19 embryonic carcinoma cells, and scrambled Aβ25-35 peptides do not have any effect on AChE (Sberna, Saez-Valero, Beyreuther, Masters, & Small, 1997). So far, there is no agreement on Aβ’s effects on nAchRs and mAchRs, with most studies focusing on α7AchR and m1AchR. α7AchR is highly expressed in the CBF and modulate Ach release (Breese et al., 1997; Perry, Court, Johnson, Piggott, & Perry, 1992). Aβ and α7AchR are present in the neuritic plaques, and Aβ is capable of binding to α7AchR with high affinity in co-immunoprecipitation assays (Wang, Lee, D'Andrea, et al., 2000), which have been confirmed by two other studies (Wang, Lee, Davis, & Shank, 2000; Wang et al., 2000). Intriguingly, it has been reported that Aβ facilitates the neurotransmission through α7AchR (Dineley et al., 2001; Fodero et al., 2004) and that it also works as an antagonist for this receptor (Grassi et al., 2003; Pettit, Shao, & Yakel, 2001). A more recent study suggests that Aβ can form a stable and soluble complex (BAβACs) with AChE, butyrylcholinesterase (BuChE), and ApoE proteins, and such complex facilitates the catalyzing capabilities of both AChE and BuChE (Kumar, Nordberg, & Darreh-Shori, 2016). Exposing cortical neurons to nontoxic levels of Aβ25-35 and Aβ1-40 impairs carbachol-induced activation of G proteins through mAchRs in a concentration- and time- dependent manner (Kelly et al., 1996). Thus, the available data suggests Aβ not only negatively affects cholinergic function but also modulates its signal transmission mostly through postsynaptic receptors. 41 1.5.4 Selective neurodegeneration in PD The importance of the nigrostriatal dopaminergic pathway was valued in the history of PD studies based on these converging lines of evidence: 1) pharmacological depletion of striatal dopamine by reserpine produced parkinsonism (Carlsson, 1959); 2) restoration of dopaminergic system by levodopa relieves motor symptoms in PD (Birkmayer & Hornykiewicz, 1961; Cotzias, Papavasiliou, & Gellene, 1969); 3) the nigrostriatal dopaminergic pathway became more well-charaterized (Anden et al., 1964; Dahlstroem & Fuxe, 1964; Hornykiewicz, 1963); and 4) there were a number of novel findings concerning alterations of biochemical and molecular markers in nigrostriatal dopaminergic pathway in PD (Bernheimer, Birkmayer, Hornykiewicz, Jellinger, & Seitelberger, 1973; Ehringer & Hornykiewicz, 1960; Hornykiewicz, 1963, 1998). A study published in 1960 first reported severe dopamine deficiency in the striatum of at the advanced stage of PD (Ehringer & Hornykiewicz, 1960). With the identification of the nigrostriatal pathway, dopaminergic defects were traced back to cell bodies in the SN which innervates the striatum (Anden et al., 1964; Bernheimer, Birkmayer, Hornykiewicz, Jellinger, & Seitelberger, 1973; Hornykiewicz, 1963). It has been suggested that around 50% of nigral neurons and 80% of striatal dopamine is lost by the time when clinical signs appear (Bernheimer et al., 1973; Fearnley & Lees, 1991; Marsden, 1990). Apart from the great decrease of striatal dopamine, L-aromatic amino acid decarboxylase (AADC) is also significantly reduced in PD patients. Similarly, levels of homovanillic acid (a major dopamine metabolite) and TH (the rate-limiting enzyme in dopamine synthesis) are also decreased, albeit to a lesser extent (Lloyd, Davidson, & Hornykiewicz, 1975). Furthermore, the dopamine transporter, which mediates dopamine reuptake from the cleft after release, was found to have a lower binding ability to its 42 agonist in PD cases (Pimoule et al., 1983), which has been confirmed in a PET imaging study (Frost et al., 1993). Different studies looking at the postsynaptic alterations within dopaminergic system in PD have reported many discrepancies, stirring up more controversy in the field. The binding ability of D2 dopamine receptors was found to be elevated in the putamen where dopamine concentration was decreased by more than 85%, but not in the caudate nucleus where dopamine deficiency was less severe (Bokobza, Ruberg, Scatton, Javoy-Agid, & Agid, 1984). Findings on D1 dopamine receptor activation in PD are inconsistent, with one study reporting increased dopamine-stimulated adenylate cyclase activity (Nagatsu, Kanamori, Kato, Iizuka, & Narabayashi, 1978), and another study showing reduced adenylate cyclase activity (Riederer, Rausch, Birkmayer, Jellinger, & Danielczyk, 1978). Additionally, a study looking at postmortem PD brains revealed a decrease in D2 receptor density with no changes in D1 receptor density (Pierot et al., 1988); however, a separate study did not detect any change in D2 receptor density (Guttman et al., 1986). Whether these divergent findings can be reconciled remains to be seen. Previous studies suggest that neurodegeneration in the dopaminergic system exhibit regional selectivity in PD. By computer visualization, SN (A9) suffers the most significant cell loss in the three measured dopaminergic groups (A8-A10) and the nearby ventral tegmental area (VTA, A10) is spared (German, Manaye, Smith, Woodward, & Saper, 1989). Similarly, in brain samples from 8 PD patients, there was a complete depletion of dopamine in the putamen of dorsal striatum but not in the caudate nucleus of dorsal striatum. More importantly, this unique pattern was not detected in any other secondary parkinsonism conditions (Kish, Shannak, & 43 Hornykiewicz, 1988). Fearnley’s group discovered that the SN is divided into two parts, the pars reticulate (SNpr) and the pars compacta (SNpc) with the latter being further suborganized into ventral and dorsal. In 20 examined PD cases, an average of 91% cells are lost in the lateral ventral tier followed by 71% in the medial ventral tier and 56% in the dorsal tier. Above all, this pattern differentiates PD from normal aging (Fearnley & Lees, 1991). 1.5.5 Interaction between αSyn and dopaminergic neurons αSyn is an abundant protein in the brain, and is the predominant component of LBs and LNs (Jakes, Spillantini, & Goedert, 1994; Spillantini et al., 1997). Since the neurodegeneration in PD preferentially occurs to dopaminergic neurons (Fearnley & Lees, 1991), it is reasonable to speculate whether there are links between αSyn and particular properties of dopaminergic neurons. A large body of evidence suggests that αSyn is able to modulate dopaminergic homeostasis and in turn, dopamine can affect the kinetics of αSyn oligomerization (Venda, Cragg, Buchman, & Wade-Martins, 2010). αSyn has been reported to inhibit the promoter activity and expression of TH, the rate-limiting enzyme in dopamine synthesis (Gao et al., 2007; Yu et al., 2004). αSyn is also able to bind to TH and decrease its activity (Kirik et al., 2002; Peng et al., 2005; Perez et al., 2002). Furthermore, αSyn has been implicated in inhibiting the activity of AADC, the enzyme responsible for catalyzing levodopa to dopamine (Tehranian, Montoya, Van Laar, Hastings, & Perez, 2006). αSyn is also involved in dopaminergic transmission as αSyn-deficient mice have increased levels of dopamine release (Yavich, Tanila, Vepsalainen, & Jakala, 2004), while overexpression of wildtype or mutant αSyn leads to reduced dopamine release, possibly through regulation of vesicle pools in dopaminergic neurons (Gorbatyuk et al., 2010; Nemani et al., 2010; Yavich et al., 2005). Taken together, αSyn not only modulates 44 dopamine synthesis and content, but it also regulates different aspects of dopaminergic transmission. The effects of dopamine on αSyn are unequivocally related to its aggregation process. An initial study indicated that αSyn forms an adduct with dopamine, leading to the inhibition of fibril formation and promotion of protofibril generation (Conway, Rochet, Bieganski, & Lansbury, 2001). This was confirmed by a following study, which demonstrated that incubation of recombinant wildtype human αSyn with dopamine promotes the formation of SDS-stable αSyn oligomers in a dosage- dependent manner in vitro (Cappai et al., 2005). Dopamine can also facilitate the oligomerization of both wildtype and mutant αSyn in a temperature-dependent way (Moussa, Mahmoodian, Tomita, & Sidhu, 2008). In addition to dopamine itself, its toxic metabolite, 3,4-dihydroxyphenylacetaldehyde (DOPAL), forms adducts with αSyn lysine residues and contribute to the formation of αSyn oligomers, which are further cross-linked by DOPAL (Follmer et al., 2015). Considering the toxicity of αSyn oligomers, enhanced oligomerization by αSyn-DA adducts would exert even more deleterious effects. Finally, it seems that modification of αSyn by DA directly blocks αSyn degradation by chaperone-mediated autophagy (CMA) in ventral medial neuron cultures and SH-SY5Y cells (Martinez-Vicente et al., 2008), suggesting the existence of other mechanisms besides oligomerization. 1.6 LRRK2 and PD With the discovery of SNCA in PD genetics, an intense investigation was undertaken to reveal other genetic contributors to the etiology of PD. LRRK2 was found to be linked to both autosomal-dominant familial PD and sporadic PD, accounting for ~10% of familial PD cases and 45 0.6% of sporadic PD cases (Berg et al., 2005; Mata et al., 2005). According to Parkinson Disease Mutation Database, over 100 mutations have been reported in the LRRK2 gene so far. However, only six of these mutations definitively produce pathogenic effects, including R1441G/C/H, Y1699C, G2019S, and I2020T (Healy et al., 2008). Regarding the clinical presentations, LRRK2 mutation carriers are indistinguishable from idiopathic PD patients, suggesting that understanding the pathophysiological functions of LRRK2 will advance our knowledge of PD etiology (Aasly et al., 2005; Healy et al., 2008). 1.6.1 The expression and cellular functions of LRRK2 In 2002, Funayama and colleagues identified a novel locus, chromosome 12q12, in hereditary parkinsonism in a Japanese family, by genome-wide linkage analysis and this locus was named PARK8 (Funayama et al., 2002). Afterwards, two groups successfully cloned a novel gene, LRRK2, which contains several mutations segregating with PARK8-associated familial PD cases (Paisan-Ruiz et al., 2004; Zimprich et al., 2004). LRRK2 mRNA is widely expressed throughout multiple organs and tissues, such as heart, kidney, lung, liver, pancreas, and brain (Paisan-Ruiz et al., 2004; Zimprich et al., 2004). Interestingly, LRRK2 mRNA is strikingly enriched in dopamine-innervated brain areas (namely striatum and certain areas of the cortex), but not in dopaminergic neurons (Galter et al., 2006). Regarding LRRK2 protein, it is widely expressed in various brain regions, whereas it is weakly expressed in the SNpc dopaminergic neurons (Mandemakers, Snellinx, O'Neill, & de Strooper, 2012). Ultrastructurally, LRRK2 is a predominantly a cytoplasmic protein and is particularly abundant in membrane-associated fractions in cell fractionation assays, indicating its interaction with synaptic vesicles, or vesicular membranes, or both (Biskup et al., 2006). 46 Over a decade has passed since the cloning of the LRRK2 gene. Although the cellular functions of LRRK2 and how they are linked to the pathogenesis of PD is still unknown, accumulating data have given us some clues. Knockdown of LRRK2 in primary cortical cultures promotes neurite growth and branching (MacLeod et al., 2006). In Caenorhabditis elegans, LRK1, a homolog of human LRRK2, is essential for regulating the axonal-dendritic polarity of synaptic vesicles, as deletion of LRK1 leads to the distribution of a synaptic vesicle protein marker to both axons and dendrites (Sakaguchi-Nakashima, Meir, Jin, Matsumoto, & Hisamoto, 2007). Another study further demonstrated that deficiency of LRRK2 decreases the phosphorylation of ezrin/radixin/moesin (ERM) proteins and F-actin in filopodia, thereby contributing to enhanced neurite growth (Parisiadou et al., 2009). Besides neurite growth, LRRK2 is also involved in protein translation. In Drosophila, LRRK2 induces robust phosphorylation of eukaryotic initiation factor 4E (eIF4E)-binding protein (4E-BP) at T37/46, which negatively regulates protein translation. After being phosphorylated, it confers less inhibition of eIF4E and therefore eIF4E-mediated protein translation is stimulated, leading to attenuated resistance to oxidative stress and greater age-dependent dopaminergic neuron loss (Imai et al., 2008). However, 4E-BP may not be an authentic substrate for LRRK2 under more physiological conditions, as overexpression of LRRK2 in HEK293 cells fails to increase 4E-BP phosphorylation (Kumar et al., 2010). Moreover, LRRK2 deletion mice display a striking accumulation of αSyn and ubiquitinated proteins in the kidney by 20-months old, suggesting its role in protein degradation (Tong et al., 2010). In summary, the recent accumulation of evidence suggests that LRRK2 is involved in vesicular trafficking, neurite growth, protein translation, and protein degradation. 47 1.6.2 LRRK2 protein domains The LRRK2 gene has 51 exons, encoding a large multi-domain protein with 2527 amino acids and weighing 286 kDa (Figure 1.3). This complex protein belongs to the Roco protein family and comprises of a central catalytic machinery- Ras of complex (ROC) GTPase domain, C-terminal of Ras (COR), and kinase domain- that is flanked by several protein-protein interaction domains, including ankyrin repeat region, leucine-rich repeat, and WD40 (Mata, Wedemeyer, Farrer, Taylor, & Gallo, 2006). Since all-known pathogenic mutations are concentrated within the central enzymatic domain, this region has garnered extensive attention. Figure 1.3 Protein domains and pathogenic mutations of LRRK2 The LRRK2 gene encodes a large multi-domain protein with 2527 amino acids. This complex protein belongs to the Roco protein family. It comprises a central catalytic ROC GTPase domain, COR, and kinase domains and is flanked by several protein-protein interaction domains, including ankyrin repeat region (ANK), leucine-rich repeat (LRR), and WD40. The well-known pathogenic mutations are presented in the box concentrated in the central domain. LRRK2 has a serine/ threonine kinase activity and preferentially phosphorylates threonine residues (Nichols et al., 2009; Pungaliya et al., 2010; West et al., 2005; West et al., 2007). Several groups have reported that purified full-length LRRK2 protein from mammalian cells has a kinase activity in autophosphorylation assays and phosphorylates a generic substrate- myelin basic protein (MBP), while an artificial kinase-dead mutant has only 10%-20% activity of the 48 wildtype protein (Gloeckner et al., 2006; Greggio et al., 2006; Smith et al., 2006; West et al., 2005; West et al., 2007). Furthermore, purified LRRK2 protein from mouse brain has a robust kinase activity in autophosphorylation assays (Li et al., 2007). Unfortunately, the use of autophosphorylation as a suitable readout of kinase activity is still under debate, and it is therefore critical to find its physiological substrates to address this issue (Greggio & Cookson, 2009). So far, two potential substrates have been proposed, moesin (Jaleel et al., 2007) and 4E-BP (Imai et al., 2008). Collectively, converging lines of evidence suggest that LRRK2 is an active kinase, but how these results relate to the physiological conditions is still questionable as most of these studies used LRRK2 protein purified under in vitro conditions. Besides the kinase domain, the Roc GTPase domain has also become the topic of intensive investigation. Purified LRRK2 from mammalian cells binds to guanosine-5’-triphosphate (GTP), but LRRK2 lacks GTPase activity, possibly due to the absence of co-factors (West et al., 2007). Likewise, Ito et al. reported that LRRK2 only bound to GTP without GTPase activity in both HEK293 and N2A cells (Ito et al., 2007). However, other groups have revealed that LRRK2 purified from bacterial lysates and mouse brain harbored GTPase activity (Guo et al., 2007; Lewis et al., 2007; Li et al., 2007). The kinase domain of LRRK2 resembles the tyrosine kinase-like (TKL) family, the members of which are commonly activated by small GTPase (Bosgraaf & Van Haastert, 2003). Therefore, it has been hypothesized that LRRK2’s kinase activity is regulated by its Roc GTPase domain; however, the data appears to be conflicting on this point. Some studies conclude that either GTP binding or GTPase activity can regulate kinase activity (Guo et al., 2007; Ito et al., 2007; Li et al., 2007; Smith et al., 2006; West et al., 2007), while other studies report that binding of GTP, GDP, or GMP does not affect kinase activity as long as 49 it has an intact Roc GTPase domain (Liu, Dobson, Glicksman, Yue, & Stein, 2010; Taymans et al., 2011). More recent findings of ARHGEF7 and ArfGAP1 as the putative guanine nucleotide exchange factor (GEF) and GTPase activating protein (GAP) for LRRK2, respectively, provide further support for its role as a GTPase (Haebig et al., 2010; Stafa et al., 2012; Xiong, Yuan, Chen, Dawson, & Dawson, 2012). Last but not least, there are multiple protein-protein interaction domains at the N- and C-terminals of LRRK2 protein. The WD40 deletion mutant decreases LRRK2 kinase activity and fails to induce caspase-3 activation when transfected into SH-SY5Y cells even in the presence of R1441C mutation, implying a potential role of WD40 domain in regulating kinase activity and protecting cells from apoptosis (Iaccarino et al., 2007). Other studies also point out that the minimal catalytic machinery of LRRK2 consists of the central Roc-Cor-kinase domain, WD40, and a C-terminal tail. Deletion of the WD40 domain abolishes dimer formation, autophosphorylation, and neurotoxicity induced by PD-linked LRRK2 mutations (Greggio et al., 2008; Jaleel et al., 2007; Jorgensen et al., 2009). Collectively, the central Roc-Cor-kinase domain is a pivotal regulatory machinery for LRRK2 and the kinase activity acts as a readout of its overall cellular function, which can be further modulated by GTP binding, GTP hydrolysis, or other domains on its protein sequence through intramolecular or intermolecular mechanisms. 1.6.3 LRRK2 mutations and their effects As noted above, there are at least six pathogenic mutations in the LRRK2 gene contributing to PD, with the G2019S mutation being the most common one as it is responsible for 2-6% of hereditary PD and 1-2% of sporadic PD (Brice, 2005; Healy et al., 2008; Lesage et al., 2006). In 50 North American Arabs, the frequency of this mutation reaches as high as 37% in familial cases and 41% in sporadic case (Lesage et al., 2006). The mutation was first discovered in 2005 by three groups (Di Fonzo et al., 2005; Gilks et al., 2005; Nichols et al., 2005) and the steady accrual of data by the following studies help elucidate its underlying pathogenic mechanisms. It has been consistently found that the G2019S mutation substantially increases kinase activity measured by a variety of in vitro kinase assays (Greggio & Cookson, 2009; Greggio et al., 2006; Imai et al., 2008; Jaleel et al., 2007; Smith et al., 2006; West et al., 2005). However, it is uncertain whether another mutation in the kinase domain, I2020T, can elevate kinase activity, as some studies reported significant increase in kinase activity (Gloeckner et al., 2006; Imai et al., 2008; West et al., 2007), while other studies showed no change (Anand et al., 2009; Luzon-Toro, Rubio de la Torre, Delgado, Perez-Tur, & Hilfiker, 2007) or even slightly decreased activity (Jaleel et al., 2007). Besides mutations in the kinase domain, R1441C/G mutants in the GTPase domain have been reported to be associated with reduced GTP hydrolysis (Guo et al., 2007; Lewis et al., 2007; Li et al., 2007) or increased GT hydrolysis (West et al., 2007). Several studies have also reported that these two mutations are linked to enhanced kinase activities (Guo et al., 2007; Iaccarino et al., 2007; Smith et al., 2006; West et al., 2005; West et al., 2007). Therefore, it is possible that any mutation with a decreased GTPase activity will extend the ‘GTP-bound’ state for LRRK2, and this helps to stimulate its overall cellular function, resulting in an increased kinase activity as a readout. Although this proposal has not been fully validated, a recent study sheds light on its reasonability. The R1441H mutation has a decreased ability to hydrolyze GTP without affecting its protein structure, thermal stability, and GDP-binding ability, but it elevates GTP-binding 51 affinity and keeps it in an ‘active’ state (Liao et al., 2014). Lastly, data on the Y1699C mutation is more ambiguous, with some studies showing significant increase in both GTPase and kinase activity (West et al., 2007), and other studies showing almost the same level of activity as LRRK2 wildtype in the kinase assays (Anand et al., 2009; Greggio et al., 2006; Jaleel et al., 2007; Luzon-Toro et al., 2007). Overall, whether kinase activity or GTPase activity or both is the readout for the pathogenic effects of these mutations still need to be determined. 1.6.4 The effects of LRRK2 protein level on cellular toxicity When examining the enzymatic activities of LRRK2, most studies demonstrate that LRRK2 mutations either increase LRRK2 kinase activity or decrease GTPase activity without affecting the stability of LRRK2 itself (Gloeckner et al., 2006; West et al., 2005). However, one particular study made an artificial GDP/GTP binding-deficient mutant, K1347A, and observed a marked decrease in LRRK2 protein levels. Thus, it is possible that changes in LRRK2 protein levels could be involved in its pathophysiological functions (Lewis et al., 2007). Similar effects have been observed in other studies where mice expressing a kinase-dead mutant show a dramatic reduction in the steady-state level of LRRK2 protein (Lin et al., 2009). These results were also confirmed by treating wildtype mice with a LRRK2-selective kinase inhibitor (Herzig et al., 2011). Overexpression of wildtype LRRK2 protein induces moderate toxicity in neuronal cells (Greggio et al., 2006; Smith et al., 2006). Primary neurons transfected with wildtype LRRK2 showed significant cell death, with the effects being potentiated by hydrogen peroxide treatment (Smith et al., 2006; West et al., 2007; Yao et al., 2013). The results were mixed from animal models to 52 study the toxicity of wildtype LRRK. Several studies have demonstrated that wildtype LRRK2 transgenic Drosophila fail to induce cell loss, although minor abnormalities were observed, such as increased lipid peroxidation in hydrogen peroxide treatment measured by 4-hydroxynonenal (4-HNE) immunostaining, defects in endocytic vesicular trafficking, and autophagic dysfunctions (Imai et al., 2008; Venderova et al., 2009; Xiong et al., 2010). However, Liu et al. demonstrated that, in transgenic Drosophila, expression of wildtype LRRK2 in photoreceptors causes retinal degeneration while expression of wildtype LRRK2 in the neurons leads to late-onset dopaminergic neuron loss, motor dysfunction, and shorter lifespan (Liu et al., 2008). These results were replicated by another study performed in wildtype LRRK2 transgenic Caenorhabditis elegans (Yao et al., 2010). Additionally, it has been reported that overexpression of LRRK2 in transgenic mice is not sufficient to cause neurodegeneration, but its presence significantly potentiates the neuropathological features seen in αSyn A53T transgenic mice (Lin et al., 2009). Collectively, overexpression of wildtype LRRK2 promotes toxicity as supported by various models and can even accelerate phenotypes in αSyn A53T transgenic mice, suggesting that modulation of LRRK2 protein levels could be another way to regulate its overall functions. Heat shock protein 90 (Hsp90) was found to be co-immunoprecipitated with hemagglutinin (HA)-tagged LRRK2 protein from transgenic mouse brain, indicating their interaction in vivo. Application of a Hsp90 inhibitor decreases the level of overexpressed wildtype and G2019S LRRK2 mutant in HEK293 cells as well as endogenous LRRK2 in primary cortical neurons. More intriguingly, Hsp90 rescues the deficits in neurite growth caused by the G2019S mutation by decreasing its expression (Wang et al., 2008). A subsequent study revealed that carboxyl terminus of HSP70-interacting protein (CHIP), a ubiquitin ligase that interacts with Hsp90, 53 ubiquitinates and facilitates the degradation of LRRK2 through the ubiquitin proteasome pathway. Overexpression of CHIP protects SH-SY5Y cells from mutant LRRK2-induced cell death (Ko et al., 2009). A more recent study has discovered that overexpression of wildtype or mutant LRRK2 (G2019S and Y1699C) in cortical neurons leads to dosage-dependent cellular toxicity (i.e. higher levels of expression results in more cell death). The kinase activities of neurons overexpressing either a kinase-active or kinase-dead form of LRRK2 were normalized to their protein expression levels. After normalization, both kinase-active and kinase-dead LRRK2 constructs appear to have a similar influence on toxicity. This strongly supports the notion that the toxicity of mutant LRRK2 in neurons is dependent on its levels rather than its kinase activity (Skibinski, Nakamura, Cookson, & Finkbeiner, 2014). 1.7 Overall goal of this research This thesis addresses two projects, the transcriptional regulation of the human LRRK2 gene, and the molecular mechanisms underlying selective neurodegeneration in AD and PD. Chapter 2 discusses the first project; Chapter 3 and Chapter 4 focus on the second project. 1.7.1 Sp1 enhances LRRK2 promoter activity and gene expression LRRK2 mutations contribute to the most common genetic factors in both familial and sporadic PD (Trinh & Farrer, 2013). In pursuit of explaining the pathological role of LRRK2 in PD, the central enzymatic domains have garnered attention in the field. The impressive accumulation of evidence suggests that pathogenic mutations in the LRRK2 gene either affect its kinase activity or GTPase activity (Gómez-Suaga et al., 2014). In addition, recent discoveries also postulate a role of its expression level in determining its toxicity in neurons, and indicate modulation of its 54 expression level is able to partially rescue cytotoxicity induced by LRRK2 mutations (Lewis et al., 2007, Skibinski, Nakamura, Cookson, & Finkbeiner, 2014). Therefore, the elucidation of mechanisms underlying the regulation of its expression would reveal possible modulators, which could be beneficial to alleviate the toxic effects of LRRK2. As one of the mechanisms to control gene expression, a gene promoter containing a complex array of cis-acting elements is required to efficiently initiate gene expression and control gene expression pattern in a spatial- and temporal- specific manner (Orphanides & Reinberg, 2002). So far, the features of the LRRK2 gene promoter remain unknown, and the regulation of its transcription is still an undetermined question. Therefore, we hypothesized that the transcription of human LRRK2 gene is regulated by certain transcription factors, and its expression can be regulated by modulating such transcription factor signaling pathway. To address our hypothesis, we firstly planned to clone the human LRRK2 gene promoter and characterize its transcriptional activity by identifying various regulatory regions. Secondly, the putative binding sites of the transcription factors in the promoter would be searched by an online software. Based on the pathophysiological functions of the transcription factors, we would further test the role of certain transcription factors in the regulation of the LRRK2 gene, focusing on the transcription factors associated with PD or other neurodegenerative disorders. Thirdly, the binding between transcription factors and cis-acting elements located in the LRRK2 gene promoter would be validated. Fourthly, the effects of the transcription factors on the LRRK2 gene expression would be investigated. Finally, if any pharmaceutical drug or reagent is available to modulate the transcription factor’s signaling pathway, its impact on LRRK2 gene expression would be further determined. 55 1.7.2 Cell type- specific effects of the APPSWE and SNCAA53T mutations contributing to the selective neurodegeneration in AD and PD The pathological hallmark of AD is neuritic plaques, which primarily consist of extracellular deposits of Aβ generated from APP (Glenner & Wong, 1984b; Goldgaber et al., 1987; Kang et al., 1987; Robakis et al., 1987; Tanzi et al., 1987). PD is pathologically characterized by intracellular LBs and LNs. They are primarily composed of aggregated αSyn protein, which is encoded by the SNCA gene (Spillantini et al., 1997; Uéda et al., 1993). In addition, both AD and PD patients exhibit extensive neuron loss within certain brain regions (Fearnley & Lees, 1991; Whitehouse et al., 1982). An invariant feature of AD is the degeneration of cholinergic neurons in the basal forebrain and the associated cholinergic deficiency in the neocortex and hippocampus (Bartus, Dean, Pontecorvo, & Flicker, 1985). In PD, dopaminergic neurodegeneration in the SN is prominent, leading to the loss of dopaminergic transmission in the striatum (Ehringer & Hornykiewicz, 1960; Hornykiewicz, 1963). Various mutations have been found in the APP and SNCA genes in familial AD and PD, respectively (Bekris, Yu, Bird, & Tsuang, 2010; Houlden & Singleton, 2012). APPSWE mutation results in increased Aβ production (Cai, Golde, & Younkin, 1993; Mullan et al., 1992), whereas SNCAA53T mutation makes αSyn more prone to be aggregated (Conway, Harper, & Lansbury, 1998). Previous studies support the interplay between Aβ and cholinergic neurons as well as the interaction between αSyn and dopaminergic neurons (Bellucci et al., 2012; Pakaski & Kalman, 2008). APP and αSyn proteins are ubiquitously expressed in the brain, but how these two proteins carrying the Swedish and A53T mutants, respectively, lead to selective neurodegeneration remains to be addressed. Here, we hypothesized that the cell type- specific effects of APPSWE and SNCAA53T mutations contributes to the selective neurodegeneration in AD and PD. SN56 and MN9D cells 56 represent cholinergic and dopaminergic neuronal phenotypes (Choi et al., 1991; Hammond, Wainer, Tonsgard, & Heller, 1986), respectively, which have been extensively used as in vitro cell models in AD (Heinitz, Beck, Schliebs, & Perez-Polo, 2006; Hicks et al., 2013; Joerchel, Raap, Bigl, Eschrich, & Schliebs, 2008; Kwakowsky et al., 2016) and PD (Galvin, 2006; Holtz & O'Malley, 2003; Su et al., 2008; Jianyong Wang et al., 2007). Specifically, we proposed to firstly determine the differential effects of APPSWE mutation on Aβ production and SNCAA53T mutation on αSyn aggregation in cholinergic SN56 and dopaminergic MN9D cells. As a major role of APP proteolysis and Aβ generation in AD pathogenesis, CTF patterns indicative of APP processing would be investigated by immunoblotting and Aβ production would be determined by ELISA in cholinergic SN56 cells and dopaminergic MN9D cells stably overexpressing wildtype or Swedish APP695 (SN56-APPWT, SN56-APPSWE, MN9D-APPWT, and MN9D-APPSWE). To determine the difference of SNCAA53T mutation in promoting aggregate formation between two neuronal cell lines, immunostaining would be applied to SN56 and MN9D cells stably overexpressing wildtype or mutated SNCA (SN56-SNCAWT, SN56-SNCAA53T, MN9D-SNCAWT, and MN9D-SNCAA53T. To study the effect of the same mutation on cell vulnerability in these two cell lines in response to oxidative stress, all stable cell lines would be treated with hydrogen peroxide. Cell death would be determined by LDH assays and apoptosis would be examined by the caspase activation. Subsequently, Aβ oligomer would be applied to APP-related stable cells, whereas αSyn oligomers would be applied to SNCA-related stable cells. Cell death and apoptosis would be evaluated by the aforementioned methods. 57 1.7.3 The effects of syndecan 3 (SDC3) and fibroblast growth factor like 1 (FGFRL1) on the neurodegeneration in APPSWE-associated AD and SNCAA53T-associated PD In the previous chapter, we have already discussed cell type- specific effects of the APPSWE and SNCAA53T mutations in cholinergic and dopaminergic cells. Following these, we aim to investigate the differential effects of APPSWE and SNCAA53T mutations on cholinergic and dopaminergic neurons in a systemic way. Since the discovery of DNA microarray in 1995 (Schena, Shalon, Davis, & Brown, 1995), its application in studying AD and PD offers another option to determine the molecular mechanisms from a novel perspective by studying large amounts of genes simultaneously, and it also helpful to gain a broad picture of these diseases in a systematic view (Lovén et al., 2012). Although microarray studies using AD or PD postmortem brain samples or animal models are fruitful (Cooper-Knock et al., 2012; Greene, 2012), there are limited results to reveal the gene expression profiling within specific neuronal populations in AD and PD by microarrays. When processing homogenous brain samples, neurons and glia are mixed together in the microarray samples without adequate purity. Thus, the information about specific effects within certain neuronal groups can hardly be obtained. With the technical advancement, LCM is employed to prepare samples containing certain neuronal groups with higher purity, but the RNA quality is compromised due to the sophisticated procedures (Lovén et al., 2012). Under such circumstance, we will use the established APP-related and SNCA-related stable cell lines and animal models to study APPSWE-associated AD and SNCAA53T-linked PD in parallel. We hypothesized that APPSWE and SNCAA53T mutations differentially alter the gene expression in cholinergic and dopaminergic cells and certain target genes can modify the cell susceptibility of cholinergic and dopaminergic neurons affected by APPSWE and SNCAA53T mutations in response to cytotoxic insults. Firstly, we aimed to determine the role 58 of APPSWE and SNCAA53T mutations in changing the gene expression in cholinergic and dopaminergic cells by performing whole-genome expression profiling assay in the established stable cells and their parental SN56 and MN9D cells. Genes differentially expressed in cholinergic and dopaminergic cells carrying the same mutation would be identified. Gene ontology enrichment analysis and pathway analysis would be applied to determine the target genes involved in the pathways related to cell survival or cell death. Secondly, qRT-PCR and immunoblotting analysis would be conducted to validate the expression of target genes in our cellular models and animal models. Finally, overexpression or knockdown of the target genes would be performed to examine their effects on the cell viability of dopaminergic and cholinergic cells carrying the APPSWE or SNCAA53T mutations in response to the cytotoxic insults. 59 Chapter 2: Sp1 enhances LRRK2 promoter activity and gene expression 2.1 Introduction LRRK2 gene was first identified as a PD-associated locus, termed as PARK8 (Funayama et al., 2002). Two years later, its mutations were found to be segregated with familial PD patients (Funayama et al., 2002; Paisan-Ruiz et al., 2004; Zimprich et al., 2004). As the most frequent genetic cause of PD, LRRK2 mutations accounts for 13% of autosomal dominant PD and 0.6% of sporadic PD cases (Berg et al., 2005). Besides, the clinical presentations of LRRK2 mutation carriers are indistinguishable from the idiopathic PD cases, and most of them are positive for LBs (Aasly et al., 2005; Healy et al., 2008). With the help of accumulating discoveries, it becomes more evident that LRRK2 plays a central role in the PD pathogenesis by connecting several other key players, such as SNCA, PINK1, DJ-1, and Parkin (Houlden & Singleton, 2012). The LRRK2 gene consists of 51 exons and encodes a large complex protein with the weight of 286 kDa. This multi-domain protein belongs to a Roco protein family, featuring by a catalytic ROC GTPase domain, a COR domain, and a kinase domain, which phosphorylates serine/ threonine residues (Mata et al., 2006). LRRK2 mRNA is highly expressed in several organs and tissues, such as heart, kidney, lung, liver, and pancreas (Paisan-Ruiz et al., 2004; Zimprich et al., 2004). In the brain, the expression of LRRK2 mRNA is enriched in dopamine-receiving areas, including the striatum and certain areas of the cortex, whereas almost no expression in the dopaminergic neurons (Galter et al., 2006). Furthermore, LRRK2 protein is weakly expressed in the dopaminergic neuron in the SNpc (Biskup et al., 2006; Mandemakers, Snellinx, O'Neill, & de Strooper, 2012). Although the physiological functions of the LRRK2 gene is not well-defined, 60 endogenous LRRK2 knockdown or knockout studies help to deepen our understanding. It has been implicated in the regulation of neurite growth (MacLeod et al., 2006; Parisiadou et al., 2009), modulation of vesicle trafficking (Sakaguchi-Nakashima, Meir, Jin, Matsumoto, & Hisamoto, 2007), protein degradation (Tong et al., 2010), and global gene expression (Habig, Walter, Poths, Riess, & Bonin, 2008; Nikonova et al., 2012). Over 75 substitutions have been discovered in LRRK2 protein, but only seven of them have confirmed pathogenic effects (G2019S, I2020T, N1437H, R1441G/C/H and Y1699C). They are concentrated in the central catalytic domains, indicating GTPase and kinase domains are of essential importance in studying PD pathogenesis (Justus & Matthew, 2010; Martin, Kim, Dawson, & Dawson, 2014). Converging lines of evidence suggest that overexpression of WT LRRK2 reduces cell viability and makes cells more susceptible to oxidative stress (Greggio et al., 2006; West et al., 2007). In a Drosophila model, overexpression of WT LRRK2 in neurons results in age-dependent dopaminergic neuron loss, motoric dysfunction, and shorter lifespan (Liu et al., 2008). Both Hsp90 and CHIP affect the stability of LRRK2. Hsp90 forms complex with LRRK2 and inhibitors of Hsp90 alleviates the deficits of axonal growth induced by overexpressing LRRK2 G2019S mutation (Wang et al., 2008). CHIP binds to and ubiquitinates LRRK2 in order to facilitate its degradation through proteasome. Accordingly, CHIP overexpression rescues LRRK2 mutant-associated toxicity, while knockdown of CHIP promotes toxicity (Ko et al., 2009). A recent study has discovered that both wildtype and mutant LRRK2 (G2019S and Y1699C) manifest dosage-dependent cellular toxicity in cortical neurons. Interestingly, kinase-active and kinase-dead LRRK2 constructs have similar influence on toxicity after normalizing to their expression level, strongly supporting that toxicity of mutant LRRK2 is dependent on its 61 levels other than kinase activity (Skibinski, Nakamura, Cookson, & Finkbeiner, 2014). In addition, transcriptional dysregulation has been observed in neurodegenerative disorders (Anthone et al., 2002; Chen et al., 2012; Christensen et al., 2004; Katharine & David, 2003; Ly et al., 2013; Santpere, Nieto, Puig, & Ferrer, 2006; Wang et al., 2014; Wang et al., 2012a; Xu, Guo, et al., 2012). For example, LRRK2 mRNA is reduced in PD patients in comparison to the control subjects (Simunovic et al., 2009). However, it is still not clear how LRRK2 is regulated at the transcriptional and translational level. In this chapter, we firstly aimed to identify the transcription start site (TSS) of human LRRK2 gene promoter and characterize promoter activity. Secondly, we planned to search for putative transcriptional factor binding sites in the promoter region and test whether they are functional. Finally, the level of identified transcriptional factor would be manipulated by either genetic methods or pharmacological treatment, and its effects on LRRK2 promoter activity would be further examined. 2.2 Methods 2.2.1 Primers and plasmids LRRK2 human gene promoter, from -1738 base pair (bp) to +133bp relative to the TSS, was amplified from genomic DNA(gDNA) extracted from HEK293 cells by Polymerase chain reaction (PCR) and cloned into pGL3-Basic by restriction enzymes in the upstream of firefly luciferase reporter gene to generate the promoter plasmids. A serial deletion fragment was amplified from the longest fragment by PCR and inserted into pGL3-Basic vector (Promega) by the same strategy. The primers with the restriction enzyme sites were synthesized as follows : 62 forward, 1) -1738NheI: ctagctagcgaaacaacttagaaaataatacactg, 2) -794NheI: ctagctagccccaagtatcaggatcctgcc, 3) -495 BglII: cttagatctggagataggcggc, 4) -413Nhe: ctagctagcggtcgcggagggtggccggc, 5) -118XhoI: ccgctcgagtcgtttttgggcctgagt, and 6) -34XhoI: ccgctcgagtccttcctcataaacaggcg; reverse, 1) -794HindIII: cccaagcttggcaggatcctgatacttggg, 2) -413HindIII: cccaagcttgccggccaccctccgcgacc, 3) -34HindIII: cccaagcttaggcagctccccgccccgcgt, 4) -4HindIII: cccaagcttgcgcccacgcccgcctgttta, and 5) +133HindIII: cccaagctttggcacctgcttccaaccc gccg. Sp1 expression plasmid (pCGN-Sp1) was inserted Sp1 cDNA with the hemagglutinin (HA) tag under the control of the cytomegalovirus promoter (Parks & Shenk, 1996). Each plasmid was first confirmed by both restriction enzyme digestion and the digested products were examined by 1%-1.5% agarose gel. Gel images were captured using the GelDoc-It Gel Imaging System. Sequencing PCR was performed to further check the nucleotide sequence of various LRRK2 promoter fragments inserted on the pGL3-Basic vector. 2.2.2 Cell culture and transfection HEK293 cells and dopaminergic MN9D cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 1 mM of sodium pyruvate (SP), 2 mM of L-glutamine, 50 units of penicillin and 50 μg of streptomycin (Invitrogen). The culture dishes (Corning) for maintaining MN9D cells were coated with 10 μg/mL poly-D-lysine (PDL, Sigma). Both HEK293 and MN9D cells were maintained at 37°C in an incubator containing 5% CO2 and cell medium was changed when necessary. Before transfection, cells were seeded on the plate at 50%-70% confluency and all transfection experiments were performed by lipofectamine 2000 (Invitrogen) following manufacturer’s instruction. 63 2.2.3 Dual-luciferase reporter assay For luciferase assays (Promega), cells were cotransfected with 500 ng firefly luciferase plasmid (pGL3-Basic) containing various promoter fragments as described before and 1 ng renilla luciferase plasmid (pCMV-Luc), which was for normalizing transfection efficiency. 24 hours after transfection, cells were then harvested by passive lysis buffer. The activities of firefly and renilla luciferase were measured by illuminance meters and the firefly luciferase activities were normalized to renilla luciferase activities and the promoter activities of various fragments were represented as relative luciferase unit (RLU) after normalizing to pGL3-Basic. 2.2.4 5’-Rapid amplification of cDNA end (5’-RACE) assay Total RNA was extracted from HEK293 cells by TRI reagent following the product’s manual (Sigma). By using the Smarter RACE cDNA amplification kit (Clotech), a patent SMARTScribe reverse transcriptase was applied to synthesize first-strand full-length cDNA and the SMARTer oligonucleotide was annealed to the extended cDNA tail. This amplified first-strand with specialized oligonucleotide was then worked as a template to generate double-strand complete cDNA copy of the original RNA with the additional SMARTer sequence at the end. The outer and inner forward primers were provided in the kit and the outer and inner reverse primers were designed based on human LRRK2 gene sequence, which were 5’-atcccagccatcatccagacc and 5’- caggatttggaccagcgtttct, respectively. PCR product for nested PCR was sequenced to locate the TSS of the human LRRK2 gene, which was the first base pair after SMARTer oligonucleotide sequence. 64 2.2.5 Electrophoretic mobility shift assay (EMSA) The EMSA performed in this experiment was modified from a previously published research (Wang et al., 2012b). HEK293 cells were transfected with pCGN-Sp1 expression plasmid and lysed in a series of hypotonic buffers to extract Sp-1 enriched nuclear protein 24 hours post-transfection. Synthesized oligonucleotides were labelled with IR700 Dye (LI-COR Biosciences) and annealed to produce double-stranded probes. The labelled probes were incubated with or without nuclear extract at 22°C for 20 min in the EMSA binding buffer (4% glycerol, 1 mM MgCl2, 0.5 mM EDTA, 0.5 mM DTT, 50 mM NaCl, 10 mM Tris-HCl (pH 7.4), and 50 μg/mL poly(dI-dC)). For the competition assay, nuclear extract was first incubated with 100 fmol (2 times excess) or 10 pmol (20 times excess) of unlabeled competition oligonucleotides for 10 min followed by adding 50 fmol labelled probes. The EMSA samples were analyzed on 4% non-denaturing polyacrylamide gels and the gels were scanned using the Odyssey scanner (LI-COR Biosciences) at a wavelength of 700 nm. The sequences of the oligonucleotides were synthesized as follows: consensus Sp1-forward: 5’-attcgatcggggcggggcgagc; consensus Sp1-reverse: 5’-gctcgccccgccccgatcgaat; mutant Sp-1 forward: 5’- attcgatcggggcggggcgagc; mutant Sp-1 reverse: 5’- gctcgccccgaaccgatcgaat (Tamaki, Ohnishi, Hartl, LeRoy, & Trojanowska, 1995); LRRK2-Sp1-probe1-forward: 5’-gccatctgggcggtgtcctc; LRRK2-Sp1-probe1-reverse: 5’gaggacaccgcccagatggc; LRRK2-Sp1- probe2-forward: 5’-gcggcgtccgcccggggtcc; LRRK2-Sp1- probe2-reverse: 5’-ggaccccgggcggacgccgc; LRRK2-SP1- probe3-forward: 5’-caacgcggggcggggagctg; LRRK2-Sp1- probe3-reverse: 5’-cagctccccgccccgcgttg. 65 2.2.6 Sp1 knockdown HEK293 cells were plated at 50% confluence for siRNA transfection the day before. For luciferase assay, 50nM Sp1siRNA or negative control siRNA was cotransfected with LRRK2 promoter plasmids into HEK293 cells by Lipofectamine 2000 (Invitrogen) following manufacturer’s instruction. Luciferase activities were examined 48 hours post-transfection. For reverse transcription polymerase chain reaction (RT-PCR) and immunoblotting, either 50nM Silencer® Select negative control siRNA or Sp1 siRNA was transfected into HEK293 cells by Lipofectamine 2000. Cells were harvested 48 hours afterwards. The sense sequence of Sp1 siRNA is 5’-gcaacaugggaauuaugaatt and the antisense sequence is 5’-uucauaauucccauguugctg. 2.2.7 RT-PCR ThermoScriptTM RT-PCR system (Invitrogen) was applied to perform this experiment. Total RNA was first extracted from HEK293 or MN9D cells by TRI reagent (Sigma). Following this, ThermoScript™ reverse transcriptase was applied to amplify cDNA from 1.0-1.5 μg total RNA. Subsequently, the newly synthesized cDNA was used as the template to further amplified human LRRK2 gene by Taq DNA polymerase. The specific primers for human LRRK2 gene were as follows: forward, 5’- gagcacgcctccaagttat, and reverse, 5’- gtgattttacctgaagttag. This pair of primers was designed to yield a 302 bp fragment, corresponding to the human LRRK2 gene coding sequence in the HEK293 cells. Additionally, another pair of primers for amplifying a 115 bp fragment of mouse LRRK2 gene in MN9D cells was as follows: forward, 5’- aggagctgcccccttgaagaca, and reverse, 5’- tgtgccacaccctccccatgt. Human and mouse β-actin was used as internal controls, and two pairs of gene specific primers for HEK293 and MN9D cells were: forward, 5’- ggacttcgagcaagagatgg, reverse, 5’-gaagcatttgcggtggag, forward, 5’-66 gacaggatgcagaaggagat, and reverse, 5’-ttgctgatccacatctgctg, respectively. All PCR products were analyzed on 1.5% agarose gels. Gel images were captured using the GelDoc-It Gel Imaging System (UVP) and quantified with the Image J software. 2.2.8 Immunoblotting HEK293 and MN9D cells were lysed in triton lysis buffer containing 150 mM sodium chloride, 1.0% Triton X-100, 50 mM Tris-HCl (pH 8.0) and protease inhibitor cocktail (Roche), followed by brief sonication. Protein concentration was determined by Bradford assay (Bio-rad) and each sample was diluted with 4x SDS loading buffer (200 mM Tris-Cl pH 6.8, 400 mM DTT, 8% SDS, 0.4% bromophenol blue, 40% glycerol). For detecting LRRK2 and Sp1, samples were resolved on 6% and 8% Tris–glycine SDS–polyacrylamide gel electrophoresis (PAGE), respectively. Samples were then transferred to polyvinylidene fluoride (PVDF-FL) membranes and membranes were blocked in phosphate-buffered saline (PBS) containing 5% non-fat milk at room temperature for 1 hour. Primary antibodies diluted in the blocking medium were incubated overnight at 4℃. The primary antibodies were detailed as follows: Rabbit anti-LRRK2 monoclonal antibody MJFF C81-8 (Abcam, 1:1000), rabbit anti-Sp1 polyclonal antibody PEP2 (Santa Cruz Biotechnology, 1:1000), mouse anti-β-actin monoclonal antibody AC-15 (Sigma, 1:5000) and mouse anti-HA monoclonal antibody 12CA5 (Abcam; 1:200). IRDye 680RD-labelled goat anti-rabbit antibodies and IRDye 800CW-labelled goat anti-mouse antibodies were applied as secondary antibodies. The gels were scanned in the Odyssey system (LI-COR Biosciences). 67 2.2.9 Mithramycin A (MTM) treatment MTM (Sigma) was dissolved in 100% methanol to make 250 mM stock preparation. In order to test dosage-dependent effects, HEK293 cells were treated with MTM at 0, 25, 75 and 125 nM for 24 hours after transfecting with various LRRK2 promoter plasmids for 1 day. Similarly, for time course experiments, HEK293 cells were first transfected with LRRK2 promoter plasmids for 1 day and treated with MTM at 125 nM for another 24 or 48 hours. For RT-PCR and immunoblotting, HEK293 and MN9D cells were treated with 125 nM MTM for 24 hours and then lysed for mRNA and protein extraction. 2.2.10 Statistical analyses All results were presented as means ± SEM. For two-group comparison, the results were analyzed by 2-tailed Student’s t-test. For multiple-group comparison, the data were analyzed by one-way analysis of variance (ANOVA) or two-way ANOVA with Bonferroni’s multiple comparison test. Statistical significance was accepted when p<0.05 (*p<0.05, **p<0.01, ***p<0.001). 2.3 Results 2.3.1 Characterization of the human LRRK2 gene promoter The human LRRK2 gene is located on chromosome 12 (12q12), spanning the region of 144274 base pair (bp). It contains 51 exons and encodes a large complex protein weighing 286 kDa, but how it is transcriptionally regulated is poorly defined. In order to identify the human LRRK2 gene promoter, its transcriptional start site was first determined by 5’-RACE assays. The full-length human LRRK2 cDNA was amplified from HEK293 cells. Two pairs of inner and outer 68 primers were applied to perform nested PCR, and a ~300bp PCR product was yielded on the 1.5% agarose gel after inner PCR reactions. After sequencing this PCR product, the TSS was mapped to 135bp upstream of translational start site (ATG), which was started with guanine and designated as +1 shown in Figure 2.1. To perform functional analyses for human LRRK2 gene promoter, a 5’-flanking region of the LRRK2 gene was first cloned from human HEK293 cells, which covered 1873bp upstream of ATG, and the full sequence of this region was presented in Figure 2.1. 69 Figure 2.1 Identification of TSS and sequence features of the human LRRK2 gene promoter. (A) The ~300bp product from nested PCR of 5’-RACE was resolved on 1.5% agarose gel. (B) The ~300bp PCR product from 5’-RACE was sequenced and the arrow indicates the TSS of the human LRRK2 gene promoter. (C) The nucleotide sequence of the human LRRK2 gene promoter. The 5’-flanking region of the human LRRK2 gene was cloned from gDNA extracted from HEK293 cells, and TSS is begun with guanine and designated as +1. The putative binding sites for various transcription factors are underlined in bold face. 2.3.2 Functional analyses of the human LRRK2 gene promoter To examine the activity of the human LRRK2 gene promoter, ten deletion fragments covering varying lengths of the 5’-flanking region were cloned into a luciferase reporter vector, pGL3-Basic. This vector lacks the eukaryotic promoter and enhancer, and the expression of firefly 70 luciferase protein is driven by the correctly inserted functional promoter upstream of it. Therefore, the bioluminescent measurement of the firefly luciferase protein indicates the inserted promoters’ activities in a quantitative way. After cloning the desired promoter fragments into the pGL3-Basic (Figure 2.2A), the construction of all plasmids was verified by restriction enzyme digestion. The digested inserts were visualized on 1.5% agarose gel (Figure 2.2B) and confirmed by sequencing each nucleotide. The promoter activity of pLRRK2-A fragment from -1738 bp to +133 bp was 8.27 ± 0.35 RLU, significantly higher than pGL3-Basic (p< 0.0001, Figure 2.2C), demonstrating it worked as a functional promoter. In order to validate luciferase assay, four deletion fragments without TSS in their promoter region, including pLRRK2-B, pLRRK2-D, pLRRK2-H, and pLRRK2-I, did not have significant promoter activities when comparing with pGL3-Basic (p> 0.05). A significant increment of luciferase activity was detected in pLRRK2-C after deleting 944 bp from pLRRK2-A, from 8.27 ± 0.35 RLU for pLRRK2-A to 11.25 ± 0.38 RLU for pLRRK2-C (p< 0.0001). Compared with pLRRK2-C, there was a significant decrease in the promoter activity of pLRRK2-E (7.70 ± 0.29 RLU, p< 0.0001), which contained the fragment from -495 bp to +133 bp. Promoter activity of pLRRK2-F, covering the region of -413 bp to +133 bp, was 15.52 ± 0.23 RLU, significantly higher than pLRRK2-E (p< 0.0001). A further deletion of 259 bp from pLRRK2-F to pLRRK2-G lowered promoter activity to 9.89 ± 0.56 RLU in a significant way (p< 0.0001). Moreover, the promoter activity of pLRRK2-J was negligible (0.78 ± 0.02 RLU) after an 84bp fragment was deleted from pLRRK2-G. Taken together, the luciferase reporter assay suggests that the fragment -118 bp to +133 bp has the minimum promoter activity necessary for initiating human LRRK2 gene transcription. Additionally, there are certain negatively regulatory cis-acting elements located in the promoter 71 regions from -1738 bp to -794 bp and -495 bp to -413 bp, and promoter regions covering -794 bp to -495 bp and -413 bp to -118 bp have positively regulatory cis-acting elements. Figure 2.2 Deletion analyses of the human LRRK2 promoter activity. (A) The human LRRK2 gene promoter was cloned from HEK293 gDNA, and a serial deletion within a 5’-flanking region was performed. The promoter with varying length was inserted into luciferase reporter vector, pGL3-Basic, to generate plasmids for luciferase assay. The numbers illustrate the start and end point of each fragment relative to the TSS and the arrows represent the direction of transcription. (B) The construction of plasmids used in luciferase assay was verified by restriction enzyme digestion, and the digested products were resolved on 1.5% agarose gel. (C) LRRK2 promoter plasmids were cotransfected into HEK293 cells with pCMV-Luc. After 24 hours of transfection, cell lysates were harvested and the luciferase activity of each promoter fragment was normalized to that of pCMV-Luc. The RLU of pGL3-Basic (marked as N) was designated as 1. The values represent means ± SEM. N=3, ***p< 0.001, by one-way ANOVA with post-hoc Bonferroni's multiple comparison test. Comparisons were made between all other columns and the pGL3-Basic control column. 2.3.3 Upregulation of the LRRK2 promoter activity by Sp1 To explore the putative regulatory cis-acting elements for the LRRK2 gene promoter, computational transcription factor search (PROMO, online tool) was performed for the 5’-flanking regions of the human LRRK2 gene. And the results showed that the human LRRK2 72 promoter contained several putative binding sites for various transcription factors, including Sp1, GATA1/2, c-Jun, HNF-3α, and NF-AT1 (Figure2.1C). More specifically, three putative Sp1 binding sites were found in the human LRRK2 gene promoter, including -537 bp to -529 bp, -263 bp to -254 bp, and -51 bp to -43 bp, and first two of them were positioned in the regions having predicted upregulating cis-acting elements (Figure 2.1C). To determine the effect of Sp1 on the human LRRK2 gene promoter, the promoter activity of the pLRRK2-C with all three putative Sp1 binding sites was evaluated in HEK293 cells after being co-transfected with a Sp1 expression plasmid (pCGN-Sp1). The plasmid pLRRK2-J containing the fragment without putative Sp1 binding sites was served as a negative control in the luciferase reporter assay (Figure 2.3A). As expected, the promoter activity of pLRRK2-C significantly increased from 10.43 ± 0.68 RLU to 34.79 ± 2.01 RLU after Sp1 overexpression (p< 0.0001), but no increase was observed after transfecting the vector control nor for the promoter activity of pLRRK2-J (p> 0.05), suggesting that promoter activity of LRRK2 gene was upregulated by Sp1 in HEK293 cells. To confirm the effect of Sp1 on the LRRK2 promoter, the endogenous Sp1 gene was knocked down by the small interfering RNA (siRNA), targeting all three isoforms of human Sp1, and a scrambled siRNA was used as a control. The success of the knockdown experiment was validated by the significantly decreased expression of endogenous Sp1 protein after siRNA treatment in HEK293 cells (Figure 2.4I). After knocking down endogenous Sp1 gene, the promoter activity of pLRRK2-C significantly dropped from 11.57 ± 0.46 RLU to 2.53 ± 0.01 RLU (p< 0.0001), but the promoter activity of pLRRK2-J and pGL3-Basic did not significantly change. (p> 0.05, Figure 2.3B). 73 Figure 2.3 The binding between Sp1 and cis-acting elements on the LRRK2 gene promoter promotes LRRK2 promoter activity. (A) pLRRK2-C (with three putative Sp1 binding sites), pLRRK2-J (without any putative Sp1 binding sites) or pGL3-Basic were cotransfected with vector or Sp1 expression plasmid into HEK293 cells, and luciferase assay was performed as mentioned before. Sp1 overexpression significantly increased the promoter activity of pLRRK2-C but had no effect on pLRRK2-J nor pGL3-Basic. The values represent means ± SEM. N =3, ***p< 0.001 by two-way ANOVA with Bonferroni’s multiple comparison test. (B) pLRRK2-C, pLRRK2-J or pGL3-Basic were transfected into HEK293 cells treated with either Sp1 siRNA or a scrambled siRNA. Luciferase assay was performed and Sp1 siRNA treatment significantly reduced promoter activity of pLRRK2-C, but had no effect on pLRRK2-J nor pGL3-Basic. The values represent means ± SEM. N =3, ***p< 0.001 by two-way ANOVA with Bonferroni’s multiple comparison tests. (C) EMSA was applied to test the binding of Sp1 with each putative binding site in the LRRK2 gene promoter. Sp1 consensus binding motif was labeled with fluorescent dye to use as a probe. By adding Sp1-enriched nuclear extract, the binding of Sp1 and the probe produced a shifted band as shown in lane 2. Competition assays were conducted by adding various concentrations of the unlabeled competitive oligonucleotides, including consensus Sp1 binding sequence (lane 3 and 4), mutant Sp1 binding sequence (lane 5 and 6) and putative Sp1-responsive elements in the human LRRK2 promoter (lane 7 to 12). To discriminate the function of all three putative Sp1 binding sites, EMSA was performed to examine the binding between Sp1 and each cis-acting element in vitro (Figure 2.3C). pCGN-Sp1 74 was transfected into HEK293 cells to extract Sp1-enriched nuclear proteins from cell lysate. Double-stranded nucleotides containing the Sp1 consensus binding motif (attcgatcgGGGCGGGgcgagc) were synthesized as a probe and labeled with IRD700 dye. Therefore, a free probe band was resolved on the DNA PAGE gel as shown in the first lane of the Figure 2.3C. A shifted band was observed after adding Sp1- enriched nuclear extract (Figure 2.3C, Lane 2), indicating the binding between Sp1 and its consensus sequence. To ensure the binding specificity, unlabeled WT oligonucleotides, containing the Sp1 consensus binding motif, and mutant oligonucleotides, mutating two nucleotides in the Sp1 consensus binding motif, were added to the EMSA system. Consistent with the expectation, unlabeled WT oligonucleotides successfully competed with the probe binding to Sp1 in a concentration-dependent manner (Figure 2.3C, Lane 3 and 4). On the contrary, mutant oligonucleotides with 2-fold and 20-fold concentration of labeled probe had a barely competing effect (Figure 2.3C, Lane 5 and 6). Followed by these, the competitive effects of all three putative Sp1 binding sites were tested separately in the EMSA. The unlabeled oligonucleotides containing the first binding site decreased the intensities of the shifted bands in a concentration-dependent manner (Figure 2.3C, Lane 7 and 8). Similarly, the third Sp1 binding site containing- oligonucleotides of two concentrations reduced the intensities of shifted bands (Figure 2.3C, Lane 9 and 10). However, the second binding site did not have obvious competitive effect (Figure 2.3C, Lane 11 and 12). Overall, it is suggested that there are two functional Sp1 binding sites in the human LRRK2 gene promoter and the binding between Sp1 and cis-acting elements in the human LRRK2 promoter upregulates its promoter activities. 75 2.3.4 Upregulation of the LRRK2 gene expression by Sp1 To further examine the effect of Sp1 on LRRK2 gene expression, endogenous LRRK2 mRNA levels were evaluated after transfecting either pCGN-Sp1 expression plasmid or control vector into HEK293 cells. As shown in the Figure 2.4A and Figure 2.4B, after normalizing to β-actin, Sp1 overexpression significantly elevated LRRK2 mRNA level by 83.1 ± 18.4% in comparison to the control (p= 0.021). Besides, the experiment was performed in dopaminergic MN9D cells to test whether there was a cell type- specific effect. In consistent with the previous results, a significant increment of 44.10 ± 2.28% (p= 0.0006) was detected for the relative LRRK2 mRNA expression level after overexpressing Sp1 in MN9D cells (Figure 2.4C-D). On the contrary, knockdown of endogenous Sp1 was achieved by treating HEK293 cells with Sp1 siRNA (Figure 2.4E), and it resulted in significant decrease of the endogenous Sp1 protein (p= 0.031, Figure 2.4F). Since Sp1 overexpression enhanced LRRK2 mRNA expression, it is necessary to examine whether Sp1 overexpression can upregulate the expression of LRRK2 gene at the protein level. By transfecting Sp1 expression plasmid into HEK293 cells, the endogenous LRRK2 protein was significantly elevated by 81.7 ± 6.21% (p= 0.0003) compared with vector transfection measured by immunoblotting (Figure 2.4G-H). On the contrary, endogenous Sp1 gene was knocked down by siRNA, and it led to significantly declined expression of LRRK2 protein in HEK293 cells by 74.52 ± 9.09% (p= 0.0005, Figure 2.4I-J). However, when Sp1 was overexpressed in MN9D cells, endogenous LRRK2 protein cannot be detected in this cell line by any antibodies we tried (data not shown). 76 Figure 2.4 Sp1 upregulates LRRK2 gene expression at both mRNA and protein levels. (A-D) Sp1 expression plasmid was transfected to HEK293 cells A-B) or MN9D cells C-D). Cell lysates were harvested 48 hours post-transfection and endogenous LRRK2 mRNA was measure by RT-PCR. The products from RT-PCR were resolved on 1.5% agarose gel. Quantification was performed by ImageJ software and endogenous LRRK2 mRNA level was normalized against β-actin. (E) HEK293 cells were treated with either Sp1 siRNA or a scrambled siRNA for 48 hours and mRNA was extracted from cell lysate for RT-PCR. The amplified LRRK2 and β-actin genes were analyzed on 1.5% agarose gel. (F) Quantification was performed by ImageJ software and endogenous LRRK2 mRNA level was normalized against β-actin. (G-J) HEK293 cells were either transfected with Sp1 expression plasmid (G) or treated with Sp1 siRNA for 48 hours. Endogenous LRRK2 protein and Sp1 expression at the protein level was detected by immunoblotting. Quantification was completed by ImageJ software and endogenous LRRK2 protein was normalized against β-actin. The values represent means ± SEM. N =3, *p< 0.05, **p< 0.01, and ***p< 0.001 analyzed by Student’s t-test. 77 2.3.5 Inhibition of the LRRK2 promoter activity and gene expression by MTM MTM is a selective Sp1 inhibitor, competing with Sp1 to bind to a GC-rich DNA sequence (Gidoni et al., 1985). To investigate whether transcriptional activation of LRRK2 gene can be inhibited by interfering with the cellular function of Sp1, MTM was applied to either HEK293 cells or MN9D cells, and its effects on LRRK2 promoter activation and gene expression were examined. The plasmid pLRRK2-C or pLRRK2-J was transfected into HEK293 cells and MTM treatment was applied for varying duration. 125nM MTM treatment was able to significantly decrease the promoter activities of the pLRRK2-C, which contained Sp1 binding sites, from 9.72 ± 0.22 RLU to 4.40 ± 0.16 RLU after 24 hours and further to 3.53 ± 0.08 RLU after 48 hours (p< 0.0001 for both 24 and 48 hours). Neither pLRRK2-J nor pGL3-Basic’s promoter activities were affected by MTM treatment (p> 0.05, Figure 2.5A). Concerning the dosage effect of MTM treatment, various concentrations of MTM were added to the culture medium of HEK293 cells, ranging from 25nM to 125nM. After 24 hours, the promoter activity of pLRRK2-C was significantly declined from 9.35 ± 0.41 RLU to 3.53 ± 0.14 RLU in 25nM MTM treatment, further decreased to 1.37 ± 0.09 RLU in 75nM, and 0.38 ± 0.02 RLU in 125nM (p< 0.0001 for 25nM, 75nM and 125nM. Figure 2.5B). In general, the data suggests that MTM treatment interferes with LRRK2 promoter activities in a time- and dosage-dependent manner. To determine the role of MTM treatment in regulating LRRK2 mRNA expression, HEK293 cells were administrated with 125nM MTM for 24 hours, and endogenous LRRK2 mRNA level was measured by RT-PCR. It was demonstrated that significant downregulation of endogenous LRRK2 mRNA level (63.43 ± 1.66%, p< 0.0001, Figure 2.5C-D) was observed after normalizing to β-actin. Consistent with findings for LRRK2 mRNA, endogenous LRRK2 78 protein expression in HEK293 cell was decreased to 38.30 ± 2.18% (p< 0.0001, Figure 2.5E-F) after 125nM MTM treatment for 24 hours. Similarly, the reduction of LRRK2 mRNA expression after MTM treatment was replicated in MN9D cells, from 100.00 ± 4.96% to 52.44 ± 2.07% (p= 0.0009, Figure 2.5G-H). Taken together, inhibition of Sp1’s activity by MTM downregulates the LRRK2 promoter activity and gene expression. Figure 2.5 MTM treatment inhibits LRRK2 promoter activity and gene expression. (A-B) pLRRK2-C, pLRRK2-J, or pGL3-Basic was transfected into HEK293 cells. The next day, MTM was applied to the transfected cell with A)125mM MTM for varying duration or with B) various concentrations for 24 hours. Luciferase assay was performed as mentioned before. The values represent the means ± SEM. N =3, ***p< 0.001 by two-way ANOVA with Bonferroni’s multiple comparison test. (C-F) 125 nM MTM or vehicle was administrated to HEK293 cells for 24 hours. Endogenous LRRK mRNA level was examined by RT-PCR C-D) and endogenous LRRK2 and β-actin protein level were measured by immunoblotting E-F). The intensities of the bands were quantified by ImageJ software. (G-H) 125 nM MTM or vehicle was administrated to MN9D cells for 24 hours. LRRK mRNA levels were determined by RT-PCR and normalized against the levels of β-actin. The intensities of the bands were quantified by ImageJ software. The values in D), F) and H) represent the means ± SEM. N =3, ***p< 0.001 by Student’s t-test. 79 2.4 Discussion Since the successful cloning of LRRK2 in 2004, it has become one of the hot topic in studying the pathogenesis of PD (Paisan-Ruiz et al., 2004; Zimprich et al., 2004). LRRK2 is expressed in many organs and tissues, such as lung, heart, liver, pancreas, and kidney (Paisan-Ruiz et al., 2004). However, it is weakly expressed in the dopaminergic neurons, but is enriched in the dopamine-innervated brain areas (Galter et al., 2006). It has been reported that the overexpression of LRRK2 WT results in cytotoxicity and the toxic effects of LRRK2 mutant is dependent on its expression level (Liu et al., 2008; Martin, Kim, Lee, et al., 2014; Skibinski, Nakamura, Cookson, & Finkbeiner, 2014). Therefore, the clarification of the mechanisms underlying its transcriptional and translational regulation could help to provide potential strategies to modulate its level. In the present study, we cloned the LRRK2 promoter, and identified its TSS and minimal promoter required for transcriptional initiation. We mapped two cis-acting Sp1-responsive elements in the human LRRK2 promoter region. Furthermore, Sp1 overexpression enhanced endogenous LRRK2 mRNA and protein expression in the cellular models. On the contrary, MTM treatment downregulated LRRK2 gene expression. LRRK2 is a large multifunctional protein, containing a GTPase ROC domain, a kinase domain, and a COR domain in between. Most of the pathogenic mutations are located in these central domains. Thus, the roles of kinase and GTPase activities in the pathogenesis have been investigated extensively (Greggio & Cookson, 2009). It is consistently found that LRRK2 G2019S mutation elevates the kinase activities and its overexpression leads to cell demise and inclusion formation in neuronal cells and primary neurons (Greggio et al., 2006; Jaleel et al., 2007; Smith et al., 2005; West et al., 2005). Oppositely, the LRRK2 mutants with decreased 80 kinase activities result in reduced cytotoxicity and the kinase-dead mutants rescue cell death and inhibit inclusion formation (Greggio et al., 2006; Smith et al., 2006). However, it is still undetermined that whether kinase activity is the major mediator of cell toxicity, since other PD-linked LRRK2 mutants did not affect or decrease kinase activities (Greggio & Cookson, 2009). Another line of evidence supports that the pathogenic effects of LRRK2 mutations are mediated by the GTPase ROC domain. It has been found that PD-associated mutations in the ROC (R1441G/C/H) and COR domains (Y1699C) reduce the rate of GTP hydrolysis (Daniels et al., 2011; Guo et al., 2007; Lewis et al., 2007; Li et al., 2007; Liao et al., 2014). To compare the detrimental effects of different functional domains in a yeast model, expression of GTPase domain alone is sufficient to impair yeasts’ viability, but the expression of kinase domain was much less toxic. Notably, the fragment containing the ROC-COR-kinase domain renders the most toxic effect, indicating kinase and COR domains may modulate GTPase- mediated toxicity (Xiong et al., 2010). Taken together, it is more likely that both of the kinase and GTPase domains work together and regulates the overall pathological functions of LRRK2. Compared with the intensive works on LRRK2’s pathophysiological functions, the studies focusing on its promoter are relatively limited. A previous study has proven that at least six TSSs, ranging from 48 bp to 120 bp upstream of the ATG, are located in the LRRK2 promoter region by using cDNA from human brain (West et al., 2005). However, our results showed that the TSS was mapped to 135bp upstream of the ATG, which was 15 bp further. It is possibly due to the cell type- or tissue type- specific effect (Consortium et al., 2014). In our study, we focused on human LRRK2 promoter, so Another study has reported that LRRK2 gene expression is upregulated by a nuclear receptor, glucocorticoid receptor (GR), after binding to its ligand (Park 81 et al., 2013). In the present study, we identified multiple putative binding sites in the human LRRK2 promoter region, including Sp1, c-Jun, HNF-3α, GATA-1/2, and NFAT1. Sp1 was one of the first transcription factors being cloned in the 1980s (Kadonaga, Carner, Masiarz, & Tjian, 1987). It was originally discovered as a transcription activator for simian virus 40 (SV40) (Dynan & Tjian, 1983). Sp1 is widely expressed and plays a role in the embryogenesis (Brandeis et al., 1994; Marin, Karis, Visser, Grosveld, & Philipsen, 1997) as well as cell cycle regulation and cancer (Black, Black, & Azizkhan-Clifford, 2001; Black, Jensen, Lin, & Azizkhan, 1999). Sp1 binds to GC-box and GT/CACCC-box (Giglioni, Comi, Ronchi, Mantovani, & Ottolenghi, 1989; Kadonaga, Jones, & Tjian, 1986), and its consensus binding sequence is (G/T)GGGCGG(G/A)(G/A)(C/T) (Kadonaga et al., 1986). In our study, the sequence of the first putative binding site in the human LRRK2 promoter was GGGCGGTGC, the second was CGTCCGCCCG, and the third was GGGGCGGGGA. There were two mismatched base pairs in the first binding site, three in the second binding site, and only one in the third binding site. As expectedly, only the first and the third binding sites were functional. The expression of Sp1 is induced under oxidative stress (Ryu et al., 2003). And its mRNA and protein levels are increased in the frontal cortex of AD brains as well as in the frontal cortex and hippocampus of AD model mouse (Bruce, John, Ross, & Valentina, 2008). In a yeast two-hybrid assay, huntingtin interacts with Sp1 and coactivator TAFII130, and mutant huntingtin inhibits the binding of Sp1 to DNA (Anthone et al., 2002). Moreover, previous studies in our lab suggest Sp1 is able to regulate the expression of BACE1, huntingtin, and SNAP-25 (Cai et al., 2008; Christensen et al., 2004; Wang et al., 2012b). 82 MTM works as an antitumoral and antibiotic drug with high binding affinity to GC-rich sequence (Van Dyke & Dervan, 1983). It is a well-known site-specific inhibitor for Sp1, as it competitively binds to the Sp1 consensus binding site on the SV40 promoter (Ray, Snyder, Thomas, Koller, & Miller, 1989). As expectedly, MTM treatment significantly decreased LRRK2 promoter activity and gene expression in our study. It has been reported that transgenic mice overexpressing LRRK2 WT or G2019S alone do not develop neuronal degeneration. However, SNCAA53T / LRRK2 WT double transgenic mice exhibit elevated reactive astrocytosis, microgliosis, and neuronal death, the severity of which is correlated with the expression level of LRRK2 proteins. Therefore, its suggests that LRRK2 expression level plays an essential role in facilitating SNCAA53T-induced neuropathology (Lin et al., 2009). In this case, downregulation of the LRRK2 level by MTM offers another possibility to alleviate the pathological alterations caused by PD-related mutations, although the off-target effects of Sp1 inhibition should be taken into consideration. Future studies will further explore the effects of MTM on modulating the phenotypes of LRRK2 transgenic mice and/or LRRK2/SNCA double transgenic mice. 2.5 Conclusion In summary, we cloned the human LRRK2 gene promoter, identified its TSS, and characterized its promoter region by deletion assays. Moreover, there were three putative Sp1-binding sites located in its promoter sequence, and two of them were functional. Sp1 overexpression resulted in enhanced LRRK2 gene promoter activity and gene expression at the transcription and translational level. The application of an Sp1 inhibitor, MTM, reduced the promoter activity and gene expression. This is the first study to describe that LRRK2 gene expression is modulated by Sp1 signaling. 83 Chapter 3: Cell type- specific effects of the APPSWE and SNCAA53T mutations in cholinergic and dopaminergic cells 3.1 Introduction In the previous chapter, we discussed that the expression of PD-associated LRRK2 gene was upregulated by the transcription factor Sp1 though enhancing its promoter activity. In the next two chapters, we focused on both AD and PD, two of the most common neurodegenerative disorders. They are featured by the prominent cell loss within certain neuronal populations (Holtzman, Morris, & Goate, 2011; Massano & Bhatia, 2012). Specifically, AD is characterized by the severe cholinergic deficiency in the basal forebrain as well as its innervated cortex and hippocampus, which is closely related to memory and cognitive deficits in the disease (Pinto, Lanctot, & Herrmann, 2011). Regarding PD, the prominent neuron loss is firstly identified in the dopaminergic neurons of SNpc and accordingly the substantial loss of dopamine content in the striatum, which results in motor symptoms in the PD patients (Sulzer, 2007). Besides neurodegeneration, both of AD and PD have their own pathological characteristics as the criteria for diagnosis. The neuropathological features of AD also include extracellular β-amyloid plaque and intracellular neurofibrillary tangles, which are formed by hyperphosphorylated tau protein (Brion, Couck, Passareiro, & Flament-Durand, 1985; Gorevic et al., 1986; Masters et al., 1985). The well-established pathological hallmark of PD is LBs and LNs, which are mainly composed of αSyn (Irizarry et al., 1998; Spillantini et al., 1997). The majority of the AD and PD cases are idiopathic, without clearly-defined causes. Therefore, tremendous efforts in the field have been put to disclose the etiology underlying a small subset of 84 AD and PD patients, which is termed as familial AD and PD, as these patients have known genetic contributors. These patients generally have similar phenotypes as the sporadic cases, but the age of onset is much younger (Bekris et al., 2010; Houlden & Singleton, 2012). The knowledge gained from familial cases is beneficial to deepen our understanding of the disease progression and molecular culprits behind neurodegeneration and other pathological changes, and eventually foster the development of therapeutic strategies. APPSWE mutation was first discovered in Sweden AD families in 1992 (Mullan et al., 1992), which significantly increases Aβ levels (Cai, Golde, & Younkin, 1993; Citron et al., 1992; Citron et al., 1994). Our lab further demonstrates that APPSWE mutation is able to shift major β-cleavage site from Glu11 to Asp1 within the Aβ sequence, resulting in the elevated C99 generation and Aβ production (Deng et al., 2013). Additionally, it has been reported that APPSWE mutation exerts neurotoxic effects and increases cell susceptibility to oxidative stress (Eckert, Steiner, Marques, Leutz, Romig, Haass, & Muller, 2001; Marques et al., 2003). SNCAA53T mutation was the first genetic mutation identified in PD (Polymeropoulos et al., 1997), and increasing evidence supports that SNCAA53T mutation accelerates αSyn polymerization and aggregation (Conway et al., 1998; K. Conway et al., 2000; B. Giasson, Uryu, Trojanowski, & Lee, 1999; Narhi et al., 1999). In addition, overexpression of SNCAA53T increases cell death under oxidative stress (Jiang et al., 2007; Prasad et al., 2004). Although both APP protein and αSyn are ubiquitously expressed in the brain (Jakes et al., 1994; Neve, Finch, & Dawes, 1988; Uéda et al., 1993), the APPSWE and SNCAA53T mutations lead to a selective pattern of neuronal demise. The underlying mechanisms are still unknown. 85 Steady accrual of data suggests that the soluble oligomeric species rather than the insoluble fibrils are the more toxic structures in both AD and PD (Benilova, Karran, & De Strooper, 2012; Kalia, Kalia, McLean, Lozano, & Lang, 2013). Aβ oligomers are secreted from cells overexpressing APP mutations (Podlisny et al., 1995; Xia et al., 1997), and also exist in the brain of APPSWE transgenic mice (Cheng et al., 2007; Lesne et al., 2006). In AD patients, Aβ oligomers are detectable in the CSF and postmortem brain samples (Mc Donald et al., 2010; Santos et al., 2012). Previous studies suggests that Aβ oligomers induce neuronal death (Kim et al., 2003; Lambert et al., 1998), reduce synaptic plasticity (Decker et al., 2010; Walsh et al., 2002), impair cognitive function and memory in mice (Cleary et al., 2005; Lesne et al., 2006). The novel understanding about the extracellular existence of αSyn is opposite to the notion that LBs and LNs are located intracellularly (Lee, Bae, & Lee, 2014). Like Aβ oligomer, αSyn oligomers are released from cultured cells (Danzer et al., 2012; Danzer et al., 2011; Emmanouilidou et al., 2010) and can be detected in the CSF and brain samples of PD patients (Paleologou et al., 2009; Tokuda et al., 2010). Application of αSyn oligomers results in neuronal death, synaptic dysfunction, and LTP deficit (Danzer, Krebs, Wolff, Birk, & Hengerer, 2009; Danzer et al., 2011; Diogenes et al., 2012). Collectively, these impressive studies indicate the existence of extracellular Aβ and αSyn oligomeric assemblies and their cytotoxic effects. In this study, hybrid SN56 and MN9D cell lines were employed. Immortalized SN56 cell lines were derived from the fusion of N18TG2 neuroblastoma cells with mice septal neurons (Hammond et al., 1986), and the MN9D cell line was generated by the fusion of mice embryonic mesencephalic dopaminergic neurons and N18TG2 cells (Choi et al., 1991). The neuronal properties of these two cell lines were validated by expression of specific neuronal markers, 86 neurite growth, and electrophysiology (Choi et al., 1991; Hammond et al., 1986). The cell type- specific effects of APPSWE mutation on amyloidogenesis and SNCAA53T mutation on fibrillogenesis were examined in these two cell lines. Furthermore, we tested whether the mutation could change the cell susceptibility to oxidative stress and oligomeric insults and whether the two cell lines carrying the same mutation would respond differentially to the toxic insults. 3.2 Methods 3.2.1 Construction of plasmids The expression plasmid of pcDNA4-hAPPWT and pcDNA4-hAPPSWE was generated by cloning human APP695 cDNA and human APP695 cDNA with K595N/M596L double mutation into pcDNA4-mycHis vector without mycHis tag (Qing et al., 2004).Human SNCA140 cDNA and SNCA140 cDNA with A53T mutation was amplified from pcDNA3-SNCAWT-myc and pcDNA3-SNCAA53T-myc by PCR, respectively (Xu et al., 2002). The pair of primers for PCR was SNCA-HindIII F (5’- gggaagcttgccaccatggatgtattc) and SNCA-XhoI R (5’- ccgctcgagggcttcaggttcgtagtc). The PCR products were purified and cloned into pcDNA4-mycHis vector (Invitrogen) at the HindIII and XhoI sites with mycHis tag. Mammalian expression vector pAPP-C99, pAPP-C89, and pAPP-C83 were generated by cloning APP C99, C89 and C83 cDNA into pcDNA3 expression vector (Invitrogen) (Qing et al., 2004). 3.2.2 Cell culture and generation of stable cell lines Cholinergic SN56 cells and dopaminergic MN9D cells were cultured in DMEM supplemented with 10% FBS, 1 mM of SP, 2 mM of L-glutamine, 50 units of penicillin and 50 μg of 87 streptomycin (Invitrogen). The culture dishes (Corning) for maintaining MN9D cells were coated with 10 μg/mL PDL (Sigma). Both SN56 and MN9D cells were maintained at 37°C in an incubator containing 5% CO2 and cell medium was changed when necessary. Before transfection, cells were seeded on the plate at 50% - 70% confluency and all transfection experiments were performed by lipofectamine 2000 (Invitrogen) following manufacturer’s instruction. To establish stable cell lines, SN56 and MN9D cells were transfected with pcDNA4-hAPPWT, pcDNA4-hAPPSWE, pcDNA4-hSNCAWT-mycHis, or pcDNA4-hSNCAA53T-mycHis vectors, and transfected cells were selected with 400 μg/mL zeocin (Invitrogen). Stable cell lines were maintained in medium containing 50 μg/mL zeocin. 3.2.3 Immunoblotting APP-related stable cells, SNCA-related stable cells, and their parental SN56 and MN9D cells were lysed in Radioimmunoprecipitation assay (RIPA) lysis buffer containing 1% triton X-100, 0.1% SDS, 1% sodium deoxycholate, 150 mM sodium chloride, 50 mM Tris-HCl (pH 7.2) and protease inhibitor cocktail (Roche), followed by brief sonication. Protein concentration was determined by Bradford assay (Bio-rad) and each sample was diluted with 4x SDS loading buffer (200 mM Tris-Cl pH 6.8, 400 mM DTT, 8% SDS, 0.4% bromophenol blue, 40% glycerol). For detecting caspase-3, cells were lysed with Chaps cell extract buffer (Cell Signaling) supplemented with 5 mM DTT and 1 mM phenylmethane sulfonyl fluoride (PMSF). Lysates were frozen and thawed three times followed by centrifuging at 14,000 rpm for 10 min. For detecting full-length APP, CTFs and cleaved caspase3, samples were diluted in 4x SDS-sample 88 buffer and resolved on 8% Tris–glycine SDS–PAGE and 16% tris-tricine SDS-PAGE. Details of the immunoblotting procedures were indicated in Section 2.2.8. Full-length APP and CTFs were detected by using rabbit C20 antibody, which recognizes the last 20 C-terminal amino acids of APP (Li, Zhou, Tong, He, & Song, 2006). Other antibodies included mouse monoclonal anti-myc antibody 9E10 (1:100, Santa Cruz Biotechnology), rabbit polyclonal anti-caspase3 antibody (1:1000, Cell Signaling) and mouse monoclonal anti-β-actin AC-15 antibody (1:5000, Sigma). 3.2.4 Aβ40/42 enzyme-linked immunosorbent assay (ELISA) Conditioned cell culture media was collected from four APP-related stable cell lines. Protease inhibitor cocktail was added to prevent degradation of Aβ proteins. The concentrations of Aβ40 and Aβ42 were detected by human Aβ40 and Aβ42 colorimetric ELISA kit (Covance) following manufacturer’s instructions. 3.2.5 Immunocytochemistry SN56-related stable cells were seeded onto PDL-coated glass cover slips in 24 well plates the day before immunocytochemistry. Then cells were rinsed in PBS, and fixed in 4% PFA for 30 mins at RT. After permeabilizing the cells with 0.1% Triton X-100 in PBS for 5 mins, cells were blocked in 5% BSA in PBS for 30 min, and incubated with primary antibodies in blocking solution 1hr at RT. The primary antibodies included rabbit monoclonal anti-αSyn antibody (Santa Cruz Biotechnology, 1:100), mouse monoclonal anti-Hsp70 antibody (Thermo Fisher Scientific, 1:50), and mouse monoclonal anti-Ubiquitin antibody (Millipore, 1:100). For detecting αSyn aggregates by Thioflavin-S, fixed cells were incubated with 0.05% Thioflavin-S for 5mins without permeabilization, and then washed with 80% ethanol for 10mins three times 89 (Ostrerova-Golts et al., 2000). After thoroughly washing with PBS, cells were incubated with goat anti-mouse Alexa Fluor 488 (green) or goat anti-rabbit Alexa Fluor 594 (red) for 1hr, rinsed in PBS-Tx, and mounted using Fluoromount-G with DAPI (Southern Biotech). Cells were imaged with a 100x oil objective lens on a Carl Zeiss Axiovert-200 epi fluorescent microscope. 3.2.6 Lactase dehydrogenase (LDH) assay and MTS assay APP-related stable cells, SNCA-related stable cells and their parental SN56 and MN9D cells were seeded on 96-well plate one day before treatment. On the next day, cells were treated with varying concentration of hydrogen peroxide (H2O2) or oligomers for 6hours or 12hours, and cell medium was collected to perform LDH assay (Promega). When the cell membrane was damaged in nonviable cells, LDH was released into the cell medium. LDH in the cell medium was measured with a fluorescently coupled enzymatic assay, resulting in the conversion of resazurin into resorufin (Promega). The fluorescence was recorded with an excitation wavelength of 560nm and an emission wavelength of 590nm. Cells were cultured on 96-well plates the same as the LDH assay, and MTS assays were performed following manufacturer’s instruction (Promega) to measure cell viability after H2O2 treatment. MTS was bioreduced by viable cells into a formazan, the absorbance of which can be measured at the wavelength of 490nm by a bioreader (Biotek, Synergy HT). 3.2.7 Preparation of Aβ oligomers and αSyn oligomers Aβ oligomers were prepared according to the protocols developed previously by two studies (Fa et al., 2010; Stine, Dahlgren, Krafft, & LaDu, 2003). Briefly, synthetic Aβ1-42 was resuspended in 90 1,1,1,3,3,3-Hexafluoro-2-Propanol (HFIP, Fluka) to monomerize Aβ1-42 at the concentration of 1mM. The Aβ/ HFIP solution was vacuum dried into a film and then dissolved in DMSO as a 5mM stock. Aβ oligomer was prepared by diluting the stock Aβ/DMSO in sterile PBS to achieve 100 μM and incubated at 4°C for 12 hours. For treating the cells. the oligomers were further diluted to 1 μM and 5 μM in culture medium, and DMSO was used as vehicle control. Preparation of αSyn oligomers was followed the protocol previously described (Danzer et al., 2007). Recombinant αSyn protein (rPeptide) was dissolved in 50 mM PBS containing 20% ethanol to a final concentration of 7 μM. After shaking for 4 hours, dissolved αSyn was lyophilized and resuspended in 50mM PBS containing 10% ethanol to the concentration of 14 μM with 24 hours shaking. The oligomer was formed after 6 days incubation at room temperature. When treating the cells, αSyn was further diluted into 7 μM in the cell culture medium, and 50mM PBS containing 10% ethanol was used as vehicle control. 3.2.8 Caspase-3/7 activity assay Cells subject to caspase-3/7 activity assay (Promega) were seeded on 96-well plate. After receiving H2O2 treatment, caspase-3/7 in the remaining cells were able to cleave the non-fluorescent caspase substrate, Z-DEVD-R110, to create the fluorescent Rhodamine 110, the excitation of which was measured at the wavelength of 499nm. 3.2.9 Statistical analyses All results were presented as means ± SEM. For multiple-group comparison, the data were analyzed by one-way ANOVA or two-way ANOVA with/ without post-hoc Bonferroni's 91 multiple comparisons test. Statistical significance was accepted when p<0.05 (*p<0.05, **p<0.01, ***p<0.001). 3.3 Results 3.3.1 APPSWE mutation enhances the amyloidogenesis in cholinergic SN56 cells APPSWE mutation has been consistently found to increase Aβ generation (Cai et al., 1993; Citron et al., 1992), and our lab also identified it can shift the APP processing from the non-amyloidogenic to the amyloidogenic pathway, elevating the C99/ C83 ratio and providing more γ-secretase substrates to generate Aβ (Deng et al., 2013). Therefore, the pattern of APP processing was first identified in the two cell lines, cholinergic SN56 and dopaminergic MN9D cells, to determine whether there were cell type- specific effects. Cholinergic SN56 and dopaminergic MN9D cells were stably transfected with either pcDNA4-APPWT or pcDNA4-APPSWE vector to overexpress comparable levels of human full-length APP protein (Figure 3.1A), and their CTF patterns were examined by western blotting (Figure 3.1B). The cells transfected with pcDNA4 vector were used as negative controls, and the cells transfected with C99, C89, and C83 expression vectors were used as the size markers for CTFs. In SN56 cells, APPWT overexpression resulted in C83 as the major CTF product, representing the predominant α-cleavage. Similarly, the predominant CTF generated in MN9D-APPWT cells was C83. The results were consistent with the previous findings in our lab, which suggest the majority of APPWT undergoes α-cleavage in HEK293 cells (Deng et al., 2013). While APPSWE was stably overexpressed in either SN56 or MN9D cells, besides the major CTFα C83 band, other two bands appeared, known as CTFβ C89 and C99. More importantly, the C99/ C83 ratio was increased in cholinergic SN56 cells compared with dopaminergic MN9D cells, suggesting 92 APPSWE mutation might potentiate the amyloidogenic pathway to a more extent in cholinergic SN56 cells than in dopaminergic MN9D cells. To further test the products of APP processing by the sequential β- and γ- cleavage, Aβ40 (Figure 3.1C) and Aβ42 (Figure 3.1D), were measured by Elisa. The conditioned medium from APPWT or APPSWE stable cells was collected for Aβ Elisa. The reading of Aβ level was normalized to β-actin and expression levels of the full-length APP proteins. Stable overexpression of APPWT in SN56 cells produced barely detectable Aβ40 in the culture medium (15.16 ± 0.27 pg/ ml), while APPSWE-overexpressing SN56 cells significantly elevated Aβ40 to 2500.00 ± 147.70 pg/ml (p= 0.0090). With respect to MN9D cells, APPSWE stable overexpression (1065.00 ± 57.75 pg/ ml) also significantly increased Aβ40 compared with APPWT (p= 0.0082). Additionally, the level of Aβ40 in SN56-APPSWE stable cells was significantly higher than that in MN9D-APPSWE stable cells (p= 0.0121, Figure 3.1C). As Aβ42 is more fibrillogenic than Aβ40 (Burdick et al., 1992; J. Kim et al., 2007), the generation of Aβ42 was further examined by Elisa. SN56-APPSWE (131.70 ± 5.47 pg/ ml) produced significantly more Aβ42 compared with SN56-APPWT (0.58 ± 0.07 pg/ ml, p= 0.0041), and the level of Aβ42 was significantly higher in MN9D-APPSWE (25.34 ± 2.67 pg/ ml) than MN9D-APPWT (0.07 ± 0.01 pg/ ml, p= 0.0276). Furthermore, APPSWE stable overexpression produced ~ 5.1 folds Aβ42 in cholinergic SN56 cells compared to dopaminergic MN9D cells (p= 0.0004, Figure 3.1D). Collectively, the predominant Aβ species in our established APPWT and APPSWE stable cell lines was Aβ40, and APPSWE increased the C99/ C83 ratio in cholinergic SN56 cells and correspondingly promoted Aβ40 and Aβ42 generation compared with dopaminergic MN9D cells. 93 Figure 3.1 APPSWE mutation increases C99/ C83 ratio and promotes Aβ generation in cholinergic SN56 cells. (A) Cholinergic SN56 and dopaminergic MN9D cells were stably transfected with plasmids to express human APPWT or APPSWE. The empty vector was transfected into cells as the negative control. The established four stable cell lines overexpressing comparable levels of full-length APP were confirmed by 8% glycine SDS-PAGE gel. (B) The CTF patterns were examined in SN56-APPWT, SN56-APPSWE, MN9D-APPWT, and MN9D-APPSWE four stable cell lines by 16% tricine SDS-PAGE gel. APPWT overexpression produced the prominent CTFα C83 in both SN56 and MN9D cell lines, while APPSWE overexpression resulted in the appearance of CTFβ C89 and C99 in both SN56 and MN9D cells. More importantly, the C99/ C83 ratio was higher in SN56-APPSWE stable cells compared with MN9D-APPSWE stable cells. (C-D) The conditioned medium from cultured APPWT or APPSWE stable cells was collected to perform Aβ40 C) and Aβ42 D) Elisa. Aβ40 and Aβ42 species were barely detected in APPWT-overexpressing cells. APPSWE significantly promoted Aβ40 and Aβ42 production in both SN56 and MN9D cells. For both Aβ40 and Aβ42, their levels were significantly higher in SN56-APPSWE cells compared with MN9D-APPSWE cells (Aβ40 p< 0.05; Aβ42 p< 0.001). The values in C-D) represent the means ± SEM. N =3, *p< 0.05, **p< 0.01, and ***p< 0.001 by one-way ANOVA followed by post-hoc Turkey multiple comparison test. 3.3.2 SNCAA53T mutation promotes the formation of cytoplasmic aggregates in dopaminergic MN9D cells SNCAA53T mutation has been found to make αSyn protein more prone to form insoluble fibrils under in vitro conditions (Conway et al., 1998; Giasson et al., 1999; Narhi et al., 1999), and αSyn overexpression in the cells induces the formation of inclusion-like aggregates, which are 94 positively stained for thioflavin-S (Ostrerova-Golts et al., 2000), ubiquitin (Ostrerova-Golts et al., 2000; Spillantini, Crowther, Jakes, Hasegawa, & Goedert, 1998), and Hsp70 (Auluck, Chan, Trojanowski, Lee, & Bonini, 2002). Therefore, we tested whether the SNCAA53T mutation behaved differently in cholinergic SN56 and dopaminergic MN9D cells in terms of its aggregation-accelerating effects. Cholinergic SN56 and dopaminergic MN9D cells were stably transfected with either pcDNA4-SNCAWT-mycHis or pcDNA4-SNCAA53T-mycHis vector to overexpress comparable levels of human wildtype or mutant αSyn protein as shown in the Figure 3.2A. To determine the properties of αSyn aggregates formed in our cellular models, the immunocytochemistry was applied to the established SNCAWT- or SNCAA53T- stable cell lines to examine the profiles of the molecular markers mentioned before. As presented in Figure 3.2B, overexpression of either wildtype αSyn or αSyn protein bearing A53T mutation induced cytoplasmic aggregates formation co-immunostained for human αSyn protein and thioflavin-S, which detects the highly ordered β-sheet structure in the amyloid-like fibrils (Lee & Lee, 2002). The aggregates positive for both αSyn and thioflavin-S were shown as condensed round structures in stably transfected cells, while thioflavin-S staining was negative in both SN56 and MN9D cells (data not shown). Only ~10% stable cells were observed with 1~2 aggregates per cell without significant differences between the 4 stable cell lines (data not shown). Intriguingly, in terms of the aggregates size, the formation of the largest cytoplasmic aggregates was detected in dopaminergic MN9D cells overexpressing SNCAA53T mutation. As ubiquitin was identified as a component of LBs-like inclusions in cell culture by previous research (Ostrerova-Golts et al., 95 2000; Spillantini et al., 1998), we further used the same strategy to examine the existence of ubiquitin in the aggregates formed in our cell models. As expected, the αSyn aggregates were co-stained with ubiquitin and αSyn, and the size of aggregates was much larger in the MN9D-SNCAA53T stable cells compared with the other three stable cell lines (Figure 3.2C). Following this, Hsp70 antibody was used to detect these αSyn aggregates. Interestingly, the αSyn aggregates appeared in SNCAWT- and SNCAA53T- overexpressing SN56 cells were devoid of Hsp70 staining, while the αSyn aggregates present in both SNCAWT- and SNCAA53T- overexpressing MN9D cells were positive for Hsp70 staining, suggesting cell type- specific effects on the molecular properties of αSyn aggregates. Overall, SNCAA53T mutation increased the size of the cytoplasmic αSyn aggregates in dopaminergic MN9D cells compared with SN56 cells, and the aggregates formed in SN56 and MN9D cells presented differential profiles of presenting molecular markers, suggesting SNCAA53T mutation facilitates LBs-like inclusion formation only in MN9D cells other than in SN56 cells. 96 Figure 3.2 SNCAA53T mutation increases the size of LBs-like inclusions in dopaminergic MN9D cells. (A) Cholinergic SN56 and dopaminergic MN9D cells stably overexpressed human wildtype or mutant αSyn protein tagged with mycHis. The empty vector was transfected into the cells as the negative control. The established four stable cell lines overexpressing comparable levels of αSyn protein were confirmed by 16% tricine SDS-PAGE gel. (B-D) Cholinergic SN56 and dopaminergic MN9D cells with stable overexpression of either SNCAWT or SNCAA53T were seeded on PDL-coated slice. Once the cells reached the proper confluency, intercellular αSyn aggregates were co-immunostained with human αSyn and Thioflavin-S B), Ubiquitin C) or Hsp70 D). Note that the size of αSyn aggregates were much larger in the MN9D-SNCAA53T stable cells compared with the other three SNCA-related stable cell lines by all these three co-immunostaining methods. Additionally, Hsp70 staining was only positive for αSyn aggregates observed in dopaminergic MN9D cells, but not in cholinergic SN56 cells. Scale bar = 5 μm. 97 3.3.3 APPSWE mutation elevates oxidative stress-induced cell death in cholinergic SN56 cells through the apoptotic pathway As discussed before, we found APPSWE and SNCAA53T mutations had more severe pathogenic effects in the cholinergic SN56 and dopaminergic MN9D cells, respectively, which were represented as the neuronal populations suffering more severe neurodegeneration occurred in AD and PD. Furthermore, we were interested to examine whether these two mutations could change the susceptibility of cholinergic and dopaminergic cells respond to oxidative stress, as oxidative stress was implied as one of the mechanisms underlying cell death in both AD and PD (Praticò, 2008; William, 1997; Zhang, Dawson, & Dawson, 2000). Cholinergic SN56 and dopaminergic MN9D cells overexpressing APPWT or APPSWE were seeded on 96-well plate on the first day at the same density, and the cells were treated with varying concentrations of H2O2 for 6 hours on the second day. After briefly centrifuging the cell culture plate, cell medium was collected for LDH assay, and cells still attached to the bottom of cell culture plate were subject to MTS assay. In the LDH assay, the activity of lactate dehydrogenase released from the damaged plasma membrane of nonviable cells was measured by a fluorometric kit. The cells treated with 0 μM H2O2 were used as the negative control and the cells treated with 0.1% triton were the positive control. The readings from all groups were normalized to the positive control group (100%) and represented as a percentage. In MTS assay, MTS is bioreduced by viable cells into a formazan product, the absorbance of which can be measured at 490mm absorbance directly from 96-well assay plates. The readings from all groups were normalized to 0 μM H2O2 group, which was assigned as 100% viability. After 6 hours treatment, both SN56-APPWT and SN56-APPSWE cells showed dosage- dependent increase of cell death, 98 from 4.74 ± 0.23% in 0 μM to 19.56 ± 1.93% in 300 μM for SN56-APPWT, and from 8.27 ± 0.89% in 0μM to 27.91 ± 1.12% in 300 μM for SN56-APPSWE. The cytotoxicity was significantly elevated for SN56-APPSWE compared with SN56-APPWT under 50 μM (p= 0.0034) and 100 μM (p< 0.0001) H2O2 treatment (Figure 3.3A). Consistently, APPSWE decreased cell viability measured by the MTS assays in 50 μM (p< 0.0001) and 100 μM (p= 0.0051) H2O2 treatment with the statistical significance (Figure 3.3C). As for MN9D cells, APPSWE did not significantly increase cell vulnerability under any tested concentrations of H2O2 treatment in comparison to APPWT (Figure 3.3B), but APPSWE significantly decreased cell viability only under 100 μM (p< 0.0001) H2O2 treatment (Figure 3.3D). Therefore, APPSWE mutation is able to decrease cell viability and makes cholinergic SN56 cells more vulnerable to the oxidative stress, while these effects were barely observed in dopaminergic MN9D cells. Subsequently, it was necessary to determine the cell death pathways of APPWT- or APPSWE- stable cells after receiving H2O2 treatment, through the apoptotic pathway or non-apoptotic pathway. It is well accepted that caspase-3 cleavage is the primary activator of apoptosis (Porter & Janicke, 1999), so the caspase-3 cleavage was measured in our established stable cells as the indicator of apoptosis. The cell lysates from APPWT- or APPSWE- stable cells were harvested after 6 hours of various concentrations of H2O2 treatment, and the cleaved caspase-3 bands were resolved on 16% tricine SDS-PAGE gel as represented in Figure 3.3E-F. The intensities of the cleaved caspase-3 bands were quantified and normalized to β-actin, and the expression of cleaved caspase-3 bands in 0 μM groups was subjectively assigned as 100% (Figure 3.3G-H). In cholinergic SN56 cells, overexpression of APP bearing Swedish mutation resulted in substantial increase of caspase-3 cleavage in a dosage-dependent manner for both 50 μM and 100 μM 99 groups compared with wildtype APP (Figure 3.3G, 50 μM p=0.0002; 100 μM p< 0.0001). Similarly, overexpression of APPSWE in the dopaminergic MN9D cells lead to a significant increase of caspase-3 cleavage in both 50 μM (Figure 3.3H p< 0.0001) and 100 μM H2O2 (p< 0.0001) treatment, although the increase was less drastic compared with that in cholinergic SN56 cells. Altogether, APPSWE sensitizes SN56 cells with higher potential than MN9D cells to H2O2-induced apoptotic cell death. 100 Figure 3.3 APPSWE mutation enhances oxidative stress-induced cytotoxicity in cholinergic SN56 cells through the apoptotic pathway. (A-B) Cholinergic SN56 cells A) and dopaminergic MN9D cells B) stably overexpressing APPWT or APPSWE were treated with 0 μM to 300 μM H2O2 for 6 hours. The cell medium was collected for LDH assay. The 0.1% triton treatment was used as the positive control, and its cytotoxicity was assigned as 100%. The cytotoxicity of other groups was normalized to the positive control as a percentage. (C-D) SN56 C) and MN9D cells D) stably overexpressing either APPWT or APPSWE were treated with 0 to 300 μM H2O2 for 6 hours. The cell viability of different treatment groups within each stable cell line was normalized to 0 μM H2O2 group, which was designated as 100%. (E-F) Cell lysate from four APP-related stable cell lines was harvested 6 hours after H2O2 treatment. Prepared protein samples were resolved on 16% tricine SDS-PAGE gel for detecting caspase-3 cleavage. β-actin was used as a loading control. (G-H) The intensities of the cleaved caspase-3 bands in E-F) were quantified by ImageJ software and normalized to 0 μM group after adjusting the loading. The values in A-D) and G-H) represent the means ± SEM. N =3, *p< 0.05, **p< 0.01, and ***p< 0.001 by two-way ANOVA followed by post-hoc Bonferroni's multiple comparisons test. 101 3.3.4 SNCAA53T mutation sensitizes dopaminergic MN9D cells to oxidative stress through caspase-3 independent pathway By applying the same strategy of studying the impact of APPSWE on oxidative stress-induced cytotoxicity in SN56 and MN9D cells, either SNCAWT- or SNCAA53T- overexpressing SN56 and MN9D cells were administered with different concentrations of H2O2. After 6 hours treatment, stable cells seeded on 96-well plate were subjected to LDH assay and MTS assay to examine cell death and cell survival, respectively. In cholinergic SN56 cells, overexpression of SNCAA53T was unexpectedly decreased cell death from 27.93 ± 2.74% to 12.34 ± 0.84% in 50 μM H2O2 (p= 0.0150), from 41.89 ± 4.01% to 23.88 ± 5.33% in 100 μM H2O2 (p= 0.0044), and from 62.83 ± 5.02% to 42.99 ± 3.61% (p= 0.0018) in 300 μM H2O2 compared with WT overexpressing cells (Figure 3.4A). Consistent with the LDH assay, SNCAA53T accordingly upregulated cell viability in 50 μM (p= 0.0105) and 100 μM (p= 0.0065) concentration groups compared with SNCAWT by MTS assays (Figure 3.4C). For dopaminergic MN9D cells, 100 μM and 300 μM H2O2 triggered significantly more cell death in MN9D-SNCAA53T than MN9D-SNCAWT measured by LDH assay (p< 0.001, Figure 3.4B). Meanwhile, SNCAA53T showed more toxic effects for MN9D cells treated with 50 μM (p= 0.0019), 100 μM (p= 0.0147), and 300 μM H2O2 (p= 0.0023) in comparison to SNCAWT (Figure 3.4D). Collectively, SNCAA53T makes MN9D cells more susceptible to H2O2-induced oxidative stress than SN56 cells, suggesting cell type- dependent effects of SNCAA53T mutation on oxidative stress-related cytotoxicity. In order to investigate the cell death pathways involved in H2O2-induced cytotoxicity, the cleavage of caspase-3 protein was detected by 16% SDS-PAGE gel (Figure 3.4 E-F) and quantified for four SNCA-related stable cells after 6 hours treatment (Figure 3.4G-H). Compared 102 with wildtype SNCA, SNCAA53T mutation substantially inhibited the activation of caspase-3 in cholinergic SN56 cells under both 50 μM and 100 μM H2O2 treatment (Figure 3.4G, 50 μM, p= 0.0015; 100 μM, p< 0.0010). Similarly, even with increased concentration of H2O2 treatment shown in Figure 3.4F, the cleaved caspase-3 bands were very faint in MN9D-SNCAA53T stable cells. However, caspase-3 cleavage was drastically increased in SNCAWT- overexpressing MN9D cells after receiving H2O2 treatment (Figure 3.4H 50 μM, p= 0.0019; 100 μM, p< 0.0010). The results indicate that SNCAA53T inhibits caspase-3 cleavage induced by oxidative stress observed in both SN56 and MN9D cells. Taken together, SNCAA53T elevated oxidative stress-induced cell death of MN9D cells instead of SN56 cells, which was probably through caspase-3 independent pathway. 103 Figure 3.4 SNCAA53T mutation promotes oxidative stress-induced cytotoxicity in dopaminergic MN9D cells through caspase-3 independent pathway. (A-B) SN56 cells A) and MN9D cells B) stably overexpressing SNCAWT or SNCAA53T were treated with 0 μM to 300 μM HO2O2 for 6hours. The cell medium after treatment was collected for LDH assay. The cytotoxicity of other groups was normalized to the 0.1% triton group as percentage. (C-D) The cholinergic SN56 C) and dopaminergic MN9D cells D) with stable overexpression of either SNCAWT or SNCAA53T were treated with 0 to 300 μM H2O2 for 6 hours. The cell viability of different treatment groups was normalized to 0 μM H2O2 group, which was designated as 100%. (E-F) Cell lysate from four SNCA-related stable cells were harvested after 6hours of varying concentration of H2O2 treatment. After measuring the protein concentration, prepared protein samples were resolved on 16% tricine SDS-PAGE gel for detecting cleaved caspase-3. β-actin was used as a loading control. (G-H) The intensities of the cleaved caspase-3 bands in E-F) were quantified by ImageJ software and normalized to β-actin. The values in A-D) and G-H) represent the means ± SEM. N =3, *p< 0.05, **p< 0.01, and ***p< 0.001 by two-way ANOVA followed by post-hoc Bonferroni's multiple comparisons test. 104 3.3.5 APPSWE mutation makes SN56 cells more vulnerable to Aβ42 oligomer treatment The aforementioned hydrogen peroxide treatment is a relatively non-specific oxidative stress for AD and PD, and we further aimed to test more specific disease-related insults in our cell models. As introduced in the previous sections, Aβ oligomers have been reported to be present in the cell culture medium (Podlisny et al., 1995; Walsh et al., 2002), in the brain of AD mouse models (Lesne et al., 2006; Oddo et al., 2006), and in the AD brains (Gong et al., 2003; McLean et al., 1999). Aβ oligomers can be prepared from synthetic Aβ peptide under certain conditions in vitro, and these Aβ oligomeric species are able to induce neuronal toxicity (Deshpande, Mina, Glabe, & Busciglio, 2006; Isaacs, Senn, Yuan, Shine, & Yankner, 2006; Lambert et al., 1998) and negatively affect electrical activities of the neurons (Hartley et al., 1999; Ye, Selkoe, & Hartley, 2003). Aβ42 is consistently found to be more toxic than Aβ40 (Burdick et al., 1992; Jarrett, Berger, & Lansbury, 1993; Kim et al., 2007). Therefore, we prepared the Aβ42 oligomers from synthetic peptides by a method as mentioned in previous studies (Fa et al., 2010; Stine et al., 2003), and explored the APPSWE mutation’s effects on the cell viability of SN56 and MN9D cells treated with Aβ42 oligomers. APPWT- and APPSWE- stable cells were applied with 1 μM and 5 μM Aβ42 oligomers for 15 hours, as we found extended duration lowered their toxicity (data not shown). By measuring cell death by LDH assay represented in Figure 3.5A, Aβ42 oligomers increased cell death in SN56 cells overexpressing APPWT to 173.33 ± 0.51% in 1 μM (p=0.0029), and to 164.55 ± 9.29% in 5 μM (p=0.0071, Figure 3.4A) relative to 0 μM. Similarly, Aβ42 oligomers elevated the cytotoxicity of SN56-APPSWE stable cells to 177.08 ± 18.98% in 1 μM (p= 0.0020), and to 184.190 ± 29.64% in 5 μM (p=0.0010, Figure 3.5A). It is necessary to mention that Aβ42 oligomers treatment caused much more cell death in SN56-APPSWE stable cells compared with SN56-APPWT stable cells treated with 5 μM oligomers (p=0.0037). 105 However, neither 1 μM nor 5 μM Aβ42 oligomers caused apparent cytotoxicity in APPWT- and APPSWE- overexpressing MN9D cells (p> 0.05), suggesting that APPSWE mutation only rendered cholinergic SN56 cells more sensitive to Aβ42 oligomer- induced cell death, but not in dopaminergic MN9D cells. Next, the cell death pathways under Aβ42 oligomer treatments were examined by detecting caspase-3/7 activity, which was indicated by their abilities to cleave a specific profluorescent substrate rhodamine 110, bis-(N-CBZL-aspartyl-L-glutamyl-L-valyl-L-aspartic acid amide; Z-DEVD-R110). As shown in the Figure 3.5B, only 5 μM Aβ42 oligomers led to increased caspase-3/7 cleavage in SN56-APPWT stable cells (p=0.0423), but not seen in the lower concentration groups (p>0.05). Regarding SN56-APPSWE stable cells, 1 μM and 5 μM Aβ42 oligomers were able to increase caspase-3/7 activation by 69.81 ± 8.65% (p= 0.001) and 50.94 ± 13.21% (p= 0.0031) compared with vehicle control group. Consistent with LDH assay, none of the tested concentration of Aβ42 oligomers had obvious effect on activating caspase-3/7 in dopaminergic MN9D cells overexpressing either APPWT or APPSWE mutation. Taken together, Aβ42 oligomer caused more severe cell death in SN56 cells bearing APPSWE mutation than its wildtype counterparts, while this effect was absent in MN9D cells overexpressing either APPWT or APPSWE. 106 Figure 3.5 Aβ42 oligomers increase cell death and caspase-3/7 activation in SN56-APPSWE cells but not MN9D-APPSWE cells. (A) 1 μM or 5 μM Aβ42 oligomers were prepared from the synthetic Aβ42 peptide and treated APPWT- or APPSWE- stable cells for 15 hours. The cell medium after treatments was collected for LDH assay. The vehicle for diluting Aβ42 oligomers was used as the negative control, and the cytotoxicity induced by Aβ42 oligomers were normalized to the negative control. 1 μM or 5 μM Aβ42 oligomers significantly promoted cell death in both SN56-APPWT and SN56-APPSWE cells, but not in any MN9D stable cells. (B) SN56 and MN9D cells with stable overexpression of either APPWT or APPSWE were treated with 1 μM or 5 μM Aβ42 oligomers for 15 hours. The caspase-3/7 activities were measured by recording the fluorescent intensities generated from cleaving a profluorescent substrate by caspase-3/7. The activities of caspase-3/7 were increased in SN56-APPWT cells treated with 5 μM Aβ42 oligomers (p< 0.05), and in SN56-APPSWE cells treated with 1 μM (p<0.001) and 5 μM (p<0.01) Aβ42 oligomers. The values in this figure represent the means ± SEM. N =3, *p< 0.05, **p< 0.01, and ***p< 0.001 by two-way ANOVA followed by post-hoc Bonferroni's multiple comparisons test. 3.3.6 Synergic effects of extracellular and intercellular αSyn species on cell death in MN9D cells With the emerging concept of Aβ oligomers, similar oligomeric species arouse the attention in studies focusing on other neurodegenerative disorders, such as PD. The soluble αSyn in the form of oligomers has been detected in cell culture system (Outeiro et al., 2008; Tetzlaff et al., 2008), in the brains of PD animal models (Tsika et al., 2010), and in the CSF and brain samples from PD patients (El-Agnaf et al., 2006; Sharon et al., 2003; Tokuda et al., 2010). The toxicity of αSyn has been extensively studied, and it is implicated in elevating mitochondrial dysfunction and oxidative stress (Hsu et al., 2000), reducing cell viability (Gosavi, Lee, Lee, Patel, & Lee, 2002), and disturbing mitochondrial and lysosomal functions (Gosavi et al., 2002; Hashimoto et 107 al., 2004). Therefore, we would like to investigate the toxicity of αSyn oligomers in our cellular system, with particular focus on the impact of the SNCAA53T mutation on cell susceptibility of cholinergic and dopaminergic cells to oligomeric treatment. We prepared αSyn oligomers according to a protocol developed by Danzer et al, as they specifically validated their effect on the cell death and caspase activation (Danzer et al., 2007). After treating the cells with 7 μM wildtype αSyn oligomers for 24 hours (Figure 3.6A), it induced significant cell death in SN56-SNCAWT (141.92 ± 8.55%, p= 0.0010), MN9D-SNCAWT (255.11 ± 0.49%, p< 0.001), and MN9D-SNCAA53T (125.35 ± 4.91%, p= 0.0472) compared with the vehicle control, but not in SN56-SNCAA53T (105.84 ± 2.04%, p> 0.05). Surprisingly, wildtype αSyn oligomers were more toxic to both cholinergic and dopaminergic cells overexpressing SNCAWT rather than SNCAA53T, with more drastic effects in dopaminergic cells. As the results were opposite to our expectations, we speculated whether there were synergic effects between extracellular oligomer species and intracellular αSyn protein with the same genotype. Therefore, we further tested how αSyn A53T oligomers would affect cell viability in our SNCA-related stable cells. As presented in Figure 3.6B, αSyn A53T oligomers only resulted in substantial cell death in dopaminergic MN9D cells with SNCAWT overexpression (197.79 ± 2.97%, p< 0.0010) and SNCAA53T overexpression (323.23 ± 11.22%, p< 0.0010), but not for SN56 cells. Further comparison between MN9D-SNCAWT and MN9D-SNCAA53T, the significance was achieved (p< 0.0010). Based on the cytotoxicity assay, it is suggested that extracellular application of wildtype and mutant αSyn oligomers cause more cell death in dopaminergic MN9D cells than cholinergic SN56 cells, and extracellular αSyn oligomers and 108 intracellular αSyn species with the same genotype have the synergic effects on reducing cell viability. Following examination of oligomers-induced cytotoxicity, the cell death pathways were further examined by aforementioned caspase-3/7 assay kit in both αSyn WT and A53T oligomeric treatments (Figure 3.6C-D). Overall, αSyn A53T oligomeric treatment lowered caspase-3/7 activities in all SNCA-associated stable cells in comparison to αSyn WT oligomeric treatment. More specifically, αSyn WT oligomers significant increased caspase-3/7 activities in SN56-SNCAWT and MN9D-SNCAWT by 42.60 ± 6.16% (p= 0.0253) and 119.38 ± 7.40% (p< 0.0010), but not in SN56-SNCAA53T (p> 0.05) nor MN9D-SNCAA53T (p> 0.05). Furthermore, the SNCAA53T mutation inhibited caspase-3/7 activities in comparison to its WT in both SN56 (p= 0.0018) and MN9D cells (p<0.0010) treated with αSyn WT oligomers (Figure 3.6C). Regarding αSyn A53T oligomeric treatment, caspase-3/7 activities were drastically increased in MN9D-SNCAWT (p< 0.0010) and decreased in MN9D-SNCAA53T (p= 0.0201), without significant changes in SN56 cells. In line with WT oligomeric treatment, the SNCAA53T mutation decreased caspase-3/7 activities in comparison to WT in both SN56 (p= 0.0017) and MN9D cells (p<0.0010) treated with αSyn A53T oligomers (Figure 3.6D). Consistent with the results from hydrogen peroxide treatment, the SNCAA53T mutation inhibited caspase-3/7 activation in both αSyn WT and A53T oligomeric treatment, indicating cell death caused by oligomers in SNCA-related stable cells probably through the caspase-3/7 independent pathway. 109 Figure 3.6 αSyn WT and A53T oligomers elevate cell death in dopaminergic MN9D cells overexpressing SNCA with the same genotype through caspase-3/7 independent pathway. (A-B) 7 μM WT A) or mutant αSyn B) oligomers were prepared from the synthetic αSyn protein in vitro and treated SNCAWT- and SNCAA53T- stable cells for 24 hours. LDH assay was applied to measure cytotoxicity. The vehicle for diluting αSyn oligomers was used as the negative control and for normalizing cytotoxicity induced by αSyn oligomers. 7 μM WT A) and mutant αSyn B) oligomers promoted cell death to the most extent in MN9D-SNCAWT and MN9D-SNCAA53T cells, respectively. (C-D) Cholinergic SN56 and dopaminergic MN9D cells with stable overexpression of either SNCAWT or SNCAA53T were treated with αSyn oligomers for 24 hours. In WT oligomer treatment C), the activities of caspase-3/7 were increased in SN56-SNCAWT cells (p< 0.05) and MN9D-SNCAWT cells (p< 0.0010) but not SN56-SNCAA53T (p> 0.05) nor MN9D-SNCAA53T(p> 0.05). In the treatment of mutant oligomer D), caspase-3/7 activation was increased more in MN9D-SNCAWT cells than SN56-SNCAWT cells, while inhibited more in MN9D-SNCAA53T than SN56-SNCAA53T. The values in this figure represent the means ± SEM. N =3, *p< 0.05, **p< 0.01, and ***p< 0.001 by two-way ANOVA followed by post-hoc Bonferroni's multiple comparisons test. 110 3.4 Discussion Seminal findings of APPSWE mutation consistently point out that it increases Aβ generation in the cellular models (Cai et al., 1993; Citron et al., 1992). In the transgenic mice, APPSWE mutation not only promotes Aβ production and plaque formation, but also causes neurodegeneration leading to impaired memory function (Calhoun et al., 1998; Hsiao et al., 1996; Lüth, Apelt, Ihunwo, Arendt, & Schliebs, 2003; Sturchler-Pierrat et al., 1997). Our results showed that APPSWE mutation enhanced both Aβ40 and Aβ42 generation in SN56 and MN9D cells, with more extent in SN56 cells. This is consistent with the initial findings that APPSWE mutation increases both Aβ40 and Aβ42 levels (Cai et al., 1993; Citron et al., 1992), while PS1/PS2 mutations increase Aβ42 levels without affecting Aβ40 or even decreasing Aβ40 (Bentahir et al., 2006; Kumar-Singh et al., 2006; Scheuner et al., 1996). Compared with CHO cells, APPSWE overexpression in our cellular system released higher levels of Aβ species (Citron et al., 1992), which may be due to higher expression of BACE1 protein in SN56 and MN9D cells with neuronal properties (Nilbratt, Friberg, Mousavi, Marutle, & Nordberg, 2007). Previous studies have demonstrated that SNCAA53T makes αSyn more prone to form aggregates and promote fibrillogenesis by examining the kinetics and morphologies of αSyn aggregation in vitro using atomic force microscopy, circular dichroism spectroscopy, and electron microscope (Conway et al., 1998; Giasson et al., 1999; Narhi et al., 1999). Several studies point out that overexpression of SNCAWT or SNCAA53T in neuroblastomas alone cannot induce cytoplasmic aggregate formation. The aggregates became detectable after treating the cells with either ferric chloride or mitochondrial inhibitor, such as rotenone (Hasegawa et al., 2004; Lee, Shin, Choi, Lee, & Lee, 2002; Lee & Lee, 2002; Ostrerova-Golts et al., 2000). Under such circumstance, 111 SNCAA53T facilitates aggregates formation in terms of both number and size (Ostrerova-Golts et al., 2000). In our cellular model, around 10% of stable cells overexpressing SNCAWT or SNCAA53T were positive for αSyn aggregates, and larger aggregates were observed in MN9D-SNCAA53T stable cells (Figure 3.2). Unfortunately, we did not find noticeable differences in the number of aggregates among these SNCA-related stable cell lines. We also tried to treat our stable cells with FeCl2, and there was no change for αSyn aggregates examined by immunofluorescent staining (data not shown). Another interesting finding in our study was that αSyn-positive aggregates were positive for Hsp70 staining in MN9D stable cells carrying SNCAWT or SNCAA53T, but negative for SN56 cells (Figure 3.2). It was mentioned by previous studies that ~70% of the LBs found in brain samples of DLB patients were positive for Hsp70 ( McLean et al., 2002; Shin, Klucken, Patterson, Hyman, & McLean, 2005). A recent study points out that different treatments are able to induce two types of inclusion in the cell culture system, including the inclusions with close proximity to the nucleus (“Juxta-Nuclear Quality control compartment (JUNQ) αSyn inclusions”) and the remaining inclusions with the relatively far distance from the nucleus (“insoluble protein deposit (IPOD) αSyn inclusions”). The treatment with the higher cytotoxicity was more related with JUNQ αSyn inclusions and immunoreactive for Hsp70, while the IPOD αSyn inclusions were found to be negative for Hsp70 staining and more associated with less toxic induction methods (Raiss et al., 2016). Therefore, the different properties of the αSyn aggregates among SNCA-related stable cell lines indicate Hsp70 positive aggregates in SNCA-overexpressing MN9D cells may represent more toxic species. 112 After investigating the pathogenic effects of APPSWE and SNCAA53T mutations in cholinergic and dopaminergic cells, we speculated whether these two mutations would change the cell vulnerability to two types of insults, hydrogen peroxide and oligomers. Hydrogen peroxide is considered as an inducer of oxidative stress, and oxidative stress is implicated in the pathogenesis of both AD and PD (Praticò, 2008; Zhang et al., 2000). In line with previous findings (Eckert, Steiner, Marques, Leutz, Romig, Haass, & Müller, 2001), APPSWE mutation rendered deleterious effects on cell viability in SN56 cells in H2O2 treatment through caspase-3 dependent pathway, but cell death was not significant in MN9D cells overexpressing APPSWE (Figure 3.3), indicating cell type- specific effects. With respect to SNCA-related stable cells, SNCA53T led to more severe cell death in MN9D cells than SNCAWT, consistent with the results in other studies (Jiang et al., 2007). However, the finding that SNCAA53T resulted in less cell death in SN56 in comparison to SNCAWT was unexpected. Another intriguing fact is that SNCAA53T inhibited caspase-3 activation in both SN56 and MN9D cells in H2O2 and αSyn oligomer treatments (Figure 3.4 and Figure 3.6), suggesting non-apoptotic mechanisms are involved. It is consistent with the findings that overexpression of SNCAA53T in PC12 cells induces nonapoptotic cell death (Stefanis, Larsen, Rideout, Sulzer, & Greene, 2001) and SH-SY5Y cells carrying SNCAA53T mutation did not enhance caspase-3 activation after applying αSyn oligomers (Danzer et al., 2007). With the recent advance in understanding the etiology of AD and PD, a more well-accepted notion is that Aβ and αSyn oligomers other than fibrils are the responsible species for neurodegeneration (Benilova et al., 2012; Kazantsev & Kolchinsky, 2008; Selkoe & Hardy, 2016). Therefore, we thought it would be interesting to test how our established stable cell lines 113 responded to oligomeric treatment. The results revealed that Aβ oligomers increased cell death in SN56-APPSWE cells compared with SN56-APPWT cells, but failed to change cell viability in both MN9D-APPWT and MN9D-APPSWE cells (Figure 3.5). In one proteomics study, SN56 cells treated with Aβ42 oligomers result in the expression of proteins associated with AD (Joerchel et al., 2008). Therefore, it is not surprising that Aβ42 oligomers induced more severe cell death in SN56-APPSWE cells. Additionally, more activation of caspase-3/7 was observed in SN56-APPSWE cells than SN56-APPWT cells treated with Aβ42 oligomers, which is consistent with the findings that Aβ42 oligomers increase cell death through caspase-dependent mechanisms (Youssef et al., 2008). In our study, extracellular WT/ mutant αSyn oligomeric species had more toxic effects in the dopaminergic cells overexpressing the same genotype of αSyn variant (Figure 3.6). As reported in a study, labeled extracellular αSyn oligomers were detected in the cytoplasm of SH-SY5Y cells stably transfected with SNCAA53T or empty vector, suggesting extracellular αSyn species have seeding effects (Danzer et al., 2007). The follow-up study done by the same group further points out that αSyn oligomers prepared in vitro causes intracellular seeding in a dosage- and time- dependent manner in both neuronal cells and primary neurons (Danzer et al., 2009). Similarly, the extracellular αSyn oligomer prepared in our study may have seeding effects apart from cytotoxicity, and these extracellular seeds may be easier to be associated the intracellular αSyn species with the same genotype, leading to more cell death. If these speculations are true, it is reasonable that αSyn WT oligomers induced more cell death in MN9D-SNCAWT stable cells, while αSyn A53T oligomers resulted in more cellular toxicity in MN9D-SNCAA53T stable cells. 114 3.5 Conclusion In summary, APPSWE mutation promoted Aβ40 and Aβ42 generation in cholinergic SN56 cells, and SNCAA53T mutation induced larger cytoplasmic LBs-like inclusions in dopaminergic MN9D cells. In the hydrogen peroxide treatment, both APPSWE and SNCAA53T mutations specifically decreased cell viability of cholinergic cells and dopaminergic cells, respectively. Moreover, APPSWE mutation elevated vulnerability of cholinergic cell to the Aβ42 oligomeric treatment, but not in dopaminergic cells. Finally, the αSyn WT oligomer caused the most cell death in dopaminergic cells overexpressing SNCAWT, while the αSyn A53T oligomer resulted in the most severe cell loss in dopaminergic cells overexpressing SNCAA53T. Overall, APPSWE and SNCAA53T mutations make cholinergic and dopaminergic cells more susceptible to the cytotoxic insults closely related to the pathogenesis of AD and PD, respectively. 115 Chapter 4: The effects of SDC3 and FGFRL1 on neurodegeneration in AD and PD 4.1 Introduction In the last chapter, we discussed that the APPSWE and SNCAA53T mutations made cholinergic and dopaminergic cells more susceptible to various cytotoxic insults, leading to cell death. Following these, we planned to further explore possible targets for mediating the effects of these two mutations on neurodegeneration. As mentioned before, although enormous efforts have been put to elucidate the molecular mechanisms underlying neurodegeneration for AD and PD, the definite answer to this question is still vague. The current treatments for AD and PD only have symptomatic relief, such as the application of anticholinesterase drugs in AD as well as levodopa and other drugs to boost dopaminergic transmission in PD (Connolly & Lang, 2014; Graham, Bonito-Oliva, & Sakmar, 2017). Although the ongoing clinical trials for AD and PD are designed to search for new disease-modifying therapeutics, none of them have offered promising clinical outcomes yet (Graham et al., 2017; Kalia, Kalia, & Lang, 2015). Since the discovery of DNA microarray in 1995 (Schena et al., 1995), it offers another option to study large amounts of genes simultaneously, obtain a ‘snapshot’ of the gene expression pattern in the biological samples, and gain undiscovered knowledge in a systematic view (Lovén et al., 2012). In the field of AD and PD, previous DNA microarray studies have enhanced our understanding of the etiology of these two diseases, which can be beneficial to search for new treatments. 116 By using DNA microarray, studies comparing the gene expression profiling of AD patients and the control subjects help to elucidates the altered pathways associated with the pathogenesis and disease progression. It has been suggested that the alteration of gene expression in temporal and prefrontal cortex is closely related to disease severity (Haroutunian, Katsel, & Schmeidler, 2009). Previous studies consistently unveil the two dysregulated pathways in AD, including neuroinflammation (Parachikova et al., 2007; Tan et al., 2010; Youn et al., 2007) and calcium signaling (Emilsson, Saetre, & Jazin, 2006; Katsel, Tan, & Haroutunian, 2009; Tan et al., 2010; Youn et al., 2007). With the advance in the methodology of dissection, laser capture microdissection (LCM) is able to isolate single cells from tissue sections, which substantially improves the purity of acquired samples. This is of pivotal importance to discover neuronal-specific effects, which could be masked in prepared samples mixed with neuronal and non-neuronal cells (Cooper-Knock et al., 2012; Ginsberg, Alldred, & Che, 2012). The application of LCM-assisted DNA microarray in AD allows the investigators to examine the gene expression profiling in cholinergic neurons and hippocampal pyramidal neurons in particular. The cholinergic neurons in the NBM of AD brain samples downregulate genes encoding the neurotrophic receptors and disturb genes involved in calcium signaling (Counts, He, Che, Ginsberg, & Mufson, 2009; Mufson, Counts, & Ginsberg, 2002), while pyramidal neurons in the hippocampal CA1 region differentially express genes implicated in endosomal function, neurotrophic receptor, and synaptic transmission (Ginsberg et al., 2010;Ginsberg et al., 2012). Regarding the microarray in PD brain tissues, most studies focused on SN and striatum – particularly putamen. Consistent findings for altered gene expression in the PD patients are related to disturbed protein quality control and mitochondrial dysfunction (Bossers et al., 2009; 117 Hauser et al., 2005; Miller et al., 2006; Moran et al., 2006; Yanli Zhang, James, Middleton, & Davis, 2005). With the help of LCM, precise dissection for single cells was applied to isolate dopaminergic neurons specifically. A study completed by Simunovic’s group employed dopaminergic neurons in the SNpc by LCM to examine the gene expression profiling in the PD patients and controls after matching the age and postmortem interval (Simunovic et al., 2009).The discovered differentially expressed genes (DEGs) are involved in programmed cell death and survival pathways, mitochondrial dysfunction, protein degradation, and neurotransmission (Simunovic et al., 2009), which have been confirmed by the findings from Elstner’s group (Elstner et al., 2011). Notably, a landmark study done by Zheng’s group performed a meta-analysis for 17 studies. After integrating data from 185 laser-captured dopaminergic neuron and substantia nigra transcriptomes with two stages of replication, ten gene sets were downregulated and associated with PD, namely mitochondrial electron transport, glucose utilization, and glucose sensing (Zheng et al., 2010). Although the microarray studies focusing on AD and PD patients flourished, the studies to uncover the effects of APPSWE and SNCAA53T mutations on gene expression are relatively limited. In this Chapter, we performed a whole-genome expression profiling in the APP-related stable cells, SNCA-related stable cells, and their parental SN56 and MN9D cells by DNA microarray. The analyses of DEGs were performed to screen interesting genes, the expression of which was affected by APPSWE or SNCAA53T mutation in a cell type- dependent manner. The GO enrichment analysis and pathways analysis were further conducted to better understand the biological interpretation of these DEGs. Among the gene list, SDC3 and FGFRL1 were selected to examine their expression at both mRNA and protein levels in vitro and in vivo. Finally, 118 knockdown experiments were completed in APP-related and SNCA-related stable cells to further test their roles in cell death induced by oxidative stress. 4.2 Methods 4.2.1 Cell culture, transfection, and knockdown of SDC3 and FGFRL1 Cholinergic SN56 and dopaminergic MN9D cells were cultured in DMEM supplemented with 10% FBS, 1 mM of SP, 2 mM of L-glutamine, 50 units of penicillin and 50 μg of streptomycin (Invitrogen). The culture dishes (Corning) for maintaining MN9D cells were coated with 10 μg/mL PDL (Sigma). Both SN56 and MN9D cells were maintained at 37°C in an incubator containing 5% CO2 and cell medium was changed when necessary. Stable cells were maintained in medium containing 50 μg/mL zeocin. Before transfection, cells were seeded on the plate at 50% - 70% confluency and all transfection experiments were performed by lipofectamine 2000 (Invitrogen) following manufacturer’s instruction. For knockdown experiments, SDC3 siRNA or negative control siRNA (Santa Cruz Biotechnology) was transfected into SN56-APPSWE and MN9D-APPSWE stable cells by Lipofectamine 2000 (Invitrogen), and ON-TARGETplus mouse FGFRL1 siRNA or control siRNA (Dharmacon) was transfected into SN56-SNCAA53T and MN9D-SNCAA53T stable cells by Lipofectamine 2000 as well. 4.2.2 Animal and genotyping Animal experiment protocols were approved by The University of British Columbia Animal Care and Use committee. Prnp-SNCAA53T transgenic mice were generated previously by Lee’s group (Lee et al., 2002) and obtained from Jackson Laboratory. Prnp-SNCAA53T hemizygous transgenic mice expressed human αSyn protein carrying A53T mutation at ~6 folds of endogenous mouse αSyn 119 driven by murine Prion promoter. All mice were genotyped at the beginning of weaning around 3 weeks of age, and confirmed by PCR from gDNA extracted from ear tissue. The tissue was digested in 300μL of lysis buffer (10 mM Tri-HCl pH8.0, 10 mM EDTA pH8.0, 150 mM NaCl, 0.5% SDS) with 3μL of 10 μg/mL Proteinase K (New England Biolabs) at 55°C overnight. The next day, samples were centrifuged and gDNA was precipitated with isopropanol. Genomic DNA was centrifuged at 16,000 x g for 15 min, washed with 70% ethanol twice, air dried, and resuspended in sterile deionized water. The gDNA was then subjected to perform PCR. The sequence of primers for conducting PCR was based on the genotyping protocol provided by Jackson Laboratory, forward: 5’- tcatgaaaggactttcaaaggc, and reverse: 5’- cctcccccagcctagacc. APPNL/NL was a recently developed APPSWE knock-in mouse line. The murine Aβ sequence was humanized and engineered to harbor the Swedish mutation. This knock-in mouse expresses the equal level of full-length APP as wildtype mice (Saito et al., 2014). The gDNA for genotyping was extracted from ear tissue as mentioned before. APPNL/NL mice were obtained from Riken BioResource Center. The sequence of primers for conducting PCR was forward: 5’-gtttgctaggtggtggttaatgg and reverse: 5’- ctttgtgaagatctaggcaggc, which amplified a 416 bp fragment located on exon 16. 4.2.3 Whole-genome expression profiling Total RNA was isolated from ten cell lines, 4 APP-related stable cell lines, 4 SNCA-related stable cell lines, SN56 and MN9D cell lines, by using TRI-Reagent (Sigma-Aldrich). Total RNA was reverse transcribed to synthesize the first strand cDNA by using a T7 Oligo (dT) Primer, which was used as the template to amplify the double-stranded DNA (dsDNA). Double-strand cDNA was processed into biotinylated complementary RNA (cRNA) by performing In vitro 120 transcription (IVT) with help of Ambion® Illumina TotalPrep RNA Amplification Kit (Thermo Fisher Scientific). 1.5 μg cRNA of each sample was used for whole-genome gene expression direct hybridization assay with Mouse MouseWG-6 v2.0 Expression BeadChip (Illumina) following the manufacturer’s instructions. Each bead in the array contains a 50-mer, sequence-specific oligo probe and each array covers 45,281 probe sets. 4.2.4 Analysis of DEGs After completing whole-genome gene expression direct hybridization assay, 20 raw datasets were background-corrected and exported from the Illumina Software ‘Beadstudio’ to R (http://www.cran.r-project.org). After log-scale transformation, quantile normalization and removal of outliners, 33,839 transcripts were detected in all samples. For screening DEGs in APP-related and SNCA-related stable cells, datasets were analyzed in Linear Models for Microarray Data (Limma) by 2 x 2 factorial design and genes satisfying false discovery rate (FDR) < 0.001 and fold change (FC) >2 were selected as DEGs for Go enrichment analysis and IPA analysis. 4.2.5 Gene Ontology (GO) enrichment analysis and Ingenuity pathway analysis (IPA) analysis To explore the GO terms enriched in our DEGs, DEGs from analyses of parental cell lines, APP-related stable cell lines and SNCA-related stable cell lines, were imported into an online bioinformatic tool, DAVID (https://david-d.ncifcrf.gov/home.jsp), by selecting ‘Functional Annotation tool’ and ‘Gene Ontology’ database following a protocol described previously (Huang da, Sherman, & Lempicki, 2009). Cellular functions and canonical pathways were 121 generated by Ingenuity Pathway Analysis (Qiagen Bioinformatics). Canonical pathway analysis determines pathways from the Ingenuity Knowledge Base that are most significant to the dataset. The significance of the association between the DEGs and the cellular functions or canonical pathway was calculated by right-tailed Fisher’s exact test, which determines the probability that the association between the genes in the data set and the canonical pathway is due to chance alone. 4.2.6 Quantitative reverse transcription PCR (qRT-PCR) Total RNA was extracted from stable cells and transcribed into cDNA by ThermoScriptTM RT-PCR system (Invitrogen) as mentioned in Section 2.2.7. The quantitative PCR was performed using SYBR® Green PCR Master Mix (Life technologies) on QuantStudio 6 Flex Real-Time PCR system (Thermo Fisher Scientific) in 96-well format according to the user’s guide. The primers, spanning an exon–exon boundary, were designed by a software, DNASTAR® PrimerSelect, and listed alphabetically in Table 4.1. Melting curves were generated at the end of the run to ensure the detection of desired amplicon. Standard curves of each amplified gene were created to monitor the PCR efficiency. The gene expression value of DEGs necessary for validation was compared with the reference genes (GAPDH or ACTB) by the comparative threshold cycle method detailed in a previous protocol (Schmittgen & Livak, 2008). 122 Table 4.1 Oligonucleotides for the qRT-PCR Transcript Forward sequence Reverse sequence ACTB ATGCTCCCCGGGCTGTATTC CTTGCTCTGGGCCTCGTCAC ADCY4 CAGAACCGGCGCAATGAGGA GGGACCGACGCGAAGAGGAC CXCL1 CATGGCTGGGATTCACCTCA AGCCTCGCGACCATTCTTG CXCL2 TCCAGAGCTTGACGGTGACG GGGGCCTTGGGGGTTGAG EGFR ATGGGCCCTGTCGCAAAGTT TGTGAAGGTCCCCGCTGATG FGFRL1 GGGCCCGCCAGGATGATG TGGCAGGCTTGTGGATGACG FN1 ATGCCCCGGAGCCTTCACA CCTTCCAGCGACCCGTAGAG GAPDH GGCCGGTGCTGAGTATGTCG GCAGAAGGGGCGGAGATGAT LTA CAGCCCATCCACTCCCTCAG CGATCCGTGCTTGCTCTCCA MRAS TGGTGGGAGATGGTGGTGTG CCTGCCCGGCTGTGTCC MYD88 AGGAACTGGGAGGCATCACC GCGGCGACACCTTTTCTCAA NTRK3 GCACGGCCAACCAGACCAT GGGCTCGCATCAGACTCAAA PDGFRα AGCGTGGGGCCTTACATCTG GGCGTGCGTCCATCTCCA PPP1R3C CCTGGGTCCTTACAATGGTTTTCA CCGCTTCTTGGCTTGGTTGTGT PPP2R1A GGTGGGAGGGCCTGAGTATG GATGGCCCGCAAGGATTCTA PTS ACTGTCGCGCCTCGTGTC CGTGGCCATTCGGATTGTT SDC3 GCCGCCCAGCTCCCTCAG CCACCACCCCACCCACGAT TLR2 CAGGGATCCGGGTGGTAAAA GCAGCCGAGGCAAGAACAAA 4.2.7 Dissection of mouse brain and preparation of mouse brain tissue homogenate Four-month-old APPNL/NL mice and age-matched wildtype C57 mice (n=3) were sacrificed and fresh brain was dissected. After washing it with ice-old PBS, each brain was placed in the pre-cooled Zivic Mouse Brain Slicers (Zivic Instrument) on dry-ice. Coronal brain slices were 123 sequentially dissected with the interval of 1mm. Brain slice containing MS, NBM, and SN brain areas were identified with the help of Mouse Brain Atlas (Franklin, 2008), and each region was further dissected as shown in Figure 4.7B. The collected brain tissues were put in the 1.5ml centrifuge tube and lysed with RIPA buffer (1% triton X-100, 0.1% SDS, 1% sodium deoxycholate, 150 mM sodium chloride, 50 mM Tris-HCl (pH 7.2) and protease inhibitor cocktail (Roche)). After sonication, brain tissue homogenates were centrifuged at 13,500 rpm for 30 mins at 4ºC, and supernatant was collected for immunoblotting. 4.2.8 Immunofluorescent staining The immunofluorescent staining was performed as previously described with modification (Ly, Cai, & Song, 2011). Briefly, Prnp-SNCAA53T (n=4) and C57 mice (n=4) were sacrificed and whole brains were fixed in 4% PFA overnight at 4ºC, followed by sequential dehydration in 15% sucrose and 30% sucrose solution. Brains were embedded in O.C.T solution and sectioned with a Leica Cryostat to 30 μm thickness slices. Brain sections were collected in 24-well plate in Tris-buffered saline (TBS) containing 50 mM Tris-Cl (pH 7.5) and 150 mM NaCl. Antigen retrieval was conducted by incubating the slices in sodium citric acid buffer (10 mM Sodium citrate, 0.05% Tween 20, pH 6.0) at 85ºC for 5mins. After permeabilizing the slices in 0.3% Triton-X100 in TBS (TBS-Tx) for 30 mins at RT, brain slice was further blocked in blocking reagents (M.O.M blocking reagent, Vector Laboratories) at room temperature for 1 hour. Goat Anti-ChAT antibody (1:100, Millipore), mouse anti-TH antibody (1:20, Thermo Fisher Scientific), and rabbit anti-FGFRL1 antibody (1:100, Abcam) in TBS-Tx were incubated for 24 hours or 48 hours. Fluorochrome-conjugated secondary antibodies: Alexa 488 labeled donkey anti-mouse IgG (Invitrogen), Alexa 568 labeled donkey anti-rabbit IgG (Invitrogen), and Cy5 labeled donkey anti-124 goat IgG (Invitrogen) (1:500 in TBS), were incubated for 1hr at RT. All images were acquired by using Leica TCS SP8 X confocal microscope (Leica Microsystems). The average fluorescent intensities were quantified by Image J software. The boundaries of TH or ChAT fluorescent signals were used as the reference to calculate the size of area in every neuron. The average signal intensity of FGFRL1 in each neuron was determined as dividing of integrated signal intensities by area size with subtraction of mean background intensity. 4.2.9 LDH assay Cells were seeded on 96-well plate one day before treatment. On the next day, cells were treated with varying concentration of H2O2 for 12 hours, and cell medium was collected to perform LDH assay (Promega). LDH was released into cell medium when the integrity of the cell membrane was compromised in nonviable cells. LDH was determined by a fluorescently coupled enzymatic assay, which was measured with an emission wavelength of 590nm. 4.2.10 RT-PCR Total RNA was extracted from SN56-SNCAA53T and MN9D-SNCAA53T stable cells in FGFRL1 knockdown experiment by TRI reagent (Sigma), and reverse transcription was detailed in section 2.2.7 using ThermoScriptTM RT-PCR system (Invitrogen). The newly synthesized cDNA was used as the template to further amplified mouse endogenous FGFRL1 gene by Taq DNA polymerase in PCR. The sequence of primer pairs was as follows: FGFRL1-F 5’-gggcccgccaggatgatg and FRGFL1-R 5’- tggcaggcttgtggatgacg. 125 4.2.11 Immunoblotting Cells were lysed in RIPA lysis buffer containing 1% triton X-100, 0.1% SDS, 1% sodium deoxycholate, 150 mM sodium chloride, 50 mM Tris-HCl (pH 7.2) and protease inhibitor cocktail (Roche), followed by brief sonication. Preparation of mouse brain tissue homogenate was described in Section 4.2.7. Procedures for performing immunoblotting were detailed in Section 3.2.3. For detecting full-length APP, SDC3 and FGFRL1, samples were resolved on 8% Tris–glycine SDS–PAGE, 16% tris-tricine SDS-PAGE was used for detecting αSyn-mycHis and β-actin. Primary antibodies included rabbit C20 antibody (1:1000), mouse monoclonal anti-myc antibody 9E10 (1:100, Santa Cruz Biotechnology), rabbit polyclonal anti-FGFRL1 antibody (1:1000, Origene), rabbit monoclonal anti-SDC3 antibody (1:1000, Abcam), and mouse monoclonal anti-β-actin AC-15 antibody (1:5000, Sigma). IRDye 680RD-labelled goat anti-rabbit antibodies (1:2000, LI-COR) and IRDye 800CW-labelled goat anti-mouse antibodies (1:2000, LI-COR) were applied as secondary antibodies. 4.2.12 Statistical analysis All results were presented as means ± SEM. For two-group comparison, the results were analyzed by 2-tailed Student’s t-test. For multiple-group comparison, the data were analyzed by one-way ANOVA or two-way ANOVA with post-hoc Bonferroni's multiple comparisons test. Statistical significance was accepted when p<0.05 (*p<0.05, **p<0.01, ***p<0.001). 126 4.3 Results 4.3.1 Data processing for DNA microarray and comparison of gene expression in SN56 and MN9D cells To examine how APPSWE and SNCAA53T affect the gene expression in cholinergic and dopaminergic cells at a systematic level, we took the advantage of our established stable cell lines, cholinergic SN56 and dopaminergic MN9D cells overexpressing human APPWT/SWE and SNCAWT/A53T. SN56 and MN9D were accepted cellular models in the microarray studies of AD and PD (Heinitz et al., 2006; Joerchel et al., 2008; Park et al., 2011; Wang et al., 2008). The whole-genome expression profiling of the eight stable cell lines, SN56-APPWT/SWE, MN9D-APPWT/SWE, SN56-SNCAWT/A53T, and MN9D-SNCAWT/A53T, and the two parental cell lines were analyzed by DNA microarray using the Illumina® MouseWG-6 v2.0 expression beadchip, which covers 45,281 probe sets corresponding to 33,843 transcripts. There were 20 raw datasets obtained from DNA microarray. The distribution of the background-corrected and log2 transformed datasets was examined by boxplots as shown in Figure 4.1A. The four outliers were identified (Figure 4.1A), and removal of the outliers was followed by quantile normalization to make the distribution identical cross all datasets (Figure 4.1B). A sample-to-sample correlation heatmap was generated after these steps of data processing (Figure 4.1C). The heatmap manifests that the best correlation existed between the biological replicates, indicating the reproducibility of these replicates. And gene expression pattern of the stable cells was correlated better with the stable cells generated from the same parental cell lines than generated from different parental cells with the same mutation. It suggests that overexpression of APPWT/SWE and SNCAWT/A53T in SN56 and MN9D cells only changed a small subset of genes without affecting their overall gene expression, making the cell type- specific effect noticeable. 127 The 20 datasets were divided into 3 parts for analyzing differentially expressed genes (DEGs), including 4 datasets of SN56 and MN9D, 8 SNCA-related datasets, and 8 APP-related datasets, and the detailed analyses were presented in the Methods of this Chapter. Compared with SN56 cells, there were 212 DEGs (116 upregulated genes and 96 downregulated) in MN9D cells, accounting for 0.63% of all tested genes when the screening criteria were set to FDR < 0.001 and FC > 2. There were 200 out of 212 the DEGs mapped in the DAVID, and the GO enrichment analyses revealed that the most overrepresented GO terms, including cytoplasmic vesicles, amine transmembrane transporter activity, catecholamine biosynthetic process, amino acid transport, symporter activity, and cell motion (Table 4.2). It suggests that both SN56 and MN9D manifest neuronal properties, and the major genetic difference between these two cell lines lies in the characteristics related with neurotransmission. 128 Table 4.2 Go analysis for DEGs in SN56 and MN9D cells Terms Enrichment Score Count % Gene Cytoplasmic vesicle 3.13 15 8.62 Slc32a1, Gpr120, Th, Egfr, Phlda1, Agtr1a, Slc17a6, Tes, Igf2r, Chic1, Serpinf1, Sh3kbp1, Dbh, Tmsb4x, 1500015O10Rik Amine transmembrane transporter activity 2.45 5 2.87 Slc32a1; Slc7a3; Slc6a2; Slc17a6; Slc1a3; Catecholamine biosynthetic process 2.38 4 2.30 Th, Dbh, Ddc, Snca Amino acid transport 2.17 5 2.87 Slc32a1, Slc38a11, Slc7a3, Slc6a7, Slc1a3 Symporter activity 1.74 5 2.87 Slc32a1, Slc6a2, Slc17a6, Slc1a3, Slc6a7 Cell motion 1.66 9 5.17 Isl1, Gbx2, Ank3, Egfr, Pf4, Dbh, Drd1, Gli3, Tes 129 Figure 4.1 Data processing and exploratory analysis for 20 datasets. (A) Boxplots of signal intensity (log2) distribution of 20 arrays before normalization, including two parental cell lines (SN56, MN9D), four APP-related stable cell lines (SN56-APPWT, SN56-APPSWE, MN9D-APPWT, MN9D-APPSWE), and four SNCA-related stable cell lines (SN56-SNCAWT, SN56-SNCAA53T, MN9D-SNCAWT, MN9D-SNCAA53T) with replication. (B) Boxplots of signal intensity distribution of all arrays after quantile normalization and removal of the outliers. (C) Heatmap of pairwise Pearson correlations of entire sample expression profiles. The replicates showed the best correlation, and the gene expression of stable cells generated from the same parental cells was correlated better with each other. 130 4.3.2 APPSWE and SNCAA53T mutations have differential effects on gene expression in cholinergic and dopaminergic neuronal cells Following the comparison between SN56 and MN9D cells, the impacts of APPSWE on altering gene expression in these two neuronal cell lines were further investigated in 4 APP-related stable cell lines. By fitting the data into linear models for microarray data (Limma), there were 537 genes manifesting interaction effects between cell lines (SN56/ MN9D) and genotypes (APPWT/ APPSWE). Additionally, there were 543 DEGs (296 upregulated and 247 downregulated) in the SN56-APPSWE compared with SN56-APPWT (FDR < 0.001 and FC >2), while there were 518 DEGs (150 upregulated and 368 downregulated) in MN9D-APPSWE compared with MN9D-APPSWE (FDR < 0.001 and FC >2). We focused on the genes having both interaction effects and differentially expressed in the comparison between SN56-APPSWE and SN56-APPWT, as presented in Figure 4.2A. The overlapping genes were 215, with 133 upregulated and 82 downregulated (FDR < 0.001 and FC > 2). The list containing 215 genes was further explored by both DAVID and Ingenuity Pathway Analysis (IPA) to gain more biological information. There were 210 out of 215 genes mapped to DAVID identifier, and the genes affected by APPSWE specific to cholinergic SN56 cells were clustered in the glycosaminoglycans (GAGs) binding and regulation of neuron differentiation and neurogenesis with statistical significance (Table 4.3). In order to gain more insight about the DEGs discovered from APP-related arrays, IPA was performed to find out what cellular functions and canonical pathways were enriched in our gene lists. As represented in Figure 4.3A, DEGs were involved in lipid metabolism, molecular transport, cell growth and proliferation, small molecule transport, and cell death and survival, to name a few (Figure 4.3A). Additionally, DEGs screened from APP-related arrays were implicated in several pathways, and the most significant five pathways includes granulocyte 131 adhesion and diapedesis, agranulocyte adhesion and diapedesis, β-alanine degradation I, LXR/ RXR activation, and PCP pathway (Figure 4.3C). By using the same strategy for studying APPSWE’s role in the gene expression, the microarray data from four SNCA-related stable cell lines were analyzed by Limma, and DEGs were screened based on FDR< 0.001 and FC >2. There were 2212 genes showing interaction effects between cell lines (SN56/ MN9D) and genotypes (SNCAWT/ SNCAA53T). There were 941 DEGs (600 upregulated and 341 downregulated) in the MN9D-SNCAA53T compared with MN9D-SNCAWT cells, while 2055 genes (848upregulated and 1207 downregulated) were found differentially expressed in SN56-SNCAA53T in comparison to SN56-SNCAWT (FDR< 0.001 and FC> 2). We focused on the 586 overlapping genes (407 upregulated and 179 downregulated) having interaction effect and existing in the gene list from the comparison between MN9D-SNCAWT and MN9D-SNCAA53T (Figure 4.2B). By comparing the DEGs discovered in APP-related stable cells (215 DEGs) and in SNCA-related stable cells (586 DEGs), 25 genes were overlapped as presented in Figure 4.2C, further suggesting that these DEGs are highly specific to the effects of APPSWE mutation or SNCAA53T mutation. Following these, the 586 DEGs in SNCA-related stable cells were performed GO enrichment analysis and significantly enriched in the following GO terms, including regulation of synaptic transmission, 'de novo' protein folding, protein-DNA complex assembly, and glycosaminoglycan binding (Table 4.4). IPA was further performed to explore potential cellular functions could be affected by this differentially expressed gene set. The top five cellular functions enriched in this gene set were cellular development, cellular movement, cellular growth and proliferation, cellular function and maintenance, and cell morphology (Figure 4.3B). Moreover, we analyzed the 586 genes for 132 significant overrepresentation in the canonical pathways using IPA. The top pathways were crosstalk between dendritic cells and natural killer cells, NFκB signaling, cleavage and polyadenylation of pre-mRNA, antigen presentation pathway, and T helper cell differentiation (Figure 4.3D), suggesting overexpression of SNCAA53T has an essential role in differentially regulating inflammation in cholinergic and dopaminergic neuronal cells. Figure 4.2 Analysis of DEGs in APP-related and SNCA-related stable cells. (A) Venn diagram to show overlapping and nonoverlapping genes after analyzing DEGs in SN56-APPWT vs. SN56-APPSWE (yellow circle) and genes having interaction effects between cell types and overexpressed genes (blue circle). (B) Venn diagram to represent overlapping and nonoverlapping genes after analyzing DEGs in MN9D-SNCAWT vs. MN9D-SNCAA53T (orange circle) and genes having interaction effects between cell types and overexpressed genes (grey circle). (C) Venn diagram to compare the gene lists obtained from analyzing APP-related arrays and SNCA-related arrays. 133 Table 4.3 GO analysis for DEGs in APP-related stable cells Table 4.4 GO analysis for DEGs in SNCA-related stable cells Terms Enrichment Score Count % Genes Regulation of synaptic transmission 2.46 5 1.05 Snca, Gria2, Ltbr, Tnf, Lta 'de novo' protein folding 2.43 4 0.84 Cd74, Hsph1, H2-DMa, Hspa8 Protein-DNA complex assembly 1.75 7 1.47 Hist1h2bl, Smarca2, Hist2h2be, Hist1h2bk, Hist1h2bh, Hist1h2bg, Hist1h1c Glycosaminoglycan binding 1.64 9 1.89 Hapln4, Cd44, Lamc2, Fn1, Lpl, Rpl29, Serpine2, Ptn Antigen processing and presentation of exogenous peptide antigen 1.59 5 1.05 Cd74, H2-M3, Tap1, H2-DMa, H2-Ab1 Regulation of cell death 1.34 18 3.78 Clu, Cx3cl1, Tbx3, Qars, Snca, Lhx4, Cd74, Msx1, Nme5, Tnf, Plcg2, Ltb, Sox9, Glo1, Bdnf, Gclm, Bcl6, Ltb Terms Enrichment Score Counts % Genes Glycosaminoglycan binding 2.98 7 3.74 Chrd, Col5a1, Fn1, Lpl, PF4, Bgn, Fstl1 Regulation of neuron differentiation and neurogenesis 1.69 5 2.67 Sema3f, Nefm, 3110039M20Rik, Phox2b, Six1 134 Figure 4.3 Cellular functions and top canonical pathways overrepresented in the DEGs for analyzing APP- and SNCA-related arrays by IPA. (A-B) The DEGs identified in APP-related arrays and SNCA-related arrays were performed IPA to discover cellular functions significantly overrepresented in DEGs. Bar graph represents the significant cellular functions were differentially affected by APPSWE mutation A) and SNCAA53T mutation B) in cholinergic SN56 and dopaminergic MN9D cells using IPA software. (C-D) Bar graph represents the top canonical pathways were differentially affected by APPSWE mutation C) and SNCAA53T mutation D) in cholinergic SN56 and dopaminergic MN9D cells using IPA software. The p value was calculated by right-tailed Fisher’s exact test. 4.3.3 Validation of selected genes differentially affected by APPSWE and SNCAA53T mutations in cholinergic and dopaminergic neuronal cells by qRT-PCR To confirm the altered gene expression in APP-related and SNCA-related cells by microarray, qRT-PCR was performed by using three sets of independent samples. Based on IPA pathway analysis, there were multiple overlapping genes in the first and third canonical pathways from the results of the APP-related array, both of which are involved in inflammation. Therefore, we focused on the DEGs in the top 3 canonical pathways. According to the results from microarray, the expression levels of adenylate cyclase 4 (ADCY4), C-X-C motif chemokine ligand 1 (CXCL1), C-X-C motif chemokine ligand 2 (CXCL2), fibronectin 1 (FN1), protein phosphatase 135 1 regulatory subunit 3C (PPP1R3C), and 6-pyruvoyl-tetrahydropterin synthase (PTS) in SN56-APPSWE were increased to 357.76 ± 8.95% (p< 0.001), 337.19 ± 27.10% (p< 0.001), 435.43 ± 49.02%, 286.89 ± 17.03% (p< 0.001), 464.08 ± 34.88% (p< 0.001), and 320.81 ± 46.80% (p< 0.01), respectively, but not in the MN9D-APPSWE cells (Figure 4.4A). Protein phosphatase 2 regulatory subunit A, alpha (PPP2R1A) and syndecan 3 (SDC3) were significantly decreased to 42.61 ± 2.61% (p< 0.001) and 31.47 ± 2.81% (p< 0.05) in SN56-APPSWE cells, while substantially increased to 124.71 ± 0.29% (p< 0.01) and 171.39 ± 26.23% (p< 0.05) in MN9D-APPSWE. Consistent with these findings, the gene expression of ADCY4, CXCL2, FN1, PPP1R3C, and PTS in SN56-APPSWE was increased to 610.36 ± 125.71%, 366.22 ± 42.59%, 489.63 ± 42.03%, 748.40 ± 101.74%, and 301.49 ± 43.50% (p<0.001), respectively, but no significant changes were observed in MN9D-APPSWE (Figure 4.4B). Additionally, the gene expression of CXCL1 was increased to 280.19 ± 39.34% in SN56-APPSWE (p< 0.001), and decreased to 58.21 ± 1.85% in MN9D-APPSWE (p< 0.05). On the contrary, the expression of PPP2R1A and SDC3 was inhibited by 42.75 ± 1.00% (p< 0.001) and 81.34 ± 2.2% (p< 0.01) in SN56-APPSWE, while enhanced by 119.24 ± 2.63% (p< 0.001) and 59.80 ± 21.24% (p<0.01) in MN9D-APPSWE. Overall, the expression levels of DEGs determined by microarray and qRT-PCR were correlated with each other for the APP-related arrays. 136 Figure 4.4 Validation of gene expression by qRT-PCR for selected DEGs from analyzing APP-related stable cells. (A) The relative expression levels of 8 genes in the significantly altered canonical pathways from microarray were summarized in bar graph. The signal intensities obtained from microarray in SN56-APPSWE and MN9D-APPSWE cells were normalized to SN56-APPWT and MN9D-APPWT, respectively. (B) Significantly 8 altered genes identified by microarray experiments were validated by qRT-PCR at mRNA level. The expression level of genes in the SN56-APPSWE and MN9D-APPSWE cells were represented as relative percentage after normalizing to SN56-APPWT and MN9D-APPWT, respectively. The values in this figure represent the means ± SEM. N =3, *p< 0.05, **p< 0.01, and ***p< 0.001 by two-way ANOVA followed by post-hoc Bonferroni's multiple comparisons test. Regarding the DEGs determined in the top canonical pathways for SNCA-related arrays, we focused on the genes involved in the NFκB signaling pathway, as it is implicated in neuronal inflammation and PD (Hirsch & Hunot, 2009). In the microarray, the gene expression of fibroblast growth factor receptor-like 1(FGFRL1), myeloid differentiation primary response gene 88 (MYD88), MRAS muscle RAS oncogene homolog (MRAS), neurotrophic receptor tyrosine kinase 3 (NTRK3), platelet derived growth factor receptor alpha (PDGFRA), and toll-like receptor 2 (TLR2) was increased to 232.29 ± 8.29% (p< 0.001), 302.63 ± 0.60% (p< 0.001), 137 573.64 ± 15.08% (p< 0.001), 241.93 ± 13.62% (p< 0.001), 314.55 ± 34.44% (p< 0.01), and 236.37 ± 1.44% (p< 0.001) in MN9D-SNCAA53T cells, respectively, while significantly decreased to 16.96 ± 1.87% (p< 0.001) , 23.11 ± 4.20% (p< 0.001), 29.15 ± 2.97% (p< 0.01), 29.06 ± 1.6% (p< 0.01), 6.91 ± 0.44% (p< 0.05), and 18.14 ± 1.53% (p< 0.001) in SN56-SNCAA53T (Figure 4.5A). The gene expression of epidermal growth factor receptor (EGFR) and lymphotoxin alpha (LTA) manifested 521.65 ± 9.78% (p< 0.001) and 314.06 ± 12.60% (p< 0.001) increase in MN9D-SNCAA53T cells, but no obvious change in SN56-SNCAA53T cells. To confirm these results, qRT-PCR was performed by using independent samples. Consistently, the expression of FGFRL1, MYD88, MRAS, NTRK3, PDGFRA, TLR2 was significantly elevated to 390.11 ± 44.16% (p< 0.001), 261.96 ± 61.96% (p< 0.001), 604.06 ± 44.53% (p< 0.001), 405.93 ± 49.13% (p< 0.001), 381.95 ± 8.69% (p< 0.01), and 428.60 ± 55.85% (p< 0.001) in MN9D-SNCAA53T cells, while lowered to 17.46 ± 0.75% (p< 0.001), 26.82 ± 2.99% (p< 0.001), 24.87 ± 1.19% (p< 0.01), 2.46 ± 0.09% (p< 0.01), 17.06 ± 0.89% (p< 0.05), and 42.56 ± 9.63% (p< 0.001) in SN56-SNCAA53T cells. Taken together, gene expression of selected DEGs in the SNCA-related array was validated by qRT-PCR. The molecules associated with the NFκB signaling pathway were upregulated in MN9D-SNCAA53T cells, but no change or downregulated in SN56-SNCAA53T cells, suggesting the NFκB signaling pathway may play a role in mediating the cell type- specific effect of SNCAA53T on gene expression. 138 Figure 4.5 Validation of gene expression by qRT-PCR for selected DEGs from analyzing SNCA-related stable cells. (A) The relative expression levels of 8 genes in the NFκB signaling pathways from microarray were summarized in the bar graph. The signal intensities of genes in SN56-SNCAA53T and MN9D-SNCAA53T cells were normalized to that in SN56-SNCAWT and MN9D-SNCAWT cells, respectively. (B) Eight significantly altered genes identified by microarray were validated by qRT-PCR at the mRNA level. The expression level of genes in SN56-SNCAA53T and MN9D-SNCAA53T cells was represented as a relative percentage after normalizing to SN56-SNCAWT and MN9D-SNCAWT, respectively. The values in this figure represent the means ± SEM. N =3, *p< 0.05, **p< 0.01, and ***p< 0.001 by two-way ANOVA followed by post-hoc Bonferroni's multiple comparisons test. 4.3.4 Differential expression of SDC3 and FGFRL1 proteins in cholinergic and dopaminergic cells carrying APPSWE or SNCAA53T mutations In the analyses of microarray and qRT-PCR, it was noticed that PPP2R1A and SDC3 were the two genes having statistical significance both in SN56-APPSWE and MN9D-APPSWE. SDC3 has been reported to be a γ-secretase substrate in the mouse embryonic fibroblasts (MEF) cells (Schulz et al., 2003). More importantly, it is accumulated in the glia with Aβ deposits in the APPSWE transgenic mice (Tg2576) (O'Callaghan et al., 2008), indicating it may play a role in the 139 AD pathogenesis. Therefore, we took the advantage of established cell models, and the endogenous SDC3 protein was measured by immunoblotting. It was found that SDC3 protein was lower in SN56-APPSWE cells (114.18 ± 5.00%) compared with SN56-APPWT (210.36 ± 14.34%, p< 0.05) after normalizing to their parental SN56 cells (Figure 4.6A-B). The endogenous SDC3 expression level was higher in MN9D cells (339. 54 ± 14.11%, p< 0.001) than SN56 cells. Overexpression of APPWT in MN9D cells enhanced its expression to 521.28 ± 31.50% (p< 0. 001), and overexpression of APPSWE further elevated its expression to 724.10 ± 28.66% (p< 0.001), suggesting that APPSWE affects the expression of endogenous SDC3 in an opposite way in cholinergic and dopaminergic neuronal cells. Among the validated genes by qRT-PCR in the NFκB signaling pathway, the physiological functions of FGFRL1 intrigues us based on the following evidence. FGFRL1 is relatively new member in the family of FGFR, also termed as FGFR5 (Sleeman et al., 2001; Wiedemann & Trueb, 2000). It is different from the rest of the family members, FGFR1-4, as it lacks the tyrosine kinase domain and works as a decoy receptor- antagonizing the functions of other FGFRs (Trueb, 2011). It is well-accepted that FGFR signaling play an essential role in regulating cell proliferation and cancer (Turner & Grose, 2010), and it has been postulated that FGFRL1 could inhibit cell proliferation, which was reported before (Steinberg, Gerber, Rieckmann, & Trueb, 2010). In order to explore whether FGFRL1 could be a mediator in SNCAA53T-associated pathology, endogenous FGFRL1 protein level was first examined by immunoblotting in SNCAWT/A53T- overexpressing cholinergic SN56 and dopaminergic MN9D cells. It was barely detectable in the SN56, SN56-SNCAWT, and SN56-SNCAA53T cells (Figure 4.6 C-D). However, its expression was significantly elevated to 2370.63 ± 314.63% in MN9D cells (p< 0.001), 140 2691.61 ± 55.06% in MN9D-SNCAWT (p< 0.001), and 3328.02 ± 74.95% in MN9D-SNCAA53T (p< 0.001). It suggests that FGFRL1 may mediate differential impacts of SNCAA53T on its pathological functions observed in cholinergic and dopaminergic neuronal cells. Figure 4.6 Differential expression of endogenous SDC3 and FGFRL1 protein in APPSWE- and SNCAA53T- overexpressing SN56 and MN9D cells. (A) The cell lysates from APP-related stable cells, SN56 cells, and MN9D cells, were prepared for immunoblotting to detect endogenous SDC3 protein levels. The full-length APP protein was examined to show its overexpression and β-actin was used as a loading control. (B) The intensities of SDC3 protein bands were quantified by Image J software and normalized to β-actin. The relative protein level of SDC3 in different cell lines was represented as a percentage with respect to SN56 cells. (C) Endogenous FGFRL1 protein levels were examined in SNCA-related stable cells, SN56 cells, and MN9D cell by immunoblotting. The αSyn-mycHis protein was examined to show its overexpression. (D) The intensities of FGFRL1 protein bands were quantified by Image J software and normalized to β-actin. The relative protein level of SDC3 in different cell lines was represented as a percentage with respect to SN56 cells. The values in this figure represent the means ± SEM. N =3, *p< 0.05, **p< 0.01, and ***p< 0.001 by two-way ANOVA followed by post-hoc Bonferroni's multiple comparisons test. 141 Overall, the expression of endogenous SDC3 and FGFRL1 protein responded differentially to APPSWE and SNCAA53T mutation overexpressed in cholinergic and dopaminergic cells, suggesting they could be potential targets for studying cell type- selective pathology induced by APPSWE and SNCAA53T mutations. 4.3.5 Differential expression of endogenous SDC3 and FGFRL1 protein in AD and PD mouse models In pursuit of extending our findings from cellular models to animal models, we took the advantage of available APPSWE knock-in mice (APPNL/NL) and SNCAA53T transgenic mice (Prnp-SNCAA53T). The recent APPNL/NL mice were developed by Saito’s group, featured by its murine Aβ sequence being engineered into the human sequence with APPSWE mutation (Saito et al., 2014). Its phenotypes are milder compared with APPSWE transgenic mice, such as APP23 mice (Saito et al., 2014; Sturchler-Pierrat et al., 1997). We aimed to compare the endogenous SDC3 expression in the cholinergic neuron and dopaminergic neuron in APPSWE animal models, so we wanted to eliminate the dosage effect of full-length APP. However, the expression of full-length APP protein in APP23 mice was not equal across varying brain regions detected by immunoblotting (data not shown). Under such circumstance, APPNL/NL mice were more suitable for this study. ChAT and TH are commonly used markers for identifying cholinergic and dopaminergic neurons, respectively (Choi et al., 2013; Ryan et al., 2013). We tried to detect the expression of SDC3 in cholinergic and dopaminergic neurons by co-immunofluorescent staining, but all tried SDC3 antibodies were failed (data not shown). Another alternative method was employed. The components of the basal forebrain cholinergic system, MS and NBM, and a 142 major dopaminergic nucleus, substantial nigra, were dissected from fresh mouse brains (Figure 4. 7B). The expression of SDC3 protein was analyzed by immunoblot analyses from brain homogenates of MS, NBM and SN regions and then compared with age-matched wildtype C57BL/6J (C57) mice (n= 3) (Figure 4.7A). In the C57 mice, the relative expression of endogenous SDC3 was 14.39 ± 8.87% (p> 0.05) in NBM, and 33.63 ± 25.11% in SN (p>0.05) after normalizing to MS (100 ± 13.83%) (Figure 4.7C). Unexpectedly, a significantly increased expression of SDC3 in the APPNL/NL mice was observed in MS and NBM (p< 0.05), but not in SN. With respect to the three regions in APPNL/NL mice, SDC3 was substantially higher in MS than SN (Figure 4.7C). These results were contradictory to the findings from the cellular models. The difference between the APPNL/NL mice and stable cell lines made us speculate that whether there was a dosage effect of APP on SDC3’s expression. To answer this question, we transiently transfected SN56 and MN9D cells with varying dosages of APPWT and APPSWE expression vector followed by examining SDC3 protein level (Figure 4.7D). In the lower dosage groups (1.0μg and 2.0μg), the expression pattern of SDC3 was almost the same as the results from APPNL/NL and C57 (Figure 4.7C). But in the highest dosage group (3.0μg), the pattern was consistent with the results from stable cell lines (Figure 4.6B), further suggesting APPSWE plays a role in regulating SDC3 expression. 143 Figure 4.7 Differential expression of the endogenous SDC3 protein in the MS and SN of APPNL/NL mice. (A) The cell lysate from MS, NBM, and SN of APPNL/NL (n=3) and C57 mice (n=3) was prepared to detect endogenous SDC3 by immunoblotting, and the results from 3 independent samples were presented. β-actin was used as a loading control. The SDC3 bands in APPNL/NL mice samples were acquired under low exposure and high exposure conditions, and the latter was the same setting as C57 samples. (B) The images of brain sections containing MS, NBM, and SN, were taken after dissecting the fresh mouse brains in a brain matrix. The black circles were illustrated as dissected areas for immunoblotting. (C) The intensities of bands in A) were quantified by Image J software. After normalizing to β-actin, the SDC3 protein levels in different regions were represented as a percentage relative to its level in MS of C57 mice. (D) Cholinergic SN56 and dopaminergic MN9D cells were transiently transfected with human APPWT/ APPSWE expression vectors, and harvested 48 hours post-transfection. Cell lysates were harvested to detect full-length APP and SDC3 protein by immunoblotting. SN56 and MN9D cells transfected with empty vector were the negative controls and β-actin was utilized as the loading control. (E) The intensities of SDC3 bands with APPWT/ SWE overexpression were quantified by Image J software. The level of SDC3 in each dosage group was represented as a relative percentage after comparing with SDC3 protein level in SN56 cells transfected with APPWT after normalizing to β-actin. The values in this figure represent the means ± SEM. N =3, *p< 0.05, **p< 0.01, and ***p< 0.001 by two-way ANOVA followed by post-hoc Bonferroni's multiple comparisons test. 144 Next, we further examined how SNCAA53T mutation regulated the expression of endogenous FGFRL1 in cholinergic and dopaminergic neurons of Prnp-SNCAA53T transgenic mice, which exhibit pathological accumulation of αSyn aggregates, neuron loss, and progressive motoric dysfunction (Lee et al., 2002; Martin et al., 2006). To confirm αSyn expression was equal in the cholinergic and dopaminergic nucleus, MS, NBM, and SN brain regions were dissected from fresh brains of Prnp-SNCAA53T transgenic mice and control C57 mice. αSyn was expressed at the similar level across theses three regions (data not shown). As mentioned before, ChAT and TH were chosen as the molecular markers for identifying cholinergic and dopaminergic neurons, respectively. FGFRL1 was detected by co-immunofluorescent staining with either ChAT or TH in the basal forebrain cholinergic neurons and SN dopaminergic neurons (Figure 4.8). We did not observe a difference in the number of ChAT-positive neurons and TH-positive neurons for transgenic and control mice at the age of 4 month-old and 10 month-old, which was consistent with the findings in a previous study (Daher et al., 2012). The overall FGFRL1 expression was lower in the basal forebrain cholinergic neurons than in the SN dopaminergic neurons for both Prnp-SNCAA53T transgenic mice and C57 mice by quantifying the average immunofluorescence intensity. In the basal forebrain cholinergic neurons, the average immunofluorescence intensity of FGFRL1 was 4.65 ± 1.23 in C57 mice and 1.95 ± 0.31 in transgenic mice (p< 0.05, Figure 4.8A-B). Contrarily, the average immunofluorescence intensity of FGFRL1 in the SN dopaminergic neurons was significantly increased from the 11.42 ± 0.87 in C57 mice to 22.72 ± 1.35 in transgenic mice (Figure 4.8C-D). Collectively, SNCAA53T downregulated FGFRL1 expression in the cholinergic neurons and upregulated its expression in the dopaminergic neurons, which was consistent with the discovery from stable cells. 145 Figure 4.8 Differential expression of the endogenous FGFRL1 protein in the cholinergic and dopaminergic neurons of Prnp-SNCAA53T mice. (A) The basal forebrain cholinergic neurons in transgenic and wildtype mice were identified by ChAT, which was co-immunostained with DAPI and FGFRL1. The images in the second and fourth rows were zoomed in from correlated images from the first and third rows. Scale bar: first and third row, 50 μm; second and fourth row, 10 μm. (B) Quantification of average immunofluorescent intensities of FGFRL1 in ChAT+ neurons after subtracting image background. (C) The SN dopaminergic neurons in transgenic and wildtype mice were identified by TH, which was co-immunostained with DAPI and FGFRL1. The images in the second and fourth rows were zoomed in from correlated images from the first and third row. Scale bar: first and third row, 50 μm; second and fourth row, 10 μm. (D) Quantification of average immunofluorescent intensities of FGFRL1 in TH+ neurons after subtracting image background. The values in this figure represent the means ± SEM. N =4, *p< 0.05, **p< 0.01, and ***p< 0.001 by Student’s t-test. 146 4.3.6 The effects of SDC3 and FGFRL1 on oxidative stress-induced cell death After discovering the expression of endogenous SDC3 protein was specifically higher in the cholinergic nucleus of APPNL/NL transgenic mice than that in the dopaminergic nucleus, we further examined whether the cell susceptibility of cholinergic cells carrying APPSWE mutation to oxidative stress could be alleviated by SDC3 knockdown. Considering the dosage effect of APPSWE on regulating SDC3 expression (Figure 4.7D), we transiently transfected SN56 and MN9D cells with APPSWE expression vector with relative low dosage to establish a cellular model, which resembles APPNL/NL mice in terms of the SDC3 expression pattern. The transiently transfected cells were treated with SDC3 or control siRNA to manipulate the endogenous SDC3 level, which was confirmed by immunoblotting (Figure 4.9A). After applying H2O2 for 12 hours, knockdown of SDC3 in SN56 cells carrying the APPSWE mutation significantly rescued cell death in all three dosage groups, from 66.08 ± 6.64% to 44.91 ± 1.93% in 50 μM (p< 0.05), from 73.60 ± 3.93% to 52.11 ± 1.69% in 100 μM (p< 0.01), and from 81.34 ± 1.60% to 59.38 ± 1.55% in 300 μM (p< 0.01). However, the significant change was not observed in MN9D cells carrying the same mutation (Figure 4.9B). For the purpose of testing FGFRL1’s influence on oxidative stress-introduced cell death, SN56 and MN9D cells stably overexpressing SNCAA53T were treated with FGFRL1 siRNA or control, and received H2O2 treatment. The knockdown experiment was verified by detecting endogenous FGFRL1 mRNA level by RT-PCR (Figure 4.9C), as its protein was barely detectable in SN56 cell by immunoblotting (Figure 4.6C). Cell death in MN9D-SNCAA53T stable cells with FGFRL1 knockdown was substantially decreased from 76.31 ± 6.72% to 51.51 ± 1.30% in 100 μM group (p< 0.01), and from 93.44 ± 0.69% to 67.38 ± 1.81% in 300 μM group (p< 0.01), 147 although the difference was also shown in the 0 μM group (p< 0.05, Figure 4.9D). Taken together, downregulation of SDC3 and FGFRL1 rescued oxidative stress-induced cytotoxicity in APPSWE-overexpressing cholinergic cells and SNCAA53T-overexpressing dopaminergic cells, respectively, suggesting SDC3 and FGFRL1 may be potential targets in alleviating APPSWE and SNCAA53T-associated cytotoxic effects. Figure 4.9 . Knockdown of SDC3 and FGFRL1 alleviates oxidative stress- induced cell death in SN56-APPSWE and MN9D-SNCAA53T cells, respectively. (A) APPSWE was transiently expressed in SN56 or MN9D cells, and the transfected cells were treated with SDC3 siRNA or control siRNA. Cell lysate was subject to immunoblotting analysis, and knockdown of SDC3 was confirmed by detecting its protein expression level. (B) SN56 and MN9D cells overexpressing APPSWE with SDC3 knockdown were treated with varying concentrations of H2O2 for 12 hours and cell death was measured by LDH assay as mentioned before. (C) SN56-SNCAA53T and MN9D-SNCAA53T stable cells were transfected with FGFRL1 or control siRNA, and the knockdown experiment was verified by measuring endogenous FGFRL1 mRNA by RT-PCR. (D) SN56-SNCAA53T and MN9D-SNCAA53T stable cells transfected FGFRL1 siRNA or control were applied varying concentration of H2O2 for 12 hours. The cell death was examined by LDH assay and 0.1% triton was the positive control (100% cytotoxicity). The values in this figure represent the means ± SEM. N =3, *p< 0.05, **p< 0.01, and ***p< 0.001 by two-way ANOVA followed by post-hoc Bonferroni's multiple comparisons test. 148 4.4 Discussion DNA microarray provides an invaluable tool to examine the whole genome profiling in the neurodegenerative disorders, although it is challenging to interpret the results obtained from varying platforms and samples. In our study, we compared the gene expression profiling in cholinergic and dopaminergic cells overexpressing either APP variants or SNCA variants. The DEGs were the genes affected by APPSWE and SNCAA53T mutations in a cell type- dependent manner- specifically in cholinergic and dopaminergic cells, respectively. In this way, we focused on how the two neuronal populations respond differentially to the same mutation at a systematic level. Previous studies have compared the gene expression profiling of AD or PD patients with the healthy controls, but the postmortem brain samples were obtained without dissecting specific neuronal population (Ginsberg, Hemby, Lee, Eberwine, & Trojanowski, 2000; Grünblatt et al., 2004; Hauser et al., 2005; Miller, Oldham, & Geschwind, 2008; Nagasaka et al., 2005; Yanli Zhang et al., 2005). More precise dissection was achieved with the help of LCM in the AD studies, but the genotypes of these AD patients are unknown (Counts, Che, Ginsberg, & Mufson, 2011; Counts et al., 2007). The same situation is applicable to the microarray studies in PD (Cantuti-Castelvetri et al., 2007; Chung et al., 2005; Elstner et al., 2011; Simunovic et al., 2009). In a microarray study of AD, the gene expression profiling of the cerebral cortex in APPSWE transgenic mice was compared with WT mice at three disease stages, genes involved in apoptosis and mitochondrial energy metabolism are consistently found to be upregulated for all three disease stages (Reddy et al., 2004). By overexpressing APPSWE or APPWT in H4 cells, kinase and phosphatase signaling is overrepresented in the gene list (Shin, Yu, Yu, Jo, & Ahn, 2010). Compared with the controls, the altered genes in the AD brains are enriched in the following 149 canonical pathways: inflammation (Miller et al., 2008; Tan et al., 2010), calcium signaling (Emilsson et al., 2006), and synaptic process (Ginsberg et al., 2000; Miller et al., 2008). These canonical pathways were not present in our results, which is probably due to different comparison strategies. It is reasonable to speculate the common pathways both affected in cholinergic neurons and dopaminergic neurons in AD patients will not be the potential targets in our study. As for PD-related microarray studies, the effects of SNCAA53T on gene expression are explored in either cellular models or transgenic mice. Overexpression of SNCAA53T in the BE-M17 neuroblastoma cells significantly changes the expression of genes in stress response, transcription, and apoptosis (Baptista et al., 2003). By using Prnp-SNCAA53T transgenic mice and C57 control mice, the significantly altered genes are implicated in the following cellular functions: cell-cell signaling, immune response, as well as transcription and translation (Miller et al., 2007). Furthermore, it has been pointed out that dopaminergic signaling-related genes and DA-inducible genes are changed in two lines of SNCAA53T transgenic mice compared with WT mice (Kurz et al., 2010). In an elegant study, gene expression profiling was compared between dopaminergic neurons carrying SNCAA53T mutation and isogenic mutation-corrected neurons, which were generated from human induced pluripotent stem cell (hiPSC). The data reveals that prosurvival genes and genes associated metabolic function are downregulated (S. D. Ryan et al., 2013). SNCAA53T mutation has been found to affect genes enriched in apoptosis (Baptista et al., 2003), neurotransmitter signaling (Kurz et al., 2010), and inflammation (Miller et al., 2007). In our study, we also identified several canonical pathways associated with inflammation, such as crosstalk between dendritic cells and natural killer cells, NFκB signaling, antigen presentation pathway, and T helper cell differentiation. And affected cellular functions in IPA indicates the 150 involvement of cell death and survival, and cell-to-cell signaling and interaction, although cell development and cell movement were valued more in the list. Among the validated genes by qRT-PCR, SDC3 and FGFRL1 were selected to be investigated in AD and PD mouse models, respectively. SDC3 is a member of transmembrane heparan sulfate proteoglycan (HSPGs), partially cloned by using Schwann cells in 1992 (Carey et al., 1992). And its full sequence was completed in 1997 (Carey et al., 1997; Carey et al., 1992). Its expression is enriched in the central nervous system (Carey et al., 1992) and dependent on the developmental stages, with high level in neonatal brain and barely detectable level in the adult brain (Carey et al., 1997; Carey et al., 1992). There are three other closely-related members, SDC1, SDC2, and SDC4. Their functions are not clearly defined, but initial findings suggest they mediate cell-cell or cell-matrix interaction, work as co-receptors for growth factors and protease, and function as low-affinity receptors for some enzymes (Bernfield et al., 1992). SDC3 has a short cytoplasmic sequence, a single-span transmembrane domain, and a large extracellular domain with several attachment sites for heparan and chondroitin sulfate chains (Tkachenko, Rhodes, & Simons, 2005). The SDC3 deficient mice display enhanced LTP and impaired hippocampal-dependent memory, suggesting it is physiologically involved in memory (Kaksonen et al., 2002). Additionally, SDC3 is found to be expressed in glia associated with Aβ deposits, and its expression is increased in Aβ-treated glia in APPSWE transgenic mice (O'Callaghan et al., 2008). In the APPSWE knock-in mice, we determined that its expression was significantly elevated in a cholinergic nucleus - medium septum, but not in the SN (Figure 3.5). Our knockdown experiment showed that its reduction partially rescued H2O2-induced cytotoxicity in SN56 cell 151 transiently overexpressing APPSWE, which was not observed in MN9D cells (Figure3.6). Taken together, SDC3 may be a mediator underlying selective neuronal death in AD. FGFRL1 was first discovered from cartilage in 2000, and it was named after the close relation to the other fibroblast growth factor receptors (FGFRs) 1-4 (Wiedemann & Trueb, 2000). Structurally, it contains a large extracellular domain with three immunoglobin-like structure, a transmembrane domain with a single span, and an intracellular domain (Kim, Moon, Yu, Kim, & Koh, 2001; Sleeman et al., 2001; Wiedemann & Trueb, 2000). It differs from other members in the lack of tyrosine kinase activity in its cytoplasmic domain, thus it could be a decoy receptor. Another possibility is to antagonize the cellular functions of other FGFRs, as it can bind to several genuine fibroblast growth factors (FGFs) without inducing downstream effectors (Trueb, 2011). It is widely expressed in multiple tissues and organs, such as pancreas, brain, heart, liver, lung, kidney, and skeletal muscle (Kim et al., 2001; Sleeman et al., 2001; Trueb, Zhuang, Taeschler, & Wiedemann, 2003; Wiedemann & Trueb, 2000). Although its biological functions are not well-defined, several studies shed light on its role in cell adhesion, proliferation, and differentiation (Rieckmann, Kotevic, & Trueb, 2008; Steinberg, Gerber, et al., 2010). The knockout studies consistently observe that homozygous FGFRL1-/- mice display severe abnormality in kidney and die prematurely, indicating its essential role in kidney development (Catela et al., 2009; Gerber, Steinberg, Beyeler, Villiger, & Trueb, 2009). In our IPA results, FGFRL1 was identified in the NFκB pathway and upregulated in MN9D-SNCAA53T stable cells instead of SN56-SNCAA53T stable cells. The direct link between FGFRL1 and the NFκB signaling is not available, but the indirect evidence provides informative 152 knowledge. FGFRL1 is consistently found to bind to FGFs (FGF2, FGF3, FGF4, FGF8, and FGF10) (Sleeman et al., 2001; Steinberg, Zhuang, et al., 2010), and the binding of FGFs to classic FGFRs activate the downstream NFκB and other signaling pathways (Bushdid et al., 2001; Drafahl, McAndrew, Meyer, Haas, & Donoghue, 2010; Kanegae, Tavares, Izpisua Belmonte, & Verma, 1998). In our findings, endogenous FGFRL1 expressed significantly higher in MN9D cells and dopaminergic neurons of SNCAA53T transgenic mice. Besides, knockdown of FGFRL1 alleviated cell death specifically in SNCAA53T- overexpressing MN9D cells other than SN56 cells. If FGFRL1 can be linked with NFκB signaling, which has a well-established role in inflammation, cell death and survival (Mattson & Meffert, 2006; Tansey & Goldberg, 2010), FGFRL1 may be an attractive target in exploring the mechanisms underlying neurodegeneration of PD. 4.5 Conclusion In summary, APPSWE and SNCAA53T mutations changed gene expression profiling of cholinergic SN56 cell and dopaminergic MN9D cells in a cell type- dependent manner. DEGs found in two cell lines carrying APPSWE mutation were involved in granulocyte adhesion and diapedesis, agranulocyte adhesion and diapedesis, β-alanine degradation, and LXR/RXR activation. DEGs affected in a cell type- dependent manner by SNCAA53T mutation were implicated in the three top pathways: crosstalk between dendritic cells and natural killer cells, NFκB signaling, cleavage and polyadenylation of pre-mRNA, suggesting mechanisms related with inflammation and RNA processing. SDC3 and FGFRL1 proteins were differentially expressed in APP-related stable cells and SNCA-related stable cells, respectively. Moreover, SDC3 was increased in MS brain region of APPSWE knock-in mice, while FGFRL1 was elevated in dopaminergic neurons in 153 SNCAA53T transgenic mice. Finally, by knocking down of SDC3 and FGFRL1 in APP-related and SNCA-related stable cells, respectively, it rescued oxidative stress-induced cell death in cholinergic cells with APPSWE mutation and dopaminergic cells carrying SNCAA53T mutation. Taken together, APPSWE and SNCAA53T affected the gene expression profiling in a cell type- dependent manner, and SDC3 and FGFRL1 could be potential targets to alleviate selective neurodegeneration caused by APPSWE and SNCAA53T mutation, respectively. 154 Chapter 5: Conclusions and discussions 5.1 General discussion The works described in this thesis are from two projects. In Chapter 2, the first project focused on characterizing the promoter of human LRRK2 gene and how this gene was regulated by the transcription factor Sp1 at the transcriptional and translational level. Although accumulating studies suggests the essential role of LRRK2 in both familial and sporadic PD, the study of its gene promoter is absent. Therefore, we aimed to explore the features of the LRRK2 gene promoter and the regulatory mechanisms underlying its gene expression. After cloning the human LRRK2 gene promoter, its TSS was located and its promoter activity was detected. In its promoter sequence, three putative Sp1 binding sites were identified and only two of them were functional. Finally, Sp1 overexpression upregulated LRRK2 gene expression, while a specific Sp1 inhibitor, MTM, reduced LRRK2 gene expression. The second project was aimed to discover novel targets for understanding the selective neurodegeneration in AD and PD. As familial and sporadic cases share common molecular mechanisms (Cookson, Hardy, & Lewis, 2008; Schellenberg & Montine, 2012), we took the advantage of APPSWE and SNCAA53T mutations to discover the pathogenic mechanisms in AD and PD, respectively. Since cholinergic neurons and dopaminergic neurons undergo severe cell demise in AD and PD, respectively (Fearnley & Lees, 1991; Whitehouse et al., 1981), the mutant or wildtype APP and SNCA genes were overexpressed in SN56 cholinergic and MN9D dopaminergic neuronal cells. In Chapter 3, we examined whether well-validated pathogenic effects of APPSWE and SNCAA53T were different in cholinergic and dopaminergic cells. Therefore, Aβ generation and αSyn aggregates formation were investigated in APP-related and 155 SNCA-related stable cells. The results support that more deleterious effects of the APPSWE and SNCAA53T mutations were observed in cholinergic cell and dopaminergic cells, respectively. Following this, we evaluated how the mutations changed the cell viability of cholinergic and dopaminergic cells in response to oxidative stress and oligomeric treatment. Consistent with our expectations, cholinergic cells overexpressing APPSWE mutation were more sensitive to H2O2-induced oxidative stress than dopaminergic cell carrying the same mutation. In contrast, more dopaminergic cells overexpressing SNCAA53T mutation underwent cell death in H2O2 treatment than cholinergic cells with the same mutation. Since oligomers are more disease-related stressors, our stable cells were treated with Aβ and αSyn oligomers to examine how mutations changed the responses of cholinergic and dopaminergic cells to such insults. Our results demonstrated that Aβ oligomers only caused cell death in SN56 cells but not in MN9D cells, with a stronger effect in SN56 cells carrying the APPSWE mutation. Additionally, αSyn WT oligomers resulted in the most severe cell loss in MN9D-SNCAWT stable cells, while αSyn A53T oligomers led to the most cell death in MN9D-SNCAA53T stable cells. Collectively, APPSWE and SNCAA53T mutations exerted stronger pathogenic effects in cholinergic and dopaminergic cells, respectively. On the other hand, APPSWE and SNCAA53T mutations made cholinergic and dopaminergic cells more susceptible to cytotoxic insults. In Chapter 4, we determined to examine the gene expression profiling changed by the APPSWE and SNCAA53T mutations in cholinergic and dopaminergic cells to gain more knowledge in the hope of finding novel targets to explain the selective neurodegeneration in AD and PD. After performing DNA microarray, DEGs analyses were conducted to discover genes differentially expressed in two neuronal cell lines carrying the same mutation. With the help of IPA, DEGs 156 showing cell type- specific effects of APPSWE overexpression were overrepresented in granulocyte adhesion and diapedesis, agranulocyte adhesion and diapedesis, β-alanine degradation I, LXR/ RXR activation, and PCP pathway. For SNCA-related stable cells, genes differentially altered by SNCAA53T in two neuronal cell lines were involved in crosstalk between dendritic cells and natural killer cells, NFκB signaling, cleavage and polyadenylation of pre-mRNA, antigen presentation pathway, and T helper cell differentiation. Based on our knowledge, we selected two molecules, SDC3 and FGFRL1, from the gene list for further investigation. Their expression affected by the APPSWE and SNCAA53T mutations was examined in vitro and in vivo. It was shown that the APPSWE mutation affected the expression of SDC3 in a dosage-dependent manner in cholinergic cells. With a relatively low level of overexpression, APPSWE increased SDC3 expression in cholinergic cells (Figure 4.7). In line with this, SDC3 was found to be elevated in the MS of APPSWE knock-in mice. Regarding FGFRL1, its expression was significantly enhanced by overexpression of the SNCAA53T mutation in dopaminergic cells and reduced in cholinergic cells. Consistently, the expression of FGFRL1 protein was much higher in dopaminergic neurons than cholinergic neurons in SNCAA53T transgenic mice. To determine the potential roles of SDC3 and FGFRL1 in neurodegeneration, we knocked down SDC3 and FGFRL1 in APP-related and SNCA-related stable cells, respectively, and determined their impacts on the oxidative stress-induced cell death. Interestingly, SDC3 deficiency partially rescued cell death for cholinergic cells overexpressing APPSWE mutation compared with its WT in H2O2 treatment, but not for dopaminergic cells. Furthermore, FGFRL1 knockdown substantially decreased cell susceptibility to H2O2 due to SNCAA53T overexpression in dopaminergic cells, but this is not the case for cholinergic cells. Taken together, APPSWE and SNCAA53T altered the gene expression in a cell type- dependent manner, and SDC3 and FGFRL1 157 could be potential targets to alleviate selective neurodegeneration caused by APPSWE and SNCAA53T mutation, respectively. 5.2 Novelty and significance of research The project described in Chapter 2 is the first study to reveal features of the human LRRK2 promoter and report the role of Sp1 in regulating LRRK2 gene expression. Application of MTM was demonstrated to decrease LRRK2 gene expression. MTM is the well-known antibiotic having antitumor activity and was used to treat testicular cancer, leukemia, and hypercalcemia (De Souza e Silva & Wilson, 1973; Perlia et al., 1970; Ream, Perlia, Wolter, & Taylor, 1968). However, its hepatic toxicity impedes its clinical use (Zojer, Keck, & Pecherstorfer, 1999). Recently, its application to neurodegenerative diseases provides a novel possibility to make it into a neuroprotective drug (Osada, Kosuge, Ishige, & Ito, 2013), which will be further discussed later. Our research reveals its involvement in regulating the expression of the human LRRK2 gene, a major player in both sporadic and familial PD cases. It offers a candidate for alleviating toxic effects of overexpressing wildtype LRRK2, and even attenuating the pathogenic effects of PD-associated LRRK2 mutations. Of course, this speculation needs to be tested in future experiments. And it is of vital importance to define the dosage range to assure minimal toxicity without affecting the general protein synthesis. In the second project, we established our cellular models by stably overexpressing APPWT/SWE or SNCAWT/A53T in cholinergic SN56 and dopaminergic MN9D cells. As SN56 and MN9D cells have neuronal properties (Choi et al., 1991; Hammond et al., 1986), they can be used as cellular models to test the effects of potential drugs or modifiers on attenuating APPSWE- and SNCAA53T- 158 associated pathologies. Another benefit of using these cellular models is able to study AD and PD in parallel and to discover molecular mechanisms common to both of them or specific to only one of them. This is the first study to compare the cytotoxic effects of the APPSWE and SNCAA53T mutations in cholinergic and dopaminergic cells at the same time. By treating our stable cells with Aβ oligomers or αSyn oligomers, it was observed that the treatment only induced cell death in one of the two neuronal cell lines carrying either APPSWE or SNCAA53T mutation, suggesting a selective effect of oligomer-induced cytotoxicity in different neuronal cells. Therefore, these cellular models can be used for screening possible reagents or drugs capable of reversing oligomer-induced toxicity. In Chapter 4, this is the first study to compare the whole-genome expression profiling in two neuronal cell lines overexpressing either APPSWE or SNCAA53T mutation in order to find disease-related pathways having cell type- specific effects. Previous studies support that inflammation-related mechanisms are implicated in AD (Parachikova et al., 2007; Tan et al., 2010; Youn et al., 2007). Our results also suggest that inflammation could be the mechanism underlying selective neuronal death in AD. As for PD, NFκB signaling could be an interesting pathway for revealing molecular identities responsible for selective neurodegeneration. Notably, the expression of SDC3 and FGFRL1 was examined both in vitro and in vivo. Their expression was differentially regulated by the APPSWE and SNCAA53T mutations in cholinergic and dopaminergic neurons, indicating neuronal-type specific effects for the first time. More importantly, knockdown of SDC3 and FGFRL1 rescued cell death of certain neuronal cells carrying the disease-related mutations in response to the oxidative stress. Although SDC3 has been investigated in AD before, this is the first study to show its drastically increased expression in the MS of APPSWE 159 knock-in mice. Since it has been pointed out that SDC3 is barely detected in adult brains (Carey et al., 1997; Carey et al., 1992), its substantial increases in the cholinergic nucleus in the APPSWE knock-in mice shed light on its possible role in the AD pathogenesis. Previous studies focusing on FGFRL1 were aimed to disclose its physiological functions (Catela et al., 2009; Gerber et al., 2009; Rieckmann et al., 2008; Steinberg, Gerber, et al., 2010). In our study, FRFGL1 was linked to pathogenic effects of SNCAA53T for the first time, and its possible role in neurodegeneration was tested for the first time, although the conclusion is tentative. 5.3 Limitation and future directions One limitation of our first project is that we did not test the effects of Sp1 signaling on LRRK2 in vivo. Since MTM is a potential target for alleviating LRRK2-associated toxicity, we will test its different dosage in LRRK2 transgenic mice or LRRK2 mutation knock-in mice (Xu, Shenoy, & Li, 2012). The safety and efficacy of MTM on attenuating PD-related phenotypes in these mice will be determined, such as numbers of dopaminergic neurons in SN, striatal dopaminergic neurotransmission, and motor activities. If the toxicity of MTM limits its application in mice, the available analogs of MTM offer other alternatives (Albertini et al., 2006). Regarding SDC3 and FGFRL1, it is necessary to further examine their roles in vivo. Local knockdown of SDC3 and FGFRL1 genes could be conducted by viral-mediated RNA interference in the cholinergic and dopaminergic nuclei of APPSWE knock-in mice and SNCAA53T transgenic mice, although this requires a highly experienced technician. Following the generation of knockdown animal models, AD- and PD- related pathological and behavioral phenotypes will be examined. However, the relatively mild phenotypes of APPSWE knock-in mice should be taken into consideration (Saito et al., 2014; Sturchler-Pierrat et al., 1997). Maybe the effects of knocking 160 down SDC3 in this mouse model is hard to be detected. Under such circumstance, another knock-in mouse model harboring three APP mutations can be employed, which shows more aggressive phenotypes than APP23, the classic APPSWE transgenic mice (Saito et al., 2014). One of the major limitations of our study is the neuronal cell lines we used. Although they have the neuronal properties, they are not primary neurons. Therefore, the conclusions drawn from SN56 and MN9D cells may not be applicable to cholinergic and dopaminergic neurons. To make up the weakness of cell lines we used, SDC3 and FGFRL1 were tested in vivo. Ideally, the recent development of induced pluripotent stem cell (iPSC) technique will be much better to be used as the in vitro model. Dopaminergic and cholinergic neurons can be differentiated from iPSC, and APP and SNCA variant can be transduced into neurons by virus. Another alternative method is to obtain iPSCs from APPSWE-associated AD and SNCAA53T-associated PD patients. And isogenic corrected iPSC-differentiated WT neurons will be used as the control. However, the availability of such iPSCs is relatively limited, and their neuronal phenotypes after differentiation should be examined carefully and comprehensively. Similarly, it would be better to dissect cholinergic and dopaminergic neurons from APPSWE and SNCAA53T mouse models by LCM technique in the DNA microarray study. In this way, more in vivo-applicable information can be extracted from the microarray data, and it will be easier to correlate the findings from microarray with the following validation in vivo. 161 5.4 Further discussion 5.4.1 The application of MTM in treating neurodegenerative diseases As an antibiotic, MTM was discovered almost 50 years ago, and has been well-known for its antitumor activity (Kennedy, Yarbro, Kickertz, & Sandberg-Wollheim, 1968). It was tested for treating testicular cancer, leukemia, and hypercalcemia before (De Souza e Silva & Wilson, 1973; Perlia et al., 1970; Ream et al., 1968). However, its safety has attracted the attention from researchers, especially its severe hepatic toxicity (Zojer et al., 1999). Nowadays, its application is limited to the research use, and several clinical trials to test its efficacy in treating various cancers are under way without drawing a final conclusion (Taylor, Parsons, Han, Jayaraman, & Rege, 2011). Additionally, recent studies highlight its translational significance in neurodegenerative diseases. In a transgenic mouse model of Huntington's Disease (HD), MTM treatment in the range of 50 μg/ kg/ day to 150 μg/ kg/ day extended lifespans, improved motor performance, and delayed pathological events in a dosage-dependent manner. However, the higher dosages of 300 μg/ kg/ day and 600 μg/ kg/ day increased morbidity and death. It has been suggested that MTM prevents the inhibition of gene expression resulting from the hypermethylation of histone H3 in HD mouse model (Ferrante et al., 2004). Another study reported that pretreating the mice with 300 μg/ kg MTM prevented hyperlocomotion caused by repeated METH administration. Pretreatment of 300 μg/ kg MTM in combination with the subsequent administration of MTM (75, 150 or 300 μg/kg) significantly attenuated deficiency of dopamine, 3,4-dihydroxyphenylacetic acid (DOPAC), and dopamine transporters associated with extended METH usage (Hagiwara, Iyo, & Hashimoto, 2009). Moreover, MTM is able to alleviate ER stressors-induced neurotoxicity in organotypic hippocampal slice cultures by decreasing the expression of a transcription factor, C/EBP homologous protein (CHOP) (Kosuge, 162 Taniguchi, Imai, Ishige, & Ito, 2011). It is exciting that several MTM analogues with lower toxicity are already available, such as MTM SD and MTM SDK. They have stronger inhibitory effects on the tumor growth with lower effective dosage (Albertini et al., 2006). More importantly, their maximum tolerated dosages are higher than MTM tested in mice (Previdi et al., 2010). Collectively, it is still not clear whether MTM is a druggable target for delaying the pathogenesis of LRRK2-related PD or even all PD cases. However, it is worth our efforts to explore the therapeutic potentials of MTM and its analogs in consideration of both safety and efficacy. 5.4.2 Oligomer-based treatment in AD and PD Since the discoveries of Aβ and αSyn oligomers in the pathogenesis of AD and PD, it has started a new area of exploring disease-modifying drugs by targeting oligomeric species. Such oligomer-targeting strategies can be generally divided into several categories, including the inhibition of fibrillogenesis by reagents or modulators, enhanced clearance of oligomeric species by immunotherapy, and blockage of potential receptors for oligomers (Hefti et al., 2013; Kalia et al., 2013). Among these, immunotherapy has attracted the most attention, and mixed results generate a debate over the oligomer-based therapies. In 1999, the first study was reported to test the active Aβ immunotherapy in APP V717F transgenic mice (PDAPP) with positive effects on preventing plaque formation (Schenk et al., 1999). Afterwards, the first clinical trial for the active Aβ immunization was terminated due to meningoencephalitis (Gilman et al., 2005). In pursuit of finding an effective passive immunotherapy, the recently completed Phase III trials for testing two Aβ-targeting antibodies (Solanezumab and Bapineuzumab) were relatively disappointing, as they did not offer clear benefits for the determined primary outcomes. 163 Solanezumab is a humanized monoclonal antibody against the middle region of Aβ and recognizes soluble monomeric Aβ, while Bapineuzumab is raised against the N-terminal region of Aβ and binds to both soluble and insoluble fibrillar Aβ species (Doody et al., 2014; Salloway et al., 2014). Strictly speaking, both of them are not oligomer-specific antibodies. The oligomer-specific antibodies have been developed, presenting protective results in blocking Aβ-induced synaptotoxicity (Hillen et al., 2010; Ryan, Narrow, Federoff, & Bowers, 2010). Inspired by the findings from Aβ immunization therapy, more studies start to develop anti-αSyn immunotherapy, but they are less advanced. It has been reported that αSyn transgenic mice actively immunized with recombinant human αSyn successfully produced antibodies against αSyn, and neurodegeneration was partially reversed with decreasing αSyn deposition in the neuronal cell bodies and synapses (Masliah et al., 2005). A following study from the same group updated that they generated a novel monoclonal αSyn antibody against its C-terminus region, which alleviates behavioral deficits in the water maze, reduces αSyn accumulation in the axons and synapses, and decrease neurodegeneration in the same αSyn transgenic mice (Masliah et al., 2011). At first, the mechanisms underlying αSyn-based immunotherapy are not clarified, as αSyn was originally thought as a cytoplasmic protein. With more understanding about the secretion of αSyn, the picture becomes clear to the researchers (Borghi et al., 2000; El-Agnaf et al., 2003; Lee, Patel, & Lee, 2005). Extracellular αSyn is suggested to be internalized and degraded in microglial (Lee, Suk, Bae, & Lee, 2008). Therefore, antibody administration is beneficial to increase glial-assisted clearance of extracellular αSyn, and thus reduces its toxicity (Bae et al., 2012). However, there is no available evidence of testing the effects of antibodies against αSyn oligomer in vivo. 164 In the search for effective and efficacious oligomer-base therapies for AD and PD, several issues should be taken into consideration. Firstly, oligomer is a concept referring to heterogenous species with diverse properties due to various preparation protocols. Moreover, not all oligomers are toxic, which can exemplified by the ‘off-pathway’ oligomers resistant to aggregation (Ehrnhoefer et al., 2008). Therefore, in the studies trying to develop oligomer-specific antibodies, it is of essential importance to clearly define the identities of the toxic oligomers at the first place. Secondly, the oligomers generated in vitro is hard to recapitulate the comprehensive features of oligomers existed physiologically, especially when considering the quite high concentration used in in vitro aggregation assays. Therefore, the antibodies generated against such non-physiological oligomeric species may not work properly after immunizing patients, although humanized antibodies could provide more promising results. Finally, the ability of oligomer-specific antibodies to penetrate brain-blood barrier is another challenge. 165 Bibliography Aasly, J. O., Toft, M., Fernandez-Mata, I., Kachergus, J., Hulihan, M., White, L. R., & Farrer, M. (2005). Clinical features of LRRK2-associated Parkinson's disease in central Norway. Ann Neurol, 57(5), 762-765. doi: 10.1002/ana.20456 Albertini, V., Jain, A., Vignati, S., Napoli, S., Rinaldi, A., Kwee, I., . . . Catapano, C. V. (2006). Novel GC-rich DNA-binding compound produced by a genetically engineered mutant of the mithramycin producer Streptomyces argillaceus exhibits improved transcriptional repressor activity: implications for cancer therapy. Nucleic Acids Res, 34(6), 1721-1734. doi: 10.1093/nar/gkl063 Anthone, W. Dunah, Hyunkyung, Jeong, April, Griffin, Yong-Man, Kim, David, G. Standaert, Steven, M. Hersch, . . . Dimitri, Krainc. (2002). Sp1 and TAFII130 Transcriptional Activity Disrupted in Early Huntington's Disease. Science. doi: 10.1126/science.1072613 Asai, M., Hattori, C., Szabo, B., Sasagawa, N., Maruyama, K., Tanuma, S., & Ishiura, S. (2003). Putative function of ADAM9, ADAM10, and ADAM17 as APP alpha-secretase. Biochem Biophys Res Commun, 301(1), 231-235. Auluck, P. K., Chan, H. Y., Trojanowski, J. Q., Lee, V. M., & Bonini, N. M. (2002). Chaperone suppression of alpha-synuclein toxicity in a Drosophila model for Parkinson's disease. Science, 295(5556), 865-868. doi: 10.1126/science.1067389 Bae, E. J., Lee, H. J., Rockenstein, E., Ho, D. H., Park, E. B., Yang, N. Y., . . . Lee, S. J. (2012). Antibody-aided clearance of extracellular alpha-synuclein prevents cell-to-cell aggregate transmission. J Neurosci, 32(39), 13454-13469. doi: 10.1523/JNEUROSCI.1292-12.2012 Baptista, Melisa, O'Farrell, Casey, Daya, Sneha, Ahmad, Rili, Miller, David, Hardy, John, . . . Cookson, Mark. (2003). Co-ordinate transcriptional regulation of dopamine synthesis genes by alpha-synuclein in human neuroblastoma cell lines. Journal of neurochemistry, 85(4), 957-968. doi: 10.1046/j.1471-4159.2003.01742.x Bartus, R. T., Dean, R. L., Pontecorvo, M. J., & Flicker, C. (1985). The cholinergic hypothesis: a historical overview, current perspective, and future directions. Ann N Y Acad Sci, 444, 332-358. Bekris, L. M., Yu, C. E., Bird, T. D., & Tsuang, D. W. (2010). Genetics of Alzheimer disease. J Geriatr Psychiatry Neurol, 23(4), 213-227. doi: 10.1177/0891988710383571 Bellucci, A., Zaltieri, M., Navarria, L., Grigoletto, J., Missale, C., & Spano, P. (2012). From alpha-synuclein to synaptic dysfunctions: new insights into the pathophysiology of Parkinson's disease. Brain Res, 1476, 183-202. doi: 10.1016/j.brainres.2012.04.014 Benilova, I., Karran, E., & De Strooper, B. (2012). The toxic Abeta oligomer and Alzheimer's disease: an emperor in need of clothes. Nat Neurosci, 15(3), 349-357. doi: 10.1038/nn.3028 Bentahir, M., Nyabi, O., Verhamme, J., Tolia, A., Horre, K., Wiltfang, J., . . . De Strooper, B. (2006). Presenilin clinical mutations can affect gamma-secretase activity by different mechanisms. J Neurochem, 96(3), 732-742. doi: 10.1111/j.1471-4159.2005.03578.x Berg, D., Schweitzer, K. J., Leitner, P., Zimprich, A., Lichtner, P., Belcredi, P., . . . Gasser, T. (2005). Type and frequency of mutations in the LRRK2 gene in familial and sporadic Parkinson's disease*. Brain, 128(Pt 12), 3000-3011. doi: 10.1093/brain/awh666 Bernfield, M., Kokenyesi, R., Kato, M., Hinkes, M. T., Spring, J., Gallo, R. L., & Lose, E. J. (1992). Biology of the syndecans: a family of transmembrane heparan sulfate 166 proteoglycans. Annu Rev Cell Biol, 8, 365-393. doi: 10.1146/annurev.cb.08.110192.002053 Biskup, S., Moore, D. J., Celsi, F., Higashi, S., West, A. B., Andrabi, S. A., . . . Dawson, V. L. (2006). Localization of LRRK2 to membranous and vesicular structures in mammalian brain. Ann Neurol, 60(5), 557-569. doi: 10.1002/ana.21019 Black, A. R., Black, J. D., & Azizkhan-Clifford, J. (2001). Sp1 and kruppel-like factor family of transcription factors in cell growth regulation and cancer. J Cell Physiol, 188(2), 143-160. doi: 10.1002/jcp.1111 Black, A. R., Jensen, D., Lin, S. Y., & Azizkhan, J. C. (1999). Growth/cell cycle regulation of Sp1 phosphorylation. J Biol Chem, 274(3), 1207-1215. Borghi, R., Marchese, R., Negro, A., Marinelli, L., Forloni, G., Zaccheo, D., . . . Tabaton, M. (2000). Full length alpha-synuclein is present in cerebrospinal fluid from Parkinson's disease and normal subjects. Neurosci Lett, 287(1), 65-67. Bossers, K., Meerhoff, G., Balesar, R., van Dongen, J. W., Kruse, C. G., Swaab, D. F., & Verhaagen, J. (2009). Analysis of gene expression in Parkinson's disease: possible involvement of neurotrophic support and axon guidance in dopaminergic cell death. Brain Pathol, 19(1), 91-107. doi: 10.1111/j.1750-3639.2008.00171.x Brandeis, M., Frank, D., Keshet, I., Siegfried, Z., Mendelsohn, M., Nemes, A., . . . Cedar, H. (1994). Sp1 elements protect a CpG island from de novo methylation. Nature, 371(6496), 435-438. doi: 10.1038/371435a0 Brion, J. P., Couck, A. M., Passareiro, E., & Flament-Durand, J. (1985). Neurofibrillary tangles of Alzheimer's disease: an immunohistochemical study. J Submicrosc Cytol, 17(1), 89-96. Bruce, A. Citron, John, S. Dennis, Ross, S. Zeitlin, & Valentina, Echeverria. (2008). Transcription factor Sp1 dysregulation in Alzheimer's disease. Journal of neuroscience research. doi: 10.1002/jnr.21695 Burdick, D., Soreghan, B., Kwon, M., Kosmoski, J., Knauer, M., Henschen, A., . . . Glabe, C. (1992). Assembly and aggregation properties of synthetic Alzheimer's A4/beta amyloid peptide analogs. J Biol Chem, 267(1), 546-554. Bushdid, P. B., Chen, C. L., Brantley, D. M., Yull, F., Raghow, R., Kerr, L. D., & Barnett, J. V. (2001). NF-kappaB mediates FGF signal regulation of msx-1 expression. Dev Biol, 237(1), 107-115. doi: 10.1006/dbio.2001.0356 Cai, F., Chen, B., Zhou, W., Zis, O., Liu, S., Holt, R. A., . . . Song, W. (2008). SP1 regulates a human SNAP-25 gene expression. J Neurochem, 105(2), 512-523. doi: 10.1111/j.1471-4159.2007.05167.x Cai, X., Golde, T., & Younkin, S. (1993). Release of excess amyloid beta protein from a mutant amyloid beta protein precursor. Science (New York, N.Y.), 259(5094), 514-516. doi: 10.1126/science.8424174 Calhoun, M., Wiederhold, K., Abramowski, D., Phinney, A., Probst, A., Sturchler-Pierrat, C., . . . Jucker, M. (1998). Neuron loss in APP transgenic mice. Nature, 395(6704), 755-756. doi: 10.1038/27351 Cantuti-Castelvetri, I., Keller-McGandy, C., Bouzou, B., Asteris, G., Clark, T. W., Frosch, M. P., & Standaert, D. G. (2007). Effects of gender on nigral gene expression and parkinson disease. Neurobiol Dis, 26(3), 606-614. doi: 10.1016/j.nbd.2007.02.009 167 Cappai, Roberto, Leck, Su-Ling, Tew, Deborah, Williamson, Nicholas, Smith, David, Galatis, Denise, . . . Hill, Andrew. (2005). Dopamine promotes alpha-synuclein aggregation into SDS-resistant soluble oligomers via a distinct folding pathway. FASEB journal : official publication of the Federation of American Societies for Experimental Biology, 19(10), 1377-1379. doi: 10.1096/fj.04-3437fje Carey, D. J., Conner, K., Asundi, V. K., O'Mahony, D. J., Stahl, R. C., Showalter, L., . . . Rothblum, L. I. (1997). cDNA cloning, genomic organization, and in vivo expression of rat N-syndecan. J Biol Chem, 272(5), 2873-2879. Carey, D. J., Evans, D. M., Stahl, R. C., Asundi, V. K., Conner, K. J., Garbes, P., & Cizmeci-Smith, G. (1992). Molecular cloning and characterization of N-syndecan, a novel transmembrane heparan sulfate proteoglycan. J Cell Biol, 117(1), 191-201. Catela, C., Bilbao-Cortes, D., Slonimsky, E., Kratsios, P., Rosenthal, N., & Te Welscher, P. (2009). Multiple congenital malformations of Wolf-Hirschhorn syndrome are recapitulated in Fgfrl1 null mice. Dis Model Mech, 2(5-6), 283-294. doi: 10.1242/dmm.002287 Chen, C. H., Zhou, W., Liu, S., Deng, Y., Cai, F., Tone, M., . . . Song, W. (2012). Increased NF-kappaB signalling up-regulates BACE1 expression and its therapeutic potential in Alzheimer's disease. Int J Neuropsychopharmacol, 15(1), 77-90. doi: 10.1017/S1461145711000149 Cheng, I. H., Scearce-Levie, K., Legleiter, J., Palop, J. J., Gerstein, H., Bien-Ly, N., . . . Mucke, L. (2007). Accelerating amyloid-beta fibrillization reduces oligomer levels and functional deficits in Alzheimer disease mouse models. J Biol Chem, 282(33), 23818-23828. doi: 10.1074/jbc.M701078200 Choi, H. K., Won, L. A., Kontur, P. J., Hammond, D. N., Fox, A. P., Wainer, B. H., . . . Heller, A. (1991). Immortalization of embryonic mesencephalic dopaminergic neurons by somatic cell fusion. Brain Res, 552(1), 67-76. Choi, J. H., Kaur, G., Mazzella, M. J., Morales-Corraliza, J., Levy, E., & Mathews, P. M. (2013). Early endosomal abnormalities and cholinergic neuron degeneration in amyloid-beta protein precursor transgenic mice. J Alzheimers Dis, 34(3), 691-700. doi: 10.3233/JAD-122143 Christensen, M. A., Zhou, W., Qing, H., Lehman, A., Philipsen, S., & Song, W. (2004). Transcriptional regulation of BACE1, the beta-amyloid precursor protein beta-secretase, by Sp1. Mol Cell Biol, 24(2), 865-874. Chung, C. Y., Seo, H., Sonntag, K. C., Brooks, A., Lin, L., & Isacson, O. (2005). Cell type-specific gene expression of midbrain dopaminergic neurons reveals molecules involved in their vulnerability and protection. Hum Mol Genet, 14(13), 1709-1725. doi: 10.1093/hmg/ddi178 Citron, M., Oltersdorf, T., Haass, C., McConlogue, L., Hung, A. Y., Seubert, P., . . . Selkoe, D. J. (1992). Mutation of the beta-amyloid precursor protein in familial Alzheimer's disease increases beta-protein production. Nature, 360(6405), 672-674. doi: 10.1038/360672a0 Citron, M., Vigo-Pelfrey, C., Teplow, D. B., Miller, C., Schenk, D., Johnston, J., . . . Selkoe, D. J. (1994). Excessive production of amyloid beta-protein by peripheral cells of symptomatic and presymptomatic patients carrying the Swedish familial Alzheimer disease mutation. Proc Natl Acad Sci U S A, 91(25), 11993-11997. 168 Cleary, James, Walsh, Dominic, Hofmeister, Jacki, Shankar, Ganesh, Kuskowski, Michael, Selkoe, Dennis, & Ashe, Karen. (2005). Natural oligomers of the amyloid-beta protein specifically disrupt cognitive function. Nature neuroscience, 8(1), 79-84. doi: 10.1038/nn1372 Connolly, B. S., & Lang, A. E. (2014). Pharmacological treatment of Parkinson disease: a review. JAMA, 311(16), 1670-1683. doi: 10.1001/jama.2014.3654 Consortium, Fantom, the, Riken Pmi, Clst, Forrest, A. R., Kawaji, H., Rehli, M., . . . Hayashizaki, Y. (2014). A promoter-level mammalian expression atlas. Nature, 507(7493), 462-470. doi: 10.1038/nature13182 Conway, K. A., Harper, J. D., & Lansbury, P. T. (1998). Accelerated in vitro fibril formation by a mutant alpha-synuclein linked to early-onset Parkinson disease. Nat Med, 4(11), 1318-1320. doi: 10.1038/3311 Conway, K., Lee, S., Rochet, J., Ding, T., Williamson, R., & Lansbury, P. (2000). Acceleration of oligomerization,not fibrillization, is a shared property of both alpha-synuclein mutations linked to early-onset Parkinson's disease: implications for pathogenesis and therapy. Proceedings of the National Academy of Sciences of the United States of America, 97(2), 571-576. doi: 10.1073/pnas.97.2.571 Conway, K., Rochet, J., Bieganski, R., & Lansbury, P. (2001). Kinetic stabilization of the alpha-synuclein protofibril by a dopamine-alpha-synuclein adduct. Science (New York, N.Y.), 294(5545), 1346-1349. doi: 10.1126/science.1063522 Cookson, M. R., Hardy, J., & Lewis, P. A. (2008). Genetic neuropathology of Parkinson's disease. Int J Clin Exp Pathol, 1(3), 217-231. Cooper-Knock, J., Kirby, J., Ferraiuolo, L., Heath, P. R., Rattray, M., & Shaw, P. J. (2012). Gene expression profiling in human neurodegenerative disease. Nat Rev Neurol, 8(9), 518-530. doi: 10.1038/nrneurol.2012.156 Corder, E. H., Saunders, A. M., Strittmatter, W. J., Schmechel, D. E., Gaskell, P. C., Small, G. W., . . . Pericak-Vance, M. A. (1993). Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset families. Science, 261(5123), 921-923. Counts, S. E., Che, S., Ginsberg, S. D., & Mufson, E. J. (2011). Gender differences in neurotrophin and glutamate receptor expression in cholinergic nucleus basalis neurons during the progression of Alzheimer's disease. J Chem Neuroanat, 42(2), 111-117. doi: 10.1016/j.jchemneu.2011.02.004 Counts, S. E., He, B., Che, S., Ginsberg, S. D., & Mufson, E. J. (2009). Galanin fiber hyperinnervation preserves neuroprotective gene expression in cholinergic basal forebrain neurons in Alzheimer's disease. J Alzheimers Dis, 18(4), 885-896. doi: 10.3233/JAD-2009-1196 Counts, S. E., He, B., Che, S., Ikonomovic, M. D., DeKosky, S. T., Ginsberg, S. D., & Mufson, E. J. (2007). Alpha7 nicotinic receptor up-regulation in cholinergic basal forebrain neurons in Alzheimer disease. Arch Neurol, 64(12), 1771-1776. doi: 10.1001/archneur.64.12.1771 Daher, J. P., Pletnikova, O., Biskup, S., Musso, A., Gellhaar, S., Galter, D., . . . Moore, D. J. (2012). Neurodegenerative phenotypes in an A53T alpha-synuclein transgenic mouse model are independent of LRRK2. Hum Mol Genet, 21(11), 2420-2431. doi: 10.1093/hmg/dds057 169 Daniels, V., Vancraenenbroeck, R., Law, B. M., Greggio, E., Lobbestael, E., Gao, F., . . . Taymans, J. M. (2011). Insight into the mode of action of the LRRK2 Y1699C pathogenic mutant. J Neurochem, 116(2), 304-315. doi: 10.1111/j.1471-4159.2010.07105.x Danzer, K. M., Haasen, D., Karow, A. R., Moussaud, S., Habeck, M., Giese, A., . . . Kostka, M. (2007). Different species of alpha-synuclein oligomers induce calcium influx and seeding. J Neurosci, 27(34), 9220-9232. doi: 10.1523/JNEUROSCI.2617-07.2007 Danzer, K. M., Kranich, L. R., Ruf, W. P., Cagsal-Getkin, O., Winslow, A. R., Zhu, L., . . . McLean, P. J. (2012). Exosomal cell-to-cell transmission of alpha synuclein oligomers. Mol Neurodegener, 7, 42. doi: 10.1186/1750-1326-7-42 Danzer, K. M., Krebs, S. K., Wolff, M., Birk, G., & Hengerer, B. (2009). Seeding induced by alpha-synuclein oligomers provides evidence for spreading of alpha-synuclein pathology. J Neurochem, 111(1), 192-203. doi: 10.1111/j.1471-4159.2009.06324.x Danzer, K. M., Ruf, W. P., Putcha, P., Joyner, D., Hashimoto, T., Glabe, C., . . . McLean, P. J. (2011). Heat-shock protein 70 modulates toxic extracellular alpha-synuclein oligomers and rescues trans-synaptic toxicity. FASEB J, 25(1), 326-336. doi: 10.1096/fj.10-164624 De Souza e Silva, N. A., & Wilson, D. M. (1973). Hypercalcemia associated with chronic lymphocytic leukemia. Treatment with mithramycin. Postgrad Med, 53(7), 129-134. Decker, H., Jurgensen, S., Adrover, M. F., Brito-Moreira, J., Bomfim, T. R., Klein, W. L., . . . Ferreira, S. T. (2010). N-methyl-D-aspartate receptors are required for synaptic targeting of Alzheimer's toxic amyloid-beta peptide oligomers. J Neurochem, 115(6), 1520-1529. doi: 10.1111/j.1471-4159.2010.07058.x Deng, Yu, Wang, Zhe, Wang, Ruitao, Zhang, Xiaozhu, Zhang, Shuting, Wu, Yili, . . . Song, Weihong. (2013). Amyloid-β protein (Aβ) Glu11 is the major β-secretase site of β-site amyloid-β precursor protein-cleaving enzyme 1(BACE1), and shifting the cleavage site to Aβ Asp1 contributes to Alzheimer pathogenesis. The European journal of neuroscience, 37(12), 1962-1969. doi: 10.1111/ejn.12235 Deshpande, A., Mina, E., Glabe, C., & Busciglio, J. (2006). Different conformations of amyloid beta induce neurotoxicity by distinct mechanisms in human cortical neurons. J Neurosci, 26(22), 6011-6018. doi: 10.1523/JNEUROSCI.1189-06.2006 Dineley, K. T., Westerman, M., Bui, D., Bell, K., Ashe, K. H., & Sweatt, J. D. (2001). Beta-amyloid activates the mitogen-activated protein kinase cascade via hippocampal alpha7 nicotinic acetylcholine receptors: In vitro and in vivo mechanisms related to Alzheimer's disease. J Neurosci, 21(12), 4125-4133. Diogenes, M. J., Dias, R. B., Rombo, D. M., Vicente Miranda, H., Maiolino, F., Guerreiro, P., . . . Outeiro, T. F. (2012). Extracellular alpha-synuclein oligomers modulate synaptic transmission and impair LTP via NMDA-receptor activation. J Neurosci, 32(34), 11750-11762. doi: 10.1523/JNEUROSCI.0234-12.2012 Doody, R. S., Thomas, R. G., Farlow, M., Iwatsubo, T., Vellas, B., Joffe, S., . . . Solanezumab Study, Group. (2014). Phase 3 trials of solanezumab for mild-to-moderate Alzheimer's disease. N Engl J Med, 370(4), 311-321. doi: 10.1056/NEJMoa1312889 Drafahl, K. A., McAndrew, C. W., Meyer, A. N., Haas, M., & Donoghue, D. J. (2010). The receptor tyrosine kinase FGFR4 negatively regulates NF-kappaB signaling. PLoS One, 5(12), e14412. doi: 10.1371/journal.pone.0014412 170 Dubois, B., Feldman, H. H., Jacova, C., Dekosky, S. T., Barberger-Gateau, P., Cummings, J., . . . Scheltens, P. (2007). Research criteria for the diagnosis of Alzheimer's disease: revising the NINCDS-ADRDA criteria. Lancet Neurol, 6(8), 734-746. doi: 10.1016/S1474-4422(07)70178-3 Dynan, W. S., & Tjian, R. (1983). The promoter-specific transcription factor Sp1 binds to upstream sequences in the SV40 early promoter. Cell, 35(1), 79-87. Eckert, A., Steiner, B., Marques, C., Leutz, S., Romig, H., Haass, C., & Müller, W. (2001). Elevated vulnerability to oxidative stress-induced cell death and activation of caspase-3 by the Swedish amyloid precursor protein mutation. Journal of neuroscience research, 64(2), 183-192. doi: 10.1002/jnr.1064 Eckert, A., Steiner, B., Marques, C., Leutz, S., Romig, H., Haass, C., & Muller, W. E. (2001). Elevated vulnerability to oxidative stress-induced cell death and activation of caspase-3 by the Swedish amyloid precursor protein mutation. J Neurosci Res, 64(2), 183-192. doi: 10.1002/jnr.1064 Ehrnhoefer, D. E., Bieschke, J., Boeddrich, A., Herbst, M., Masino, L., Lurz, R., . . . Wanker, E. E. (2008). EGCG redirects amyloidogenic polypeptides into unstructured, off-pathway oligomers. Nat Struct Mol Biol, 15(6), 558-566. doi: 10.1038/nsmb.1437 El-Agnaf, O. M., Salem, S. A., Paleologou, K. E., Cooper, L. J., Fullwood, N. J., Gibson, M. J., . . . Allsop, D. (2003). Alpha-synuclein implicated in Parkinson's disease is present in extracellular biological fluids, including human plasma. FASEB J, 17(13), 1945-1947. doi: 10.1096/fj.03-0098fje El-Agnaf, Omar, Salem, Sultan, Paleologou, Katerina, Curran, Martin, Gibson, Mark, Court, Jennifer, . . . Allsop, David. (2006). Detection of oligomeric forms of alpha-synuclein protein in human plasma as a potential biomarker for Parkinson's disease. FASEB journal : official publication of the Federation of American Societies for Experimental Biology, 20(3), 419-425. doi: 10.1096/fj.03-1449com Elstner, M., Morris, C. M., Heim, K., Bender, A., Mehta, D., Jaros, E., . . . Prokisch, H. (2011). Expression analysis of dopaminergic neurons in Parkinson's disease and aging links transcriptional dysregulation of energy metabolism to cell death. Acta Neuropathol, 122(1), 75-86. doi: 10.1007/s00401-011-0828-9 Emilsson, L., Saetre, P., & Jazin, E. (2006). Alzheimer's disease: mRNA expression profiles of multiple patients show alterations of genes involved with calcium signaling. Neurobiol Dis, 21(3), 618-625. doi: 10.1016/j.nbd.2005.09.004 Emmanouilidou, Evangelia, Melachroinou, Katerina, Roumeliotis, Theodoros, Garbis, Spiros, Ntzouni, Maria, Margaritis, Lukas, . . . Vekrellis, Kostas. (2010). Cell-produced alpha-synuclein is secreted in a calcium-dependent manner by exosomes and impacts neuronal survival. The Journal of neuroscience : the official journal of the Society for Neuroscience, 30(20), 6838-6851. doi: 10.1523/JNEUROSCI.5699-09.2010 Fa, Mauro, Orozco, Ian, Francis, Yitshak, Saeed, Faisal, Gong, Yimin, & Arancio, Ottavio. (2010). Preparation of oligomeric beta-amyloid 1-42 and induction of synaptic plasticity impairment on hippocampal slices. Journal of visualized experiments : JoVE(41). doi: 10.3791/1884 Fearnley, J. M., & Lees, A. J. (1991). Ageing and Parkinson's disease: substantia nigra regional selectivity. Brain, 114 ( Pt 5), 2283-2301. 171 Ferrante, R. J., Ryu, H., Kubilus, J. K., D'Mello, S., Sugars, K. L., Lee, J., . . . Ratan, R. R. (2004). Chemotherapy for the brain: the antitumor antibiotic mithramycin prolongs survival in a mouse model of Huntington's disease. J Neurosci, 24(46), 10335-10342. doi: 10.1523/JNEUROSCI.2599-04.2004 Fodero, L. R., Mok, S. S., Losic, D., Martin, L. L., Aguilar, M. I., Barrow, C. J., . . . Small, D. H. (2004). Alpha7-nicotinic acetylcholine receptors mediate an Abeta(1-42)-induced increase in the level of acetylcholinesterase in primary cortical neurones. J Neurochem, 88(5), 1186-1193. Follmer, C., Coelho-Cerqueira, E., Yatabe-Franco, D. Y., Araujo, G. D., Pinheiro, A. S., Domont, G. B., & Eliezer, D. (2015). Oligomerization and Membrane-binding Properties of Covalent Adducts Formed by the Interaction of alpha-Synuclein with the Toxic Dopamine Metabolite 3,4-Dihydroxyphenylacetaldehyde (DOPAL). J Biol Chem, 290(46), 27660-27679. doi: 10.1074/jbc.M115.686584 Forno, L. S. (1996). Neuropathology of Parkinson's disease. J Neuropathol Exp Neurol, 55(3), 259-272. Franklin, Keith; Paxinos, George (2008). The Mouse Brain in Stereotaxic Coordinates, Compact 3rd Edition. Funayama, M., Hasegawa, K., Kowa, H., Saito, M., Tsuji, S., & Obata, F. (2002). A new locus for Parkinson's disease (PARK8) maps to chromosome 12p11.2-q13.1. Ann Neurol, 51(3), 296-301. Galter, D., Westerlund, M., Carmine, A., Lindqvist, E., Sydow, O., & Olson, L. (2006). LRRK2 expression linked to dopamine-innervated areas. Ann Neurol, 59(4), 714-719. doi: 10.1002/ana.20808 Galvin, J. E. (2006). Interaction of alpha-synuclein and dopamine metabolites in the pathogenesis of Parkinson's disease: a case for the selective vulnerability of the substantia nigra. Acta Neuropathol, 112(2), 115-126. doi: 10.1007/s00401-006-0096-2 Gao, N., Li, Y. H., Li, X., Yu, S., Fu, G. L., & Chen, B. (2007). Effect of alpha-synuclein on the promoter activity of tyrosine hydroxylase gene. Neurosci Bull, 23(1), 53-57. Gasser, T. (2009). Molecular pathogenesis of Parkinson disease: insights from genetic studies. Expert Rev Mol Med, 11, e22. doi: 10.1017/S1462399409001148 Gerber, S. D., Steinberg, F., Beyeler, M., Villiger, P. M., & Trueb, B. (2009). The murine Fgfrl1 receptor is essential for the development of the metanephric kidney. Dev Biol, 335(1), 106-119. doi: 10.1016/j.ydbio.2009.08.019 Giasson, B. I., Duda, J. E., Quinn, S. M., Zhang, B., Trojanowski, J. Q., & Lee, V. M. (2002). Neuronal alpha-synucleinopathy with severe movement disorder in mice expressing A53T human alpha-synuclein. Neuron, 34(4), 521-533. Giasson, B., Uryu, K., Trojanowski, J., & Lee, V. (1999). Mutant and wild type human alpha-synucleins assemble into elongated filaments with distinct morphologies in vitro. The Journal of biological chemistry, 274(12), 7619-7622. doi: 10.1074/jbc.274.12.7619 Gidoni, D., Kadonaga, J. T., Barrera-Saldana, H., Takahashi, K., Chambon, P., & Tjian, R. (1985). Bidirectional SV40 transcription mediated by tandem Sp1 binding interactions. Science, 230(4725), 511-517. Giglioni, B., Comi, P., Ronchi, A., Mantovani, R., & Ottolenghi, S. (1989). The same nuclear proteins bind the proximal CACCC box of the human beta-globin promoter and a similar sequence in the enhancer. Biochem Biophys Res Commun, 164(1), 149-155. 172 Gilman, S., Koller, M., Black, R. S., Jenkins, L., Griffith, S. G., Fox, N. C., . . . Team, A. N. Study. (2005). Clinical effects of Abeta immunization (AN1792) in patients with AD in an interrupted trial. Neurology, 64(9), 1553-1562. doi: 10.1212/01.WNL.0000159740.16984.3C Ginsberg, S. D., Alldred, M. J., & Che, S. (2012). Gene expression levels assessed by CA1 pyramidal neuron and regional hippocampal dissections in Alzheimer's disease. Neurobiol Dis, 45(1), 99-107. doi: 10.1016/j.nbd.2011.07.013 Ginsberg, S. D., Alldred, M. J., Counts, S. E., Cataldo, A. M., Neve, R. L., Jiang, Y., . . . Che, S. (2010). Microarray analysis of hippocampal CA1 neurons implicates early endosomal dysfunction during Alzheimer's disease progression. Biol Psychiatry, 68(10), 885-893. doi: 10.1016/j.biopsych.2010.05.030 Ginsberg, S., Hemby, S., Lee, V., Eberwine, J., & Trojanowski, J. (2000). Expression profile of transcripts in Alzheimer's disease tangle-bearing CA1 neurons. Annals of neurology, 48(1), 77-87. doi: 10.1002/1531-8249(200007)48:1<77::AID-ANA12>3.0.CO;2-A Gispert, S., Del Turco, D., Garrett, L., Chen, A., Bernard, D. J., Hamm-Clement, J., . . . Nussbaum, R. L. (2003). Transgenic mice expressing mutant A53T human alpha-synuclein show neuronal dysfunction in the absence of aggregate formation. Mol Cell Neurosci, 24(2), 419-429. Glenner, G. G., & Wong, C. W. (1984a). Alzheimer's disease and Down's syndrome: sharing of a unique cerebrovascular amyloid fibril protein. Biochem Biophys Res Commun, 122(3), 1131-1135. Glenner, G. G., & Wong, C. W. (1984b). Alzheimer's disease: initial report of the purification and characterization of a novel cerebrovascular amyloid protein. Biochem Biophys Res Commun, 120(3), 885-890. Goldgaber, D., Lerman, M. I., McBride, O. W., Saffiotti, U., & Gajdusek, D. C. (1987). Characterization and chromosomal localization of a cDNA encoding brain amyloid of Alzheimer's disease. Science, 235(4791), 877-880. Gómez-Suaga, P., Fdez, E., Fernández, B., Martínez-Salvador, M., Blanca Ramírez, M., Madero-Pérez, J., . . . Hilfiker, S. (2014). Novel insights into the neurobiology underlying LRRK2-linked Parkinson's disease. Neuropharmacology. doi: 10.1016/j.neuropharm.2014.05.020 Gong, Y., Chang, L., Viola, K. L., Lacor, P. N., Lambert, M. P., Finch, C. E., . . . Klein, W. L. (2003). Alzheimer's disease-affected brain: presence of oligomeric A beta ligands (ADDLs) suggests a molecular basis for reversible memory loss. Proc Natl Acad Sci U S A, 100(18), 10417-10422. doi: 10.1073/pnas.1834302100 Gorbatyuk, O. S., Li, S., Nash, K., Gorbatyuk, M., Lewin, A. S., Sullivan, L. F., . . . Muzyczka, N. (2010). In vivo RNAi-mediated alpha-synuclein silencing induces nigrostriatal degeneration. Mol Ther, 18(8), 1450-1457. doi: 10.1038/mt.2010.115 Gorevic, P. D., Goni, F., Pons-Estel, B., Alvarez, F., Peress, N. S., & Frangione, B. (1986). Isolation and partial characterization of neurofibrillary tangles and amyloid plaque core in Alzheimer's disease: immunohistological studies. J Neuropathol Exp Neurol, 45(6), 647-664. Gosavi, N., Lee, H. J., Lee, J. S., Patel, S., & Lee, S. J. (2002). Golgi fragmentation occurs in the cells with prefibrillar alpha-synuclein aggregates and precedes the formation of fibrillar inclusion. J Biol Chem, 277(50), 48984-48992. doi: 10.1074/jbc.M208194200 173 Graham, W. V., Bonito-Oliva, A., & Sakmar, T. P. (2017). Update on Alzheimer's Disease Therapy and Prevention Strategies. Annu Rev Med, 68, 413-430. doi: 10.1146/annurev-med-042915-103753 Grassi, F., Palma, E., Tonini, R., Amici, M., Ballivet, M., & Eusebi, F. (2003). Amyloid beta(1-42) peptide alters the gating of human and mouse alpha-bungarotoxin-sensitive nicotinic receptors. J Physiol, 547(Pt 1), 147-157. doi: 10.1113/jphysiol.2002.035436 Greene, James. (2012). Current status and future directions of gene expression profiling in Parkinson's disease. Neurobiology of disease, 45(1), 76-82. doi: 10.1016/j.nbd.2010.10.022 Greffard, S., Verny, M., Bonnet, A. M., Beinis, J. Y., Gallinari, C., Meaume, S., . . . Duyckaerts, C. (2006). Motor score of the Unified Parkinson Disease Rating Scale as a good predictor of Lewy body-associated neuronal loss in the substantia nigra. Arch Neurol, 63(4), 584-588. doi: 10.1001/archneur.63.4.584 Greggio, E., & Cookson, M. R. (2009). Leucine-rich repeat kinase 2 mutations and Parkinson's disease: three questions. ASN Neuro, 1(1). doi: 10.1042/AN20090007 Greggio, E., Jain, S., Kingsbury, A., Bandopadhyay, R., Lewis, P., Kaganovich, A., . . . Cookson, M. R. (2006). Kinase activity is required for the toxic effects of mutant LRRK2/dardarin. Neurobiol Dis, 23(2), 329-341. doi: 10.1016/j.nbd.2006.04.001 Grünblatt, E., Mandel, S., Jacob-Hirsch, J., Zeligson, S., Amariglo, N., Rechavi, G., . . . Youdim, M. B. H. (2004). Gene expression profiling of parkinsonian substantia nigra pars compacta; alterations in ubiquitin-proteasome, heat shock protein, iron and oxidative stress regulated proteins, cell adhesion/cellular matrix and vesicle trafficking genes. Journal of Neural Transmission, 111, in doi: 10.1007/s00702-004-0212-1 Guo, L., Gandhi, P. N., Wang, W., Petersen, R. B., Wilson-Delfosse, A. L., & Chen, S. G. (2007). The Parkinson's disease-associated protein, leucine-rich repeat kinase 2 (LRRK2), is an authentic GTPase that stimulates kinase activity. Exp Cell Res, 313(16), 3658-3670. doi: 10.1016/j.yexcr.2007.07.007 Habig, K., Walter, M., Poths, S., Riess, O., & Bonin, M. (2008). RNA interference of LRRK2-microarray expression analysis of a Parkinson's disease key player. Neurogenetics, 9(2), 83-94. doi: 10.1007/s10048-007-0114-0 Hagiwara, H., Iyo, M., & Hashimoto, K. (2009). Mithramycin protects against dopaminergic neurotoxicity in the mouse brain after administration of methamphetamine. Brain Res, 1301, 189-196. doi: 10.1016/j.brainres.2009.09.010 Hammond, D. N., Wainer, B. H., Tonsgard, J. H., & Heller, A. (1986). Neuronal properties of clonal hybrid cell lines derived from central cholinergic neurons. Science, 234(4781), 1237-1240. Hardy, John, & Selkoe, Dennis. (2002). The amyloid hypothesis of Alzheimer's disease: progress and problems on the road to therapeutics. Science (New York, N.Y.), 297(5580), 353-356. doi: 10.1126/science.1072994 Haroutunian, V., Katsel, P., & Schmeidler, J. (2009). Transcriptional vulnerability of brain regions in Alzheimer's disease and dementia. Neurobiol Aging, 30(4), 561-573. doi: 10.1016/j.neurobiolaging.2007.07.021 Hartley, D. M., Walsh, D. M., Ye, C. P., Diehl, T., Vasquez, S., Vassilev, P. M., . . . Selkoe, D. J. (1999). Protofibrillar intermediates of amyloid beta-protein induce acute 174 electrophysiological changes and progressive neurotoxicity in cortical neurons. J Neurosci, 19(20), 8876-8884. Hasegawa, Takafumi, Matsuzaki, Michiko, Takeda, Atsushi, Kikuchi, Akio, Akita, Hirotoshi, Perry, George, . . . Itoyama, Yasuto. (2004). Accelerated alpha-synuclein aggregation after differentiation of SH-SY5Y neuroblastoma cells. Brain research, 1013(1), 51-59. doi: 10.1016/j.brainres.2004.04.018 Hashimoto, M., Kawahara, K., Bar-On, P., Rockenstein, E., Crews, L., & Masliah, E. (2004). The Role of alpha-synuclein assembly and metabolism in the pathogenesis of Lewy body disease. J Mol Neurosci, 24(3), 343-352. doi: 10.1385/JMN:24:3:343 Hauser, Michael, Li, Yi-Ju, Xu, Hong, Noureddine, Maher, Shao, Yujun, Gullans, Steven, . . . Vance, Jeffery. (2005). Expression profiling of substantia nigra in Parkinson disease, progressive supranuclear palsy, and frontotemporal dementia with parkinsonism. Archives of neurology, 62(6), 917-921. doi: 10.1001/archneur.62.6.917 Healy, Daniel G., Falchi, Mario, O'Sullivan, Sean S., Bonifati, Vincenzo, Durr, Alexandra, Bressman, Susan, . . . Roberts, J. W. (2008). Phenotype, genotype, and worldwide genetic penetrance of LRRK2-associated Parkinson's disease: a case-control study. The Lancet Neurology, 7(7), 583-590. Hebert, L. E., Weuve, J., Scherr, P. A., & Evans, D. A. (2013). Alzheimer disease in the United States (2010-2050) estimated using the 2010 census. Neurology, 80(19), 1778-1783. doi: 10.1212/WNL.0b013e31828726f5 Hefti, F., Goure, W. F., Jerecic, J., Iverson, K. S., Walicke, P. A., & Krafft, G. A. (2013). The case for soluble Abeta oligomers as a drug target in Alzheimer's disease. Trends Pharmacol Sci, 34(5), 261-266. doi: 10.1016/j.tips.2013.03.002 Heinitz, Katrin, Beck, Martin, Schliebs, Reinhard, & Perez-Polo, J. (2006). Toxicity mediated by soluble oligomers of beta-amyloid(1-42) on cholinergic SN56.B5.G4 cells. Journal of neurochemistry, 98(6), 1930-1945. doi: 10.1111/j.1471-4159.2006.04015.x Hicks, D. A., Makova, N. Z., Gough, M., Parkin, E. T., Nalivaeva, N. N., & Turner, A. J. (2013). The amyloid precursor protein represses expression of acetylcholinesterase in neuronal cell lines. Journal of Biological Chemistry. doi: 10.1074/jbc.M113.461269 Hillen, H., Barghorn, S., Striebinger, A., Labkovsky, B., Muller, R., Nimmrich, V., . . . Ebert, U. (2010). Generation and therapeutic efficacy of highly oligomer-specific beta-amyloid antibodies. J Neurosci, 30(31), 10369-10379. doi: 10.1523/JNEUROSCI.5721-09.2010 Hirsch, E. C., & Hunot, S. (2009). Neuroinflammation in Parkinson's disease: a target for neuroprotection? Lancet Neurol, 8(4), 382-397. doi: 10.1016/S1474-4422(09)70062-6 Holtz, W. A., & O'Malley, K. L. (2003). Parkinsonian mimetics induce aspects of unfolded protein response in death of dopaminergic neurons. J Biol Chem, 278(21), 19367-19377. doi: 10.1074/jbc.M211821200 Holtzman, D. M., Morris, J. C., & Goate, A. M. (2011). Alzheimer's disease: the challenge of the second century. Sci Transl Med, 3(77), 77sr71. doi: 10.1126/scitranslmed.3002369 Hornykiewicz, O. (1963). [The tropical localization and content of noradrenalin and dopamine (3-hydroxytyramine) in the substantia nigra of normal persons and patients with Parkinson's disease]. Wien Klin Wochenschr, 75, 309-312. Houlden, H., & Singleton, A. B. (2012). The genetics and neuropathology of Parkinson's disease. Acta Neuropathol, 124(3), 325-338. doi: 10.1007/s00401-012-1013-5 175 Hsiao, K., Chapman, P., Nilsen, S., Eckman, C., Harigaya, Y., Younkin, S., . . . Cole, G. (1996). Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice. Science, 274(5284), 99-102. Hsu, L. J., Sagara, Y., Arroyo, A., Rockenstein, E., Sisk, A., Mallory, M., . . . Masliah, E. (2000). alpha-synuclein promotes mitochondrial deficit and oxidative stress. Am J Pathol, 157(2), 401-410. Huang da, W., Sherman, B. T., & Lempicki, R. A. (2009). Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc, 4(1), 44-57. doi: 10.1038/nprot.2008.211 Hussain, I., Powell, D., Howlett, D. R., Tew, D. G., Meek, T. D., Chapman, C., . . . Christie, G. (1999). Identification of a novel aspartic protease (Asp 2) as beta-secretase. Mol Cell Neurosci, 14(6), 419-427. doi: 10.1006/mcne.1999.0811 Hyman, B. T., Phelps, C. H., Beach, T. G., Bigio, E. H., Cairns, N. J., Carrillo, M. C., . . . Montine, T. J. (2012). National Institute on Aging-Alzheimer's Association guidelines for the neuropathologic assessment of Alzheimer's disease. Alzheimers Dement, 8(1), 1-13. doi: 10.1016/j.jalz.2011.10.007 Irizarry, M. C., Growdon, W., Gomez-Isla, T., Newell, K., George, J. M., Clayton, D. F., & Hyman, B. T. (1998). Nigral and cortical Lewy bodies and dystrophic nigral neurites in Parkinson's disease and cortical Lewy body disease contain alpha-synuclein immunoreactivity. J Neuropathol Exp Neurol, 57(4), 334-337. Isaacs, A. M., Senn, D. B., Yuan, M., Shine, J. P., & Yankner, B. A. (2006). Acceleration of amyloid beta-peptide aggregation by physiological concentrations of calcium. J Biol Chem, 281(38), 27916-27923. doi: 10.1074/jbc.M602061200 Jakes, R., Spillantini, M. G., & Goedert, M. (1994). Identification of two distinct synucleins from human brain. FEBS Lett, 345(1), 27-32. Jaleel, M., Nichols, R. J., Deak, M., Campbell, D. G., Gillardon, F., Knebel, A., & Alessi, D. R. (2007). LRRK2 phosphorylates moesin at threonine-558: characterization of how Parkinson's disease mutants affect kinase activity. Biochem J, 405(2), 307-317. doi: 10.1042/BJ20070209 Jankovic, J. (2008). Parkinson's disease: clinical features and diagnosis. J Neurol Neurosurg Psychiatry, 79(4), 368-376. doi: 10.1136/jnnp.2007.131045 Jarrett, J. T., Berger, E. P., & Lansbury, P. T., Jr. (1993). The carboxy terminus of the beta amyloid protein is critical for the seeding of amyloid formation: implications for the pathogenesis of Alzheimer's disease. Biochemistry, 32(18), 4693-4697. Jiang, Haibing, Wu, Yen-Ching, Nakamura, Masayuki, Liang, Yideng, Tanaka, Yuji, Holmes, Susan, . . . Smith, Wanli. (2007). Parkinson's disease genetic mutations increase cell susceptibility to stress: mutant alpha-synuclein enhances H2O2- and Sin-1-induced cell death. Neurobiology of aging, 28(11), 1709-1717. doi: 10.1016/j.neurobiolaging.2006.07.017 Joerchel, S., Raap, M., Bigl, M., Eschrich, K., & Schliebs, R. (2008). Oligomeric beta-amyloid(1-42) induces the expression of Alzheimer disease-relevant proteins in cholinergic SN56.B5.G4 cells as revealed by proteomic analysis. Int J Dev Neurosci, 26(3-4), 301-308. doi: 10.1016/j.ijdevneu.2008.01.004 Justus, C. Dächsel, & Matthew, J. Farrer. (2010). LRRK2 and Parkinson disease. Archives of neurology. 176 Kadonaga, J. T., Carner, K. R., Masiarz, F. R., & Tjian, R. (1987). Isolation of cDNA encoding transcription factor Sp1 and functional analysis of the DNA binding domain. Cell, 51(6), 1079-1090. Kadonaga, J. T., Jones, K. A., & Tjian, R. (1986). Promoter-Specific Activation of Rna Polymerase-Ii Transcription by Sp1. Trends in Biochemical Sciences, 11(1), 20-23. doi: Doi 10.1016/0968-0004(86)90226-4 Kaksonen, M., Pavlov, I., Voikar, V., Lauri, S. E., Hienola, A., Riekki, R., . . . Rauvala, H. (2002). Syndecan-3-deficient mice exhibit enhanced LTP and impaired hippocampus-dependent memory. Mol Cell Neurosci, 21(1), 158-172. Kalia, L. V., Kalia, S. K., & Lang, A. E. (2015). Disease-modifying strategies for Parkinson's disease. Mov Disord, 30(11), 1442-1450. doi: 10.1002/mds.26354 Kalia, L. V., Kalia, S. K., McLean, P. J., Lozano, A. M., & Lang, A. E. (2013). alpha-Synuclein oligomers and clinical implications for Parkinson disease. Ann Neurol, 73(2), 155-169. doi: 10.1002/ana.23746 Kanegae, Y., Tavares, A. T., Izpisua Belmonte, J. C., & Verma, I. M. (1998). Role of Rel/NF-kappaB transcription factors during the outgrowth of the vertebrate limb. Nature, 392(6676), 611-614. doi: 10.1038/33429 Kang, J., Lemaire, H. G., Unterbeck, A., Salbaum, J. M., Masters, C. L., Grzeschik, K. H., . . . Muller-Hill, B. (1987). The precursor of Alzheimer's disease amyloid A4 protein resembles a cell-surface receptor. Nature, 325(6106), 733-736. doi: 10.1038/325733a0 Katharine, L. Sugars, & David, C. Rubinsztein. (2003). Transcriptional abnormalities in Huntington disease. Trends in genetics : TIG. Katsel, P., Tan, W., & Haroutunian, V. (2009). Gain in brain immunity in the oldest-old differentiates cognitively normal from demented individuals. PLoS One, 4(10), e7642. doi: 10.1371/journal.pone.0007642 Kazantsev, Aleksey, & Kolchinsky, Alexander. (2008). Central role of alpha-synuclein oligomers in neurodegeneration in Parkinson disease. Archives of neurology, 65(12), 1577-1581. doi: 10.1001/archneur.65.12.1577 Kelly, J. F., Furukawa, K., Barger, S. W., Rengen, M. R., Mark, R. J., Blanc, E. M., . . . Mattson, M. P. (1996). Amyloid beta-peptide disrupts carbachol-induced muscarinic cholinergic signal transduction in cortical neurons. Proc Natl Acad Sci U S A, 93(13), 6753-6758. Kennedy, B. J., Yarbro, J. W., Kickertz, V., & Sandberg-Wollheim, M. (1968). Effect of mithramycin on a mouse glioma. Cancer Res, 28(1), 91-97. Kim, H. J., Chae, S. C., Lee, D. K., Chromy, B., Lee, S. C., Park, Y. C., . . . Hong, S. T. (2003). Selective neuronal degeneration induced by soluble oligomeric amyloid beta protein. FASEB J, 17(1), 118-120. doi: 10.1096/fj.01-0987fje Kim, I., Moon, S., Yu, K., Kim, U., & Koh, G. Y. (2001). A novel fibroblast growth factor receptor-5 preferentially expressed in the pancreas(1). Biochim Biophys Acta, 1518(1-2), 152-156. Kim, J., Onstead, L., Randle, S., Price, R., Smithson, L., Zwizinski, C., . . . McGowan, E. (2007). Abeta40 inhibits amyloid deposition in vivo. J Neurosci, 27(3), 627-633. doi: 10.1523/JNEUROSCI.4849-06.2007 Kirik, Deniz, Rosenblad, Carl, Burger, Corinna, Lundberg, Cecilia, Johansen, Teit, Muzyczka, Nicholas, . . . Björklund, Anders. (2002). Parkinson-like neurodegeneration induced by targeted overexpression of alpha-synuclein in the nigrostriatal system. The Journal of 177 neuroscience : the official journal of the Society for Neuroscience, 22(7), 2780-2791. doi: 20026246 Ko, H. S., Bailey, R., Smith, W. W., Liu, Z., Shin, J. H., Lee, Y. I., . . . Dawson, V. L. (2009). CHIP regulates leucine-rich repeat kinase-2 ubiquitination, degradation, and toxicity. Proc Natl Acad Sci U S A, 106(8), 2897-2902. doi: 10.1073/pnas.0810123106 Koike, H., Tomioka, S., Sorimachi, H., Saido, T. C., Maruyama, K., Okuyama, A., . . . Ishiura, S. (1999). Membrane-anchored metalloprotease MDC9 has an alpha-secretase activity responsible for processing the amyloid precursor protein. Biochem J, 343 Pt 2, 371-375. Kosuge, Y., Taniguchi, Y., Imai, T., Ishige, K., & Ito, Y. (2011). Neuroprotective effect of mithramycin against endoplasmic reticulum stress-induced neurotoxicity in organotypic hippocampal slice cultures. Neuropharmacology, 61(1-2), 252-261. doi: 10.1016/j.neuropharm.2011.04.009 Kumar-Singh, S., Theuns, J., Van Broeck, B., Pirici, D., Vennekens, K., Corsmit, E., . . . Van Broeckhoven, C. (2006). Mean age-of-onset of familial alzheimer disease caused by presenilin mutations correlates with both increased Abeta42 and decreased Abeta40. Hum Mutat, 27(7), 686-695. doi: 10.1002/humu.20336 Kumar, R., Nordberg, A., & Darreh-Shori, T. (2016). Amyloid-beta peptides act as allosteric modulators of cholinergic signalling through formation of soluble BAbetaACs. Brain, 139(Pt 1), 174-192. doi: 10.1093/brain/awv318 Kurz, A., Double, K. L., Lastres-Becker, I., Tozzi, A., Tantucci, M., Bockhart, V., . . . Gispert, S. (2010). A53T-alpha-synuclein overexpression impairs dopamine signaling and striatal synaptic plasticity in old mice. PLoS One, 5(7), e11464. doi: 10.1371/journal.pone.0011464 Kwakowsky, A., Potapov, K., Kim, S., Peppercorn, K., Tate, W. P., & Abraham, I. M. (2016). Treatment of beta amyloid 1-42 (Abeta(1-42))-induced basal forebrain cholinergic damage by a non-classical estrogen signaling activator in vivo. Sci Rep, 6, 21101. doi: 10.1038/srep21101 Lambert, M. P., Barlow, A. K., Chromy, B. A., Edwards, C., Freed, R., Liosatos, M., . . . Klein, W. L. (1998). Diffusible, nonfibrillar ligands derived from Abeta1-42 are potent central nervous system neurotoxins. Proc Natl Acad Sci U S A, 95(11), 6448-6453. Lammich, S., Kojro, E., Postina, R., Gilbert, S., Pfeiffer, R., Jasionowski, M., . . . Fahrenholz, F. (1999). Constitutive and regulated alpha-secretase cleavage of Alzheimer's amyloid precursor protein by a disintegrin metalloprotease. Proc Natl Acad Sci U S A, 96(7), 3922-3927. Lang, A. E., & Lozano, A. M. (1998a). Parkinson's disease. First of two parts. N Engl J Med, 339(15), 1044-1053. doi: 10.1056/NEJM199810083391506 Lang, A. E., & Lozano, A. M. (1998b). Parkinson's disease. Second of two parts. N Engl J Med, 339(16), 1130-1143. doi: 10.1056/NEJM199810153391607 Lashuel, H. A., Overk, C. R., Oueslati, A., & Masliah, E. (2013). The many faces of alpha-synuclein: from structure and toxicity to therapeutic target. Nat Rev Neurosci, 14(1), 38-48. doi: 10.1038/nrn3406 Lee, H. J., Bae, E. J., & Lee, S. J. (2014). Extracellular alpha--synuclein-a novel and crucial factor in Lewy body diseases. Nat Rev Neurol, 10(2), 92-98. doi: 10.1038/nrneurol.2013.275 178 Lee, H. J., Patel, S., & Lee, S. J. (2005). Intravesicular localization and exocytosis of alpha-synuclein and its aggregates. J Neurosci, 25(25), 6016-6024. doi: 10.1523/JNEUROSCI.0692-05.2005 Lee, H. J., Suk, J. E., Bae, E. J., & Lee, S. J. (2008). Clearance and deposition of extracellular alpha-synuclein aggregates in microglia. Biochem Biophys Res Commun, 372(3), 423-428. doi: 10.1016/j.bbrc.2008.05.045 Lee, He-Jin J., Shin, Soon Y., Choi, Chan, Lee, Young H., & Lee, Seung-Jae J. (2002). Formation and removal of alpha-synuclein aggregates in cells exposed to mitochondrial inhibitors. The Journal of biological chemistry, 277(7), 5411-5417. doi: 10.1074/jbc.M105326200 Lee, He-Jin, & Lee, Seung-Jae. (2002). Characterization of cytoplasmic alpha-synuclein aggregates. Fibril formation is tightly linked to the inclusion-forming process in cells. The Journal of biological chemistry, 277(50), 48976-48983. doi: 10.1074/jbc.M208192200 Lee, M. K., Stirling, W., Xu, Y., Xu, X., Qui, D., Mandir, A. S., . . . Price, D. L. (2002). Human alpha-synuclein-harboring familial Parkinson's disease-linked Ala-53 --> Thr mutation causes neurodegenerative disease with alpha-synuclein aggregation in transgenic mice. Proc Natl Acad Sci U S A, 99(13), 8968-8973. doi: 10.1073/pnas.132197599 Lesne, S., Koh, M. T., Kotilinek, L., Kayed, R., Glabe, C. G., Yang, A., . . . Ashe, K. H. (2006). A specific amyloid-beta protein assembly in the brain impairs memory. Nature, 440(7082), 352-357. doi: 10.1038/nature04533 Lewis, P. A., Greggio, E., Beilina, A., Jain, S., Baker, A., & Cookson, M. R. (2007). The R1441C mutation of LRRK2 disrupts GTP hydrolysis. Biochem Biophys Res Commun, 357(3), 668-671. doi: 10.1016/j.bbrc.2007.04.006 Li, X., Tan, Y. C., Poulose, S., Olanow, C. W., Huang, X. Y., & Yue, Z. (2007). Leucine-rich repeat kinase 2 (LRRK2)/PARK8 possesses GTPase activity that is altered in familial Parkinson's disease R1441C/G mutants. J Neurochem, 103(1), 238-247. doi: 10.1111/j.1471-4159.2007.04743.x Li, Y., Zhou, W., Tong, Y., He, G., & Song, W. (2006). Control of APP processing and Abeta generation level by BACE1 enzymatic activity and transcription. FASEB J, 20(2), 285-292. doi: 10.1096/fj.05-4986com Liao, J., Wu, C. X., Burlak, C., Zhang, S., Sahm, H., Wang, M., . . . Hoang, Q. Q. (2014). Parkinson disease-associated mutation R1441H in LRRK2 prolongs the "active state" of its GTPase domain. Proc Natl Acad Sci U S A, 111(11), 4055-4060. doi: 10.1073/pnas.1323285111 Lin, X., Parisiadou, L., Gu, X. L., Wang, L., Shim, H., Sun, L., . . . Cai, H. (2009). Leucine-rich repeat kinase 2 regulates the progression of neuropathology induced by Parkinson's-disease-related mutant alpha-synuclein. Neuron, 64(6), 807-827. doi: 10.1016/j.neuron.2009.11.006 Liu, Z., Wang, X., Yu, Y., Li, X., Wang, T., Jiang, H., . . . Smith, W. W. (2008). A Drosophila model for LRRK2-linked parkinsonism. Proc Natl Acad Sci U S A, 105(7), 2693-2698. doi: 10.1073/pnas.0708452105 Lovén, Jakob, Orlando, David, Sigova, Alla, Lin, Charles, Rahl, Peter, Burge, Christopher, . . . Young, Richard. (2012). Revisiting global gene expression analysis. Cell, 151(3), 476-482. doi: 10.1016/j.cell.2012.10.012 179 Lüth, Hans-Joachim, Apelt, Jenny, Ihunwo, Amadi, Arendt, Thomas, & Schliebs, Reinhard. (2003). Degeneration of beta-amyloid-associated cholinergic structures in transgenic APP SW mice. Brain research, 977(1), 16-22. doi: 10.1016/S0006-8993(03)02658-1 Ly, P. T., Cai, F., & Song, W. (2011). Detection of neuritic plaques in Alzheimer's disease mouse model. J Vis Exp(53). doi: 10.3791/2831 Ly, P. T., Wu, Y., Zou, H., Wang, R., Zhou, W., Kinoshita, A., . . . Song, W. (2013). Inhibition of GSK3beta-mediated BACE1 expression reduces Alzheimer-associated phenotypes. J Clin Invest, 123(1), 224-235. doi: 10.1172/JCI64516 64516 [pii] MacLeod, D., Dowman, J., Hammond, R., Leete, T., Inoue, K., & Abeliovich, A. (2006). The familial Parkinsonism gene LRRK2 regulates neurite process morphology. Neuron, 52(4), 587-593. doi: 10.1016/j.neuron.2006.10.008 Mandemakers, W., Snellinx, A., O'Neill, M. J., & de Strooper, B. (2012). LRRK2 expression is enriched in the striosomal compartment of mouse striatum. Neurobiol Dis, 48(3), 582-593. doi: 10.1016/j.nbd.2012.07.017 Marin, M., Karis, A., Visser, P., Grosveld, F., & Philipsen, S. (1997). Transcription factor Sp1 is essential for early embryonic development but dispensable for cell growth and differentiation. Cell, 89(4), 619-628. Marques, C. A., Keil, U., Bonert, A., Steiner, B., Haass, C., Muller, W. E., & Eckert, A. (2003). Neurotoxic mechanisms caused by the Alzheimer's disease-linked Swedish amyloid precursor protein mutation: oxidative stress, caspases, and the JNK pathway. J Biol Chem, 278(30), 28294-28302. doi: 10.1074/jbc.M212265200 Martin, I., Kim, J. W., Dawson, V. L., & Dawson, T. M. (2014). LRRK2 pathobiology in Parkinson's disease. J Neurochem, 131(5), 554-565. doi: 10.1111/jnc.12949 Martin, I., Kim, J. W., Lee, B. D., Kang, H. C., Xu, J. C., Jia, H., . . . Dawson, V. L. (2014). Ribosomal protein s15 phosphorylation mediates LRRK2 neurodegeneration in Parkinson's disease. Cell, 157(2), 472-485. doi: 10.1016/j.cell.2014.01.064 Martin, L. J., Pan, Y., Price, A. C., Sterling, W., Copeland, N. G., Jenkins, N. A., . . . Lee, M. K. (2006). Parkinson's disease alpha-synuclein transgenic mice develop neuronal mitochondrial degeneration and cell death. J Neurosci, 26(1), 41-50. doi: 10.1523/JNEUROSCI.4308-05.2006 Martinez-Vicente, Marta, Talloczy, Zsolt, Kaushik, Susmita, Massey, Ashish, Mazzulli, Joseph, Mosharov, Eugene, . . . Cuervo, Ana. (2008). Dopamine-modified alpha-synuclein blocks chaperone-mediated autophagy. The Journal of clinical investigation, 118(2), 777-788. doi: 10.1172/JCI32806 Masliah, E., Rockenstein, E., Adame, A., Alford, M., Crews, L., Hashimoto, M., . . . Schenk, D. (2005). Effects of alpha-synuclein immunization in a mouse model of Parkinson's disease. Neuron, 46(6), 857-868. doi: 10.1016/j.neuron.2005.05.010 Masliah, E., Rockenstein, E., Mante, M., Crews, L., Spencer, B., Adame, A., . . . Schenk, D. (2011). Passive immunization reduces behavioral and neuropathological deficits in an alpha-synuclein transgenic model of Lewy body disease. PLoS One, 6(4), e19338. doi: 10.1371/journal.pone.0019338 Massano, J., & Bhatia, K. P. (2012). Clinical approach to Parkinson's disease: features, diagnosis, and principles of management. Cold Spring Harb Perspect Med, 2(6), a008870. doi: 10.1101/cshperspect.a008870 180 Masters, C. L., Simms, G., Weinman, N. A., Multhaup, G., McDonald, B. L., & Beyreuther, K. (1985). Amyloid plaque core protein in Alzheimer disease and Down syndrome. Proc Natl Acad Sci U S A, 82(12), 4245-4249. Mata, I. F., Kachergus, J. M., Taylor, J. P., Lincoln, S., Aasly, J., Lynch, T., . . . Farrer, M. J. (2005). Lrrk2 pathogenic substitutions in Parkinson's disease. Neurogenetics, 6(4), 171-177. doi: 10.1007/s10048-005-0005-1 Mata, I. F., Wedemeyer, W. J., Farrer, M. J., Taylor, J. P., & Gallo, K. A. (2006). LRRK2 in Parkinson's disease: protein domains and functional insights. Trends Neurosci, 29(5), 286-293. doi: 10.1016/j.tins.2006.03.006 Mattson, M. P., & Meffert, M. K. (2006). Roles for NF-kappaB in nerve cell survival, plasticity, and disease. Cell Death Differ, 13(5), 852-860. doi: 10.1038/sj.cdd.4401837 Mc Donald, J. M., Savva, G. M., Brayne, C., Welzel, A. T., Forster, G., Shankar, G. M., . . . Ageing, Study. (2010). The presence of sodium dodecyl sulphate-stable Abeta dimers is strongly associated with Alzheimer-type dementia. Brain, 133(Pt 5), 1328-1341. doi: 10.1093/brain/awq065 McLean, C., Cherny, R., Fraser, F., Fuller, S., Smith, M., Beyreuther, K., . . . Masters, C. (1999). Soluble pool of Abeta amyloid as a determinant of severity of neurodegeneration in Alzheimer's disease. Annals of neurology, 46(6), 860-866. doi: 10.1002/1531-8249(199912)46:6<860::AID-ANA8>3.0.CO;2-M McLean, P. J., Kawamata, H., Shariff, S., Hewett, J., Sharma, N., Ueda, K., . . . Hyman, B. T. (2002). TorsinA and heat shock proteins act as molecular chaperones: suppression of alpha-synuclein aggregation. J Neurochem, 83(4), 846-854. Miller, J. A., Oldham, M. C., & Geschwind, D. H. (2008). A systems level analysis of transcriptional changes in Alzheimer's disease and normal aging. J Neurosci, 28(6), 1410-1420. doi: 10.1523/JNEUROSCI.4098-07.2008 Miller, R. M., Kiser, G. L., Kaysser-Kranich, T., Casaceli, C., Colla, E., Lee, M. K., . . . Federoff, H. J. (2007). Wild-type and mutant alpha-synuclein induce a multi-component gene expression profile consistent with shared pathophysiology in different transgenic mouse models of PD. Exp Neurol, 204(1), 421-432. doi: 10.1016/j.expneurol.2006.12.005 Miller, R. M., Kiser, G. L., Kaysser-Kranich, T. M., Lockner, R. J., Palaniappan, C., & Federoff, H. J. (2006). Robust dysregulation of gene expression in substantia nigra and striatum in Parkinson's disease. Neurobiol Dis, 21(2), 305-313. doi: 10.1016/j.nbd.2005.07.010 Moran, L. B., Duke, D. C., Deprez, M., Dexter, D. T., Pearce, R. K., & Graeber, M. B. (2006). Whole genome expression profiling of the medial and lateral substantia nigra in Parkinson's disease. Neurogenetics, 7(1), 1-11. doi: 10.1007/s10048-005-0020-2 Mousavi, M., & Hellström-Lindahl, E. (2009). Nicotinic receptor agonists and antagonists increase sAPPalpha secretion and decrease Abeta levels in vitro. Neurochemistry international, 54(3-4), 237-244. doi: 10.1016/j.neuint.2008.12.001 Moussa, C. E., Mahmoodian, F., Tomita, Y., & Sidhu, A. (2008). Dopamine differentially induces aggregation of A53T mutant and wild type alpha-synuclein: insights into the protein chemistry of Parkinson's disease. Biochem Biophys Res Commun, 365(4), 833-839. doi: 10.1016/j.bbrc.2007.11.075 Mufson, E. J., Counts, S. E., & Ginsberg, S. D. (2002). Gene expression profiles of cholinergic nucleus basalis neurons in Alzheimer's disease. Neurochem Res, 27(10), 1035-1048. 181 Mullan, M., Crawford, F., Axelman, K., Houlden, H., Lilius, L., Winblad, B., & Lannfelt, L. (1992). A pathogenic mutation for probable Alzheimer's disease in the APP gene at the N-terminus of beta-amyloid. Nat Genet, 1(5), 345-347. doi: 10.1038/ng0892-345 Nagasaka, Yosuke, Dillner, Karin, Ebise, Hayao, Teramoto, Reiji, Nakagawa, Hiroyuki, Lilius, Lena, . . . Graff, Caroline. (2005). A unique gene expression signature discriminates familial Alzheimer's disease mutation carriers from their wild-type siblings. Proceedings of the National Academy of Sciences of the United States of America, 102(41), 14854-14859. doi: 10.1073/pnas.0504178102 Narhi, L., Wood, S. J., Steavenson, S., Jiang, Y., Wu, G. M., Anafi, D., . . . Citron, M. (1999). Both familial Parkinson's disease mutations accelerate alpha-synuclein aggregation. J Biol Chem, 274(14), 9843-9846. Nemani, V. M., Lu, W., Berge, V., Nakamura, K., Onoa, B., Lee, M. K., . . . Edwards, R. H. (2010). Increased expression of alpha-synuclein reduces neurotransmitter release by inhibiting synaptic vesicle reclustering after endocytosis. Neuron, 65(1), 66-79. doi: 10.1016/j.neuron.2009.12.023 Neve, R. L., Finch, E. A., & Dawes, L. R. (1988). Expression of the Alzheimer amyloid precursor gene transcripts in the human brain. Neuron, 1(8), 669-677. Nichols, R. J., Dzamko, N., Hutti, J. E., Cantley, L. C., Deak, M., Moran, J., . . . Alessi, D. R. (2009). Substrate specificity and inhibitors of LRRK2, a protein kinase mutated in Parkinson's disease. Biochem J, 424(1), 47-60. doi: 10.1042/BJ20091035 Nichols, W. C., Pankratz, N., Hernandez, D., Paisan-Ruiz, C., Jain, S., Halter, C. A., . . . Parkinson Study Group, Progeni investigators. (2005). Genetic screening for a single common LRRK2 mutation in familial Parkinson's disease. Lancet, 365(9457), 410-412. doi: 10.1016/S0140-6736(05)17828-3 Nikonova, E. V., Xiong, Y., Tanis, K. Q., Dawson, V. L., Vogel, R. L., Finney, E. M., . . . Dawson, T. M. (2012). Transcriptional responses to loss or gain of function of the leucine-rich repeat kinase 2 (LRRK2) gene uncover biological processes modulated by LRRK2 activity. Hum Mol Genet, 21(1), 163-174. doi: 10.1093/hmg/ddr451 Nilbratt, Mats, Friberg, Linda, Mousavi, Malahat, Marutle, Amelia, & Nordberg, Agneta. (2007). Retinoic acid and nerve growth factor induce differential regulation of nicotinic acetylcholine receptor subunit expression in SN56 cells. Journal of neuroscience research, 85(3), 504-514. doi: 10.1002/jnr.21156 O'Callaghan, P., Sandwall, E., Li, J. P., Yu, H., Ravid, R., Guan, Z. Z., . . . Zhang, X. (2008). Heparan sulfate accumulation with Abeta deposits in Alzheimer's disease and Tg2576 mice is contributed by glial cells. Brain Pathol, 18(4), 548-561. doi: 10.1111/j.1750-3639.2008.00152.x Oddo, S., Caccamo, A., Tran, L., Lambert, M. P., Glabe, C. G., Klein, W. L., & LaFerla, F. M. (2006). Temporal profile of amyloid-beta (Abeta) oligomerization in an in vivo model of Alzheimer disease. A link between Abeta and tau pathology. J Biol Chem, 281(3), 1599-1604. doi: 10.1074/jbc.M507892200 Olson, M. I., & Shaw, C. M. (1969). Presenile dementia and Alzheimer's disease in mongolism. Brain, 92(1), 147-156. Orphanides, G., & Reinberg, D. (2002). A unified theory of gene expression. Cell, 108(4), 439-451. 182 Osada, N., Kosuge, Y., Ishige, K., & Ito, Y. (2013). Mithramycin, an agent for developing new therapeutic drugs for neurodegenerative diseases. J Pharmacol Sci, 122(4), 251-256. Ostrerova-Golts, Natalie, Petrucelli, Leonard, Hardy, John, Lee, John M., Farer, Matthew, & Wolozin, Benjamin. (2000). The A53T -synuclein mutation increases iron-dependent aggregation and toxicity. The Journal of Neuroscience, 20(16), 6048-6054. Outeiro, T. F., Putcha, P., Tetzlaff, J. E., Spoelgen, R., Koker, M., Carvalho, F., . . . McLean, P. J. (2008). Formation of toxic oligomeric alpha-synuclein species in living cells. PLoS One, 3(4), e1867. doi: 10.1371/journal.pone.0001867 Paisan-Ruiz, C., Jain, S., Evans, E. W., Gilks, W. P., Simon, J., van der Brug, M., . . . Singleton, A. B. (2004). Cloning of the gene containing mutations that cause PARK8-linked Parkinson's disease. Neuron, 44(4), 595-600. doi: 10.1016/j.neuron.2004.10.023 Pakaski, M., & Kalman, J. (2008). Interactions between the amyloid and cholinergic mechanisms in Alzheimer's disease. Neurochem Int, 53(5), 103-111. doi: 10.1016/j.neuint.2008.06.005 Paleologou, K. E., Kragh, C. L., Mann, D. M., Salem, S. A., Al-Shami, R., Allsop, D., . . . El-Agnaf, O. M. (2009). Detection of elevated levels of soluble alpha-synuclein oligomers in post-mortem brain extracts from patients with dementia with Lewy bodies. Brain, 132(Pt 4), 1093-1101. doi: 10.1093/brain/awn349 Parachikova, A., Agadjanyan, M. G., Cribbs, D. H., Blurton-Jones, M., Perreau, V., Rogers, J., . . . Cotman, C. W. (2007). Inflammatory changes parallel the early stages of Alzheimer disease. Neurobiol Aging, 28(12), 1821-1833. doi: 10.1016/j.neurobiolaging.2006.08.014 Parisiadou, L., Xie, C., Cho, H. J., Lin, X., Gu, X. L., Long, C. X., . . . Cai, H. (2009). Phosphorylation of ezrin/radixin/moesin proteins by LRRK2 promotes the rearrangement of actin cytoskeleton in neuronal morphogenesis. J Neurosci, 29(44), 13971-13980. doi: 10.1523/JNEUROSCI.3799-09.2009 Park, Bokyung, Oh, Chang-Ki, Choi, Won-Seok, Chung, In, Youdim, Moussa, & Oh, Young. (2011). Microarray expression profiling in 6-hydroxydopamine-induced dopaminergic neuronal cell death. Journal of neural transmission (Vienna, Austria : 1996), 118(11), 1585-1598. doi: 10.1007/s00702-011-0710-x Park, J. M., Ho, D. H., Yun, H. J., Kim, H. J., Lee, C. H., Park, S. W., . . . Seol, W. (2013). Dexamethasone induces the expression of LRRK2 and alpha-synuclein, two genes that when mutated cause Parkinson's disease in an autosomal dominant manner. BMB Rep, 46(9), 454-459. Parks, C. L., & Shenk, T. (1996). The serotonin 1a receptor gene contains a TATA-less promoter that responds to MAZ and Sp1. J Biol Chem, 271(8), 4417-4430. Peng, Xiangmin, Peng, Xiangmin, Tehranian, Roya, Dietrich, Paula, Stefanis, Leonidas, & Perez, Ruth. (2005). Alpha-synuclein activation of protein phosphatase 2A reduces tyrosine hydroxylase phosphorylation in dopaminergic cells. Journal of cell science, 118(Pt 15), 3523-3530. doi: 10.1242/jcs.02481 Perez, Ruth, Waymire, Jack, Lin, Eva, Liu, Jen, Guo, Fengli, & Zigmond, Michael. (2002). A role for alpha-synuclein in the regulation of dopamine biosynthesis. The Journal of neuroscience : the official journal of the Society for Neuroscience, 22(8), 3090-3099. doi: 20026307 Perlia, C. P., Gubisch, N. J., Wolter, J., Edelberg, D., Dederick, M. M., & Taylor, S. G., 3rd. (1970). Mithramycin treatment of hypercalcemia. Cancer, 25(2), 389-394. 183 Pettit, D. L., Shao, Z., & Yakel, J. L. (2001). beta-Amyloid(1-42) peptide directly modulates nicotinic receptors in the rat hippocampal slice. J Neurosci, 21(1), RC120. Pinto, T., Lanctot, K. L., & Herrmann, N. (2011). Revisiting the cholinergic hypothesis of behavioral and psychological symptoms in dementia of the Alzheimer's type. Ageing Res Rev, 10(4), 404-412. doi: 10.1016/j.arr.2011.01.003 Podlisny, M. B., Ostaszewski, B. L., Squazzo, S. L., Koo, E. H., Rydell, R. E., Teplow, D. B., & Selkoe, D. J. (1995). Aggregation of secreted amyloid beta-protein into sodium dodecyl sulfate-stable oligomers in cell culture; Start of abeta oligomer hypothesis. J Biol Chem, 270(16), 9564-9570. Polymeropoulos, M. H., Lavedan, C., Leroy, E., Ide, S. E., Dehejia, A., Dutra, A., . . . Nussbaum, R. L. (1997). Mutation in the alpha-synuclein gene identified in families with Parkinson's disease. Science, 276(5321), 2045-2047. doi: 10.1126/science.276.5321.2045 Porter, A. G., & Janicke, R. U. (1999). Emerging roles of caspase-3 in apoptosis. Cell Death Differ, 6(2), 99-104. doi: 10.1038/sj.cdd.4400476 Prasad, Judith, Kumar, Bipin, Andreatta, Cynthia, Nahreini, Piruz, Hanson, Amy, Yan, Xiang, & Prasad, Kedar. (2004). Overexpression of alpha-synuclein decreased viability and enhanced sensitivity to prostaglandin E(2), hydrogen peroxide, and a nitric oxide donor in differentiated neuroblastoma cells. Journal of neuroscience research, 76(3), 415-422. doi: 10.1002/jnr.20058 Praticò, Domenico. (2008). Oxidative stress hypothesis in Alzheimer's disease: a reappraisal. Trends in pharmacological sciences, 29(12), 609-615. doi: 10.1016/j.tips.2008.09.001 Previdi, S., Malek, A., Albertini, V., Riva, C., Capella, C., Broggini, M., . . . Catapano, C. V. (2010). Inhibition of Sp1-dependent transcription and antitumor activity of the new aureolic acid analogues mithramycin SDK and SK in human ovarian cancer xenografts. Gynecol Oncol, 118(2), 182-188. doi: 10.1016/j.ygyno.2010.03.020 Qing, H., Zhou, W., Christensen, M. A., Sun, X., Tong, Y., & Song, W. (2004). Degradation of BACE by the ubiquitin-proteasome pathway. FASEB J, 18(13), 1571-1573. doi: 10.1096/fj.04-1994fje Raiss, C. C., Braun, T. S., Konings, I. B., Grabmayr, H., Hassink, G. C., Sidhu, A., . . . Claessens, M. M. (2016). Functionally different alpha-synuclein inclusions yield insight into Parkinson's disease pathology. Sci Rep, 6, 23116. doi: 10.1038/srep23116 Ray, R., Snyder, R. C., Thomas, S., Koller, C. A., & Miller, D. M. (1989). Mithramycin blocks protein binding and function of the SV40 early promoter. J Clin Invest, 83(6), 2003-2007. doi: 10.1172/JCI114110 Ream, N. W., Perlia, C. P., Wolter, J., & Taylor, S. G., 3rd. (1968). Mithramycin therapy in disseminated germinal testicular cancer. JAMA, 204(12), 1030-1036. Rebeck, G. W., Reiter, J. S., Strickland, D. K., & Hyman, B. T. (1993). Apolipoprotein E in sporadic Alzheimer's disease: allelic variation and receptor interactions. Neuron, 11(4), 575-580. Reddy, P., McWeeney, Shannon, Park, Byung, Manczak, Maria, Gutala, Ramana, Partovi, Dara, . . . Quinn, Joseph. (2004). Gene expression profiles of transcripts in amyloid precursor protein transgenic mice: up-regulation of mitochondrial metabolism and apoptotic genes is an early cellular change in Alzheimer's disease. Human molecular genetics, 13(12), 1225-1240. doi: 10.1093/hmg/ddh140 184 Rieckmann, T., Kotevic, I., & Trueb, B. (2008). The cell surface receptor FGFRL1 forms constitutive dimers that promote cell adhesion. Exp Cell Res, 314(5), 1071-1081. doi: 10.1016/j.yexcr.2007.10.029 Robakis, N. K., Ramakrishna, N., Wolfe, G., & Wisniewski, H. M. (1987). Molecular cloning and characterization of a cDNA encoding the cerebrovascular and the neuritic plaque amyloid peptides. Proc Natl Acad Sci U S A, 84(12), 4190-4194. Ryan, D. A., Narrow, W. C., Federoff, H. J., & Bowers, W. J. (2010). An improved method for generating consistent soluble amyloid-beta oligomer preparations for in vitro neurotoxicity studies. J Neurosci Methods, 190(2), 171-179. doi: 10.1016/j.jneumeth.2010.05.001 Ryan, S. D., Dolatabadi, N., Chan, S. F., Zhang, X., Akhtar, M. W., Parker, J., . . . Lipton, S. A. (2013). Isogenic human iPSC Parkinson's model shows nitrosative stress-induced dysfunction in MEF2-PGC1alpha transcription. Cell, 155(6), 1351-1364. doi: 10.1016/j.cell.2013.11.009 Ryu, Hoon, Lee, Junghee, Zaman, Khalequz, Kubilis, James, Ferrante, Robert J., Ross, Brian D., . . . Ratan, Rajiv R. (2003). Sp1 and Sp3 are oxidative stress-inducible, antideath transcription factors in cortical neurons. The Journal of neuroscience, 23(9), 3597-3606. Saito, T., Matsuba, Y., Mihira, N., Takano, J., Nilsson, P., Itohara, S., . . . Saido, T. C. (2014). Single App knock-in mouse models of Alzheimer's disease. Nat Neurosci, 17(5), 661-663. doi: 10.1038/nn.3697 Sakaguchi-Nakashima, A., Meir, J. Y., Jin, Y., Matsumoto, K., & Hisamoto, N. (2007). LRK-1, a C. elegans PARK8-related kinase, regulates axonal-dendritic polarity of SV proteins. Curr Biol, 17(7), 592-598. doi: 10.1016/j.cub.2007.01.074 Salloway, S., Sperling, R., Fox, N. C., Blennow, K., Klunk, W., Raskind, M., . . . Clinical Trial, Investigators. (2014). Two phase 3 trials of bapineuzumab in mild-to-moderate Alzheimer's disease. N Engl J Med, 370(4), 322-333. doi: 10.1056/NEJMoa1304839 Santos, A. N., Ewers, M., Minthon, L., Simm, A., Silber, R. E., Blennow, K., . . . Hampel, H. (2012). Amyloid-beta oligomers in cerebrospinal fluid are associated with cognitive decline in patients with Alzheimer's disease. J Alzheimers Dis, 29(1), 171-176. doi: 10.3233/JAD-2012-111361 Santpere, G., Nieto, M., Puig, B., & Ferrer, I. (2006). Abnormal Sp1 transcription factor expression in Alzheimer disease and tauopathies. Neuroscience letters. Saunders, A. M., Strittmatter, W. J., Schmechel, D., George-Hyslop, P. H., Pericak-Vance, M. A., Joo, S. H., . . . et al. (1993). Association of apolipoprotein E allele epsilon 4 with late-onset familial and sporadic Alzheimer's disease. Neurology, 43(8), 1467-1472. Schena, M., Shalon, D., Davis, R. W., & Brown, P. O. (1995). Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science, 270(5235), 467-470. Schenk, D., Barbour, R., Dunn, W., Gordon, G., Grajeda, H., Guido, T., . . . Seubert, P. (1999). Immunization with amyloid-beta attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature, 400(6740), 173-177. doi: 10.1038/22124 Scheuner, D., Eckman, C., Jensen, M., Song, X., Citron, M., Suzuki, N., . . . Younkin, S. (1996). Secreted amyloid beta-protein similar to that in the senile plaques of Alzheimer's disease is increased in vivo by the presenilin 1 and 2 and APP mutations linked to familial Alzheimer's disease. Nat Med, 2(8), 864-870. 185 Schmittgen, T. D., & Livak, K. J. (2008). Analyzing real-time PCR data by the comparative C(T) method. Nat Protoc, 3(6), 1101-1108. Schulz, J. G., Annaert, W., Vandekerckhove, J., Zimmermann, P., De Strooper, B., & David, G. (2003). Syndecan 3 intramembrane proteolysis is presenilin/gamma-secretase-dependent and modulates cytosolic signaling. J Biol Chem, 278(49), 48651-48657. doi: 10.1074/jbc.M308424200 Selkoe, D. J., & Hardy, J. (2016). The amyloid hypothesis of Alzheimer's disease at 25 years. EMBO Mol Med. doi: 10.15252/emmm.201606210 Sharon, R., Bar-Joseph, I., Frosch, M. P., Walsh, D. M., Hamilton, J. A., & Selkoe, D. J. (2003). The formation of highly soluble oligomers of alpha-synuclein is regulated by fatty acids and enhanced in Parkinson's disease. Neuron, 37(4), 583-595. Shin, J., Yu, S. B., Yu, U. Y., Jo, S. A., & Ahn, J. H. (2010). Swedish mutation within amyloid precursor protein modulates global gene expression towards the pathogenesis of Alzheimer's disease. BMB Rep, 43(10), 704-709. doi: 10.5483/BMBRep.2010.43.10.704 Shin, Y., Klucken, J., Patterson, C., Hyman, B. T., & McLean, P. J. (2005). The co-chaperone carboxyl terminus of Hsp70-interacting protein (CHIP) mediates alpha-synuclein degradation decisions between proteasomal and lysosomal pathways. J Biol Chem, 280(25), 23727-23734. doi: 10.1074/jbc.M503326200 Simunovic, F., Yi, M., Wang, Y., Macey, L., Brown, L. T., Krichevsky, A. M., . . . Sonntag, K. C. (2009). Gene expression profiling of substantia nigra dopamine neurons: further insights into Parkinson's disease pathology. Brain, 132(Pt 7), 1795-1809. doi: 10.1093/brain/awn323 Sinha, S., Anderson, J. P., Barbour, R., Basi, G. S., Caccavello, R., Davis, D., . . . John, V. (1999). Purification and cloning of amyloid precursor protein beta-secretase from human brain. Nature, 402(6761), 537-540. doi: 10.1038/990114 Skibinski, Gaia, Nakamura, Ken, Cookson, Mark R., & Finkbeiner, Steven. (2014). Mutant LRRK2 Toxicity in Neurons Depends on LRRK2 Levels and Synuclein But Not Kinase Activity or Inclusion Bodies. The Journal of Neuroscience, 34(2), 418-433. doi: 10.1523/JNEUROSCI.2712-13.2014 Sleeman, M., Fraser, J., McDonald, M., Yuan, S., White, D., Grandison, P., . . . Murison, J. G. (2001). Identification of a new fibroblast growth factor receptor, FGFR5. Gene, 271(2), 171-182. Smith, W. W., Pei, Z., Jiang, H., Dawson, V. L., Dawson, T. M., & Ross, C. A. (2006). Kinase activity of mutant LRRK2 mediates neuronal toxicity. Nat Neurosci, 9(10), 1231-1233. doi: 10.1038/nn1776 Smith, W. W., Pei, Z., Jiang, H., Moore, D. J., Liang, Y., West, A. B., . . . Ross, C. A. (2005). Leucine-rich repeat kinase 2 (LRRK2) interacts with parkin, and mutant LRRK2 induces neuronal degeneration. Proc Natl Acad Sci U S A, 102(51), 18676-18681. doi: 10.1073/pnas.0508052102 Spillantini, M., Crowther, R., Jakes, R., Hasegawa, M., & Goedert, M. (1998). alpha-Synuclein in filamentous inclusions of Lewy bodies from Parkinson's disease and dementia with lewy bodies. Proceedings of the National Academy of Sciences of the United States of America, 95(11), 6469-6473. doi: 10.1073/pnas.95.11.6469 Spillantini, M., Schmidt, M., Lee, V., Trojanowski, J., Jakes, R., & Goedert, M. (1997). Alpha-synuclein in Lewy bodies. Nature, 388(6645), 839-840. doi: 10.1038/42166 186 Stefanis, L., Larsen, K. E., Rideout, H. J., Sulzer, D., & Greene, L. A. (2001). Expression of A53T mutant but not wild-type alpha-synuclein in PC12 cells induces alterations of the ubiquitin-dependent degradation system, loss of dopamine release, and autophagic cell death. J Neurosci, 21(24), 9549-9560. Steinberg, F., Gerber, S. D., Rieckmann, T., & Trueb, B. (2010). Rapid fusion and syncytium formation of heterologous cells upon expression of the FGFRL1 receptor. J Biol Chem, 285(48), 37704-37715. doi: 10.1074/jbc.M110.140517 Steinberg, F., Zhuang, L., Beyeler, M., Kalin, R. E., Mullis, P. E., Brandli, A. W., & Trueb, B. (2010). The FGFRL1 receptor is shed from cell membranes, binds fibroblast growth factors (FGFs), and antagonizes FGF signaling in Xenopus embryos. J Biol Chem, 285(3), 2193-2202. doi: 10.1074/jbc.M109.058248 Stine, W. B., Jr., Dahlgren, K. N., Krafft, G. A., & LaDu, M. J. (2003). In vitro characterization of conditions for amyloid-beta peptide oligomerization and fibrillogenesis. J Biol Chem, 278(13), 11612-11622. doi: 10.1074/jbc.M210207200 Strittmatter, W. J., Saunders, A. M., Schmechel, D., Pericak-Vance, M., Enghild, J., Salvesen, G. S., & Roses, A. D. (1993). Apolipoprotein E: high-avidity binding to beta-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc Natl Acad Sci U S A, 90(5), 1977-1981. Sturchler-Pierrat, C., Abramowski, D., Duke, M., Wiederhold, K. H., Mistl, C., Rothacher, S., . . . Sommer, B. (1997). Two amyloid precursor protein transgenic mouse models with Alzheimer disease-like pathology. Proc Natl Acad Sci U S A, 94(24), 13287-13292. Su, X., Maguire-Zeiss, K. A., Giuliano, R., Prifti, L., Venkatesh, K., & Federoff, H. J. (2008). Synuclein activates microglia in a model of Parkinson's disease. Neurobiol Aging, 29(11), 1690-1701. doi: 10.1016/j.neurobiolaging.2007.04.006 Sulzer, David. (2007). Multiple hit hypotheses for dopamine neuron loss in Parkinson's disease. Trends in neurosciences, 30(5), 244-250. doi: 10.1016/j.tins.2007.03.009 Sun, X., He, G., & Song, W. (2006). BACE2, as a novel APP theta-secretase, is not responsible for the pathogenesis of Alzheimer's disease in Down syndrome. FASEB J, 20(9), 1369-1376. doi: 10.1096/fj.05-5632com Tamaki, T., Ohnishi, K., Hartl, C., LeRoy, E. C., & Trojanowska, M. (1995). Characterization of a GC-rich region containing Sp1 binding site(s) as a constitutive responsive element of the alpha 2(I) collagen gene in human fibroblasts. J Biol Chem, 270(9), 4299-4304. Tan, M. G., Chua, W. T., Esiri, M. M., Smith, A. D., Vinters, H. V., & Lai, M. K. (2010). Genome wide profiling of altered gene expression in the neocortex of Alzheimer's disease. J Neurosci Res, 88(6), 1157-1169. doi: 10.1002/jnr.22290 Tanabe, C., Hotoda, N., Sasagawa, N., Sehara-Fujisawa, A., Maruyama, K., & Ishiura, S. (2007). ADAM19 is tightly associated with constitutive Alzheimer's disease APP alpha-secretase in A172 cells. Biochem Biophys Res Commun, 352(1), 111-117. doi: 10.1016/j.bbrc.2006.10.181 Tansey, M. G., & Goldberg, M. S. (2010). Neuroinflammation in Parkinson's disease: its role in neuronal death and implications for therapeutic intervention. Neurobiol Dis, 37(3), 510-518. doi: 10.1016/j.nbd.2009.11.004 Tanzi, R. E., Gusella, J. F., Watkins, P. C., Bruns, G. A., St George-Hyslop, P., Van Keuren, M. L., . . . Neve, R. L. (1987). Amyloid beta protein gene: cDNA, mRNA distribution, and genetic linkage near the Alzheimer locus. Science, 235(4791), 880-884. 187 Taylor, D. J., Parsons, C. E., Han, H., Jayaraman, A., & Rege, K. (2011). Parallel screening of FDA-approved antineoplastic drugs for identifying sensitizers of TRAIL-induced apoptosis in cancer cells. BMC Cancer, 11, 470. doi: 10.1186/1471-2407-11-470 Tehranian, Roya, Montoya, Susana, Van Laar, Amber, Hastings, Teresa, & Perez, Ruth. (2006). Alpha-synuclein inhibits aromatic amino acid decarboxylase activity in dopaminergic cells. Journal of neurochemistry, 99(4), 1188-1196. doi: 10.1111/j.1471-4159.2006.04146.x Tetzlaff, J. E., Putcha, P., Outeiro, T. F., Ivanov, A., Berezovska, O., Hyman, B. T., & McLean, P. J. (2008). CHIP targets toxic alpha-Synuclein oligomers for degradation. J Biol Chem, 283(26), 17962-17968. doi: 10.1074/jbc.M802283200 Tkachenko, E., Rhodes, J. M., & Simons, M. (2005). Syndecans: new kids on the signaling block. Circ Res, 96(5), 488-500. doi: 10.1161/01.RES.0000159708.71142.c8 Tokuda, T., Qureshi, M., Ardah, M., Varghese, S., Shehab, S., Kasai, T., . . . El-Agnaf, O. (2010). Detection of elevated levels of α-synuclein oligomers in CSF from patients with Parkinson disease. Neurology, 75(20), 1766-1772. doi: 10.1212/WNL.0b013e3181fd613b Tong, Y., Yamaguchi, H., Giaime, E., Boyle, S., Kopan, R., Kelleher, R. J., 3rd, & Shen, J. (2010). Loss of leucine-rich repeat kinase 2 causes impairment of protein degradation pathways, accumulation of alpha-synuclein, and apoptotic cell death in aged mice. Proc Natl Acad Sci U S A, 107(21), 9879-9884. doi: 10.1073/pnas.1004676107 Trinh, Joanne, & Farrer, Matt. (2013). Advances in the genetics of Parkinson disease. Nature reviews Neurology, 9(8), 445-454. Trueb, B. (2011). Biology of FGFRL1, the fifth fibroblast growth factor receptor. Cell Mol Life Sci, 68(6), 951-964. doi: 10.1007/s00018-010-0576-3 Trueb, B., Zhuang, L., Taeschler, S., & Wiedemann, M. (2003). Characterization of FGFRL1, a novel fibroblast growth factor (FGF) receptor preferentially expressed in skeletal tissues. J Biol Chem, 278(36), 33857-33865. doi: 10.1074/jbc.M300281200 Tsika, E., Moysidou, M., Guo, J., Cushman, M., Gannon, P., Sandaltzopoulos, R., . . . Mazzulli, J. R. (2010). Distinct region-specific alpha-synuclein oligomers in A53T transgenic mice: implications for neurodegeneration. J Neurosci, 30(9), 3409-3418. doi: 10.1523/JNEUROSCI.4977-09.2010 Turner, N., & Grose, R. (2010). Fibroblast growth factor signalling: from development to cancer. Nat Rev Cancer, 10(2), 116-129. doi: 10.1038/nrc2780 Uéda, K., Fukushima, H., Masliah, E., Xia, Y., Iwai, A., Yoshimoto, M., . . . Saitoh, T. (1993). Molecular cloning of cDNA encoding an unrecognized component of amyloid in Alzheimer disease. Proceedings of the National Academy of Sciences of the United States of America, 90(23), 11282-11286. doi: 10.1073/pnas.90.23.11282 van der Putten, H., Wiederhold, K. H., Probst, A., Barbieri, S., Mistl, C., Danner, S., . . . Bilbe, G. (2000). Neuropathology in mice expressing human alpha-synuclein. J Neurosci, 20(16), 6021-6029. Van Dyke, M. W., & Dervan, P. B. (1983). Chromomycin, mithramycin, and olivomycin binding sites on heterogeneous deoxyribonucleic acid. Footprinting with (methidiumpropyl-EDTA)iron(II). Biochemistry, 22(10), 2373-2377. Vassar, R., Bennett, B. D., Babu-Khan, S., Kahn, S., Mendiaz, E. A., Denis, P., . . . Citron, M. (1999). Beta-secretase cleavage of Alzheimer's amyloid precursor protein by the transmembrane aspartic protease BACE. Science, 286(5440), 735-741. 188 Venda, Lara, Cragg, Stephanie, Buchman, Vladimir, & Wade-Martins, Richard. (2010). α-Synuclein and dopamine at the crossroads of Parkinson's disease. Trends in neurosciences, 33(12), 559-568. doi: 10.1016/j.tins.2010.09.004 Vigo-Pelfrey, C., Lee, D., Keim, P., Lieberburg, I., & Schenk, D. B. (1993). Characterization of beta-amyloid peptide from human cerebrospinal fluid. J Neurochem, 61(5), 1965-1968. Walsh, Dominic, Klyubin, Igor, Fadeeva, Julia, Cullen, William, Anwyl, Roger, Wolfe, Michael, . . . Selkoe, Dennis. (2002). Naturally secreted oligomers of amyloid beta protein potently inhibit hippocampal long-term potentiation in vivo. Nature, 416(6880), 535-539. doi: 10.1038/416535a Wang, J., Duhart, H. M., Xu, Z., Patterson, T. A., Newport, G. D., & Ali, S. F. (2008). Comparison of the time courses of selective gene expression and dopaminergic depletion induced by MPP+ in MN9D cells. Neurochem Int, 52(6), 1037-1043. doi: 10.1016/j.neuint.2007.10.017 Wang, J., & Song, W. (2016). Regulation of LRRK2 promoter activity and gene expression by Sp1. Mol Brain, 9, 33. doi: 10.1186/s13041-016-0215-5 Wang, Jianyong, Xu, Zengjun, Fang, Hong, Duhart, Helen, Patterson, Tucker, & Ali, Syed. (2007). Gene expression profiling of MPP+-treated MN9D cells: a mechanism of toxicity study. Neurotoxicology, 28(5), 979-987. doi: 10.1016/j.neuro.2007.02.013 Wang, K., Liu, S., Wang, J., Wu, Y., Cai, F., & Song, W. (2014). Transcriptional regulation of human USP24 gene expression by NF-kappa B. Journal of neurochemistry, 128(6), 818-828. doi: 10.1111/jnc.12626 Wang, L., Xie, C., Greggio, E., Parisiadou, L., Shim, H., Sun, L., . . . Cai, H. (2008). The chaperone activity of heat shock protein 90 is critical for maintaining the stability of leucine-rich repeat kinase 2. J Neurosci, 28(13), 3384-3391. doi: 10.1523/JNEUROSCI.0185-08.2008 Wang, R., Luo, Y., Ly, P. T., Cai, F., Zhou, W., Zou, H., & Song, W. (2012a). Sp1 regulates human huntingtin gene expression. Journal of molecular neuroscience : MN, 47(2), 311-321. doi: 10.1007/s12031-012-9739-z Wang, R., Luo, Y., Ly, P. T., Cai, F., Zhou, W., Zou, H., & Song, W. (2012b). Sp1 regulates human huntingtin gene expression. J Mol Neurosci, 47(2), 311-321. doi: 10.1007/s12031-012-9739-z West, A. B., Moore, D. J., Biskup, S., Bugayenko, A., Smith, W. W., Ross, C. A., . . . Dawson, T. M. (2005). Parkinson's disease-associated mutations in leucine-rich repeat kinase 2 augment kinase activity. Proc Natl Acad Sci U S A, 102(46), 16842-16847. doi: 10.1073/pnas.0507360102 West, A. B., Moore, D. J., Choi, C., Andrabi, S. A., Li, X., Dikeman, D., . . . Dawson, T. M. (2007). Parkinson's disease-associated mutations in LRRK2 link enhanced GTP-binding and kinase activities to neuronal toxicity. Hum Mol Genet, 16(2), 223-232. doi: 10.1093/hmg/ddl471 Whitehouse, P. J., Price, D. L., Struble, R. G., Clark, A. W., Coyle, J. T., & Delon, M. R. (1982). Alzheimer's disease and senile dementia: loss of neurons in the basal forebrain. Science, 215(4537), 1237-1239. Whitehouse, P., Price, D., Clark, A., Coyle, J., & DeLong, M. (1981). Alzheimer disease: evidence for selective loss of cholinergic neurons in the nucleus basalis. Annals of neurology, 10(2), 122-126. doi: 10.1002/ana.410100203 189 Wiedemann, M., & Trueb, B. (2000). Characterization of a novel protein (FGFRL1) from human cartilage related to FGF receptors. Genomics, 69(2), 275-279. doi: 10.1006/geno.2000.6332 William, R. Markesbery. (1997). Oxidative Stress Hypothesis in Alzheimer's Disease. Free Radical Biology and Medicine, 23. doi: 10.1016/S0891-5849(96)00629-6 Xia, W., Zhang, J., Kholodenko, D., Citron, M., Podlisny, M. B., Teplow, D. B., . . . Selkoe, D. J. (1997). Enhanced production and oligomerization of the 42-residue amyloid beta-protein by Chinese hamster ovary cells stably expressing mutant presenilins. J Biol Chem, 272(12), 7977-7982. Xiong, Y., Coombes, C. E., Kilaru, A., Li, X., Gitler, A. D., Bowers, W. J., . . . Moore, D. J. (2010). GTPase activity plays a key role in the pathobiology of LRRK2. PLoS Genet, 6(4), e1000902. doi: 10.1371/journal.pgen.1000902 Xu, J., Kao, S. Y., Lee, F. J., Song, W., Jin, L. W., & Yankner, B. A. (2002). Dopamine-dependent neurotoxicity of alpha-synuclein: a mechanism for selective neurodegeneration in Parkinson disease. Nat Med, 8(6), 600-606. doi: 10.1038/nm0602-600 Xu, Q., Guo, H., Zhang, X., Tang, B., Cai, F., Zhou, W., & Song, W. (2012). Hypoxia regulation of ATP13A2 (PARK9) gene transcription. Journal of neurochemistry, 122(2), 251-259. doi: 10.1111/j.1471-4159.2012.07676.x Xu, Q., Shenoy, S., & Li, C. (2012). Mouse models for LRRK2 Parkinson's disease. Parkinsonism Relat Disord, 18 Suppl 1, S186-189. doi: 10.1016/S1353-8020(11)70058-X Yan, R., Bienkowski, M. J., Shuck, M. E., Miao, H., Tory, M. C., Pauley, A. M., . . . Gurney, M. E. (1999). Membrane-anchored aspartyl protease with Alzheimer's disease beta-secretase activity. Nature, 402(6761), 533-537. doi: 10.1038/990107 Yavich, L., Oksman, M., Tanila, H., Kerokoski, P., Hiltunen, M., van Groen, T., . . . Jakala, P. (2005). Locomotor activity and evoked dopamine release are reduced in mice overexpressing A30P-mutated human alpha-synuclein. Neurobiol Dis, 20(2), 303-313. doi: 10.1016/j.nbd.2005.03.010 Yavich, L., Tanila, H., Vepsalainen, S., & Jakala, P. (2004). Role of alpha-synuclein in presynaptic dopamine recruitment. J Neurosci, 24(49), 11165-11170. doi: 10.1523/JNEUROSCI.2559-04.2004 Ye, C. P., Selkoe, D. J., & Hartley, D. M. (2003). Protofibrils of amyloid beta-protein inhibit specific K+ currents in neocortical cultures. Neurobiol Dis, 13(3), 177-190. Youn, H., Jeoung, M., Koo, Y., Ji, H., Markesbery, W. R., Ji, I., & Ji, T. H. (2007). Kalirin is under-expressed in Alzheimer's disease hippocampus. J Alzheimers Dis, 11(3), 385-397. Youssef, I., Florent-Bechard, S., Malaplate-Armand, C., Koziel, V., Bihain, B., Olivier, J. L., . . . Pillot, T. (2008). N-truncated amyloid-beta oligomers induce learning impairment and neuronal apoptosis. Neurobiol Aging, 29(9), 1319-1333. doi: 10.1016/j.neurobiolaging.2007.03.005 Yu, Shun, Zuo, Xiaohong, Li, Yaohua, Zhang, Chen, Zhou, Ming, Zhang, Yu, . . . Chan, Piu. (2004). Inhibition of tyrosine hydroxylase expression in alpha-synuclein-transfected dopaminergic neuronal cells. Neuroscience letters, 367(1), 34-39. doi: 10.1016/j.neulet.2004.05.118 Zhang, Y., Dawson, V., & Dawson, T. (2000). Oxidative stress and genetics in the pathogenesis of Parkinson's disease. Neurobiology of disease, 7(4), 240-250. doi: 10.1006/nbdi.2000.0319 190 Zhang, Yanli, James, Michael, Middleton, Frank, & Davis, Richard. (2005). Transcriptional analysis of multiple brain regions in Parkinson's disease supports the involvement of specific protein processing, energy metabolism, and signaling pathways, and suggests novel disease mechanisms. American journal of medical genetics. Part B, Neuropsychiatric genetics : the official publication of the International Society of Psychiatric Genetics, 137B(1), 5-16. doi: 10.1002/ajmg.b.30195 Zheng, B., Liao, Z., Locascio, J. J., Lesniak, K. A., Roderick, S. S., Watt, M. L., . . . Global, P. D. Gene Expression Consortium. (2010). PGC-1alpha, a potential therapeutic target for early intervention in Parkinson's disease. Sci Transl Med, 2(52), 52ra73. doi: 10.1126/scitranslmed.3001059 Zimprich, A., Biskup, S., Leitner, P., Lichtner, P., Farrer, M., Lincoln, S., . . . Gasser, T. (2004). Mutations in LRRK2 cause autosomal-dominant parkinsonism with pleomorphic pathology. Neuron, 44(4), 601-607. doi: 10.1016/j.neuron.2004.11.005 Zojer, N., Keck, A. V., & Pecherstorfer, M. (1999). Comparative tolerability of drug therapies for hypercalcaemia of malignancy. Drug Saf, 21(5), 389-406.