This electronic thesis or dissertation has been downloaded from Explore Bristol Research, http://research-information.bristol.ac.uk

Author: Khazim, Mahmoud Title: Investigating synucleinopathies using translating ribosome affinity purification in cellular Parkinson’s disease models

General rights Access to the thesis is subject to the Creative Commons Attribution - NonCommercial-No Derivatives 4.0 International Public License. A copy of this may be found at https://creativecommons.org/licenses/by-nc-nd/4.0/legalcode This license sets out your rights and the restrictions that apply to your access to the thesis so it is important you read this before proceeding. Take down policy Some pages of this thesis may have been removed for copyright restrictions prior to having it been deposited in Explore Bristol Research. However, if you have discovered material within the thesis that you consider to be unlawful e.g. breaches of copyright (either yours or that of a third party) or any other law, including but not limited to those relating to patent, trademark, confidentiality, data protection, obscenity, defamation, libel, then please contact [email protected] and include the following information in your message:

•Your contact details •Bibliographic details for the item, including a URL •An outline nature of the complaint

Your claim will be investigated and, where appropriate, the item in question will be removed from public view as soon as possible.

Investigating synucleinopathies using translating ribosome affinity purification in cellular Parkinson’s disease models

Mahmoud Rabi Khazim Bristol Medical School January 2019

A dissertation submitted to the University of Bristol in accordance with the requirements for award of the degree of Doctorate of Philosophy in the Faculty of Translational Health Sciences

Word count: 46,111

i

Abstract Parkinson’s disease (PD) is a neurodegenerative disease characterised by accumulation of α- synuclein protein and loss of midbrain dopaminergic (mDA) neurons of the Substantia nigra (SN), resulting in motor dysfunction. Mutations and multiplications of the SNCA locus, which encodes α-synuclein, have been associated with familial Parkinson’s Disease. Using lentiviruses to overexpress wild-type and mutant α-synuclein in mDA neuron models, the transcriptomic and functional changes associated with increased α-synuclein burden were investigated.

The SH-SY5Y cell line was employed to study gene expression, by profiling actively translated mRNAs isolated by translating ribosome affinity purification (TRAP). Differential expression analysis in α-synuclein overexpressing cells identified actin cytoskeleton, mitochondrial and oxidative stress response proteins, as well as a key protein involved in miRNA biogenesis, DGCR8, to be key pathways affected by α-synuclein overexpression. Validation using molecular biology techniques was able to partially validate some of these targets. mDA neurons derived from human induced pluripotent stem cells (iPSC) were employed to study functional phenotypes, differentiation capacity and neurite dynamics, in response to overexpression of wild-type and mutant α-synuclein. Whilst the capacity for iPSC cells to differentiate into mDA neurons expressing mature neuronal and dopaminergic markers was not affected, both neurite extension and axonal branching was seen to be reduced in α- synuclein overexpressing cultures. Furthermore, these models were used to interrogate alterations in expression of microRNAs (miRNA) associated with mDA function and PD pathogenesis. Intracellular miRNA analysis resulted in the identification of differential expression of miRs-34b, -34c, -155 and -494, known to be associated with regulation of neuroinflammation, oxidative stress response and post-transcriptional regulation of α- synuclein mRNA.

The data presented implicates several pathways, genes and miRNA that are perturbed in response to α-synuclein overexpression. Further study of these pathways will contribute to our understanding of mechanisms underlying α-synuclein pathology and may represent potential therapeutic targets for PD.

ii

Dedication & Acknowledgements First and foremost, I would like to thank my supervisors Dr. Liang-Fong Wong and my secondary supervisor Professor James Uney for the opportunity to study for this PhD in their lab. I feel so lucky to have joined a lab with such a supportive and collaborative environment. Thanks also to Dr. Oscar Cordero Llana who became an unofficial supervisor. You all truly inspired me, and I appreciate the guidance, time and the sharing of your knowledge and expertise with me so much.

Thank you to my colleagues in the Uney/Wong/Llana lab. Every single one of you has helped me in some way or another and made this such a positive experience. Thanks to Dr. Helen Scott and Dr. Jess Gaunt for their advice throughout TRAP. Dr Hadil Al-Ahdal and Dr. Jalilah Idris for teaching me were everything was in the lab. And to my mentor Dr. Liang-Fong Wong, I am so grateful for all your help in the planning and implementation of the experiments throughout. Thanks also to all the students I’ve supervised in the lab who were such a pleasure to teach.

A big thanks also to the Caldwell group and stem cell lab colleagues. Thanks particularly to Dr. Paul Nistor, Dr. Petros Stathakos and Dr. Fella Hammachi for their warm advice and the collaborative atmosphere they created. Thanks to Dr. Abdelkader Essafi and his lab for all the advice and help with RNA-sequencing and making it feel simple. Thanks to Lorena Sueiro Ballesteros and Dr Andrew Herman for their help with all things flow cytometry and the Wolfson facility staff for their help with all things microscopy. I was always happy to book a session at both facilities. Thanks to all the admin and support staff for their help whenever anything needed to be fixed, purchased or prepared. Without the help of Doug, Dave, Bill and Andy, nothing could have been achieved. A massive thanks to all the CMM colleagues in B and C floor offices for making waking up and coming in everyday fun and exciting. And a special mention to my friends Dr. Alice Sherrard and Dr. Ahmed Basri for all the laughs.

To my family and community who have shown me so much love and support, I am eternally grateful. To my friends Owen, Chi, Gareth, Chris, Jason, Siobhan and Khobir. Finally, this thesis is dedicated to my family. To my little sisters Tala and Lina who I am so lucky to have had supporting and cheering for me every step of the way. And to my father Dr. Rabi Khazim and my mother Riaf Fahs, all of this was possible because of you. Thanks to my mother for all her spiritual and logistical support throughout. And to my father who was the one who believed in me and encouraged me to do this PhD. I can never thank you enough for everything.

iii

Authors declaration

The CMV-EGFP, PGK-EGFP lentiviral backbone plasmids were gifts from Dr. Liang-Fong Wong and the CMV-α-synucleinWT, CMV -α-synucleinA53T and the CMV-L10a-EGFP lentiviral backbone plasmids were gifts from Dr. Helen Scott. .

NAS2 iPSC line was a gift from Dr. Tilo Kunath’s lab in Edinburgh University

TEM preparation and staining was performed by an undergraduate student Bryony McPherson

NTA was performed by collaborator Dr. Andrew Shearn

FACS sorting was performed with Dr. Andrew Herman

Neurite analysis immunostaining and neurite analysis was performed by undergraduate student Maria Otero Jimenez

I declare that the work in this dissertation was carried out in accordance with the requirements of the University's Regulations and Code of Practice for Research Degree Programmes and that it has not been submitted for any other academic award. Except where indicated by specific reference in the text, the work is the candidate's own work. Work done in collaboration with, or with the assistance of, others, is indicated as such. Any views expressed in the dissertation are those of the author.

SIGNED: ...... DATE: 14/01/2018

iv

Table of contents Abstract ...... ii Dedication & Acknowledgements ...... iii Authors declaration ...... iv Table of contents ...... v List of figures ...... xi List of tables ...... xiii Chapter 1 – General Introduction ...... 1 1.1 Parkinson’s disease: A historical perspective ...... 1 1.1.1 Parkinson’s disease classification and disorder-anatomy associations ...... 1 1.1.2 The discovery of dopamine ...... 4 1.1.3 The development of the first targetted Parkinson’s disease treatment ...... 5 1.1.4 Identification of the nigro-striatal dopaminergic pathway ...... 6 1.2 Parkinson’s disease classification and current treatments ...... 8 1.2.1 Parkinson’s disease symptoms and incidence ...... 8 1.2.2 Current treatment options for PD...... 10 1.3 Midbrain Dopaminergic (mDA) neurons ...... 11 1.3.1 Development of the midbrain dopaminergic neurons ...... 11 1.3.1.1 Neuralisation ...... 11 1.3.1.2 Specification of differentiating neural tube cells towards a midbrain fate ...... 14 1.4 Alpha-Synuclein...... 19 1.4.1 Association of α-synuclein with familial PD and Lewy bodies ...... 19 1.4.2 Clinical heterogeneity of α-synuclein mutations ...... 20 1.4.3 Modifications: truncation, phosphorylation and nitration ...... 23 1.4.4 Spreading of α-synuclein ...... 24 1.4.5 Structure and aggregation properties of α-synuclein ...... 27 1.4.6 Protein degradation pathways ...... 29 1.4.7 Function of α-synuclein ...... 30 1.4.7.1 Neurotransmitter release ...... 30 1.4.7.2 Mitochondria and Reactive oxygen species ...... 32 1.4.7.3 ER-Golgi trafficking ...... 32 1.4.7.4 Axonal transport and neurite dynamics ...... 33 1.4.7.5 Neuroinflammation ...... 34 1.5 Genetic loci associated with PD ...... 35

v

1.5.1 LRRK2 ...... 36 1.5.2 Parkin ...... 36 1.5.3 PINK1 ...... 37 1.5.4 DJ1 ...... 38 1.6 Modelling of Parkinson’s disease ...... 38 1.6.1 Classic neurotoxin models ...... 39 1.6.2 Genetic mouse models ...... 40 1.6.3 In vitro modelling of Parkinson’s disease ...... 41 1.6.4 Pluripotent stem cell models: hESC and hiPSC ...... 42 1.6.4.1 Differentiation of stem cells towards a midbrain dopaminergic fate ...... 44 1.6.5 Utility of lentiviral vectors for disease modelling ...... 47 1.7 Summary and objectives ...... 50 Chapter 2 - Materials and Methods ...... 52 2.1 Generating lentiviral plasmids and particles for affinity purification, promoter testing and disease modelling ...... 52 2.1.1 Polymerase chain reaction (PCR) ...... 53 2.1.2 Electrophoresis of PCR product ...... 53 2.1.3 DNA extraction from Gel and determination of concentrations ...... 54 2.1.4 Restriction enzyme digestion of DNA for recombinant DNA cloning ...... 55 2.1.5 Ligation of DNA ...... 56 2.1.6 Transformation into chemically competent Escherichia coli ...... 57 2.1.7 Maxi Preparations ...... 58 2.1.8 Culture of HEK293T cells and lentiviral preparation ...... 60 2.1.9 Lentiviral titering by fluorescence activated cell sorting (FACS) ...... 62 2.1.10 Lentiviral titering by Taqman qPCR ...... 63 2.2 Generating cellular PD models ...... 63 2.2.1 Culturing SH-SY5Y cells...... 64 2.2.2 Differentiation of SH-SY5Y cultures ...... 64 2.2.3 Culturing of iPSCs ...... 65 2.2.4 Differentiation of iPSCs towards a midbrain dopaminergic fate ...... 66 2.2.5 Immunoflourescence staining in SH-SY5Y and iPSC cells ...... 68 2.2.6 Western Blot for characterisation of -synuclein expression in cellular models ...... 69 2.2.7 RNA extraction and cDNA generation ...... 71 2.3 Statistical analysis ...... 72

vi

Chapter 3 – Cellular models of Parkinson’s disease ...... 73 3.1 Introduction ...... 73 3.1.1 – Aims ...... 76 3.2 Materials and Methods ...... 77 3.2.1 Promoter testing for stable long-term expression of transgenes in SH-SY5Y cells and iPSCs ...... 77 3.2.2 Transduction of SH-SY5Y and iPSC-derived neurons cells with α-synuclein ...... 78 3.2.3 qPCR analysis of differentiation markers in differentiating iPSC-derived neurons ...... 79 3.2.4 Functional analysis of neurite dynamics of iPSC-derived neurons ...... 80 3.3 Results ...... 81 3.3.1 Generating a SH-SY5Y model of PD for translating ribosome affinity purification ...... 81 3.3.1.1 Optimisation of promoter usage in lentiviral vector constructs...... 81 3.3.1.2 Promoter optimistation in differentiated SH-SY5Y cells ...... 83 3.3.1.3 Expression of WT and A53T mutant α-synuclein in SH-SY5Y cells ...... 85 3.3.1.4 Production of synuclein SH-SY5Y cells expressing L10a-EGFP ...... 86 3.3.2 Generating an iPSC model of PD ...... 88 3.3.2.1 Preliminary differentiation of iPSCs towards a midbrain dopaminergic cell fate ...... 88 3.3.2.2 Lentiviral vector transduction of iPSC-derived dopaminergic neurons ...... 89 3.3.2.3 Transduction of WT, A53T and G51D α-synuclein into differentiating iPSCs and characterising differentiation for miRNA profiling ...... 92 3.3.3 Neurite dynamics in iPSC-derived dopaminergic neurons...... 96 3.4 Discussion ...... 99 3.5 Conclusion ...... 104 Chapter 4 – Translating ribosome affinity purification and RNA sequencing to study genome-wide gene expression changes induced by α-synuclein ...... 106 4.1 Introduction ...... 106 4.1.1 Next generation sequencing and Parkinson’s disease research ...... 106 4.1.2 Translatome profiling techniques and advantages over transcriptome analysis ...... 108 4.1.3 Library preparation for raw data sequencing acquisition ...... 110 4.1.4 RNA-seq analysis analysis platform and pipelines ...... 111 4.1.5 Aims...... 114 4.2 Materials and Methods ...... 116 4.2.1 Lentiviral transduction of L10a-EGFP into SH-SY5Y PD models ...... 116 4.2.2 Translating ribosome affinity purification and mRNA purification ...... 116

vii

4.2.3 RNA amplification and cDNA library preparation for next generation sequencing on illumina platform ...... 118 4.2.4 Gene expression analysis ...... 121 4.2.5 – Validation of differentially expressed genes by transcript expression analysis by RT-qPCR using SYBR Green ...... 122 4.2.6 – Protein expression analysis using Western Blot ...... 123 4.3 Results ...... 125 4.3.1 TRAP and library preparation ...... 125 4.3.1.1 Translating ribosome affinity purification ...... 125 4.3.1.2 NGS library preparation ...... 125 4.3.2 RNA-seq Analysis ...... 127 4.3.2.1 Raw read general read statistics ...... 127 4.3.2.2 Alignment and general alignment statistics...... 128 4.3.2.3 Differential expression analysis ...... 130 4.3.2.4 Pathway analysis ...... 134 4.3.3 Validation of Differentially expressed genes determined by RNA-seq ...... 138 4.3.3.1 Evidence based choice of suitable housekeeping gene ...... 138 4.3.3.2 – Validation of DEGs ...... 139 4.3.3.2.1 – Cytoskeletal protein encoding genes ...... 140 4.3.3.2.2 – Mitochondrial protein encoding genes ...... 143 4.3.3.2.3 – Protein transport and dopamine signalling protein encoding genes ...... 146 4.3.3.2.4 – Oxidative stress response and apoptosis signalling encoding genes ...... 149 4.3.3.2.5 – microRNA biogenesis encoding gene ...... 152 4.3.3.2.6 – Conclusions from validation studies ...... 153 4.4 Discussion ...... 154 4.5 - Conclusions ...... 158 Chapter 5 - miRNA profiling and differential expression analysis in iPSC-derived neurons ...... 160 5.1 Introduction ...... 160 5.1.1 miRNA biogenesis and function in mDA neurons ...... 160 5.1.2 miRNAs associated with PD mechanisms ...... 162 5.1.2.1 miRNA regulation of familial PD-associated genes ...... 162 5.1.2.2 miRNAs in mDA development and function ...... 164 5.1.2.3 miRNAs associated with neuroinflammation ...... 165 5.1.3 Extracellular miRNA ...... 167

viii

5.1.3.1 Exosomes ...... 168 5.1.4 - Aims ...... 171 5.2 Materials and methods ...... 173 5.2.1 Isolation of exosomes from iPSC-derived neuronal cultures ...... 173 5.2.2 Characterisation of isolated exosomes by Transmission Electron Microscopy ...... 174 5.2.3 Isolation of intracellular and exosomal miRNA ...... 175 5.2.4 cDNA generation and qPCR expression profiling by Taqman small RNA assays for intracellular and extracellular miRNA ...... 175 5.2.5 Nanoparticle tracking analysis (NTA) for determining isolated exosomes from iPSC-derived PD models ...... 176 5.3 Results ...... 177 5.3.1 Differential miRNA expression in iPSC-derived PD models ...... 177 5.3.1.1 – Selection of miRNA for profiling ...... 177 5.3.1.2 – Selection of housekeeping gene for normalisation of differential expression analysis ...... 178 5.3.1.3 – Differential expression analysis of miRNAs involved in direct regulation of familial PD-associated genes ...... 179 5.3.1.4 – Differential expression analysis of miRNAs involved in mDA development and function ...... 181 5.3.1.5 – Differential expression analysis of miRNAs involved in neuroinflammatory processes ...... 181 5.3.1.6 Summary of intracellular miRNA analysis ...... 182 5.3.2 Characterisation of extracellular vesicles isolated from cultured media from iPSC-derived neuronal PD models ...... 183 5.3.2.1 Transmission electron microscopy ...... 183 5.3.2.2 Nanoparticle tracking analysis ...... 184 5.3.3 Exosomal miRNA profiling isolated from iPSC-derived neuronal culture conditioned media ...... 186 5.4 Discussion ...... 188 5.4.1 Intracellular miRNA profiling in iPSC-derived PD models ...... 188 5.4.2 Extracellular miRNA profiling in iPSC-derived PD models ...... 192 5.5 Conclusion ...... 193 Chapter 6 General Discussion ...... 195 6.1 - Parkinson’s disease and α-synuclein ...... 195 6.2 - Development of a SH-SY5Y model for translating ribosome affinity purification ...... 195 6.3 - Translating ribosome affinity purification, RNA-seq library preparation and analysis ...... 196

ix

6.4 Development of an iPSC-model of synucleinopathy...... 199 6.5 Functional characterisation of neurite dynamics ...... 200 6.6 Intracellular and extracellular miRNA profiling ...... 200 6.7 Conclusions and future work ...... 201 References ...... 203 Appendix 1 - SHSY5Y transduction with EGFP – flow cytometry - promoter test ...... 241 Appendix 2 – SHSY5Y promoter test – FACS sorting for EGFP expression ...... 246 Appendix 3 – SHSY5Y PD model and control cells transduced with CMV-L10a-EGFP lentivirus particles (MOI 2) – FACS sorted for EGFP expression ...... 250 Appendix 4 – RNA integrity of affinity purified samples from SHSY5Y PD model ...... 258 Appendix 5 – RNA amplification – dsDNA...... 259 Appendix 6 – Universal/index adapters, amplification primers and P5/P7 adapter sequences .... 265 Appendix 7 – Adapter ligated Amplicon QC ...... 266

x

List of figures Figure 1.1: Symptoms and anatomical abnormalities in PD……………………………………………………………….3 Figure 1.2: Catecholamine synthesis and metabolism pathways………………………………………………………..5 Figure 1.3: Reduction of dopamine in PD…………………………………………………………………………………………..7 Figure 1.4: Dopaminergic neurons produce dense dendritic arborisations………………………………………..7 Figure 1.5: Immunohistochemical labelling of midbrain dopaminergic neurons for α-synuclein………..9 Figure 1.6: Schematic representing the three embryonic vesicles of the neural tube that develop into the five secondary vesicles and their adult derivatives…………………………………………………………………….12 Figure 1.7: Developmental pathways directing towards a neural fate……………………………………………..14 Figure 1.8: Floorplate and mDA development………………………………………………………………………………….18 Figure 1.9: Mutations in SNCA………………………………………………………………………………………………………….21 Figure 1.10: Schematic representing the spreading and propagation of α-synuclein LN/LB pathology in idiopathic PD………………………………………………………………………………………………………………………………..25 Figure 1.11: Amyloid fibril formation and aggregation dynamics……………………………………………………..29 Figure 1.12: Pathways associated with α-synuclein pathology…………………………………………………………31 Figure 1.13: Potential interactions between PD-associated disease genes……………………………………….35 Figure 1.14: Sources for cell replacement therapies and modelling of Parkinson’s disease………………44 Figure 1.15: Differentiation towards specific neural cell types…………………………………………………………46 Figure 1.16: Schematic representing the structure of a lentivirus and the plasmids used to express the different components………………………………………………………………………………………………………………..50 Figure 2.1: Lentiviral backbone plasmids………………………………………………………………………………………….52 Figure 2.2: Cloning strategy for cloning EF1α promoter into a plasmid which contains the CMV promoter to express EGFP……………………………………………………………………………………………………………….56 Figure 2.3: Schematic of the self-inactivating (SIN) lentiviral backbone plasmid………………………………60 Figure 2.4: Schematic outlining the SH-SY5Y differentiation protocol towards an mDA phenotype…65 Figure 2.5: Schematic outlining the iPSC differentiation protocol towards an mDA phenotype……….67 Figure 3.1: Neurite analysis in iPSC-derived neurons……………………………………………………………………….80 Figure 3.2: Lentivirus promoter transduction efficiencies………………………………………………………………..82 Figure 3.3: Differentiation of SH-SY5Y cells………………………………………………………………………………………83 Figure 3.4: Promoter testing in SH-SY5Y cells…………………………………………………………………………………..84 Figure 3.5: α-synuclein protein expression analysis in SH-SY5Y cells………………………………………………..86 Figure 3.6: Representative FACS tracing showing gating for the isolation of EGFP expressing cells from experimental samples……………………………………………………………………………………………………………..87 Figure 3.7: Differentiation of NAS2 iPSCs into mDA neurons……………………………………………………………89 Figure 3.8: Differentiated iPSC neurons expressing EGFP from lentiviral vectors containing different promoters……………………………………………………………………………………………………………………………………….91 Figure 3.9: Characterisation of day 30 differentiating iPSC-models………………………………………………….94 Figure 3.10: Immunofluorescent staining of experimental cultures at day 30 of differentiation………95 Figure 3.11: Neurite analysis of iPSC-derived dopaminergic neurons and characterisation of day 35 differentiating iPSC neurons…………………………………………………………………………………………………………….98 Figure 4.1: Schematic outlining the processing and steps towards the generation of differential expression analysis from TRAP-derived mRNA……………………………………………………………………………….111 Figure 4.2: Schematic outlining pipeline for RNA-seq raw data towards differential expression analysis using Tuxedo pipeline……………………………………………………………………………………………………….113 Figure 4.3: Schematic outlining protocol for extracting mRNAs in translating polysomes……………...116 Figure 4.4: Schematic outlining the processing of TRAP samples from amplification of RNA to library preparation and pooling towards sequencing………………………………………………………………………………..119 Figure 4.5: Sequences of the universal adapter and index adapter………………………………………………..120 Figure 4.6: Schematic outlining pipeline for RNA-seq raw data towards differential expression analysis using Partek Flow and Galaxy platform using the Tuxedo pipeline……………………………………122 Figure 4.7: Fragment size selection on pooled adapter ligated amplicon……………………………………….126

xi

Figure 4.8: Outline of FASTQC module summary output page……………………………………………………….127 Figure 4.9: Graph and table showing overall alignment statistics including percentage unique, multi- mapped and unmapped reads……………………………………………………………………………………………………….129 Figure 4.10: Pipeline analysis………………………………………………………………………………………………………...131 Figure 4.11: Top 10 significantly (p<0.05) up-regulated and down-regulated protein coding DEGs………………………………………………………………………………………………………………………………………………134 Figure 4.12: KEGG pathway analysis for DEGs which were >2 fold significant differentially expressed………………………………………………………………………………………………………………………………………136 Figure 4.13: Bestkeeper regression analysis of 3 candidate housekeeping genes: GAPDH, ACTB and HRPLP…………………………………………………………………………………………………………………………………………….138 Figure 4.14: RNA-seq analysis of differential expression of SNCA in A53T compared to WT overexpression and qPCR validation………………………………………………………………………………………………140 Figure 4.15: Gene expression analysis for DEGs which encode cytoskeletal proteins FGD3 and SPON2……………………………………………………………………………………………………………………………………………142 Figure 4.16: Gene expression and protein expression analysis for DEGs which encode mitochondrial proteins TOMM40 and SLC25A12………………………………………………………………………………………………….145 Figure 4.17: Gene expression and protein expression analysis for DEGs which encode protein transport and dopamine signalling proteins UNC199B and RGS8…………………………………………………..148 Figure 4.18: Gene expression and protein expression analysis for DEGs which encode oxidative stress response and apoptotic proteins FOXO1 and CRADD…………………………………………………………………….151 Figure 4.19: Gene expression and protein expression analysis for DGCR8 which encode a protein involved in miRNA biogenesis………………………………………………………………………………………………………..153 Figure 5.1: The miRNA processing pathway……………………………………………………………………………………161 Figure 5.2: miRNA biogenesis and release via exosomes………………………………………………………………..170 Figure 5.3: Schematic of the timeline of generation of iPSC-derived mDA neurons and characterisation of transduction and differentiation……………………………………………………………………..172 Figure 5.4: Bestkeeper regression analysis of 2 candidate housekeeping genes: RNU6B and RNU24……………………………………………………………………………………………………………………………………………179 Figure 5.5: Differential expression analysis for miRNAs involved in direct binding of familial PD- associated gene targets (miR-7, -34b, -34c and -494)…………………………………………………………………….180 Figure 5.6: Differential expression analysis for miRNAs involved in mDA development and function (miR-132, -133b and -137)……………………………………………………………………………………………………………..181 Figure 5.7: Differential expression analysis for miRNAs associated with modulation of neuroinflammation (miR-21 and -155)…………………………………………………………………………………………..182 Figure 5.8: Transmission electron microscopy (TEM) images of isolated extracellular vesicle-like structures………………………………………………………………………………………………………………………………………184 Figure 5.9: NTA analysis of conditioned media samples from day 32 differentiating iPSC-derived cultures………………………………………………………………………………………………………………………………………….185 Figure 5.10: Extracellular miRNA expression profiling…………………………………………………………………….187

xii

List of tables Table 1.1: A history of PD………………………………………………………………………………………………………………….1 Table 1.2: SNCA multiplications and mutations associated with PD and clinical presentations……….21 Table 1.3: The staging of α-synuclein pathology as described by Braak et al……………………………………24 Table 1.4: Table of neurotoxins used to generate models of PD……………………………………………………...39 Table 2.1: List of lentiviral backbone plasmids and lentiviral particles cloned and generated……….….52 Table 2.2 : Fast start high fidelity PCR kit reagents used for PCR of fragments for cloning and conditions used for the PCR reaction……………………………………………………………………………………………….53 Table 2.3: LB Broth components for growing transformed E. coli and Agar components for plating transformed E. Coli………………………………………………………………………………………………………………………….58 Table 2.4: List of solutions used to isolate plasmid DNA transformed and amplified by transformation into Stabl3 competent E. Coli by Maxi preparation………………………………………………………………………….60 Table 2.5: Quantities of each plasmid and media for lentiviral particle generation in HEK293T cells.62 Table 2.6: Taqman qPCR reaction components and cycling conditions for qPCR-based lentiviral titre determination………………………………………………………………………………………………………………………………….63 Table 2.7: Components used to make N2 supplement for differentiation of SH-SY5Y cells………………65 Table 2.8: Factors added to NVIM for specification of midbrain dopaminergic fate in iPSCs and factors added to NVMM to maintain neurons………………………………………………………………………………….67 Table 2.9: List of antibodies used for immunofluorescence analysis of cultured SH-SY5Y and iPSC- derived cultures……………………………………………………………………………………………………………………………….69 Table 2.10: Table listing components of RIPA buffer for lysis of cells and 2 x loading buffer for mixing with protein samples for loading onto SDS-PAGE gel……………………………………………………………………….70 Table 2.11: Composition of resolving and stacking gels for SDS-PAGE……………………………………………..71 Table 2.12: Composition of running buffer for resolution of SDS-PAGE proteins, transfer buffer for electrophoretic transfer of proteins onto PVDF membranes and blocking buffer…………………………….71 Table 2.13: antibodies used to probe protein expression with Western blotting……………………………..71 Table 2.14: Protocol for cDNA generation using Superscript III cDNA kit………………………………………….72 Table 2.15: SYBR green qPCR reaction and thermal cycling conditions…………………………………………….72 Table 3.1: List of primers used for qPCR expression profiling of differentiation markers in iPSC- derived neurons………………………………………………………………………………………………………………………………80 Table 3.2: Literature review of studies that have employed lentiviral vectors to express transgenes in neuronal models………………………………………………………………………………………………………………………………81 Table 4.1: Lysis and wash buffers utilised in TRAP methodology……………………………………………………117 Table 4.2: List of primers for validation of DEGs from RNA-seq analysis………………………………………..123 Table 4.3: Table shows the concentration and RIN of TRAP-derived mRNA, average dscDNA insert fragment size and average adapter ligated amplicon size and amplified dscDNA fragment analysis by capillary gel electrophoresis…………………………………………………………………………………………………………..125 Table 4.4: Gene ontology gene set enrichment analysis for cellular component related genes………137 Table 4.5: Gene ontology gene set enrichment analysis for biological function related genes...... 137 Table 5.1: Studies that have assessed differentially expressed extracellular miRNA profiles for PD.168 Table 5.2: List of control miRNAs and miRNAs of interest with roles associated with PD, development of mDA neurons and neuroinflammation……………………………………………………………………………………...176 Table 5.3: List of PD-associated miRNA targets chosen from literature with their proposed mechanism, models used and respective publications for screening in iPSC-derived PD models……177

xiii

Chapter 1 – General Introduction 1.1 Parkinson’s disease: A historical perspective

1.1.1 Parkinson’s disease classification and disorder-anatomy associations

Parkinson’s Disease (PD) is the second most prevalent neurodegenerative disorder after

Alzheimer’s disease (AD). The current understanding of the pathophysiology of PD has taken many years to unfold and is probably best understood from a historical perspective (Table

1.1).

Observation Implication Publication Patient with midbrain Identification of Substantia nigra (SNc) (Blocq et al. 1893) tuberculoma caused as a pathological site in PD Parkinsonism Synthesis of Dopamine Dopamine only thought of as precursor (Barger and Ewins to Adrenaline and Noradrenaline 1910) Described the characterstic Abnormal eosinophilic cytoplasmic (Lewy 1912) anatomical feature of PD ‘the inclusions consisting of radiating fibrils Lewy Body’ surrounding a dense core Detailed descriptions of Midbrain atrophy and destruction of (Foix et al. 1925) decrease in size and nerve cells associated with PD pigmentation of SNcp neurons in PD Discovery of substance “X” Substance “X” found to be identical to (Montagu 1957) found in varying concentrations dopamine in the brain Drug respirine causes severe loss L-DOPA (dopamine precursor) injection (Carlsson et al. 1957) of catecholamines, causing alleviates respirine induced tranquilisation tranquilisation Highest levels of dopamine in Distribution of dopamine in brain (Bertler et al. 1959; the mammalian brain were implicates a role for the corpus striatum Sano et al. 1959) found in the corpus striatum in the control of motor function Patients with PD showed severe Confirmation of the association of (Ehringer et al. 1960) dopamine deficits in the caudate dopamine deficiency within the caudate and putamen and putamen in PD Intravenous injections of Complete abolition or reduction in (Birkmayer et al. 1961) levodopa to 20 PD patients akinesia by drug therapy relieved symptoms Dopamine markedly reduced in Identification of nigrostriatal (Andén et al. 1964) neostriatum following lesions to dopaminergic pathway SNpc Table 1.1 – A history of PD. A history of the main discoveries which have led to the understanding of the anatomical and chemical disruptions associated with Parkinson’s disease

PD was first expounded in an essay entitled ‘An Essay on the Shaking Palsy” by James

Parkinson in 1817 (Parkinson 1817). Up to that point, “the shaking palsy” had been employed

1

by writers in general to describe cases of palsy accompanied by involuntary tremors. In his treatise, Parkinson defined this slow progressing disease by symptoms apparent at different stages of disease progression beginning with a “slight sense of weakness” with a “proneness to trembling in some particular part” with symptoms gradually increasing over the usually long duration of the disease. The “malady” is defined as:

“Involuntary tremulous motion, with lessened muscular power, in parts not in action

and even when supported; with a propensity to bend the trunk forward, and to pass

from a walking to a running pace: the senses and intellects being uninjured”

The work of the French neurologist Jean-Martin Charcot was instrumental to the characterisation of PD, identifying bradykinesia as a cardinal clinical feature, replacing weakness, as it was noted that patients had a slowness of execution and not marked weakness (Figure 1.1A). This was one of the reasons Charcot rejected the previous label of

“shaking palsy” and assigned it the eponymous title “Parkinson’s Disease”.

In 1893, a case study reported a 38 year old with PD, at autopsy, as having a hazelnut size enucleated tuberculoma which selectively damaged the substantia nigra (SN) (Blocq et al.

1893). This midbrain lesion was thought to cause the Parkinsonian symptoms. This led to the hypothesis that the SN was the major pathological site in PD and that lesions of the SN was directly related to aetiology of PD (Figure 1.1B). In the early 20th century, Friedrich Lewy identified and described the characteristic anatomical feature of PD, by performing exhaustive post-mortem examinations on PD patients. This was the abnormal eosinophilic cytoplasmic inclusions consisting of radiating fibrils surrounding a dense core (Lewy 1912).

2

In 1919 a Soviet scientist, Constantin Tretiakoff, noticed a marked loss of pigmented substantia nigra pars compacta (SNpc) neurons, as well as cell body swelling and neurofibrillary alterations whilst examining 54 autopsied brain sections. Surviving SN cells were noted as having inclusions which he named Lewy bodies, in recognition of Freidrich Lewy

(Trétiakoff 1919). These results were confirmed in 1925 (Foix et al. 1925) and SN lesions were studied in detail in idiopathic and post-encephalitic Parkinsonism in 1953 (Greenfield et al.

1953). However, it would not be until the discovery of the dopaminergic nature of the nigrostriatal neurons, and the characteristic depletion of dopamine in the basal ganglia in general, that a further understanding of the anatomical pathways relating to movement could be reached.

Figure 1.1 – Symptoms and anatomical abnormalities in PD. A) Motor abnormalities characteristic of Parkinson’s disease (Figure adapted from (Gowers 1901)). Illustration of a PD patient from William Gower seminal book ‘Manual of the Diseases of the Nervous System’. B) Schematic diagram representing normal, healthy midbrain dopaminergic neuron projections (red line) from the SN to the striatum (putamen and caudate nucleus) with black arrows point to normal neuromelanin pigmentation of midbrain SN dopaminergic neurons (left). Parkinson’s disease brains with a loss of midbrain SN dopaminergic neurons and subsequent loss of synaptic innervation to the striatum (dashed red line). Black arrows point to a loss of the neuromelanin containing midbrain SN dopaminergic neurons (Figure taken from (Dauer et al. 2003)).

3

1.1.2 The discovery of dopamine

Many important discoveries that would lead to the discovery of the dopaminergic nature of the basal ganglia, and in turn lead to the development of the first directed treatments for PD symptoms, were undertaken in parallel to the anatomy-disorder association studies

(described in section 1.1.1) (Table 1.1). The first major milestone in the understanding of the pathophysiology of PD was the synthesis of Dopamine (Barger and Ewins 1910) which was found to have weak sympathomimetic properties (Barger and Dale 1910). A year later, D,L- dopa was synthesised chemically (Funk 1911) and 2 years later, the dopamine precursor 3:4-

Dihydroxyphenylalanine (ʟ-Dopa) was isolated from fava beans (Vicia faba) (Guggenheirn

1913). However, for many years after its initial synthesis, this newly discovered catecholamine was thought to be only an intermediate product for the metabolism of adrenaline and noradrenaline (Figure 1.2). It was not until 1938, that the enzyme dopa decarboxylase (DDC) was discovered to produce dopamine from ʟ-Dopa in cells taken from mammalian tissue

(Holtz et al. 1938). Thus, it was in 1957, whilst determining levels of catecholamines in mammalian brain extracts using a sensitive quantitative fluorometric technique, that dopamine was identified as present in varying concentrations in brain tissue (Montagu 1957).

Up to this point, dopamine loss in the striatum was not yet considered to be related to the pathogenesis of PD. Later, two independent studies characterising regional distribution of dopamine in the mammalian brain found highest levels in the striatum (Bertler et al. 1959;

Sano et al. 1959). This observation led the investigators to conclude that dopamine may have a function, other than as an intermediate, associated with the function of the striatum and motor function.

4

Figure 1.2 – Catecholamine synthesis and metabolism pathways. Phenylalanine is converted to tyrosine by phenylalanine hydroxylase which, in turn, is converted to L-dopa by tyrosine hydroxylase (TH). Further processing by dopa decarboxylase (DDC) converts L-dopa to dopamine. Noradrenaline can then be synthesised from dopamine by the action of dopamine- β-hydroxylase. Finally, noradrenaline can be converted to adrenaline by the action of phenylethanolamine N- methytransferase (adapted from (Daubner et al. 2011)).

1.1.3 The development of the first targetted Parkinson’s disease treatment

The discovery of the anatomical and chemical disruptions associated with parkinsonism set the stage for the development of the first therapy for PD. In 1957 researchers investigating the drug Reserpine, which causes catecholamine depletion and induces tranquilisation in mice, discovered an important clue. Intraperitoneal injection of L-Dopa into catecholamine depleted mice was found to antagonise the action of Reserpine such that they were normal or even overactive upon injection (Carlsson et al. 1957). Furthermore, the intravenous addition of L-Dopa resulted in a marked increase in dopamine brain content and central excitation (Carlsson et al. 1958).

5

Further analyses of dopamine content in freshly dissected autopsied brain tissue from patients with motor symptoms (6 PD patients, 6 patients with symptoms of unknown aetiology and 2 Huntington’s patients) showed that only the PD patients had severe dopamine deficits in the striatum (Figure 1.1B)(Ehringer et al. 1960). Following this, a clinical trial was conducted to deliver intravenous injections of levodopa to 20 PD patients. The results were astonishing, as patients who received levodopa showed a complete abolition or reduction in akinesia. The effects lasted for 3 hours at full intensity and then dissipated over the next 24 hours (Birkmayer et al. 1961). It was not until 1967, with the publication of a seminal clinical paper that described effective administration of L-dopa therapy, that the era of dopamine replacement dawned. Here, Cotzias used larger, sustained oral doses of L-dopa, administered to induce sustained beneficial improvements in Parkinsonism symptoms such as tremor and rigidity (Cotzias et al. 1967). This was a huge milestone in clinical research and for the first time, perturbations in a chemical in a specific brain region was associated with a characterised degenerative disorder.

1.1.4 Identification of the nigro-striatal dopaminergic pathway

The discovery that PD is characterised by loss of dopamine in the basal ganglia greatly stimulated research on the basal ganglia, in particular the SNpc. Thus, it was in 1961, that the very first descriptions of striatopallidal-nigral degeneration in a case of PD was reported

(Adams et al. 1961). It was later found that lesions in the SNpc induced a reduction in striatal dopamine and the existence of a nigro-striatal dopaminergic pathway was identified, originating in the substantia nigra, through the internal capsule and terminating in the neostriatum (Figure 1.3) (Andén et al. 1964; Poirier et al. 1965). It was noted specifically that the reduction in dopamine content following SNpc lesions was akin to the reduction in dopamine that is observed in PD.

6

Figure 1.3 – Reduction of dopamine in PD. 18F-dopa PET images of the striatum in a healthy individual (left) and a patient with Parkinson’s disease (right) which shows a loss of the dopaminergic tracer in the putamen and caudate nucleus (Figure taken from (Piccini et al. 2004)).

In 1964, the first accurate maps of anatomical organisation of dopaminergic neuronal systems and their projections was mapped by Dahlstrom and Fuxe (Dahlström et al. 1964), who studied the localisation, distribution and morphology of amine-containing neurons (Falck et al. 1962). The catecholaminergic neurons of the mesencephalon were divided topographically into groups A8-A10, with the A9 group being of importance, as these were the neurons that send fibres to the neostriatum via the nigro-striatal pathway. Together, these midbrain dopaminergic (mDA) neurons represent approximately 75% of all DA neurons in the adult brain, with a single mDA neuron being able to form wide-spread and dense axonal branching in the neostriatum (Figure 1.4)(Matsuda et al. 2009).

Figure 1.4 – Dopaminergic neurons produce dense dendritic arborisations. Camera lucida reconstruction of a single rat SNpc neuron. 75,000 striatal neurons are estimated to be influenced by a single dopaminergic neuron (Matsuda et al. 2009).

7

Together, these studies represent the academic discoveries of the past 200 years that have led to a preliminary clinical understanding of the pathophysiology of PD. The genetic era would bring with it a tide of fast-paced academic literature on PD. These contemporary studies would be accompanied by advances in multiple disciplines, such as molecular biology, imaging and computer modelling, to bring about a deeper understanding of the aetiology and the molecular and cellular mechanisms responsible for the development of PD.

1.2 Parkinson’s disease classification and current treatments

1.2.1 Parkinson’s disease symptoms and incidence

PD is a clinical syndrome characterised by motor disturbances of which the 3 main symptoms are: a slowness of movement (bradykinesia), resting tremor and muscle stiffness (rigidity).

The estimated prevalence of PD in Europe varies considerably and ranges from 66-

12,500/100000, probably due to issues with diagnosis and the clinical heterogeneity of the disease (von Campenhausen et al. 2005). A higher incidence has been associated with increased age and in males compared to females (Van Den Eeden et al. 2003; Pringsheim et al. 2014; Hirsch et al. 2016). Environmental factors such as pesticide exposure and head injury have been observed to correlate with increased risk (Noyce et al. 2012). Motor symptoms arise when approximately 30% of mDA neurons are lost (Fearnley et al. 1991; Ma et al. 1997) with a corresponding reduction in striatal dopamine levels of approximately 80% (Scherman et al. 1989).

These parkinsonian symptoms arise because of loss of neuromelanin-containing A9 midbrain dopaminergic neurons of the SNpc, resulting in a loss of dopamine to the neostriatum. It is characterised histologically by the presence of abnormal, insoluble protein aggregates, termed Lewy bodies (LBs) and Lewy neurites (LNs). These aggregates contain proteins such as

8

synphilin, ubiquitin, neurofilament protein and occasionally tau but are predominantly composed of α-synuclein protein (Figure 1.5). These α-synuclein Lewy bodies are also found in patients with Dementia with Lewy Bodies (DLB) and α-synuclein is also found in glial cytoplasmic inclusions that are characteristic of Multiple Systems Atrophy (MSA). Together, these diseases are grouped together into a family of neurodegenerative disorders termed synucleinopathies.

Figure 1.5 - Immunohistochemical labelling of midbrain dopaminergic neurons for α-synuclein and ubiquitin, identifying Lewy bodies which have a distinct halo shape (Figure taken from (Dauer et al. 2003)).

The motor symptoms experienced by patients are often accompanied by secondary non- motor symptoms such as autonomic, gastrointestinal and neuropsychiatric symptoms (Zhang et al. 2016). LB/LN pathology is thought to be able to spread through the brain caudo-rostrally through the vagal connections of the gut (described in section 1.4.4). These non-motor symptoms may be linked to the spreading of LB/LN pathology through different anatomical regions. Parkinson’s disease associated dementia, for example, has been previously correlated to LB distribution and density in the cortex and is thought to affect 30% of PD patients, representing approximately a 4-6-fold increase from the general population

(Cummings 1988; Mattila et al. 2000).

9

1.2.2 Current treatment options for PD

As mentioned in section 1.1.3, the first targeted treatment for PD motor symptoms was dopamine replacement therapy. Alternative pharmacological treatments have been developed such as dopamine agonists that function like dopamine to activate dopamine receptors. Other treatments include monoamine oxidase B or catechol-O-methyltransferase inhibitors that can prevent the degradation of dopamine, thus increasing striatal dopamine levels. These drugs can relieve motor symptoms associated with PD but cannot slow, stop or reverse PD pathological changes like SNpc neuron degeneration (reviewed in (Connolly et al.

2014)).

More recently, brain derived neurotrophic factors such as nerve growth factor (NGF), brain- derived neurotrophic factor (BDNF) and glial-derived neurotrophic factor (GDNF) have shown promising results in pre-clinical and clinical studies as neuroprotective agents for neurodegenerative diseases. These treatments can slow down disease progression by enhancing survival and regeneration of mDA neurons but require invasive procedures to direct proteins to affected regions (reviewed in (Allen et al. 2013)). In an α-synuclein overexpression rat model, it was shown that increased α-synuclein burden resulted in the reduction of the GDNF receptor component RET tyrosine kinase, revealing a possible mechanism of pathology (Decressac, Kadkhodaei, et al. 2012). In a recent human trial, 41 PD patients were administered intraputamenal GDNF via a transcutaneous port and monitored over 40 weeks. The study reported a large, clinically relevant improvement in motor function in 43% of PD patients given GDNF compared to PD patients given a placebo, whilst none of the patients given placebo showed any significant improvement. However, overall, there was no significant difference in motor function between the GDNF and placebo groups (Whone et al. 2019).

10

As motor symptoms arise after 30% of mDA neurons have degenerated, the prodromal phase of PD offers a therapeutic window to delay or prevent the disease. c-Abl tyrosine kinase inhibitors, which can prevent phosphorylation of α-synuclein, has had some preclinical success in protecting dopaminergic neurons from degeneration and improving motor behaviour (Hebron et al. 2013; Karuppagounder et al. 2014). Direct stimulation of basal ganglia activity via deep brain stimulating electrodes in the subthalamic nucleus has also shown some efficacy in improving motor symptoms (Volkmann et al. 2001).

These therapeutic strategies are not as effective in the later stages of the disease when a substantial number of neurons have degenerated. Fetal mesencephalic grafts into PD patient brains have previously shown some evidence of functional motor improvements and, encouraged by results, stem cell-derived transplantation strategies are currently being developed for early phase clinical trials (Barker et al. 2017).

1.3 Midbrain Dopaminergic (mDA) neurons

1.3.1 Development of the midbrain dopaminergic neurons

1.3.1.1 Neuralisation

The mDA neurons of the SNpc are derived from the ventral midline, termed the floorplate, of the neural tube. Formation of the early nervous system begins during the gastrulation event early in embryogenesis and the appearance of a more complex three-layered structure consisting of endoderm, mesoderm and ectoderm, as well as a structure termed the notochord. The notochord secretes growth factors that stimulate the ectoderm layer, which begins to differentiate into neuroectoderm cells. It is at this point that the process of neurulation begins with the formation of a neural plate that separates from the surface ectoderm at the neural folds, and folds into a tube-like structure. This tube will give rise to

11

nearly all the neurons and glia of the central nervous system (CNS) of vertebrates. The anterior portion of the neural tube then begins to swell and develop the three primary vesicles which give rise to the forebrain (prosencephalon), midbrain (mesencephalon) and hindbrain (rhombencephelon) (Figure 1.6).

Figure 1.6 - Schematic representing the three embryonic vesicles of the neural tube that develop into the five secondary vesicles and their adult derivatives. The prosencephalon gives rise to the telencephalon and diencephalon, and subsequently forebrain structures such as the cerebral hemispheres and the thalamus, respectively. The mesencephalon gives rise to midbrain structures including the substantia nigra. The Rhombencephalon gives rise to the metencephalon and myelencephalon, and subsequently hindbrain structures such as the pons/cerebellum and medulla, respectively (Figure adapted from (Jinkins 2000))

Early investigations into the processes that drive stem cell specification used isolated embryonic tissue at different stages of early embryogenesis to attempt to discover the prospective fate of the isolated tissue. Hans Spemann and Hilde Mangold in 1924 (Spemann et al. 2001) transplanted the dorsal lip of the Triton species of newt blastopore embryo at gastrulation stage into the ventral side of a different pigmented embryo and showed that instead of developing into mesodermal notochord, the graft induced the formation of neural tissue from ectoderm and dorsalisation of the ventral mesoderm. This resulted in the formation of a twin embryo at the graft site, containing cells of both pigmentations. This experiment provided the first evidence for a process termed induction, whereupon cells adopt their developmental fate based on interactions with other cells relative to their position

12

within the developing embryo. In 1991, Smith and Harland systematically screened for mRNAs that encode proteins which have a role in neural induction in ventralised Xenopus embryos

(Smith et al. 1991). They identified a number of potential candidates, the most interesting of which was a novel mRNA that they named Noggin (Smith et al. 1992). This mRNA was later identified in mammals and was shown to be expressed in the dorsal lip and the notochord and could recapitulate neural induction and formation of the neural plate without the presence of the mesoderm (Smith et al. 1993; Valenzuela et al. 1995).

Around the same time, other potential neural induction candidates were being studied. For example, Activin, a member of the TGF-B family, was shown to inhibit neural induction

(Hemmati-Brivanlou et al. 1994). Interfering with Activin signalling disrupted normal mesoderm development but also induced neuralisation. Subsequently, Follistatin was identified from the reproductive system as an antagonist of Activin and could induce neuralisation (Hemmati-Brivanlou et al. 1994).

Finally, an important neural induction signalling network was identified using Drosophila models in 1995 by De Robertis et al. Chordin was shown to have neuralising activity that was antagonised by bone morphogenic factor BMP-4 (a member of the TGF-B superfamily) (Sasai et al. 1995; Piccolo et al. 1996). Importantly, inhibition of endogenous BMP-4 was able to drive neuralisation. Thus, the critical Neural-inducing factors identified in the Mangold-

Spemann organizer were bone morphogenic protein (BMP) inhibitors, including Chordin,

Follastatin and Noggin (Figure 1.7).

In 2000, a role for fibroblast growth factor (FGF) signalling in neuralisation was identified in chick embryo development studies. Fgf8 was able to enhance expression of an early activated gene, the early response to neural induction (ERNI) gene, that is expressed in presumptive

13

neurons (Streit et al. 2000). This response was not associated with BMP antagonism, and an

FGF inhibitor blocked early neural induction prior to gastrulation.

Figure 1.7 – Developmental pathways directing towards a neural fate. Activation of the BMP pathway induces ectoderm transition to epidermis, termed the default model. Inhibition of BMP, Activin and TGF-β signalling by Noggin, Chordin and Follistatin inactivates downstream signalling, resulting in transition to neuroectoderm, directing cells to a neural fate. (Adapted from (Muñoz-Sanjuán et al. 2002))

1.3.1.2 Specification of differentiating neural tube cells towards a midbrain fate

Specification along the dorsal-ventral axis is mediated by the notochord ventrally whilst dorsal specification is mediated by the epidermis. In an attempt to identify lethal mutations in

Drosophila melanogaster, Nusslein-Volhard identified 15 genetic loci that altered the segmentation pattern of developing larvae and generated distinct anatomical phenotypes

(Nüsslein-Volhard et al. 1980). One of these genetic loci, the hedgehog locus, was excreted by the notochord and regulated CNS polarity in vertebrate models (Roelink et al. 1994;

Echelard et al. 1993). Sonic Hedgehog (Shh), a member of the hedgehog protein family and a ligand of the hedgehog signalling pathway, is cleaved to release the amino-terminal active form which induces formation of the floorplate of the neural tube (Figure 1.8A). The floor plate, once formed, can also secrete Shh in a gradient and acts as a secondary organiser

(Roelink et al. 1995; Briscoe et al. 1999) that directs the specification of different cell types

14

within the neural tube (Hynes et al. 1995). In mouse embryos harbouring a GFP tagged Shh from the endogeneous locus shows that Shh generates a dynamic gradient along the dorso- ventral axis (Chamberlain et al. 2008). The graded concentration of Shh protein induces different groupings of neurons to arise, expressing divergent combinations of transcription factors and activated genes to specify cell fate (Ribes et al. 2010; Dessaud et al. 2007) (Figure

1.8A). This is thought to be mediated by combinations of genes which either require Shh signalling and expressed ventrally such as NKX.2 (Lei et al. 2006) or are repressed by Shh signalling such as PAX6 which is expressed dorsally (Ericson et al. 1997). SHH knockout mice have shown pronounced defects in floor plate and notochord establishment (Chiang et al.

1996).

An important transcription factor involved in specification is the hepatocyte nuclear factor

HNF-3β (FOXA2) that is expressed in the notochord and the floorplate (Figure 1.8B). FOXA2 has been associated with the regulation, by a feedback loop, of SHH transcription to specify ventral midbrain progenitor fate (Epstein et al. 1999; Mavromatakis et al. 2011; Sasaki et al.

1997). Furthermore, FOXA2 has recently been shown to directly induce Shh expression in the floor plate by binding to an enhancer region in the regulatory region of Shh (Cho et al. 2014).

Deletion of the FOXA2 gene resulted in the loss of notochord development, disrupted dorsal- ventral patterning and caused embryonic lethality in a mouse model (Ang et al. 1994).

FOXA1/FOXA2 double knock-out mice identified a dramatic reduction in Neurog2 levels (Ferri et al. 2007), which is required for mDA progenitor maturation during specification (Kele et al.

2006). A study of FOXA2 binding sites identified sites associated with 5409 genes, with positive regulation identified for key determinants of mDA neurons (Metzakopian et al. 2012).

Later in development FOXA2 has also been shown to regulate Nurr1 expression and act

15

synergistically with Nurr1 in mDA neurogenesis from progenitors (Yi et al. 2014; Lee et al.

2010). Additionally, FOXA2 overexpression in mouse embryonic stem cells (mESCs) promoted the generation of dopaminergic neurons (Kittappa et al. 2007). Furthermore, a one allele deletion of FOXA2 gives rise to a mouse model with parkinsonism and disrupted mDA development (Ferri et al. 2007; Kittappa et al. 2007).

Importantly, a role for Wnt1, which is expressed in the presumptive midbrain, has been proposed for regulation of anterior-posterior specification along the midbrain-hindbrain axis

(Figure 1.8). Mouse models with global disruption of the WNT1 gene resulted in mice with severe abnormalities in midbrain, hindbrain and cerebellar development, elucidating a role for Wnt signalling in early specification (McMahon et al. 1990; Thomas et al. 1990). Disruption of Wnt/β-catenin signalling in the ventral midbrain in a conditional mouse mutant showed an expansion of progenitors and reduced expression of both SHH and FOXA2 and a reduction in generation of dopaminergic neurons and thus a role in neurogenesis (Joksimovic et al. 2009;

Tang et al. 2010). Interestingly, midbrain activation of Wnt/β-catenin signalling in mouse models showed a pronounced disruption of anterior-posterior and dorso-ventral patterning and a further reduction of dopaminergic neurons (Chilov et al. 2010; Nouri et al. 2015). These studies suggest that endogeneous Wnt signalling is optimised to establish an mDA progenitor character and allow for proper mDA neurogenesis.

Wnt1 has been shown to antagonise Shh signalling and forms an autoregulatory feedback loop with the transcription factor LIM Homeobox Transcription Factor 1 Alpha (LMX1A) to regulate Otx2 and ventral midbrain cell fate during development (Chung, Leung, et al. 2009).

LMX1A is also expressed in the ventral midbrain, with transfection of vectors to express

16

LMX1A sufficient to induce ectopic expression of mDA neurons in chick embryos and (Figure

1.8) (Andersson et al. 2006).

In conclusion, there is evidence that stage-specific expression of these extrinsic and intrinsic early patterning factors can induce genes involved in midbrain development. Shh and Wnt1 are the main extrinsic signals involved in mDA development, with an activation of the Wnt1 pathway resulting in loss of Shh expression (Tang et al. 2010). This suggests that a balance between Wnt and Shh signalling mediates progression of ventral midbrain progenitors to dopaminergic neurons. These extrinsic factors initiate the expression of intrinsic transcription factors such as Lmx1a and Foxa2 which can act cooperatively to induce activation of mDA genes and mDA neuron specification (Figure 1.8B)(Lin et al. 2009; Nakatani et al. 2010). Other key factors expressed in mDA neurons at different developmental stages are shown in Figure

1.8B.

17

Figure 1.8 – Floorplate and mDA development. A) Development of the neural tube and the floor plate. The notochord secretes Shh and BMP antagonists such as Chordin, Noggin and Folastatin whilst Shh, Foxa2 and Lmx1a are co-expressed in the ventral floorplate of embryos. Neural progenitor subtypes are orderly patterned along the dorso-ventral axis by the gradients of Shh from the notochord (N) and floor plate (FP) and Wnts/Bmps produced by the roof plate (RP) (from reference (Ulloa et al. 2010)). B) Molecular factors involved at different stages of development of mDA neurons (adapted from reference (Hegarty et al. 2013)). Factors are colour coded according to their role. In red are factors which play roles in establishing the midbrain/hindbrain junctions and floor plate. In green are diffusible signalling factors which induce mDA neurogenesis in radial glial-like cells. In blue are the induced factors expressed due to inductive signalling in green. Shh-Foxa2 and Wnt1-Lmx1a autoregulatory loops are involved in mDA neurogenesis in the floor plate. In pink are factors which promote differentiation of neural progenitor cells to post-mitotic neurons and induce expression of proteins that give the dopaminergic neurotransmitter phenotype of mDA neurons. Molecules in black are not in the correct timepoint. Anterior-posterior patterning is regulated by Wnt1 and Fgf8, restricting expression of Otx2 to the ventral midbrain. Lmx1a transcription factor is involved in the regulation of Wnt1 signaling. Dorso-ventral patterning is regulated by the graded expression of Shh, which is regulated by FOXA2 transcription factor.

18

1.4 Alpha-Synuclein

1.4.1 Association of α-synuclein with familial PD and Lewy bodies

Following its clinical characterisation, PD was thought of as an idiopathic disease with mainly non-genetic etiological factors. The majority of twin studies showed no evidence for heritability (Tanner et al. 1999; Duvoisin et al. 1981; Ward et al. 1983). Reports of familial aggregation had been documented (Elbaz et al. 1999; Spellman 1962) and clinical genetic studies reported some evidence of increased risk of transmission of PD (Lazzarini et al. 1994;

Spellman 1962; Maraganore et al. 1991). These studies suggested that, if familial PD has a genetic basis, segregation and pedigree analyses suggest an autosomal dominant inheritance of a gene or genes with reduced penetrance.

In 1996 scientists first reported the identification of genetic markers on chromosome 4q21- q23 that was linked to autosomal dominant PD in a large kindred of Italian descent (The

Contursi Kindred) (Polymeropoulos et al. 1996; Golbe et al. 1990; Golbe et al. 1996). This loci contained the SNCA gene that encodes α-synuclein and sequence analysis of the fourth exon revealed a single base pair change at position 209 from G to A (G209A), which resulted in an alanine to threonine substitution at position 53 (Ala53Thr). This mutation showed complete segregation with the PD phenotype in the Italian kindred as well as 3 independent PD families

(Polymeropoulos et al. 1997). Based on these observations and the identification of α- synuclein in AD amyloid plaques (Uéda et al. 1993), the authors hypothesised that this alanine to threonine substitution may disrupt the α helix structure, inducing formation of β-pleated sheets and aggregation into amyloid-like structures. The Italian kindred PD clinical phenotype was thought to be typical for sporadic PD, except for a relatively earlier age of onset of illness

(46 ± 13 years). It was estimated that the A53T mutation in this family resulted in 85%

19

penetrance and suggested that a single gene defect was sufficient to induce the PD phenotype.

In the same year, a group reported strong staining of LBs from idiopathic Parkinson’s and DLB patients using two antibodies, which correspond to α-synuclein residues (Jakes et al. 1994;

Spillantini et al. 1997). The group suggested that α-synuclein might be the main component of PD LBs. Importantly, the PD cases studied were non-familial, therefore concluding that aggregation of α-synuclein is common in the aetiology and pathogenesis of all cases of PD.

Since LBs are diagnostic of synucleinopathies, these discoveries set the foundations for the research into mechanisms of cellular dysfunction and pathological pathways.

1.4.2 Clinical heterogeneity of α-synuclein mutations

The discovery of a missense mutation in SNCA that can cause rare autosomal dominant inheritance of familial PD and the finding that synuclein aggregates are the main component of Lewy bodies has fuelled a great deal of research in this area. More missense mutations as well as duplication, triplication and multiplication of the SNCA gene have been reported to be linked to the aetiology of the disease (Table 1.3; Figure 1.9). Additionally, a multitude of posttranslational modifications, including phosphorylation, ubiquitination, nitration and truncation, as well as alternative splicing events, have been discovered and the myriad of ways in which these modifications affect α-synuclein function has been studied.

20

Figure 1.9 – Mutations in SNCA. Schematic showing protein structure of α-synuclein and PD-related mutations A30P, E46K, H50Q, G51D, A53T/E. α-synuclein, a 140-amino-acid protein, is composed of three domains including an amphipathic N- terminal domain, a non-amyloidogenic component (NAC) with a hydrophobic core and an acidic C-terminal domain. The amphipathic and NAC domains contain KTKEGV repeats. Figure taken from (Devoto et al. 2017).

Age of onset Mutation Clinical symptoms/Biological effect Early onset, rapid progression, dementia, LB and tau pathology, 19-60 G51D glial cytoplasmic α-synuclein inclusions Early onset, rapid progression, cognitive decline and dementia, 20-50 Triplication hallucinations and autonomic dysfunction, LB and tau pathology Early onset, rapid progression, dementia, LB/tau pathology, glial 36-62 A53E cytoplasmic α-synuclein inclusions Early onset, rapid progression, dementia, myoclonus, LB and tau 20-85 A53T pathology 31-77 Duplication Rapid progression, dementia, LB pathology, hallucinations 54-76 A30P Hallucinations, mild dementia, LB pathology 59-69 E46K REM sleep disorder, cognitive decline, hallucinations, LB pathology Cognitive decline, various psychiatric symptoms, LB and tau 60-71 H50Q pathology Table 1.2 - SNCA multiplications and mutations associated with PD and their clinical presentations.

In 1998, another missense mutation in the SNCA gene was linked to autosomal dominant PD

(AD-PD). Researchers identified a G88C substitution in German familial PD patients, which caused an alanine to proline change at amino acid 30 (A30P). Clinically, these patients developed signs of “typical” PD in their 50s. The authors noted that proline is involved in the acquisition of protein secondary structure, and hypothesised that this mutation would be involved in induction of PD by influencing α-synuclein aggregation propensity (Krüger et al.

1998).

21

In 2003, studies revealed that a family with an apparently autosomal dominant PD showed triplication of the SNCA gene leading to increased dosage (Spellman 1962; Muenter et al.

1998). Clinically, the pathological course of this pedigree was rapidly progressive and characterised by young onset of parkinsonism (average 34 years of age), followed by dementia, with widespread LB pathology (Singleton et al. 2003). A year later in 2004, the gene dosage mechanism was further elaborated on by the identification of SNCA duplications which resulted in familial parkinsonism with a clinical phenotype reminiscent of idiopathic PD in age of onset and disease progression and lack of atypical features on autopsy (Chartier-

Harlin et al. 2004; Ibanez et al. 2004). Cognitive decline and dementia were not prominent features of the progression of the disease compared to SNCA triplication carriers (Singleton et al. 2003).

Another missense mutation associated with AD-PD, with a base pair change at 188 from G to

A (G188A) was identified, causing a glutamate shift to lysine at position 46 (E46K) in individuals from a large Spanish family (Zarranz et al. 2004). Clinically, the patients from this kindred presented with PD with an age of onset typical for sporadic PD, dementia and later onset visual hallucinations, with cognitive deficits found in the majority of patients. Strikingly,

Lewy body pathology was found in cortical and subcortical areas that are associated with cognitive symptoms (Gomez-Tortosa et al. 1999; Harding et al. 2001).

In 2013, another 2 SNCA missense mutations were discovered. The first, a base pair change at 150 from T to G encoding a histidine to glutamine substitution at amino acid 50 (H50Q) in exon 4 of SNCA (Appel-Cresswell et al. 2013). This was only found in one patient, of

English/Welsh origin, who presented at age 60 and showed combination motor and cognitive decline. The other mutation, found in two separate families, showed a G to A mutation at

22

base 152 (G152A) causing a glycine to aspartic acid shift at residue 51 (G51D) (Lesage et al.

2013; Kiely et al. 2013). Patients with this mutation presented with a very young onset of motor symptoms (19-39 years of age), tremors and dementia. Most interestingly, post- mortem showed unique striatal and neocortical pathology, as well as LB pathology and glial cytoplasmic inclusions of MSA. This represented the first genetic link to MSA, another synucleinopathy.

More mutations have been found in familial PD patients including a C to A mutation leading to a alanine to glutamate shift (A53E) which shows a severity of pathology most comparable to that of G51D mutations (Pasanen et al. 2014). A summary of the SNCA mutations is provided in Table 1.3.

1.4.3 Modifications: truncation, phosphorylation and nitration

Partially truncated and high molecular weight aggregates of α-synuclein have also been recovered from LBs, suggesting other mechanisms by which α-synuclein may be modified and rendered either insoluble or more prone to aggregate (Baba et al. 1998). Further research has elucidated an association between C-terminal truncated wild-type and mutated α-synuclein and promotion of aggregation and fibril formation (Murray et al. 2003; Ulusoy et al. 2010).

Additionally, there is evidence for the existence of numerous and diverse post-translational modifications of α-synuclein such a nitration (Yu et al. 2010; Liu et al. 2011), phosphorylation

(Fujiwara et al. 2002; Anderson et al. 2006) and acetylation (Maltsev et al. 2012; Bu et al.

2017).

23

1.4.4 Spreading of α-synuclein

The Lewy body pathology staging classification for PD was outlined by Heiko Braak in 2004

(Braak et al. 2004). Braak and colleagues noted 6 stages of LB pathology outlined in Table 1.4 and Figure 1.10.

Stage Affected tissues Description of pathology The least affected brains, Stage I, showed LB/LN (mainly LNs at this stage) pathology restricted entirely to the dorsal IX/X motor nucleus and/or intermediate reticular zone. Olfactory Medulla Pre-clinical bulb and anterior olfactory nucleus lesions also begin to Oblongata & Stage I & II develop. The key feature of Stage II is the appearance of LBs Olfactory bulb and LNs in neurons of the caudal raphe nuclei and the reticular formation, particularly the gigantocellular reticular nucleus, with LNs preceeding the appearance of LBs. Stage III is characterised by the appearance of LNs and LBs in a subset of midbrain neuromelanin-containing neurons of the substantia nigra, initially restricted to posterolateral and Stage III & IV Midbrain posteromedial subnuclei. Stage IV shows marked degeneration of the substantia nigra neurons and involvement of the magnocellular nuclei of the basal forebrain. By this stage severe damage to the anterior olfactory bulb has occurred. Neurodegeneration in structures noted in Stages I-IV continues. Stage V shows LB/LN pathology in the allocortex with severe LNs pathology in first, second and third Ammon’s Horn sectors. Cortical pathology appears to originate from the Stage V & VI Neocortex mesocortex into adjacent neocortex, insular cortex, cingulate cortex and prefrontal areas. Finally, stage VI involves the entire neocortex with milder pathology in primary motor area and primary motor field as well as first order sensory association and primary sensory areas. Table 1.3 – The staging of α-synuclein pathology as described by Braak et al. Propagation of α-synuclein pathology beginning in the olfactory bulb and medulla oblongata in early stages I & II. Later, α-synuclein pathology was observed to spread into the midbrain and substantia nigra in stages III & IV. Finally, spreading was observed in cortical regions.

24

Figure 1.10 - Schematic representing the spreading and propagation of α-synuclein LN/LB pathology in idiopathic PD. A) Pathology is hypothesised to start in the olfactory bulb (yellow arrows) or from the gastrointestinal tract via the vagus nerve (red arrows). As the disease progresses spreading is seen to the midbrain and cortical regions. Figure taken from (Hansen et al. 2012). B) Caudorostral spreading of α-synuclein LN/LB pathology related to early pre-symptomatic and symptomatic stages with dark colours representing early stages of PD and the light colours representing later stages of disease progression. Figure taken from (Pacheco et al. 2012).

Pathology is thought to begin in medullary structures in the brain stem and olfactory system in Stage I & II. LB pathology is then thought to cross into the SNpc, characterised by mDA degeneration, and the amygdala and thalamus, as well as in the anterior olfactory nucleus in

Stage III & IV. Finally, LB pathology spreads into the neocortex affecting motor and sensory

25

areas in Stage V & VI. The staging scheme proposed by Braak, with its depiction of a temporal sequence of pathological events, has had a huge influence on the development of early diagnosis and identification of potential biomarkers for PD (Figure 1.10).

The spreading of α-synuclein outlined by Braak et al. hinted at a mechanism of cell-to-cell propagation similarly to prions, infectious proteins whose pathological folding properties is transmissible to other prion proteins (Braak et al. 2003). Significant to this hypothesis, both aggregated and non-aggregated forms of α-synuclein can be isolated from cerebro-spinal fluid (CSF) and blood of PD patients (Borghi et al. 2000; El-Agnaf et al. 2003). This theory was given further credence by the discovery of a host-to-graft propagation of α-synuclein in post- mortem sections of PD human patients who had undergone fetal mesencephalic dopaminergic neuron grafting (Kordower et al. 2008; Li et al. 2008; Kurowska et al. 2011).

Extracellular α-synuclein is released via exocytosis (Lee et al. 2005) and taken up by cells via endocytosis or exosomal transport, with these mechanisms thought to be important to progression of disease pathology (Desplats et al. 2009; Hansen et al. 2011; Angot et al.

2012)(Desplats 2009; Hansen 2011; Angot 2012). Moreover, it was found that spreading of

LN/LB pathology could be initiated by α-synuclein transfer into the vagal nerve and halted by resectioning of the vagal nerve connecting the gut to the CNS (Ulusoy et al. 2013; Pan-

Montojo et al. 2012).

Another potential mechanism for α-synuclein secretion is via exosomes, membrane-bound vesicles which are released by many types of cells and found in body fluids. It has been shown that certain prion proteins can be released within exosomes which can be transferred to recipient cells, initiating propagation of pathology (Fevrier et al. 2004; Vella et al. 2008).

Extracellular α-synuclein has also been shown to be released via exosomes (Emmanouilidou

26

et al. 2010; Angot et al. 2012) with an increase in release linked with lysosomal dysfunction

(Alvarez-Erviti et al. 2011). Exosome-associated α-synuclein has been shown to be more readily taken up by recipient cells and capable of inducing increased toxicity (Danzer et al.

2012). Indeed, exosomes from PD patients injected into immune-deficient mice induced mild features of parkinsonism compared to exosomes from healthy control (Han et al. 2016).

1.4.5 Structure and aggregation properties of α-synuclein

The discovery of a link between mutations of SNCA and familial PD, as well as the inclusion of

α-synuclein protein aggregates in LBs, the histological hallmark of PD, spurred research into the physiological function of α-synuclein. α-synuclein is a 140-amino-acid protein that contains a hydrophobic NAC domain, a positively charged N-terminal containing a repeating

11-amino-acid sequence with a conserved hexameric KTKEGV motif as well as an acidic C- terminal tail (Figure 10) (Uéda et al. 1993). Although thought of as a “natively unfolded protein” (Weinreb et al. 1996), the N-terminal is thought to be able to induce an α-helical structure with this conformational adoption linked to a phospholipid binding function (Jenco et al. 1998; Ahn et al. 2002).

α-synuclein belongs to a family of synucleins which includes α, β (Nakajo et al. 1993) and γ

(Akopian et al. 1995) synucleins. Whilst α and β synucleins are concentrated at nerve terminals of the central nervous system, γ-synuclein is present in the peripheral nervous system (Akopian et al. 1995). β and γ synucleins are 62% and 55% identical in amino acid sequence to α-synuclein respectively. Importantly, it has been shown that these proteins have differential, cell-specific expression profiles and only α-synuclein is present at a high intensity in areas of the CNS affected by PD such as the SN and basal ganglia (Li et al. 2002). β-synuclein has been shown to function in a neuroprotective role against α-synuclein induced

27

neurodegeneration, inhibiting its aggregation and fibrilisation (Hashimoto et al. 2001; Park et al. 2003; Fan et al. 2006). Although β and γ synucleins have not been associated with PD, recent studies have suggested that β-synuclein may have aggregative properties

(Taschenberger et al. 2013) and a link between a mutation in β-synuclein and neurodegeneration in a mouse model has been shown (Fujita et al. 2010). Α α and β-synuclein double Knock-out (KO) mouse model has been generated and shows a mild phenotype, with a reduction in striatal dopamine levels, whilst a single KO of either α or β-synuclein did not show this phenotype (Chandra et al. 2004).

α-synuclein can be conformationally organised into β-pleated sheets that are resistant to proteolysis thus increasing its probability of polymerisation (Serpell et al. 2000; Conway et al.

2000). Notably, a study identified α-synuclein multimers in brain lysates of PD patients who harbour the A53T mutation, though not in sporadic PD or control samples, hinting at an association between mutant A53T α-synuclein and a potential pathological mechanism

(Langston et al. 1998). Later studies have reported that the predominant soluble form of α- synuclein exists in the native tetramer form which resists aggregation (Bartels et al. 2011;

Wang et al. 2011). Thus, it may be that reduced stability of oligomers increases the availability of more readily aggregated monomer and fibrils associated with the development of LB pathology (Bucciantini et al. 2002; Volpicelli-Daley et al. 2011). Furthermore, it has been noted that normal brain samples contain abundant α-synuclein tetramers, with PD-associated mutations linked to a shift from tetramer to monomer and subsequent neurotoxicity

(Dettmer et al. 2015). This oligomeric species has been observed to have specific membrane binding affinity and a functional role distinct to that of monomeric α-synuclein (van Rooijen

28

et al. 2008). Thus, formation of toxic oligomers is associated with protein misfolding and oligomeric secondary structures (Figure 1.11).

Figure 1.11 – Amyloid fibril formation and aggregation dynamics. Polymerisation of α-synuclein can be divided into “on- pathway” or “off-pathway” in relation to toxic fibril formation. Oligomer toxicity is thought to be associated with higher levels of β-pleated sheet structure and fibril elongation. Blue circles- Little or no β-structure; Red circles- High β-structure. Figure taken from (Roberts et al. 2015).

1.4.6 Protein degradation pathways

Protein degradation is an important process regulating intracellular homeostasis, which is undertaken by both the autophagy-lysosomal pathway (ALP) and the ubiquitin proteasome system (UPS). Autophagy itself consists of 3 main mechanisms of autophagic degradation, namely chaperone-mediated autophagy (CMA), macroautophagy and microautophagy.

Suppression of autophagy in transgenic mouse models leads to a neurodegenerative phenotype in the absence of disease-associated proteins (Hara et al. 2006; Komatsu et al.

2006), with ultrastructural evidence of autophagic degeneration reported in PD patients

(Anglade et al. 1997). Thus, it is not surprising that degradation of monomeric α-synuclein has been reported to be processed by both the proteasome and autophagy in an in vitro PC12 model (Webb et al. 2003). More specifically, it was reported that wild-type α-synuclein degradation via autophagy is specifically mediated by the CMA (Vogiatzi et al. 2008) by direct

29

binding to lysosome-associated membrane protein 2 (LAMP2A) autophagy receptor, with poor degradation of α-synuclein pathogenic mutations (Cuervo et al. 2004). Furthermore, when CMA is blocked, the macroautophagy pathway is activated to compensate and clear aggregates.

The UPS is also known to degrade misfolded proteins by a process in which ubiquitinated substrates are cleaved into smaller peptides before passing through the proteasome.

Wildtype and mutant α-synuclein was reported to both be degraded by the UPS in a neuroblastoma model, with slower degradation of mutant pathogenic α-synuclein (Bennett et al. 1999). Stable expression of both mutant and wild-type α-synuclein have been recently been shown to impair proteasome activity directly in another neuroblastoma model

(Lindersson et al. 2004; Zondler et al. 2017). Moreover, an autosomal dominant point mutation in the ubiquitin carboxy-terminal hydrolase L1 (UCHL1) gene, associated with the generation and stabilisation of monoubiquitin in neurons (Osaka et al. 2003), has been associated with familial PD (Leroy et al. 1998). Perturbing UCHL1 function induces a build-up of α-synuclein (Liu et al. 2002; Cartier et al. 2012). One study even reported a direct interaction between UCHL1 and LAMP2A, the lysosomal receptor for CMA (Kabuta et al. 2008) and a more recent study reported the existence of feedback between the two systems (Yang et al. 2013).

1.4.7 Function of α-synuclein

1.4.7.1 Neurotransmitter release

How α-synuclein may affect the pathological process (Figure 1.12) has been the subject of intense study. The presynaptic localisation of α-synucleins, its association with mDA neuronal function and interaction with phospholipids suggested a role in neurotransmitter release and

30

synaptic vesicle binding. The first SNCA KO mouse model reported perturbations in striatal dopamine (Abeliovich et al. 2000). Overexpression of synuclein led to an inhibition of neurotransmitter release in murine PC12 cells (Larsen et al. 2006) and in primary neuronal cultures (Nemani et al. 2010), by perturbing vesicle docking or synaptic vesicle reclustering

(Diao et al. 2013). Synaptic vesicle membrane docking and fusion requires membrane fusion machinery, a main component of which is soluble N-ethylmaleimide-sensitive factor attachment protein receptor (SNARE). In an elegant transgenic mouse study, α-synuclein was found to display chaperone activity and promoted SNARE complex assembly by direct binding to SNARE proteins (Burré et al. 2010). It can also function downstream of another presynaptic chaperone protein, CSPα, and rescue the neurodegenerative phenotype caused by CSPα-KO

(Chandra et al. 2005).

Figure 1.12 – Pathways associated with α-synuclein pathology. Functional disruption of organelles: nucleus, mitochondria, ER/Golgi, lysosome and synaptic dysfunction (purple box). Mitochondria-associated ER membrane and subsequent mitochondrial fragmentation dynamics (blue box). Axonal transport disruption (green box). Figure taken from (Wong et al. 2017)

31

1.4.7.2 Mitochondria and Reactive oxygen species

α-synuclein has been reported to directly bind to mitochondrial lipid membranes influencing mitochondrial fission and fusion dynamics associated with the regulation of mitochondrial health (Nakamura et al. 2008; Nakamura et al. 2011). Nuclear α-synuclein has been reported to bind to the promoter of the master mitochondrial transcriptional activator, PGC1α, leading to repression of transcription and regulating mitochondrial function (Siddiqui et al. 2012). iPSC-derived neurons expressing mutant A53T α-synuclein was also shown to induce a dysfunction in PGC1α transcriptional networks (Ryan et al. 2013).

Major sources of cellular reactive oxygen species (ROS) in mDA neurons are thought to be both mitochondria and the metabolism of dopamine. ROS is thought to be linked to an increase in mitochondrial DNA (mtDNA) damage in high-energy demand tissues (Bender et al.

2006). Increased ROS has been reported to influence α-synuclein conformation and aggregation (Esteves et al. 2009), with dopamine possibly playing a role in these conformational changes (Outeiro et al. 2009). α-synuclein protein has been shown to be localised to the mitochondrial membrane, in dopaminergic neurons, with accumulation of α- synuclein protein to the mitochondria impairing normal mitochondrial function and increasing mitophagy (Li et al. 2007; Chinta et al. 2010) as well as mitochondrial dysfunction association with α-synuclein mutant transgenic mouse models (Martin et al. 2006).

1.4.7.3 ER-Golgi trafficking

Anterograde transport of synthesised proteins from the endoplasmic reticulum, a major site of protein folding, to the Golgi has been shown to be disrupted by α-synuclein in a dose- dependent manner (Gitler et al. 2008), including trafficking of the dopamine transporter

DAT(Oaks et al. 2013). α-synuclein has been shown to aggregate within the ER inducing a

32

stress response (Colla et al. 2012). α-synuclein can also be localised to mitochondria- associated ER membranes (Guardia-Laguarta et al. 2014).

1.4.7.4 Axonal transport and neurite dynamics

Axonal transport dysfunction and changes in expression of motor proteins, kinesin and dynein, have been reported in response to mutant A30P and A53T α-synuclein overexpression in rat PD models (Chung, Koprich, et al. 2009; Chu et al. 2012). Overexpression of synuclein has been shown to impair microtubule-dependent trafficking (Lee et al. 2006) and it has been shown that α-synuclein can interact with tubulin, influencing its polymerisation and microtubule dynamics in neurons (Zhou et al. 2010). Different pathological mutations can impair microtubule assembly, affecting tubulin-kinesin interplay, and neurite network morphology (Prots et al. 2013) which has a downstream effect on neurite outgrowth

(Takenouchi et al. 2001; Liu et al. 2013). In a recent study, neurite analysis of rat primary dopaminergic neurons showed that neurons transfected with wild-type, A53T and A30P α- synuclein had the effect of impairing neurite length whilst increasing the number of primary neurites from neuronal soma (Koch et al. 2015). However, a similar study showed early changes, with increased neurite length associated with an increase of wildtype synuclein but not for mutant strains (Liu et al. 2013). These results suggest that α-synuclein may be involved in neurite elongation or initiation with a possible role for wild-type α-synuclein in facilitation of tubulin polymerisation. It was later demonstrated that synuclein can interact with tubulin tetramers inducing a structural reconfiguration which promotes nucleation and growth rate of microtubules. Synuclein mutations associated with PD do not undergo this folding and induce aggregation of tubulin. This study suggested that synuclein can act as a novel microtubule dynamase regulating neurite dynamics (Cartelli et al. 2016).

33

1.4.7.5 Neuroinflammation

Post-mortem studies in PD patient brains identified increased neuroinflammation and microglial activation (McGeer et al. 1988) as well as in 1-methyl-4-phenyl-1,2,3,6- tetrahydropyridine (MPTP)-induced (mechanism described in section 1.6.1) parkinsonism in humans (Langston et al. 1999) and animals (Członkowska et al. 1996). This was supported by positron emission tomography (PET) on PD patient brains that showed a significant increase in activated microglia (Ouchi et al. 2005; Gerhard et al. 2006). Furthermore, pro-inflammatory cytokines have been identified as being upregulated in PD post-mortem SNpc (Mogi, Harada,

Kondo, et al. 1994; Mogi, Harada, Riederer, et al. 1994) as well as in serum (Brodacki et al.

2008) and CSF (Blum-Degena et al. 1995).

In vivo and in vitro evidence suggests that α-synuclein itself can activate microglia (Zhang,

Wang, et al. 2005; Su et al. 2008; Austin et al. 2006). Furthermore, pathogenic mutant synucleins (A30P and A53T) can produce an elevated microglial activation, with associated increases in ROS and inflammatory cytokines (Zhang et al. 2007; Rojanathammanee et al.

2011; Hoenen et al. 2016). It is thought that extracellular α-synuclein, as a result of apoptosis or other mechanisms of release, can be engulfed by microglia, by phagocytosis in an attempt to clear deposits (Lee et al. 2008; Cao et al. 2012), or can be processed as an antigen on direct contact with cell surface receptors such as toll-like receptors (TLRs) (Fellner et al. 2013; Kim,

Ho, et al. 2013). Thus, increased α-synuclein load and increased degeneration and release of

α-synuclein may stimulate increased microglial activation. Although this microglial activation is necessary for clearing extracellular deposits, it may also contribute to increased inflammatory cytokine production and ROS, further inducing pathology.

34

In summary, α-synuclein can function in a variety of cellular processes and molecular pathways including neurotransmitter release, mitochondrial function, ER-Golgi trafficking, axonal dynamics and inflammation. As well as this, α-synuclein is targeted by several protein degradation mechanisms, with aggregation inducing the formation of multimeric α-synuclein species which are resistant to these proteolytic pathways. Thus, an accumulation of these misfolded degradation-resistant proteins is thought to contribute to pathological mechanisms of PD.

1.5 Genetic loci associated with PD

Mutations in other genetic loci have been associated with PD. These mutations are associated with several molecular pathways which can lead to the development of PD and associated Lewy body pathology. The potential interactions between genes known to be mutated in familial forms of PD and α-synuclein pathology is outlined in Figure 1.13 and discussed further in this section.

Figure 1.13 – Potential interactions between PD-associated disease genes. Mutations and multiplications of the SNCA gene leads to the accumulation of α-synuclein. This promotes oligomerisation of α-synuclein and subsequent activation of the ubiquitin-proteasome system (UPS) and endosomal-lysosomal pathways. Parkin and DJ1 are involved in UPS function and DJ1 is further linked to oxidative stress response. UCHL1 is a ubiquitin hydrolase and ligase which is also associated with UPS function. Parkin, PINK1 and DJ1 are associated with mitochondrial function which synthesis the ATP required for normal UPS function. The Leucine-rich repeat kinase 2 (LRRK2) plays a role in cellular trafficking, phosphorylation and intracellular signalling. Cytoskeletal dysregulation may also perturb clearance of accumulated protein further leading to α-synuclein pathology. Figure taken from reference (Farrer 2006).

35

1.5.1 LRRK2

Identified by genome-wide linkage analysis, mutations in the Leucine-rich repeat kinase 2

(LRRK2; PARK8), located in chromosome 12p11.2-q13.1, have also been associated with autosomal dominant PD (Hasegawa et al. 2001; Funayama et al. 2002; Zimprich et al. 2004;

Paisán-Ruıź et al. 2004). Using high-resolution recombination mapping and candidate gene sequencing, six disease-associated mutations in the LRRK2 gene locus were identified

(Zimprich et al. 2004).

LRRK2 was found to co-immunoprecipitate with PD-associated protein parkin (Smith et al.

2005). LRRK2 has been shown to regulate autophagy via a calcium-dependent mechanism

(Gomez-Suaga et al. 2011) and mutant LRRK2 has been shown to increase mitochondrial clearance by LC3-II mediated autophagy due to a deficit in calcium recovery after depolarisation (Cherra III et al. 2013). Thus, LRRK2 function is associated with regulation of mitochondria and its pathology linked to oxidative stress (Heo et al. 2010).

1.5.2 Parkin

As well as the autosomal dominant transmission seen with the SNCA and LRRK2 mutations, many accounts of familial juvenile parkinsonism (fJP) have been published (Martin et al. 1971;

Matsumine et al. 1997) with a higher risk in offspring of consanguineous marriages (Ota et al.

1958; Takahashi et al. 1994; Ishikawa et al. 1996), suggesting autosomal recessive transmission. Genome-wide linkage analysis in probands with fJP identified a microdeletion in chromosome 6q25.2-27 which encodes a 12 exon, 465 amino acid protein, aptly named

Parkin (Matsumine et al. 1997; Kitada et al. 1998). Parkin protein is thought to interact with

E2 ubiquitin-conjugating enzymes as a E3 ubiquitin-ligating enzyme essential to the formation

36

of polyubiquitin and proteasomal degradation pathways (Morett et al. 1999; Shimura et al.

2000).

Parkin mutations are associated with a slow progression and a typical age of onset of under

40, with post-mortem histological analysis showing selective loss of nigral neurons and irregular appearance of LB pathology (Farrer et al. 2001). Parkin mutations account for up to

50% of fJP cases and a wide variety of parkin mutations are associated with different clinical phenotypes (Abbas et al. 1999; Lücking et al. 2000; Lohmann et al. 2003; Hedrich et al. 2001;

Hedrich et al. 2002; Kann et al. 2002). A significant role for Parkin in the pathological process is its recruitment to damaged mitochondria to promote autophagy (Narendra et al. 2008).

1.5.3 PINK1

Association mapping of other autosomal recessive fJP patients, with clinical features similar patients with Parkin mutations, identified mutations in the PTEN-induced putative kinase 1

(PINK1; PARK6), located on chromosome 1p35-36 (Valente et al. 2001; Valente et al. 2004;

Hatano et al. 2004). The PINK1 gene was further found to contain 8 exons and encode for a protein of 581 amino-acids, which localises to mitochondrial membranes.

PINK1 can act as a sensor of damaged mitochondria and its kinase activity can induce the translocation and recruitment of Parkin, from the cytoplasm, to depolarised mitochondria to initiate mitophagy (Matsuda et al. 2010; Geisler et al. 2010). Overexpression of WT, but not mutant PD-linked PINK1 has been found to have a neuroprotective effect in neurons (Valente et al. 2004). Silencing of PINK1 has been found to induce mitophagy, dysregulation of mitochondrial morphology dynamics, membrane potential loss and an increase in oxidative stress in an in vitro dopaminergic model (Cui et al. 2011). Furthermore, mutant PINK1 or

37

PINK1 knock-down induces mitochondrial respiration deficits and can impair proteasome function along with an increase of α-synuclein aggregation (Liu et al. 2009).

1.5.4 DJ1

Homozygosity mapping from probands with autosomal recessive fJP led to localisation of a deletion and a point mutation in a gene locus encoding DJ1 protein (Bonifati et al. 2003), previously associated with a redox sensor activated in response to endogenous ROS

(Mitsumoto et al. 2001). Later studies would elucidate the neuroprotective role of DJ1 in response to oxidative stress (Taira et al. 2004; Guzman et al. 2010). DJ1 function has also been linked to modulation of α-synuclein aggregation and response to α-synuclein-induced toxicity

(Zhou et al. 2005; Batelli et al. 2008; Batelli et al. 2015). DJ1 has been noted to localise to mitochondria (Canet-Avilés et al. 2004; Junn, Jang, et al. 2009) with loss of DJ1 leading to perturbed mitochondrial function and dynamics (Krebiehl et al. 2010). Furthermore, DJ1 is thought to act with PINK1/Parkin to regulate mitophagy pathways by forming a ubiquitin E3 ligase complex (Xiong et al. 2009) or in parallel pathways (Thomas et al. 2010).

1.6 Modelling of Parkinson’s disease

PD results from the degeneration and loss of the dopaminergic neurons of the SNpc that synthesise dopamine and project to the striatum, with leading to striatal dopamine loss. The etiology of PD remains unclear with a range of diverse factors implicated including age, environmental exposures and genetic factors associated with susceptibility. As such, models have been generated to recapitulate mechanisms of PD induction and initiation of the PD neurodegenerative process. Pre-clinical modelling of PD has generally tended to use either neurotoxin-induced models to induce selective degeneration of these neurons or the perturbed expression of PD-associated genes.

38

1.6.1 Classic neurotoxin models

Early modelling of the degeneration of mDA neurons mostly used neurotoxin induced animal models (Table 1.5). One of the first of these models to study parkinsonism was the 6-

Hydroxydopamine (6-OHDA) model. Developed by Ungerstedt in 1968 (Ungerstedt 1968), administration of this neurotoxin into the substantia nigra of rats induced degeneration within the nigro-striatal system causing perturbations in motor behaviour (Luthman et al.

1989). 6-OHDA is preferentially taken up by catecholaminergic neurons via catecholaminergic transporters including the dopamine transporter (DAT) and has been shown to interact with mutated forms of α-synuclein, increasing the degenerative phenotype (Lehmensiek et al.

2006).

Neurotoxin Type Mechanism 6-OHDA Hydroxylated analogue of dopamine Neurotoxin contaminant produced Interferes with Complex I of the mitochondrial MPTP during synthesis of the synthetic electro transport chain producing increased reactive (MPP+) opioid MPPP oxygen species and dopaminergic cell death. Uptake by DAT or noradrenergic transporter NAT (no Rotenone Insecticide, piscicide, pesticide evidence of this for Rotenone) Paraquat Herbicide, Pesticide Effects proton gradient in electron transport chain – Chemical inhibitor of oxidative CCCP interfering with mitochondrial membrane potential. phosphorylation Induces the PINK1-Parkin pathway Table 1.4 – Table of neurotoxins used to generate models of PD

The neurotoxin MPTP, a by-product of the synthesis of a synthetic heroin substitute 1-methyl-

4-phenyl-propion-oxypiperidine (MPPP), was accidentally administered to drug users inducing irreversible idiopathic parkinsonism (Langston et al. 1983). MPTP is the most commonly used neurotoxin in animal models. Unlike 6-OHDA, MPTP is able to cross the blood- brain barrier (BBB) and is taken up by astrocytes where it is converted to its active MPP+ form and can selectively target the nigro-striatal system via uptake by DAT (Javitch et al. 1985;

Gainetdinov et al. 1997).

39

Agricultural chemicals such as rotenone (Sherer et al. 2003), and N,N’-dimethyl 1-4-4-

4’bypiridinium (paraquat) (Rappold et al. 2011) which are also able to cross the BBB can also induce parkinsonism and selective loss of nigral neurons in animal models with evidence of formation of LB-like inclusions in rats (Manning-Bog et al. 2002). Uncoupling reagents such as carbonyl cyanide m-chlorophenylhydrazone (CCCP), which causes mitochondrial membrane depolarisation and induces recruitment of the PD-associated ubiquitin ligase Parkin to mitochondria, has been used to assess mitophagy (Narendra et al. 2008). However, neurotoxin models rarely mimic the etiology and progression of PD and thus insights from studies using these models can be limited.

1.6.2 Genetic mouse models

The discovery of a mutation in the SNCA locus that was associated with familial autosomal dominant PD, and other familial PD associated loci spurred a genetic approach to PD modelling recapitulating disease associated genes. The mouse is the most commonly used transgenic model for pre-clinical study disease modelling and will be the subject of focus here, however there are many other non-vertebrate models and vertebrate models which are generated using the same principles outlined here.

The SNCA KO mouse model that has been generated showed a very subtle phenotype with an altered release of DA and a reduction in striatal DA levels (Abeliovich et al. 2000). Mouse models overexpressing WT α-synuclein reported accumulation of α-synuclein (Rockenstein et al. 2002) and olfactory dysfunction (Fleming et al. 2008). More recently, α-synuclein overexpression models have been able to recapitulate the age-dependant mDA neuron loss and motor impairments seen in humans (Janezic et al. 2013). A53T mutant expressing transgenic mice showed evidence of a more severe phenotype with motor disturbances

40

(Gispert et al. 2003) and aggregation of the mutant synuclein (Giasson et al. 2002; Kahle et al.

2000). Mouse models expressing truncated α-synuclein, which has been linked to increased aggregation propensity, have been characterised as having reduced striatal dopamine, whilst mDA loss was not apparent (Michell et al. 2007; Daher et al. 2009).

These studies represent some of the different α-synuclein KO, overexpression and mutation murine models which recapitulate distinct degenerative phenotypes resembling the familial human forms and can be used to characterise specific molecular neurodegenerative pathways. Many models have been generated to perturb expression of WT or mutant forms of different PD-associated genetic loci such as LRRK2 (Li et al. 2009; Kitada et al. 2009;

Ramonet et al. 2011), Parkin (Goldberg et al. 2003), PINK1 (Moisoi et al. 2014), and DJ-1 (Pham et al. 2010). Furthermore, a triple-KO parkin/DJ-1/PINK1 mouse model has been generated, although surprisingly, no evidence of mDA neuron loss or reduction in striatal dopamine levels in aged triple-KO was observed (Kitada et al. 2009).

1.6.3 In vitro modelling of Parkinson’s disease

Compared to animal models, cellular models are generally less costly, avoids the ethics of animal research and can be easier to manipulate, pharmacologically or genetically, and analysed using techniques such as microscopy. With these characteristics, cell lines can be used more efficiently to model disease for basic research as well as for high-throughput drug screening.

Both the pheochromocytoma cell line PC12, which expresses high levels of the catecholaminergic enzyme tyrosine hydroxylase (Greene et al. 1976), and the human neuroblastoma cell line SH-SY5Y which expresses dopamine-β-methyltransferase (Biedler et al. 1978) have been used extensively for in vitro PD modelling in the past. Both can be

41

differentiated towards a neuronal phenotype and can take up and release catecholamines

(Balasooriya et al. 2007). The SH-SY5Y line is preferred as it is a human cell line and has been shown to upregulate neuronal and dopaminergic markers upon differentiation including

Tyrosine hydroxylase (TH) and DAT (Lopes et al. 2010; Korecka et al. 2013). SH-SY5Y have been used to model disease using the classic neurotoxin models such as MPP+ exposure (Wu 2009), as well as to generate genetic models such as α-synuclein PD models (Hasegawa et al. 2004;

Desplats et al. 2009; Ding et al. 2013; Melo et al. 2017).

A noteworthy alternative line is the mouse immortalised neural progenitor line MN9D which has also been shown to have dopaminergic (Hermanson et al. 2003) and neuronal properties

(Choi et al. 1999). This line is advantageous compared to primary neurons from embryonic mammalian mesencephalon as they are easier to obtain and can be maintained with reproducibility and consistency.

1.6.4 Pluripotent stem cell models: hESC and hiPSC

Mammalian genomes can be highly conserved and animal models have proven invaluable to study disease however, humans and mice for example harbour developmental, genetic and physiological differences. An alternative to the use of the animal models and cell lines mentioned in section 1.6.3 presented itself when embryonic stem cells (ESCs) were first isolated from the inner cell mass of pre-implantation mouse embryos by Martin in 1981.

These cells were able to form teratocarcinomas containing derivatives of all three germ layers, including undifferentiated stem cells, endoderm, connective and epithelial tissues (Evans et al. 1981). It was much later, in 1998, that researchers were able to isolate human embryonic stem cells (hESCs) from cleavage stage embryos produced from in vitro fertilisation, generating the first hESC lines (Thomson et al. 1998). The ability to culture primary, self-

42

renewing pluripotent cells offered scientists an unlimited quantity of easily obtainable cells with the potential to be differentiated into any cell fate for in vitro modelling (Figure 1.14).

These cells were shown to express cell surface markers SSEA3 (Shevinsky et al. 1982), SSEA4

(Kannagi et al. 1983), TRA-1-60 and TRA-1-81 (Andrew et al. 1984), characteristic of human embryonal carcinoma lines and nonhuman primate stem cells (Thomson et al. 1995). These

ESCs have been found to express the core transcriptional regulatory circuitry characteristic of pluripotency (Boyer et al. 2005) which include the pluripotency genes OCT4 (Nichols et al.

1998), SOX2 (Yuan et al. 1995; Avilion et al. 2003), NANOG (Chambers et al. 2003; Mitsui et al. 2003).

A method for reprogramming adult somatic cells to induce a phenotype characterised by the expression of stem cell specific pluripotency genes and cell surface markers was discovered by Takahashi et al. in 2006, originally in mouse cells (Takahashi et al. 2006) and later from human somatic cells (Takahashi et al. 2007). These cells were reprogrammed by retroviral induction of 4 genes (Oct3/4, Sox2, Klf4 and c-Myc). They were phenotypically characteristic of embryonic stem cells with normal karyotypes and the ability to express stem cell markers, self-renew and differentiate into any cell type. Whilst ESCs require genetic manipulation or pre-implantation genetic screening to model diseases, iPSCs offered the use of somatic cells from patients who harbour mutations and therefore more physiologically-relevant disease modelling. Many patient-derived iPSC-lines have been generated with PD-associated gene defects for SNCA (Devine et al. 2011; Byers et al. 2011) as well as LRRK2 (Nguyen et al. 2011),

PINK1 and Parkin (Chung et al. 2016). Modern genome editing technologies has also enabled the generation of isogenic controls, differing exclusively in disease-associated gene expression (Soldner et al. 2011; Reinhardt et al. 2013; Ryan et al. 2013). The ability to produce

43

autologous patient-derived cells and differentiate them offered the opportunity to study disease mechanisms in vitro with the added benefit of providing a model in which cell-specific phenotypes could be analysed.

Figure 1.14 - Sources for cell replacement therapies and modelling of Parkinson’s disease. Sources of cells include human fetal mesencephalic tissue, pluripotent stem cells (ESCs and iPSCs) differentiated towards mDA neurons and direct fibroblast to mDA neuron reprogramming. Fetal cells and mature neurons can them be transplanted directly into the adult brain. mDA neurons can then be used for PD modelling for investigating disease aetiology or for novel drug discovery. Neural plate (NP), neuroepithelial cells (NEP), midbrain floor plate (mFP), midbrain dopaminergic (mDA), midbrain dopaminergic progenitors (mDP), mDA postmitotic neuroblasts (mDNb). Figure taken from (Arenas et al. 2015)

1.6.4.1 Differentiation of stem cells towards a midbrain dopaminergic fate

PD is characterised clinically by loss of dopaminergic neurons of the substantia nigra and thus, to model PD, ESCs and hiPSCs would need to be driven towards a generalised neural fate and specified to a ventral midbrain fate. To achieve this, researchers have attempted to recapitulate developmental cues (described in section 1.3.1). Earlier studies using mESCs used co-culture with stromal cell lines (Kawasaki et al. 2000) or transplantation of undifferentiated cells into mouse rat striatum (Björklund et al. 2002) to achieve a neural fate. Other methods included the sequential addition of FGF8 and SHH patterning molecules, recapitulating regional specification in vivo (Lee et al. 2000; Barberi et al. 2003) or exogenous overexpression

44

of transgenes to express mDA-specific transcription factors such as Nurr1 or FOXA2 (Kim et al. 2002). Concurrently, derivation of dopaminergic neurons from hESCs was being evaluated and early studies showed promising mDA differentiation potential of hESCs with the use of stromal co-culture lines as well as sequential addition of patterning molecules such as FGF8 and SHH (Perrier et al. 2004) or neurotrophic factors (Park et al. 2004) and demonstrated that

ESC fates can be relatively easily manipulated by the addition of exogeneous factors.

As in normal development, inhibition of BMP antagonist noggin, found in the Spemann organiser, was shown to promote neuralisation in hESC cultures (Pera et al. 2004; Gerrard et al. 2005; Chiba et al. 2008). Inhibition of Activin/Nodal signalling was also able to drive hESCs towards a neuroectoderm fate using a Nodal (TGF-β family) receptor antagonist (SB431542)

(Smith et al. 2008). The combination of Noggin and inhibition of Activin/Nodal signalling by

SB431542 generated a highly efficient process of neuralisation in hESC cultures (Chambers et al. 2009). SMAD signalling is the main signalling transducer for TGF-β receptors and combining these two compounds was able to block SMAD 2/3 and SMAD1/5/8. Modest mDA patterning was also reported by activation of SHH/FGF8 signalling.

Highly efficient generation of mDA neurons from pluripotent stem cells has recently been described from hESCs and hiPSCs (Figure 1.15). This was achieved by the dual SMAD inhibition protocol combined with dose-specific activation of SHH signalling alongside a potent GSK3β inhibitor and activator of Wnt signalling. Canonical Wnt played an important role in generating FOXA2/LMX1A expressing mDA progenitors whilst Fgf8 signalling did not play a significant role in specifying mDA progenitors from pluripotent stem cell-derived cultures

(Kriks et al. 2011; Kirkeby et al. 2012). The specified neural precursors were terminally differentiated to express neuronal markers microtubule associated protein 2 (MAP2) and

45

neuron-specific class III β-tubulin (TUJ1) and dopaminergic markers TH. The induction of mature neuronal morphologies was achieved by the synergistic action of the Notch inhibitor

DAPT (Crawford et al. 2007) and dibutyryl-cAMP (db-cAMP) (Sharma et al. 1990; Stachowiak et al. 2003; Kim et al. 2011). For functional maturation, growth factors, GDNF and BDNF were included after specification to increase the survival (Sautter et al. 1998; Erickson et al. 2001).

In conclusion, to generate mDA neurons from pluripotent stem cells, soluble factors can be used to control the main signalling pathways implicated in the specification of mDA neurons

(BMP, SHH, WNT) during normal development and to achieve a functional maturation to neurons a mix of neurotrophic factors, Notch and CREB signalling can enhance and promote mDA neuron development in ESC cultures in vitro.

Figure 1.15 – Differentiation towards specific neural cell types.Schematic showing generalised principles of A) specification of pluripotent stem cells towards specific neural lineages by in vitro recapitulation of neurodevelopmental molecular signalling pathways. B) The in vitro process is thought to recapitulate organisation related to neural tube development. C) Specified cells can be matured in culture to generate distinct neuronal sub-types as well as somatic lineages. Figure taken from (Broccoli et al. 2014).

46

1.6.5 Utility of lentiviral vectors for disease modelling

Overexpression of synuclein by viral vector gene delivery, commonly adeno-associated viruses (AAV) and lentiviruses (LV), offers an alternative to generate models and for gene- therapy therapeutic approaches. These vectors are highly efficient in transducing neurons and stem cells for stable, sustained transgene expression in both dividing and non-dividing cells.

Although transgenic mice have been generated to model synuclein associated-PD, the mouse models do not always produce the same prominent, progressive neurodegenerative phenotype of the human disease. For example, some synuclein models lacked LB pathology or TH neuron degeneration (Masliah et al. 2000). The same is true for patient-derived iPSC, where mild phenotypes have been reported for patient-derived lines with an SNCA triplication (Byers et al. 2011) and the A53T mutation (Ryan et al. 2013; Haenseler et al. 2017), probably due to the progressive, age-dependent changes usually associated with the human disease.

Rodent models of synucleinopathy generated with the use of viral vectors have been able to recapitulate the main SNpc-associated characteristic phenotype of PD including progressive mDA cell loss, LB-like inclusions and motor impairments (Kirik et al. 2003; Bianco et al. 2002;

St Martin et al. 2007; Decressac, Mattsson, et al. 2012; Oliveras-Salvá et al. 2013). In vivo disease modelling using viral vectors requires injection of vectors to the site of required expression which can introduce some variability, however, the use of cell-specific promoters can be used to drive cell-type specific expression and give increased control of gene expression. Due to the heterogeneity of many models, neuron-specific promoters have been employed to drive transgene expression in neurons (Dittgen et al. 2004). Furthermore, the use of viral vectors has been employed in in vitro cellular models such as SH-SY5Y cells (Bianco

47

et al. 2002; Han et al. 2011), primary neurons (Tönges et al. 2014) and iPSCs (Mazzulli et al.

2016) to model disease or for testing of efficacy of therapeutic agents.

In this study, 3rd generation lentivirus vectors were used to as delivery tools of mutated α- synuclein for PD modelling in SH-SY5Y cells and iPSC-derived neurons (Naldini et al. 1996).

Lentiviruses, a genus of retroviruses, are able to stably integrate into host genome of both dividing and non-dividing mammalian cells and, as such, have been widely used as gene delivery vectors. Lentiviral vectors have been used to transduce a wide variety of CNS neurons

(Consiglio et al. 2004; Jakobsson et al. 2003; Dittgen et al. 2004). Vectors contain 2 copies of positive viral ssRNA, have been generated to be safe and provide sustained expression of transgenes for gene therapy and overexpression models. The 3rd generation lentiviral vectors include only 3 of the 9 HIV-1 genes and these are contained within the 4 separate trans acting plasmid vectors (Figure 1.16) that include:

1) a packaging plasmid which contains the gag (group-specific antigen) HIV-1 gene that

produces matrix, capsid and nucleocapsid structural proteins and the pol (DNA

polymerase) HIV-1 gene whose products include reverse transcriptase, integrase and

protease.

2) An envelope plasmid containing a heterologous env gene of another virus encoding

the G-protein of the vesicular stomatitis virus (VSV-G) protein in a separate vector.

The envelope protein enables attachment to host cell, fusion with the endosomal

membrane and viral entry. This particular G-protein confers upon the lentiviral particle

a broader tropism, as VSV-G interacts with a ubiquitous cellular receptor

(Mastromarino et al. 1987; Mifune et al. 1982). VSV-G can also increase the stability

of lentiviruses allowing for generation of higher viral concentration (Burns et al. 1993).

48

3) A regulatory construct containing the rev (regulator of expression of virion proteins)

HIV-1 gene, important for synthesis of viral proteins and promotes the cytoplasmic

accumulation of viral transcripts thus aiding in production of viral proteins. Rev

connects a Rev-responsive element (RRE) RNA motif to nuclear export machinery.

4) A lentiviral backbone or transfer vector plasmid, driven by a strong promoter, with

long terminal repeats (LTRs), which mediates integration of retroviral DNA into host

genome. The lentiviral backbone used in this study was the pRRL self-inactivating

recombinant system, wherein the U3 promoter in the 3’-LTR is deleted, which is

integrated into the genome resulting in transcriptional inactivation. This backbone

contains a chimeric Rous sarcoma virus (RSV)-HIV-1, truncated 5’ LTR, the packaging

signal Psi (ψ) which regulates packaging of retroviral RNA genome into the viral capsid

and the RRE motif. It contains a simian virus 40 (SV40) polyadenylation and origin of

replication sequence downstream of the 3’ LTR and a Woodchuck hepatitis virus

posttranscriptional regulatory elements (WPRE) inserted just before the 3’ LTR to

enhance transgene expression.

49

Figure 1.16 – Schematic representing the structure of a lentivirus and the plasmids used to express the different components to generate lentiviruses. Essential components include structural proteins encoded by a packaging plasmid (rev, gag and pol), an envelope protein encoded by the envelope plasmid, as well as the protein of interest encoded by a a lentiviral transfer plasmid flanked by an LTR which is integrated into the host genome during transduction. 1.7 Summary and objectives

The neurodegenerative disorder PD is an incurable disease, causing devastating motor symptoms which can have a severe impact on a patient’s health and wellbeing, with limited therapies. Thus, understanding the etiology of the disease and the pathways which are perturbed during pathology is paramount to the development of new therapies, especially ones which may help to prevent or slow disease progression. Research has identified many aspects of the pathological pathways associated with PD, including associated genetic loci and functional disturbances in mitochondrial, protein degradation and neurite dynamics.

Animal and cellular PD models have been developed to determine disease-associated pathways. These models can be generated using different pharmacological or genetic manipulations including the use of lentiviral vectors to perturb different systems. Most recently, human pluripotent stem cells, differentiated towards midbrain dopaminergic neurons have been shown to provide unrivalled physiologically relevant models to investigate

50

molecular mechanisms of disease progression. Recent advancements in genome-wide analysis of gene expression including large dataset analysis pipelines have been developed and employed to characterise disease-associated molecular perturbations in order to identify novel biomarkers or new therapies. Furthermore, gene expression may be regulated at the translational levels and thus techniques to study actively translating mRNA have been developed in conjunction with next generation sequencing (NGS) technologies. These translatomic preparations reflect the actual functional genomic readout of a given cell type.

In order to further understand the molecular mechanisms and pathways that are activated by

-synuclein to contribute PD pathology, this PhD will describe a body of work with the following aims:

1) to establish cellular models of -synuclein overexpression using genetic tools and iPSC

technology;

2) to determine genome-wide translatomic changes in cellular human PD models by NGS

technology, in conjunction with translating mRNA purification; and

3) to profile miRNA levels in human iPSc neuron cultures that overexpress wild type and

mutant -synuclein.

51

Chapter 2 - Materials and Methods 2.1 Generating lentiviral plasmids and particles for affinity purification, promoter testing and disease modelling

For this project numerous plasmids were generated for the production of lentiviral particles to infect both SH-SY5Y cells and differentiated iPSC cells. Lentiviruses were generated to test different promoters in their ability in maintaining long-term expression of transgenes in these cell types (Table 2.1; Figure 2.1). Once the optimal promoter was ascertained, this was cloned into a lentiviral backbone plasmid (Naldini et al. 1996) for disease modelling and driving expression of L10a-EGFP fusion protein for TRAP (Figure 2.1).

SH-SY5Y iPSC-derived Disease modelling and Disease modelling in promoter neuron promoter TRAP in SH-SY5Y iPSC-derived neurons testing testing CAG-EGFP CAG-EGFP CMV-L10a-EGFP SYN1-Empty control CMV-EGFP CMV-EGFP CMV-α-synucleinWT SYN1-α-synucleinWT EF1α-EGFP EF1α-EGFP CMV-α-synucleinA53T SYN1-α-synucleinA53T PGK-EGFP PGK-EGFP SYN1-α-synucleinG51D SYN1-EGFP Table 2.1 -List of lentiviral backbone plasmids and lentiviral particles cloned and generated for promoter testing and disease modelling in both SH-SY5Y cells and iPSC-derived neurons. The CMV-EGFP, PGK-EGFP lentiviral backbone plasmids were gifts from Dr. Liang-Fong Wong and the CMV-α-synucleinWT, CMV -α-synucleinA53T and the CMV-L10a-EGFP lentiviral backbone plasmids were gifts from Dr. Helen Scott. All other plasmids were cloned and amplified, and lentiviral particles generated.

Figure 2.1 – Lentiviral backbone plasmids. Schematic of the lentiviral backbone plasmids used to generate lentiviral particles to drive A) EGFP expression for promoter testing and L10a-EGFP expression for TRAP and B) α-synuclein for disease modelling in SH-SY5Y cells and iPSC-derived neurons.

52

The 3rd generation lentiviruses were produced using 4 separate trans acting plasmid vectors to express the necessary machinery to create lentiviral particles. These are the packaging, envelope and regulatory plasmids as well as a lentiviral backbone plasmid (Figure 2.1).

2.1.1 Polymerase chain reaction (PCR)

To produce double stranded DNA fragments for cloning into the lentiviral backbone plasmid, the Fast start high fidelity PCR kit (Roche) was employed. A no template control was always included, and reaction reagents and conditions are included (Table 2.1 and 2.2).

PCR reagents Quantity Supplier 10x Reaction Buffer (+Mg2+) 5μl Roche Template 50ng (1μl) Roche DMSO 4μl Roche Forward Primer (10μM) 1μl Eurofins genomics Reverse Primer (10μM) 1μl Eurofins genomics dNTPs 1μl New England Biolabs Taq Polymerase enzyme mix 0.5μl Roche H2O 36.5μl Sigma Total 50μl PCR cycle conditions Temperature Time Initial denaturing 95°C 3 mins 40 cycles: Denature 95°C 30 secs Anneal 55-65°C 30 secs Extension 72°C 90 secs Final extension 72°C 10 mins Table 2.2 – Fast start high fidelity PCR kit reagents used for PCR of fragments for cloning and conditions used for the PCR reaction. 50ng of template was used with forward and reverse primers designed to add restriction sites to the end of the fragment insert of interest.

Annealing temperatures used were between 55-65°C according to primer pair melting temperature. Final extension was 10 mins as the plasmid used was approximately 10kb in size (1min/kb).

2.1.2 Electrophoresis of PCR product

To separate and isolate the PCR insert fragment for cloning, PCR product was loaded into a

1%w/v agarose gel and run at 100V for approximately 1 hour. The gel was formed by

53

dissolving Ultrapure™ Agarose (Invitrogen; 15510-027) in 1x 40mM tris-acetate, 1mM EDTA

(TAE buffer; Geneflow; B9-0030) diluted in ultrapure H2O and heated in a microwave.

Ethidium bromide (Sigma; E1510; 0.5μg/ml) was added to the gel mix prior to casting and a comb inserted for loading of PCR product. The gel was left to set and then placed in an electrophoresis tank filled with 1x TAE buffer and the comb removed. The PCR product was then mixed with a 6x DNA gel loading dye (NEB; B7025S) and loaded into the wells alongside a 1kb molecular weight marker ladder (NEB; N3232L). The bands were visualised with a UV transilluminator and captured using GelDoc-It bioimaging system (UVP) with VisionWorks analysis software.

2.1.3 DNA extraction from Gel and determination of concentrations

Once PCR product was visualised with a UV transilluminator and molecular weight ladder used to confirm appropriate size, the bands were excised with a scalpel and weighed in a 1.5ml

Eppendorf using the QIAquick gel extraction kit (QIAgen; #28706) was used to isolate fragment of interest according to manufacturer’s instructions. Briefly, 3x volume of buffer

QG was added (100mg gel ~ 100μl) and incubated at 50°C until gel dissolved. 1 gel volume of isopropanol was added and DNA bound to a QIAquick centrifuge column by centrifugation and washed with buffer PE with another centrifugation step. A quick centrifugation was applied to remove residual ethanol from buffer PE and 30-50μl of H2O was applied to the spin column to elute the DNA into a fresh tube. All centrifugation steps were performed at 10,000g and flow-through was discarded. The DNA concentration was determined using the Nanodrop

2000 UV-Vis spectrophotometer (Thermo Fisher).

54

2.1.4 Restriction enzyme digestion of DNA for recombinant DNA cloning

Restriction enzymes (NEB; Roche) were used on PCR product (insert fragment) and plasmid vector DNA to produce linearised plasmid vector backbone DNA with complimentary sticky ends to the PCR product that can be manipulated for cloning by ligation by formation of hydrogen bonds (Figure 2.2). Thus, DNA was digested by suitable restriction enzymes using

>1 Unit of restriction enzyme for each 1μg of DNA, in recommended reaction buffer provided by the manufacturer, for a minimum of 2 hours to overnight at 37°C. Digested plasmid vector

DNA was electrophoresed on a 1% agarose gel as in section (2.1.2) and the linearised backbone DNA extracted from gel as in section (2.1.3) and eluted in 30-50μl H2O. Digested

PCR product was purified with the QIAquick PCR purification kit (QIAgen; #28106) according to manufacturer’s instruction. Briefly, 5 volumes of buffer PB was added to 1 volume of PCR product and bound to QIAquick spin column by centrifugation. DNA was washed with buffer

PE and then centrifuged for another minute to remove excess ethanol from buffer PE and 30-

50μl H2O applied to the spin column and eluted by centrifugation into a fresh tube. The DNA concentration was determined using the Nanodrop 2000 UV-Vis spectrophotometer.

55

Figure 2.2 – Cloning strategy for cloning EF1α promoter into a plasmid which contains the CMV promoter to express EGFP. A) Schematic of the lentiviral backbone plasmid with cut sites. B) SerialCloner software virtual cutter tool showing the two restriction enzymes used, cut sites and fragments generated. After linearising the plasmid and cutting out the promoter, the linearised plasmid was run on an electrophoresis gel and extracted for ligation of different promoters. C) The primers used to PCR the EF1α promoter, whilst adding complimentary restriction enzyme cut sites for ligation with the linearised backbone plasmid for cloning.

2.1.5 Ligation of DNA

To create a stable circular plasmid, insert fragments and linearised lentiviral backbone plasmids with complimentary sticky ends were ligated together prior to transformation.

Ligations were performed using the Rapid DNA Ligation Kit (Roche; #11635379001) according to manufacturer’s instructions. The equation used to calculate the ratio of vector to insert concentrations to use for a 1:1, 1:3 or 1:5 was as follows:

56

Vector(150ng)/Size 1 1 1 = 표푟 표푟 Insert (x ng)/Size 1 3 5

Briefly, a 1:1, 1:3 and 1:5 ratio of vector (150ng starting concentration) as well as an insert only and a vector only controls were mixed together in a centrifuge tube and H2O added to

8μl. 2μl of 5x DNA dilution buffer was added to bring to a 1x concentration and mixed. 10μl of DNA ligation buffer was added to this and mixed. 1μl of the T4 DNA ligase enzyme was then added and incubated at room temperature (RT) for 10 minutes and then placed on ice to stop the reaction.

2.1.6 Transformation into chemically competent Escherichia coli

DH5α (Invitrogen; #18265-017) bacterial cells (or SCS110 Competent Cells (Stratagene;

#200247) for plasmids with methylated sticky end sites) were used for transforming ligated plasmid DNS (section 2.1.5). Transformations were performed around a Bunsen burner, using aseptic technique. For minipreps 50μl aliquots of DH5α bacterial cells in plastic centrifuge tubes were thawed on ice and 10 μl of the ligase reaction was mixed in and incubated on ice for 30 minutes. Following this the cells and ligation mix was heat shocked at 42°C in a water bath for up to 60 seconds and placed back on ice for 2 minutes. 100pg of pUC18 control plasmid was applied to an aliquot of cells and the same protocol performed in parallel as a positive control. 1ml of Lysogeny broth was added and incubated at 37°C for an hour with frequent mixing every 10 minutes (Table 2.3). Following this the cell/LB mix was centrifuged at 5000rpm for approximately 45 seconds and 900μl removed with the remaining mix used to resuspend bacterial pellet. The mixture was plated onto agar plates with ampicillin

(100μg/ml) and plates left in an incubator overnight at 37°C (Table 2.3).

After overnight culture, single bacterial colonies were picked and transferred into 5ml LB with

100μg/ml of ampicillin and placed in a rotating incubator 180rpm at 37oC. LB overnight

57

cultures were then centrifuged and minipreparation was continued using the Qiaprep spin miniprep kit (Qiagen; #27104) as per manufacturer’s guidelines. Briefly, overnight cell/LB mix was centrifuged at 6800g for 3 minutes to harvest cells. Pelleted bacterial cells were lysed in

250μl buffer P1 with RNase A. 250μl of buffer P2 was then added and mixed by inversion followed by 350μl of buffer N3 and mixed likewise. This mix was centrifuged at 17,900g and

800μl of supernatant applied to a QIAquick spin column and centrifuged for 60 seconds to bind DNA. Buffer PE was used to wash the DNA and centrifuged again for 1 minute to remove excess ethanol from buffer PE and then 30-50μl of H2O applied to the spin column and centrifuged in a new centrifuge tube to collect purified DNA. For SCS110 competent cells, the same procedure was used but 1.7μl of 50mM β-mercaptoethanol added to each aliquot of cells and heat pulse was for 45 seconds.

Lysogeny Broth Quantity Tryptone 1g Yeast Extract 0.5g NaCl 0.5g 1mM NaOH 0.1ml Agar (for plating) 1.5g Water 100ml Table 2.3 – LB Broth components for growing transformed E. coli and Agar components for plating transformed E. Coli. For Agar, LB broth components had Agar added and mixed prior to plating.

2.1.7 Maxi Preparations

Maxi preparations were used to isolate larger concentration yields of lentiviral plasmid DNA encoding the different promoter-EGFP constructs, as well as L10-EGFP, and α-synuclein. Maxi preparations were also used to amplify packaging plasmids (Gag/pol & Rev) and envelope plasmids (VSV-G) from bacterial cultures. A plasmid encoding L10a-EGFP was generated and kindly provided by Dr. Helen Scott.

For maxi preparations, 200ng miniprep product was transformed into an aliquot of OneShot

Stbl3 (Thermo Fisher; #C7373-03) E. coli bacterial cells as with DH5α (section 2.1.6) with heat

58

shock for 45 seconds and plated onto agar plates with 100μg/ml ampicillin and incubated at

37˚C overnight. The next day, a single colony was picked and transferred into 5ml LB with

100μg/ml of ampicillin and placed in a rotating incubator at 180rpm at 37˚C for 8 hours.

Following this, the incubated LB/cells culture was transferred into 500ml LB with ampicillin in a large conical flask and incubated in a rotating incubator at 180rpm at 37˚C overnight.

The next day, overnight LB culture was pelleted at 4000rpm for 30mins at RT. Following pelleting, supernatant was disposed, and pellet resuspended in 18ml Solution 1 and 2ml

Lysozyme (Table 2.4) and 40ml Solution 2 and mixed by inversion 6 times and incubated for

10mins at RT. After 10mins, 20ml of ice cold Solution 3 was added and mixed and incubated on ice for 10mins. After a further 10 minutes incubation, the mixture was centrifuged at

4000rpm for 20mins at 4˚C. Supernatant was filtered through Whatman filter paper, and 0.6 volumes of isopropanol was added and mixed gently and left to incubate at RT for 10mins.

After 10mins, filtrate was centrifuged at 6000rpm for 20mins at RT. Supernatant was then disposed and pellet was allowed to dry for 10mins and resuspended in 3ml TE buffer (10mM

Tris-Cl, pH8.0, 1mM EDTA) on a shaking platform.

Once dissolved, 2-step Caesium Chloride-Ethidium Bromide gradient centrifugation was used to purify DNA in TE buffer (Garger et al. 1983). Briefly, 3g of CsCl was added to the 3ml of DNA in TE buffer followed by the addition of ethidium bromide (0.08ml per ml of DNA in TE/CsCl).

Once, CsCl was dissolved, the mixture was centrifuged at 6000rpm for 12 minutes at RT.

Supernatant was placed into an OptiSeal Ultracentrifuge tube and TE/CsCl added to fill the tube. The tubes were centrifuged at 65000rpm for 16 hours at 20°C using an NVTi65 Rota,

Optimal LE-80k Ultracentrifuge; Beckman Coulter). Once removed from the centrifuge, the

DNA band, bound to EtBr was excised and placed into a fresh OptiSeal Ultracentrifuge tube

59

and again TE/CsCl used to fill to the top and, this time, centrifuged for 6 hours with the same parameters. The purified band was excised with a syringe and H2O saturated butanol added to equal volume of DNA/EtBr and mixed vigorously followed by transferring of the purified

DNA band to a fresh tube 3 times. H2O was added to a volume of 10ml and 35ml Ethanol was mixed with inversions. This mixture was left to stand at 4°C in a fridge for an hour and spun at 6000rpm for 20mins at 4°C. After spin, supernatant was decanted and 750ul of 70% ethanol was added to the pellet and centrifuged at 6000rpm at 4°C for a further 5mins. Ethanol was then removed and pellet allowed to air dry to remove excess ethanol. The pellet was finally resuspended in 300μl H2O by allowing pellet to slowly dissolve overnight. Concentration was determined by spectrophotometry.

Solution 1 100ml Lysozyme 10ml Solution 2 100ml Solution 3 100ml 5M Potassium Glucose 0.9g Lysozyme 0.1g 10M NaOH 2ml 60ml Acetate pH4.8 1M Tris Glacial Acetic 1M Tris pH8.0 2.5ml 100ul 20% SDS 93ml 11.5ml pH8.0 Acid O.5M EDTA 2.0ml H2O 9.9ml H20 5ml H2O 28.5ml H2O 95.5ml Table 2.4 -List of solutions used to isolate plasmid DNA transformed and amplified by transformation into Stabl3 competent E. Coli by Maxi preparation.

2.1.8 Culture of HEK293T cells and lentiviral preparation

Figure 2.3 – Schematic of the self-inactivating (SIN) lentiviral backbone plasmid with ampicillin resistance used as visualised by SerialCloner graphics visualisation tool.

60

Transgenes and promoters were cloned into a self-inactivating (non-replicating) 3rd generation lentiviral backbone vector (pRRL.SIN.cppt) (Naldini et al. 1996)(Figure 2.3).

HEK293T cells (passage < 25), maintained in media containing DMEM with 10% FBS supplemented with 1:100 Glutamax, Non-Essential Amino Acids (NEAA) (Gibco; #11140-050) were trypsinised (Trypsin-EDTA; Gibco; #15400-054) and 8 x 106 cells were counted on a

2 haemocytometer plated onto 15cm plates, and incubated at 37°C/5% CO2. Lentiviral backbone plasmid DNA (10µg), Gag/Pol plasmid DNA (10µg), Rev plasmid DNA (2µg) and VSV-

G (3.4µg) were mixed in 1.125ml and added 140µl 2M CaCl2. 1.125ml of Hepes-buffered saline

(HBS; 50mM HEPES, 280mM NaCl, 1.5mM Na2HPO4) was mixed in dropwise and mixed and mixture was left at RT for 30mins to create DNA precipitate of which 2.4ml of DNA/HBSS mix is added onto HEK293T cells in 15cm2 plates, dropwise and dispersed, and incubated overnight at 37°C for uptake by endo/phagocytosis (Table 2.5).

The following day, media was replaced with HEK293T maintenance media with 10mM Sodium

Butyrate for 6 hours to increase DNA release into the cytoplasm. The media was harvested and placed at 4°C overnight and media was replaced with normal HEK293T maintenance media. The next day, media was harvested and pooled and centrifuged at 1000rcf for 5 mins

(Megafuge-20R; Heraeus Sepatech) and filtered supernatant using a filter bottle. The filtrate was spun at 4°C at 6000g overnight in a JLA10.5 rotor (Beckman Coulter). Supernatant was disposed safely and pellet resuspended in ice cold Phosphate Buffered Saline (PBS; Thermo

Fisher; #14190-144) and transferred into ultraclear ultracentrifuge tubes and spun at

20,000rpm for 90mins at 4°C, in a SW40Ti rotor (Beckman Coulter). Supernatant was gently poured off and pellet resuspended by covering with 100μl TSSM buffer consisting of

20mM Tris, pH 7.3, 100mM sodium chloride, 10mg/ml sucrose and 10mg/ml mannitol for 2-

61

5hrs followed by the use of gentle pipetting and making up to 192μl and aliquoted for storage at -80°C.

Backbone Gag/Pol Env Rev 2M CaCl2 H2O HBS 2x plasmid Make up For 1 10ug 10ug 3.4ug 2ug 140ul to 1.125ml plate 1.125ml For Make up 12.5x 125ug 125ug 42.5ug 25ug 1.75ml 15ml to 15ml plates Table 2.5 -Quantities of each plasmid and media for lentiviral particle generation in HEK293T cells for 1 plate and the 12.5x master mix used for 12 plates.

2.1.9 Lentiviral titering by fluorescence activated cell sorting (FACS)

HEK293T cells were cultured as per section 2.1.8 and then seeded into 10 wells of a 12 well plate at 7.5 x 104/well for each lentivirus for titering and incubated overnight. The next day these wells were transduced using serial dilutions of 10-3–10-6 of concentrated lentivirus in

TSSM buffer, as well as an untransduced control (all in duplicates) in 500μl HEK293T culture media for 2 hours and then topped up with 1ml media. An untransduced control well was trypsinised and dissociated to single cells suspension and counted for subsequent calculation.

On day of flow cytometry, cells from each well individually were trypsinised, centrifuged at

1000rcf for 2mins, had supernatant removed and pellet washed with PBS and centrifuged again. Following PBS wash, cells were fixed by resuspending in 1ml of 4% paraformaldehyde solution (Sigma; #P6148) in PBS for 15 minutes. Cells were then pelleted, washed and resuspended in 1 ml of PBS and transferred to flow cytometry tubes (BDBioscience; #352052) for flow cytometric analysis of expression of EGFP using the FACS Canto II (BD Bioscience).

After recording the percentage of GFP positive cells for each dilution, titre was calculated using the following equation, and averaged for the dilutions:

푇푟푎푛푠푑푢푐푖푛푔 푢푛푖푡푠 푡표푡푎푙 푐푒푙푙 푐표푢푛푡 ∗%퐺퐹푃 푝표푠푖푡푖푣푒 = 푚푙 푑푖푙푢푡푖표푛

62

2.1.10 Lentiviral titering by Taqman qPCR

To determine the titre of the non-EGFP expressing lentiviruses, a Taqman (ThermoFisher) qPCR-based approach was used in which the number of integration sites was analysed using a standard curve from plasmid DNA of known concentrations (10-4-10-9). HEK293T were cultured (as described in section 2.1.8), and plated at 7.5 x 10-4 cells per well of a 12 well plate.

Wells were transduced, in duplicate, with the lentivirus for titring and with a lentivirus of known titre (by FACS) at a dilution of 1:1000 and 1:5000 diluted in 500μl of HEK293T culture media and after 2 hours topped up with 1ml of culture media. An untransduced control was also included. After 3 days, wells were passaged at a ratio of 1:5 and passaged further every

2-3 days at this ratio 3 more times to remove any unintegrated lentiviral particles in suspension. These cells were collected by centrifugation and washed with 1ml PBS and finally extracted using the Wizard DNA extraction kit (Promega) using manufacturer’s instructions.

PCR reactions were set up, in triplicate, using 100ng of genomic DNA from duplicates of each dilution of lentivirus and diluted plasmid DNA to create a standard curve (Table 2.6).

1X Stock Volume Cycling conditions concn (µL) TaqMan Universal 2X 12.5 Temp Time Cycles PCR MasterMix Forward Primer 10µM 1.25 95°C 10 min 1 Reverse Primer 10 µM 1.25 95°C 15 sec 40 Probe 6.25 µM 1 60°C 2 min H2O - 4 20ng/ DNA 5 µL Total 25 Table 2.6 -Taqman qPCR reaction components and cycling conditions for qPCR-based lentiviral titre determination 2.2 Generating cellular PD models

To test our TRAP platform we are using a well-established cellular model for PD: the SH SY-5Y neuroblastoma cell line. The SH-SY5Y cell line has become a popular cell model for

63

neurodegenerative disease research because this cell line possesses many characteristics of

DAergic neurons. Indeed, most of the major PD disease pathways are intact in the SH-SY5Y genome. For TRAP, having large quantities of phenotypically homogenous cells is a major advantage. Furthermore, this cell line can be differentiated into a functionally mature neuronal phenotype, with increased neuritic outgrowth, in the presence of Retinoic Acid (RA).

Upon differentiation, SH-SY5Y cells stop proliferating and a constant cell number is subsequently maintained. Gene overexpression of SNCA would allow the study of these genes in a model which preserves many neuronal and dopaminergic pathways.

2.2.1 Culturing SH-SY5Y cells

SH-SY5Y cells were cultured in media consisting of Dulbecco’s Modified Eagle’s Medium

(DMEM; Sigma; #D6546) supplemented with 10% Fetal Bovine Serum (FBS; ThermoFisher;

#10500064), 1% Glutamax (Gibco; #35050-061) and 1% non-essential amino acids

(ThermoFisher; #11140035), incubated at 37°C/5% CO2. Cells were allowed to proliferate until they reached approximately 80% confluency with culture media changed once every 3 days.

At 80% confluency, media was aspirated and cells washed with PBS and incubated with

Trypsin (Trypsin-EDTA; Gibco; #15400-054) for 3-5 minutes. Cells were then mechanically lifted off the plate and added to 4x volume of culture media, to inactivate the Trypsin, in a

15ml centrifuge tube (Thermo Fisher; #05-539-12). Cells were pelleted at 1000rpm for 5 minutes. Supernatant was aspirated and cells resuspended in culture media and replated onto appropriate tissue culture vessels.

2.2.2 Differentiation of SH-SY5Y cultures

Cultured, undifferentiated SH-SY5Y cells were passaged at a density of 12 x 103 cells/cm2 and incubated at 37°C and allowed 48 hours to adhere. After 48 hours, differentiation was

64

initiated by changing media to SH-SY5Y differentiation media consisting of 1:1 DMEM

Ham’s/F12 (Thermo Fisher; #21331-020) and Neurobasal (Thermo Fisher; #21331-020) supplemented with 0.5% Glutamax and 0.5% N2 supplement (Table 2.7) and 10 µM Retinoic

Acid (Sigma; #R2625) for 14 days (Figure 2.4).

Day 0 differentiation: Cells seeded at 12 x 103 Day 15 cells/cm2 TRAP RNA isolation

DMEM DMEM:Ham’s F12 1:1

10% FBS N2 supplement

10µM Retinoic Acid

Figure 2.4 - Schematic outlining the SH-SY5Y differentiation protocol towards an mDA phenotype. Culture media was replaced with differentiation media supplemented with RA for 14 days prior to TRAP RNA isolation

Component Manufacturer Catalogue No. Stock Final Conc. In 10ml Insulin Sigma I-1882 25mg/ml 2.5mg/ml

Apo-transferrin Sigma T-1147 100mg/ml 10mg/ml

Bovine serum albumin Gibco 15260-037 75mg/ml 7.5mg/ml

Progesterone Sigma P-8783 0.6mg/ml 0.00198mg/ml

Putrescine Sigma P-5780 160mg/ml 1.6mg/ml

Na Selenite Sigma S-5261 3mM 1:100

DMEM:F12 Gibco 21331-020 6.875ml

Table 2.7 – Components used to make N2 supplement for differentiation of SH-SY5Y cells

2.2.3 Culturing of iPSCs

Undifferentiated iPSC lines derived from reprogrammed dermal fibroblasts of a member of the Iowa kindred carrying a triplication in the SNCA locus (AST23) and an unaffected member without the mutation as a control (NAS2) were kindly donated by Dr. Tilo Kunath (Edinburgh

University). iPSCs were cultured on vitronectin (Life Technologies; #A14700) coated tissue culture vessels in Essential 8 (E8) and E8 supplement (Gibco; #A1517001) culture media and

65

incubated at 37°C/5% CO2. Cells were allowed to proliferate until they reached approximately

80% confluency with culture media changed every day.

Cells were allowed to reach 80% confluency and then passaged by removing of culture media, followed by washing with PBS and adding UltraPure 0.5 mM EDTA pH 8.0 (Sigma, #15575-

020) in PBS until colonies began to lift off (2-3 minutes). At this stage, the EDTA solution was aspirated and E8 media with E8 supplement was replaced and cells mechanically lifted off the well by mechanical action with a P1000 pipette.

2.2.4 Differentiation of iPSCs towards a midbrain dopaminergic fate

To initiate differentiation, 3 wells of a 6 well plate of iPSC cultures, at 80% confluency, were dissociated from the culture plate using StemPro Accutase (Gibco; #A11105-01). After cells had begun to lift off the surface of the well, Accutase solution was aspirated and replaced with differentiation media. iPSC colonies were mechanically resuspended in neural ventral initiation media (NVIM) consisting of a 1:1 ratio of DMEM Ham’s F12 /Neurobasal supplemented with 1:100 glutamax, 1:50 N2 (Gibco; #17502-048) and 1:50 B27 without

Vitamin A) (Gibco; #12587-010) and pooled together in non-adherent culture conditions to start embryoid body (EB) formation (Figure 2.5). Along with this base media, factors which direct differentiation towards a neural ventral midbrain (Table 2.8), as well as Y-27632 (10µM) for the first media change to increase cell viability, was added to NVIM.

66

Figure 2.5 - Schematic outlining the iPSC differentiation protocol towards an mDA phenotype. NB – Neurobasal; SHH- Sonic Hedgehog; BDNF – Brain-derived neurotrophic factor; GDNF – Glial-derived neurotrophic factor; cAMP – cyclic adenosine monophosphate.

Manufacturer NVIM supplemented Involvement in Concentration and catalogue Pathways factors differentiation number Selleckchem; inhibition of TGF-β Dual inhibition for SB431542 10µM S1067 signalling improved Stemgent; 04- inhibition of BMP4 neuroectoderm LDN-193189 100nM 0074-10 signalling induction Floor plate Recombinant Sonic R&D Systems; specification and 200ng/ml Hedgehog (Shh) cat. no.1845-SH dorso-ventral patterning Axon Medchem; GSK3β inhibitor activates rostro-caudal CHIR99021 0.8mM 1386 WNT/β-catenin signalling patterning Axon Medchem; Prevents dissociation Y-27632 10µM 1683 induced apoptosis. NVMM Manufacturer Involvement in supplemented Concentration and catalogue Pathways differentiation factors number Peprotech; 450- Neurotrophic growth BDNF 20ng/ml Regulation of growth, 02 factor maintenance and Peprotech; 450- Neurotrophic growth GDNF 10ng/ml survival of neurons 10 factor Removes damaging ascorbic Acid (AA; Sigma, cat. no. 0.2mM Anti-oxidant reactive oxygen Vitamin C) A5960 species Sigma cat no. activates cAMP dependent db-cAMP 500µM induce mature D0627 protein kinase (PKA) neuronal γ-secretase/ Notch DAPT 1µM Sigma; D5942 morphologies signalling inhibitor Table 2.8 – Factors added to NVIM for specification of midbrain dopaminergic fate in iPSCs and factors added to NVMM to maintain neurons

After four days of EB formation, EBs were plated on 6 well plates coated in Poly-L-Ornithine

(PLO)(Sigma #P3655), Laminin (10µg/ml; Invitrogen; #23017-015) and Fibronectin (10µg/ml;

Invitrogen; #33010-018) (PLO-L-F) in NVIM with half B27 and N2. On day 9 of differentiation,

67

factors were stopped. Two days later, EBs had grown large monolayers of outgrowing cells, and these were dissociated thoroughly using accutase and replated in a new well of a 6 well plate coated with PLO-L-F with NVIM changed to Neurobasal supplemented with B27 (-Vit. A)

(neural ventral maintenance media; NVMM). NVMM was supplemented with factors to maintain neuronal cultures for the rest of the duration of differentiation (Figure 2.5).

2.2.5 Immunoflourescence staining in SH-SY5Y and iPSC cells

For determining optimal promoter for stable long-term expression of transgenes in SH-SY5Y and iPSC-derived cultures, cells on coverslips had culture media removed and were fixed with

4% Paraformaldehyde (PFA) for 15 minutes (Day 14 for SH-SY5Y cells and Day 30/35 for iPSC- derived cells). Fixed cells on coverslips were then washed with PBS twice. For immunofluorescent staining, coverslips in 4 well plates were blocked in PBS with 0.1% Triton-

X (Sigma; X100; PBS-TX) with 10% Normal Donkey Serum (NDS) or Normal Goat Serum and

2% BSA for 1 hour. Primary antibodies (Table 2.9) were added to antibody diluent consisting of PBTX with 1% NDS and 0.2% BSA and incubated on a shaker overnight at 4°C. After overnight incubation, cells were washed with PBS-TX 3 times and incubated in secondary antibody (Table 2.9) in antibody diluent for 1 hour at RT. Cells were then washed 3 times with

PBS-TX and then mounted onto slides with vectashield mounting medium with 4’, 6-

Diamidino-Phenylindole Dihydrochloride (DAPI; Vector laboratories; #H-1200) which stains all cell nuclei. Immunoflourescent images were taken using a Leica microscope (LEITZ

DMRD/DMRB) equipped with a Leica Digital FireWire Camera System for Fluorescence

Microscopy (DFC-340) and Leica application suite (V3.3.0). Cell counting from 3 fields of view from 3 coverslips was performed using imageJ cell counting module. Thus, for analysing the

68

differentiation capacity of iPSC-derived neurons, cells were stained for TH and TUJ1 and DAPI to analyse percentage of dopaminergic neurons in each culture on day 35 of differentiation.

Primary antibody Host Cat. No. Manufacturer anti-FOXA2 (HNF-3β) Mouse sc-101060 Santa Cruz anti-dopaminergic marker Rabbit AB152 Millipore Tyrosine Hyrdroxylase (TH) anti-Neuron-specific class III Mouse MS435P Covance; beta-tubulin (TUJ1) anti-Enhanced Green Chicken ab13970 Abcam Flourescent Protein (EGFP) anti-6His tag Chicken PA1-9531 ThermoFisher Secondary antibody Host Cat. No. Manufacturer Green 488 - anti-mouse Goat A11029 Invitrogen Green 488 – anti-chicken Goat A11039 Invitrogen Red 568 – anti-mouse Goat A21124 Invitrogen Red 568 – anti-rabbit Donkey A10042 Invitrogen Far-red 660 - anti-mouse Goat A21054 Invitrogen Table 2.9 – List of antibodies used for immunofluorescence analysis of cultured SH-SY5Y and iPSC-derived cultures

2.2.6 Western Blot for characterisation of -synuclein expression in cellular models

Differentiating iPSC-derived PD models were collected on day 30 of differentiation and lysed with 100µl RIPA buffer (Table 2.10) by vortexing (Corning LSE; #10-320-807) and incubated on ice for 30mins. The suspension was then centrifuged at 12000rpm for 20mins at 4°C, on a table-top centrifuge (Eppendorf; #5424R) and supernatant collected in fresh microcentrifuge tube on ice. Protein quantification was determined by Pierce BCA protein assay kit

(ThermoFisher; #23225), and equal quantities of sample mixed with equal parts 2x loading buffer (Table 9) by vortexing and then heated on a heating block (Techne; #DB2D) at 100°C for 5 minutes. Protein samples were loaded and resolved on a polyacrylamide gel (Table 2.10) alongside a protein molecular weight ladder (Geneflow; #S6-0024), using the mini-PROTEAN cell (Bio-rad; 1#658000EDU). The tank was filled with running buffer (Table 2.11) and then sealed and connected to a power supply (Bio-Rad Powerpack 200; #1655052) and run until loading dye had migrated to the bottom of the gel.

69

Electrotransfer of resolved proteins on SDS-PAGE gels onto polyvinylidene difluoride (PVDF;

Roche; #12886700) membrane was achieved using the 2-gel tetra Mini-transblot cell and blotting module (Bio-Rad; #1660827EDU). The tank filled with transfer buffer (table 11) and run at 200A for 90 minutes. Membrane was then blocked in 5% w/v milk in PBS-T (Table 2.11).

After blocking, the membrane was placed in 1% w/v milk in PBS-T (antibody diluent) with primary antibody, and incubated overnight at 4°C. Membrane was then washed 3 times with

PBS-T and film was incubated in antibody diluent with secondary antibody for 1 hour at room temperature (RT). Membrane was washed again 3 times in PBS-T and treated with Pierce

Enhanced Chemiluminescent (ECL) Western Blotting Substrate (ThermoFisher; #32209) for detection of HRP, following manufacturer’s instructions. Scientific imaging film (Amersham hyperfilm; #GE 28906837) was exposed to the membrane and film was developed using a medical film processor (Konica Minolta; #SRX-101A). The membrane was then stripped in restore stripping buffer (Thermo Fisher; 21059), washed 3 times in PBS-T and rebrobed with housekeeping proteins (-actin, -tubulin) and reimaged. Analysis of gel and normalisation was determined using the ImageJ gel module by relative density.

RIPA buffer Manufacturer 2x Loading buffer Manufacturer

reagents and cat. No. reagents and cat. No.

IGEPAL CA-630 Sigma; I3021 1% Tris-HCl pH 6.8 100mM

sodium Sigma; D6750 0.5% SDS Sigma; L37711 4% deoxycholate Bromophenol SDS Sigma; L37711 0.1% Sigma; B-806 0.2% Blue Thermo Fisher; PBS to 10ml Glycerol Sigma; G5516 10% 14190-144

Protease 1 tablet for Roche; DTT (added just Melford; inhibitor cocktail 10 ml RIPA 200mM 11836170001 before use) MB1015 tablet (added fresh) Table 2.10 – Table listing components of RIPA buffer for lysis of cells and 2 x loading buffer for mixing with protein samples for loading onto SDS-PAGE gel

70

Volume 10% Resolving gel reagents Stacking gel reagent Volume (ml) Manufacturer (ml) Double distilled Double distilled H O 4 2.7 2 H2O 37.5:1 37.5:1 Acrylamide/Bis- 3.3 Acrylamide/Bis- 0.67 Geneflow (EC-890) acrylamide acrylamide 1M Tris in ddH2O, 1.5M Tris in ddH2O, pH 8.8 2.5 0.5 pH 6.8 10% SDS 0.1 10% SDS 0.04 Sigma (L3771) Sigma- 10% ammonium persulphate 0.1 10% APS 0.04 Aldrich A3678 APS

N, N, N’, N’- tertramethylethylenediamine 0.004 TEMED 0.004 Sigma (T9281) TEMED Total Volume 10ml Total Volume 4ml Table 2.11 – Composition of resolving and stacking gels for SDS-PAGE

1x running 1x transfer Blocking Manufacturer buffer buffer buffer Dried milk Marvel Tris-base 25mM Tris-base 48mM Sigma; T1503 5% w/v (skimmed) original Sigma; Glycine 250mM Glycine 39mM Sigma; G0709 PBS tablets 5 tablets 18912014 Sigma; SDS 0.1% SDS 0.037% Sigma; L37711 Tween-20 0.1% P5927 Methanol 20% Sigma; 32213 in 1L ddH2O in ddH2O To 1L in ddH2O To 1L Table 2.12 – Composition of running buffer for resolution of SDS-PAGE proteins, transfer buffer for electrophoretic transfer of proteins onto PVDF membranes and blocking buffer for immunoblotting

For -synuclein protein expression Manufacturer Concentration Mouse anti-α-synuclein antibody BDBioscience; 610787 1:1000 Loading control antibodies Manufacturer Concentration β-actin-HRP conjugated antibody abcam; ab20272 1:5000 Mouse anti--Tubulin Sigma; 1:1000 Secondary antibodies Manufacturer Concentration anti-mouse Horseradish peroxidase conjugate HRP; Promega; W4021 1:7000 Table 2.13 – antibodies used to probe protein expression with Western blotting

2.2.7 RNA extraction and cDNA generation RNA was extracted using the mirVana RNA extraction kit (Invitrogen; #AM1560) as per manufacturer’s instructions. RNA concentration was measured and 2μg of cDNA was generated with the SuperScriptIII reverse transcriptase cDNA kit (Invitrogen; 4376357) using random hexamer priming. cDNA reverse transcribed. The cDNA generation protocol is briefly outlined in table 2.14 and quantitative PCR (qPCR) reaction and thermal cycling parameters are outlined in table 2.15.

71

Added following Quantity RNA template 2μg Random primers 1μl 10mM dNTP mix 1μl H2O To 13μl Mixture heated to 65˚C for 5 minutes and incubated on ice for 1 minute 5x First Strand Buffer 4μl 0.1 M DTT 1μl RNaseOUT recombinant RNase inhibitor 1μl Superscript III RT enzyme 1μl Enzyme mixture added to primer/template mixture and incubated at 25˚C for 5 minutes Incubated at 50˚C for 60 minutes Reaction inactivated by heating mixture to 70˚C for 15 minutes 1μl (2 RNase H Units) Incubated at 37˚C for 20 minutes Table 2.14 -Protocol for cDNA generation using Superscript III cDNA kit

Component Volume Power SYBR Green PCR Master Mix (2x) 25μl Forward Primer (10μM) 2μl Reverse Primer (10μM) 2μl Template 10ng Water To 25μl qPCR cycling conditions Step Hold Temperature 95˚C Step PCR Cycles x40 Denature Anneal/Extend Temperature 95˚C 60˚C Time 15 seconds 1 minute Table 2.15 -SYBR green qPCR reaction and thermal cycling conditions

2.3 Statistical analysis All statistical analyses were performed using Prism software. Where appropriate, one-way

ANOVA with the suitable post-hoc test was utilised. A p-value of below 0.05 was considered statistically significant and all probabilities were detail. Further statistical details will be described for the relevant data where it is presented.

72

Chapter 3 – Cellular models of Parkinson’s disease 3.1 Introduction

Parkinson’s disease (PD) is a neurodegenerative disease that currently has no cure. PD is a clinically and etiologically heterogeneous disease that eventually leads to degeneration of the dopaminergic neurons in the SNpc. The histopathological marker common to most PD cases, whether idiopathic or of known genetic cause, is the appearance of α-synuclein aggregates.

Familial PD is thought to account for only 15% of all PD cases and is associated with a number of genetic loci, including SNCA, the gene that encodes α-synuclein. Perturbations in α- synuclein expression have been shown to disrupt several cellular pathways but its physiological function is, as of yet, not fully elucidated. Multiplications and mutations of the

SNCA gene locus have been associated with a younger age of onset and more aggressive pathology compared to idiopathic PD.

Given the relative importance of α-synuclein in PD etiology, many models of α-synuclein have been explored, ranging from in vitro cellular models to in vivo rodent models. In this dissertation we chose to use a well-established cellular model for studying PD-associated neurodegeneration: the SH-SY5Y neuroblastoma cell line (described in section 1.6.3). This cell line was originally isolated as a subclone from a neuroblastoma bone marrow biopsy with dopamine-β-hydroxylase activity (Biedler et al. 1973; Biedler et al. 1978).

The SH-SY5Y cell line is a popular cell model for neurodegenerative disease research because it possesses many characteristics of dopaminergic neurons. These cells also have the advantage of being proliferative in their undifferentiated state, are relatively homogeneous, easy to culture and manipulate. Having a homogenous cell line has the advantage of providing more consistent and reproducible results. Studies have shown that the SH-SY5Y cells express

73

the dopamine transporter and dopamine receptors with differentiation inducing an increase in levels of dopaminergic markers DAT and TH as well as neuronal markers such as MAP2

(Presgraves et al. 2004; McMillan et al. 2007; Lopes et al. 2010; Constantinescu et al. 2007).

As in substantia nigra neurons, this cell line is sensitive to compounds which induce parkinsonism (Chen et al. 2005; Wei et al. 2013). SH-SY5Y cells can be differentiated into a functionally mature neuronal phenotype, with increased neuritic outgrowth, in the presence of Retinoic acid (RA). PD modelling in this cell line can thus allow for the study of disease- associated genes in a model which preserves many neuronal or dopaminergic pathways.

Indeed, studies have previously explored phenotypes relating to wild-type and mutant α- synuclein overexpression in this cell line (Hasegawa et al. 2004; Desplats et al. 2009; Ding et al. 2013; Melo et al. 2017).

Given the usefulness of the SH-SY5Y model, we developed overexpressing wildtype (WT) and

A53T mutant α-synuclein SH-SY5Y cultures by lentiviral transduction. These models can be used in further experiments to determine differential gene expression as a result of a- synuclein overexpression (described in Chapter 4). In order to facilitate gene expression analyses using the TRAP technique, the cells would need to express the L10a ribosomal protein (RBP) fused to EGFP protein; this was also achieved by lentiviral vector transduction.

This will enable the actively translated mRNAs at the time of harvest to be affinity purified with anti-EGFP antibodies.

Although the SH-SY5Y cell line has many features which make it a useful PD model, it has some limitations. The cell line is not strictly dopaminergic and studies have shown that RA differentiation generates phenotypes with dopaminergic, cholinergic and adrenergic characteristics (Adem et al. 1987; Xicoy et al. 2017). Examples of other more pertinent models

74

are human dopaminergic neurons that can be derived from hESCs and hiPSCs (described in section 1.6.4 & 1.6.4.1). Furthermore, stem cell-derived neurons have previously been shown to demonstrate pathological phenotypes when α-synuclein is overexpressed via viral vectors

(Schneider et al. 2007; Zasso et al. 2018). Thus, within the scope of this dissertation was the development of an iPSC-derived dopaminergic neurons as an alternative PD model.

Since the advent of iPSC technology, protocols have been developed to differentiate pluripotent stem cells towards defined cellular phenotypes. Thus, iPSC-derived neurons offer an unlimited source of cells in which to study disease mechanisms in vitro, due to their pluripotency and ability to proliferate indefinitely (1.6.4). Recently, two labs have developed a robust “floorplate” method to differentiate iPSCs towards midbrain dopaminergic neurons, the neurons that degenerate (Kirkeby et al. 2012; Kriks et al. 2011), leading to motor symptoms associated with PD (described in section 1.6.4.1). These iPSC-derived neurons offer the advantage of being more accessible than primary post-mitotic neurons and can be generated from human somatic cells.

iPSC-derived midbrain dopaminergic (mDA) neurons can be obtained from patient specific iPSCs and these can be compared to age and sex matched controls or non-affected family members. However, results from these comparisons can be confounded by normal genetic variability. To minimise genetic variability, viral transduction of disease-associated genes can be used to produce relevant models with the original, non-transduced line used as a control.

It has been previously reported that the exogeneous expression of PD-associated α-synuclein mutants can perturb normal neurite dynamics (described in detail in section 1.4.6.5).

Moreover, α-synuclein overexpression has been shown to impair differentiation of iPSCs towards a mDA fate, identified by a decrease in number of TH+ neurons (Schneider et al. 2007;

75

Oliveira et al. 2015; Zasso et al. 2018). In this chapter, the iPSC-derived midbrain dopaminergic neurons will be analysed for expression of TH and TUJ1, as a proxy for differentiation capacity, as well as neurite dynamics including length and number of primary neurites and branching neurites, in response to perturbed α-synuclein expression.

3.1.1 – Aims The aim of the experiments described in this chapter relate to the generation of SH-SY5Y and iPSC-derived neurons that overexpress wild-type and mutant α-synuclein, specifically the aims of this chapter are to:

1) generate lentiviral constructs expressing α-synuclein and L10a-EGFP from different

promoters;

2) assess the efficiency of transgene expression from the various lentiviral vectors in

SH-SY5Y and iPSC-derived neurons; and

3) characterise the virally transduced cell lines, including levels of α-synuclein protein,

and for the iPSC-derived neurons, differentiation capacity and neurite dynamics.

These cell lines were used for further evaluation of α-synuclein-associated perturbations in later chapters (Chapter 4 and 5).

76

3.2 Materials and Methods

3.2.1 Promoter testing for stable long-term expression of transgenes in SH-SY5Y cells and iPSCs

For SH-SY5Y cultures, undifferentiated SH-SY5Y cells were cultured (section 2.2.1) and were seeded at a density of 2 x 105 cells onto culture plates and separately transduced with lentiviral particles to express Enhanced Green Flourescent Protein (EGFP) under the control of different promoters (CMV, CAG, EF1α, PGK) at a multiplicity of infection (MOI) of 2 and 5

(generation of lentiviral particles is described in section 2.1). SH-SY5Y cells were then dissociated and pelleted and fixed by resuspending in 1ml of 4% paraformaldehyde solution

(Sigma; #P6148) in PBS for 15 minutes. Cells were then pelleted, washed and resuspended in

1 ml of PBS and transferred to flow cytometry tubes (BDBioscience; #352052) for flow cytometric analysis of expression of EGFP using the FACS Canto II (BD Bioscience)(n=1). An

MOI of 2 was deemed adequate at driving EGFP expression in SH-SY5Y cells sufficiently for subsequent FACS sorting (Appendix 1 Table).

To prepare for FACS sorting, undifferentiated SH-SY5Y cells were transduced at an MOI of 2

(n=1) with lentiviruses with different promoters on T25 culture flasks (Thermo-Fisher;

#156367), were allowed to reach 80% confluency, and then washed, dissociated thoroughly in Accutase and pelleted at 1000rpm for 5 minutes in 15ml tubes. Cells were then washed with PBS, pelleted and resuspended in FACS buffer consisting of 25mM HEPES (Sigma;

#H3375), 2.5mM EDTA (Sigma; #E5134) and 0.5% Bovine Serum Albumin (BSA; Sigma;

#A7906) in PBS and transferred into flow cytometry tubes. Cells were then sorted for EGFP expression using a BD Influx high speed flouresence activated cell sorter (BDBioscience)

(Appendix 2) and collected in 15ml polypropylene falcon tubes with warmed, filter-sterilised

77

FBS and then resuspended in SH-SY5Y culture media supplemented with 1% Penicillin-

Streptomycin (Thermo Fisher; #15070-063). 250,000 cells per sample were collected for each condition and seeded onto a T25 flask.

FACS sorted samples were passaged onto coverslips, in triplicate, at a density of 12 x 103 cells/cm2 in 4 well plates (Thermo Fisher; #176740) in culture media and incubated for 48 hours to adhere. Differentiation was then initiated by changing media to SH-SY5Y differentiation media as described in section 2.2.2, for 14 days. Cells were immunostained for

EGFP and DAPI and cell counts analysed (described in section 2.2.5). iPSCs were cultured and differentiated as described in section 2.2.3 and 2.2.4. On day 20 of differentiation cells were passaged onto coverslips, in triplicate, in 4 well plates at a density of 1.5 x 105 per coverslip and allowed 48 hours to settle. On day 22 of differentiation cells were separately transduced with lentiviral particles which drive expression of EGFP under the control of different promoters (CMV, CAG, EF1α, PGK and hSYN1) at an MOI of 2, and allowed

3 days for expression on a protein level (generation of lentiviral particles is described in section 2.1). Cells were further cultured for 10 days, followed by immunofluorescent analysis of EGFP expression on day 35 (described in section 2.2.5).

3.2.2 Transduction of SH-SY5Y and iPSC-derived neurons cells with α-synuclein

To generate SH-SY5Y synucleinopathy models, we transduced cells with the lentiviral particles to drive expression of either wildtype or mutant α-synuclein, under the control of the CMV promoter. SH-SY5Y cells in culture media were trypsinised and 2 x 105 cells were seeded into a 6 well plate and transduced with lentivirus particles expressing wildtype α-synuclein

(wildtype overexpression; WT1-3), mutant A53T α-synuclein (A53T mutant overexpression;

AT1-AT3) at a MOI of 5, or untransduced (control C1-C3) (n=3 in each group). Once the wells

78

had reached 80% confluency, they were passaged onto 2 x T25 plates, separately for each sample; one for protein isolation and western characterisation and one for further processing with L10a-EGFP transduction for TRAP.

To generate and characterise an iPSC PD model, day 11 differentiating NAS2 iPSCs, were passaged onto 6 well plates at a density of 2.5 x 105 cells/well. These cells were transduced with lentiviral particles that express Wildtype, A53T and G51D mutant α-synuclein, as well as an empty vector control, at an MOI of 5 (n=4 for each group). Western blotting and densitometry analysis was used (section 2.2.6) to analyse expression of α-synuclein. Statistical testing used was one-way ANOVA with Tukey’s post-hoc test.

To analyse the differentiation capacity of iPSC-derived neurons synucleinopathy models, cells were stained for TH and TUJ1 and DAPI to analyse percentage of dopaminergic neurons in each culture on day 30 of differentiation.

3.2.3 qPCR analysis of differentiation markers in differentiating iPSC-derived neurons

RNA was extracted and cDNA generated as per section 2.2.7. qPCR was performed on differentiating iPSC-derived neurons on day 0, 11, 23 and 35 for the genes LMX1A, FOXA2,

MAP2, DDC and ACTB was used as a housekeeping gene (primers listed in table 3.1). qPCR was performed on the StepOnePlus Real Time PCR system (ThermoFisher) and results analysed using the ΔΔCT method.

79

Marker Forward/Reverse Primer sequence LMX1A Forward GCAAAGGGGACTATGAGAAGGA Reverse CGTTTGGGGCGCTTATGGT FOXA2 Forward GGAGCAGCTACTATGCAGAGC Reverse CGTGTTCATGCCGTTCATCC MAP2 Forward CTCAGCACCGCTAACAGAGG Reverse CATTGGCGCTTCGGACAAG DDC Forward TGGGGACCACAACATGCTG Reverse TCAGGGCAGATGAATGCACTG ACTB Forward ACAGAGCCTCGCCTTTGC Reverse CATCACGCCCTGGTGCC Table 3.1 - List of primers used for qPCR expression profiling of differentiation markers in iPSC-derived neurons

3.2.4 Functional analysis of neurite dynamics of iPSC-derived neurons iPSCs PD models were generated as previously described (3.2.2) with 4 biological replicates for each condition. Differentiating iPSC-derived neurons were plated onto 2 coverslips for each biological replicate group. Overall, 5 cells per biological replicate groups (20 for each condition) were analysed for neurite length, number of primary neurites and number of branching points. Neurons were stained for TH, TUJ1 and 6His-tag on day 35 of differentiation and neurite dynamics were analysed using the Neurite tracer plugin for ImageJ (Figure 3.1).

Cells which co-expressed TH, TUJ1 and 6His-tag were analysed and one-way ANOVA was performed to determine statistical significance of the results.

Figure 3.1 – Neurite analysis in iPSC-derived neurons. A) Neurite analysis of hiPSC-derived cultures on day 35 of differentiation. Immunoflourescent merge and constituent panels from top to bottom: 6His-tag (Green) for exogeneous α-synuclein, Tyrosine Hydroxylase (Red) and β-tubulin (Purple) is shown, as well as the 8-bit image used for analysis. B) Neurite tracer module on ImageJ with masks on neurites derived from one soma.

80

3.3 Results

3.3.1 Generating a SH-SY5Y model of PD for translating ribosome affinity purification

3.3.1.1 Optimisation of promoter usage in lentiviral vector constructs

Promoter optimisation experiments were performed to establish the best promoter for long- term, stable expression of transgenes in SH-SY5Y cells. A literature search was conducted to identify a list of suitable promoters for use in transduction of neuronal or specifically SH-SY5Y cells (Table 3.2).

Cells transduced Promoter used Percentage transduction Paper Primary mouse CAG/EF1α 45%/67% (Kumar et al. 2015) cerebellar neurons Differentiated SH- Unquantified but suggests PGK/CMV (Li et al. 2010) SY5Y cells highest efficiency by CMV Differentiated hESCs EF1α 30-48% (Gropp et al. 2003) hiPSC-derived PGK 70-90% (Seibler et al. 2011) dopamine neurons Rat substantia nigra PGK Unquantified (Bianco et al. 2002) neurons hESCs-derived PGK Unquantiied (Grealish et al. 2014) dopamine neurons Table 3.2 - Literature review of studies that have employed lentiviral vectors to express transgenes in neuronal models and percentage transduction where available. CMV, cytomegalovirus; PGK, phosphoglycerate kinase; EF1- α, elongation Factor 1-α; CAG, CMV/chicken-β Actin; hESC, human embryonic stem cells; hiPSC, human induced pluripotent stem cells.

Based on the above literature, lentiviral transfer plasmids were cloned to express enhanced green fluorescent protein (EGFP) using the cytomegalovirus (CMV), phosphoglycerate kinase

(PGK), elongation Factor 1-α (EF1- α) and CMV/chicken-β Actin (CAG) promoters (Figure 3.2A).

After cloning, transfer vector plasmids with different promoters were co-expressed, separately, with packaging, envelope and regulatory plasmids in HEK293T cells to generate lentiviral vector particles (described in section 2.1). Once lentiviral vector particles were isolated and titred by flow cytometry, lentiviral vector transduction efficiency was performed to determine a suitable MOI for efficient transduction (Figure 3.2B). Undifferentiated SH-SY5Y cells were transduced at MOIs of 2 and 5 and flow cytometry was employed to analyse the efficiency of transduction. Flow cytometric analysis showed no EGFP expression in the

81

negative untransduced control and transduction efficiencies ranging between 27.1-58.0% for transductions using an MOI of 2 and between 55.6-84.6% for transductions using an MOI of

5 (Figure 3.2B).

Figure 3.2 – Lentivirus promoter transduction efficiencies. A) Schematic for the 4 viral plasmids used to generate lentiviral particles for testing in SH-SY5Y lines. Each plasmid contains a different promoter (CMV, CAG, EF1α, PGK) to express EGFP. B) Table shows the percentage of cells expressing EGFP, as determined by flow cytometric analysis, for each viral promoter, used at an MOI of 2 and 5 in undifferentiated SH-SY5Y cells (n=1).

The results suggest that there is increased transduction efficiency with increasing MOI.

However, as these experiments will be used to inform later experiments such as overexpressing a-synuclein or L10a ribosomal subunit protein fused to EGFP, the MOI of 2 was deemed sufficient for further analyses of stable long-term expression of EGFP. At an MOI of 2, a moderate level of transduction efficiency would be achieved without drastically

82

perturbing cell function, for example the translation machinery due to elevated levels of exogeneous EGFP protein in the ribosomal compartment.

3.3.1.2 Promoter optimistation in differentiated SH-SY5Y cells

SH-SY5Y cells were transduced with lentiviral vectors harbouring the 4 different promoters used to express EGFP at an MOI of 2 (Figure 3.3A). EGFP expressing cells were sorted by flow cytometry and differentiated by Retinoic Acid (RA) treatment for 14 days (Figure 3.3 andFigure 3.4B) and analysed for EGFP expression by immunofluorescence at the end of the differentiation protocol (Figure 3.4C). Cell counts were performed, and mean percentage expression was determined for each vector at the end of the differentiation protocol (Figure

3.4D).

Figure 3.3 -Differentiation of SH-SY5Y cells. FACS sorted SHSY5Y cells transduced with a lentivirus that drives L10a-EGFP fusion protein under the control of the CMV promoter, at an MOI of 2 on day 0 and day 15 of differentiation with Retinoic acid. Scale bar 100µm.

83

Figure 3.4 –Promoter testing in SH-SY5Y cells. A) Schematic of the protocol for testing promoters for long-term stable transduction in SH-SY5Y cells. Cells were transduced with lentiviral vectors expressing an EGFP reporter at an MOI of 2. Transduced cells were FACS sorted to generate a population of EGFP expressing cells and then cultured and taken through a 14-day differentiation protocol to produce neuron-like cells. Immunoflourescent staining at day 14 was employed to determine the percentage of cells expressing EGFP at the end of the experiment. B) FACS sorting was performed on transduced cultures of SH-SY5Y cells, under regulation by different promoters. C) Representative images from day 14 differentiated SH-SY5Y cells transduced using different promoters. DAPI (blue) stained cells, representing total number of cells, and EGFP expression (green) by different promoters in SH-SY5Y cultures. D) Graph shows average number of cells expressing EGFP (±SEM) from cultures (n = 3) at the end of the 14 day differentiation period. Scale bar = 100μm.

84

Analysis shows that the highest percentage of cells (73.2 ±3.0%) expressing EGFP at the end of the 14-day SH-SY5Y differentiation protocol was achieved with the CMV-EGFP construct as compared to CAG-EGFP (59.9 ± 6.6%), EF1α-EGFP (65.0 ± 3.0%) and PGK-EGFP (44.4 ± 7.5%) promoter constructs (Figure 3.4C, D). This suggests that the CMV-EGFP construct was best able to maintain long-term stable expression of transgenes in SH-SY5Y cells. Although cells were FACS sorted prior to differentiation, the percentage of cells expressing EGFP at the end of the differentiation period was observed to decline in all promoter conditions. Some of this loss may be from silencing of transgene expression during differentiation and technical error such as cell aggregates being harder to visualise individually for EGFP due to its presence in the cytoplasm. Nonetheless, from these results the CMV promoter was significantly more efficient for long term expression of transgenes in SH-SY5Y cells and was subsequently used to generate the α-synuclein and the L10a-EGFP fusion protein constructs.

3.3.1.3 Expression of WT and A53T mutant α-synuclein in SH-SY5Y cells

Having established that the CMV promoter was most efficient for expressing transgenes in our SH-SY5Y cells, independent cultures were transduced with lentiviral particles expressing

WT and A53T mutant α-synuclein at an MOI of 5. To validate α-synuclein protein production in these cells, Western blotting was performed. Protein analysis showed expression of α- synuclein in transduced SH-SY5Y cells. No endogenous α-synuclein expression was detected, which is in keeping with literature suggesting low expression rate in the SH-SY5Y cell line

(Figure 3.5A)(Kumar et al. 2017). Densitometry analysis showed that WT (41.8±0.91) and A53T

(42.3±0.22) α-synuclein protein expression was comparable and both were significantly higher than to control untransduced cultures (p<0.001, one-way anova with Tukey post-hoc test) (Figure 3.5B).

85

Figure 3.5 – α-synuclein protein expression analysis in SH-SY5Y cells. A) Western blot of α-synuclein and α-tubulin from undifferentiated SH-SY5Y cultures. C1-C3 represent untransduced cells, while and WT 1-WT 3 and A53T 1 – A53T 3 represent cells transduced with lentiviral particles expressing WT and mutant A53T α-synuclein respectively. B) ImageJ densitometry analysis of Western blot showed that there was no significant difference between mean WT and A53T α- synuclein expression (±SD) and both WT and A53T α-synuclein expression were significantly higher than to control untransduced cultures. One-way ANOVA, Tukey post-hoc test (n=3). *** p<0.001

3.3.1.4 Production of synuclein SH-SY5Y cells expressing L10a-EGFP

Having established that the WT and A53T synuclein expressing SH-SY5Y cells were producing similar and robust levels of synuclein, they were transduced with lentiviral vectors to express

L10a-EGFP fusion protein, under the regulation of the CMV promoter at an MOI of 2. 3 days after transduction, cells were harvested in FACS sorting buffer and 200,000 cells were isolated based on expression of EGFP from each individual sample (Figure 3.6). Gating was kept constant such that these cells would produce similar expression levels of L10a-EGFP fusion protein. These lines were then ready to be used for the translating ribosome affinity purification protocol and next generation sequencing experiments (described in chapter 4).

86

Figure 3.6 – Representative FACS tracing showing gating for the isolation of EGFP expressing cells from experimental samples (one per condition Control 1, WT1, A53T1). The gating for EGFP expression was kept constant between the different samples.

87

3.3.2 Generating an iPSC model of PD

3.3.2.1 Preliminary differentiation of iPSCs towards a midbrain dopaminergic cell fate

The differentiation protocol established by Kirkeby et al. was used to obtain dopaminergic neurons from NAS2 iPSCs. Cells were monitored by microscopy throughout the specification of iPSC towards mDA neurons (Figure 3.7A). Cells were passaged onto coverslips, at day 22, in technical triplicate. At day 30 and day 45 of differentiation, cultures were stained for the floorplate marker FOXA2 and dopaminergic marker TH (Figure 3.7B). Cell counts were undertaken, and analysis showed 76.9±6.2% and 87.4±4.3% of cells were seen to express

FOXA2 and 17.6±4.40% and 28.7±5.0% to express TH on day 30 and 45 respectively (Figure

3.7C). A t-test was employed and showed no significant differences between day 30 and 45 for both differentiation markers (p>0.05). It was noted that at high densities, these cells had a tendency to form clusters and thus cell counts were difficult to undertake. However, most cells were clearly seen to express FOXA2 and cells with neurite-like processes stained positively for the mature dopaminergic marker TH.

88

Figure 3.7 - Differentiation of NAS2 iPSCs into mDA neurons. A) Brightfield images monitoring the differentiation process through specification stages (day 0-14). Scale bars from left to right: 400μm, 200 μm, 200μm and 100μm. B) Immunoflourescent analysis of floorplate marker FOXA2 (green), mature DA neuronal maker TH (Red) and nuclear marker DAPI (Blue). Scale bar = 100μm. C) Mean cell counts (±SD) showed a high proportion of cells coexpressing FOXA2 and TH, indicative of successful differentiation protocol (n=3).

3.3.2.2 Lentiviral vector transduction of iPSC-derived dopaminergic neurons

To test promoters for stable expression of transgenes in iPSC-derived cultures, NAS2 iPSCs were transduced at an MOI of 2, on day 22 of differentiation and analysed for EGFP expression on day 35. The promoters that were tested included CAG, CMV, EF1α and PGK promoters as well as the human neuron specific synapsin1 promoter (hSYN1). The hSYN1 promoter was included in these experiments to ensure that transgene expression was confined to neuronal

89

populations within the heterogenous cell population in an iPSC culture. Cell cultures were stained for EGFP, TUJ1 (neuronal marker), TH (dopaminergic marker) and DAPI (nuclear staining) (Figure 3.8A). This allowed the analyses of EGFP expression within neuronal and non- neuronal cells. The percentage of neuronal and non-neuronal cells that expressed EGFP were assessed separately to ascertain whether EGFP was preferentially expressed in certain populations, due to either preferential transduction or promoter activation. The percentage of TUJ1+ cells expressing GFP ranged from 69.1±10.5% for the CMV-GFP lentiviral construct to 48.2±5.3% for the PGK-GFP construct, with no significant differences in expression between the constructs. As expected, viral particles with the hSYN1 promoter showed expression of GFP in TUJ1 expressing cells on day 10 (58.2%±3.1) whilst only a small fraction of TUJ1 negative cells expressed EGFP under the control of the hSYN1 promoter on day 10

(8.6%±1.4), which was statistically significantly lower than all other promoters (p<0.05)

(Figure 3.8B). Thus, for transducing differentiating iPSCs, the hSYN1 promoter will be used as it offers the study a means of controlling and restricting genetic expression to within the neuronal population.

90

Figure 3.8 –Differentiated iPSC neurons expressing EGFP from lentiviral vectors containing different promoters. A) Representative immunofluorescent staining of differentiating iPSC cultures transduced with various lentiviral constructs expressing EGFP 13 days post transduction. DAPI nuclear stain is shown in blue, EGFP staining in green, TH staining in red and TUJ1 staining in purple. Scale bar = 100μm. B) Left: Mean percentage of cells co-expressing TUJ1+ and EGFP concordantly (±SD) and Right: percentage of TUJ1- cells expressing EGFP 10 days post-transduction. One-way ANOVA, Tukey post-hoc test (n=3). *p<0.05, ** p<0.01, *** p<0.001

91

3.3.2.3 Transduction of WT, A53T and G51D α-synuclein into differentiating iPSCs and characterising differentiation for miRNA profiling

Although control and PD (α-synuclein triplication) iPSC lines are available, to control for variation on a genetic and technical level, it was decided to generate an alternative iPSC PD model using lentiviral vectors to overexpress WT as well as mutant A53T and G51D synuclein

(section 3.2.2). Thus, lentiviral particles were generated to express WT and mutant α- synuclein under the control of the hSYN1 promoter (lentivirus generation is described in section 2.1). Generation of lentiviral vectors to express wild-type and mutant α-synuclein involved cloning both hSYN1 promoters and synucleins into lentiviral transfer vectors and co- expressing along with packaging, envelope and regulatory lentiviral plasmids in HEK293T cells, to generate lentiviral particles. Isolated lentiviral vector particles were titred by a qPCR-based method.

Cells were transduced at day 11, upon completion of the specification stage of the protocol and entry into terminal differentiation, with lentiviral particles expressing α-synuclein under the control of the hSYN1 promoter, at an MOI of 5 (n=4). Cultures were passaged onto plates on day 25 for protein isolation and onto coverslips for immunofluorescent analysis. Cells were harvested on day 30, 19 days post transduction, and Western blotting was used to analyse protein expression in transduced and untransduced lines (Figure 3.9A). Western blot analysis showed elevated relative expression of α-synuclein in WT (41.4±17.1), A53T (34.0±11.9) and

G51D (47.1±10.1) cultures compared to control cultures. There was significantly higher expression of synuclein in WT and G51D-transduced cultures compared to control culture

(p<0.0017 and p<0.0006 respectively, one-way ANOVA Tukey post-hoc test) (Figure 3.9B).

There was higher in synuclein expression from the A53T construct, but this was not significant

92

(p<0.067), possibility due to large variation; however, there were no significant differences in

α-synuclein expression between the 3 α-synuclein constructs.

In parallel, a non-transduced control was differentiated towards an mDA fate and cells harvested on days 0, 11, 23 and 35 for qPCR analysis of floorplate (LMX1A, FOXA2), dopaminergic (Dopa-decarboxylase (DDC) and neuronal (microtubule-associated protein 2

(MAP2)) markers (Figure 3.9C). All markers were seen to increase from day 0 to day 23 and continued to increase to day 35, except for FOXA2 which showed a small decrease in expression from day 23 to day 35.

Immunofluorescent staining was used to probe for expression of TUJ1 and TH in all cultures, independently at day 30 of differentiation (Figure 3.9D). From experience, further culturing would result in dense clustering of cells due to neuronal migration and proliferation making cell counting difficult. To identify any perturbations to the differentiation process owing to expression of transgenes, cells from individual samples were passaged onto coverslips on day

25 of differentiation, and cell counts were performed on day 30. Figure 3.9E shows average percentage expression of TUJ1 and percentage of TUJ1 co-expressing the dopaminergic marker TH. The percentage of TUJ1 expressing cells ranged from 26.7% to 31.8% whilst the percentage of these cells which expressed TH was 33.6% to 40.7% and was not significantly changed between conditions, as determined by one-way Anova (p<0.05) with Tukey’s multiple comparisons test.

93

Figure 3.9 - Characterisation of day 30 differentiating iPSC-models. A) Western blot of synuclein expression in day 30 iPSC differentiated cultures overexpressing WT (W1-4), A53T (A1-4), and G51D (G1-4) α-synuclein, as well as an empty controls (C1-4), at an MOI of 5 (n=4). B) Densitometry analysis of mean expression of the 3 α-synuclein constructs (±SD). One-way ANOVA, Tukey post-hoc test (n=4). *p<0.05, ** p<0.01, *** p<0.001. C) qPCR to determine differentiation potential of iPSCs towards a midbrain dopaminergic fate (LMX1A, FOXA2, MAP2, DDC) (n=1). D) Immunofluorescent staining of experimental cultures at day 30 of differentiation for DAPI, TUJ1 and TH (n=4; scale bar = 100μm). E) Differentiation characterisation by cell counts to identify average percentage (±SD) of cells expressing TUJ1 and, of these, percentage also expressing TH (n=4). No statistically significant differences (p>0.05) as determined by One-way ANOVA, Tukey post-hoc test (n=4).

94

As approximately only 30% of cells stained for a neuronal marker, this begs the question as to what these other cells may be, and it was thought that they may be of a glial origin, as these stem from the same progenitor population. Thus, immunofluorescent staining was performed for glial fibrillary acid protein (GFAP), a marker for astrocytes, which have the ability to self-renew and proliferate (Bruttger et al. 2015). Indeed, staining showed widespread GFAP expression in our differentiating iPSC culture (Figure 3.10) which is typical of iPSC-derived mDA differentiated cultures (Kriks et al. 2011; Martí et al. 2017). The cultures seem to recreate the normal CNS environment of a heterogenous population of cells, in a neuronal network which includes both neurons and glial cells. These glial cells allow for the recapitulation of the activation of neuroinflammatory pathways in cultures (Hui et al. 2016).

Figure 3.10 – Immunofluorescent staining of experimental cultures at day 30 of differentiation for DAPI (nuclear marker), TUJ1 (neuronal marker) and GFAP (glial marker) (n=1; scale bar = 100μm).

95

3.3.3 Neurite dynamics in iPSC-derived dopaminergic neurons

It has been reported in literature that perturbed α-synuclein expression can result in dysfunction in neuronal microtubule dynamics. Thus, we thought to analyse neurite dynamics in our iPSC-derived PD models which includes WT as well as A53T and G51D synuclein overexpression. As our exogeneous α-synuclein also contains a 6xhistidine (6xHis) tag, neurite dynamics can be monitored using endpoint immunofluorescent analysis for DAPI (nuclear),

6xHis (exogenous α-synuclein), TUJ1 (neuronal) and TH (dopaminergic). In this way, it was possible to trace neurites from cells which co-stain for all markers, i.e. transduced dopaminergic neurons (Figure 3.11A).

In parallel, a non-transduced control was differentiated towards an mDA fate and cells harvested on days 0, 11, 23 and 35 for qPCR analysis of floorplate (LMX1A, FOXA2), dopaminergic (DDC) and neuronal (MAP2) markers (Figure 3.11B). All markers were seen to increase from day 0 to day 23 and either continued to increase (LMX1A), plateau (MAP2, DDC) or showed a slight decrease from day 23 to day 35.

Cells that were positive for 6xHis, TUJ1 and TH were analysed for neurite length, number of primary and number of branching neurites (Figure 3.11C-F). Here we found that there was no statistically significant increase in the number of primary neurites, i.e. neurites originating from the soma, in cells overexpressing WT (1.5±0.3), A53T (1.4±0.3) and G51D (1.3±0.1) α- synuclein compared to control (1.1±0.1) (Figure 3.11C).

Total primary neurite length per cell was significantly increased in control lines (454.0±55.6) compared to WT (253.4±23.8) and G51D (254.4±41.0) overexpression models (Figure 3.11D).

Although there was a trend to a decrease in A53T α-synuclein expressing cultures

(366.4±58.3), this was not significant. Total neurite length per cell, as a metric, does not take

96

into account how many primary neurites are stem from a single neuron. Thus, it was decided to also look at total neurite length per cell normalised to the number of primary neurites; i.e. mean neurite length (Figure 3.11E). Similar to total neurite length, mean neurite length was also significantly decreased in WT (168.9±8.4) and G51D (193.2±24.2) cultures compared to control (408.6±30.9), with a non-statistically significant decrease noted for A53T (278.9±49.6) cultures (Figure 3.11E).

The total number of branching neurites was also seen to be higher in control (1.8±0.5) compared to G51D (0.2±0.1) overexpression models (Figure 3.11F). The number of branching neurites in control cultures was also noted to be higher than in WT (0.7±0.2) and A53T

(1.2±0.2) cultures compared, this was not statistically significant (Figure 3.11F). The mean number of branching neurites per primary neurite was significantly higher in control (1.6±04) compared to both WT (0.5±0.2) and G51D (0.2±0.1) overexpression cultures. Although control culture mean branching neurites were higher than A53T (0.9±0.2), this was not statistically significant (Figure 3.11G).

In summary, expression of both WT and mutant α-synuclein protein in iPSC-derived neuronal cultures had no effect on the number of average primary neurites nor the total neurite length.

However, upon closer examination of the data, the complexity of neurite branching was affected by overexpression of both WT and G51D mutant synucleins, as the number of branching neurites and total length per cell was lower compared to control cultures. Although there was an observed trend for reduced complexity in both these metrics for the A53T α- synuclein overexpressing cells, this did not reach significance.

97

98

Figure 3.11 – Neurite analysis of iPSC-derived dopaminergic neurons and characterisation of day 35 differentiating iPSC neurons transduced with lentiviral vectors to overexpress WT (W1-4), A53T (A1-4), and G51D (G1-4) α-synuclein, at an MOI of 5 (n=4). A) Immunofluorescent staining of experimental cultures at day 35 of differentiation for DAPI, 6xHis, TUJ1 and TH (n=4; scale bar = 100μm). B) qPCR to determine differentiation potential of iPSCs towards a midbrain dopaminergic fate using primers for LMX1A, FOXA2, MAP2 and DDC (n=1). C-G) Graphs and tables showing average (±SD) (n=4) results for: C) the average number of primary neurites, D) total neurite length, E) average neurite length, F) total branching neurites and G) average number of branching neurites per primary neurite. One-way ANOVA, Tukey post-hoc test (n=4). *p<0.05, ** p<0.01, *** p<0.001. 3.4 Discussion

The aim of this study was to develop a cellular PD model for use in translating ribosome affinity purification (TRAP) for the investigation of the pathogenesis of PD associated with α- synuclein perturbations. Lentiviral particles were generated and used to express transgenes, specifically α-synuclein in SH-SY5Y and iPSC-derived neurons. In generating the lentiviral constructs, promoter usage was considered and some promoters that have been used in literature to express transgenes in neuronal/neuron-like cells include the CAG promoter,

EF1α, PGK and hSYN1 (Kumar et al. 2017)(Table 3.2). Here we show that SH-SY5Y cells were able to express EGFP from the variety of promoters used, and that the CMV promoter was the most robust at producing EGFP expression throughout the culture and differentiation of

SH-SY5Y cells.

One study specifically explored the ability of different promoters to induce transgene expression in SH-SY5Y cells using the pRRL.SIN lentiviral backbone with CMV, PGK and FCIV promoters to express EGFP in undifferentiated and 7 days differentiated SH-SY5Y cells (Li et al. 2010). It was noted that the CMV promoter was best able to produce robust EGFP expression. Our studies thus support this observation.

To further these studies, lentiviral particles were generated to stably express exogeneous WT or A53T mutant α-synuclein, under the control of the CMV promoter. SH-SY5Y cultures were transduced, at a MOI of 5, and overexpression characterised by protein and transcript analysis. The SH-SY5Y α-synuclein overexpression models showed a significant increase in

99

levels of α-synuclein protein compared to untransduced controls. The MOI of 5 was chosen as the initial test using CMV-EGFP lentiviral particles at an MOI of 5 produced ~80% transduction efficiency and lentiviral MOI ranges between 2-30 have previously used in studies using iPSC-derived neuronal cultures and neuronal-like cells, such as SH-SY5Y cells

(Desplats et al. 2009; Prots et al. 2013; Mazzulli et al. 2016; Vogiatzi et al. 2008; Lee et al.

2017). These cells were subsequently transduced with the L10a-EGFP fusion protein construct. Enrichment of these cells via FACS produced a population of SH-SY5Y cells that stably expressed the fusion protein and WT or mutant synuclein, and this provided a source of material for subsequent TRAP experiments as described in Chapter 4.

The use of iPSC-derived mDA neurons overexpressing synuclein was also explored. Lentiviral promoters that have been used to express transgenes in undifferentiated and differentiated stem cells include EF1α (Gropp et al. 2003) and PGK (Seibler et al. 2011)(Table 3.2). iPSC cultures were differentiated with the Kirkeby protocol to derive mDA neurons that were also tested with the lentiviral constructs containing the above 4 promoters with the inclusion of the hSYN1 to provide cell-specific expression in neurons. hSYN1 was chosen as the promoter to express transgenes specifically in differentiating iPSC-derived neurons, though it was not clear if the levels of expression would be comparable to constitutive promoters. This construct was observed to express the reporter gene EGFP at a similar efficiency to the other promoters used, and as expected, non-neuronal cells were transduced at a statistically significant lower level compared to neurons. Based on this data, lentiviral particles were generated to express WT, A53T and G51D α-synuclein, under the control of the hSYN1 promoter. iPSC-derived neurons were transduced separately with these lentiviral vectors, and synuclein overexpression was demonstrated. While it should be noted that synuclein

100

overexpression in the A53T synuclein overexpressing cells was not significantly increased compared to control untransduced cells, this was not significantly different to α-synuclein protein levels in WT and G51D overexpression models.

Differentiation of iPSCs to mDA neurons was characterised by a transcriptomic increase of

LMX1A and FOXA2, two critical floorplate markers, as well as DDC, a dopaminergic marker and MAP2, a neuronal marker. Furthermore, cells were shown to express FOXA2, TH and TUJ1 at a protein level by immunofluorescent analysis. Characterisation of differentiation capacity was further analysed in these iPSC-derived neurons by determining the number of TUJ1 and

TH positive cells, as differentiation potential and maturation capacity have been reported to be perturbed in α-synuclein stem cell models, within 30 days of differentiation (Schneider et al. 2007; Oliveira et al. 2015). However, no statistically significant changes in average number of neurons or dopaminergic neurons were noted. It may be that the differentiated cells were not left for a sufficient amount of time to show reduced maturation capacity in these cultures.

Further studies could be undertaken to determine the effect of perturbed α-synuclein expression on differential expression of dopaminergic neuron differentiation markers, such as nuclear receptor related 1 protein (NURR1) and G-protein-regulated inward-rectifier potassium channel 2 (GIRK-2).

As increased age is one of the main factors in PD susceptibility, another limitation of iPSC- derived cultures is that reprogramming is thought to revert cells to a younger phenotype, based on analysis of age-related phenotypic markers such as telomere length (Lanza et al.

2000). Whilst donor age has been found to be linked to a decrease in reprogramming efficiency (Sardo et al. 2017), iPSCs from older donor samples do not show dysfunction in differentiation capacity and show a reduction in other cellular hallmarks of ageing compared

101

to parental fibroblasts (Lapasset et al. 2011; Miller et al. 2013). Thus, it is difficult to model late-onset neurodegenerative diseases derived from iPSCs in vitro. Some researchers have identified pharmacological agents to reduce telomerase activity in mDA neurons, recapitulating age-associated phenotypes (Vera et al. 2016). Furthermore, the use of proteins associated with premature ageing have been developed to recapitulate cellular age in stem cell-derived models for modelling age-related diseases (Miller et al. 2013; Zhang, Wang, et al.

2005). Another strategy would be to age cells in culture and longer-term culturing and maturation of human iPSC-derived neurons has been described for over 100 days (Lam et al.

2017).

The heterogeneity of iPSC-derived neurons could present a limitation to our experiments.

Cellular heterogeneity is attributed to the presence of non-neuronal cells such as astrocytes and oligodendrocytes, and indeed we have shown that some of these non-neuronal cells in our cultures were glial associated protein GFAP positive indicating the presence of astrocytes.

However, one advantage of a heterogenous culture is that the interactions between multiple cell types can be studied and this may be a more accurate reflection of the cellular events that take place during disease progression. In support of this a recent study has shown that glial cells are required to activate neuroinflammatory processes in mouse primary cultures

(Hui et al. 2016). As α-synuclein is thought to activate microglia, these interactions are thought to be important for the initiation or progression of pathology in PD (Zhang, Wang, et al. 2005; Su et al. 2008).

Finally, neurite dynamics in iPSC-derived mDA neurons were functionally analysed by staining differentiated cultures for markers for exogeneous α-synuclein (6xHis), TUJ1 and TH. Analysis showed a non-significantly higher number of primary neurites in α-synuclein overexpressing

102

neurons compared to controls. This is in keeping with a recent report that suggested that overexpression of WT, A30P and A53T α-synuclein in primary midbrain neurons led to a significantly higher number of primary neurites (Koch et al. 2015). The results presented in this chapter suggest that overexpression of both WT and mutant G51D α-synuclein resulted in a significant decrease in the total length of all primary neurites, as well as the mean length of primary neurites. And, whilst a decrease by trend was observed for A53T overexpressing cells in both metrics, this may be related to the decreased expression of the A53T mutant α- synuclein relative to control cultures as compared to WT and G51D overexpression, as determined by protein expression analysis. Studies have shown reduced neurite outgrowth in WT overexpression models (Takenouchi et al. 2001) including in an iPSC model (Oliveira et al. 2015), in support of our results.

Furthermore, mDA neurons are thought to form dense axonal arborisations and a recent study has reported evidence for a decrease in synaptic connections in an A53T α-synuclein iPSC-derived PD model (Kouroupi et al. 2017). If branching neurites can be taken as a proxy for synaptic connections, our results suggest decreased branching in response to WT and

G51D α-synuclein overexpression. However, overexpression of A53T α-synuclein did not result in a significant decrease in branching neurites in these iPSC-derived cultures although there was a decrease by trend, possibly associated with the relatively lower expression of

A53T α-synuclein compared to the other overexpression models. These results suggest that overexpression of WT and mutant G51D α-synuclein can result in perturbations in axonal elongation and branching dynamics which may represent early degenerative phenotypes in

PD.

103

Some limitations that were presented during this study included the need to choose cells in areas of lower density to be able to manually follow neurites accurately. As our cultures were of relatively high density and produced complex neuronal networks as well as intertwined axons, this was challenging. Studies have shown that lower density neuronal cultures induce an increase in synaptic sprouting compared to high density cultures, producing an increased synapse to neuron ratio due to cell-cell interactions (Cullen et al. 2010; Biffi et al. 2013).

Although we have developed a method to analyse only transduced neurons, the levels of expression, at a single cell level, could not be determined using this platform. Furthermore, studies have also identified a role for glial cells in neurite complexity (Lieth et al. 1990;

Kanemaru et al. 2007). Future studies should look at the expression of microtubule associated proteins such as kinesin and dynein to further elucidate the molecular mechanisms associated with perturbed α-synuclein expression.

3.5 Conclusion

SH-SY5Y PD models were generated using lentiviral vectors expressing α-synuclein and characterised for use in subsequent translating ribosome affinity purification (TRAP) experiments described in the next chapter. The CMV promoter was found to be the optimal promoter for use in long-term expression of exogenous transgenes in SH-SY5Y cultures and was used in generating the α-synuclein overexpression models. As yields of mRNA from SH-

SY5Y cells from TRAP were very low, and producing enough cells for use of TRAP in an iPSC- derived model would be both much more expensive and challenging, the use of iPSC-derived neurons as a source of TRAP was not further developed. iPSC-derived PD models were also generated for analysis of functional phenotypes associated with perturbed α-synuclein expression. The hSYN promoter was demonstrated to be the

104

optimal promoter for use in long-term and cell-specific expression of transgenes in neurons.

Functional phenotypes included metrics such as differentiation capacity and neurite dynamics. Functionally, whilst differentiation capacity was unchanged in the iPSC-derived models, neurite dynamics were observed to be perturbed in iPSC-derived models of PD. These models will be further employed in the study of miRNA differential expression associated with perturbed α-synuclein expression.

105

Chapter 4 – Translating ribosome affinity purification and RNA sequencing to study genome-wide gene expression changes induced by α-synuclein 4.1 Introduction

4.1.1 Next generation sequencing and Parkinson’s disease research

Advancements in methodological and computational techniques, such as the advent of next generation sequencing (NGS) have revolutionised the field of biological research. This was accompanied by improvements in the generation of sequencing data and computational algorithm for data analysis, as well as improved molecular biology methods to isolate diverse species of DNA such as genomic DNA, histone-bound DNA (ChIP) and cDNA from RNA sources ranging from total cellular RNA to small RNAs such as miRNA and ribosome-bound mRNA.

Furthermore, the GRCh38 (Genome Reference Consortium Human genome build 38), which is generated and assembled from whole genome reference sequences of multiple individuals, provided a valuable map to which sequencing data can be aligned.

Microarray hybridisation technologies were the first to be employed to generate large datasets of transcriptomic information (Lockhart et al. 1996). Using this platform, studies have identified genome-wide differential expression from diverse PD models identifying the involvement of multiple pathways including: Ubiquitin-Proteasome system (UPS), oxidative stress and vesicle trafficking (Scherzer et al. 2003; Grünblatt et al. 2004; Zhang, James, et al.

2005; Miller et al. 2006; Yoo et al. 2003; Simunovic et al. 2008; Simunovic et al. 2010).

Concordance analyses, to evaluate the reproducibility of microarray PD study data sets, have been able to identify a transcriptomic profile which may be ‘characteristic’ of PD (Oerton et al. 2017; Feng et al. 2017; Tan et al. 2018; Mariani et al. 2016). However, microarrays are

106

designed to only contain known genetic sequences and may exclude novel transcripts due to

RNA splicing or editing.

RNA sequencing (RNA-seq) technology allows for sequencing of individual transcripts without the need for hybridisation and associated biases such as non-specific hybridisation and probe redundancy (Zhao et al. 2014). As it does not rely on pre-designed complementary sequences, this allows for the detection of novel transcripts and transcript isoforms. RNA-seq further has the advantages of allowing for quantification of larger data sets, limited only by sequencing depth, thus able to achieve a larger dynamic range of expression levels, and greater ability to detect low abundance transcripts (Wilhelm et al. 2009). However, with these larger datasets, a considerable bioinformatics challenge is created, requiring more complex data analysis.

An early study employing RNA-seq in neurodegenerative research identified differential expression of apolipoprotein splice variants in Alzheimer’s disease brains compared to controls (Twine et al. 2011). Another early RNA-seq study analysed control human iPSC- derived neurons, at different stages of differentiation, identifying several novel long non- coding RNA (lncRNA) whose expression was drastically changed through the differentiation process, suggesting a role in neurogenesis (Lin et al. 2011). In the study of PD, RNA-seq analyses of PD patient blood leukocytes (Soreq et al. 2014)(Yasuo 2018), skin (Planken et al.

2017), CSF (Hossein-Nezhad et al. 2016) and patient derived iPSC neurons (Kouroupi et al.

2017; Woodard et al. 2014; Lin et al. 2011) as well as non-human models (Hook et al. 2018) elucidated systemic disease processes that are not restricted to neurons. PD post-mortem brain samples have been analysed using RNA-seq and analysis has found links between disease and a number of pathways including response to abnormal unfolded protein, synaptic

107

vesicle release, oligodendrocyte function and immune function (Dumitriu et al. 2015; Riley et al. 2014; Henderson-Smith et al. 2016).

4.1.2 Translatome profiling techniques and advantages over transcriptome analysis

One of the limitations of RNA-seq transcriptomic analysis is that total mRNA levels do not always correlate with the levels of the protein encoded by the mRNAs. Protein abundance is regulated by post-transcriptional and translational regulatory mechanisms as well as protein degradation (Schwanhäusser et al. 2011; Vogel et al. 2012). In fact, a recent study that performed an integrative analysis of the transcriptome and proteome from prefrontal cortex of PD patient and control brains showed relatively little overlap between the generated transcriptomic data and proteomics data. The authors suggest that this discrepancy could be due to the sensitivity of the genome-wide mass spectrometry proteomics technique used which has a smaller dynamic range of detection and is unable to characterise and identify as many protein coding genes (3558 unique proteins) compared to RNA-seq (17,580 protein coding genes) (Dumitriu et al. 2015). Furthermore, the authors mention that post- transcriptional regulation of mRNAs may also play a role in discordance between the two datasets.

Methods have been developed to analyse the translatome and the actively translating mRNA by NGS technologies; these include polysomal profiling, ribosomal profiling and ribosome affinity purification techniques. These techniques can be employed to identify disease associated translational pathways and are more likely to generate data that correlate better to the proteome than whole transcriptome profiling (Wang et al. 2013). Protein synthesis is regulated by ribosomal complexes, with each eukaryotic ribosome composed of a large 60S and small 40S ribosomal subunit. Polysomal profiling and ribosomal profiling rely on the

108

separation of ribosome bound mRNAs by sucrose gradient centrifugation, using specialised equipment and a highly labour-intensive methodology. Furthermore, these techniques require a high amount of cellular starting material, and do not allow for running samples in parallel (Mašek et al. 2011; Ingolia et al. 2012).

To enable cell type-specific profiling of ribosomes from heterogeneous tissues, translating ribosome affinity purification (TRAP) was developed (Heiman et al. 2008). This technique relies on immunoprecipitation of ribosomes that are tagged by expression of L10a ribosomal subunit protein fused with EGFP. This enables affinity purification of the mRNA associated with ribosomes. A further refinement to this technique is the expression of the L10a-GFP in defined cell populations using cell-specific promoters. This method can be quick and efficient and require much less starting material than polysomal or ribosomal profiling. An early application of this technique was the study of translational profiles of nigrostriatal neurons by use of the dopamine receptor promoter to drive expression of L10a-EGFP fusion protein

(Doyle et al. 2008). This resulted in the identification of distinct subsets of genes which distinguish nigrostriatal neurons from closely related neurons. More recently, a transgenic mouse line that expresses L10a-EGFP fusion protein under the control of the TH promoter has been generated to monitor mRNA translational profiles of dopaminergic neurons (Dougherty

2017). mDA neurons are thought to be able to co-release GABA (Tritsch et al. 2012) and, using this transgenic mouse model in conjunction with microarray technologies, researchers were able to identify an upregulated novel enzyme possibly linked to GABA production in these cells (Dougherty 2017).

Studying mRNAs that are actively being translated should lead to a better understanding of gene expression regulation and be a bridge between transcriptomic and proteomic analysis.

109

Identifying actively translated transcripts that are differentially expressed in PD models will help us gain a better understanding of the mechanisms involved in α-synuclein pathology, with the ultimate aim of developing therapeutic gene targets or novel biomarkers for PD.

4.1.3 Library preparation for raw data sequencing acquisition

NGS analysis of the differentially expressed transcripts requires building of a fragment library for sequencing. Library preparation is dependent on research goals and experimental set-up such as the species of origin and cellular compartment of the starting material and amount of raw data reads required for specific analyses (differential gene expression, novel isoform identification). These then inform the decision on choice of sequencing platform used i.e. the chemistry of the sequencing technology, with many different commercially available library preparation protocols available.

The choice of sequencing platform should consider research goals and budget, with longer reads and paired end reads costing significantly more but achieving more reliable information, for example for isoform identification, ease of mapping and transcript identification.

However, for well annotated genomes, shorter reads are usually readily mapped and identified thereby longer reads may not be necessary. Longer read lengths and paired-end reads allow for the identification of transcript rearrangements such as single nucleotide variants, novel transcripts and gene fusions. As the aim of this study was to look at globally differentially expressed genes (DEGs), 50bp single-end reads are sufficient. Furthermore, a large data-set is required to generate enough raw reads for DEG analysis, especially for low abundance transcripts. However, it is important to note that studies have shown that increasing biological replicates is more efficient at achieving robust differential expression analysis compared to increasing read coverage, with increased reads over 10 million per

110

sample giving diminishing returns on ability to discern differential expression (Liu 2014;

Schurch 2016). The RNA-seq platform used in this study resulted in approximately 20 million reads of 50bp length per sample with 3 biological repeats for each condition, to produce robust differential gene expression analysis.

4.1.4 RNA-seq analysis analysis platform and pipelines

Interpretation of RNA-seq data requires analysis pipelines which include computer tools and algorithms to guarantee quality of raw reads, accurate mapping to the reference genome, estimation of transcript expression levels, normalisation and DEG identification (Figure 4.1).

Figure 4.1 – Schematic outlining the processing and steps towards the generation of differential expression analysis from TRAP-derived mRNA. mRNA is first amplified and reverse transcribed to generate a cDNA fragment library. Library preparation is performed by ligating universal and index adapters and amplicons sequenced. Sequencing raw data reads are checked for integrity and mapped to a reference genome. A transcriptome assembly is created and merged together and used to normalise expression levels and normalised to produce a list of differentially expressed genes.

For differential gene expression analysis, many RNA-seq analysis pipelines have been developed to produce efficient and reproducible analysis of data. The tools to analyse genomic data can be run on more complex coding software such as R statistical programming language (Gentleman et al. 2004) or more user-friendly platforms like the widely used open

111

source web-based platform Galaxy (Giardine et al. 2005) and commercial packages such as the Partek suite of analysis tools. With such a compendium of softwares available for the complicated analysis required for DEG analysis, there is yet no consensus about which methodology produces the most robust and reproducible analysis data. Studies that have compared analysis packages have shown the importance of package choice, as their integration of different normalisation algorithms can markedly affect data analysis in differential gene analysis in terms of specificity and sensitivity of differentially expressed gene detection (Conesa et al. 2016; Costa-Silva et al. 2017; Zhang, Jhaveri, et al. 2014; Rapaport et al. 2013). These comparative studies note that no single method produces clearly superior analyses and have further suggested that the integration and consensus between different packages may be best analysed for robust specificity and sensitivity of detection. In this study, the Tuxedo package of analysis software (Figure 4.2) was used primarily as it has been developed for wet-lab researchers without necessitating the learning of complex programming languages. The Tuxedo suite (Trapnell et al. 2012) pipeline (outlined in Figure

4.2) was used with different combinations of alignment algorithms and reference annotation datasets (Figure 4.2 in red) to test the reproducibility of our analysis which is a strategy used in RNA-seq, i.e. to compare the results of analysis by different pipelines (Guo et al. 2014). The published pipeline consists of software 3 main modules; an alignment module (Tophat), a transcript assembly module (Cufflinks) and a differential gene expression module (CuffDiff).

112

Figure 4.2 -– Schematic outlining pipeline for RNA-seq raw data towards differential expression analysis using Tuxedo pipeline (adapted from (Trapnell et al. 2012)) using the Galaxy or Partek platforms. Input files were parsed and quality of reads quality determined using FASTQC. Reference annotation used was the ENCODE project’s Gencode human hg38 library which was used to align the raw reads with Tophat or STAR splice-junction aware aligner to produce general alignment statistics. Reads were assembled into a transcriptome assembly and merged by Cufflinks. Differential expression and statistics were determined by the CuffDiff module.

Raw reads from RNA-seq are initially aligned to a reference genome, with the newer generation of aligners such as STAR (Dobin et al. 2013) and Tophat (Kim, Pertea, et al. 2013) being able to recognise splice-junctions and map exon-intron junctions to detect alternatively spliced transcripts. Currently, the Genome Reference Consortium human reference 38

(GRCh38) build is the primary genome assembly used for aligning raw reads and contains alternative sequences for genomic regions where variability and complexity has been noted.

Once raw reads have been aligned, they require assembling together to full-length transcripts, a process known as transcriptome assembly or reconstruction. Tools that have been developed to assemble transcripts include Cufflinks (Trapnell et al. 2012), Scripture

(Guttman et al. 2010) and Stringtie (Pertea et al. 2015). Transcript assembly and merging can be undertaken with or without a reference annotation, usually depending on the availability of well-annotated transcriptomes for the species being studied. For humans, several reference transcript annotation datasets exist, including Refseq, Ensembl and Gencode.

Gencode is an additive set of annotations which merges the European EMBL-EBI Ensembl automated (Hubbard et al. 2002) and Havana manual gene annotations (Harrow et al. 2012)

113

as part of the ENCODE project’s follow up to the Human Genome Project whilst RefSeq annotations have been developed by the American NCBI (Pruitt et al. 2006). The individual transcriptome assemblies are then merged together giving a single merged assembly to normalise transcript expression. These full-length transcripts are then quantified and normalised to measure relative abundances.

Finally, for differential expression analysis, the reads and the merged transcriptome assembly are run through an algorithm to calculate expression levels in multiple sample conditions and tests the statistical significance of these changes. Programs commonly used for differential expression analysis include Cuffdiff, DESeq and EdgeR (Anders et al. 2010; Trapnell et al. 2012;

Robinson edgeR 2009). These programs differ in the type of statistical models used to calculate relative abundance and expression with Cufflinks using a read-normalisation strategy whereas DESeq and EdgeR use ‘read count-based methods’ (see review (Guo et al.

2014)), although there is no consensus on the best statistical strategy for DEG analysis (Anders et al. 2010).

4.1.5 Aims

In this Chapter, I will determine differential gene expression in α-synuclein overexpressing

SH-SY5Y models that were generated as described in Chapter 3. The aims of the experiments are to:

1) successfully prepare an RNA-seq library from an enriched population of translating

mRNAs using the TRAP platform;

2) perform NGS to determine genes that are differentially expressed due to wild-type or

mutant α-synuclein overexpression;

114

3) validate differentially expressed genes using molecular and cellular biology techniques

including qPCR and Western Blotting.

115

4.2 Materials and Methods

4.2.1 Lentiviral transduction of L10a-EGFP into SH-SY5Y PD models

Lentiviral particles that express the L10a-EGFP fusion protein from the CMV promoter were used to transduce undifferentiated control, wild-type and mutant A53T α-synuclein overexpression SH-SY5Y cells (generated as described in section 3.2.2), at an MOI of 2. FACS sorting was performed as described in section 3.2.1 (Appendix 3). 250,000 cells per sample were collected and re-plated onto T25 flasks, separately for each sample.

4.2.2 Translating ribosome affinity purification and mRNA purification

FACS sorted samples were passaged and cultured and differentiated (section 2.2.2) on 10cm2 plates (Corning; #430591). Cells were differentiated for 14 days prior to TRAP processing.

After 14 days, TRAP was started using the protocol adapted from the original publication

(Figure 4.3)(Heiman et al. 2014).

Figure 4.3 - Schematic outlining protocol for extracting mRNAs in translating polysomes. L10a, a 60S ribosomal protein fused to EGFP is transduced into undifferentiated SH-SY5Y cells. Cells are lysed in buffer spiked with RNase-inhibitors, cycloheximide to stall translation and DHCPs to retain polysomes in complex. These isolated polysome complexes can be purified by adding EGFP conjugated magnetic beads to isolated polysome complexes from total RNA Briefly, on the day of TRAP, cells had 100µg/ml cycloheximide (Sigma; #C7698) added to culture media and incubated for 15mins. Cells were then washed, on ice, with PBS with

100µg/ml cycloheximide and then lysed with lysis buffer (Table 4.1). Cells were incubated on ice for 10mins and pipetted vigorously and transferred to a prechilled microcentrifuge tube

116

and centrifuged at 2000g for 10mins at 4°C. The supernatant was transferred into a fresh chilled microcentrifuge tube with 1/9 volume 300mM DHPC (Avanti Polar Lipids; #850306P) and mixed by inversion and incubated for 5mins on ice. To clear the lysate, samples were centrifuged at 20000g for 15mins at 4°C and transferred supernatant into fresh chilled microcentrifuge tubes.

High Salt (0.35M KCl) Low Salt (0.15M KCl) Manufacturer and Cat. Cell-lysis Buffer Wash Buffer Wash Buffer No. 20mM HEPES (pH7.3) 20mM HEPES (pH7.3) 20mM HEPES (pH7.3) VRW; 169241 Applied Biosystems; 150mM KCl 350mM KCl 150mM KCl AM9640G Applied Biosystems; 10mM MgCl 10mM MgCl 10mM MgCl AM9530G NP-40 1% (vol/vol) NP-40 1% (vol/vol) NP-40 1% (vol/vol) AG Scientific P1505 RNase-free H2O RNase-free H2O RNase-free H2O Ambion; AM9938 Add fresh: Add fresh: Add fresh:

0.5mM DTT 0.5mM DTT 0.5mM DTT Sigma; D9779 100g/ml cycloheximide 100g/ml Cycloheximide 100g/ml Cycloheximide Sigma; C7698 10l/ml RNasin RNasin 1l/1ml RNasin 1l/1ml Promega; N2515 Applied Biosystems; 10l/ml Superasin AM2694 EDTA-free protease inhibitor (1 Roche; 11836170001 tablet in 10ml) Table 4.1 – Lysis and wash buffers utilised in TRAP methodology

300μl of Pierce Protein L magnetic beads (ThermoFisher; #88849) were washed three times with 3% w/v nuclease-free BSA in nuclease-free PBS, using a magnetic tube holder. Beads were then resuspended in high-salt buffer (Table 4.1). To this, 50µg of each of anti-EGFP antibodies 19F7 & 19C8 (Rockafeller Institute) was incubated on a slow-tilt rotator for 1 hour.

Beads were then collected on a magnet and washed 2 more times with 1ml low-salt buffer

(Table 4.1). Antibody coated beads were then resuspended in 180µl low-salt buffer with 20µl

300mM DHPC. Beads in low-salt buffer were added to freshly extracted RNA samples and incubated at 4°C overnight on a rocker.

After overnight incubation of antibody coated beads and RNA samples, beads were collected by a magnet and washed 3 times with 1ml high-salt buffer (Table 4.1). Finally, beads were

117

collected by magnet and resuspended in 100µl Agilent absolute RNA Nanoprep kit lysis buffer

(Agilent; #400753), vortexed and incubated for 10mins at RT to remove beads from polysomal complexes. Beads were collected by magnet and RNA, now in lysis buffer, was transferred into a fresh chilled microcentrifuge tube and purified using the Agilent absolute RNA nanoprep kit, according to manufacturer’s instructions, and resuspended in 30µl RNase-free

H2O and quantified on a nanophotometer (Geneflow; IMPLEN P300). RNA quality was determined using the high sensitivity RNA ScreenTape (Appendix 4)(Agilent; #5067-5579) platform on the Agilent 2200 TapeStation (Agilent; #G2964AA).

4.2.3 RNA amplification and cDNA library preparation for next generation sequencing on illumina platform cDNA library preparation for NGS requires a choice of sequencing platform, as workflow will be determined by this choice. The well-established lllumina platform technology was chosen for this study. To begin preparation for entry into the NGS workflow, for the lllumina platform requires 4 steps: 1) Having enough starting material; 2) Blunting and A-Tailing of mRNA 3)

Ligation of adapters that can be read by the RNA-seq machinery and 4) Amplification of ligated amplicons (Figure 4.4).

118

Figure 4.4 - Schematic outlining the processing of TRAP samples from amplification of RNA to library preparation and pooling towards sequencing

A limitation of the TRAP technique is the low yields of RNA isolated that is not sufficient for preparation of cDNA libraries. Thus, in this study, the Seqplex RNA amplification kit (Sigma;

SEQR) was employed using 10ng of starting material from each sample. The resultant dsDNA

(150-400 base pairs (bp) fragments) was amplified with PCR in a 96 well qPCR plate (Applied

Biosystems; #4346906) with SYBR Green (Sigma; #S9430) diluted in 10mM Tris-HCl (Sigma;

T3038) to monitor the amplification process, in the StepOne quantitative qPCR system

(Applied Biosystems; #4376357). At this point, the samples were cleaned up with the

GenElute PCR Cleanup Kit (Sigma; #NA1020), according to manufacturer’s instructions, and eluted in 50µl nuclease-free H2O. Amplified samples were treated with primer removal enzyme and QC by qPCR as per manufacturer’s instructions and quantified by nanophotometry and assayed on the D1000 ScreenTape (Agilent; #5067-558), using the

Agilent 2200 TapeStation system (Appendix 5).

Blunting/filling-in and A-tailing was performed in one reaction mixture. Here, 150ng of cDNA fragments from RNA amplification product was mixed with 1µl of T4 DNA polymerase (New

119

England Biolabs; #M0203S) and 0.5µl Taq polymerase with 3µl 10x Taq buffer (New England

Biolabs; #M0273S) with 1µl of 10mM dNTPs (equimolar solution of dATP, dCTP, dGTP and dTTP; New England Biosciences; #N0440S, #N0441S, #N0442S, #N0443S) in a 30µl reaction in

PCR tubes, for each sample separately. This mix was assembled on ice and then incubated at

12°C for 30 minutes (end polishing), 75°C for 20 minutes (heat inactivate the T4 DNA polymerase) and 72°C for 10 minutes (activate A-tailing by Taq polymerase) in a gradient thermal cycler (Techne; TC512).

Adapters compatible with the Illumina hiseq platform were formed by annealing equimolar

(500µM) concentrations of a universal and index oligonucleotide with 1x Fast digest buffer

2.1 (New England Biolabs; #B7202S) (Figure 4.5) made up to 50µl in nuclease-free H2O

(Ambion; #AM9938). The tubes were placed on a heatblock (Techne; DB2D) at 95°C for 3 minutes. The heat block was removed from the apparatus and allowed to cool to RT slowly for 60 minutes. 100ng of end repaired, A-tailed DNA fragments was mixed with 1µl DNA ligase

(New England Biolabs; #M0202S) and 1µl 10µM annealed oligonucleotides and 10µl 10x ligase buffer (provided with DNA ligase) made up to 100µl in nuclease-free H2O.and incubated for 1 hour at 16°C. Finally, the reaction mix was column purified using the Genelute PCR Clean-up kit and eluted in 30µl H2O, ready to proceed to further amplification and library QC.

AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTACACGACGCTCTTCCGATCT

GTTCGTCTTCTGCCGTATGCTCTAXXXXXXCACTGACCTCAAGTCTGCACACGAGAAGGCTAG/5Phos/

Figure 4.5 -Sequences of the universal adapter (top) and index adapter (bottom). In blue is the complementary portion of the 2 sequences and the unique 6bp index region in the index adapter in red. 2 of these are ligated to flank A tailed fragment from TRAP samples. The index adapter includes a 5’ phosphate group to aid in ligation of the adapter to the DNA fragments.

Adapter ligated fragments were amplified using primers which recognise the first 20bps of the universal adapter sequence (Amp1) and the first 20bps of the index adapter (Amp2)

(Appendix 6). A 5x PCR master mix was made by mixing 0.5µl of 10µM of each primer with

120

50µl Phusion master mix enzyme (ThermoFisher; #F548S) with 100ng of adapter ligated fragments, made up to 100µl with nuclease-free H2O and divided into 5 separate PCR tubes for each sample. The PCR tubes (Alpha Labs; #LW2510) were placed in a thermal cycler and cycles set to denature at 98°C for 1s, anneal at 60°C for 5s and extend at 72°C for 15s. The

PCR tubes were removed separately at cycle 9, 11, 13, 15, and 17.

Molar concentrations were determined for the range of cycles using the NEBnext library quant kit for Illumina (New England Biolabs; #E7630) which contains primers which target the

P5 and P7 Illumina adaptor sequences and as such will quantify only those fragments which have had adapters ligated successfully (Appendix 6). The amplification was carried out using high ROX conditions in the StepOne qPCR system. Cycle 15 was chosen for pooling and analysed on the TapeStation 2200 using the D1000 ScreenTape to work out average library size for use in quantification (Appendix 7). Equal quantities of each sample were pooled together and purified with the Genelute PCR clean-up system and eluted in 32µl nuclease- free H2O and concentration determined for the pooled library using the NEBNext library quant kit. This sample was stored at -20C and sent away for sequencing.

4.2.4 Gene expression analysis

When raw reads are generated, the first step is to assess the raw data for its integrity for quality control purposes by programs such as FastQC (Andrews 2010). This program gives an overview of potential problems with raw sequencing data before further analysis by producing: basic statistics, sequencing quality per base, GC content, sequence duplication, overrepresented sequences and kmer content analysis.

Sequencing results were aligned to a reference genome, GRCh38, using 2 validated alignment programs, STAR (Dobin et al. 2013) and Tophat2 (Kim, Pertea, et al. 2013), using Partek Flow

121

Software (Ion Torrent; #4485102), to give alignment statistics of mapped reads (Figure 4.6).

Mapped reads were then assembled to create a final transcriptome assembly using Cufflinks and differential expression of genes and transcripts was quantified using Cuffdiff, which form the Tuxedo Suite protocol for RNA-seq (Trapnell et al. 2012). In parallel, the Galaxy platform for RNA-seq analysis was also used with TopHat and following the tuxedo protocol (Goecks et al. 2010) (Figure 4.6).

Figure 4.6 - Schematic outlining pipeline for RNA-seq raw data towards differential expression analysis using Partek Flow and Galaxy platform using the Tuxedo pipeline

4.2.5 – Validation of differentially expressed genes by transcript expression analysis by

RT-qPCR using SYBR Green

RNA was extracted from day 14 differentiated SH-SY5Y PD models using the mirVana RNA extraction kit (Invitrogen; #AM1560) as per manufacturer’s instructions. For cytoplasmic RNA purification and isolation, the RNeasy minikit, according to manufacturer’s instructions. RNA concentration was measured and 2μg of cDNA was generated with the SuperScriptIII reverse transcriptase cDNA kit (Invitrogen; #4376357) using random hexamer priming. Differentially expressed genes (DEGs) were chosen from RNA-seq analysis and primers prepared (Sigma) for validation by qPCR expression analysis (Table 4.2).

Choice of normalisation housekeeping gene was determined by running qPCR reactions, for all samples, using 3 separate housekeeping genes commonly used for normalisation in SH-

122

SY5Y cells: β-Actin (ACTB), human 60S acidic ribosomal protein P0 (hRPLP0) and

Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Španinger et al. 2013). Thus, these housekeeping genes were analysed for using Bestkeeper algorithm, which calculates the coefficient of variance and standard deviation (SD) of the raw quantitation cycle (Cq) values and works out the geometric mean of the Cq values (Pfaffl et al. 2004). This averaged geometric mean is then compared with each housekeeping gene separately to establish the housekeeping gene with the best correlation to the geometric mean of all selected housekeeping genes. Thus, in this study gene expression was normalised to β-Actin (ACTB) housekeeping gene.

Table 4.2 -List of primers for validation of DEGs from RNA-seq analysis

4.2.6 – Protein expression analysis using Western Blot

Western Blotting was employed to analyse protein expression in SH-SY5Y synucleinopathy models for DEGs, as determined by RNA-seq analysis and as described in section 4.2.4.

Antibodies used include DGCR8 (Proteintech; #10996-1-AP; 1:1000), SLC25A12 (Santa Cruz;

123

#sc271056; 1:1000), UNC119B (Proteintech; #26201-1-AP; 1:1000), CRADD (Santa Cruz; #sc-

377080; 1:1000), SNCA (BDBioscience; #610787; 1:1000), TOMM40 (gift from Prof. Collinson,

Bristol University; 1:500) and β-Actin (Abcam; #ab20272; 1:2000) used as a normalisation control.

124

4.3 Results

4.3.1 TRAP and library preparation

4.3.1.1 Translating ribosome affinity purification

At the end of the 14-day differentiation of the L10a-EGFP transduced SH-SY5Y PD models

(described in Chapter 3), mRNA in translating polysome complexes were immunoprecipitated by TRAP and mRNA was extracted as described in Section 4.2.2. The RNA structural integrity in the samples were analysed on the Agilent RNA tapescreen. All samples had a RNA integrity number (RIN) (Appendix 4) of over 7 which indicates relatively intact RNA and were deemed suitable for further processing (Table 4.3).

TRAP RNA Average dscDNA Average amplicon Sample RIN conc. (ng/µl) fragment size size Control 1 6.0 8.2 195 213 Control 2 12.0 8.0 190 232 Control 3 17.2 7.9 197 239 Wildtype 1 13.2 7.9 194 325 Wildtype 2 13.2 7.9 188 319 Wildtype 3 18.8 7.6 193 320 Mutant A53T 1 33.5 7.3 184 320 Mutant A53T 2 18.8 7.8 189 318 Mutant A53T 3 18.8 7.5 195 236 Table 4.3 – Table shows the concentration and RIN of TRAP-derived mRNA, average dscDNA insert fragment size and average adapter ligated amplicon size and amplified dscDNA fragment analysis by capillary gel electrophoresis.

4.3.1.2 NGS library preparation

RNA amplification was employed to produce double stranded cDNA for processing into a sequencing library. The resultant dsDNA was made up of random, overlapping 150-400 base pair (bp) fragments and cDNA concentration was quantified by a nanophotometer and Agilent

DNA Tapestation was used to visualise average size of the fragments (Table 4.3). Thus, average cDNA library insert size after amplification ranged between 184-197 bps, within the

125

recommended 150-200bp range, and concentrations were sufficient to proceed with library preparation. The library preparation protocol is outlined in section 4.2.3. cDNA amplified from TRAP-derived RNA was taken through the library preparation protocol and an equimolar concentration of from each sample was pooled together after quantification of adapter-ligated amplicons using qPCR. The RNA-seq library was analysed using Agilent Bioanalyzer prior to sequencing and this showed a small peak which corresponds to primer or primer-dimer contamination (Figure 4.7A). This type of contamination can be especially detrimental to achieving accurate sequencing data as smaller sized amplicons are preferentially sequenced, thus yielding low quality data. Therefore, a bead-based size selection protocol was applied to the pooled sample to remove smaller fragments (<100bp)

(Figure 4.7B).

Figure 4.7 - Fragment size selection on pooled adapter ligated amplicon. A) Before size selection with 2 small sized peaks approximately (yellow arrows) 50-100bp in size, corresponding to primers and primer dimers. B) After removal of peaks.

126

4.3.2 RNA-seq Analysis

4.3.2.1 Raw read general read statistics

Raw reads were returned as multiple “.fastq” files that contain both the sequencing and quality scores. The FASTQC (Babraham Bioinformatics; https://www.bioinformatics.babraham.ac.uk/index.html) module on Galaxy was used to assess the quality of raw reads and identify any flags in quality of raw data (Figure 4.8).

Overall, average Phred quality scores for all samples was >38. These scores are logarithmically linked to error probability such that a Phred score of 30 represents a 99.9% (1:1000) base call accuracy and a score of 40 (1:10000) represents 99.99% base call accuracy. A Phred score of over 30 suggests high quality of reads.

Figure 4.8 – Outline of FASTQC module summary output page. Raw data quality are analysed including: Basic statistics, per base sequencing quality, per tile sequence quality, per sequence quality scores, per base sequence content, per sequence GC content, per base N content, sequence length distribution, sequence duplication levels, overrepresented sequences, adapter content and Kmer content.

The FASTQC module did however identify flags for per base sequence content at the start of reads which, in a random library should show no difference however, is a normal artefact in

127

the first few bases due to our amplification method. Duplicated sequences and overrepresented sequences were also flagged, with sequence BLAST

(https://blast.ncbi.nlm.nih.gov/Blast.cgi) showing some of the duplicated sequencing reads as a result of rRNA overrepresentation associated with ribosomal preparations such as TRAP.

Kmer content was also flagged which suggested an enrichment or overrepresentation of certain motifs which is linked to duplicated signals. These general read statistics suggested good quality of reads but with some overrepresented/overamplified sequences. This is most likely due to the unavoidable need for initial amplification of RNA from a less diverse source such as the ribosomal compartment compared to overall total mRNA.

4.3.2.2 Alignment and general alignment statistics

In the first instance, both STAR and Tophat splice-aware aligner were used to align the raw data to the human genome (GRChg38) and generate mapping alignment statistics for individual samples (Figure 4.9). 240 million Illumina HiSeq raw data reads were sequenced and showed comparable total reads generated from individual samples ranging from

2.13x107-2.72x107 with an average of 2.44x107±2.00x106. Mapping and alignment of reads to human genome was consistently above 90% for all samples (91.8-94.5% for Tophat and 94.1-

97.0% for STAR). The majority of these reads were uniquely aligned (54.2-62.0% for Tophat and 57.8-65.8% for STAR). Multi-aligned reads made up a significant portion of the non- uniquely aligned reads for both aligners. It is normal for some reads to align to several genomic loci due to repetitive sequences or common domains of gene paralogues (Conesa et al. 2016).

128

Figure 4.9 – Graph and table showing overall alignment statistics including percentage unique, multi-mapped and unmapped reads using both A) Tophat and B) STAR alignment tool, aligned to the hg38 human genome reference consortium build 38 (hg38).

129

4.3.2.3 Differential expression analysis

Differential expression analysis was determined using the Tuxedo RNA-seq pipeline, using the

Galaxy and Partek flow platforms with slight modifications in aligner and reference transcript sequence databases (Figure 4.10A). A list of DEGs was generated (q-value <0.05) which compared two experimental conditions: 1) Control compared to WT overexpression (effect of increased α-synuclein) and 2) WT compared to A53T overexpression models (effect of mutant synuclein) from each pipeline. The modified pipelines were compared to determine the influence of both aligner and reference annotation on DEG analysis.

130

Figure 4.10 – Pipeline analysis. A) Schematic of the processing of raw data using the Tuxedo suite, with modifications used, for determining DEGs from raw data. B,C) A comparison of DEG lists, generated by pipelines outlined in A, for B) Control compared to WT α-synuclein overexpression and C) WT α-synuclein compared to A53T α-synuclein (n=3, q>0.05). Tables show Total number of DEGs and number of DEGs showing greater than 2 fold change as well as number and percentage of DEGs common between 2 pipelines for total DEGs and DEGs showing greater than 2 fold change.

131

The different pipelines (1, 2, 3) revealed 197, 184 and 161 DEGs respectively when control samples were compared to WT-synuclein expressing cells (Figure 4.10B). When the WT population was compared to the mutant A53T population, the highest number of DEGs in analysis for WT compared to A53T was observed in pipeline 2 (425), followed by pipeline 1

(360) and then pipeline 3 (343) (Figure 4.10C). Regardless of the pipeline used, there were more DEGs between WT and A53T populations, compared to control versus WT populations.

The use of Tophat aligner with either Genode or Ensembl reference transcript annotations yielded more DEGs compared to the use of STAR aligner with the Ensembl reference dataset.

Although the Gencode dataset contains more annotated genes such as pseudogenes that are not found in the Ensembl reference annotation, this was not associated with an increase in observed DEGs.

DEGs common between pipelines were also determined. When WT cells were compared to control samples, pipeline 1 shared 54.3% and 39.1% of overall DEGs with pipelines 2 and 3 respectively whilst pipelines 2 and 3 shared the highest amount of common overall DEGs,

56.5%. Therefore, the use of the same reference transcript annotation here yielded the highest number and percentage of common overall DEGs although the use of the same aligner with the different reference transcript annotation also showed similar levels of common

DEGs. Thus, the synergistic effect of the use of a different reference annotation and aligner created less reproducibility of DEGs.

When WT synuclein cells were compared to A53T cells, pipeline 1 shared 50.8% and 42.8% of overall DEGs with pipelines 2 and 3 respectively, whilst pipelines 2 and 3 shared 45.0% of overall DEGs. Here, the use of a common aligner yielded the highest number and percentage of common DEGs, followed by the use of a common aligner with different reference transcript

132

annotation. Again, the use of both a different aligner and reference genome annotation created the least reproducibility in common DEGs. This analysis shows how the use of different reference transcript annotation datasets and aligners can affect DEG analysis.

Analysis also suggests that the synergistic effects of the use of a different reference transcript annotation and aligner yields the lowest reproducibility in DEGs.

Overall, DEG analysis identified 33.5% of DEGs common to all 3 pipelines for control compared to WT cells and 32.9% for WT compared to A53T populations. However, there were also many

DEGs identified by only 1 pipeline. As Gencode annotations are the most highly curated, the lists generated using this pipeline (pipeline #1) were used as the basis for choice of DEGs taken forward and validated by further techniques, with DEGs that were reproduced by more than one pipeline prioritised for validation. The top 10 upregulated and downregulated genes between the two experimental conditions are outlined (Figure 4.11).

133

Figure 4.11 – Top 10 significantly (p<0.05) up-regulated (Green) and down-regulated (Blue) protein coding DEGs, as determined using Tuxedo pipeline on the Galaxy platform with Gencode reference annotation, with fold changes and functions, between Control and WT (top) and WT and A53T overexpression (bottom) conditions (n=3).

4.3.2.4 Pathway analysis

Gene Ontology (GO) groups together genes of a common pathway to determine biological phenotypes (Ashburner et al. 2000). These enrichment analyses included GO terms related to cellular and biological function or mapping onto protein interaction and signalling networks using the Kyoto Encyclopedia of Genes and Genomes (KEGG). DEGs were filtered before

134

enrichment analysis for statistical significance (q value <0.05) and for fold changes (> +2 or -

2). DEGs generated from this study were categorised based on their related functions by GO and KEGG enrichment using Enrichr online software (Figure 4.12 & Tables 4.3 & 4.4) (Chen,

Tan, et al. 2013; Kuleshov et al. 2016). KEGG analyses showed non-statistically significant enrichment of genes associated with Lysine degradation, Glycerolipid metabolism and synthesis and degradation of ketone bodies as well as genes involved in protein processing differentially expressed in WT overexpression cultures compared to control (Figure 4.12A).

KEGG analysis showed a statistically significant enrichment of genes associated with ECM- receptor interaction, as well as non-statistically significant differential expression of genes associated with Notch signalling and Alzheimer’s disease from A53T overexpression compared to WT overexpression analysis (Figure 4.12B). Cellular component and biological function GO pathways are outlined in Table 4.4 and 4.5.

135

Figure 4.12 - KEGG pathway analysis for DEGs which were >2 fold significant differentially expressed determined by Tuxedo with Gencode transcript annotation for conditions A) Control compared to WT α-synuclein overexpression, B) WT α-synuclein overexpression compared to A53T α-synuclein overexpression with a p value assigned using Fisher’s exact test proportion test.

136

Table 4.4 – Gene ontology gene set enrichment analysis for cellular component related genes. Genes were grouped together based on common pathways with a p value assigned using Fisher’s exact test proportion test.

Table 4.5 - Gene ontology gene set enrichment analysis for biological function related genes. Genes were grouped together based on common pathways with a p value assigned using Fisher’s exact test proportion test.

137

4.3.3 Validation of Differentially expressed genes determined by RNA-seq

4.3.3.1 Evidence based choice of suitable housekeeping gene

Selection of reference gene is important for qPCR analysis and commonly used housekeeping genes such as the 60S acidic ribosomal protein large P1 (RPLP1), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and beta-actin (ACTB) have been identified as being suitable endogenous controls in SH-SY5Y cells (Španinger et al. 2013). Many studies have suggested that reference housekeeping genes may be differentially expressed between cell types and disease conditions and thus it is best to determine the most reliable and stable reference gene for use in these cells and experimental conditions. Thus, 3 commonly used housekeeping genes were analysed using Bestkeeper (Pfaffl et al. 2004). BestKeeper calculates the coefficient of variance and standard deviation (SD) of the raw quantitation cycle (Cq) values and establishes the BestKeeper index, which is the geometric mean of the Cq values of candidate housekeeping genes. Linear regression analysis showed that the candidate reference gene with the highest correlation to the BestKeeper index, with a coefficient of correlation of 0.767 and p value of 0.016 was ACTB and the standard deviation was acceptable to the algorithm (Figure 4.13).

Figure 4.13 – Bestkeeper regression analysis of 3 candidate housekeeping genes: GAPDH, ACTB and HRPLP. GAPDH (p 0.047) and ACTB (p 0.016) were determined to have adequate coefficient of correlation, 0.674 and 0.767 respectively, to the Bestkeeper index. Column data analysis including range of raw Cq values, SD and Coefficient of variation. Graph to visualise raw Cq variation. (n=1)

138

4.3.3.2 – Validation of DEGs cDNA generated from TRAP-derived RNA was used to validate DEGs with >2 fold change that were statistically significant (q value <0.05), based on potential links to pathological processes and pathways (Figure 4.11). Selected DEGs were confirmed by independent qPCR analysis using total RNA and TRAP-RNA and cytoplasmic-RNA which excludes nuclear RNA and represents a closer compartment to the TRAP samples. As TRAP measures changes in actively translating mRNAs, Western blot analysis was used to measure changes in protein expression where an appropriate antibody was available.

Unexpectedly, one of the DEGs that was identified when WT and A53T samples were compared was α-synuclein itself, which showed an approximately 2-fold change increase in average α-synuclein transcript levels in the A53T mutant cultures compared to WT cultures

(Figure 4.14A). To determine whether this was a bona fide result, qPCR analysis was employed to analyse transcripts in our cultures (Figure 4.14B). It does appear as though there is a difference between α-synuclein transcript expression levels in the A53T models (116.2±28.4) and the WT models (47.4±1.75). The difference between the WT and A53T α-synuclein transcript levels was not significant (p<0.06), as determined by one-way ANOVA using Tukey multiple comparisons testing. As this was determined subsequent to TRAP isolation and RNA sequencing, this does mean that downstream analysis should be interpreted with this result in mind.

139

Figure 4.14 -A) RNA-seq analysis of differential expression of SNCA showing a 2 fold increase α-synuclein transcript expression in A53T compared to wild-type overexpression. B) qPCR transcript analysis (±SD) of SH-SY5Y cultures showing a significant increase in α-synuclein transcript levels between A53T and non-transduced cultures (p<0.006) and non- significant change between WT and A53T and control and WT cultures. One-way ANOVA, Tukey post-hoc test (n=4). ** p<0.01 (±SD)

Thus, despite care during the experiment to ensure that the SH-SY5Y cells were transduced at equivalent MOI for the different vectors, approximately 2-fold increase was detected in

A53T compared to WT overexpression models. The implication of this observation is that these models are not “perfect” and that DEGs that are detected between these two groups may be attributed to either increased synuclein expression or mutant synuclein expression, or both.

4.3.3.2.1 – Cytoskeletal protein encoding genes

There were a number of cytoskeletal protein coding genes that were found to be differentially expressed in our analysis. As neurite dynamics were perturbed in iPSC-derived PD models

(described in Chapter 3.3.3), two cytoskeletal genes that were highly differentially expressed were chosen for validation. The first gene FGD3, which encodes the FYVE, RhoGEF and PH

140

domain containing 3 protein, was found to be increased in WT and A53T samples compared to control by 14.72 and 7.21-fold, respectively (Figure 4.15A). FGD3 is involved in p75 neutrophin receptor-mediated signalling and survival, and process formation in neurons

(Hayakawa et al. 2008) and has previously been found to be upregulated in whole blood samples from AD patients compared to control samples by whole-genome microarray (Bai et al. 2014).

The second DEG was SPON2, which was found to be decreased in WT and A53T samples compared to control by 4.63 and 5.35-fold, respectively (Figure 4.15A). SPON2 encodes an extracellular matrix protein that has been shown to promote outgrowth of neurites in hippocampal embryonic neurons (Feinstein et al. 1999).

141

Figure 4.15 – Gene expression analysis for DEGs which encode cytoskeletal proteins FGD3 and SPON2. A) RNA-seq results table showing Log2 fold change, fold change, q value and significance as well as number of different pipelines the DEG was identified in (n=3). Total, cytoplasmic and TRAP-RNA gene expression analysis (±SD) for B) FGD3 and C) SPON2 using the ΔΔCt method with ACTB used as a housekeeping control (n=3). One-way ANOVA, Tukey post-hoc test (n=4). *p<0.05, ** p<0.01, *** p<0.001 qPCR gene expression analysis for FGD3 showed a significant upregulation in A53T overexpressing cells (2.05±0.21) compared to both control (1.00±0.42) and WT overexpressing cells (0.80±0.08), in terms of total RNA samples (Figure 4.15B). However, no significant differences were observed in cytoplasmic RNA samples. Contradicting TRAP-RNA gene expression analysis showed a significant decrease in A53T overexpressing cells

(0.73±0.04) compared to control (1.00±0.10) (Figure 4.15B). qPCR gene expression analysis for SPON2 showed a significant increase in A53T cells

(2.14±0.11) compared to control (1.00±0.33) and WT (0.83±0.06) cells in total RNA-derived

142

samples (Figure 4.15C). Expression of SPON2 in cytoplasmic RNA A53T overexpressing cells was also increased significantly (1.47±0.03) compared to control (1.00±0.12) and WT overexpressing cells (0.96±0.19) in cytoplasmic RNA samples. Both of these results contradicted RNA-seq analysis results where SPON2 was observed to be decreased, whilst no significant changes were observed in TRAP-RNA samples (Figure 4.15C).

4.3.3.2.2 – Mitochondrial protein encoding genes

PD has long been associated with mitochondrial dysfunction (described in section 1.4.7.2), with several disease association genes having been identified including Parkin and PINK1

(described in section 1.5.2 and 1.5.3). Several mitochondrial protein encoding genes were differentially expressed in RNA-seq analysis and two of these were taken forward to validation.

The first DEG chosen for validation was the translocase of outer mitochondrial membrane 40,

TOMM40, which was identified to be increased in A53T overexpressing cells compared to control and WT overexpressing cells by 52.35 and 60.97-fold, respectively (Figure 4.16A).

TOMM40 encodes the channel forming mitochondrial protein of the translocase of mitochondrial outer membrane complex. Downregulation of this protein has been associated with reduced mitochondrial protein import and increased mitophagy of outer mitochondrial membrane proteins by PINK/Parkin (Bertolin et al. 2013). Decreased TOMM40 mRNA has also been observed in PD patient brain and α-synuclein WT and mutant overexpression B103 rat neuroblastoma models (Bender et al. 2013).

The second DEG SLC25A12 was identified to be increased in both WT and A53T overexpressing cells compared to control by 9.95 and 15.02-fold, respectively (Figure 4.16A). SLC25A12, encodes AGC1, a mitochondrial aspartate-glutamate carrier protein which is a component of

143

malate-aspartate shuttle (MAS) activity (Ramos et al. 2003) and expressed in neurons (del

Arco et al. 1998) and SH-SY5Y cells (Menga et al. 2015). SLC25A12-KO mouse models have been described to have deficits in motor coordination and a decrease in striatal dopamine levels at day 20 post-natal (Llorente-Folch et al. 2013).

144

Figure 4.16 - Gene expression and protein expression analysis for DEGs which encode mitochondrial proteins TOMM40 and SLC25A12. A) RNA-seq results showing Log2 fold change, fold change, q value and significance as well as number of different pipelines the DEG was identified in. Total, cytoplasmic and TRAP-RNA gene expression(±SD) analysis for B) TOMM40 and C) SLC25A12 using the ΔΔCt method with ACTB used as a housekeeping control (n=3). Western blot protein expression analysis for D) TOMM40 and E) SLC25A12 as well as F) their respective densonometric analysis with ACTB (n=3). One-way ANOVA, Tukey post-hoc test (n=4). *p<0.05.

145

qPCR gene expression analysis for TOMM40 showed a significant increase in the A53T overexpressing cells (1.96±0.34) compared to both control (1.00±0.38) and WT overexpressing cells (0.85±0.12) when total RNA was analysed (Figure 4.16B). However, both cytoplasmic and TRAP-derived RNA samples did not show any significant differences in

TOMM40 expression between the three groups (Figure 4.16B). Furthermore, protein expression analysis did not show any significant changes (Figure 4.16D, F). qPCR gene expression analysis for SLC25A12 showed a significant increase in A53T overexpressing cells (1.63±0.34) compared to control (1.00±0.06) and an increase nearing significance in WT overexpressing cells (1.48±0.06) compared to control cells when total RNA was analysed, replicating the RNA-seq results although with more modest fold changes

(Figure 4.16C). No significant changes were observed however in both cytoplasmic or TRAP-

RNA samples although TRAP-RNA gene expression analysis did show a non-significant increase in WT (1.22±0.35) and A53T (1.76±0.54) overexpressing cells compared to control

(1.00±0.07)(Figure 4.16C). Protein expression analysis showed no significant differences between samples (Figure 4.16E, F).

4.3.3.2.3 – Protein transport and dopamine signalling protein encoding genes

Impaired intracellular trafficking (reviewed in (Hunn et al. 2015) and described in section

1.4.6.5) as well as perturbations in dopamine signalling pathways have been implicated in the pathology of PD. Thus, one DEG chosen for validation was UNC119B, which was identified to be increased in A53T overexpressing cells compared to both control and WT overexpressing cells by 54.57 and 47.18-fold (Figure 4.17A), respectively. This gene encodes a cargo adaptor transport factor (lipid-binding chaperone B) protein whose homologue in C. elegans is associated with neuronal axonal branching (Knobel et al. 2001) and has been previously

146

identified as co-aggregating with poly-GA dipeptide repeats in a primary rat cortical neuron model of amyotrophic lateral sclerosis (May et al. 2014).

The other DEG chosen for validation was RGS8, which was significantly decreased in A53T overexpressing cells compared to WT overexpressing cells by 8.63-fold (Figure 4.17A). This gene encodes the regulator of G protein signalling protein which is expressed in the brain and densely so in Purkinje cells (Kuwata et al. 2008), and is thought to regulate potassium currents

(Saitoh et al. 1997) and signalling via the G-protein-coupled receptors including the dopamine receptor DRD2 (Werner et al. 1996). Previously, NGS analysis showed reduced expression of

RGS8 in a transgenic mouse model of Spinocerebellar ataxia type 2 neurodegeneration

(Dansithong et al. 2015) and was predicted to cause perturbation, specifically an increase, in intracellular Ca2+ (Marambaud et al. 2009). qPCR gene expression analysis for UNC119B showed a significant increase in A53T overexpressing cells (1.69±0.32) compared to control (1.00±0.21) in total RNA samples as well as a non-significant increase between A53T (1.43±0.22) and control (1.00±0.22) in TRAP-RNA samples, as observed in the RNA-seq results (Figure 4.17B). Protein expression analysis also showed non-significant increases in WT and A53T overexpressing cells compared to control

(Figure 4.17D, E). qPCR gene expression analysis for RGS8 showed a significant decrease in WT (0.58±0.18) and

A53T (0.40±0.00) overexpressing cells compared to control in total RNA samples (Figure

4.17C). However, cytoplasmic RNA samples showed a significant increase in WT overexpressing cells (1.81±0.38) compared to control (1.00±0.10) (Figure 4.17C). TRAP-RNA gene expression analysis showed non-significant decreases in both WT and A53T overexpressing cells compared to control (Figure 4.17C).

147

Figure 4.17 - Gene expression and protein expression analysis for DEGs which encode protein transport and dopamine signalling proteins UNC199B and RGS8. A) RNA-seq results showing Log2 fold change, fold change, q value and significance as well as number of different pipelines the DEG was identified in. Total, cytoplasmic and TRAP-RNA gene expression analysis (±SD) for B) UNC119B and C) RGS8 using the ΔΔCt method with ACTB used as a housekeeping control (n=3). D) Western blot protein expression analysis for UNC119B as well as E) respective densonometric analysis (n=3). One-way ANOVA, Tukey post-hoc test (n=4). *p<0.05.

148

4.3.3.2.4 – Oxidative stress response and apoptosis signalling encoding genes

Oxidative stress response (described in section 1.4.7.2) as well as apoptotic pathways have been implicated in cellular loss in PD. FOXO1 was identified to be significantly decreased in

WT cells compared to control by 59.95-fold (Figure 4.18A). FOXO1 encodes the forkhead box protein 1 which is a member of the FOXO transcription factors that are thought to promote expression of genes associated with autophagy and UPS (Webb et al. 2014). Autophagy genes are activated in neurons in response to FOXO1 nuclear localisation in mice (Xu et al. 2011).

Whole human genome microarray of prefrontal cortex tissue from PD patients showed FOXO1 was significantly increased in PD patients, compared to control (Dumitriu et al. 2012).

Another DEG chosen for further validation was CRADD which was found to be significantly higher in A53T compared to WT overexpressing cells by 12.13-fold (Figure 4.18A). CRADD encodes an adapter protein which contains a caspase recruiting domain and is a part of the caspase-2 activating complex (Ribe et al. 2012). Mutations in CRADD that result in loss of

CRADD/caspase-2 mediated apoptosis are associated with megalencephaly due to perturbations in developmental neuronal apoptosis (Di Donato et al. 2016). Overexpression of CRADD is associated with increased caspase-2 dependent apoptosis as well as formation of perinuclear aggregates (Jabado et al. 2004) and a recent study suggested a role for CRADD- targetting peptides in protecting against rotenone-induced neurotoxicity in SH-SY5Y cells with possible therapeutic use as a novel neuronal cell apoptosis inhibitors (Jang et al. 2016). qPCR gene expression analysis for FOXO1 showed a significant increase in A53T cells

(1.90±0.21) compared to WT overexpressing cells (0.76±0.05) and control (1.00±0.41) in total

RNA samples, however these were contradictory to the decreases identified from RNA-seq experiments (Figure 4.18B). TRAP-RNA gene expression analysis however showed no significant changes between conditions (Figure 4.18B).

149

qPCR gene expression analysis for CRADD showed very significant upregulation in A53T overexpressing cells (2.95±0.33) compared to both control (1.00±0.32) and WT overexpressing (0.91±0.05) in total RNA samples (Figure 4.18C). TRAP-RNA gene expression analysis and protein analysis showed a modest yet non-significant increase in A53T overexpressing cells (1.45±0.42) compared to control (1.00±0.16) and WT overexpressing cells (1.20±0.15) (Figure 4.18C). Furthermore, protein analysis also showed a similar, modest yet non-significant increase in A53T overexpressing cells (1.45±0.84) compared to control

(1.00±0.35) and WT overexpressing cells (1.17±0.54) (Figure 4.18D, E). This increase in CRADD replicates the increase in A53T overexpressing cells from RNA-seq analysis.

150

Figure 4.18 - Gene expression and protein expression analysis for DEGs which encode oxidative stress response and apoptotic proteins FOXO1 and CRADD. A) RNA-seq results showing Log2 fold change, fold change, q value and significance as well as number of different pipelines the DEG was identified in. Total and TRAP-RNA gene expression analysis (±SD) for B) FOXO1 and C) CRADD using the ΔΔCt method with ACTB used as a housekeeping control (n=3). Western blot protein expression analysis for D) CRADD as well as F) respective densonometric analysis (n=3). One-way ANOVA, Tukey post-hoc test (n=4). *p<0.05, ** p<0.01, *** p<0.001.

151

4.3.3.2.5 – microRNA biogenesis encoding gene microRNA (miRNA) dysregulation has been previously identified as associating with disease and many miRNAs have been found to be differentially expressed in neurodegenerative conditions including PD (described in chapter 5). A very important component of the miRNA biogenesis pathway is DGCR8. Deletion of DGCR8 in 22q11.2 deletion multisystem syndrome has been associated with schizophrenia (Merico et al. 2014) and early onset PD with mDA loss and LB formation (Butcher et al. 2017). Furthermore, a recent study using a neural stem cell line from DGCR8-KO mouse embryonic brains observed that the cells were unable to differentiate towards any neuronal phenotypes, with RNA-seq analysis identifying a perturbation in multiple cellular functions (Liu et al. 2017). DGCR8 was identified as being significantly downregulated in WT overexpressing cells by 9.58-fold compared to control as well as non-significantly lower in A53T overexpressing cells by 3.94-fold compared to control

(Figure 4.19A). qPCR gene expression analysis in total RNA samples showed a significant increase in A53T overexpressing cells (1.85±0.08) compared to both control (1.00±0.46) and WT overexpressing cells (0.74±0.07), which was contradictory to the RNA-seq data (Figure 4.19B).

However, no significant changes were detected in either cytoplasmic RNA samples and TRAP

RNA-samples (Figure 4.19B). Furthermore, protein expression analysis showed a modest decrease in both WT (0.88±0.57) and A53T (0.52±0.41) overexpressing cells compared to control (1.00±0.04), although these were non-significant (Figure 4.19C, D).

152

Figure 4.19 – Gene expression and protein expression analysis for DGCR8 which encode a protein involved in miRNA biogenesis. A) RNA-seq results table showing Log2 fold change, fold change, q value and significance as well as number of different pipelines the DEG was identified in. B) Total, cytoplasmic and TRAP-RNA gene expression analysis (±SD) for DGCR8 using the ΔΔCt method with ACTB used as a housekeeping control (n=3). C) Western blot protein expression analysis for DGCR8 as well as D) respective densonometric analysis (n=3). One-way ANOVA, Tukey post-hoc test (n=4). *p<0.05, ** p<0.01.

4.3.3.2.6 – Conclusions from validation studies

RNA-seq validation produced some replication of the results, although the majority of analyses were inconclusive or conflicting. For instance, the increase in FGD3 expression in WT and A53T overexpressing cells from RNA-seq analysis was only partially validated in total RNA samples, with an increase in expression in A53T but not WT overexpressing cells, however, an

153

opposite result was observed in TRAP-RNA, with a decrease in expression in A53T overexpressing cells compared to control.

Overall, the observed changes in TOMM40, SLC25A12, UNC119B and CRADD genes in the

RNA-seq experiments were only partially validated in total RNA samples, but not in qPCR analyses of cytoplasmic or TRAP RNA, nor protein expression analyses. In the TRAP-RNA samples, SLC25A12, UNC119B and CRADD showed a trend towards significance in replicating the observed RNA-seq data. In the case of RGS8, qPCR analyses of total and cytoplasmic RNA partially replicated the RNA-seq results and there were also non-significant trends in the same direction using TRAP-RNA.

Gene expression analysis for FOXO1 showed contradictory results in comparison to RNA-seq data in both total and TRAP-RNA whilst the observed changes in DGCR8 in RNA-seq experiments could be not validated in any of the RNA or protein samples.

4.4 Discussion

In this chapter, RNA sequencing was undertaken to identify mRNAs that may be differentially expressed when wild-type (WT) or mutant (A53T) a-synuclein is overexpressed in SH-SY5Y cells, using the TRAP technique to isolate putative mRNAs that were being actively translated in translating ribosome complexes. The lists of DEGs were analysed bioinformatically and a selection of genes were validated by qPCR and Western blot analyses.

During the bioinformatic analysis, raw reads from the sequencing were validated using the

FastQC module. As mentioned, several flags were identified, most notably duplicated sequences and overrepresented sequences. These sequences were mostly rRNA sequences, suggesting some over-amplification of rRNA sequences common to translatome preparations.

154

Duplication is also a common artefact of starting NGS experiments using low input translatome samples, with lower duplication indicating a higher complexity of the samples

(Song et al. 2018). Notably, high fold changes (up to Log2x15 i.e. >30000 fold change) were observed in all three pipelines and furthermore, the different pipelines showed only partial overlap of DEGs between the different groups. This raised some concern regarding the quality of the RNA-seq analysis however, research has suggested that ribosomal preparations, as compared to total RNA NGS experiments, are likely to create decreased diversity and complexity of transcripts as pre-mature mRNAs are eliminated (Zhang et al. 2015). Whilst these flags suggested that there may be some compromise to the raw data quality, power and false discovery rate (FDR) a recent study has suggested that PCR duplication from pre- amplification and amplicon amplification, paired with single-end reads produces up to 66% of reads are PCR-duplicates (Parekh et al. 2016).

Another possible explanation as to why the duplication and overrepresented sequences flags were raised is the fact that pre-amplification was required due to the low starting yield of

RNA from TRAP. Amplified low input RNA-seq is thought to further decrease the overall complexity of library preparations. High duplication rates have been reported, across all amplification platforms tested, in recent studies comparing commercial RNA amplification kits. The SeqPlex kit that was used in this experiment has been shown to generate relatively high PCR duplicates and rRNA duplication (Shanker et al. 2015). However, according to the manufacturer’s literature, the SeqPlex amplification kit amplifies rRNA less efficiently than mRNA. Furthermore, SeqPlex performed well in rRNA depletion, read alignments as well as other measures such as coverage in a comparative study of RNA amplification kits (Mao et al.

2014). Thus, whilst the choice of amplification kit has been validated as being adequate,

155

duplications and rRNA contamination is an expected and common result of both translatome profiling and the amplification processes required for taking such low yield RNA samples towards NGS sequencing. A recent study comparing pre-amplification kits suggested that biases arising from RNA amplification protocols can easily confuse or mask true biological effects (Mao et al. 2014).

Differential gene expression analysis was performed using slight modifications on the Tuxedo

RNA-seq analysis pipeline, which is a commonly used, well characterised pipeline. The Tuxedo pipeline uses the most well annotated reference transcript assembly (Gencode) and this generated a list of DEGs for further validation. Differential expression analysis using this pipeline showed that α-synuclein perturbations resulted in the up/down-regulation of many genes: 197 DEGs for Control versus WT cells and 360 DEGs for WT versus A53T cells.

Unfortunately, as mentioned in this chapter, RNA-seq analysis also detected an increase in expression of A53T mutant α-synuclein compared to expression of WT α-synuclein in the WT overexpression model which may confound the analysis.

Analysis of the reproducibility of DEGs between pipelines showed only some common DEGs between all 3 pipelines employed. Overall 33.5% of genes were detected in all 3 pipelines for

Control compared to WT analysis and 32.9% of genes were detected by all 3 pipelines in WT and A53T analysis. The least reproducibility was observed when using both a different aligner and a different reference transcript annotation dataset together. Furthermore, there are many more factors which can affect the reproducibility of analysis which were not tested.

Increasing the number of biological repeats has been identified as being important to improving the robustness of DEG analysis. A recent study comparing two conditions has shown that, with the use of only 3 biological replicates, only 20-40% of the total DEGs

156

identified by the total 48 biological replicates used in the study were detected (using 11 different pipelines) of (Schurch et al. 2016). To reproduce >85% of DEGs determined by 48 replicates required over 20 biological replicates. Thus, for increased DEG detection sensitivity, further biological repeats would be advantageous.

Enrichment analysis (or functional enrichment analysis) was performed on statistically significantly differentially expressed gene sets which showed >2 fold differential expression.

This is a statistical method used to identify classes or pathways of genes which may be functionally related with disease phenotypes. KEGG pathway enrichment for control compared to WT synuclein overexpressing cells showed only one statistically significant enriched pathway, Lysine degradation. GO cellular pathway analysis showed significant enrichment of actin filament branch point genes/proteins, whilst GO biological function pathway analysis showed significant enrichment of regulation of transcription, dendrite extension and cellular response to UV-A pathways. Thus, enrichment analysis suggests a role for the lysine amino acid degradation process as well as actin filament and dendrite extension pathways, which are known to be perturbed in PD. As noted in section 3.2.3, neurite dynamics were determined to be perturbed in iPSC-derived mDA cultures. These results suggest that formation of neurites may play a part in α-synuclein-associated pathology, as has been previously described (Liu et al. 2013; Cartelli et al. 2016; Takenouchi et al. 2001).

KEGG pathways enrichment analysis using genes differentially expressed in WT versus A53T overexpressing cells showed one statistically significant enriched pathway, the extracellular matrix-receptor interaction pathway. Furthermore, GO cellular pathway analysis showed significant enrichment of perinuclear endoplasmic reticulum lumen (where proteins are synthesised and collected) and actin body pathways, whilst GO biological function pathway

157

analysis showed significant enrichment of biofilm matrix organisation, paranodal junction assembly and ceramide metabolic processes. Thus, enrichment analysis for this analysis suggests a role for extracellular matrix related pathways as well as protein synthesis, actin body and assembly of the paranode junction (where myelinating glia meets neuronal axons).

Validation was performed on some interesting targets that may be involved in variable pathological pathways, using qPCR transcript and Western blotting protein analysis techniques. Overall, these analyses were unable to replicate the high fold changes seen in

RNA-seq analysis. An attempt was made to analyse the amplified cDNA product, prior to adapter ligation and further amplification, to attempt to identify whether this discrepancy arose due to amplification, but did not produce meaningful results (data not shown; very high error bars), most likely owing to the fragmented nature of the amplification product.

Interestingly, the expression of α-synuclein at both the protein and transcript level was increased significantly using qPCR and Western blotting and this correlated well to RNA-seq analysis.

4.5 - Conclusions

The results presented suggest that adequate control, such as control of fragment size, amplicon concentration, primer-dimer removal, etc., was taken to produce the pooled library for sequencing. This resulted in the generation of raw data with relatively little differences in number of read, read quality control checks, and alignment statistics between samples.

However, there may have been some overamplification associated with our library preparation protocol.

RNA-seq DEG analysis showed differential expression of genes that may be related to α- synuclein associated pathology. Furthermore, enrichment analysis showed

158

overrepresentation of several molecular, cellular and pathological pathways. Although some of the RNAseq data could not be reproducibly validated using gene expression or protein expression techniques, these could be studied in a more physiologically relevant model such as iPSC-derived neurons to confirm these results. Further analysis of the RNAseq data may also reveal other DEGs that warrant further investigation.

159

Chapter 5 - miRNA profiling and differential expression analysis in iPSC-derived neurons 5.1 Introduction

5.1.1 miRNA biogenesis and function in mDA neurons miRNA are a class of small RNAs, originally discovered in C. elegans, that have been shown to be involved in a myriad of cellular regulatory mechanisms (Lee et al. 1993). miRNA biogenesis is initiated by RNA polymerase II mediated transcription to produce primary miRNAs that are processed by a miRNA processing complex containing the RNase III enzyme Drosha and

DGCR8 into pre-miRNA that is subsequently exported from the nucleus by Exportin-5 (Yi et al.

2003)(Figure 5.1). In the cytoplasm, pre-miRNAs are processed into ~22nt double stranded miRNAs by the RNase III enzyme Dicer (Bernstein et al. 2001). Strands unwind to produce one strand which functions as a mature miRNA by incorporating target mRNA into a ribonucleoprotein complex known as the RNA-induced silencing complex (RISC). The mature miRNA strand is usually associated with Argonaute 2 (Ago2) protein which is involved in target mRNA cleavage and degradation (Okamura et al. 2004; Ender et al. 2010). Mature miRNAs can downregulate target gene activity by binding to target sequences in 3’ untranslated region (UTR) or open reading frame (ORF) region of mRNAs, leading to inhibition of translation or initiating the cleavage and degradation of target mRNA (Figure 5.1). Inhibition or cleavage mechanisms are alternatively employed based on the degree of complementarity between the miRNA and target mRNA. A single miRNA may target multiple, possibly hundreds, of targets simultaneously and may act together synergistically to inhibit translation (Selbach et al. 2008). Likewise, protein encoding mRNA may have binding sites for several miRNAs.

Furthermore, several miRNA have been observed to be contained together within clusters on single polycistronic transcripts (Tanzer et al. 2004).

160

Figure 5.1 – The miRNA processing pathway. Primary miRNA are transcribed by RNA polymerase II and then cleaved by the Drosha-DGCR8 complex to generate a hairpin pre-miRNA. Pre-miRNA is exported from the nucleus by Exportin-5 to the cytoplasm where the RNAse Dicer cleaves pre-miRNA to mature miRNA duplex. The passenger strand is degraded and the functional miRNA strand can then be incorporated into the RNA-induced silencing complex (RISC) with Ago2 to target mRNA for translation repression or cleavage (Winter et al. 2009).

The first reported association between miRNA regulation and mDA neuron function was described in a study in which a Cre-mediated deletion of Dicer in mDA neurons derived from mouse ESCs was generated (Kim et al. 2007). This deletion resulted in a near complete loss of dopaminergic neurons, due in part to increased apoptosis and reduced neurogenesis. These observations were replicated in vivo in a mouse model harbouring an mDA-specific deletion of Dicer, with behavioural studies revealing reduced locomotion in an open-field assay.

Further studies in murine Dicer mDA-specific deletion models have observed additional evidence of dysregulation of mDA function with mDA neuron loss and behavioural deficits

(Cuellar et al. 2008; Pang et al. 2014). These results suggested that miRNA regulatory proteins are essential for the terminal differentiation and/or maintenance of mDA neurons.

Other miRNA biogenesis proteins have also been associated with perturbed mDA function.

Chromosome 22q11.2 deletion syndrome is a result of a chromosomal deletion of DGCR8 and

161

post-mortem analysis of patients with this syndrome exhibit mDA neuron loss (Butcher et al.

2013). The syndrome has also been found to result in reduced motor function and parkinsonian symptoms (Roizen et al. 2010).

Currently, studies have linked miRNAs to diverse PD-associated pathways, with the discovery of miRNAs associated with α-synuclein regulation (Junn, Lee, et al. 2009; Doxakis 2010), midbrain dopaminergic neuron differentiation (Yang et al. 2012), neuroinflammation (Thome et al. 2016) and mitochondrial function (Minones-Moyano 2011). Furthermore, differential miRNA expression has been observed in PD patients (Miñones-Moyano et al. 2011; Hoss et al. 2016) and various α-synuclein PD models (Kong et al. 2015; Thome et al. 2016). Specific mechanisms of potential miRNA-associated PD pathology are discussed in detail in the next section (5.1.2).

5.1.2 miRNAs associated with PD mechanisms

5.1.2.1 miRNA regulation of familial PD-associated genes miR-7 is mainly expressed in neurons and was the first miRNA shown to bind to the 3’UTR of

α-synuclein mRNA inducing a downregulation of α-synuclein protein (Junn, Lee, et al. 2009).

This observation was replicated in another study that showed that overexpression of miR-7 significantly reduced α-synuclein mRNA and protein levels in murine cortical neurons (Doxakis

2010). Furthermore, miR-7 has been shown to be reduced in PD patient brains and loss of miR-7 leads to accumulation of α-synuclein in an in vivo mouse model (McMillan et al. 2017).

This miRNA has also been shown to inhibit neuronal apoptosis in SH-SY5Y PD models (Li et al.

2016) and can target the Nod-like receptor protein 3 inflammasome and inhibit microglial activation (Zhou et al. 2016).

162

miR34b and miR34c are located on chromosome 11q23 and transcribed as a single non-coding precursor. In a microarray screen, these two mature miRNAs were both found to be depleted in SNpc brain tissue of PD patients compared to control brains (Miñones-Moyano et al. 2011).

Although the authors noted that this could be as a result of a loss of dopaminergic neurons in the PD patients, a reduction was also noted in frontal cortex and cerebellum, where little degeneration had occurred. This was coupled with a reduction in Parkin and DJ1 protein, two proteins associated with familial PD, as well as mitochondrial function and oxidative stress response pathways, respectively. Furthermore, more recent studies have also shown that inhibition of these miRNAs in SH-SY5Y cells can result in increased α-synuclein expression and that these miRNA can directly target the 3’UTR of α-synuclein mRNA (Kabaria et al. 2015).

Direct binding of miRNA has been described for other PD-associated protein mRNAs. Loss of

DJ-1 protein function is associated with autosomal recessive familial PD and miR-494 has been observed to directly bind to DJ-1 mRNA. miR-494 is highly expressed in mouse brain tissue and overexpression has been noted to repress expression of DJ1 protein in 3T3 cells, making them more susceptible to oxidative stress (Xiong et al. 2014). Furthermore, LRRK2 is upregulated in PD patient brains compared to controls and miR-205 was determined to regulate LRRK2 expression by direct targeting of its mRNA in primary neuronal cultures, with its overexpression downregulating LRRK2 protein expression (Cho et al. 2012). Moreover, miRNAs have been identified that may act on pathways that affect α-synuclein aggregation, for example miR-16-1 which is thought to target heat shock protein 70, a protein that is known to negatively regulate α-synuclein aggregation (Zhang and Cheng 2014).

163

5.1.2.2 miRNAs in mDA development and function

The discovery of differential expression of miR133b in PD patient brain samples was coupled with a hypothetical model whereby miR-133b expression functions within a negative feedback circuit that suppresses paired-like homeodomain transcription factor 3 Pitx3 expression and regulates mDA neuron differentiation or maintenance (Kim et al. 2007).

Exosome-mediated miR-133b exchange, from rat mesenchymal stromal cells to primary cortical neurons, has also been reported to contribute to the regulation of neurite outgrowth

(Xin et al. 2012). Furthermore, levels of miR-133b are reportedly reduced in the plasma of PD patients compared to control (Zhang, Yang, Hu, et al. 2017). miRNA expression profiling of brainstem samples from A30P transgenic mice identified miRNAs that were differentially expressed compared to control mice, including miR-132 which was significantly downregulated in the transgenic mice (Gillardon et al. 2008). This miRNA was identified as potentially being involved in mDA function and has been shown to be enriched in neurons, with overexpression of this miRNA in rat primary cortical neurons able to induce neurite outgrowth and conversely inhibition attenuated outgrowth by regulation of levels of GTPase-activating protein, p250GAP (Vo et al. 2005). Another study conducted a miRNA array screen in mouse ESCs (mESCs), differentiated towards mDA neurons, and showed a significant increase in miR-132 expression during later stages of differentiation compared to progenitor stage mESCs. It was further shown that miR-132 is able to directly regulate Nurr1 expression, an important transcription factor during dopamine neuron development. Thus, inhibition of miR-132 was associated with a promotion of mDA differentiation, as assessed by number of TH positive cells (Yang et al. 2012).

164

Furthermore, miRNA sequencing in A30P α-synuclein Drosophila overexpression models identified another miRNA, miR-137, as being overexpressed in α-synuclein overexpression

Drosophila models. This miRNA was shown to affect expression of several neuroactive-ligand receptors transcriptional targets including the dopamine receptor, which was downregulated in α-synuclein overexpression Drosophila models, and thus perturbing mDA function (Kong et al. 2015).

5.1.2.3 miRNAs associated with neuroinflammation

PD is characterised by chronic neuroinflammation which exacerbates mDA neuron degeneration. Using an array-based approach to identify miRNAs associated with neuroinflammation in an AAV2-α-synuclein overexpression mouse model of PD, enhanced expression of miR-155 was detected as early as 2 weeks post transfection of AAV2-α-synuclein in SNpc tissue (Thome et al. 2016). Furthermore, a mouse model with complete deletion of miR-155 was found to have a reduced inflammatory response to AAV2-induced α-synuclein overexpression (Thome et al. 2016). miR-155 has further been shown to be released in exosomes from immune cells, namely bone marrow-derived dendritic cells, and taken up by recipient dendritic cells, in vivo and in vitro, inducing an upregulation of expression of pro- inflammatory genes in response to activating lipopolysaccharide treatment (Alexander et al.

2015). Furthermore, miR-155 has also been shown to be upregulated in the hippocampus

(Arena et al. 2017) and cortex (Guedes et al. 2014) of mouse models of Alzheimer’s disease corresponding with an increase in microglial and astrocytic activation. These observations link miR-155 with the pathophysiology of neurodegeneration through neuroinflammatory pathways.

165

miR-21 was originally shown to be upregulated in tumours and has been shown to be increased in the cortices of a rat stroke model (Buller et al. 2010), acting to reduce apoptosis in cortical neurons of mice (Han et al. 2014; Huang et al. 2016). However, in the context of

PD, miR-21 overexpression in an MPP+ SH-SY5Y model resulted in significant reduction of lysosome-associated membrane protein 2 (LAMP2A) mRNA and protein levels and markedly increased levels of α-synuclein (Su et al. 2016). LAMP2A is the rate-limiting step in the chaperone-mediated autophagy pathway, known to be involved in the clearance of α- synuclein (Vogiatzi et al. 2008). Furthermore, miR-21 has been reported to be secreted via exosomes from lung tumour cells and bind as ligands to Toll-like receptor (TLR) in immune cells to induce a prometastatic inflammatory response (Fabbri et al. 2012). miR-21 has also been observed to bind to TLRs in neurons, reducing hippocampal axon and dendrite growth

(Liu et al. 2015). As PD is characterised by neuroinflammation, the prospect that these miRNAs may be involved in immune activation that is manifested prior to motor symptoms caused by degeneration is intriguing.

Collectively, these observations indicate that miRNA dysregulation is involved in mDA development and function, inflammation, neurite outgrowth processes as well as regulating several proteins involved in the pathology of PD including α-synuclein itself. They also suggest that α-synuclein perturbations could induce dysregulation of specific miRNAs. These studies rely mostly on animal models, cell lines such as HEK and SH-SY5Y cells, or human tissue or plasma/serum. In this chapter we will study miRNA dysregulation in hiPSC-derived neuronal cultures that overexpressed WT, A53T and G51D α-synuclein.

166

5.1.3 Extracellular miRNA

While the majority of miRNAs are found intracellularly, extracellular miRNAs have been identified in a variety of extracellular fluids such as serum, plasma, saliva and urine (Weber et al. 2010). Moreover, expression profiles of extracellular miRNA have been linked to disease states and as such they have been studied for their potential use as novel clinical biomarkers

(Alevizos et al. 2010), including in PD (Khoo et al. 2012; Cao et al. 2017). Extracellular miRNAs appear to be remarkably stable and protected from degradation by endogeneous RNase activity. Mechanisms that are thought to protect and stabilise extracellular miRNA include loading into high density lipoprotein (Tabet et al. 2014; Vickers et al. 2011), binding to Ago2 protein (Arroyo et al. 2011; Turchinovich et al. 2011) and packaging into microvesicles (MVs) such as exosomes (Valadi et al. 2007).

The majority of extracellular miRNA have been reported to be concentrated in either exosomes (Gallo et al. 2012) or associated with Ago2 protein (Turchinovich et al. 2012).

Although Ago2 has been identified within MVs (Collino et al. 2010), Ago2-associated miRNAs outside of vesicles are thought to be released non-specifically, as a result of cell death

(Turchinovich et al. 2012). Exosome-bound miRNA however are thought to be sorted specifically into exosomes and have been reported to be endocytosed by recipient cells, transferring miRNA and inducing target gene transcriptional regulation (Sohel et al. 2013;

Simeoli et al. 2017). Thus, these mechanisms protect and stabilise miRNA in bodily fluids, enabling the use of these novel extracellular miRNA as biomarkers of disease (Table 5.1).

167

Differentially Model expressed Assay type Normalisation Reference miRNA 85 miRNA PD patients miR-1, miR-22*, 2-ΔCt and RNA (Margis et al. Taqman qPCR (blood) miR-29 concentration 2011) array 384 miRNA PD patients (Cardo et al. miR-331-5p TaqMan qPCR miR-191 (plasma) 2013) array PD patients 17 in serum, 5 in Small RNA (Burgos et al. Bioinformatics (Serum/CSF) CSF sequencing 2014) 27 miRNA: miR-1, PD patients (CSF 746 Taqman RNU6B and 19b, 153, 409, (Gui et al. 2015) exosomes) qPCR array RNU44 10a and let-7g miR-195, miR- Small RNA PD patients 185, miR-15b, Bioinformatics sequencing and (Ding et al. 2016) (serum) miR-221, miR- and Let-7d/g/i RT-qPCR 181a A53T mouse miR-144, miR- Small RNA Bioinformatics (Mo et al. 2017) model (CSF) 200a, miR-542 sequencing PD patients 24 miRNA miR19b, miR23 let-7d/g/I and (blood Taqman qPCR (Cao et al. 2017) and miR195 cel-miR39-3p exosomes) array Table 5.1 - Studies that have assessed differentially expressed extracellular miRNA profiles for PD. Including differentially expressed miRNA, miRNA detection, normalisation and source. Cerebrospinal fluid (CSF).

5.1.3.1 Exosomes

Exosomes are lipid raft enriched extracellular 30-120nm vesicles released from cells by exocytosis upon fusion of multivesicular bodies with the plasma membrane (Mathivanan et al. 2010) (Figure 5.2). These vesicles can be secreted by most cell types and are present in most biological fluids. They have been found to contain various proteins and nucleic acids such as mRNAs and miRNAs. It has been widely reported that α-synuclein can be secreted and transmitted in a cell-to-cell manner leading to the spread of the neurodegenerative phenotype. Thus, it is not surprising that α-synuclein protein has been identified within extracellular exosomes, and cell-to-cell transmission of α-synuclein oligomers by this mechanism has been demonstrated to propagate α-synuclein pathology (Danzer et al. 2012;

Shi et al. 2014; Ngolab et al. 2017)Shi 2014; Ngolab 2017). Exosome-mediated transfer of RNA species, including miRNA, has been reported and the ability of these extracellular vesicles to

168

deliver RNA cargo to recipient cells, to effect transcriptional regulation in recipient cells, has been demonstrated (Valadi et al. 2007; Hunter et al. 2008; Skog et al. 2008) (Figure 5.2A). In fact, pluripotent stem cell-derived exosomes have been shown to induce a protective effect on cardiac insult models through miRNA-associated mechanisms and are being developed as an adjuvant therapy for myocardial infarction (Arslan et al. 2013; Khan et al. 2015).

169

Figure 5.2 miRNA biogenesis and release via exosomes. A) miRNA biogenesis and extracellular release of miRNA. After transcription by RNA pol II and processing by Drosha, pre-miRNA are exported to from the nucleus to cytoplasm, where they are processed by Dicer into mature miRNAs. Mature miRNA can be bound to Ago2, actively sorted into exosomes or associated with high-density lipoprotein prior to extracellular release. Figure taken from (Sohel 2016). B) Exosomes are contained within multivesicular bodies, released upon fusion with the cell membrane. Exosomes carry diverse cargo which includes various proteins and nucleic acids. The bi-lipid membrane is composed of tetraspanin (CD9, CD63, CD81) transmembrane proteins. Figure taken from (Jung et al. 2017).

As mentioned, it has been suggested that specific endogenous miRNAs are selectively exported within extracellular vesicles. Comparing miRNA profiles from parent cells and their vesicles demonstrated that certain miRNA are preferentially exported and enriched in

170

exosomes compared to parent cells (Guduric-Fuchs et al. 2012). Furthermore, individual exosomal miRNAs have been shown to be differentially expressed in various physiological states, for example exosomes derived from PD patient cerebro-spinal fluid (CSF) compared to healthy control CSF-derived exosomes (Gui et al. 2015). Furthermore, an active sorting mechanism for miRNA into exosomes has been proposed, though this remains to be fully elucidated (Villarroya-Beltri et al. 2013; Kosaka et al. 2013; Guduric-Fuchs et al. 2012).

Studies that have attempted to identify novel extracellular miRNA biomarkers have used different sources, assay types and normalisation techniques, generating variable results

(Table 5.1). Many of these studies have used patient blood, CSF, plasma, serum. Moreover, there is no consensus on how to normalise extracellular miRNA levels, although one study identified a combination of Let-7d, Let-7g and Let-7i (Let-7d/g/i) as being stably expressed in human serum (Chen, Liang, et al. 2013). Normalisation is especially challenging in the case of exosomes, as miRNA are thought to be sorted specifically, but may be more relevant to cell- to-cell signalling processes than non-exosomal miRNA. Thus, it is important when studying extracellular, and specifically exosomal miRNA, to have a robust normalisation strategy.

5.1.4 - Aims miRNA have been shown to regulate several biological pathways associated with mDA function and disease progression. The results of the RNA-seq analysis, described in Chapter

4, showed differential expression of a key gene involved in the biogenesis of miRNA. The aim of the experiments described in this chapter are to:

1) employ iPSC-derived mDA neurons, overexpressing wild-type and mutants A53T and

G51D α-synuclein, to determine intra- and extracellular exosomal miRNA profiles in

response to perturbed α-synuclein expression

171

2) isolate and characterise extracellular exosomes by transmission electron microscopy

(TEM) and nanoparticle tracking analysis (NTA) to determine whether perturbed α-

synuclein expression in these cultures resulted in a change in exosome release

dynamics, and to aid in normalisation of exosomal miRNA expression.

Thus, the identification of dynamics of miRNA expression, in a physiologically relevant human model such as iPSC-derived mDA neurons, may help in the understanding of the role of miRNA in PD disease progression and the identification of novel disease biomarkers.

A schematic for the generation of iPSC-derived mDA neurons overexpressing α-synuclein, characterisation and timeline of characterisation of miRNA expression profiles is shown in

Figure 5.3.

Figure 5.3 - Schematic of the timeline of generation of iPSC-derived mDA neurons and characterisation of transduction and differentiation. Characterisation of exosomes by nanoparticle tracking analysis (NTA) and intracellular and extracellular miRNA expression profiling.

172

5.2 Materials and methods

5.2.1 Isolation of exosomes from iPSC-derived neuronal cultures

NAS2 iPSC lines were cultured and subjected to the mDA differentiation protocol described in section 2.2.4. Conditioned media from these cultures was collected, rather than aspirated, for subsequent exosome isolation and characterisation (day 32), and miRNA isolation (day

35), and stored at -80°C. To isolate exosomes, a rapid purification kit was employed, the miRCURY Exosome Kit (Exiqon; #300102) on 1ml of conditioned media as per manufacturer’s instructions. This protocol uses a buffer which precipitates exosomes by diminishing the hydration of subcellular particles, which can then be collected by centrifugation. Briefly, 1.1ml of conditioned media from both day 32 or day 35 differentiated iPSC-derived PD models was spun down at 10,000 x g, to remove cellular debris, and 1.0ml of supernatant transferred into a new centrifuge tube. 400μl Precipitation Buffer B was then added and mixed by vortexing and tube incubated for 60 minutes at 4˚C. The mixture was then centrifuged at 10,000 x g for

30 minutes at room temperature, to collect exosomes, which were resuspended in 100μl of

PBS. The protocol was originally performed on a control non-transduced day 30 differentiated iPSC-derived mDA culture sample of conditioned media, to test the ability of this protocol to isolate exosomes by TEM characterisation. The supernatant from iPSC-derived PD models

(described in section 2.2.4) with isolated exosomes were either lysed (day 35) by the mirVana kit (as per section 5.2.4) and prepared for miRNA characterisation or resuspended in PBS (Day

32) for NTA analysis (section 5.2.5). Furthermore, day 30 samples from non-transduced control cells was used for TEM characterisation and stored at -20°C.

173

5.2.2 Characterisation of isolated exosomes by Transmission Electron Microscopy

2µl of exosomes in PBS was transferred onto a formvar-carbon coated 300 mesh electron microscopy grids (Agar Scientific) and left for 2 minutes. Once specimen had adsorbed onto formvar-carbon coated grid, a 100µL drop of H2O was placed on a metal plate covered with parafilm on ice, and grid was washed by transfer of grid with the sample membrane side facing down using clean forceps for 20 seconds. Sample on grid was transferred to a 20 µL drop of a 1:10 ratio of a mixture of 3% uranyl acetate and 1.8% methyl cellulose, respectively, for 5 minutes. The grids were removed with stainless steel loops and excess fluid dragged gently, at a 90° angle, on Whatman no.1 filter paper. Grids were left to dry and stored in grid storage boxes.

Grid-mounted exosomes, sample side facing down, were incubated in 100μl of 0.1% acetylated BSA (Aurion) in PBS for 10 minutes on parafilm. Grids were then incubated in primary antibody for CD63 and CD81 in 0.1% BSA in PBS for 1 hour at room temperature and then washed in PBS for 2 minutes, 5 times. Grids were then incubated in a 1:20 dilution of

10nm secondary gold antibody (Aurion) in 100μl 0.1% BSA in PBS. Grids were finally washed in PBS for 2 minutes, 5 times and then sample on grid was transferred to a 20 µL drop of a

1:10 ratio of a mixture of 3% uranyl acetate and 1.8% methyl cellulose, respectively, for 5 minutes. The grids were removed with stainless steel loops and excess fluid dragged gently, at a 90° angle, on Whatman no.1 filter paper. Grids were left to dry and stored in grid storage boxes. Grids were observed with a TEM (Tecnai 12 - FEI 120kV BioTwin Spirit) equipped with an

FEI Eagle 4k x4k CCD camera. Images were analysed for the presence of lipid membranes of

20-120nm by negative staining. (Exosome preparation and staining was performed by student

Bryony McPherson).

174

5.2.3 Isolation of intracellular and exosomal miRNA

The isolation of small RNA molecules from cell cultures and isolated exosomes, in this study, was performed using mirVana™ miRNA Isolation Kit (ThermoFisher; #AM1561) with acid- phenol (ThermoFisher; #AM9720), which is specifically designed for the purification of miRNA in natural populations. Briefly, 1ml of conditioned media from differentiating iPSCs was subjected to the exosome isolation as described in section 5.2.1. Once exosomes had been pelleted samples were disrupted in a denaturing lysis buffer. At this stage, 2 fmole of a synthetic miRNA found only in Caenorhabditis elegans, cel-miR-39-3p (Exiqon; #203203), was spiked into lysis buffer to act as a technical control. Next, samples are subjected to Acid-

Phenol:Chloroform extraction, as per manufacturer’s instructions. Total RNA was eluted in

50µl nuclease-free H2O and quantified using the Nanodrop 2000 UV-Vis spectrophotometer

(ThermoFisher).

5.2.4 cDNA generation and qPCR expression profiling by Taqman small RNA assays for intracellular and extracellular miRNA

TaqMan Small RNA Assays were employed to detect and quantify mature microRNAs

(miRNAs) in a two-step PCR reaction. Firstly, cDNA from 2ng of total RNA is reverse transcribed (RT) using a small RNA-specific, stem-loop RT primer from the TaqMan Small RNA

Assays (Applied Biosystems; #4427975) and reagents from the TaqMan MicroRNA Reverse

Transcription Kit (Applied Biosystems; #4366596). In the PCR step, PCR products are amplified from cDNA using the TaqMan Small RNA Assay together with the Taqman Universal PCR

Master Mix II (Applied Biosystems; #4440040).

In the first instance, qPCR was performed on two prospective small RNA normalisation genes,

RNU6B and RNU24, and analysed using the Bestkeeper algorithm as described in section

175

4.2.5. miRNA small RNA assays for use in this study are shown in Table 5.2. ΔΔCt method used relative to RNU24 for analysis and one-way ANOVA with Tukey’s post-hoc test used to test significance.

Catalogue miRNA of Catalogue Control miRNAs Number interest Number RNU6B 001093 7 000386 RNU24 001001 21 000397 cel-miR-39-3p 000200 34b 000427 34c 000428 132 000457 133b 002247 137 000593 155 002623 494 002365

Table 5.2 - List of control miRNAs and miRNAs of interest with roles associated with PD, development of mDA neurons and neuroinflammation

5.2.5 Nanoparticle tracking analysis (NTA) for determining isolated exosomes from iPSC-derived PD models

The NanoSight NS300 NTA (Malvern) uses laser of particles in suspension samples, passing through the laser at a constant rate. A video is taken and scattered light from particles is analysed for concentration and size based on Brownian motion, using the Stokes-Einstein equation. NTA concentration determination from exosomal samples isolated from day 32 differentiated iPSC-derived neurons was performed by a collaborator Dr. Andrew Shearn.

176

5.3 Results

5.3.1 Differential miRNA expression in iPSC-derived PD models

5.3.1.1 – Selection of miRNA for profiling

Target Mechanism/Function Model Publications

HEK and rat/murine mir-7 Overexpression of miR-7 suppresses α- (Doxakis 2010) cortical neurons Direct binding to α- synuclein expression at protein and (Junn, Lee, et synuclein mRNA mRNA levels HEK cells al. 2009) (Miñones- Decreased miR-34b/c in PD brain even at miR-34b/c PD brain samples Moyano et al. premotor stages Direct binding to α- 2011) synuclein mRNA Inhibition of miR-34b/c increased α- Human dopaminergic (Kabaria et al. synuclein levels SH-SY5Y cells 2015) miR-494 Overexpression of miR-494 decreased 3T3 and MPTP mouse (Xiong et al. Direct binding to DJ1 DJ1 on a protein level models 2014) mRNA Reduced in SNpc of PD patients. PD patients miR-133b (Kim et al. Overexpression reduces expression of Overexpression: Targets Pitx3 & RhoA 2007) dopamine transporter mouse primary mDA (mDA development Exosome mediated transfer of mir-133b & regulation of Rat MSCs to primary (Xin et al. from MSCs to primary cortical cultures neurite outgrowth) rat cortical cultures 2012) increased neurite outgrowth miR-132 α-synuclein (A30P)- (Gillardon et Increased in A30P α- synuclein Direct binding of transgenic al. 2008; Vo et transgenic mice (neurite outgrowth) Nurr1 mRNA mice al. 2005) (regulation of Increased during mESC mDA development and mDA differentiated (Yang et al. differentiation. Downregulation of miR- neurite outgrowth) mESCs 2012) 132 promoted mDA differentiation miR-137 Increased in PD α-synuclein (Kong et al. (Dopamine receptor overexpressing flies. Brain enriched and Drosophila model 2015) function) highly conserved. Mouse AAV2-α- miR-155 Overexpression of α-synuclein induces (Thome et al. synuclein & miR 155-/- (Inflammation) expression miR-155 in mouse model 2016) transgenic mice miR-21 Direct binding to Increased miR-21 reduced levels of Mouse models or SH- LAMP2A mRNA LAMP2A protein and mRNA levels and (Su et al. 2016) SY5Y cells (protein degradation increased α-synuclein protein and inflammation) Table 5.3 – List of PD-associated miRNA targets chosen from literature with their proposed mechanism, models used and respective publications for screening in iPSC-derived PD models. Blue represents direct binding to mRNA which encode familial PD-associated proteins, Green represents miRNA involved in mDA differentiation and function and Red represents miRNA associated with inflammatory processes. MSC, Mesenchymal stem cells.

177

iPSC-derived PD models that overexpress WT, A53T and G51D α-synuclein were developed and described in Chapter 3. These cells were differentiated to dopamine neurons for 35 days and then harvested for total RNA isolation. Differential miRNA expression analysis was subsequently performed by small RNA qPCR (Taqman). The miRNAs that were selected for analysis were determined following a literature review and illustrated in Table 5.3.

5.3.1.2 – Selection of housekeeping gene for normalisation of differential expression analysis

As differential regulation of miRNAs in certain disease conditions and cell types has been reported, the choice of a standardised housekeeping or reference gene is critical prior to quantitative polymerase chain reaction (qPCR) analysis. Generally, small non-coding RNAs such as small nuclear and small nucleolar RNAs are used as housekeeping genes. One of the most commonly used reference genes, RNU6B, has been seen to fluctuate and be highly variable in certain cell types (Xiang et al. 2014; Benz et al. 2013; Lamba et al. 2014). Another small RNA, RNU24 has been identified as a more suitable housekeeping gene in PD blood samples (Serafin et al. 2014). We determined the housekeeping gene of choice for our specific experimental requirements by performing qPCR with RNU6B and RNU24. Linear regression analysis of these candidate housekeeping genes was determined using Bestkeeper software.

There was relatively good correlation between both candidate genes and the BestKeeper index, with a coefficient of correlation of 0.996 for RNU6B and 0.964 for RNU24 (Figure 5.4).

However, as RNU6B was expressed at a relatively lower level (average Cq 33.41) compared to

RNU24 (average Cq 25.49), this resulted in high SD in RNU6B analysis (±1.28) which excluded the use of this reference gene for differential expression analysis in experimental samples.

178

Figure 5.4 – Bestkeeper regression analysis of 2 candidate housekeeping genes: RNU6B and RNU24. The BestKeeper index is determined using the geometric mean of raw Cq values of all candidate housekeeping genes. Both RNU6B and RUN24 candidate reference genes had high coefficient of correlation, 0.996 and 0.964 respectively, with a p value of 0.001 based on the Pearson’s correlation coefficient analysis. Column data analysis including range of raw Cq values, SD and Coefficient of variation. Graph to visualise raw Cq variation (n=1).

5.3.1.3 – Differential expression analysis of miRNAs involved in direct regulation of familial PD- associated genes qPCR analysis was undertaken to determine differential expression of selected miRNAs (Table

5.3 in blue). Analysis of intracellular miRNA showed no significant differences in miR-7 expression between all groups (p>0.05)(Figure 5.5A).

There was however, a significant increase in miR-34b expression in iPSC-derived mDA neurons overexpressing WT α-synuclein (3.82±0.67) compared to control (1±0.32), A53T (1.14±0.50) and G51D (0.37±0.07) samples (Figure 5.5B). There was also a similar trend in miR-34c, with

WT overexpressing iPSC-derived cultures expressing significantly elevated levels of miR-34c

(4.74±0.58) compared to control (1±0.34), A53T (1.38±0.53) and G51D (0.46±0.13) cultures

(Figure 5.5C). Although there was also a decrease in both miRNAs in G51D overexpressing cultures compared to control and A53T, this was only significant for miR34c expression in

G51D expressing cells compared to A53T cells (Figure 5.5C). Thus, analysis showed a significant upregulation in both cluster miRNA (miR-34b and miR-34c) in WT α-synuclein cultures.

179

miR-494 expression was also, more modestly, increased in iPSC-derived mDA neurons overexpressing WT α-synuclein (1.59±0.55) compared to A53T (0.77±0.28) and G51D

(0.89±0.17) cultures (Figure 5.5D). However, miR-494 was not significantly decreased in control iPSC-derived mDA cultures (1±0.15) compared to overexpressing WT α-synuclein cultures.

Thus, miRNAs associated with direct binding of both α-synuclein and DJ1 were observed to be perturbed with significant upregulation associated with the overexpression of WT α- synuclein. These perturbations may be associated with endogeneous cellular responses to increased α-synuclein load.

Figure 5.5 - Differential expression analysis for miRNAs involved in direct binding of familial PD-associated gene targets (miR-7, -34b, -34c and -494). Day 35 qPCR analysis, using Taqman small RNA assays, for iPSC-differentiated cultures overexpressing WT (W1-4), A53T (A1-4), and G51D (G1-4) α-synuclein, as well as controls (C1-4), at an MOI of 5 (n=4) and analysed using the ΔΔCt method relative to the small nucleolar RNA RNU24. (n=4). One-way ANOVA, Tukey post-hoc test (n=4). *p<0.05, *** p<0.001.

180

5.3.1.4 – Differential expression analysis of miRNAs involved in mDA development and function qPCR analysis was also undertaken to determine differential expression of selected miRNAs that have been previously associated with regulation of mDA development and functional pathways such as neurite outgrowth and dopamine receptor function (Table 5.3 in green).

The miRNAs analysed were miR-132, miR-133b and miR-137. Interestingly, none of these miRNAs were found to be differentially expressed in cells overexpressing WT or mutant α- synuclein, compared to control cells (Figure 5.6A-C).

Figure 5.6 - Differential expression analysis for miRNAs involved in mDA development and function (miR-132, -133b and - 137). Day 35 qPCR analysis, using Taqman small RNA assays, for iPSC-differentiated cultures overexpressing WT (W1-4), A53T (A1-4), and G51D (G1-4) α-synuclein, as well as controls (C1-4), at an MOI of 5 (n=4) and analysed using the ΔΔCt method relative to the small nucleolar RNA RNU24. (n=4). One-way ANOVA, Tukey post-hoc test (n=4).

5.3.1.5 – Differential expression analysis of miRNAs involved in neuroinflammatory processes qPCR analysis was undertaken to determine differential expression of selected miRNAs that have been previously associated with the regulation of neuroinflammatory pathways (Table

181

5.3 in red). Whilst expression of miR-21 was not changed between controls and wild-type, and mutant α-synuclein overexpressing cells (Figure 5.7A), miR-155 expression was significantly increased in iPSC-derived mDA neurons overexpressing WT α-synuclein

(2.58±0.74) to control (1±0.21), A53T (0.81±0.40) and G51D (0.81±0.16) samples (Figure 5.7B)

WT α-synuclein overexpressing cultures.

Figure 5.7 - Differential expression analysis for miRNAs associated with modulation of neuroinflammation (miR-21 and - 155). Day 35 qPCR analysis, using Taqman small RNA assays, for iPSC-differentiated cultures overexpressing WT (W1-4), A53T (A1-4), and G51D (G1-4) α-synuclein, as well as controls (C1-4), at an MOI of 5 (n=4) and analysed using the ΔΔCt method relative to the small nucleolar RNA RNU24. (n=4). One-way ANOVA, Tukey post-hoc test (n=4). *p<0.05, ** p<0.01, *** p<0.001.

5.3.1.6 Summary of intracellular miRNA analysis

In conclusion, these results suggest that overexpression of WT α-synuclein induces upregulation of miRNAs involved in α-synuclein direct post-transcriptional regulation (miRs

34b and 34c) and neuroinflammation (miR-155), compared to both control and A53T and

G51D mutant overexpressing iPSC-derived mDA cultures. Further, miR-34c was also downregulated in G51D samples compared to A53T suggesting some differential effect of these two mutant α-synuclein species. Furthermore, miR-494 expression appears to be downregulated in iPSC-derived mDA neurons overexpressing mutant α-synuclein compared

182

to WT α-synuclein overexpression, although there is a non-significant trend for lower expression in control samples.

5.3.2 Characterisation of extracellular vesicles isolated from cultured media from iPSC- derived neuronal PD models

5.3.2.1 Transmission electron microscopy

Exosomal miRNA can be released from cells and affect expression of target mRNAs in recipient cells. Exosomes are characterised by the presence of common proteins such as the tetraspanin proteins CD9, CD63 and CD81 by biochemistry techniques and visualised by negative staining transmission electron microscopy as having a cup-shaped morphology

(Simons et al. 2009; Mathivanan et al. 2010). We isolated exosomes produced by iPSC-derived dopaminergic neurons in culture and characterised these by negative staining transmission electron microscopy (TEM).

Firstly, negative staining TEM was used to identify the presence of extracellular vesicles isolated from conditioned media from untransduced day 30 differentiated iPSC-cultures

(Figure 5.8). The appearance of structures with an extracellular vesicle-like, cup-shaped morphology object was observed (Figure 5.8A). Further characterisation of these vesicle-like structures using immunogold staining for exosomal transmembrane proteins CD63 and CD81 showed that exosomes isolated from differentiating stem cells were clearly identifiable with

TEM and immunogold staining showed localisation with extracellular vesicle-like objects

(Figure 5.8B, C). Thus, isolation of exosomes from differentiating iPSC-derived neuronal cultures was successful in isolating CD63, CD81 expressing extracellular vesicles by immunogold staining.

183

Figure 5.8 - Transmission electron microscopy (TEM) images of isolated extracellular vesicle-like structures isolated from 1ml cultured media by miRCURY exosome isolation kit from NAS2 iPSC cultures on day 30 of mDA differentiation (yellow arrows). Structures show bilayered membranes characteristic of extracellular vesicles and average size was approximately 90nm. (A) TEM without staining and TEM showing co-localisation of EV-like structures with (B) CD63 and (C) CD81 antibodies by immunogold staining.

5.3.2.2 Nanoparticle tracking analysis

Nanoparticle Tracking Analysis (NTA) was performed to further characterise exosomes that were secreted by iPSC-derived mDA neurons expressing wild-type and mutant synuclein

(Figure 5.9) (NTA was performed by collaborator Dr. Andrew Shearn from the Emanueli lab in

Bristol).

184

Figure 5.9 – NTA analysis of conditioned media samples from day 32 differentiating iPSC-derived cultures overexpressing WT (W1-4), A53T (A1-4), and G51D (G1-4) α-synuclein, as well as controls (C1-4), at an MOI of 5 (n=4). A-E) NTA analysis showing distribution of concentration based on specific size ranges A) 30-60nm, B) 60-90nm, C) 90-120nm, D) 120-210nm and the cumulative concentration and size of all particles E) 30-120nm in length representing exosome size ranges (n = 4). One-way ANOVA, Tukey post-hoc test (n=4). *p<0.05, ** p<0.01.

NTA analysis revealed a decrease in the concentration of 30-60nm (Figure 5.9A) and 60-90nm

(Figure 5.9B) particles in iPSC-derived mDA neurons overexpressing A53T and G51D mutant

α-synuclein compared to control iPSC-derived mDA cultures, on day 32 of differentiation.

Furthermore, a decrease in concentration of particles ranging between 90-120nm (Figure

5.9C) and 120-210nm (Figure 5.9D)(larger microvesicles) was observed from the conditioned media of iPSC-derived neurons overexpressing WT α-synuclein as compared to control iPSC- derived mDA cultures. For all particles ranging between 30-120nm (Figure 5.9E) in size, which

185

is the prospective exosome size range, there was a significant decrease in all experimental synuclein overexpression models compared to control. Thus, it appears that in our PD models, there is a decrease in exosome number as a result of perturbed α-synuclein expression, suggesting that increased α-synuclein leads to a change in exosomal release dynamics.

Furthermore, these changes appear within different subpopulations of exosomes, as determined by size of particles, in response to different strains of α-synuclein such that overexpression of mutant strains induce a deficiency in the release of smaller (30-90nm) particles whilst overexpression of wild-type α-synuclein induces a deficiency in the release of larger (90-120nm and 120-210nm) particles.

5.3.3 Exosomal miRNA profiling isolated from iPSC-derived neuronal culture conditioned media

Exosomes were purified from conditioned media of iPSC-derived neuronal cultures on day 35 of differentiation and qPCR was performed on miRNA isolated from exosomes. All 9 selected miRNAs listed in Table 1 were tested for differential expression using a spiked-in C. elegans exogenous reference miRNA (cel-39-3p). Expression changes were also normalised to the average exosomal concentration for each culture condition, as determined by NTA analysis.

Results showed that many of these miRNAs (miRNAs 7, 34b, 133b, 155 and 494) were undetectable in the exosomal preparations, suggesting that they were expressed at low concentrations that were beyond detection thresholds.

A few miRNAs were within the detectable range for qPCR analysis (miRNAs 21, 34c, 132, 137).

The results showed that there was a trend for miR-21 and miR-132 expression to be elevated in iPSC-derived neurons overexpressing G51D compared to control neurons, although this did not reach significance (Figure 5.10 A, C). miR-137 was expressed in these cells but there were

186

no statistically significant differences in levels between control and α-synuclein- overexpressing neurons (Figure 5.10D). Interestingly, there was a large increase in miR-34c in

WT α-synuclein overexpressing cultures (7.50±6.10), compared to control (1±0.77), A53T

(1.46±0.89) and G51D (1.34±0.99) cultures (Figure 5.10B) nearing significance compared to control (p=0.07). However, this did not reach statistical significance due to the large variation between samples. This observation replicated observed intracellular increase of miR-34c in response to WT α-synuclein overexpression. Thus, it appears as though a specific set of miRNA are enriched in exosomes derived from iPSC-derived neuronal cultures.

Figure 5.10 – Extracellular miRNA expression profiling. Day 35 iPSC cultures differentiated towards an mDA fate, overexpressing WT (W1-4), A53T (A1-4), and G51D (G1-4) α-synuclein, as well as controls (C1-4), at an MOI of 5 (n=4). Exosomes were isolated from 1ml of conditioned media using Exiqon exosome isolation kit. miRNA differential expression analysis was performed using Taqman small RNA assays for all 9 selected miRs and analysed using the ΔΔCt method relative to an exogeneous spiked in control, cel-miR-3p. One-way ANOVA, Tukey post-hoc test (n=4).

187

5.4 Discussion miRNAs have been shown to play critical roles in the development of various diseases such as

PD; implicating their potential use as novel therapeutic agents or biomarkers. Translatomic changes associated with perturbed α-synuclein expression have been studied, using TRAP and

RNA-sequencing, identifying significant differential expression of a critical protein component involved in miRNA biogenesis, DGCR8, however, this could not be validated using other analysis techniques. As post-transcriptional regulation of mRNA can be regulated by miRNA, miRNA analysis was employed to determine prospective mechanisms of post-transcriptional regulation of PD-associated pathways, from intracellular and exosomal fractions derived from iPSC-derived PD models.

5.4.1 Intracellular miRNA profiling in iPSC-derived PD models

Intracellular miRNA profiles showed differential expression of miRs 34b/c, -155 and -494.

Previous studies have identified an association between decreased miR-34b/c levels with PD patient SNpc tissues, and this was thought to affect the post-transcriptional regulation of α- synuclein. Here we showed an increase of miR-34b/c in WT synuclein overexpression cultures compared to control and mutant α-synuclein overexpressing cells. A very recent study of miRNA expression using whole transcriptome miRNA sequencing in iPSC-derived neurons from patients with 22q11.del syndrome, which includes a loss of DGCR8, showed a downregulation of nearly all miRNA except for miR-34b/c which were upregulated, although this did not reach significance (Zhao et al. 2015). Thus, our results may be indicative of early regulatory processes in response to increased wild-type α-synuclein load or that increased α- synuclein over time is part of a negative feedback loop inducing a decrease in levels of this miRNA over time.

188

Similarly, miR155 was also seen to be expressed at significantly higher levels in wild-type α- synuclein overexpressing cultures compared to control and mutant α-synuclein overexpressing cells. miR155 has been identified as a regulator of microglial inflammatory responses and has been found to be upregulated in microglia and macrophages in response to proinflammatory cytokines (Cardoso et al. 2012). In the context of PD, an AAV-WT α- synuclein overexpression mouse model identified an upregulation of miR-155 in the SNpc of

α-synuclein overexpressing mice compared to control mice (Thome et al. 2016). Furthermore, it is thought that macrophages can regulate miR-155 expression during an immune response

(Androulidaki et al. 2009). The observed increase in miR-155 expression in this study concurs with the observations of elevated miR-155 levels in the α-synuclein overexpression mouse model (Thome et al. 2016). As mentioned in Chapter 3, the iPSC-derived cultures showed a heterogenous mixture of neuronal and glial cells, as identified by GFAP expression.. It has been shown that neuronal cultures require non-neuronal cells to modulate the effects of neuroinflammation (Hui et al. 2016) and, a limitation of this study is that the non-neuronal cells were not characterised sufficiently to ascertain cell types due to lack of sample material.

Further studies should aim to identify if cytokine expression is changed in response to perturbed α-synuclein in these models alongside miR-155 upregulation. Interestingly, overexpression of mutant α-synuclein species did not affect miR-155 levels, suggesting an ability of the WT species specifically to produce an inflammatory response at the stage of differentiation analysed. Inhibition of miR-155 may represent a therapeutic avenue for neurodegenerative diseases and is currently being developed as a therapy for amyotrophic lateral sclerosis (ALS) (Koval et al. 2013).

189

As described in section 1.5.4 1, loss-of-function mutations in DJ1 protein are thought to be associated with autosomal recessive early-onset PD, related to a perturbed cellular response to oxidative stress (Bonifati et al. 2003). Moreover, DJ1 protein levels have been found to be reduced in human sporadic PD cortex samples (Nural et al. 2009). Overexpression of miR-494 has been identified as being able to significantly downregulate DJ1 protein, by directly binding

DJ1 mRNA and thus leaving cells more prone to oxidative stress and degeneration (Xiong et al. 2014). Our results showed that iPSC-derived mDA neurons overexpressing WT α-synuclein resulted in a statistically significant upregulation of miR-494 compared to mutant species models. Thus, compared to the mutant species, WT α-synuclein appears to induce an upregulation of a miRNA associated with downregulation of DJ1 protein and reduced cellular oxidative stress response. Therefore, an increase in miR-494, as observed in our analysis, may be able to induce a decrease in DJ1 protein levels which, can increase the susceptibility of neurons to oxidative stress-mediated degeneration. Protein expression analysis for DJ1 could be undertaken to validate this hypothesis, however there was not enough time to perform this analysis.

Neurite outgrowth and differentiation potential were analysed in chapter 3 and showed no significant changes in differentiation potential, as determined by expression of neuronal and dopaminergic markers, whilst neurite growth was attenuated in cultures expressing WT and mutant α-synuclein as compared to control cultures. Thus, it is not surprising that expression of miRNAs 133b and 137, which are associated with developmental regulation of mDA neurons, were not changed between cultures. Furthermore, miRNA expression profiling suggests that the changes in neurite dynamics are not associated with perturbed expression of either miRNAs 132 or 133b.

190

miRNA profiling in iPSC-derived PD models was originally undertaken to interrogate RNA-seq results, showing differential expression of a key protein involved in miRNA biogenesis, DGCR8.

Relating these results to RNA-seq differential expression analysis results, DGCR8 showed a significant downregulation in expression in WT overexpression models compared to control

(Chapter 4). A non-significant decrease was also noted in A53T overexpression compared to control. However, DGCR8 is involved in the biogenesis of all miRNA and expression profiling results suggest that not all miRNAs are differentially expressed. If differential expression was linked to perturbed expression of DGCR8, a generalised perturbation of all miRNA, processed by the canonical pathway, would be expected. Studies employing DGCR8-KO mice have reported a downregulation of over 90% of miRNA compared to control mice (Zou et al. 2018).

DGCR8-KO mESCs have been studied and showed a reduction in intermediate pre-miRNA and fully mature miRNA, giving further evidence of DGCR8-dependent miRNA processing (Wang et al. 2007). Thus, the pattern of specific miRNA differential expression may be due to another activated pathway resulting from perturbed α-synuclein expression in iPSC-derived cultures.

As the model employed in this study appears to produce robust differential expression in certain miRNA, it would be very beneficial to attempt small RNA-sequencing, or newer platforms like the NanoString nCounter technology, to validate these results and look at genome-wide miRNA differential expression. Our results however do validate previous literature that has identified a role for the neuroinflammation-associated miR-155 and the α- synuclein regulating cluster miRs-34b and 34c. An α-synuclein transgenic overexpression animal model would be ideal to characterise and validate these findings over different stages of disease progression.

191

5.4.2 Extracellular miRNA profiling in iPSC-derived PD models

Several studies have attempted to identify novel extracellular miRNA biomarkers, as well as miRNA within exosomes, in bodily fluids of PD patients compared to controls (Table 5.1). In this study, exosomal concentration was seen to decrease in α-synuclein overexpression cultures compared to control cultures. Although there is no evidence for exosome concentration changes in disease in literature, increased α-synuclein load may be dysregulating cellular functions involved in production or release of exosomes. Further mechanistic investigations would be interesting in the future.

Extracellular miRNA profiles have been related to staging of the pathological process and may be employed clinically to identify disease progression in both PD and Alzheimer’s disease

(Burgos et al. 2014). iPSCs can be an ideal source of biological material for these types of studies as they can be differentiated into the cells affected in PD and disease modelling can be achieved using genetic manipulation techniques, thus recapitulating certain disease phenotypes in-vitro. It is also important to make sure that the in vitro models used recapitulate the in vivo environment and processes for production of extracellular miRNA. In this case, miR-155 was upregulated in wild-type α-synuclein overexpressing cultures, and this may be able to mediate a response from GFAP expressing glial cells also observed in our cultures. Further studies could thus determine levels of inflammatory cytokines as a proxy for activation of microglia. Furthermore, whether exosome-mediated transfer is involved could be determined by culturing exosomes from wild-type overexpressing cells with control cultures and analysing activation of microglia.

Thus, this study further aimed to identify miRNA expression profiles from exosomes isolated from conditioned media from iPSC-derived PD model cultures by qPCR after successful

192

characterisation by TEM and NTA analysis. Results suggest that only a subset of the assayed miRNA are expressed at detectable levels in exosome fractions (miRs 21, 34c, 132 and 137).

Profiling of exosomal miRNA may thus require the collection of increased volume of conditioned media for more robust expression analysis. Detectable miRNAs are thought to regulate a number of cellular processes such as protein degradation and inflammation, direct regulation of α-synuclein, differentiation and mDA function. Importantly, miR-21 is a miRNA which has been identified to be trafficked via exosomes and produce a pro-inflammatory response in immune cells (Fabbri et al. 2012) . As this miRNA is relatively highly expressed in exosomes from iPSC-derived models, miR-21 may have the potential to be regulated therapeutically to reduce neuroinflammation via cell-to-cell signalling pathways.

Furthermore, although normalisation was controlled by exosome count and a spiked-in exogenous control, other factors such as method of exosome isolation and RNA isolation methods may hinder the identification of true biological effects (Eldh et al. 2012; Ding et al.

2018). Future studies should identify the best normalisation strategy for the cell type being studied, which may require global miRNA profiling. Unfortunately, due to the low amount of miRNA which can be extracted from exosomes, small RNA sequencing may not be possible without prior amplification or upscaling of cultures to produce more exosomes for expression profiling.

5.5 Conclusion

In conclusion, the results contained within this chapter suggest that perturbed α-synuclein is associated with differential expression of multiple miRNA involved in regulation of mDA function or PD pathological processes. Exosome isolation from conditioned media of iPSC- derived cultures was characterised by TEM and NTA analysis. This allowed for identification

193

of exosomal miRNA. Furthermore, exosome concentration analysis by NTA suggests that levels of exosomes are decreased in α-synuclein overexpression models. Finally, certain miRNA were detected in exosomal isolates and these may mediate signalling between cells in culture. Thus, evidence is presented for a dysregulation of intracellular miRNA and exosome dynamics.

194

Chapter 6 General Discussion 6.1 - Parkinson’s disease and α-synuclein

Parkinson’s disease (PD) is an etiologically heterogenous disease, with a diversity of causative molecular mechanisms proposed, ultimately resulting in the degeneration of midbrain dopaminergic (mDA) neurons of the substantia nigra pars compacta (SNpc). The death of these mDA neurons leads to dopamine deficiency in the striatum which clinically presents in patients as motor symptoms, although many other clinical non-motor symptoms have been described. As with many other neurodegenerative diseases, PD is accompanied by the presence of intracellular aggregates of misfolded protein. In the case of PD, these protein aggregates are primarily composed of misfolded α-synuclein protein (Lewy bodies and Lewy neurites). Indeed, aggregates of α-synuclein are thought to develop and propagate throughout the central nervous system (CNS), with evidence for loss of other CNS neurons during the different stages of disease development. Evidence has also been presented for the mechanisms of propagation of α-synuclein pathology, primarily by mechanisms of cell-to-cell transfer. The etiology of PD is linked to several factors including ageing, environmental exposures and several genomic loci which link PD to pathological mechanisms including mitochondrial dysfunction, response to oxidative stress, neuroinflammation and protein degradation. However, the presence of α-synuclein aggregates appears to be a pathognomonic hallmark of PD and thus perturbation of α-synuclein in cellular and animal models have been analysed using a variety of methods.

6.2 - Development of a SH-SY5Y model for translating ribosome affinity purification

The aim of the work presented was to design and implement a strategy towards the study of actively translating mRNA in cellular synucleinopathy neuronal models. To this end, TRAP was

195

employed on a neuroblastoma SH-SY5Y model of synucleinopathy which express an EGFP tagged ribosomal protein. To express the tagged ribosomal protein, lentiviral particles were generated to test different promoters for robust expression in SH-SY5Y cells. This test identified the CMV promoter as the optimal promoter for use in this cell line from a test of 4 different promoters and may aid in the choice of promoter for use in future studies employing this cell line. Furthermore, this test corroborates a previous study that proposed the CMV promoter as an ideal promoter in these cells (Li et al. 2010).

The synucleinopathy models were generated using lentiviral particles to overexpress WT and mutant species of α-synuclein. α-synuclein protein expression was identified by Western blot protein expression. Furthermore, the FACS sorting for expression of the EGFP tagged ribosomal protein was employed to produce a purified population of cells for TRAP.

6.3 - Translating ribosome affinity purification, RNA-seq library preparation and analysis

The TRAP protocol was applied to these models and actively translating mRNAs were immunoprecipitated using anti-EGFP coated magnetic beads and used to prepare a NGS library for RNA sequencing. This involved amplification of the low yield translating mRNA and reverse transcription to cDNA fragments. Ligation of adapter sequences to cDNA fragments for multiplex sequencing was performed and adapter-ligated sequences were amplified and pooled together for sequencing. Trace analysis for each processing stage was employed to identify successful processing based on average fragment size and concentration.

RNA-seq results were analysed using a commonly used RNA-seq analysis pipeline, namely the

TUXEDO analysis pipeline (Trapnell et al. 2012). This pipeline was employed using slight modifications so as to assess the contribution of pipeline algorithms to differential gene

196

expression analysis. As has been previously identified, these different pipelines lead to a slight divergence in identification of DEGs, with just over 30% of overall DEGs identified in all three pipelines. In Chapter 4, discussion of the limitations of the RNA-seq analysis was described.

Although alignment statistics showed good alignment to the most recent human genome reference, some flags were raised with respect to overamplification. However, as RNA-seq is a relatively newly adopted analysis technique in the field of research, there is no definitive criteria for a “successful” RNA-seq experiment. Although these flags may represent a reduction in the depth of coverage, and the extent to which analysis may be compromised is difficult to characterise, they are known to be associated with pre-amplification and ribosomal samples.

In this study, RNA-seq analysis produced a list of genes that were differentially expressed between control SH-SY5Y and wild-type and mutant A53T α-synuclein overexpression SH-

SY5Y cultures. Gene targets with significantly high fold changes were identified and classed based on prospective association with disease processes including: cytoskeletal, mitochondrial, protein trafficking, dopamine signalling, oxidative stress response and apoptosis regulation. Finally, these targets were validated using a variety of transcript level and protein level analyses. Unfortunately, validation using qPCR and Western blotting was only able to partially confirm the differential expression of a few targets including TOMM40,

SLC25A12, UNC119B and CRADD.

TOMM40 and SLC25A12 have both previously been associated with mitochondrial function and these results suggest that perturbed α-synuclein expression may produce a change in expression of proteins that regulate these functions. UNC119B has been linked to modulation of neuronal axon branching and mediates intracellular trafficking. Dysregulation of

197

intracellular trafficking has been associated with PD pathology and α-synuclein itself has been identified as acting as a chaperone protein for complexes that mediate synaptic vesicle exocytosis. Thus, UNC119B could potentially be a novel PD-associated target. Finally, CRADD was also partially validated in our SH-SY5Y models and is associated with the induction of caspase-2 dependant apoptosis. This protein has recently been identified as a novel therapeutic target to rescue neuronal cell death in neurodegenerative disease (Jang et al.

2016). Transcriptomics has been employed in the study of PD and has resulted in the confirmation of several pathways including but not limited to mitochondrial function, oxidative stress, protein degradation, protein chaperones and neuroinflammation

(Borrageiro et al. 2018; Cooper-Knock et al. 2012). Enrichment analysis further identified systemic changes in pathways including, significantly, extra-cellular matrix (ECM) proteins.

Brain ECM is thought to regulate the neuronal environment and perturbation is linked to activation of microglia (Bonneh-Barkay et al. 2009). Furthermore, neuronal cells in culture are known to require ECM proteins for their survival and these proteins provide neuroprotection to neurons including specifically mDA neurons (Zhang, Yang, Toledo, et al. 2017). ECM proteins have also been associated with regulation of neurite dynamics such as outgrowth

(Lamoureux et al. 1990; Tonge et al. 2012). Other altered pathways include actin filament branch point, dendrite extension and actin body pathways, with a striking number of differentially expressed genes in these pathways. These pathways and ECM pathways link together the important interplay between the actin cytoskeleton and the ECM which regulate the structure and function of synapses with possible relevance to PD pathological process

(Chapman 2014).

198

Furthermore, a differentially expressed gene was identified which is known to play an important functional role within the microRNA biosynthesis pathway, DGCR8. This prompted us to look towards studying miRNA expression profiles, this time, within a more physiologically relevant model, iPSC-derived PD models.

6.4 Development of an iPSC-model of synucleinopathy

To generate iPSC-derived synucleinopathy models, iPSCs were preliminarily subjected to an mDA differentiation protocol and characterised by expression of midbrain, neuronal and dopaminergic markers. Analysis showed an enrichment of dopaminergic neurons with floor plate markers and thus this model would represent a physiologically relevant model for the study of PD in the cells that degenerate during pathology. The original intention was to use iPSC-derived cultures for TRAP, however, the low yield of immunoprecipitated mRNA from

SH-SY5Y models, using many more cells than would be possible due to the costs involved with iPSC differentiation, this was deemed unfeasible.

Thus, again several promoters were tested for their ability to express transgenes in iPSC- derived mDA cultures. For this study, the hSYN1 promoter was used as it could be used to overexpress α-synuclein specifically in neurons. However, the other promoters were also seen to produce robust expression of transgenes in iPSC-derived neurons as well as non- neuronal cells. Thus, overexpression of wild-type and mutant α-synuclein was achieved by generating lentiviral particles to overexpress wild-type, and A53T and G51D mutant α- synuclein. Characterisation by protein analysis for α-synuclein showed robust overexpression in all models, although the expression of the A53T mutant was less than WT and G51D α- synuclein. It was noted that α-synuclein overexpression in these cultures did not produce any significant differences in differentiation capacity, although this was a previously identified

199

phenomenon in response to elevated α-synuclein in an iPSC-derived model (Oliveira et al.

2015).

6.5 Functional characterisation of neurite dynamics

As the RNA-seq results showed many differentially expressed genes related to cytoskeletal and ECM proteins, these cells were functionally characterised for neurite dynamics. Results showed significantly reduced neurite length and neurite branching in WT and G51D overexpressing iPSC-derived dopaminergic neurons. This is in keeping with previously published results showing decreased neurite extension in cells with elevated α-synuclein levels (Oliveira et al. 2015; Takenouchi et al. 2001). Thus, these results may represent a link between RNA-seq results and functional/structural disturbances leading to a reduction in neurite outgrowth and branching. Future studies should focus on biochemical expression of

ECM proteins and expression of structural proteins associated with neurite dynamics.

6.6 Intracellular and extracellular miRNA profiling

As DGCR8 was seen to be differentially expressed from RNA-seq results, both intracellular and extracellular microRNA expression profiles were interrogated. For extracellular microRNA profiling, exosomes were isolated and characterised by both transmission electron microscopy and nanoparticle tracking analysis. These results validated the use of the commercially available kit for exosome isolation from conditioned media of iPSC-derived cultures. Furthermore, exosomal fractions, based on size, were differentially produced, with an overall decrease in α-synuclein overexpressing cells compared to control in the prospective exosomal size range. However, mechanistically, these insights remain to be elucidated.

Thus, both the intracellular and extracellular miRNA profiles were interrogated via qPCR for miRNAs thought to be involved in PD pathogenesis and mDA function. The results from

200

intracellular expression profiling showed the differential expression of 4 miRNAs (miRs -

34b/c, -494 and -155). These results suggest that increased wild-type α-synuclein, in iPSC- derived mDA neurons, results in an increase in specific miRNAs. These miRNAs are involved specifically in the regulation of α-synuclein and DJ1 expression and the modulation of neuroinflammation. miR-155, an important modulator of neuroinflammation, has previously been found to be elevated in an α-synuclein overexpression mouse model, corroborating results in this study (Thome et al. 2016). Whether this increase in miR-155 results in an increased activation of microglia remains to be studied and could be the focus of future studies. Thus, intracellular expression profiling identified miRNAs which may represent therapeutic targets for PD as well as other neurodegenerative diseases. Further studies that profile differential miRNA expression in α-synuclein overexpression models (for example using miRNA microarrays) may also help identify other miRNAs that potentially play important roles in PD pathology.

Extracellular miRNA profiles showed no differential expression and only 4 miRNA (miRs -21, -

34c, -132 and -137) were within the detectable range for expression. This includes a miRNA involved in induction of neurite outgrowth (miR-132) and neuroinflammation (miR-21). miR-

21 from exosomes has been associated with activation of microglia and represents a possible mechanism for induction of neuroinflammation associated with PD. However, future studies would require the isolation of increased amounts of conditioned media for robust differential expression analysis.

6.7 Conclusions and future work

In conclusion, the results presented show an array of transcriptomic changes in response to perturbed α-synuclein levels in cellular synucleinopathy models. NGS profiling showed

201

differential expression of genes associated with several pathways previously linked to PD.

NGS technologies represent a relatively new technique and analysis algorithms are being modified and developed to produce robust, reproducible gene expression analysis. As it has been shown that reproducibility for DEG analysis requires many biological repeats (Schurch et al. 2016), the future of NGS technologies relies on the implementation of reproducible platforms which can be collected and analysed together for robust data analysis. A project similar to the 100,000 genomes project that was undertaken by the National Health Service to link genomic abnormalities to medical histories may be useful in this respect (Turnbull et al. 2018). A project of this scale in conjunction with genome-wide transcriptional and/or translational analysis may be able to produce a wealth of information (DEG analysis, alternative splicing, etc) for the development of disease diagnosis and new therapies. miRNA profiling also showed differential expression of multiple miRNAs in iPSC-derived synucleinopathy models. Transgenic mice with either knock out or overexpression of the specific miRNA may help in understanding the role of these miRNA. In situ hybridisation may be able to help in understanding dynamics of miRNA expression. Finally, multiple lines of enquiry show the involvement of neurite extension and branching in response to perturbed

α-synuclein. These may represent pre-symptomatic alterations and possible therapeutic avenues (Gu et al. 2013; Calabrese 2008).

202

References Abbas, Nacer, Christoph B Lücking, Sylvain Ricard, Alexandra Dürr, Vincenzo Bonifati, Giuseppe De Michele, Sandrine Bouley, Jenny R Vaughan, Thomas Gasser, and Roberto Human molecular genetics Marconi. 1999. 'A wide variety of mutations in the parkin gene are responsible for autosomal recessive parkinsonism in Europe', 8: 567-74. Abeliovich, Asa, Yvonne Schmitz, Isabel Fariñas, Derek Choi-Lundberg, Wei-Hsien Ho, Pablo E Castillo, Natasha Shinsky, Jose Manuel Garcia Verdugo, Mark Armanini, and Anne Neuron Ryan. 2000. 'Mice lacking α-synuclein display functional deficits in the nigrostriatal dopamine system', 25: 239-52. Adams, R., L. Van Bogaert, and H. Van Der Eecken. 1961. '[Nigro-striate and cerebello-nigro-striate degeneration. (Clinical uniqueness and pathological variability of presenile degeneration of the extrapyramidal rigidity type]', Psychiatr Neurol (Basel), 142: 219-59. Adem, Abdu, Maria EK Mattsson, Agneta Nordberg, and Sven Developmental brain research Påhlman. 1987. 'Muscarinic receptors in human SH-SY5Y neuroblastoma cell line: regulation by phorbol ester and retinoic acid-induced differentiation', 33: 235-42. Ahn, Bong-Hyun, Hyangshuk Rhim, Shi Yeon Kim, Young-Mo Sung, Mun-Yong Lee, Ju-Youn Choi, Benjamin Wolozin, Jong-Soo Chang, Young Han Lee, and Taeg Kyu Journal of Biological Chemistry Kwon. 2002. 'α-Synuclein interacts with phospholipase D isozymes and inhibits pervanadate-induced phospholipase D activation in human embryonic kidney-293 cells', 277: 12334-42. Akopian, Armen N, and John N Journal of Biological Chemistry Wood. 1995. 'Peripheral nervous system-specific genes identified by subtractive cDNA cloning', 270: 21264-70. Alevizos, Ilias, and Gabor G Nature Reviews Rheumatology Illei. 2010. 'MicroRNAs as biomarkers in rheumatic diseases', 6: 391. Alexander, Margaret, Ruozhen Hu, Marah C Runtsch, Dominique A Kagele, Timothy L Mosbruger, Tanya Tolmachova, Miguel C Seabra, June L Round, Diane M Ward, and Ryan M Nature communications O’Connell. 2015. 'Exosome-delivered microRNAs modulate the inflammatory response to endotoxin', 6: 7321. Allen, Shelley J, Judy J Watson, Deborah K Shoemark, Neil U Barua, Nikunj K Pharmacology Patel, and therapeutics. 2013. 'GDNF, NGF and BDNF as therapeutic options for neurodegeneration', 138: 155-75. Alvarez-Erviti, Lydia, Yiqi Seow, Anthony H Schapira, Chris Gardiner, Ian L Sargent, Matthew JA Wood, and J Mark Neurobiology of disease Cooper. 2011. 'Lysosomal dysfunction increases exosome-mediated alpha-synuclein release and transmission', 42: 360-67. Andén, N-E, A Carlsson, A Dahlström, K Fuxe, N-Å Hillarp, and Ko Larsson. 1964. 'Demonstration and mapping out of nigro-neostriatal dopamine neurons', Life sciences, 3: 523-30. Anders, Simon, and Wolfgang Genome biology Huber. 2010. 'Differential expression analysis for sequence count data', 11: R106. Anderson, John P, Donald E Walker, Jason M Goldstein, Rian De Laat, Kelly Banducci, Russell J Caccavello, Robin Barbour, Jiping Huang, Kristin Kling, and Michael Journal of Biological Chemistry Lee. 2006. 'Phosphorylation of Ser-129 is the dominant pathological modification of α-synuclein in familial and sporadic Lewy body disease', 281: 29739-52. Andersson, Elisabet, Ulrika Tryggvason, Qiaolin Deng, Stina Friling, Zhanna Alekseenko, Benoit Robert, Thomas Perlmann, and Johan Cell Ericson. 2006. 'Identification of intrinsic determinants of midbrain dopamine neurons', 124: 393-405. Andrew, Peter, George Banting, Ivan Damjanov, Danielle Arnaud, and Philip Avner. 1984. 'Three monoclonal antibodies defining distinct differentiation antigens associated with different high molecular weight polypeptides on the surface of human embryonal carcinoma cells', 3: 347-61. Andrews, Simon. 2010. 'FastQC: a quality control tool for high throughput sequence data'.

203

Androulidaki, Ariadne, Dimitrios Iliopoulos, Alicia Arranz, Christina Doxaki, Steffen Schworer, Vassiliki Zacharioudaki, Andrew N Margioris, Philip N Tsichlis, and Christos Immunity Tsatsanis. 2009. 'The kinase Akt1 controls macrophage response to lipopolysaccharide by regulating microRNAs', 31: 220-31. Ang, Siew-Lan, and Janet Cell Rossant. 1994. 'HNF-3β is essential for node and notochord formation in mouse development', 78: 561-74. Anglade, P, S Vyas, F Javoy-Agid, MT Herrero, PP Michel, J Marquez, A Mouatt-Prigent, M Ruberg, EC Hirsch, Y Histology Agid, and histopathology. 1997. 'Apoptosis and autophagy in nigral neurons of patients with Parkinson's disease', 12: 25-32. Angot, Elodie, Jennifer A Steiner, Carla M Lema Tomé, Peter Ekström, Bengt Mattsson, Anders Björklund, and Patrik PLoS One Brundin. 2012. 'Alpha-synuclein cell-to-cell transfer and seeding in grafted dopaminergic neurons in vivo', 7: e39465. Appel-Cresswell, Silke, Carles Vilarino-Guell, Mary Encarnacion, Holly Sherman, Irene Yu, Brinda Shah, David Weir, Christina Thompson, Chelsea Szu-Tu, and Joanne Trinh. 2013. 'Alpha- synuclein p. H50Q, a novel pathogenic mutation for Parkinson's disease', Movement Disorders, 28: 811-13. Arena, Andrea, A M Iyer, I Milenkovic, G G Kovacs, I Ferrer, M Perluigi, and E Current Alzheimer Research Aronica. 2017. 'Developmental expression and dysregulation of miR-146a and miR- 155 in Down's syndrome and mouse models of Down's syndrome and Alzheimer's disease', 14: 1305-17. Arenas, Ernest, Mark Denham, and J Carlos Development Villaescusa. 2015. 'How to make a midbrain dopaminergic neuron', 142: 1918-36. Arroyo, Jason D, John R Chevillet, Evan M Kroh, Ingrid K Ruf, Colin C Pritchard, Donald F Gibson, Patrick S Mitchell, Christopher F Bennett, Era L Pogosova-Agadjanyan, and Derek L Proceedings of the National Academy of Sciences Stirewalt. 2011. 'Argonaute2 complexes carry a population of circulating microRNAs independent of vesicles in human plasma', 108: 5003-08. Arslan, Fatih, Ruenn Chai Lai, Mirjam B Smeets, Lars Akeroyd, Andre Choo, Eissa NE Aguor, Leo Timmers, Harold V van Rijen, Pieter A Doevendans, and Gerard Stem cell research Pasterkamp. 2013. 'Mesenchymal stem cell-derived exosomes increase ATP levels, decrease oxidative stress and activate PI3K/Akt pathway to enhance myocardial viability and prevent adverse remodeling after myocardial ischemia/reperfusion injury', 10: 301-12. Ashburner, Michael, Catherine A Ball, Judith A Blake, David Botstein, Heather Butler, J Michael Cherry, Allan P Davis, Kara Dolinski, Selina S Dwight, and Janan T Nature genetics Eppig. 2000. 'Gene Ontology: tool for the unification of biology', 25: 25. Austin, Susan A, Angela M Floden, Eric J Murphy, and Colin K Journal of Neuroscience Combs. 2006. 'α-synuclein expression modulates microglial activation phenotype', 26: 10558-63. Avilion, Ariel A, Silvia K Nicolis, Larysa H Pevny, Lidia Perez, Nigel Vivian, Robin Genes Lovell-Badge, and development. 2003. 'Multipotent cell lineages in early mouse development depend on SOX2 function', 17: 126-40. Baba, Minami, Shigeo Nakajo, Pang-Hsien Tu, Taisuke Tomita, Kazuyasu Nakaya, VM Lee, John Q Trojanowski, and Takeshi Iwatsubo. 1998. 'Aggregation of alpha-synuclein in Lewy bodies of sporadic Parkinson's disease and dementia with Lewy bodies', The American journal of pathology, 152: 879. Bai, Zhouxian, Boryana Stamova, Huichun Xu, Bradley P Ander, Jiajia Wang, Glen C Jickling, Xinhua Zhan, DaZhi Liu, Guangchun Han, Lee-Way Alzheimer disease Jin, and associated disorders. 2014. 'Distinctive RNA Expression Profiles in Blood Associated with Alzheimer’s Disease after Accounting for White Matter Hyperintensities', 28: 226. Balasooriya, Inoka S, and Kandatege Wimalasena. 2007. 'Are SH-SY5Y and MN9D cell lines truly dopaminergic?'.

204

Barberi, Tiziano, Peter Klivenyi, Noel Y Calingasan, Hyojin Lee, Hibiki Kawamata, Kathleen Loonam, Anselme L Perrier, Juan Bruses, Maria E Rubio, and Norbert Nature biotechnology Topf. 2003. 'Neural subtype specification of fertilization and nuclear transfer embryonic stem cells and application in parkinsonian mice', 21: 1200. Barger, George, and Henry Hallett Dale. 1910. 'Chemical structure and sympathomimetic action of amines', The Journal of physiology, 41: 19. Barger, George, and Arthur James Ewins. 1910. 'Some phenolic derivatives of β-phenylethylamine', Journal of the Chemical Society, Transactions, 97: 2253-61. Barker, Roger A, Malin Parmar, Lorenz Studer, and Jun Cell stem cell Takahashi. 2017. 'Human trials of stem cell-derived dopamine neurons for Parkinson’s disease: dawn of a new era', 21: 569- 73. Bartels, Tim, Joanna G Choi, and Dennis J Nature Selkoe. 2011. 'α-Synuclein occurs physiologically as a helically folded tetramer that resists aggregation', 477: 107. Batelli, Sara, Diego Albani, Raffaela Rametta, Letizia Polito, Francesca Prato, Marzia Pesaresi, Alessandro Negro, and Gianluigi PloS one Forloni. 2008. 'DJ-1 modulates α-synuclein aggregation state in a cellular model of oxidative stress: relevance for Parkinson's disease and involvement of HSP70', 3: e1884. Batelli, Sara, Roberto William Invernizzi, Alessandro Negro, Eleonora Calcagno, Serena Rodilossi, Gianluigi Forloni, and Diego Neurodegenerative Diseases Albani. 2015. 'The Parkinson's disease-related protein DJ-1 protects dopaminergic neurons in vivo and cultured cells from alpha-synuclein and 6-hydroxydopamine toxicity', 15: 13-23. Bender, Andreas, Paula Desplats, Brian Spencer, Edward Rockenstein, Anthony Adame, Matthias Elstner, Christoph Laub, Sarina Mueller, Andrew O Koob, and Michael PloS one Mante. 2013. 'TOM40 mediates mitochondrial dysfunction induced by α-synuclein accumulation in Parkinson’s disease', 8: e62277. Bender, Andreas, Kim J Krishnan, Christopher M Morris, Geoffrey A Taylor, Amy K Reeve, Robert H Perry, Evelyn Jaros, Joshua S Hersheson, Joanne Betts, and Thomas Nature genetics Klopstock. 2006. 'High levels of mitochondrial DNA deletions in substantia nigra neurons in aging and Parkinson disease', 38: 515. Bennett, M Catherine, John F Bishop, Yan Leng, P Boon Chock, Thomas N Chase, and M Maral Journal of Biological Chemistry Mouradian. 1999. 'Degradation of α-synuclein by proteasome', 274: 33855-58. Benz, Fabian, Christoph Roderburg, David Vargas Cardenas, Mihael Vucur, Jérémie Gautheron, Alexander Koch, Henning Zimmermann, Jörn Janssen, Lukas Nieuwenhuijsen, Mark Experimental Luedde, and molecular medicine. 2013. 'U6 is unsuitable for normalization of serum miRNA levels in patients with sepsis or liver fibrosis', 45: e42. Bernstein, Emily, Amy A Caudy, Scott M Hammond, and Gregory J Nature Hannon. 2001. 'Role for a bidentate ribonuclease in the initiation step of RNA interference', 409: 363. Bertler, Å, and E Rosengren. 1959. 'Occurrence and distribution of dopamine in brain and other tissues', Experientia, 15: 10-11. Bertolin, Giulia, Rosa Ferrando-Miguel, Maxime Jacoupy, Sabine Traver, Karl Grenier, Andrew W Greene, Aurélien Dauphin, François Waharte, Aurélien Bayot, and Jean Autophagy Salamero. 2013. 'The TOMM machinery is a molecular switch in PINK1 and PARK2/PARKIN- dependent mitochondrial clearance', 9: 1801-17. Bianco, C Lo, JL Ridet, BL Schneider, N Deglon, and P Proceedings of the National Academy of Sciences Aebischer. 2002. 'α-Synucleinopathy and selective dopaminergic neuron loss in a rat lentiviral-based model of Parkinson's disease', 99: 10813-18. Bianco, Christophe Lo, Bernard L Schneider, Matthias Bauer, Ali Sajadi, Alexis Brice, Takeshi Iwatsubo, and Patrick Proceedings of the National Academy of Sciences Aebischer. 2004. 'Lentiviral vector delivery of parkin prevents dopaminergic degeneration in an α-synuclein rat model of Parkinson's disease', 101: 17510-15.

205

Biedler, June L, Lawrence Helson, and Barbara A Cancer research Spengler. 1973. 'Morphology and growth, tumorigenicity, and cytogenetics of human neuroblastoma cells in continuous culture', 33: 2643-52. Biedler, June L, Suzanne Roffler-Tarlov, Melitta Schachner, and Lewis S Cancer research Freedman. 1978. 'Multiple neurotransmitter synthesis by human neuroblastoma cell lines and clones', 38: 3751-57. Biffi, Emilia, Giulia Regalia, Andrea Menegon, Giancarlo Ferrigno, and Alessandra PloS one Pedrocchi. 2013. 'The influence of neuronal density and maturation on network activity of hippocampal cell cultures: a methodological study', 8: e83899. Birkmayer, Walther, and Oleh Hornykiewicz. 1961. '[The L-3, 4-dioxyphenylalanine (DOPA)-effect in Parkinson-akinesia.]', Wiener klinische Wochenschrift, 73: 787-88. Björklund, Lars M, and Ole Isacson. 2002. 'Regulation of dopamine cell type and transmitter function in fetal and stem cell transplantation for Parkinson's disease.' in, Progress in brain research (Elsevier). Blaess, Sandra, and Siew-Lan Wiley Interdisciplinary Reviews: Developmental Biology Ang. 2015. 'Genetic control of midbrain dopaminergic neuron development', 4: 113-34. Blocq, Paul, and Gheorghe Marinesco. 1893. Sur un cas de tremblement parkinsonien hémiplégique: symptomatique d'une tumeur du pédoncule cérébral (G. Masson). Blum-Degena, Doris, Thomas Müller, Wilfried Kuhn, Manfred Gerlach, Horst Przuntek, and Peter Neuroscience letters Riederer. 1995. 'Interleukin-1β and interleukin-6 are elevated in the cerebrospinal fluid of Alzheimer's and de novo Parkinson's disease patients', 202: 17-20. Bonifati, Vincenzo, Patrizia Rizzu, Marijke J Van Baren, Onno Schaap, Guido J Breedveld, Elmar Krieger, Marieke CJ Dekker, Ferdinando Squitieri, Pablo Ibanez, and Marijke Science Joosse. 2003. 'Mutations in the DJ-1 gene associated with autosomal recessive early-onset parkinsonism', 299: 256-59. Bonneh-Barkay, Dafna, and Clayton A Brain pathology Wiley. 2009. 'Brain extracellular matrix in neurodegeneration', 19: 573-85. Borghi, Roberta, Roberta Marchese, Alessandro Negro, Lucio Marinelli, Gianluigi Forloni, Damiano Zaccheo, Giovanni Abbruzzese, and Massimo Neuroscience letters Tabaton. 2000. 'Full length α-synuclein is present in cerebrospinal fluid from Parkinson's disease and normal subjects', 287: 65-67. Borrageiro, Genevie, William Haylett, Soraya Seedat, Helena Kuivaniemi, and Soraya European Journal of Neuroscience Bardien. 2018. 'A review of genome-wide transcriptomics studies in Parkinson's disease', 47: 1-16. Boyer, Laurie A, Tong Ihn Lee, Megan F Cole, Sarah E Johnstone, Stuart S Levine, Jacob P Zucker, Matthew G Guenther, Roshan M Kumar, Heather L Murray, and Richard G cell Jenner. 2005. 'Core transcriptional regulatory circuitry in human embryonic stem cells', 122: 947-56. Braak, Heiko, Estifanos Ghebremedhin, Udo Rüb, Hansjürgen Bratzke, Kelly Cell Del Tredici, and tissue research. 2004. 'Stages in the development of Parkinson’s disease-related pathology', 318: 121-34. Braak, Heiko, U Rüb, WP Gai, and Kelly Journal of neural transmission Del Tredici. 2003. 'Idiopathic Parkinson's disease: possible routes by which vulnerable neuronal types may be subject to neuroinvasion by an unknown pathogen', 110: 517-36. Briscoe, J, L Sussel, P Serup, D Hartigan-O'Connor, TM Jessell, JLR Rubenstein, and J Nature Ericson. 1999. 'Homeobox gene Nkx2. 2 and specification of neuronal identity by graded Sonic hedgehog signalling', 398: 622. Broccoli, Vania, Serena G Giannelli, and Pietro G Frontiers in neuroscience Mazzara. 2014. 'Modeling physiological and pathological human neurogenesis in the dish', 8: 183. Brodacki, Bogdan, Jacek Staszewski, Beata Toczyłowska, Ewa Kozłowska, Nadzieja Drela, Małgorzata Chalimoniuk, and Adam Neuroscience letters Stępien. 2008. 'Serum interleukin (IL-2, IL-10,

206

IL-6, IL-4), TNFα, and INFγ concentrations are elevated in patients with atypical and idiopathic parkinsonism', 441: 158-62. Bruttger, Julia, Khalad Karram, Simone Wörtge, Tommy Regen, Federico Marini, Nicola Hoppmann, Matthias Klein, Thomas Blank, Simon Yona, and Yochai Immunity Wolf. 2015. 'Genetic cell ablation reveals clusters of local self-renewing microglia in the mammalian central nervous system', 43: 92-106. Bu, Bing, Xin Tong, Dechang Li, Yachong Hu, Wangxiao He, Chunyu Zhao, Rui Hu, Xiaoqing Li, Yongping Shao, and Cong ACS chemical neuroscience Liu. 2017. 'N-Terminal Acetylation Preserves α-Synuclein from Oligomerization by Blocking Intermolecular Hydrogen Bonds', 8: 2145-51. Bucciantini, Monica, Elisa Giannoni, Fabrizio Chiti, Fabiana Baroni, Lucia Formigli, Jesús Zurdo, Niccolo Taddei, Giampietro Ramponi, Christopher M Dobson, and Massimo nature Stefani. 2002. 'Inherent toxicity of aggregates implies a common mechanism for protein misfolding diseases', 416: 507. Buller, Ben, Xianshuang Liu, Xinli Wang, Rui L Zhang, Li Zhang, Ann Hozeska-Solgot, Michael Chopp, and Zheng G Zhang. 2010. 'MicroRNA-21 protects neurons from ischemic death', The FEBS journal, 277: 4299-307. Burgos, Kasandra, Ivana Malenica, Raghu Metpally, Amanda Courtright, Benjamin Rakela, Thomas Beach, Holly Shill, Charles Adler, Marwan Sabbagh, and Stephen PloS one Villa. 2014. 'Profiles of extracellular miRNA in cerebrospinal fluid and serum from patients with Alzheimer's and Parkinson's diseases correlate with disease status and features of pathology', 9: e94839. Burns, Jane C, Theodore Friedmann, Wolfgang Driever, Michelle Burrascano, and Jiing-Kuan Proceedings of the National Academy of Sciences Yee. 1993. 'Vesicular stomatitis virus G glycoprotein pseudotyped retroviral vectors: concentration to very high titer and efficient gene transfer into mammalian and nonmammalian cells', 90: 8033-37. Burré, Jacqueline, Manu Sharma, Theodoros Tsetsenis, Vladimir Buchman, Mark Etherton, and Thomas C Science Südhof. 2010. 'α-Synuclein promotes SNARE-complex assembly in vivo and in vitro': 1195227. Butcher, Nancy J, Tim-Rasmus Kiehl, Lili-Naz Hazrati, Eva WC Chow, Ekaterina Rogaeva, Anthony E Lang, and Anne S JAMA neurology Bassett. 2013. 'Association between early-onset Parkinson disease and 22q11. 2 deletion syndrome: identification of a novel genetic form of Parkinson disease and its clinical implications', 70: 1359-66. Butcher, Nancy J, Daniele Merico, Mehdi Zarrei, Lucas Ogura, Christian R Marshall, Eva WC Chow, Anthony E Lang, Stephen W Scherer, and Anne S PloS one Bassett. 2017. 'Whole-genome sequencing suggests mechanisms for 22q11. 2 deletion-associated Parkinson’s disease', 12: e0173944. Byers, Blake, Branden Cord, Ha Nam Nguyen, Birgitt Schüle, Lief Fenno, Patrick C Lee, Karl Deisseroth, J William Langston, Renee Reijo Pera, and Theo D PloS one Palmer. 2011. 'SNCA triplication Parkinson's patient's iPSC-derived DA neurons accumulate α-synuclein and are susceptible to oxidative stress', 6: e26159. Calabrese, Edward J Critical reviews in toxicology. 2008. 'Enhancing and regulating neurite outgrowth', 38: 391-418. Canet-Avilés, Rosa M, Mark A Wilson, David W Miller, Rili Ahmad, Chris McLendon, Sourav Bandyopadhyay, Melisa J Baptista, Dagmar Ringe, Gregory A Petsko, and Mark R Proceedings of the National Academy of Sciences Cookson. 2004. 'The Parkinson's disease protein DJ-1 is neuroprotective due to cysteine-sulfinic acid-driven mitochondrial localization', 101: 9103-08. Cao, Shuwen, David G Standaert, and Ashley S Journal of neuroinflammation Harms. 2012. 'The gamma chain subunit of Fc receptors is required for alpha-synuclein-induced pro- inflammatory signaling in microglia', 9: 259.

207

Cao, Xiang-Yang, Jing-Min Lu, Zhi-Qiang Zhao, Ming-Chao Li, Ting Lu, Xu-Sheng An, and Liu-Jun Neuroscience letters Xue. 2017. 'MicroRNA biomarkers of Parkinson’s disease in serum exosome-like microvesicles', 644: 94-99. Cardo, Lucía F, Eliecer Coto, Lorena de Mena, Renée Ribacoba, Germán Moris, Manuel Menéndez, and Victoria Journal of neurology Alvarez. 2013. 'Profile of microRNAs in the plasma of Parkinson’s disease patients and healthy controls', 260: 1420-22. Cardoso, Ana L, Joana R Guedes, Luís Pereira de Almeida, and Maria C Immunology Pedroso de Lima. 2012. 'miR-155 modulates microglia-mediated immune response by down-regulating SOCS-1 and promoting cytokine and nitric oxide production', 135: 73-88. Carlsson, A, M Lindquist, T Magnusson, and B Waldeck. 1958. 'Vol. 86 CATECHOL COMPOUNDS IN BRAINS 11', Science, 127: 471. Carlsson, Arvid, Margit Lindqvist, and TOR Magnusson. 1957. '3, 4-Dihydroxyphenylalanine and 5- hydroxytryptophan as reserpine antagonists'. Cartelli, Daniele, Alessandro Aliverti, Alberto Barbiroli, Carlo Santambrogio, Enzio M Ragg, Francesca VM Casagrande, Francesca Cantele, Silvia Beltramone, Jacopo Marangon, and Carmelita Scientific reports De Gregorio. 2016. 'α-Synuclein is a novel microtubule dynamase', 6: 33289. Cartier, Anna E, Kiren Ubhi, Brian Spencer, Ruben A Vazquez-Roque, Kori Ann Kosberg, Lawrence Fourgeaud, Priya Kanayson, Christina Patrick, Edward Rockenstein, and Gentry N PLoS One Patrick. 2012. 'Differential effects of UCHL1 modulation on alpha-synuclein in PD-like models of alpha-synucleinopathy', 7: e34713. Chambers, Ian, Douglas Colby, Morag Robertson, Jennifer Nichols, Sonia Lee, Susan Tweedie, and Austin Cell Smith. 2003. 'Functional expression cloning of Nanog, a pluripotency sustaining factor in embryonic stem cells', 113: 643-55. Chambers, Stuart M, Christopher A Fasano, Eirini P Papapetrou, Mark Tomishima, Michel Sadelain, and Lorenz Nature biotechnology Studer. 2009. 'Highly efficient neural conversion of human ES and iPS cells by dual inhibition of SMAD signaling', 27: 275. Chandra, Sreeganga, Francesco Fornai, Hyung-Bae Kwon, Umar Yazdani, Deniz Atasoy, Xinran Liu, Robert E Hammer, Giuseppe Battaglia, Dwight C German, and Pablo E Proceedings of the National Academy of Sciences Castillo. 2004. 'Double-knockout mice for α-and β-synucleins: effect on synaptic functions', 101: 14966-71. Chandra, Sreeganga, Gilbert Gallardo, Rafael Fernández-Chacón, Oliver M Schlüter, and Thomas C Cell Südhof. 2005. 'α-Synuclein cooperates with CSPα in preventing neurodegeneration', 123: 383-96. Chapman, Mary Ann Medical hypotheses. 2014. 'Interactions between cell adhesion and the synaptic vesicle cycle in Parkinson’s disease', 83: 203-07. Chartier-Harlin, Marie-Christine, Jennifer Kachergus, Christophe Roumier, Vincent Mouroux, Xavier Douay, Sarah Lincoln, Clotilde Levecque, Lydie Larvor, Joris Andrieux, and Mary Hulihan. 2004. 'α-Synuclein locus duplication as a cause of familial Parkinson's disease', The Lancet, 364: 1167-69. Chen, Edward Y, Christopher M Tan, Yan Kou, Qiaonan Duan, Zichen Wang, Gabriela Vaz Meirelles, Neil R Clark, and Avi BMC bioinformatics Ma’ayan. 2013. 'Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool', 14: 128. Chen, Liu-Ji, Yan-Qin Gao, Xue-Jun Li, Di-Han Shen, and Feng-Yan Journal of pineal research Sun. 2005. 'Melatonin protects against MPTP/MPP+-induced mitochondrial DNA oxidative damage in vivo and in vitro', 39: 34-42. Chen, Xi, Hongwei Liang, Danping Guan, Cheng Wang, Xiaoyun Hu, Lin Cui, Sidi Chen, Chunni Zhang, Junfeng Zhang, and Ke PloS one Zen. 2013. 'A combination of Let-7d, Let-7g and Let-7i serves as a stable reference for normalization of serum microRNAs', 8: e79652.

208

Cherra III, Salvatore J, Erin Steer, Aaron M Gusdon, Kirill Kiselyov, and Charleen T The American journal of pathology Chu. 2013. 'Mutant LRRK2 elicits calcium imbalance and depletion of dendritic mitochondria in neurons', 182: 474-84. Chiba, Shunmei, Young Mook Lee, Wenbo Zhou, and Curt R Stem Cells Freed. 2008. 'Noggin enhances dopamine neuron production from human embryonic stem cells and improves behavioral outcome after transplantation into Parkinsonian rats', 26: 2810-20. Chinta, Shankar J, Jyothi K Mallajosyula, Anand Rane, and Julie K Neuroscience letters Andersen. 2010. 'Mitochondrial alpha-synuclein accumulation impairs complex I function in dopaminergic neurons and results in increased mitophagy in vivo', 486: 235-39. Cho, Hyun Jin, Guoxiang Liu, Seok Min Jin, Loukia Parisiadou, Chengsong Xie, Jia Yu, Lixin Sun, Bo Ma, Jinhui Ding, and Renée Human molecular genetics Vancraenenbroeck. 2012. 'MicroRNA-205 regulates the expression of Parkinson's disease-related leucine-rich repeat kinase 2 protein', 22: 608-20. Choi, Won-Seok, LMT Canzoniero, SL Sensi, Karen L O'Malley, Byung J Gwag, Seonghyang Sohn, Ji- Eun Kim, Tae H Oh, Eunhee B Lee, and Young J Experimental neurology Oh. 1999. 'Characterization of MPP+-induced cell death in a dopaminergic neuronal cell line: role of macromolecule synthesis, cytosolic calcium, caspase, and Bcl-2-related proteins', 159: 274- 82. Chu, Yaping, Gerardo A Morfini, Lori B Langhamer, Yinzhen He, Scott T Brady, and Jeffrey H Brain Kordower. 2012. 'Alterations in axonal transport motor proteins in sporadic and experimental Parkinson’s disease', 135: 2058-73. Chung, Chee Yeun, James B Koprich, Hasan Siddiqi, and Ole Journal of Neuroscience Isacson. 2009. 'Dynamic changes in presynaptic and axonal transport proteins combined with striatal neuroinflammation precede dopaminergic neuronal loss in a rat model of AAV α- synucleinopathy', 29: 3365-73. Chung, Sangmi, Amanda Leung, Baek-Soo Han, Mi-Yoon Chang, Jung-Il Moon, Chun-Hyung Kim, Sunghoi Hong, Jan Pruszak, Ole Isacson, and Kwang-Soo Cell stem cell Kim. 2009. 'Wnt1- lmx1a forms a novel autoregulatory loop and controls midbrain dopaminergic differentiation synergistically with the SHH-FoxA2 pathway', 5: 646-58. Chung, Sun Young, Sarah Kishinevsky, Joseph R Mazzulli, John Graziotto, Ana Mrejeru, Eugene V Mosharov, Lesly Puspita, Parvin Valiulahi, David Sulzer, and Teresa A Stem cell reports Milner. 2016. 'Parkin and PINK1 patient iPSC-derived midbrain dopamine neurons exhibit mitochondrial dysfunction and α-synuclein accumulation', 7: 664-77. Colla, Emanuela, Poul H Jensen, Olga Pletnikova, Juan C Troncoso, Charles Glabe, and Michael K Journal of Neuroscience Lee. 2012. 'Accumulation of toxic α-synuclein oligomer within endoplasmic reticulum occurs in α-synucleinopathy in vivo', 32: 3301-05. Collino, Federica, Maria Chiara Deregibus, Stefania Bruno, Luca Sterpone, Giulia Aghemo, Laura Viltono, Ciro Tetta, and Giovanni PloS one Camussi. 2010. 'Microvesicles derived from adult human bone marrow and tissue specific mesenchymal stem cells shuttle selected pattern of miRNAs', 5: e11803. Conesa, Ana, Pedro Madrigal, Sonia Tarazona, David Gomez-Cabrero, Alejandra Cervera, Andrew McPherson, Michał Wojciech Szcześniak, Daniel J Gaffney, Laura L Elo, and Xuegong Zhang. 2016. 'A survey of best practices for RNA-seq data analysis', Genome biology, 17: 13. Connolly, Barbara S, and Anthony E Jama Lang. 2014. 'Pharmacological treatment of Parkinson disease: a review', 311: 1670-83. Consiglio, Antonella, Angela Gritti, Diego Dolcetta, Antonia Follenzi, Claudio Bordignon, Fred H Gage, Angelo Luigi Vescovi, and Luigi Proceedings of the National Academy of Sciences Naldini. 2004. 'Robust in vivo gene transfer into adult mammalian neural stem cells by lentiviral vectors', 101: 14835-40.

209

Constantinescu, R, AT Constantinescu, H Reichmann, and B Janetzky. 2007. 'Neuronal differentiation and long-term culture of the human neuroblastoma line SH-SY5Y.' in, Neuropsychiatric disorders an integrative approach (Springer). Conway, Kelly A, James D Harper, and Peter T Biochemistry Lansbury. 2000. 'Fibrils formed in vitro from α-synuclein and two mutant forms linked to Parkinson's disease are typical amyloid', 39: 2552-63. Cooper-Knock, Johnathan, Janine Kirby, Laura Ferraiuolo, Paul R Heath, Magnus Rattray, and Pamela J Nature Reviews Neurology Shaw. 2012. 'Gene expression profiling in human neurodegenerative disease', 8: 518. Costa-Silva, Juliana, Douglas Domingues, and Fabricio Martins PloS one Lopes. 2017. 'RNA-Seq differential expression analysis: An extended review and a software tool', 12: e0190152. Cotzias, George C, Melvin H Van Woert, and Lewis M Schiffer. 1967. 'Aromatic amino acids and modification of parkinsonism', New England Journal of Medicine, 276: 374-79. Crawford, T Quinn, and Henk Developmental dynamics: an official publication of the American Association of Anatomists Roelink. 2007. 'The notch response inhibitor DAPT enhances neuronal differentiation in embryonic stem cell-derived embryoid bodies independently of sonic hedgehog signaling', 236: 886-92. Cuellar, Trinna L, Tigwa H Davis, Peter T Nelson, Gabriel B Loeb, Brian D Harfe, Erik Ullian, and Michael T Proceedings of the National Academy of Sciences McManus. 2008. 'Dicer loss in striatal neurons produces behavioral and neuroanatomical phenotypes in the absence of neurodegeneration', 105: 5614-19. Cuervo, Ana Maria, Leonidas Stefanis, Ross Fredenburg, Peter T Lansbury, and David Science Sulzer. 2004. 'Impaired degradation of mutant α-synuclein by chaperone-mediated autophagy', 305: 1292-95. Cui, Tao, Chunxiang Fan, Li Gu, Hua Gao, Qi Liu, Tao Zhang, Zhifeng Qi, Chunli Zhao, Huanying Zhao, and Qing Brain research Cai. 2011. 'Silencing of PINK1 induces mitophagy via mitochondrial permeability transition in dopaminergic MN9D cells', 1394: 1-13. Cullen, D Kacy, Meghan E Gilroy, Hillary R Irons, and Michelle C Brain research LaPlaca. 2010. 'Synapse-to-neuron ratio is inversely related to neuronal density in mature neuronal cultures', 1359: 44-55. Cummings, JL European Neurology. 1988. 'The dementias of Parkinson's disease: prevalence, characteristics, neurobiology, and comparison with dementia of the Alzheimer type', 28: 15- 23. Członkowska, Anna, Małgorzata Kohutnicka, Iwona Kurkowska-Jastrzębska, and Andrzej Neurodegeneration Członkowski. 1996. 'Microglial reaction in MPTP (1-methyl-4-phenyl-1, 2, 3, 6-tetrahydropyridine) induced Parkinson's disease mice model', 5: 137-43. Daher, João Paulo L, Mingyao Ying, Rebecca Banerjee, Rebecca S McDonald, Myriam Dumas Hahn, Lichuan Yang, M Flint Beal, Bobby Thomas, Valina L Dawson, and Ted M Molecular neurodegeneration Dawson. 2009. 'Conditional transgenic mice expressing C-terminally truncated human α-synuclein (αSyn119) exhibit reduced striatal dopamine without loss of nigrostriatal pathway dopaminergic neurons', 4: 34. Dahlström, Annica, and Kjell Fuxe. 1964. 'Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brain stem neurons', Acta Physiologica Scandinavica. Supplementum: SUPPL 232: 1. Dansithong, Warunee, Sharan Paul, Karla P Figueroa, Marc D Rinehart, Shaina Wiest, Lance T Pflieger, Daniel R Scoles, and Stefan M PLoS genetics Pulst. 2015. 'Ataxin-2 regulates RGS8 translation in a new BAC-SCA2 transgenic mouse model', 11: e1005182. Danzer, Karin M, Lisa R Kranich, Wolfgang P Ruf, Ozge Cagsal-Getkin, Ashley R Winslow, Liya Zhu, Charles R Vanderburg, and Pamela J Molecular neurodegeneration McLean. 2012. 'Exosomal cell-to-cell transmission of alpha synuclein oligomers', 7: 42.

210

Daubner, S Colette, Tiffany Le, Shanzhi Archives of biochemistry Wang, and biophysics. 2011. 'Tyrosine hydroxylase and regulation of dopamine synthesis', 508: 1-12. Dauer, William, and Serge Neuron Przedborski. 2003. 'Parkinson's disease: mechanisms and models', 39: 889-909. Decressac, Mickael, Bengt Mattsson, Martin Lundblad, P Weikop, and Anders Neurobiology of disease Björklund. 2012. 'Progressive neurodegenerative and behavioural changes induced by AAV-mediated overexpression of α-synuclein in midbrain dopamine neurons', 45: 939-53. del Arco, Araceli, and Jorgina Journal of Biological Chemistry Satrústegui. 1998. 'Molecular cloning of Aralar, a new member of the mitochondrial carrier superfamily that binds calcium and is present in human muscle and brain', 273: 23327-34. Desplats, Paula, He-Jin Lee, Eun-Jin Bae, Christina Patrick, Edward Rockenstein, Leslie Crews, Brian Spencer, Eliezer Masliah, and Seung-Jae Proceedings of the National Academy of Sciences Lee. 2009. 'Inclusion formation and neuronal cell death through neuron-to-neuron transmission of α-synuclein', 106: 13010-15. Dettmer, Ulf, Andrew J Newman, Frank Soldner, Eric S Luth, Nora C Kim, Victoria E Von Saucken, John B Sanderson, Rudolf Jaenisch, Tim Bartels, and Dennis Nature communications Selkoe. 2015. 'Parkinson-causing α-synuclein missense mutations shift native tetramers to monomers as a mechanism for disease initiation', 6: 7314. Devine, Michael J, Mina Ryten, Petr Vodicka, Alison J Thomson, Tom Burdon, Henry Houlden, Fatima Cavaleri, Masumi Nagano, Nicola J Drummond, and Jan-Willem Nature communications Taanman. 2011. 'Parkinson's disease induced pluripotent stem cells with triplication of the α-synuclein locus', 2: 440. Devoto, Victorio M Pozo, Tomas L Disease models Falzone, and mechanisms. 2017. 'Mitochondrial dynamics in Parkinson's disease: a role for α-synuclein?', 10: 1075-87. Di Donato, Nataliya, Ying Y Jean, A Murat Maga, Briana D Krewson, Alison B Shupp, Maria I Avrutsky, Achira Roy, Sarah Collins, Carissa Olds, and Rebecca A The American Journal of Human Genetics Willert. 2016. 'Mutations in CRADD result in reduced caspase-2-mediated neuronal apoptosis and cause megalencephaly with a rare lissencephaly variant', 99: 1117-29. Diao, Jiajie, Jacqueline Burré, Sandro Vivona, Daniel J Cipriano, Manu Sharma, Minjoung Kyoung, Thomas C Südhof, and Axel T Elife Brunger. 2013. 'Native α-synuclein induces clustering of synaptic-vesicle mimics via binding to phospholipids and synaptobrevin-2/VAMP2', 2: e00592. Ding, Haixia, Zhen Huang, Mengjie Chen, Cheng Wang, Xi Chen, Jiangning Chen, Junfeng Parkinsonism Zhang, and related disorders. 2016. 'Identification of a panel of five serum miRNAs as a biomarker for Parkinson's disease', 22: 68-73. Ding, Huiping, Naomi S Fineberg, Michelle Gray, and Talene A Journal of Molecular Neuroscience Yacoubian. 2013. 'α-Synuclein overexpression represses 14-3-3θ transcription', 51: 1000-09. Ding, Meng, Cheng Wang, Xiaolan Lu, Cuiping Zhang, Zhen Zhou, Xi Chen, Chen-Yu Zhang, Ke Zen, Chunni Analytical Zhang, and bioanalytical chemistry. 2018. 'Comparison of commercial exosome isolation kits for circulating exosomal microRNA profiling', 410: 3805-14. Dittgen, Tanjew, Axel Nimmerjahn, Shoji Komai, Pawel Licznerski, Jack Waters, Troy W Margrie, Fritjof Helmchen, Winfried Denk, Michael Brecht, and Pavel Proceedings of the National Academy of Sciences Osten. 2004. 'Lentivirus-based genetic manipulations of cortical neurons and their optical and electrophysiological monitoring in vivo', 101: 18206-11. Dobin, Alexander, Carrie A Davis, Felix Schlesinger, Jorg Drenkow, Chris Zaleski, Sonali Jha, Philippe Batut, Mark Chaisson, and Thomas R Gingeras. 2013. 'STAR: ultrafast universal RNA-seq aligner', Bioinformatics, 29: 15-21. Dougherty, Joseph D Scientific reports. 2017. 'Generation and characterization of a mouse line for monitoring translation in dopaminergic neurons', 7: 8117. Doxakis, Epaminondas. 2010. 'Post-transcriptional regulation of α-synuclein expression by mir-7 and mir-153', Journal of Biological Chemistry, 285: 12726-34.

211

Doyle, Joseph P, Joseph D Dougherty, Myriam Heiman, Eric F Schmidt, Tanya R Stevens, Guojun Ma, Sujata Bupp, Prerana Shrestha, Rajiv D Shah, and Martin L Cell Doughty. 2008. 'Application of a translational profiling approach for the comparative analysis of CNS cell types', 135: 749-62. Dumitriu, Alexandra, Javad Golji, Adam T Labadorf, Benbo Gao, Thomas G Beach, Richard H Myers, Kenneth A Longo, and Jeanne C BMC medical genomics Latourelle. 2015. 'Integrative analyses of proteomics and RNA transcriptomics implicate mitochondrial processes, protein folding pathways and GWAS loci in Parkinson disease', 9: 5. Dumitriu, Alexandra, Jeanne C Latourelle, Tiffany C Hadzi, Nathan Pankratz, Dan Garza, John P Miller, Jeffery M Vance, Tatiana Foroud, Thomas G Beach, and Richard H PLoS genetics Myers. 2012. 'Gene expression profiles in Parkinson disease prefrontal cortex implicate FOXO1 and genes under its transcriptional regulation', 8: e1002794. Duvoisin, Roger C, Roswell Eldridge, Adrian Williams, John Nutt, and Donald Calne. 1981. 'Twin study of Parkinson disease', Neurology, 31: 77-77. Echelard, Yann, Douglas J Epstein, Benoit St-Jacques, Liya Shen, Jym Mohler, Jill A McMahon, and Andrew P Cell McMahon. 1993. 'Sonic hedgehog, a member of a family of putative signaling molecules, is implicated in the regulation of CNS polarity', 75: 1417-30. Ehringer, Hornykiewicz, and O Hornykiewicz. 1960. 'Verteilung von Noradrenalin und Dopamin (3- Hydroxytyramin) im Gehirn des Menschen und ihr Verhalten bei Erkrankungen des extrapyramidalen Systems', Klinische Wochenschrift, 38: 1236-39. El-Agnaf, Omar MA, Sultan A Salem, Katerina E Paleologou, Leanne J Cooper, Nigel J Fullwood, Mark J Gibson, Martin D Curran, JENNIFER A COURT, DAVID MA MANN, and SHU-ICHI The FASEB Journal IKEDA. 2003. 'α-Synuclein implicated in Parkinson’s disease is present in extracellular biological fluids, including human plasma', 17: 1945-47. Elbaz, A, F Grigoletto, M Baldereschi, MM Breteler, JM Manubens-Bertran, S Lopez-Pousa, JF Dartigues, A Alperovitch, C Tzourio, and WA Rocca. 1999. 'Familial aggregation of Parkinson’s disease A population-based case-control study in Europe', Neurology, 52: 1876- 76. Eldh, Maria, Jan Lötvall, Carina Malmhäll, and Karin Molecular immunology Ekström. 2012. 'Importance of RNA isolation methods for analysis of exosomal RNA: evaluation of different methods', 50: 278-86. Emmanouilidou, Evangelia, Katerina Melachroinou, Theodoros Roumeliotis, Spiros D Garbis, Maria Ntzouni, Lukas H Margaritis, Leonidas Stefanis, and Kostas Journal of Neuroscience Vekrellis. 2010. 'Cell-produced α-synuclein is secreted in a calcium-dependent manner by exosomes and impacts neuronal survival', 30: 6838-51. Ender, Christine, and Gunter J Cell Sci Meister. 2010. 'Argonaute proteins at a glance', 123: 1819-23. Epstein, Douglas J, Andrew P McMahon, and Alexandra L Development Joyner. 1999. 'Regionalization of Sonic hedgehog transcription along the anteroposterior axis of the mouse central nervous system is regulated by Hnf3-dependent and-independent mechanisms', 126: 281-92. Erickson, Jeffery T, Teresa A Brosenitsch, and David M Journal of Neuroscience Katz. 2001. 'Brain- derived neurotrophic factor and glial cell line-derived neurotrophic factor are required simultaneously for survival of dopaminergic primary sensory neurons in vivo', 21: 581-89. Esteves, A Raquel, Daniela M Arduíno, Russell H Swerdlow, Catarina R Oliveira, Sandra M Antioxidants Cardoso, and redox signaling. 2009. 'Oxidative stress involvement in α- synuclein oligomerization in Parkinson's disease cybrids', 11: 439-48. Evans, Martin J, and Matthew H nature Kaufman. 1981. 'Establishment in culture of pluripotential cells from mouse embryos', 292: 154. Fabbri, Muller, Alessio Paone, Federica Calore, Roberta Galli, Eugenio Gaudio, Ramasamy Santhanam, Francesca Lovat, Paolo Fadda, Charlene Mao, and Gerard J Nuovo. 2012.

212

'MicroRNAs bind to Toll-like receptors to induce prometastatic inflammatory response', Proceedings of the National Academy of Sciences, 109: E2110-E16. Falck, B, N-Å Hillarp, Gr Thieme, and A al Torp. 1962. 'Fluorescence of catechol amines and related compounds condensed with formaldehyde', Journal of Histochemistry & Cytochemistry, 10: 348-54. Fan, Yuxin, Pornprot Limprasert, Ian VJ Murray, Annette C Smith, Virginia M-Y Lee, John Q Trojanowski, Bryce L Sopher, and Albert R Human molecular genetics La Spada. 2006. 'β- synuclein modulates α-synuclein neurotoxicity by reducing α-synuclein protein expression', 15: 3002-11. Farrer, Matt, Piu Chan, Rong Chen, Louis Tan, Sarah Lincoln, Dena Hernandez, Lysia Forno, Katrina Gwinn-Hardy, Leonard Petrucelli, Jennifer Annals of Neurology: Official Journal of the American Neurological Association Hussey, and the Child Neurology Society. 2001. 'Lewy bodies and parkinsonism in families with parkin mutations', 50: 293-300. Fearnley, Julian M, and Andrew J Brain Lees. 1991. 'Ageing and Parkinson's disease: substantia nigra regional selectivity', 114: 2283-301. Feinstein, Yael, Víctor Borrell, Cristina Garcia, Tal Burstyn-Cohen, Vered Tzarfaty, Ayala Frumkin, Akinao Nose, Hitoshi Okamoto, Shin-ichi Higashijima, and Eduardo Development Soriano. 1999. 'F-spondin and mindin: two structurally and functionally related genes expressed in the hippocampus that promote outgrowth of embryonic hippocampal neurons', 126: 3637- 48. Fellner, Lisa, Regina Irschick, Kathrin Schanda, Markus Reindl, Lars Klimaschewski, Werner Poewe, Gregor K Wenning, and Nadia Glia Stefanova. 2013. 'Toll-like receptor 4 is required for α- synuclein dependent activation of microglia and astroglia', 61: 349-60. Feng, Yinling, and Xuefeng Molecular medicine reports Wang. 2017. 'Systematic analysis of microarray datasets to identify Parkinson's disease‑associated pathways and genes', 15: 1252-62. Ferri, Anna LM, Wei Lin, Yannis E Mavromatakis, Julie C Wang, Hiroshi Sasaki, Jeffrey A Whitsett, and Siew-Lan Development Ang. 2007. 'Foxa1 and Foxa2 regulate multiple phases of midbrain dopaminergic neuron development in a dosage-dependent manner', 134: 2761-69. Fevrier, Benoit, Didier Vilette, Fabienne Archer, Damarys Loew, Wolfgang Faigle, Michel Vidal, Hubert Laude, and Graça Proceedings of the National Academy of Sciences Raposo. 2004. 'Cells release prions in association with exosomes', 101: 9683-88. Fleming, Sheila M, Nicole A Tetreault, Caitlin K Mulligan, Ché B Hutson, Eliezer Masliah, and Marie- Françoise European Journal of Neuroscience Chesselet. 2008. 'Olfactory deficits in mice overexpressing human wildtype α-synuclein', 28: 247-56. Foix, Ch, and J Nicolesco. 1925. Les noyaux gris centraux et la région mésencéphalo-sous-optique: suivi d'un appendice sur l'anatomie pathologique de la maladie de Parkinson (Massons). Fujita, Masayo, Shuei Sugama, Kazunari Sekiyama, Akio Sekigawa, Tohru Tsukui, Masaaki Nakai, Masaaki Waragai, Takato Takenouchi, Yoshiki Takamatsu, and Jianshe Nature Communications Wei. 2010. 'A β-synuclein mutation linked to dementia produces neurodegeneration when expressed in mouse brain', 1: 110. Fujiwara, Hideo, Masato Hasegawa, Naoshi Dohmae, Akiko Kawashima, Eliezer Masliah, Matthew S Goldberg, Jie Shen, Koji Takio, and Takeshi Nature cell biology Iwatsubo. 2002. 'α-Synuclein is phosphorylated in synucleinopathy lesions', 4: 160. Funayama, Manabu, Kazuko Hasegawa, Hisayuki Kowa, Masaaki Saito, Shoji Tsuji, Fumiya Annals of Neurology: Official Journal of the American Neurological Association Obata, and the Child Neurology Society. 2002. 'A new locus for Parkinson's disease (PARK8) maps to chromosome 12p11. 2–q13. 1', 51: 296-301. Funk, Casimir. 1911. 'Synthesis of dl-3: 4-dihydroxyphenylalanine', Journal of the Chemical Society, Transactions, 99: 554-57.

213

Gainetdinov, Raul R, Fabio Fumagalli, Sara R Jones, and Marc G Journal of neurochemistry Caron. 1997. 'Dopamine transporter is required for in vivo MPTP neurotoxicity: evidence from mice lacking the transporter', 69: 1322-25. Gallo, Alessia, Mayank Tandon, Ilias Alevizos, and Gabor G PloS one Illei. 2012. 'The majority of microRNAs detectable in serum and saliva is concentrated in exosomes', 7: e30679. Garger, SJ, OM Griffith, LK Biochemical Grill, and biophysical research communications. 1983. 'Rapid purification of plasmid DNA by a single centrifugation in a two-step cesium chloride- ethidium bromide gradient', 117: 835-42. Geisler, Sven, Kira M Holmström, Diana Skujat, Fabienne C Fiesel, Oliver C Rothfuss, Philipp J Kahle, and Wolfdieter Nature cell biology Springer. 2010. 'PINK1/Parkin-mediated mitophagy is dependent on VDAC1 and p62/SQSTM1', 12: 119. Gentleman, Robert C, Vincent J Carey, Douglas M Bates, Ben Bolstad, Marcel Dettling, Sandrine Dudoit, Byron Ellis, Laurent Gautier, Yongchao Ge, and Jeff Genome biology Gentry. 2004. 'Bioconductor: open software development for computational biology and bioinformatics', 5: R80. Gerhard, Alexander, Nicola Pavese, Gary Hotton, Federico Turkheimer, Meltem Es, Alexander Hammers, Karla Eggert, Wolfgang Oertel, Richard B Banati, and David J Neurobiology of disease Brooks. 2006. 'In vivo imaging of microglial activation with [11C](R)-PK11195 PET in idiopathic Parkinson's disease', 21: 404-12. Gerrard, Lesley, Leigh Rodgers, and Wei Stem cells Cui. 2005. 'Differentiation of human embryonic stem cells to neural lineages in adherent culture by blocking bone morphogenetic protein signaling', 23: 1234-41. Giardine, Belinda, Cathy Riemer, Ross C Hardison, Richard Burhans, Laura Elnitski, Prachi Shah, Yi Zhang, Daniel Blankenberg, Istvan Albert, and James Genome research Taylor. 2005. 'Galaxy: a platform for interactive large-scale genome analysis', 15: 1451-55. Giasson, Benoit I, John E Duda, Shawn M Quinn, Bin Zhang, John Q Trojanowski, and Virginia M-Y Neuron Lee. 2002. 'Neuronal α-synucleinopathy with severe movement disorder in mice expressing A53T human α-synuclein', 34: 521-33. Gillardon, Frank, Matthias Mack, Wolfgang Rist, Cathrin Schnack, Martin Lenter, Tobias Hildebrandt, and Bastian Hengerer. 2008. 'MicroRNA and proteome expression profiling in early- symptomatic α-synuclein (A30P)-transgenic mice', Proteomics-Clinical Applications, 2: 697- 705. Gispert, Suzana, Domenico Del Turco, Lisa Garrett, Amy Chen, David J Bernard, John Hamm-Clement, Horst-Werner Korf, Thomas Deller, Heiko Braak, Georg Molecular Auburger, and Cellular Neuroscience. 2003. 'Transgenic mice expressing mutant A53T human alpha-synuclein show neuronal dysfunction in the absence of aggregate formation', 24: 419-29. Gitler, Aaron D, Brooke J Bevis, James Shorter, Katherine E Strathearn, Shusei Hamamichi, Linhui Julie Su, Kim A Caldwell, Guy A Caldwell, Jean-Christophe Rochet, and J Michael Proceedings of the National Academy of Sciences McCaffery. 2008. 'The Parkinson's disease protein α- synuclein disrupts cellular Rab homeostasis', 105: 145-50. Goecks, Jeremy, Anton Nekrutenko, and James Taylor. 2010. 'Galaxy: a comprehensive approach for supporting accessible, reproducible, and transparent computational research in the life sciences', Genome biology, 11: R86. Golbe, Lawrence I, Giuseppe Di Iorio, Vincenzo Bonavita, Douglas C Miller, and Roger C Duvoisin. 1990. 'A large kindred with autosomal dominant Parkinson's disease', Annals of neurology, 27: 276-82. Golbe, Lawrence I, Alice M Lazzarini, Roger C Duvoisin, Giuseppe Di Iorio, Giuseppe Sanges, Vincenzo Bonavita, and Salvatore la Sala. 1996. 'Clinical genetic analysis of Parkinson's disease in the Contursi kindred', Annals of neurology, 40: 767-75. Goldberg, Matthew S, Sheila M Fleming, James J Palacino, Carlos Cepeda, Hoa A Lam, Anushree Bhatnagar, Edward G Meloni, Nanping Wu, Larry C Ackerson, and Gloria J Journal of

214

Biological Chemistry Klapstein. 2003. 'Parkin-deficient mice exhibit nigrostriatal deficits but not loss of dopaminergic neurons'. Gomez-Suaga, Patricia, Berta Luzon-Toro, Dev Churamani, Ling Zhang, Duncan Bloor-Young, Sandip Patel, Philip G Woodman, Grant C Churchill, and Sabine Human molecular genetics Hilfiker. 2011. 'Leucine-rich repeat kinase 2 regulates autophagy through a calcium-dependent pathway involving NAADP', 21: 511-25. Gomez-Tortosa, E, K Newell, MC Irizarry, M Albert, JH Growdon, and BT Hyman. 1999. 'Clinical and quantitative pathologic correlates of dementia with Lewy bodies', Neurology, 53: 1284-84. Gowers, William Richard. 1901. A manual of diseases of the nervous system (P. Blakiston's Sons & Company). Grealish, Shane, Elsa Diguet, Agnete Kirkeby, Bengt Mattsson, Andreas Heuer, Yann Bramoulle, Nadja Van Camp, Anselme L Perrier, Philippe Hantraye, and Anders Cell stem cell Björklund. 2014. 'Human ESC-derived dopamine neurons show similar preclinical efficacy and potency to fetal neurons when grafted in a rat model of Parkinson’s disease', 15: 653-65. Greene, L. A., and A. S. Tischler. 1976. 'Establishment of a noradrenergic clonal line of rat adrenal pheochromocytoma cells which respond to nerve growth factor', Proc Natl Acad Sci U S A, 73: 2424-8. Greenfield, JG, Frances D Journal of neurology Bosanquet, neurosurgery,, and psychiatry. 1953. 'The brain-stem lesions in Parkinsonism', 16: 213. Gropp, Michal, Pavel Itsykson, Orna Singer, Tamir Ben-Hur, Etti Reinhartz, Eithan Galun, and Benjamin E Molecular Therapy Reubinoff. 2003. 'Stable genetic modification of human embryonic stem cells by lentiviral vectors', 7: 281-87. Grünblatt, E, Silvia Mandel, J Jacob-Hirsch, Sharon Zeligson, Ninette Amariglo, Gideon Rechavi, Jie Li, Rivka Ravid, Wolfgang Roggendorf, and Peter Journal of neural transmission Riederer. 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', 111: 1543-73. Gu, Haigang, Shan Ping Yu, Claire-Anne Gutekunst, Robert E Gross, Ling International journal of physiology Wei, pathophysiology, and pharmacology. 2013. 'Inhibition of the Rho signaling pathway improves neurite outgrowth and neuronal differentiation of mouse neural stem cells', 5: 11. Guardia-Laguarta, Cristina, Estela Area-Gomez, Cornelia Rüb, Yuhui Liu, Jordi Magrané, Dorothea Becker, Wolfgang Voos, Eric A Schon, and Serge Journal of Neuroscience Przedborski. 2014. 'α-Synuclein is localized to mitochondria-associated ER membranes', 34: 249-59. Guduric-Fuchs, Jasenka, Anna O’Connor, Bailey Camp, Christina L O'Neill, Reinhold J Medina, and David A Simpson. 2012. 'Selective extracellular vesicle-mediated export of an overlapping set of microRNAs from multiple cell types', BMC genomics, 13: 357. Guedes, Joana R, Carlos M Custódia, Ricardo J Silva, Luís P de Almeida, Maria C Pedroso de Lima, and Ana L Human molecular genetics Cardoso. 2014. 'Early miR-155 upregulation contributes to neuroinflammation in Alzheimer's disease triple transgenic mouse model', 23: 6286-301. Guggenheirn, M. 1913. 'Dioxyphenylalanin, eine neue Aminosäure ausVicia faba', Hoppe-Seyler’s Z Physiol Chem, 88: 276-84. Gui, YaXing, Hai Liu, LiShan Zhang, Wen Lv, and XingYue Oncotarget Hu. 2015. 'Altered microRNA profiles in cerebrospinal fluid exosome in Parkinson disease and Alzheimer disease', 6: 37043. Guo, Yan, Shilin Zhao, Fei Ye, Quanhu Sheng, and Yu BioMed research international Shyr. 2014. 'MultiRankSeq: multiperspective approach for RNAseq differential expression analysis and quality control', 2014. Guttman, Mitchell, Manuel Garber, Joshua Z Levin, Julie Donaghey, James Robinson, Xian Adiconis, Lin Fan, Magdalena J Koziol, Andreas Gnirke, and Chad Nature biotechnology Nusbaum.

215

2010. 'Ab initio reconstruction of cell type–specific transcriptomes in mouse reveals the conserved multi-exonic structure of lincRNAs', 28: 503. Guzman, Jaime N, Javier Sanchez-Padilla, David Wokosin, Jyothisri Kondapalli, Ema Ilijic, Paul T Schumacker, and D James Nature Surmeier. 2010. 'Oxidant stress evoked by pacemaking in dopaminergic neurons is attenuated by DJ-1', 468: 696. Haenseler, Walther, Federico Zambon, Heyne Lee, Jane Vowles, Federica Rinaldi, Galbha Duggal, Henry Houlden, Katrina Gwinn, Selina Wray, and Kelvin C Scientific reports Luk. 2017. 'Excess α-synuclein compromises phagocytosis in iPSC-derived macrophages', 7: 9003. Han, C, J Huang, K Ma, Y Shen, L Liu, G Zhang, X Xu, J Li, H Jiang, and S Movement Disorders Guo. 2016. 'exosomes-injection to reproduce a mouse model of Parkinson's disease: A preliminary report: 872', 31: S282. Han, Ye, Christina E Khodr, Mohan K Sapru, Jyothi Pedapati, and Martha C Brain research Bohn. 2011. 'A microRNA embedded AAV alpha-synuclein gene silencing vector for dopaminergic neurons', 1386: 15-24. Han, Zhaoli, Fanglian Chen, Xintong Ge, Jin Tan, Ping Lei, and Jianning Zhang. 2014. 'miR-21 alleviated apoptosis of cortical neurons through promoting PTEN-Akt signaling pathway in vitro after experimental traumatic brain injury', Brain research, 1582: 12-20. Hansen, Christian, Elodie Angot, Ann-Louise Bergström, Jennifer A Steiner, Laura Pieri, Gesine Paul, Tiago F Outeiro, Ronald Melki, Pekka Kallunki, and Karina The Journal of clinical investigation Fog. 2011. 'α-Synuclein propagates from mouse brain to grafted dopaminergic neurons and seeds aggregation in cultured human cells', 121: 715-25. Hansen, Christian, and Jia-Yi Trends in molecular medicine Li. 2012. 'Beyond α-synuclein transfer: pathology propagation in Parkinson's disease', 18: 248-55. Hara, Taichi, Kenji Nakamura, Makoto Matsui, Akitsugu Yamamoto, Yohko Nakahara, Rika Suzuki- Migishima, Minesuke Yokoyama, Kenji Mishima, Ichiro Saito, and Hideyuki Nature Okano. 2006. 'Suppression of basal autophagy in neural cells causes neurodegenerative disease in mice', 441: 885. Harding, Antony J, and Glenda M Halliday. 2001. 'Cortical Lewy body pathology in the diagnosis of dementia', Acta neuropathologica, 102: 355-63. Harrow, Jennifer, Adam Frankish, Jose M Gonzalez, Electra Tapanari, Mark Diekhans, Felix Kokocinski, Bronwen L Aken, Daniel Barrell, Amonida Zadissa, and Stephen Searle. 2012. 'GENCODE: the reference human genome annotation for The ENCODE Project', Genome research, 22: 1760-74. Hasegawa, K, M Funayama, N Matsuura, H Furusawa, F Sakai, H Kowa, and F European neurology Obata. 2001. 'Analysis of α-synuclein, parkin, tau, and UCH-L1 in a Japanese Family with Autosomal Dominant Parkinsonism', 46: 20-24. Hasegawa, Takafumi, Michiko Matsuzaki, Atsushi Takeda, Akio Kikuchi, Hirotoshi Akita, George Perry, Mark A Smith, and Yasuto Brain research Itoyama. 2004. 'Accelerated α-synuclein aggregation after differentiation of SH-SY5Y neuroblastoma cells', 1013: 51-59. Hashimoto, Makoto, Edward Rockenstein, Michael Mante, Margaret Mallory, and Eliezer Neuron Masliah. 2001. 'β-Synuclein inhibits α-synuclein aggregation: a possible role as an anti- parkinsonian factor', 32: 213-23. Hatano, Y, K Sato, B Elibol, H Yoshino, Y Yamamura, V Bonifati, H Shinotoh, M Asahina, S Kobayashi, and AR Neurology Ng. 2004. 'PARK6-linked autosomal recessive early-onset parkinsonism in Asian populations', 63: 1482-85. Hayakawa, Makio, Masahide Matsushima, Hiroshi Hagiwara, Toshiyuki Oshima, Tomofumi Fujino, Ken Ando, Kiyomi Kikugawa, Hirofumi Tanaka, Keiji Miyazawa, and Masatoshi Genes to cells Kitagawa. 2008. 'Novel insights into FGD3, a putative GEF for Cdc42, that undergoes SCFFWD1/β-TrCP-mediated proteasomal degradation analogous to that of its homologue FGD1 but regulates cell morphology and motility differently from FGD1', 13: 329-42.

216

Hebron, Michaeline L, Irina Lonskaya, and Charbel E-H Human molecular genetics Moussa. 2013. 'Nilotinib reverses loss of dopamine neurons and improves motor behavior via autophagic degradation of α-synuclein in Parkinson's disease models', 22: 3315-28. Hedrich, K, K Marder, J Harris, M Kann, T Lynch, H Meija–Santana, PP Pramstaller, E Schwinger, SB Bressman, and S Neurology Fahn. 2002. 'Evaluation of 50 probands with early-onset Parkinson’s disease for Parkin mutations', 58: 1239-46. Hedrich, Katja, Martin Kann, Andrea J Lanthaler, Andreas Dalski, Cordula Eskelson, Olfert Landt, Eberhard Schwinger, Peter Vieregge, Anthony E Lang, and Xandra O Human molecular genetics Breakefield. 2001. 'The importance of gene dosage studies: mutational analysis of the parkin gene in early-onset parkinsonism', 10: 1649-56. Heiman, Myriam, Ruth Kulicke, Robert J Fenster, Paul Greengard, and Nathaniel Nature protocols Heintz. 2014. 'Cell type–specific mRNA purification by translating ribosome affinity purification (TRAP)', 9: 1282. Heiman, Myriam, Anne Schaefer, Shiaoching Gong, Jayms D Peterson, Michelle Day, Keri E Ramsey, Mayte Suárez-Fariñas, Cordelia Schwarz, Dietrich A Stephan, and D James Cell Surmeier. 2008. 'A translational profiling approach for the molecular characterization of CNS cell types', 135: 738-48. Hemmati-Brivanlou, Ali, Olivia G. Kelly, and Douglas A. Melton. 1994. 'Follistatin, an antagonist of activin, is expressed in the Spemann organizer and displays direct neuralizing activity', Cell, 77: 283-95. Henderson-Smith, Adrienne, Jason J Corneveaux, Matthew De Both, Lori Cuyugan, Winnie S Liang, Matthew Huentelman, Charles Adler, Erika Driver-Dunckley, Thomas G Beach, and Travis L Neurology Genetics Dunckley. 2016. 'Next-generation profiling to identify the molecular etiology of Parkinson dementia', 2: e75. Heo, Hye Young, Ji-Min Park, Cy-Hyun Kim, Baek Soo Han, Kwang-Soo Kim, and Wongi Experimental cell research Seol. 2010. 'LRRK2 enhances oxidative stress-induced neurotoxicity via its kinase activity', 316: 649-56. Hermanson, Elisabet, Bertrand Joseph, Diogo Castro, Eva Lindqvist, Piia Aarnisalo, Åsa Wallén, Gerard Benoit, Bastian Hengerer, Lars Olson, and Thomas Experimental cell research Perlmann. 2003. 'Nurr1 regulates dopamine synthesis and storage in MN9D dopamine cells', 288: 324-34. Hirsch, Lauren, Nathalie Jette, Alexandra Frolkis, Thomas Steeves, and Tamara Neuroepidemiology Pringsheim. 2016. 'The incidence of Parkinson's disease: a systematic review and meta- analysis', 46: 292-300. Hoenen, Claire, Audrey Gustin, Cindy Birck, Mélanie Kirchmeyer, Nicolas Beaume, Paul Felten, Luc Grandbarbe, Paul Heuschling, and Tony PloS one Heurtaux. 2016. 'Alpha-synuclein proteins promote pro-inflammatory cascades in microglia: stronger effects of the A53T mutant', 11: e0162717. Holtz, Peter, Rudolf Heise, and Käthe Lüdtke. 1938. 'Fermentativer Abbau von l-dioxyphenylalanin (Dopa) durch Niere', Naunyn-Schmiedeberg's Archives of Pharmacology, 191: 87-118. Hook, Paul W, Sarah A McClymont, Gabrielle H Cannon, William D Law, A Jennifer Morton, Loyal A Goff, and Andrew S The American Journal of Human Genetics McCallion. 2018. 'Single-Cell RNA-Seq of Mouse Dopaminergic Neurons Informs Candidate Gene Selection for Sporadic Parkinson Disease', 102: 427-46. Hoss, Andrew G, Adam Labadorf, Thomas G Beach, Jeanne C Latourelle, and Richard H Myers. 2016. 'microRNA profiles in Parkinson's disease prefrontal cortex', Frontiers in aging neuroscience, 8. Hossein-Nezhad, Arash, Roya Pedram Fatemi, Rili Ahmad, Elaine R Peskind, Cyrus P Zabetian, Shu- Ching Hu, Min Shi, Claes Wahlestedt, Jing Zhang, and Mohammad Ali Journal of Parkinson's disease Faghihi. 2016. 'Transcriptomic profiling of extracellular RNAs present in

217

cerebrospinal fluid identifies differentially expressed transcripts in Parkinson’s disease', 6: 109-17. Huang, Wei, Shan-Shan Tian, Peng-Zhou Hang, Chuan Sun, Jing Guo, and Zhi-Min Molecular Therapy-Nucleic Acids Du. 2016. 'Combination of microRNA-21 and microRNA-146a attenuates cardiac dysfunction and apoptosis during acute myocardial infarction in mice', 5. Hubbard, Tim, Daniel Barker, Ewan Birney, Graham Cameron, Yuan Chen, L Clark, Tony Cox, J Cuff, Val Curwen, and Thomas Down. 2002. 'The Ensembl genome database project', Nucleic acids research, 30: 38-41. Hui, Chin Wai, Yang Zhang, and Karl PloS one Herrup. 2016. 'Non-neuronal cells are required to mediate the effects of neuroinflammation: results from a neuron-enriched culture system', 11: e0147134. Hunn, Benjamin HM, Stephanie J Cragg, J Paul Bolam, Maria-Grazia Spillantini, and Richard Trends in neurosciences Wade-Martins. 2015. 'Impaired intracellular trafficking defines early Parkinson's disease', 38: 178-88. Hunter, Melissa Piper, Noura Ismail, Xiaoli Zhang, Baltazar D Aguda, Eun Joo Lee, Lianbo Yu, Tao Xiao, Jeffrey Schafer, Mei-Ling Ting Lee, and Thomas D PloS one Schmittgen. 2008. 'Detection of microRNA expression in human peripheral blood microvesicles', 3: e3694. Hynes, Mary, Jeffery A Porter, Chin Chiang, David Chang, Marc Tessier-Lavigne, Philip A Beachy, and Arnon Neuron Rosenthal. 1995. 'Induction of midbrain dopaminergic neurons by Sonic hedgehog', 15: 35-44. Ibanez, P, AM Bonnet, B Debarges, E Lohmann, F Tison, Y Agid, A Dürr, A Brice, P Pollak, and French Parkinson's Disease Genetics Study Group. 2004. 'Causal relation between α-synuclein locus duplication as a cause of familial Parkinson's disease', The Lancet, 364: 1169-71. Ingolia, Nicholas T, Gloria A Brar, Silvia Rouskin, Anna M McGeachy, and Jonathan S Nature protocols Weissman. 2012. 'The ribosome profiling strategy for monitoring translation in vivo by deep sequencing of ribosome-protected mRNA fragments', 7: 1534. Ishikawa, Atsushi, and Shoji Neurology Tsuji. 1996. 'Clinical analysis of 17 patients in 12 Japanese families with autosomal-recessive type juvenile parkinsonism', 47: 160-66. Jabado, O, Q Wang, HJ Rideout, M Yeasmin, KX Guo, K Vekrellis, S Papantonis, JM Angelastro, CM Troy, L Cell death Stefanis, and differentiation. 2004. 'RAIDD aggregation facilitates apoptotic death of PC12 cells and sympathetic neurons', 11: 618. Jakes, Ross, Maria Grazia Spillantini, and Michel Goedert. 1994. 'Identification of two distinct synucleins from human brain', FEBS letters, 345: 27-32. Jakobsson, Johan, Cecilia Ericson, Maria Jansson, Elin Björk, and Cecilia Journal of neuroscience research Lundberg. 2003. 'Targeted transgene expression in rat brain using lentiviral vectors', 73: 876-85. Janezic, Stephanie, Sarah Threlfell, Paul D Dodson, Megan J Dowie, Tonya N Taylor, Dawid Potgieter, Laura Parkkinen, Steven L Senior, Sabina Anwar, and Brent Proceedings of the National Academy of Sciences Ryan. 2013. 'Deficits in dopaminergic transmission precede neuron loss and dysfunction in a new Parkinson model': 201309143. Jang, Tae-Ho, In-Hye Lim, Chang Min Kim, Jae Young Choi, Eun-Ae Kim, Tae-Jin Lee, and Hyun Ho Scientific reports Park. 2016. 'Rescuing neuronal cell death by RAIDD-and PIDD-derived peptides and its implications for therapeutic intervention in neurodegenerative diseases', 6: 31198. Javitch, Jonathan A, Robert J D'Amato, Stephen M Strittmatter, and Solomon H Proceedings of the National Academy of Sciences Snyder. 1985. 'Parkinsonism-inducing neurotoxin, N-methyl-4- phenyl-1, 2, 3, 6-tetrahydropyridine: uptake of the metabolite N-methyl-4-phenylpyridine by dopamine neurons explains selective toxicity', 82: 2173-77. Jenco, John M, Andrew Rawlingson, Brenda Daniels, and Andrew J Biochemistry Morris. 1998. 'Regulation of phospholipase D2: selective inhibition of mammalian phospholipase D isoenzymes by α-and β-synucleins', 37: 4901-09.

218

Jinkins, J Randy. 2000. Atlas of neuroradiologic embryology, anatomy, and variants (Lippincott Williams & Wilkins). Joksimovic, Milan, Beth A Yun, Raja Kittappa, Angela M Anderegg, Wendy W Chang, Makoto M Taketo, Ronald DG McKay, and Rajeshwar B Nature neuroscience Awatramani. 2009. 'Wnt antagonism of Shh facilitates midbrain floor plate neurogenesis', 12: 125. Jung, Ji-Hye, Xuebin Fu, and Phillip C Circulation research Yang. 2017. 'Exosomes generated from iPSC-derivatives: new direction for stem cell therapy in human heart diseases', 120: 407-17. Junn, Eunsung, Won Hee Jang, Xin Zhao, Byeong Seon Jeong, and M Maral Journal of neuroscience research Mouradian. 2009. 'Mitochondrial localization of DJ-1 leads to enhanced neuroprotection', 87: 123-29. Junn, Eunsung, Kang-Woo Lee, Byeong Seon Jeong, Teresa W Chan, Joo-Young Im, and M Maral Mouradian. 2009. 'Repression of α-synuclein expression and toxicity by microRNA-7', Proceedings of the National Academy of Sciences, 106: 13052-57. Kabaria, Savan, Doo Chul Choi, Amrita Datta Chaudhuri, M Maral Mouradian, and Eunsung Junn. 2015. 'Inhibition of miR-34b and miR-34c enhances α-synuclein expression in Parkinson's disease', FEBS letters, 589: 319-25. Kabuta, Tomohiro, Akiko Furuta, Shunsuke Aoki, Koh Furuta, and Keiji Journal of Biological Chemistry Wada. 2008. 'Aberrant interaction between Parkinson disease-associated mutant UCH-L1 and the lysosomal receptor for chaperone-mediated autophagy', 283: 23731-38. Kahle, Philipp J, Manuela Neumann, Laurence Ozmen, Veronika Müller, Helmut Jacobsen, Alice Schindzielorz, Masayasu Okochi, Uwe Leimer, Herman van der Putten, and Alphonse Journal of Neuroscience Probst. 2000. 'Subcellular localization of wild-type and Parkinson's disease- associated mutant α-synuclein in human and transgenic mouse brain', 20: 6365-73. Kanemaru, Kazunori, Yohei Okubo, Kenzo Hirose, and Masamitsu Journal of Neuroscience Iino. 2007. 'Regulation of neurite growth by spontaneous Ca2+ oscillations in astrocytes', 27: 8957-66. Kann, Martin, Helfried Jacobs, Kathrin Mohrmann, Kirsten Schumacher, Katja Hedrich, Jennifer Garrels, Karin Wiegers, Eberhard Schwinger, Peter P Pramstaller, and Xandra O Annals of neurology Breakefield. 2002. 'Role of parkin mutations in 111 community-based patients with early-onset parkinsonism', 51: 621-25. Kannagi, Reiji, Nancy A Cochran, Fumitsugu Ishigami, Sen-itiroh Hakomori, PeterW Andrews, Barbara B Knowles, and Davor The embo journal Solter. 1983. 'Stage-specific embryonic antigens (SSEA-3 and-4) are epitopes of a unique globo-series ganglioside isolated from human teratocarcinoma cells', 2: 2355-61. Karuppagounder, Senthilkumar S, Saurav Brahmachari, Yunjong Lee, Valina L Dawson, Ted M Dawson, and Han Seok Scientific reports Ko. 2014. 'The c-Abl inhibitor, nilotinib, protects dopaminergic neurons in a preclinical animal model of Parkinson's disease', 4: 4874. Kawasaki, Hiroshi, Kenji Mizuseki, Satomi Nishikawa, Satoshi Kaneko, Yoshihisa Kuwana, Shigetada Nakanishi, Shin-Ichi Nishikawa, and Yoshiki neuron Sasai. 2000. 'Induction of midbrain dopaminergic neurons from ES cells by stromal cell–derived inducing activity', 28: 31-40. Khan, Mohsin, Emily Nickoloff, Tatiana Abramova, Jennifer Johnson, Suresh K Verma, Prasanna Krishnamurthy, Alexander R Mackie, Erin Vaughan, Venkata NS Garikipati, and Cynthia L Circulation research Benedict. 2015. 'Embryonic stem cell-derived exosomes promote endogenous repair mechanisms and enhance cardiac function following myocardial infarction': CIRCRESAHA. 115.305990. Khoo, Sok Kean, David Petillo, Un Jung Kang, James H Resau, Brian Berryhill, Jan Linder, Lars Forsgren, Leslie A Neuman, and Aik Choon Journal of Parkinson's disease Tan. 2012. 'Plasma-based circulating MicroRNA biomarkers for Parkinson's disease', 2: 321-31. Kiely, Aoife P, Yasmine T Asi, Eleanna Kara, Patricia Limousin, Helen Ling, Patrick Lewis, Christos Proukakis, Niall Quinn, Andrew J Lees, and John Hardy. 2013. 'α-Synucleinopathy associated

219

with G51D SNCA mutation: a link between Parkinson’s disease and multiple system atrophy?', Acta neuropathologica, 125: 753-69. Kim, Changyoun, Dong-Hwan Ho, Ji-Eun Suk, Sungyong You, Sarah Michael, Junghee Kang, Sung Joong Lee, Eliezer Masliah, Daehee Hwang, and He-Jin Nature communications Lee. 2013. 'Neuron-released oligomeric α-synuclein is an endogenous agonist of TLR2 for paracrine activation of microglia', 4: 1562. Kim, Daehwan, Geo Pertea, Cole Trapnell, Harold Pimentel, Ryan Kelley, and Steven L Salzberg. 2013. 'TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions', Genome biology, 14: R36. Kim, Howard, Tasneem Zahir, Charles H Tator, and Molly S PloS one Shoichet. 2011. 'Effects of dibutyryl cyclic-AMP on survival and neuronal differentiation of neural stem/progenitor cells transplanted into spinal cord injured rats', 6: e21744. Kim, Jong-Hoon, Jonathan M Auerbach, José A Rodríguez-Gómez, Iván Velasco, Denise Gavin, Nadya Lumelsky, Sang-Hun Lee, John Nguyen, Rosario Sánchez-Pernaute, and Krys Nature Bankiewicz. 2002. 'Dopamine neurons derived from embryonic stem cells function in an animal model of Parkinson's disease', 418: 50. Kim, Jongpil, Keiichi Inoue, Jennifer Ishii, William B Vanti, Sergey V Voronov, Elizabeth Murchison, Gregory Hannon, and Asa Abeliovich. 2007. 'A MicroRNA feedback circuit in midbrain dopamine neurons', Science, 317: 1220-24. Kirik, Deniz, Lucy E Annett, Corinna Burger, Nicholas Muzyczka, Ronald J Mandel, and Anders Proceedings of the National Academy of Sciences Björklund. 2003. 'Nigrostriatal α- synucleinopathy induced by viral vector-mediated overexpression of human α-synuclein: a new primate model of Parkinson's disease', 100: 2884-89. Kirkeby, Agnete, Shane Grealish, Daniel A Wolf, Jenny Nelander, James Wood, Martin Lundblad, Olle Lindvall, and Malin Cell reports Parmar. 2012. 'Generation of regionally specified neural progenitors and functional neurons from human embryonic stem cells under defined conditions', 1: 703-14. Kitada, Tohru, Shuichi Asakawa, Nobutaka Hattori, Hiroto Matsumine, Yasuhiro Yamamura, Shinsei Minoshima, Masayuki Yokochi, Yoshikuni Mizuno, and Nobuyoshi Nature Shimizu. 1998. 'Mutations in the parkin gene cause autosomal recessive juvenile parkinsonism', 392: 605. Kitada, Tohru, Youren Tong, Clement A Gautier, and Jie Journal of neurochemistry Shen. 2009. 'Absence of nigral degeneration in aged parkin/DJ-1/PINK1 triple knockout mice', 111: 696- 702. Kittappa, Raja, Wendy W Chang, Rajeshwar B Awatramani, and Ronald DG PLoS biology McKay. 2007. 'The foxa2 gene controls the birth and spontaneous degeneration of dopamine neurons in old age', 5: e325. Knobel, Karla M, Warren S Davis, Erik M Jorgensen, and Michael J Development Bastiani. 2001. 'UNC-119 suppresses axon branching in C. elegans', 128: 4079-92. Koch, JC, F Bitow, J Haack, Z d'Hedouville, JN Zhang, L Tönges, U Michel, LMA Oliveira, TM Jovin, J Cell death Liman, and disease. 2015. 'Alpha-Synuclein affects neurite morphology, autophagy, vesicle transport and axonal degeneration in CNS neurons', 6: e1811. Komatsu, Masaaki, Satoshi Waguri, Tomoki Chiba, Shigeo Murata, Jun-ichi Iwata, Isei Tanida, Takashi Ueno, Masato Koike, Yasuo Uchiyama, and Eiki Nature Kominami. 2006. 'Loss of autophagy in the central nervous system causes neurodegeneration in mice', 441: 880. Kong, Yan, Xijun Liang, Lin Liu, Dongdong Zhang, Chao Wan, Zhenji Gan, and Liudi Yuan. 2015. 'High throughput sequencing identifies MicroRNAs mediating α-synuclein toxicity by targeting neuroactive-ligand receptor interaction pathway in early stage of drosophila Parkinson's disease model', PLoS One, 10: e0137432. Kordower, Jeffrey H, Yaping Chu, Robert A Hauser, Thomas B Freeman, and C Warren Nature medicine Olanow. 2008. 'Lewy body–like pathology in long-term embryonic nigral transplants in Parkinson's disease', 14: 504.

220

Korecka, Joanna A, Ronald E van Kesteren, Eva Blaas, Sonia O Spitzer, Jorke H Kamstra, August B Smit, Dick F Swaab, Joost Verhaagen, and Koen PloS one Bossers. 2013. 'Phenotypic characterization of retinoic acid differentiated SH-SY5Y cells by transcriptional profiling', 8: e63862. Kosaka, Nobuyoshi, Haruhisa Iguchi, Keitaro Hagiwara, Yusuke Yoshioka, Fumitaka Takeshita, and Takahiro Journal of Biological Chemistry Ochiya. 2013. 'Neutral sphingomyelinase 2 (nSMase2)-dependent exosomal transfer of angiogenic microRNAs regulate cancer cell metastasis': jbc. M112. 446831. Kouroupi, Georgia, Era Taoufik, Ioannis S Vlachos, Konstantinos Tsioras, Nasia Antoniou, Florentia Papastefanaki, Dafni Chroni-Tzartou, Wolfgang Wrasidlo, Delphine Bohl, and Dimitris Proceedings of the National Academy of Sciences Stellas. 2017. 'Defective synaptic connectivity and axonal neuropathology in a human iPSC-based model of familial Parkinson’s disease', 114: E3679-E88. Koval, Erica D, Carey Shaner, Peter Zhang, Xavier du Maine, Kimberlee Fischer, Jia Tay, B Nelson Chau, Gregory F Wu, and Timothy M Human molecular genetics Miller. 2013. 'Method for widespread microRNA-155 inhibition prolongs survival in ALS-model mice', 22: 4127-35. Krebiehl, Guido, Sabine Ruckerbauer, Lena F Burbulla, Nicole Kieper, Brigitte Maurer, Jens Waak, Hartwig Wolburg, Zemfira Gizatullina, Frank N Gellerich, and Dirk PloS one Woitalla. 2010. 'Reduced basal autophagy and impaired mitochondrial dynamics due to loss of Parkinson's disease-associated protein DJ-1', 5: e9367. Kriks, Sonja, Jae-Won Shim, Jinghua Piao, Yosif M Ganat, Dustin R Wakeman, Zhong Xie, Luis Carrillo- Reid, Gordon Auyeung, Chris Antonacci, and Amanda Nature Buch. 2011. 'Dopamine neurons derived from human ES cells efficiently engraft in animal models of Parkinson’s disease', 480: 547. Krüger, Rejko, Wilfried Kuhn, Thomas Müller, Dirk Woitalla, Manuel Graeber, Sigfried Kösel, Horst Przuntek, Jörg T Epplen, Ludger Schols, and Olaf Riess. 1998. 'AlaSOPro mutation in the gene encoding α-synuclein in Parkinson's disease', Nature genetics, 18: 106-08. Kuleshov, Maxim V, Matthew R Jones, Andrew D Rouillard, Nicolas F Fernandez, Qiaonan Duan, Zichen Wang, Simon Koplev, Sherry L Jenkins, Kathleen M Jagodnik, and Alexander Nucleic acids research Lachmann. 2016. 'Enrichr: a comprehensive gene set enrichment analysis web server 2016 update', 44: W90-W97. Kumar, Roshan, Deepak K Jangir, Garima Verma, Shashi Shekhar, Pranita Hanpude, Sanjay Kumar, Raniki Kumari, Nirpendra Singh, Neel Sarovar Bhavesh, and Nihar Ranjan Scientific reports Jana. 2017. 'S-nitrosylation of UCHL1 induces its structural instability and promotes α- synuclein aggregation', 7: 44558. Kumar, Sujeet, Katrin Zimmermann, Hiroyuki Hioki, Alexander Pfeifer, Stephan L Histochemistry Baader, and cell biology. 2015. 'Efficient and graded gene expression in glia and neurons of primary cerebellar cultures transduced by lentiviral vectors', 143: 109-21. Kurowska, Zuzanna, Elisabet Englund, Håkan Widner, Olle Lindvall, Jia-Yi Li, and Patrik Journal of Parkinson's disease Brundin. 2011. 'Signs of degeneration in 12–22-year old grafts of mesencephalic dopamine neurons in patients with Parkinson's disease', 1: 83-92. Kuwata, Hitoshi, Kazuki Nakao, Takeshi Harada, Ikuo Matsuda, and Atsu Kobe J Med Sci Aiba. 2008. 'Generation of RGS8 null mutant mice by Cre/loxP system', 53: 275-81. Lam, Rebecca S, Felix M Töpfer, Phillip G Wood, Volker Busskamp, and Ernst PloS one Bamberg. 2017. 'Functional maturation of human stem cell-derived neurons in long-term cultures', 12: e0169506. Lamba, Vishal, Yogita Ghodke-Puranik, Weihua Guan, and Jatinder K BMC research notes Lamba. 2014. 'Identification of suitable reference genes for hepatic microRNA quantitation', 7: 129. Lamoureux, Phillip, Vivian L Steel, Clara Regal, Lynne Adgate, Robert E Buxbaum, and Steven R The Journal of cell biology Heidemann. 1990. 'Extracellular matrix allows PC12 neurite elongation in the absence of microtubules', 110: 71-79.

221

Langston, J William, Philip Ballard, James W Tetrud, and Ian Science Irwin. 1983. 'Chronic Parkinsonism in humans due to a product of meperidine-analog synthesis', 219: 979-80. Langston, J William, Savitha Sastry, Piu Chan, Lysia S Forno, Laurel M Bolin, and Donato A Di Monte. 1998. 'Novel α-synuclein-immunoreactive proteins in brain samples from the Contursi kindred, Parkinson's, and Alzheimer's disease', Experimental neurology, 154: 684-90. Langston, JW, LS Forno, J Tetrud, AG Reeves, JA Kaplan, D Annals of Neurology: Official Journal of the American Neurological Association Karluk, and the Child Neurology Society. 1999. 'Evidence of active nerve cell degeneration in the substantia nigra of humans years after 1- methyl-4-phenyl-1, 2, 3, 6-tetrahydropyridine exposure', 46: 598-605. Lanza, Robert P, Jose B Cibelli, Catherine Blackwell, Vincent J Cristofalo, Mary Kay Francis, Gabriela M Baerlocher, Jennifer Mak, Michael Schertzer, Elizabeth A Chavez, and Nancy Science Sawyer. 2000. 'Extension of cell life-span and telomere length in animals cloned from senescent somatic cells', 288: 665-69. Lapasset, Laure, Ollivier Milhavet, Alexandre Prieur, Emilie Besnard, Amelie Babled, Nafissa Aït- Hamou, Julia Leschik, Franck Pellestor, Jean-Marie Ramirez, John Genes De Vos, and development. 2011. 'Rejuvenating senescent and centenarian human cells by reprogramming through the pluripotent state', 25: 2248-53. Larsen, Kristin E, Yvonne Schmitz, Matthew D Troyer, Eugene Mosharov, Paula Dietrich, Abrar Z Quazi, Magali Savalle, Venu Nemani, Farrukh A Chaudhry, and Robert H Journal of Neuroscience Edwards. 2006. 'α-Synuclein overexpression in PC12 and chromaffin cells impairs catecholamine release by interfering with a late step in exocytosis', 26: 11915-22. Lazzarini, AM, RH Myers, TR Zimmerman, MH Mark, LI Golbe, JI Sage, WG Johnson, and RC Duvoisin. 1994. 'A clinical genetic study of Parkinson's disease Evidence for dominant transmission', Neurology, 44: 499-99. Lee, He-Jin, Smita Patel, and Seung-Jae Journal of Neuroscience Lee. 2005. 'Intravesicular localization and exocytosis of α-synuclein and its aggregates', 25: 6016-24. Lee, He-Jin, Ji-Eun Suk, Eun-Jin Bae, Seung-Jae Biochemical Lee, and biophysical research communications. 2008. 'Clearance and deposition of extracellular α-synuclein aggregates in microglia', 372: 423-28. Lee, He-Jin, Farnaz Khoshaghideh, Stephen Lee, and Seung-Jae European Journal of Neuroscience Lee. 2006. 'Impairment of microtubule-dependent trafficking by overexpression of α- synuclein', 24: 3153-62. Lee, Rosalind C, Rhonda L Feinbaum, and Victor cell Ambros. 1993. 'The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14', 75: 843-54. Lee, Sang-Hun, Nadya Lumelsky, Lorenz Studer, Jonathan M Auerbach, and Ron D Nature biotechnology McKay. 2000. 'Efficient generation of midbrain and hindbrain neurons from mouse embryonic stem cells', 18: 675. Lee, Seongmi, Seung Tack Oh, Ha Jin Jeong, Sok Cheon Pak, Hi-Joon Park, Jongpil Kim, Hyun-seok Cho, Songhee The Korean Journal of Physiology Jeon, and Pharmacology. 2017. 'MPTP- induced vulnerability of dopamine neurons in A53T α-synuclein overexpressed mice with the potential involvement of DJ-1 downregulation', 21: 625-32. Lehmensiek, Vera, Eva-Maria Tan, Stefan Liebau, Thomas Lenk, Heinz Zettlmeisl, Johannes Schwarz, and Alexander Neurochemistry international Storch. 2006. 'Dopamine transporter-mediated cytotoxicity of 6-hydroxydopamine in vitro depends on expression of mutant α-synucleins related to Parkinson's disease', 48: 329-40. Leroy, Elisabeth, Rebecca Boyer, Georg Auburger, Barbara Leube, Gudrun Ulm, Eva Mezey, Gyongyi Harta, Michael J Brownstein, Sobhanadditya Jonnalagada, and Tanya Nature Chernova. 1998. 'The ubiquitin pathway in Parkinson's disease', 395: 451. Lesage, Suzanne, Mathieu Anheim, Franck Letournel, Luc Bousset, Aurélie Honoré, Nelly Rozas, Laura Pieri, Karine Madiona, Alexandra Dürr, and Ronald Melki. 2013. 'G51D α-synuclein mutation causes a novel Parkinsonian–pyramidal syndrome', Annals of neurology, 73: 459-71.

222

Lewy, FH. 1912. 'Paralysis agitans. I. Pathologische anatomie', Handbuch der neurologie, 3: 920-33. Li, J-Y, P Henning Jensen, and A Neuroscience Dahlström. 2002. 'Differential localization of α-, β-and γ-synucleins in the rat CNS', 113: 463-78. Li, Jia-Yi, Elisabet Englund, Janice L Holton, Denis Soulet, Peter Hagell, Andrew J Lees, Tammaryn Lashley, Niall P Quinn, Stig Rehncrona, and Anders Nature medicine Björklund. 2008. 'Lewy bodies in grafted neurons in subjects with Parkinson's disease suggest host-to-graft disease propagation', 14: 501. Li, Mingjie, Nada Husic, Ying Lin, Heather Christensen, Ibrahim Malik, Sally McIver, Christine M LaPash Daniels, David A Harris, Paul T Kotzbauer, and Mark P Journal of neuroscience methods Goldberg. 2010. 'Optimal promoter usage for lentiviral vector-mediated transduction of cultured central nervous system cells', 189: 56-64. Li, Shize, Xuecheng Lv, Kaihua Zhai, Ruyan Xu, Yong Zhang, Songyao Zhao, Xiaoming Qin, Liujie Yin, and Jiyu Lou. 2016. 'MicroRNA-7 inhibits neuronal apoptosis in a cellular Parkinson’s disease model by targeting Bax and Sirt2', American journal of translational research, 8: 993. Li, Wen-Wei, Ru Yang, Jing-Chun Guo, Hui-Min Ren, Xi-Liang Zha, Jie-Shi Cheng, and Ding-Fang Neuroreport Cai. 2007. 'Localization of α-synuclein to mitochondria within midbrain of mice', 18: 1543-46. Li, Yanping, Wencheng Liu, Tinmarla F Oo, Lei Wang, Yi Tang, Vernice Jackson-Lewis, Chun Zhou, Kindiya Geghman, Mikhail Bogdanov, and Serge Nature neuroscience Przedborski. 2009. 'Mutant LRRK2 R1441G BAC transgenic mice recapitulate cardinal features of Parkinson's disease', 12: 826. Lieth, Erich, David R McClay, and Jean M Glia Lauder. 1990. 'Neuronal-glial interactions: Complexity of neurite outgrowth correlates with substrated adhesivity of serotonergic neurons', 3: 169- 79. Lin, Mingyan, Erika Pedrosa, Abhishek Shah, Anastasia Hrabovsky, Shahina Maqbool, Deyou Zheng, and Herbert M PloS one Lachman. 2011. 'RNA-Seq of human neurons derived from iPS cells reveals candidate long non-coding RNAs involved in neurogenesis and neuropsychiatric disorders', 6: e23356. Lin, Wei, Emmanouil Metzakopian, Yannis E Mavromatakis, Nan Gao, Nikolaos Balaskas, Hiroshi Sasaki, James Briscoe, Jeffrey A Whitsett, Martyn Goulding, and Klaus H Developmental biology Kaestner. 2009. 'Foxa1 and Foxa2 function both upstream of and cooperatively with Lmx1a and Lmx1b in a feedforward loop promoting mesodiencephalic dopaminergic neuron development', 333: 386-96. Lindersson, Evo, Rasmus Beedholm, Peter Højrup, Torben Moos, WeiPing Gai, Klavs B Hendil, and Poul H Journal of Biological Chemistry Jensen. 2004. 'Proteasomal inhibition by α-synuclein filaments and oligomers'. Liu, Guangwei, Peng Wang, Xin Li, Yaohua Li, Shengli Xu, Kenji Uéda, Piu Chan, and Shun Journal of Neural Transmission Yu. 2013. 'Alpha-synuclein promotes early neurite outgrowth in cultured primary neurons', 120: 1331-43. Liu, Hsin-Yu, Chiao-Ming Huang, Yun-Fen Hung, and Yi-Ping Experimental neurology Hsueh. 2015. 'The microRNAs Let7c and miR21 are recognized by neuronal Toll-like receptor 7 to restrict dendritic growth of neurons', 269: 202-12. Liu, Wencheng, Cristofol Vives-Bauza, Ai Yamamoto, Yingcai Tan, Yanping Li, Jordi Magrané, Mihaela A Stavarache, Sebastian Shaffer, Simon Chang, and Michael G PloS one Kaplitt. 2009. 'PINK1 defect causes mitochondrial dysfunction, proteasomal deficit and α-synuclein aggregation in cell culture models of Parkinson's disease', 4: e4597. Liu, Yanying, Min Qiang, Yan Wei, and Rongqiao Journal of molecular cell biology He. 2011. 'A novel molecular mechanism for nitrated α-synuclein-induced cell death', 3: 239-49. Liu, Yichin, Lara Fallon, Hilal A Lashuel, Zhihua Liu, and Peter T Cell Lansbury Jr. 2002. 'The UCH-L1 gene encodes two opposing enzymatic activities that affect α-synuclein degradation and Parkinson's disease susceptibility', 111: 209-18.

223

Liu, Zhong, Cheng Zhang, Alireza Khodadadi-Jamayran, Lam Dang, Xiaosi Han, Kitai Kim, Hu Li, Rui Stem cells Zhao, and development. 2017. 'Canonical microRNAs enable differentiation, protect against DNA damage, and promote cholesterol biosynthesis in neural stem cells', 26: 177-88. Llorente-Folch, Irene, Ignasi Sahún, Laura Contreras, María José Casarejos, Josep María Grau, Takeyori Saheki, María Angeles Mena, Jorgina Satrústegui, Mara Dierssen, and Beatriz Journal of neurochemistry Pardo. 2013. 'AGC1-malate aspartate shuttle activity is critical for dopamine handling in the nigrostriatal pathway', 124: 347-62. Lockhart, David J, Helin Dong, Michael C Byrne, Maximillian T Follettie, Michael V Gallo, Mark S Chee, Michael Mittmann, Chunwei Wang, Michiko Kobayashi, and Heidi Nature biotechnology Norton. 1996. 'Expression monitoring by hybridization to high-density oligonucleotide arrays', 14: 1675. Lohmann, Ebba, Magali Periquet, Vincenzo Bonifati, Nick W Wood, Giuseppe De Michele, Anne- Marie Bonnet, Valérie Fraix, Emmanuel Broussolle, Martin WIM Horstink, and Marie Annals of neurology Vidailhet. 2003. 'How much phenotypic variation can be attributed to parkin genotype?', 54: 176-85. Lopes, Fernanda Martins, Rafael Schröder, Mário Luiz Conte da Frota Júnior, Alfeu Zanotto-Filho, Carolina Beatriz Müller, André Simões Pires, Rosalva Thereza Meurer, Gabriela Delevati Colpo, Daniel Pens Gelain, and Flávio Brain research Kapczinski. 2010. 'Comparison between proliferative and neuron-like SH-SY5Y cells as an in vitro model for Parkinson disease studies', 1337: 85-94. Lücking, Christoph B, Alexandra Dürr, Vincenzo Bonifati, Jenny Vaughan, Giuseppe De Michele, Thomas Gasser, Biswadjiet S Harhangi, Giuseppe Meco, Patrice Denèfle, and Nicholas W New England Journal of Medicine Wood. 2000. 'Association between early-onset Parkinson's disease and mutations in the parkin gene', 342: 1560-67. Luthman, J, A Fredriksson, E Sundström, G Jonsson, and T Behavioural brain research Archer. 1989. 'Selective lesion of central dopamine or noradrenaline neuron systems in the neonatal rat: motor behavior and monoamine alterations at adult stage', 33: 267-77. Ma, Shuang Yong, Matias Röyttä, Juha O Rinne, Yrjö Collan, and Urpo K Journal of the neurological sciences Rinne. 1997. 'Correlation between neuromorphometry in the substantia nigra and clinical features in Parkinson's disease using disector counts', 151: 83-87. Maltsev, Alexander S, Jinfa Ying, and Ad Biochemistry Bax. 2012. 'Impact of N-terminal acetylation of α-synuclein on its random coil and lipid binding properties', 51: 5004-13. Manning-Bog, Amy B, Alison L McCormack, Jie Li, Vladimir N Uversky, Anthony L Fink, and Donato A Journal of Biological Chemistry Di Monte. 2002. 'The herbicide paraquat causes up- regulation and aggregation of α-synuclein in mice paraquat and α-synuclein', 277: 1641-44. Mao, Shihong, Edward Sendler, Robert J Goodrich, Russ Hauser, and Stephen A Krawetz. 2014. 'A comparison of sperm RNA-seq methods', Systems biology in reproductive medicine, 60: 308- 15. Maraganore, DM, AE Harding, and CD Marsden. 1991. 'A clinical and genetic study of familial Parkinson's disease', Movement Disorders, 6: 205-11. Marambaud, Philippe, Ute Dreses-Werringloer, and Valérie Molecular neurodegeneration Vingtdeux. 2009. 'Calcium signaling in neurodegeneration', 4: 20. Margis, Regina, Rogério Margis, and Carlos RM Journal of biotechnology Rieder. 2011. 'Identification of blood microRNAs associated to Parkinsońs disease', 152: 96-101. Mariani, Elisa, Flavia Frabetti, Andrea Tarozzi, Maria Chiara Pelleri, Fabrizio Pizzetti, and Raffaella PLoS One Casadei. 2016. 'Meta-analysis of Parkinson's disease Transcriptome data using TRAM software: whole Substantia Nigra tissue and single dopamine neuron differential gene expression', 11: e0161567. Martí, Yasmina, Elina Nürnberg, Sandra Horschitz, Mathias Hafner, Patrick Schloss, Andreas Meyer- Lindenberg, and Thorsten Journal of Cellular Biotechnology Lau. 2017. 'Refinement of a

224

neuronal differentiation protocol predominantly yields human iPS cell-derived dopaminergic neurons for the investigation of neurodegenerative pathomechanisms in vitro', 3: 61-80. Martin, Lee J, Yan Pan, Ann C Price, Wanda Sterling, Neal G Copeland, Nancy A Jenkins, Donald L Price, and Michael K Journal of Neuroscience Lee. 2006. 'Parkinson's disease α-synuclein transgenic mice develop neuronal mitochondrial degeneration and cell death', 26: 41-50. Martin, William E, Joseph A Resch, and Abe B Archives of neurology Baker. 1971. 'Juvenile parkinsonism', 25: 494-500. Mašek, Tomáš, Leoš Valášek, and Martin Pospíšek. 2011. 'Polysome analysis and RNA purification from sucrose gradients.' in, RNA (Springer). Masliah, Eliezer, Edward Rockenstein, Isaac Veinbergs, Margaret Mallory, Makoto Hashimoto, Ayako Takeda, Yutaka Sagara, Abbyann Sisk, and Lennart Science Mucke. 2000. 'Dopaminergic loss and inclusion body formation in α-synuclein mice: implications for neurodegenerative disorders', 287: 1265-69. Mastromarino, Paola, Cinzia Conti, Paola Goldoni, Bernard Hauttecoeur, and Nicola Journal of General Virology Orsi. 1987. 'Characterization of membrane components of the erythrocyte involved in vesicular stomatitis virus attachment and fusion at acidic pH', 68: 2359-69. Mathivanan, Suresh, Hong Ji, and Richard J Journal of proteomics Simpson. 2010. 'Exosomes: extracellular organelles important in intercellular communication', 73: 1907-20. Matsuda, Noriyuki, Shigeto Sato, Kahori Shiba, Kei Okatsu, Keiko Saisho, Clement A Gautier, Yu-shin Sou, Shinji Saiki, Sumihiro Kawajiri, and Fumiaki The Journal of cell biology Sato. 2010. 'PINK1 stabilized by mitochondrial depolarization recruits Parkin to damaged mitochondria and activates latent Parkin for mitophagy', 189: 211-21. Matsuda, Wakoto, Takahiro Furuta, Kouichi C Nakamura, Hiroyuki Hioki, Fumino Fujiyama, Ryohachi Arai, and Takeshi Journal of Neuroscience Kaneko. 2009. 'Single nigrostriatal dopaminergic neurons form widely spread and highly dense axonal arborizations in the neostriatum', 29: 444-53. Matsumine, Hiroto, Masaaki Saito, Satoe Shimoda-Matsubayashi, Hajime Tanaka, Atsushi Ishikawa, Yuko Nakagawa-Hattori, Masayuki Yokochi, Tomonori Kobayashi, Shuichi Igarashi, and Hiroki American journal of human genetics Takano. 1997. 'Localization of a gene for an autosomal recessive form of juvenile Parkinsonism to chromosome 6q25. 2-27', 60: 588. Mattila, PM, JO Rinne, H Helenius, DW Dickson, and M Acta neuropathologica Röyttä. 2000. 'Alpha- synuclein-immunoreactive cortical Lewy bodies are associated with cognitive impairment in Parkinson’s disease', 100: 285-90. Mavromatakis, Yannis E, Wei Lin, Emmanouil Metzakopian, Anna LM Ferri, Carol H Yan, Hiroshi Sasaki, Jeff Whisett, and Siew-Lan Mechanisms of development Ang. 2011. 'Foxa1 and Foxa2 positively and negatively regulate Shh signalling to specify ventral midbrain progenitor identity', 128: 90-103. May, Stephanie, Daniel Hornburg, Martin H Schludi, Thomas Arzberger, Kristin Rentzsch, Benjamin M Schwenk, Friedrich A Grässer, Kohji Mori, Elisabeth Kremmer, and Julia Acta neuropathologica Banzhaf-Strathmann. 2014. 'C9orf72 FTLD/ALS-associated Gly-Ala dipeptide repeat proteins cause neuronal toxicity and Unc119 sequestration', 128: 485-503. Mazzulli, Joseph R, Friederike Zunke, Ole Isacson, Lorenz Studer, and Dimitri Proceedings of the National Academy of Sciences Krainc. 2016. 'α-Synuclein–induced lysosomal dysfunction occurs through disruptions in protein trafficking in human midbrain synucleinopathy models', 113: 1931-36. McGeer, PL, S Itagaki, BE Boyes, and EG Neurology McGeer. 1988. 'Reactive microglia are positive for HLA-DR in the substantia nigra of Parkinson's and Alzheimer's disease brains', 38: 1285- 85. McMillan, Catherine R, Rohita Sharma, Tom Ottenhof, and Lennard P Neuroscience letters Niles. 2007. 'Modulation of tyrosine hydroxylase expression by melatonin in human SH-SY5Y neuroblastoma cells', 419: 202-06.

225

McMillan, Kirsty J, Tracey K Murray, Nora Bengoa-Vergniory, Oscar Cordero-Llana, Jane Cooper, Amy Buckley, Richard Wade-Martins, James B Uney, Michael J O’Neill, and Liang F Molecular Therapy Wong. 2017. 'Loss of MicroRNA-7 Regulation Leads to α-Synuclein Accumulation and Dopaminergic Neuronal Loss In Vivo', 25: 2404-14. Melo, TQ, KC Van Zomeren, MFR Ferrari, HWGM Boddeke, and JCVM Experimental brain research Copray. 2017. 'Impairment of mitochondria dynamics by human A53T α-synuclein and rescue by NAP (davunetide) in a cell model for Parkinson’s disease', 235: 731-42. Menga, Alessio, Vito Iacobazzi, Vittoria Infantino, Maria Laura Avantaggiati, Ferdinando The international journal of biochemistry Palmieri, and cell biology. 2015. 'The mitochondrial aspartate/glutamate carrier isoform 1 gene expression is regulated by CREB in neuronal cells', 60: 157-66. Merico, Daniele, Gregory Costain, Nancy J Butcher, William Warnica, Lucas Ogura, Simon E Alfred, Linda M Brzustowicz, and Anne S Frontiers in neurology Bassett. 2014. 'MicroRNA dysregulation, gene networks, and risk for schizophrenia in 22q11. 2 deletion syndrome', 5: 238. Michell, A Wetal, GK Tofaris, H Gossage, P Tyers, MG Spillantini, and RA Cell transplantation Barker. 2007. 'The effect of truncated human α-synuclein (1–120) on dopaminergic cells in a transgenic mouse model of Parkinson's Disease', 16: 461-74. Mifune, Kumato, Masanobu Ohuchi, and Kazuaki FEBS letters Mannen. 1982. 'Hemolysis and cell fusion by rhabdoviruses', 137: 293-97. Miller, Justine D, Yosif M Ganat, Sarah Kishinevsky, Robert L Bowman, Becky Liu, Edmund Y Tu, Pankaj K Mandal, Elsa Vera, Jae-won Shim, and Sonja Cell stem cell Kriks. 2013. 'Human iPSC-based modeling of late-onset disease via progerin-induced aging', 13: 691-705. Miller, Renee M, Gretchen L Kiser, TM Kaysser-Kranich, Randall J Lockner, Chockalingam Palaniappan, and Howard J Neurobiology of disease Federoff. 2006. 'Robust dysregulation of gene expression in substantia nigra and striatum in Parkinson's disease', 21: 305-13. Miñones-Moyano, Elena, Sílvia Porta, Georgia Escaramís, Raquel Rabionet, Susana Iraola, Birgit Kagerbauer, Yolanda Espinosa-Parrilla, Isidre Ferrer, Xavier Estivill, and Eulàlia Martí. 2011. 'MicroRNA profiling of Parkinson's disease brains identifies early downregulation of miR- 34b/c which modulate mitochondrial function', Human molecular genetics, 20: 3067-78. Mitsui, Kaoru, Yoshimi Tokuzawa, Hiroaki Itoh, Kohichi Segawa, Mirei Murakami, Kazutoshi Takahashi, Masayoshi Maruyama, Mitsuyo Maeda, and Shinya cell Yamanaka. 2003. 'The homeoprotein Nanog is required for maintenance of pluripotency in mouse epiblast and ES cells', 113: 631-42. Mitsumoto, Atsushi, and Yasuhito Free radical research Nakagawa. 2001. 'DJ-1 is an indicator for endogenous reactive oxygen species elicited by endotoxin', 35: 885-93. Mo, Mingshu, Yousheng Xiao, Shuxuan Huang, Luan Cen, Xiang Chen, Limin Zhang, Qin Luo, Shaomin Li, Xinling Yang, and Xian Oncotarget Lin. 2017. 'MicroRNA expressing profiles in A53T mutant alpha-synuclein transgenic mice and Parkinsonian', 8: 15. Mogi, Makio, Minoru Harada, Tomoyshi Kondo, Peter Riederer, Hirofumi Inagaki, Masayasu Minami, and Toshiharu Neuroscience letters Nagatsu. 1994. 'Interleukin-1β, interleukin-6, epidermal growth factor and transforming growth factor-α are elevated in the brain from parkinsonian patients', 180: 147-50. Mogi, Makio, Minoru Harada, Peter Riederer, Hirotaro Narabayashi, Keisuke Fujita, and Toshiharu Neuroscience letters Nagatsu. 1994. 'Tumor necrosis factor-α (TNF-α) increases both in the brain and in the cerebrospinal fluid from parkinsonian patients', 165: 208-10. Moisoi, Nicoleta, Valentina Fedele, Jennifer Edwards, and L Miguel Neuropharmacology Martins. 2014. 'Loss of PINK1 enhances neurodegeneration in a mouse model of Parkinson's disease triggered by mitochondrial stress', 77: 350-57. Montagu, KA. 1957. 'Catechol compounds in rat tissues and in brains of different animals', Nature, 180: 244-45.

226

Morett, Enrique, and Peer Trends in biochemical sciences Bork. 1999. 'A novel transactivation domain in parkin', 24: 229-31. Muenter, Manfred D, Lysia S Forno, Oleh Hornykiewicz, Stephen J Kish, Demetrius M Maraganore, Richard J Caselli, Haruo Okazaki, Frank M Howard, Barry J Snow, and Donald B Calne. 1998. 'Hereditary form of parkinsonism—dementia', Annals of neurology, 43: 768-81. Muñoz-Sanjuán, Ignacio, and Ali H Nature Reviews Neuroscience Brivanlou. 2002. 'Neural induction, the default model and embryonic stem cells', 3: 271. Murray, Ian VJ, Benoit I Giasson, Shawn M Quinn, Vishwanath Koppaka, Paul H Axelsen, Harry Ischiropoulos, John Q Trojanowski, and Virginia M-Y Biochemistry Lee. 2003. 'Role of α- synuclein carboxy-terminus on fibril formation in vitro', 42: 8530-40. Nakajo, Shigeo, Kazuhiko Tsukada, Kumiko Omata, Yasuharu Nakamura, and Kazuyasu European journal of biochemistry Nakaya. 1993. 'A new brain-specific 14-kDa protein is a phosphoprotein: Its complete amino acid sequence and evidence for phosphorylation', 217: 1057-63. Nakamura, Ken, Venu M Nemani, Farnaz Azarbal, Gaia Skibinski, Jon M Levy, Kiyoshi Egami, Larissa Munishkina, Jue Zhang, Brooke Gardner, and Junko Journal of Biological Chemistry Wakabayashi. 2011. 'Direct membrane association drives mitochondrial fission by the Parkinson disease-associated protein α-synuclein': jbc. M110. 213538. Nakamura, Ken, Venu M Nemani, Erika K Wallender, Katrin Kaehlcke, Melanie Ott, and Robert H Journal of Neuroscience Edwards. 2008. 'Optical reporters for the conformation of α- synuclein reveal a specific interaction with mitochondria', 28: 12305-17. Nakatani, Tomoya, Minoru Kumai, Eri Mizuhara, Yasuko Minaki, and Yuichi Developmental biology Ono. 2010. 'Lmx1a and Lmx1b cooperate with Foxa2 to coordinate the specification of dopaminergic neurons and control of floor plate cell differentiation in the developing mesencephalon', 339: 101-13. Naldini, Luigi, Ulrike Blömer, Philippe Gallay, Daniel Ory, Richard Mulligan, Fred H Gage, Inder M Verma, and Didier Science Trono. 1996. 'In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector', 272: 263-67. Narendra, Derek, Atsushi Tanaka, Der-Fen Suen, and Richard J The Journal of cell biology Youle. 2008. 'Parkin is recruited selectively to impaired mitochondria and promotes their autophagy', 183: 795-803. Nemani, Venu M, Wei Lu, Victoria Berge, Ken Nakamura, Bibiana Onoa, Michael K Lee, Farrukh A Chaudhry, Roger A Nicoll, and Robert H Neuron Edwards. 2010. 'Increased expression of α- synuclein reduces neurotransmitter release by inhibiting synaptic vesicle reclustering after endocytosis', 65: 66-79. Ngolab, Jennifer, Ivy Trinh, Edward Rockenstein, Michael Mante, Jazmin Florio, Margarita Trejo, Deborah Masliah, Anthony Adame, Eliezer Masliah, and Robert A Acta neuropathologica communications Rissman. 2017. 'Brain-derived exosomes from dementia with Lewy bodies propagate α-synuclein pathology', 5: 46. Nguyen, Ha Nam, Blake Byers, Branden Cord, Aleksandr Shcheglovitov, James Byrne, Prachi Gujar, Kehkooi Kee, Birgitt Schüle, Ricardo E Dolmetsch, and William Cell stem cell Langston. 2011. 'LRRK2 mutant iPSC-derived DA neurons demonstrate increased susceptibility to oxidative stress', 8: 267-80. Nichols, Jennifer, Branko Zevnik, Konstantinos Anastassiadis, Hitoshi Niwa, Daniela Klewe-Nebenius, Ian Chambers, Hans Schöler, and Austin Cell Smith. 1998. 'Formation of pluripotent stem cells in the mammalian embryo depends on the POU transcription factor Oct4', 95: 379-91. Noyce, Alastair J, Jonathan P Bestwick, Laura Silveira-Moriyama, Christopher H Hawkes, Gavin Giovannoni, Andrew J Lees, and Anette Annals of neurology Schrag. 2012. 'Meta-analysis of early nonmotor features and risk factors for Parkinson disease', 72: 893-901.

227

Nural, Hikmet, Ping He, Thomas Beach, Lucia Sue, Weiming Xia, and Yong Molecular neurodegeneration Shen. 2009. 'Dissembled DJ-1 high molecular weight complex in cortex mitochondria from Parkinson's disease patients', 4: 23. Nüsslein-Volhard, Christiane, and Eric Nature Wieschaus. 1980. 'Mutations affecting segment number and polarity in Drosophila', 287: 795. Oaks, Adam W, Nicholas Marsh-Armstrong, Jessica M Jones, Joel J Credle, and Anita PloS one Sidhu. 2013. 'Synucleins antagonize endoplasmic reticulum function to modulate dopamine transporter trafficking', 8: e70872. Oerton, Erin, and Andreas BMC neurology Bender. 2017. 'Concordance analysis of microarray studies identifies representative gene expression changes in Parkinson’s disease: a comparison of 33 human and animal studies', 17: 58. Okamura, Katsutomo, Akira Ishizuka, Haruhiko Siomi, Mikiko C Genes Siomi, and development. 2004. 'Distinct roles for Argonaute proteins in small RNA-directed RNA cleavage pathways', 18: 1655-66. Oliveira, LMA, Lisandro Jorge Falomir-Lockhart, MG Botelho, KH Lin, P Wales, JC Koch, Elizabeth Gerhardt, H Taschenberger, TF Outeiro, P Cell death Lingor, and disease. 2015. 'Elevated α- synuclein caused by SNCA gene triplication impairs neuronal differentiation and maturation in Parkinson's patient-derived induced pluripotent stem cells', 6: e1994. Oliveras-Salvá, Marusela, Anke Van der Perren, Nicolas Casadei, Stijn Stroobants, Silke Nuber, Rudi D’Hooge, Chris Van den Haute, and Veerle Molecular neurodegeneration Baekelandt. 2013. 'rAAV2/7 vector-mediated overexpression of alpha-synuclein in mouse substantia nigra induces protein aggregation and progressive dose-dependent neurodegeneration', 8: 44. Osaka, Hitoshi, Yu-Lai Wang, Koji Takada, Shuichi Takizawa, Rieko Setsuie, Hang Li, Yae Sato, Kaori Nishikawa, Ying-Jie Sun, and Mikako Human molecular genetics Sakurai. 2003. 'Ubiquitin carboxy-terminal hydrolase L1 binds to and stabilizes monoubiquitin in neuron', 12: 1945-58. Ota, Yukio, Sawako Miyoshi, Osamu Ueda, Torayoshi Mukai, Akio Psychiatry Maeda, and Clinical Neurosciences. 1958. 'Familial paralysis agitans juvenilis. A clinical, anatomical and genetic study', 12: 112-21. Ouchi, Yasuomi, Etsuji Yoshikawa, Yoshimoto Sekine, Masami Futatsubashi, Toshihiko Kanno, Tomomi Ogusu, and Tatsuo Annals of neurology Torizuka. 2005. 'Microglial activation and dopamine terminal loss in early Parkinson's disease', 57: 168-75. Outeiro, Tiago F, Jochen Klucken, Kathryn Bercury, Julie Tetzlaff, Preeti Putcha, Luis MA Oliveira, Alexandre Quintas, Pamela J McLean, and Bradley T PloS one Hyman. 2009. 'Dopamine- induced conformational changes in alpha-synuclein', 4: e6906. Pacheco, Carla, Luis G Aguayo, and Carlos Frontiers in physiology Opazo. 2012. 'An extracellular mechanism that can explain the neurotoxic effects of α-synuclein aggregates in the brain', 3: 297. Paisán-Ruız,́ Coro, Shushant Jain, E Whitney Evans, William P Gilks, Javier Simón, Marcel Van Der Brug, Adolfo López De Munain, Silvia Aparicio, Angel Martıneź Gil, and Naheed Neuron Khan. 2004. 'Cloning of the gene containing mutations that cause PARK8-linked Parkinson's disease', 44: 595-600. Pan-Montojo, Francisco, Mathias Schwarz, Clemens Winkler, Mike Arnhold, Gregory A O'sullivan, Arun Pal, Jonas Said, Giovanni Marsico, Jean-Marc Verbavatz, and Margarita Scientific reports Rodrigo-Angulo. 2012. 'Environmental toxins trigger PD-like progression via increased alpha-synuclein release from enteric neurons in mice', 2: 898. Pang, Xueyan, Eric M Hogan, Alison Casserly, Guangping Gao, Paul D Gardner, Andrew R Molecular Tapper, and Cellular Neuroscience. 2014. 'Dicer expression is essential for adult midbrain dopaminergic neuron maintenance and survival', 58: 22-28. Parekh, Swati, Christoph Ziegenhain, Beate Vieth, Wolfgang Enard, and Ines Scientific reports Hellmann. 2016. 'The impact of amplification on differential expression analyses by RNA- seq', 6: 25533.

228

Park, June-Young, and Peter T Biochemistry Lansbury. 2003. 'β-Synuclein inhibits formation of α- synuclein protofibrils: a possible therapeutic strategy against Parkinson's disease', 42: 3696- 700. Park, Sepill, Keum Sil Lee, Young Jae Lee, Hyun Ah Shin, Hwang Yoon Cho, Kyu Chang Wang, Yong Sik Kim, Hoon Taek Lee, Kil Saeng Chung, and Eun Young Neuroscience letters Kim. 2004. 'Generation of dopaminergic neurons in vitro from human embryonic stem cells treated with neurotrophic factors', 359: 99-103. Parkinson, JAMES. 1817. 'The shaking palsy', Sherwood, Neely and Jones, London. Pasanen, Petra, Liisa Myllykangas, Maija Siitonen, Anna Raunio, Seppo Kaakkola, Jukka Lyytinen, Pentti J Tienari, Minna Pöyhönen, and Anders Paetau. 2014. 'A novel α-synuclein mutation A53E associated with atypical multiple system atrophy and Parkinson's disease-type pathology', Neurobiology of aging, 35: 2180. e1-80. e5. Pera, Martin F, Jessica Andrade, Souheir Houssami, Benjamin Reubinoff, Alan Trounson, Edouard G Stanley, Dorien Ward-van Oostwaard, and Christine J Cell Sci Mummery. 2004. 'Regulation of human embryonic stem cell differentiation by BMP-2 and its antagonist noggin', 117: 1269-80. Periquet, Magali, Morwena Latouche, Ebba Lohmann, Nina Rawal, Giuseppe De Michele, Sylvain Ricard, Hélio Teive, Valérie Fraix, Marie Vidailhet, and David Brain Nicholl. 2003. 'Parkin mutations are frequent in patients with isolated early-onset parkinsonism', 126: 1271-78. Perrier, Anselme L, Viviane Tabar, Tiziano Barberi, Maria E Rubio, Juan Bruses, Norbert Topf, Neil L Harrison, and Lorenz Proceedings of the National Academy of Sciences Studer. 2004. 'Derivation of midbrain dopamine neurons from human embryonic stem cells', 101: 12543- 48. Pertea, Mihaela, Geo M Pertea, Corina M Antonescu, Tsung-Cheng Chang, Joshua T Mendell, and Steven L Nature biotechnology Salzberg. 2015. 'StringTie enables improved reconstruction of a transcriptome from RNA-seq reads', 33: 290. Pfaffl, Michael W, Ales Tichopad, Christian Prgomet, and Tanja P Biotechnology letters Neuvians. 2004. 'Determination of stable housekeeping genes, differentially regulated target genes and sample integrity: BestKeeper–Excel-based tool using pair-wise correlations', 26: 509-15. Pham, TT, F Giesert, A Röthig, T Floss, M Kallnik, K Weindl, SM Hölter, U Ahting, H Prokisch, L Genes Becker, Brain, and Behavior. 2010. 'DJ-1-deficient mice show less TH-positive neurons in the ventral tegmental area and exhibit non-motoric behavioural impairments', 9: 305-17. Piccini, Paola, and Alan The Lancet Neurology Whone. 2004. 'Functional brain imaging in the differential diagnosis of Parkinson's disease', 3: 284-90. Piccolo, Stefano, Yoshiki Sasai, Bin Lu, and Eddy M Cell De Robertis. 1996. 'Dorsoventral patterning in Xenopus: inhibition of ventral signals by direct binding of chordin to BMP-4', 86: 589-98. Planken, Anu, Lille Kurvits, Ene Reimann, Liis Kadastik-Eerme, Külli Kingo, Sulev Kõks, and Pille BMC neurology Taba. 2017. 'Looking beyond the brain to improve the pathogenic understanding of Parkinson’s disease: implications of whole transcriptome profiling of Patients’ skin', 17: 6. Poirier, LJ, and TL Sourkes. 1965. 'INFLUENCE OF THE SUBSTANTIA NIGRA ON THE CATECHOLAMEME CONTENT OF THE STRIATUM'. Polymeropoulos, Mihael H, Joseph J Higgins, Lawrence I Golbe, William G Johnson, Susan E Ide, Giuseppe Di Iorio, Giuseppe Sanges, Edward S Stenroos, Lana T Pho, and Alejandro A Schaffer. 1996. 'Mapping of a gene for Parkinson's disease to chromosome 4q21-q23', Science, 274: 1197-99. Polymeropoulos, Mihael H, Christian Lavedan, Elisabeth Leroy, Susan E Ide, Anindya Dehejia, Amalia Dutra, Brian Pike, Holly Root, Jeffrey Rubenstein, and Rebecca Boyer. 1997. 'Mutation in the α-synuclein gene identified in families with Parkinson's disease', Science, 276: 2045-47. Prakash, Nilima, Claude Brodski, Thorsten Naserke, Eduardo Puelles, Robindra Gogoi, Anita Hall, Markus Panhuysen, Diego Echevarria, Lori Sussel, and Daniela M Vogt Development

229

Weisenhorn. 2006. 'A Wnt1-regulated genetic network controls the identity and fate of midbrain-dopaminergic progenitors in vivo', 133: 89-98. Presgraves, Steve P, Sabine Borwege, Mark J Millan, and Jeffrey N Experimental neurology Joyce. 2004. 'Involvement of dopamine D2/D3 receptors and BDNF in the neuroprotective effects of S32504 and pramipexole against 1-methyl-4-phenylpyridinium in terminally differentiated SH-SY5Y cells', 190: 157-70. Pringsheim, Tamara, Nathalie Jette, Alexandra Frolkis, and Thomas DL Movement disorders Steeves. 2014. 'The prevalence of Parkinson's disease: A systematic review and meta-analysis', 29: 1583-90. Prots, Iryna, Vanesa Veber, Stefanie Brey, Silvia Campioni, Katrin Buder, Roland Riek, Konrad J Bohm, and Beate Journal of Biological Chemistry Winner. 2013. 'alpha-Synuclein oligomers impair neuronal microtubule-kinesin interplay': jbc. M113. 451815. Pruitt, Kim D, Tatiana Tatusova, and Donna R Maglott. 2006. 'NCBI reference sequences (RefSeq): a curated non-redundant sequence database of genomes, transcripts and proteins', Nucleic acids research, 35: D61-D65. Ramonet, David, João Paulo L Daher, Brian M Lin, Klodjan Stafa, Jaekwang Kim, Rebecca Banerjee, Marie Westerlund, Olga Pletnikova, Liliane Glauser, and Lichuan PloS one Yang. 2011. 'Dopaminergic neuronal loss, reduced neurite complexity and autophagic abnormalities in transgenic mice expressing G2019S mutant LRRK2', 6: e18568. Ramos, Milagros, Araceli del Arco, Beatriz Pardo, Alberto Martınez-Serrano, Juan Ramon Martınez- Morales, Keiko Kobayashi, Tomotsugu Yasuda, Elena Bogonez, Paola Bovolenta, and Takeyori Developmental Brain Research Saheki. 2003. 'Developmental changes in the Ca2+-regulated mitochondrial aspartate–glutamate carrier aralar1 in brain and prominent expression in the spinal cord', 143: 33-46. Rapaport, Franck, Raya Khanin, Yupu Liang, Mono Pirun, Azra Krek, Paul Zumbo, Christopher E Mason, Nicholas D Socci, and Doron Genome biology Betel. 2013. 'Comprehensive evaluation of differential gene expression analysis methods for RNA-seq data', 14: 3158. Rappold, Phillip M, Mei Cui, Adrianne S Chesser, Jacqueline Tibbett, Jonathan C Grima, Lihua Duan, Namita Sen, Jonathan A Javitch, and Kim Proceedings of the National Academy of Sciences Tieu. 2011. 'Paraquat neurotoxicity is mediated by the dopamine transporter and organic cation transporter-3', 108: 20766-71. Reinhardt, Peter, Benjamin Schmid, Lena F Burbulla, David C Schöndorf, Lydia Wagner, Michael Glatza, Susanne Höing, Gunnar Hargus, Susanna A Heck, and Ashutosh Cell stem cell Dhingra. 2013. 'Genetic correction of a LRRK2 mutation in human iPSCs links parkinsonian neurodegeneration to ERK-dependent changes in gene expression', 12: 354-67. Ribe, Elena M, Ying Y Jean, Rebecca L Goldstein, Claudia Manzl, Leonidas Stefanis, Andreas Villunger, and Carol M Biochemical Journal Troy. 2012. 'Neuronal caspase 2 activity and function requires RAIDD, but not PIDD', 444: 591-99. Ribes, Vanessa, Nikolaos Balaskas, Noriaki Sasai, Catarina Cruz, Eric Dessaud, Jordi Cayuso, Samuel Tozer, Lin Lin Yang, Ben Novitch, Elisa Genes Marti, and Development. 2010. 'Distinct Sonic Hedgehog signaling dynamics specify floor plate and ventral neuronal progenitors in the vertebrate neural tube', 24: 1186-200. Riley, Brigit E, Shyra J Gardai, Dorothea Emig-Agius, Marina Bessarabova, Alexander E Ivliev, Birgit Schüle, Jeff Alexander, William Wallace, Glenda M Halliday, and J William PloS one Langston. 2014. 'Systems-based analyses of brain regions functionally impacted in Parkinson's disease reveals underlying causal mechanisms', 9: e102909. Roberts, Hazel, and David Biomolecules Brown. 2015. 'Seeking a mechanism for the toxicity of oligomeric α-synuclein', 5: 282-305. Robinson edgeR, Mark. 2009. '’Methods for differential expression in digital gene expression datasets.' in, Bioconductor.

230

Rockenstein, Edward, Margaret Mallory, Makoto Hashimoto, David Song, Clifford W Shults, Ingrid Lang, and Eliezer Journal of neuroscience research Masliah. 2002. 'Differential neuropathological alterations in transgenic mice expressing α-synuclein from the platelet- derived growth factor and Thy-1 promoters', 68: 568-78. Roelink, H, JA Porter, C Chiang, Y Tanabe, DT Chang, PA Beachy, and TM Cell Jessell. 1995. 'Floor plate and motor neuron induction by different concentrations of the amino-terminal cleavage product of sonic hedgehog autoproteolysis', 81: 445-55. Roelink, Henk, Al Augsburger, J Heemskerk, Vladimir Korzh, S Norlin, A Ruiz i Altaba, Y Tanabe, Marysia Placzek, T Edlund, and Thomas M cell Jessell. 1994. 'Floor plate and motor neuron induction by vhh-1, a vertebrate homolog of hedgehog expressed by the notochord', 76: 761-75. Roizen, Nancy J, Anne M Higgins, Kevin M Antshel, Wanda Fremont, Robert Shprintzen, and Wendy R The Journal of pediatrics Kates. 2010. '22q11. 2 deletion syndrome: are motor deficits more than expected for IQ level?', 157: 658-61. Rojanathammanee, Lalida, Eric J Murphy, and Colin K Journal of neuroinflammation Combs. 2011. 'Expression of mutant alpha-synuclein modulates microglial phenotype in vitro', 8: 44. Ryan, Scott D, Nima Dolatabadi, Shing Fai Chan, Xiaofei Zhang, Mohd Waseem Akhtar, James Parker, Frank Soldner, Carmen R Sunico, Saumya Nagar, and Maria Cell Talantova. 2013. 'Isogenic human iPSC Parkinson’s model shows nitrosative stress-induced dysfunction in MEF2-PGC1α transcription', 155: 1351-64. Saitoh, Osamu, Yoshihiro Kubo, Yoshihide Miyatani, Tomiko Asano, and Hiroyasu Nature Nakata. 1997. 'RGS8 accelerates G-protein-mediated modulation of K+ currents', 390: 525. Sano, I, Takesada Gamo, Y Kakimoto, K Taniguchi, M Takesada, and Ki Nishinuma. 1959. 'Distribution of catechol compounds in human brain', Biochimica et biophysica acta, 32: 586-87. Sardo, Valentina Lo, William Ferguson, Galina A Erikson, Eric J Topol, Kristin K Baldwin, and Ali Nature biotechnology Torkamani. 2017. 'Influence of donor age on induced pluripotent stem cells', 35: 69. Sasai, Yoshiki, Bin Lu, Herbert Steinbeisser, and Eddy M. De Robertis. 1995. 'Regulation of neural induction by the Chd and Bmp-4 antagonistic patterning signals in Xenopus', Nature, 376: 333-36. Sasaki, Hiroshi, Chi-chung Hui, Masato Nakafuku, and Hisato Development Kondoh. 1997. 'A binding site for Gli proteins is essential for HNF-3beta floor plate enhancer activity in transgenics and can respond to Shh in vitro', 124: 1313-22. Sautter, Jürgen, Morten Meyer, Christian Spenger, Rolf W Seiler, and Hans R Neuroreport Widmer. 1998. 'Effects of combined BDNF and GDNF treatment on cultured dopaminergic midbrain neurons', 9: 1093-96. Scherman, Daniel, Claire Desnos, François Darchen, Pierre Pollak, France Javoy-Agid, Yves Annals of Neurology: Official Journal of the American Neurological Association Agid, and the Child Neurology Society. 1989. 'Striatal dopamine deficiency in Parkinson's disease: role of aging', 26: 551-57. Scherzer, Clemens R, Roderick V Jensen, Steven R Gullans, and Mel B Human molecular genetics Feany. 2003. 'Gene expression changes presage neurodegeneration in a Drosophila model of Parkinson's disease', 12: 2457-66. Schneider, Bernard L, Corey R Seehus, Elizabeth E Capowski, Patrick Aebischer, Su-Chun Zhang, and Clive N Human molecular genetics Svendsen. 2007. 'Over-expression of alpha-synuclein in human neural progenitors leads to specific changes in fate and differentiation', 16: 651-66. Schurch, Nicholas J, Pietá Schofield, Marek Gierliński, Christian Cole, Alexander Sherstnev, Vijender Singh, Nicola Wrobel, Karim Gharbi, Gordon G Simpson, and Tom Rna Owen-Hughes. 2016. 'How many biological replicates are needed in an RNA-seq experiment and which differential expression tool should you use?'.

231

Schwanhäusser, Björn, Dorothea Busse, Na Li, Gunnar Dittmar, Johannes Schuchhardt, Jana Wolf, Wei Chen, and Matthias Nature Selbach. 2011. 'Global quantification of mammalian gene expression control', 473: 337. Seibler, Philip, John Graziotto, Hyun Jeong, Filip Simunovic, Christine Klein, and Dimitri Journal of Neuroscience Krainc. 2011. 'Mitochondrial Parkin recruitment is impaired in neurons derived from mutant PINK1 induced pluripotent stem cells', 31: 5970-76. Selbach, Matthias, Björn Schwanhäusser, Nadine Thierfelder, Zhuo Fang, Raya Khanin, and Nikolaus nature Rajewsky. 2008. 'Widespread changes in protein synthesis induced by microRNAs', 455: 58. Serafin, Alice, Luisa Foco, Hagen Blankenburg, Anne Picard, Stefano Zanigni, Alessandra Zanon, Peter P Pramstaller, Andrew A Hicks, and Christine BMC research notes Schwienbacher. 2014. 'Identification of a set of endogenous reference genes for miRNA expression studies in Parkinson’s disease blood samples', 7: 715. Serpell, Louise C, John Berriman, Ross Jakes, Michel Goedert, and R Anthony Proceedings of the National Academy of Sciences Crowther. 2000. 'Fiber diffraction of synthetic α-synuclein filaments shows amyloid-like cross-β conformation', 97: 4897-902. Shanker, Savita, Ariel Paulson, Howard J Edenberg, Allison Peak, Anoja Perera, Yuriy O Alekseyev, Nicholas Beckloff, Nathan J Bivens, Robert Donnelly, and Allison F Gillaspy. 2015. 'Evaluation of commercially available RNA amplification kits for RNA sequencing using very low input amounts of total RNA', Journal of biomolecular techniques: JBT, 26: 4. Sharma, Shanti, John T Hansen, and Mary FD International journal of developmental neuroscience Notter. 1990. 'Effects of NGF and dibutyryl cAMP on neuronal differentiation of embryonal carcinoma cells', 8: 33-45. Sherer, Todd B, Jin-Ho Kim, Ranjita Betarbet, and J Timothy Experimental neurology Greenamyre. 2003. 'Subcutaneous rotenone exposure causes highly selective dopaminergic degeneration and α-synuclein aggregation', 179: 9-16. Shevinsky, Lynne Hamburger, Barbara B Knowles, Ivan Damjanov, and Davor Cell Solter. 1982. 'Monoclonal antibody to murine embryos defines a stage-specific embryonic antigen expressed on mouse embryos and human teratocarcinoma cells', 30: 697-705. Shi, Min, Changqin Liu, Travis J Cook, Kristin M Bullock, Yanchun Zhao, Carmen Ginghina, Yanfei Li, Patrick Aro, Romel Dator, and Chunmei Acta neuropathologica He. 2014. 'Plasma exosomal α-synuclein is likely CNS-derived and increased in Parkinson’s disease', 128: 639-50. Shimura, Hideki, Nobutaka Hattori, Shin-ichiro Kubo, Yoshikuni Mizuno, Shuichi Asakawa, Shinsei Minoshima, Nobuyoshi Shimizu, Kazuhiro Iwai, Tomoki Chiba, and Keiji Nature genetics Tanaka. 2000. 'Familial Parkinson disease gene product, parkin, is a ubiquitin-protein ligase', 25: 302. Siddiqui, Almas, Shankar J Chinta, Jyothi K Mallajosyula, Subramanian Rajagopolan, Ingrid Hanson, Anand Rane, Simon Melov, Julie K Free Radical Biology Andersen, and Medicine. 2012. 'Selective binding of nuclear alpha-synuclein to the PGC1alpha promoter under conditions of oxidative stress may contribute to losses in mitochondrial function: implications for Parkinson’s disease', 53: 993-1003. Simeoli, Raffaele, Karli Montague, Hefin R Jones, Laura Castaldi, David Chambers, Jayne H Kelleher, Valentina Vacca, Thomas Pitcher, John Grist, and Hadil Nature communications Al-Ahdal. 2017. 'Exosomal cargo including microRNA regulates sensory neuron to macrophage communication after nerve trauma', 8: 1778. Simons, Mikael, and Graca Current opinion in cell biology Raposo. 2009. 'Exosomes–vesicular carriers for intercellular communication', 21: 575-81. Simunovic, Filip, Ming Yi, Yulei Wang, Laurel Macey, Lauren T Brown, Anna M Krichevsky, Susan L Andersen, Robert M Stephens, Francine M Benes, and Kai C Brain Sonntag. 2008. 'Gene expression profiling of substantia nigra dopamine neurons: further insights into Parkinson's disease pathology', 132: 1795-809.

232

Simunovic, Filip, Ming Yi, Yulei Wang, Robert Stephens, and Kai C PloS one Sonntag. 2010. 'Evidence for gender-specific transcriptional profiles of nigral dopamine neurons in Parkinson disease', 5: e8856. Singleton, AB, M Farrer, J Johnson, A Singleton, S Hague, J Kachergus, M Hulihan, T Peuralinna, A Dutra, and R Nussbaum. 2003. 'α-Synuclein locus triplication causes Parkinson's disease', Science, 302: 841-41. Skog, Johan, Tom Würdinger, Sjoerd Van Rijn, Dimphna H Meijer, Laura Gainche, William T Curry Jr, Bob S Carter, Anna M Krichevsky, and Xandra O Nature cell biology Breakefield. 2008. 'Glioblastoma microvesicles transport RNA and proteins that promote tumour growth and provide diagnostic biomarkers', 10: 1470. Smith, Joseph R, Ludovic Vallier, Giuseppe Lupo, Morgan Alexander, William A Harris, and Roger A Developmental biology Pedersen. 2008. 'Inhibition of Activin/Nodal signaling promotes specification of human embryonic stem cells into neuroectoderm', 313: 107-17. Smith, Wanli W, Zhong Pei, Haibing Jiang, Darren J Moore, Yideng Liang, Andrew B West, Valina L Dawson, Ted M Dawson, and Christopher A Proceedings of the National Academy of Sciences Ross. 2005. 'Leucine-rich repeat kinase 2 (LRRK2) interacts with parkin, and mutant LRRK2 induces neuronal degeneration', 102: 18676-81. Smith, William C, Anne K Knecht, Mike Wu, and Richard M Nature Harland. 1993. 'Secreted noggin protein mimics the Spemann organizer in dorsalizing Xenopus mesoderm', 361: 547. Smith, William C., and Richard M. Harland. 1991. 'Injected Xwnt-8 RNA acts early in Xenopus embryos to promote formation of a vegetal dorsalizing center', Cell, 67: 753-65. Smith, William C., and Richard M. Harland. 1992. 'Expression cloning of noggin, a new dorsalizing factor localized to the Spemann organizer in Xenopus embryos', Cell, 70: 829-40. Sohel, Mahmodul Hasan Achievements in the Life Sciences. 2016. 'Extracellular/circulating microRNAs: release mechanisms, functions and challenges', 10: 175-86. Sohel, Md Mahmodul Hasan, Michael Hoelker, Sina Seifi Noferesti, Dessie Salilew-Wondim, Ernst Tholen, Christian Looft, Franca Rings, Muhammad Jasim Uddin, Thomas E Spencer, and Karl PloS one Schellander. 2013. 'Exosomal and non-exosomal transport of extra-cellular microRNAs in follicular fluid: implications for bovine oocyte developmental competence', 8: e78505. Soldner, Frank, Josée Laganière, Albert W Cheng, Dirk Hockemeyer, Qing Gao, Raaji Alagappan, Vikram Khurana, Lawrence I Golbe, Richard H Myers, and Susan Cell Lindquist. 2011. 'Generation of isogenic pluripotent stem cells differing exclusively at two early onset Parkinson point mutations', 146: 318-31. Song, Yang, Beatrice Milon, Sandra Ott, Xuechu Zhao, Lisa Sadzewicz, Amol Shetty, Erich T Boger, Luke J Tallon, Robert J Morell, and Anup BMC genomics Mahurkar. 2018. 'A comparative analysis of library prep approaches for sequencing low input translatome samples', 19: 696. Soreq, Lilach, Alessandro Guffanti, Nathan Salomonis, Alon Simchovitz, Zvi Israel, Hagai Bergman, and Hermona PLoS computational biology Soreq. 2014. 'Long non-coding RNA and alternative splicing modulations in Parkinson's leukocytes identified by RNA sequencing', 10: e1003517. Španinger, Klemen, Arthur Sytkowski, and Nataša Open Chemistry Debeljak. 2013. 'Selection of reference genes for quantitative real-time PCR evaluation of chronic erythropoietin treatment effect on the SH-SY5Y and PC12 cells', 11: 348-57. Spellman, George G. 1962. 'Report of familial cases of parkinsonism: evidence of a dominant trait in a patient's family', Jama, 179: 372-74. Spemann, H., and H. Mangold. 2001. 'Induction of embryonic primordia by implantation of organizers from a different species. 1923', Int J Dev Biol, 45: 13-38. Spillantini, Maria Grazia, Marie Luise Schmidt, Virginia M-Y Lee, John Q Trojanowski, Ross Jakes, and Michel Goedert. 1997. 'α-Synuclein in Lewy bodies', Nature, 388: 839-40.

233

St Martin, Jessie L, Jochen Klucken, Tiago F Outeiro, Paul Nguyen, Christine Keller-McGandy, Ippolita Cantuti-Castelvetri, Tom N Grammatopoulos, David G Standaert, Bradley T Hyman, and Pamela J Journal of neurochemistry McLean. 2007. 'Dopaminergic neuron loss and up- regulation of chaperone protein mRNA induced by targeted over-expression of alpha- synuclein in mouse substantia nigra', 100: 1449-57. Stachowiak, EK, X Fang, J Myers, S Dunham, and MK Journal of neurochemistry Stachowiak. 2003. 'cAMP-induced differentiation of human neuronal progenitor cells is mediated by nuclear fibroblast growth factor receptor-1 (FGFR1)', 84: 1296-312. Streit, Andrea, Alyson J. Berliner, Costis Papanayotou, Andres Sirulnik, and Claudio D. Stern. 2000. 'Initiation of neural induction by FGF signalling before gastrulation', Nature, 406: 74-78. Su, Chunhe, Xiaopeng Yang, and Jiyu Lou. 2016. 'Geniposide reduces α-synuclein by blocking microRNA-21/lysosome-associated membrane protein 2A interaction in Parkinson disease models', Brain research, 1644: 98-106. Su, Xiaomin, Kathleen A Maguire-Zeiss, Rita Giuliano, Landa Prifti, Karthik Venkatesh, and Howard J Neurobiology of aging Federoff. 2008. 'Synuclein activates microglia in a model of Parkinson's disease', 29: 1690-701. Tabet, Fatiha, Kasey C Vickers, Luisa F Cuesta Torres, Carrie B Wiese, Bassem M Shoucri, Gilles Lambert, Claire Catherinet, Leonel Prado-Lourenco, Michael G Levin, and Seth Nature communications Thacker. 2014. 'HDL-transferred microRNA-223 regulates ICAM-1 expression in endothelial cells', 5: 3292. Taira, Takahiro, Yoshiro Saito, Takeshi Niki, Sanae MM Iguchi-Ariga, Kazuhiko Takahashi, and Hiroyoshi EMBO reports Ariga. 2004. 'DJ-1 has a role in antioxidative stress to prevent cell death', 5: 213-18. Takahashi, H, E Ohama, S Suzuki, Y Horikawa, A Ishikawa, T Morita, S Tsuji, and F Neurology Ikuta. 1994. 'Familial juvenile parkinsonism clinical and pathologic study in a family', 44: 437-37. Takahashi, Kazutoshi, Koji Tanabe, Mari Ohnuki, Megumi Narita, Tomoko Ichisaka, Kiichiro Tomoda, and Shinya cell Yamanaka. 2007. 'Induction of pluripotent stem cells from adult human fibroblasts by defined factors', 131: 861-72. Takahashi, Kazutoshi, and Shinya cell Yamanaka. 2006. 'Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors', 126: 663-76. Takenouchi, Takato, Makoto Hashimoto, Leigh J Hsu, Bartek Mackowski, Edward Rockenstein, Margaret Mallory, Eliezer Molecular Masliah, and Cellular Neuroscience. 2001. 'Reduced neuritic outgrowth and cell adhesion in neuronal cells transfected with human α-synuclein', 17: 141-50. Tan, Cheng, Xiaoyang Liu, and Jiajun Parkinson’s Disease Chen. 2018. 'Microarray Analysis of the Molecular Mechanism Involved in Parkinson’s Disease', 2018. Tang, Mianzhi, J Carlos Villaescusa, Sarah X Luo, Camilla Guitarte, Simonia Lei, Yasunori Miyamoto, Makoto M Taketo, Ernest Arenas, and Eric J Journal of Neuroscience Huang. 2010. 'Interactions of Wnt/β-catenin signaling and sonic hedgehog regulate the neurogenesis of ventral midbrain dopamine neurons', 30: 9280-91. Tanner, Caroline M, Ruth Ottman, Samuel M Goldman, Jonas Ellenberg, Piu Chan, Richard Mayeux, and J William Langston. 1999. 'Parkinson disease in twins: an etiologic study', Jama, 281: 341-46. Tanzer, Andrea, and Peter F Journal of molecular biology Stadler. 2004. 'Molecular evolution of a microRNA cluster', 339: 327-35. Taschenberger, Grit, Johan Toloe, Julia Tereshchenko, Jasper Akerboom, Pauline Wales, Roland Benz, Stefan Becker, Tiago F Outeiro, Loren L Looger, and Mathias Annals of neurology Bähr. 2013. 'β-synuclein aggregates and induces neurodegeneration in dopaminergic neurons', 74: 109- 18. Thomas, Kelly Jean, Melissa K McCoy, Jeff Blackinton, Alexandra Beilina, Marcel van der Brug, Anna Sandebring, David Miller, Dragan Maric, Angel Cedazo-Minguez, and Mark R Human

234

molecular genetics Cookson. 2010. 'DJ-1 acts in parallel to the PINK1/parkin pathway to control mitochondrial function and autophagy', 20: 40-50. Thome, Aaron D, Ashley S Harms, Laura A Volpicelli-Daley, and David G Standaert. 2016. 'microRNA- 155 regulates alpha-synuclein-induced inflammatory responses in models of parkinson disease', Journal of Neuroscience, 36: 2383-90. Thomson, James A, Joseph Itskovitz-Eldor, Sander S Shapiro, Michelle A Waknitz, Jennifer J Swiergiel, Vivienne S Marshall, and Jeffrey M science Jones. 1998. 'Embryonic stem cell lines derived from human blastocysts', 282: 1145-47. Thomson, James A, Jennifer Kalishman, Thaddeus G Golos, Maureen Durning, Charles P Harris, Robert A Becker, and John P Proceedings of the National Academy of Sciences Hearn. 1995. 'Isolation of a primate embryonic stem cell line', 92: 7844-48. Tong, Youren, Antonio Pisani, Giuseppina Martella, Maha Karouani, Hiroo Yamaguchi, Emmanuel N Pothos, and Jie Proceedings of the National Academy of Sciences Shen. 2009. 'R1441C mutation in LRRK2 impairs dopaminergic neurotransmission in mice', 106: 14622-27. Tonge, David A, Hugo T de Burgh, Reginald Docherty, Martin J Humphries, Susan E Craig, and John Brain research Pizzey. 2012. 'Fibronectin supports neurite outgrowth and axonal regeneration of adult brain neurons in vitro', 1453: 8-16. Tönges, Lars, Éva M Szegö, Patrizia Hause, Kim-Ann Saal, Lars Tatenhorst, Jan Christoph Koch, Zara dHedouville, Vivian Dambeck, Sebastian Kügler, and Christoph P Frontiers in aging neuroscience Dohm. 2014. 'Alpha-synuclein mutations impair axonal regeneration in models of Parkinson's disease', 6: 239. Trapnell, Cole, Adam Roberts, Loyal Goff, Geo Pertea, Daehwan Kim, David R Kelley, Harold Pimentel, Steven L Salzberg, John L Rinn, and Lior Pachter. 2012a. 'Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks', Nature protocols, 7: 562-78. Trapnell, Cole, Adam Roberts, Loyal Goff, Geo Pertea, Daehwan Kim, David R Kelley, Harold Pimentel, Steven L Salzberg, John L Rinn, and Lior Nature protocols Pachter. 2012b. 'Differential gene and transcript expression analysis of RNA-seq experiments with TopHat and Cufflinks', 7: 562-78. Trétiakoff, Constantin. 1919. 'Contribution à l'étude de l'anatomie pathologique du locus niger de Soemmering avec quelques déductions relatives à la pathogénie des troubles du tonus musculaire et de la maladie de Parkinson'. Tritsch, Nicolas X, Jun B Ding, and Bernardo L Nature Sabatini. 2012. 'Dopaminergic neurons inhibit striatal output through non-canonical release of GABA', 490: 262. Turchinovich, Andrey, and Barbara RNA biology Burwinkel. 2012. 'Distinct AGO1 and AGO2 associated miRNA profiles in human cells and blood plasma', 9: 1066-75. Turchinovich, Andrey, Ludmila Weiz, Anne Langheinz, and Barbara Nucleic acids research Burwinkel. 2011. 'Characterization of extracellular circulating microRNA', 39: 7223-33. Turnbull, Clare, Richard H Scott, Ellen Thomas, Louise Jones, Nirupa Murugaesu, Freya Boardman Pretty, Dina Halai, Emma Baple, Clare Craig, and Angela BMJ: British Medical Journal Hamblin. 2018. 'The 100000 genomes project: Bringing whole genome sequencing to the NHS', 361. Twine, Natalie A, Karolina Janitz, Marc R Wilkins, and Michal PloS one Janitz. 2011. 'Whole transcriptome sequencing reveals gene expression and splicing differences in brain regions affected by Alzheimer's disease', 6: e16266. Uéda, Kenji, Hisashi Fukushima, Eliezer Masliah, Yu Xia, Akihiko Iwai, Makoto Yoshimoto, DA Otero, Jun Kondo, Yasuo Ihara, and Tsunao Saitoh. 1993a. 'Molecular cloning of cDNA encoding an unrecognized component of amyloid in Alzheimer disease', Proceedings of the National Academy of Sciences, 90: 11282-86. Uéda, Kenji, Hisashi Fukushima, Eliezer Masliah, Yu Xia, Akihiko Iwai, Makoto Yoshimoto, DA Otero, Jun Kondo, Yasuo Ihara, and Tsunao Proceedings of the National Academy of Sciences

235

Saitoh. 1993b. 'Molecular cloning of cDNA encoding an unrecognized component of amyloid in Alzheimer disease', 90: 11282-86. Ulusoy, Ayse, Fabia Febbraro, Poul H Jensen, Deniz Kirik, and Marina European Journal of Neuroscience Romero-Ramos. 2010. 'Co-expression of C-terminal truncated alpha-synuclein enhances full-length alpha-synuclein-induced pathology', 32: 409-22. Ulusoy, Ayse, Raffaella Rusconi, Blanca I Pérez-Revuelta, Ruth E Musgrove, Michael Helwig, Bettina Winzen-Reichert, and Donato A EMBO molecular medicine Di Monte. 2013. 'Caudo-rostral brain spreading of α-synuclein through vagal connections', 5: 1119-27. Ungerstedt, Urban European journal of pharmacology. 1968. '6-Hydroxy-dopamine induced degeneration of central monoamine neurons', 5: 107-10. Valadi, Hadi, Karin Ekström, Apostolos Bossios, Margareta Sjöstrand, James J Lee, and Jan O Nature cell biology Lötvall. 2007. 'Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells', 9: 654. Valente, Enza Maria, Patrick M Abou-Sleiman, Viviana Caputo, Miratul MK Muqit, Kirsten Harvey, Suzana Gispert, Zeeshan Ali, Domenico Del Turco, Anna Rita Bentivoglio, and Daniel G Science Healy. 2004. 'Hereditary early-onset Parkinson's disease caused by mutations in PINK1', 304: 1158-60. Valente, Enza Maria, Anna Rita Bentivoglio, Peter H Dixon, Alessandro Ferraris, Tamara Ialongo, Marina Frontali, Alberto Albanese, and Nicholas W The American Journal of Human Genetics Wood. 2001. 'Localization of a novel locus for autosomal recessive early-onset parkinsonism, PARK6, on human chromosome 1p35-p36', 68: 895-900. Valenzuela, David M, Aris N Economides, Eduardo Rojas, Teresa M Lamb, L Nunez, Pam Jones, NY Lp, R Espinosa, Camilynn I Brannan, and Debra J Journal of Neuroscience Gilbert. 1995. 'Identification of mammalian noggin and its expression in the adult nervous system', 15: 6077-84. Van Den Eeden, Stephen K, Caroline M Tanner, Allan L Bernstein, Robin D Fross, Amethyst Leimpeter, Daniel A Bloch, and Lorene M American journal of epidemiology Nelson. 2003. 'Incidence of Parkinson’s disease: variation by age, gender, and race/ethnicity', 157: 1015- 22. van Rooijen, Bart D, Mireille MAE Claessens, and Vinod FEBS letters Subramaniam. 2008. 'Membrane binding of oligomeric α-synuclein depends on bilayer charge and packing', 582: 3788-92. Vella, Laura J, Robyn A Sharples, Rebecca M Nisbet, Roberto Cappai, and Andrew F European Biophysics Journal Hill. 2008. 'The role of exosomes in the processing of proteins associated with neurodegenerative diseases', 37: 323-32. Vera, Elsa, Nazario Bosco, and Lorenz Cell reports Studer. 2016. 'Generating late-onset human iPSC- based disease models by inducing neuronal age-related phenotypes through telomerase manipulation', 17: 1184-92. Vickers, Kasey C, Brian T Palmisano, Bassem M Shoucri, Robert D Shamburek, and Alan T Nature cell biology Remaley. 2011. 'MicroRNAs are transported in plasma and delivered to recipient cells by high-density lipoproteins', 13: 423. Villarroya-Beltri, Carolina, Cristina Gutiérrez-Vázquez, Fátima Sánchez-Cabo, Daniel Pérez- Hernández, Jesús Vázquez, Noa Martin-Cofreces, Dannys Jorge Martinez-Herrera, Alberto Pascual-Montano, María Mittelbrunn, and Francisco Nature communications Sánchez- Madrid. 2013. 'Sumoylated hnRNPA2B1 controls the sorting of miRNAs into exosomes through binding to specific motifs', 4: 2980. Vo, Ngan, Matthew E Klein, Olga Varlamova, David M Keller, Tadashi Yamamoto, Richard H Goodman, and Soren Impey. 2005. 'A cAMP-response element binding protein-induced microRNA regulates neuronal morphogenesis', Proc Natl Acad Sci U S A, 102: 16426-31. Vogel, Christine, and Edward M Nature Reviews Genetics Marcotte. 2012. 'Insights into the regulation of protein abundance from proteomic and transcriptomic analyses', 13: 227.

236

Vogiatzi, Tereza, Maria Xilouri, Kostas Vekrellis, and Leonidas Journal of Biological Chemistry Stefanis. 2008. 'Wild type a-synuclein is degraded by chaperone mediated autophagy and macroautophagy in neuronal cells'. Volkmann, J, N Allert, J Voges, PH Weiss, H-J Freund, and V Neurology Sturm. 2001. 'Safety and efficacy of pallidal or subthalamic nucleus stimulation in advanced PD', 56: 548-51. Volpicelli-Daley, Laura A, Kelvin C Luk, Tapan P Patel, Selcuk A Tanik, Dawn M Riddle, Anna Stieber, David F Meaney, John Q Trojanowski, and Virginia M-Y Neuron Lee. 2011. 'Exogenous α- synuclein fibrils induce Lewy body pathology leading to synaptic dysfunction and neuron death', 72: 57-71. von Campenhausen, Sonja, Bernhard Bornschein, Regina Wick, Kai Bötzel, Cristina Sampaio, Werner Poewe, Wolfgang Oertel, Uwe Siebert, Karin Berger, and Richard European Neuropsychopharmacology Dodel. 2005. 'Prevalence and incidence of Parkinson's disease in Europe', 15: 473-90. Wang, Tong, Yizhi Cui, Jingjie Jin, Jiahui Guo, Guibin Wang, Xingfeng Yin, Qing-Yu He, and Gong Nucleic acids research Zhang. 2013. 'Translating mRNAs strongly correlate to proteins in a multivariate manner and their translation ratios are phenotype specific', 41: 4743-54. Wang, Wei, Iva Perovic, Johnathan Chittuluru, Alice Kaganovich, Linh TT Nguyen, Jingling Liao, Jared R Auclair, Derrick Johnson, Anuradha Landeru, and Alana K Proceedings of the National Academy of Sciences Simorellis. 2011. 'A soluble α-synuclein construct forms a dynamic tetramer': 201113260. Wang, Yangming, Rostislav Medvid, Collin Melton, Rudolf Jaenisch, and Robert Nature genetics Blelloch. 2007. 'DGCR8 is essential for microRNA biogenesis and silencing of embryonic stem cell self-renewal', 39: 380. Ward, Christopher D, Roger C Duvoisin, Susan E Ince, John D Nutt, Roswell Eldridge, and Donald B Calne. 1983. 'Parkinson's disease in 65 pairs of twins and in a set of quadruplets', Neurology, 33: 815-15. Webb, Ashley E, and Anne Trends in biochemical sciences Brunet. 2014. 'FOXO transcription factors: key regulators of cellular quality control', 39: 159-69. Webb, Julie L, Brinda Ravikumar, Jane Atkins, Jeremy N Skepper, and David C Journal of Biological Chemistry Rubinsztein. 2003. 'α-Synuclein is degraded by both autophagy and the proteasome', 278: 25009-13. Weber, Jessica A, David H Baxter, Shile Zhang, David Y Huang, Kuo How Huang, Ming Jen Lee, David J Galas, and Kai Clinical chemistry Wang. 2010. 'The microRNA spectrum in 12 body fluids', 56: 1733-41. Wei, Lei, Congcong Sun, Ming Lei, Guofei Li, Li Yi, Feifei Luo, Yi Li, Li Ding, Zhuolin Liu, and Shaomin Journal of Molecular Neuroscience Li. 2013. 'Activation of Wnt/β-catenin pathway by exogenous Wnt1 protects SH-SY5Y cells against 6-hydroxydopamine toxicity', 49: 105-15. Weinreb, Paul H, Weiguo Zhen, Anna W Poon, Kelly A Conway, and Peter T Biochemistry Lansbury. 1996. 'NACP, a protein implicated in Alzheimer's disease and learning, is natively unfolded', 35: 13709-15. Werner, Pia, Nicolas Hussy, Gary Buell, Kenneth A Jones, and R Alan Molecular pharmacology North. 1996. 'D2, D3, and D4 dopamine receptors couple to G protein-regulated potassium channels in Xenopus oocytes', 49: 656-61. Wilhelm, Brian T, and Josette-Renée Methods Landry. 2009. 'RNA-Seq—quantitative measurement of expression through massively parallel RNA-sequencing', 48: 249-57. Winter, Julia, Stephanie Jung, Sarina Keller, Richard I Gregory, and Sven Nature cell biology Diederichs. 2009. 'Many roads to maturity: microRNA biogenesis pathways and their regulation', 11: 228. Wong, Yvette C, and Dimitri Nature medicine Krainc. 2017. 'α-synuclein toxicity in neurodegeneration: mechanism and therapeutic strategies', 23: 1.

237

Woodard, Chris M, Brian A Campos, Sheng-Han Kuo, Melissa J Nirenberg, Michael W Nestor, Matthew Zimmer, Eugene V Mosharov, David Sulzer, Hongyan Zhou, and Daniel Cell reports Paull. 2014. 'iPSC-derived dopamine neurons reveal differences between monozygotic twins discordant for Parkinson’s disease', 9: 1173-82. Xiang, Mengqin, Ying Zeng, Ruirui Yang, Haifan Xu, Zhuo Chen, Jing Zhong, Hailong Xie, Yinghui Xu, Xi Biochemical Zeng, and biophysical research communications. 2014. 'U6 is not a suitable endogenous control for the quantification of circulating microRNAs', 454: 210-14. Xicoy, Helena, Bé Wieringa, and Gerard JM Molecular neurodegeneration Martens. 2017. 'The SH- SY5Y cell line in Parkinson’s disease research: a systematic review', 12: 10. Xin, Hongqi, Yi Li, Ben Buller, Mark Katakowski, Yi Zhang, Xinli Wang, Xia Shang, Zheng Gang Zhang, and Michael Chopp. 2012. 'Exosome-mediated transfer of miR-133b from multipotent mesenchymal stromal cells to neural cells contributes to neurite outgrowth', Stem cells, 30: 1556-64. Xiong, Hui, Danling Wang, Linan Chen, Yeun Su Choo, Hong Ma, Chengyuan Tang, Kun Xia, Wei Jiang, Ze’ev Ronai, and Xiaoxi The Journal of clinical investigation Zhuang. 2009. 'Parkin, PINK1, and DJ-1 form a ubiquitin E3 ligase complex promoting unfolded protein degradation', 119: 650-60. Xiong, Ran, Zhiquan Wang, Zongbo Zhao, Hui Li, Wei Chen, Bei Zhang, Liling Wang, Li Wu, Wen Li, and Jianqing Ding. 2014. 'MicroRNA-494 reduces DJ-1 expression and exacerbates neurodegeneration', Neurobiology of aging, 35: 705-14. Xu, Ping, Madhumita Das, Judith Reilly, Roger J Genes Davis, and development. 2011. 'JNK regulates FoxO-dependent autophagy in neurons', 25: 310-22. Yang, Dehua, Ting Li, Yi Wang, Yuanjia Tang, Huijuan Cui, Yu Tang, Xiaojie Zhang, Degui Chen, Nan Shen, and Weidong Le. 2012. 'miR-132 regulates the differentiation of dopamine neurons by directly targeting Nurr1 expression', J Cell Sci, 125: 1673-82. Yang, Fang, Ya-ping Yang, Cheng-jie Mao, Ling Liu, Hui-fen Zheng, Li-fang Hu, and Chun-feng Acta Pharmacologica Sinica Liu. 2013. 'Crosstalk between the proteasome system and autophagy in the clearance of α-synuclein', 34: 674. Yi, Rui, Yi Qin, Ian G Macara, Bryan R Genes Cullen, and development. 2003. 'Exportin-5 mediates the nuclear export of pre-microRNAs and short hairpin RNAs', 17: 3011-16. Yokochi, Masayuki European neurology. 1997. 'Familial juvenile parkinsonism', 38: 29-33. Yoo, Myung S, Hong S Chun, Jessica J Son, Lorraine A DeGiorgio, Dae J Kim, Chu Peng, and Jin H Molecular brain research Son. 2003. 'Oxidative stress regulated genes in nigral dopaminergic neuronal cells: correlation with the known pathology in Parkinson’s disease', 110: 76-84. Yu, Zhongwang, Xiaohui Xu, Zhenghua Xiang, Jianfeng Zhou, Zhaohuan Zhang, Chun Hu, and Cheng PloS one He. 2010. 'Nitrated α-synuclein induces the loss of dopaminergic neurons in the substantia nigra of rats', 5: e9956. Yuan, Huabing, Nicoletta Corbi, Claudio Basilico, Lisa Genes Dailey, and development. 1995. 'Developmental-specific activity of the FGF-4 enhancer requires the synergistic action of Sox2 and Oct-3', 9: 2635-45. Zarranz, Juan J, Javier Alegre, Juan C Gómez-Esteban, Elena Lezcano, Raquel Ros, Israel Ampuero, Lídice Vidal, Janet Hoenicka, Olga Rodriguez, and Begona Atarés. 2004. 'The new mutation, E46K, of α-synuclein causes parkinson and Lewy body dementia', Annals of neurology, 55: 164-73. Zasso, Jacopo, Mastad Ahmed, Alessandro Cutarelli, and Luciano bioRxiv Conti. 2018. 'Inducible alpha-synuclein overexpression affects human Neural Stem Cells behavior': 290619. Zhang, Dawei, Shanzheng Yang, Enrique M Toledo, Daniel Gyllborg, Carmen Saltó, J Carlos Villaescusa, and Ernest Sci. Signal. Arenas. 2017. 'Niche-derived laminin-511 promotes midbrain dopaminergic neuron survival and differentiation through YAP', 10: eaal4165.

238

Zhang, Tie-mei, Shu-yang Yu, Peng Guo, Yang Du, Yang Hu, Ying-shan Piao, Li-jun Zuo, Teng-hong Lian, Rui-dan Wang, and Qiu-jin Medicine Yu. 2016. 'Nonmotor symptoms in patients with Parkinson disease: A cross-sectional observational study', 95. Zhang, Wei, Shannon Dallas, Dan Zhang, Jian-Ping Guo, Hao Pang, Belinda Wilson, David S Miller, Biao Chen, Wanqin Zhang, and Patrick L Glia McGeer. 2007. 'Microglial PHOX and Mac-1 are essential to the enhanced dopaminergic neurodegeneration elicited by A30P and A53T mutant alpha-synuclein', 55: 1178-88. Zhang, Wei, Tongguang Wang, Zhong Pei, David S Miller, Xuefei Wu, Michelle L Block, Belinda Wilson, Wanqin Zhang, Yong Zhou, and Jau-Shyong The FASEB Journal Hong. 2005. 'Aggregated α-synuclein activates microglia: a process leading to disease progression in Parkinson’s disease', 19: 533-42. Zhang, Xingtan, Benjamin D Rosen, Haibao Tang, Vivek Krishnakumar, and Christopher D PLoS One Town. 2015. 'Polyribosomal RNA-Seq reveals the decreased complexity and diversity of the Arabidopsis translatome', 10: e0117699. Zhang, Xiong, Rui Yang, Bei-Lei Hu, Pengcheng Lu, Li-Li Zhou, Zhi-Yong He, Hong-Mei Wu, and Jian- Hong Zhu. 2017. 'Reduced circulating levels of miR-433 and miR-133b are potential biomarkers for Parkinson’s disease', Frontiers in cellular neuroscience, 11: 170. Zhang, Yanli, Michael James, Frank A Middleton, and Richard L American Journal of Medical Genetics Part B: Neuropsychiatric Genetics Davis. 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', 137: 5-16. Zhang, Zhelin, and Yan The Scientific World Journal Cheng. 2014. 'miR-16-1 promotes the aberrant α-synuclein accumulation in parkinson disease via targeting heat shock protein 70', 2014. Zhang, Zong Hong, Dhanisha J Jhaveri, Vikki M Marshall, Denis C Bauer, Janette Edson, Ramesh K Narayanan, Gregory J Robinson, Andreas E Lundberg, Perry F Bartlett, and Naomi R PloS one Wray. 2014. 'A comparative study of techniques for differential expression analysis on RNA- Seq data', 9: e103207. Zhao, Shanrong, Wai-Ping Fung-Leung, Anton Bittner, Karen Ngo, and Xuejun PloS one Liu. 2014. 'Comparison of RNA-Seq and microarray in transcriptome profiling of activated T cells', 9: e78644. Zhou, RM, YX Huang, XL Li, C Chen, Q Shi, GR Wang, C Tian, ZY Wang, YY Jing, and C Molecular biology reports Gao. 2010. 'Molecular interaction of α-synuclein with tubulin influences on the polymerization of microtubule in vitro and structure of microtubule in cells', 37: 3183- 92. Zhou, Wenbo, and Curt R Journal of Biological Chemistry Freed. 2005. 'DJ-1 up-regulates glutathione synthesis during oxidative stress and inhibits A53T α-synuclein toxicity', 280: 43150-58. Zhou, Yan, Ming Lu, Ren-Hong Du, Chen Qiao, Chun-Yi Jiang, Ke-Zhong Zhang, Jian-Hua Ding, and Gang Hu. 2016. 'MicroRNA-7 targets Nod-like receptor protein 3 inflammasome to modulate neuroinflammation in the pathogenesis of Parkinson’s disease', Molecular neurodegeneration, 11: 28. Zimprich, Alexander, Saskia Biskup, Petra Leitner, Peter Lichtner, Matthew Farrer, Sarah Lincoln, Jennifer Kachergus, Mary Hulihan, Ryan J Uitti, and Donald B Neuron Calne. 2004. 'Mutations in LRRK2 cause autosomal-dominant parkinsonism with pleomorphic pathology', 44: 601-07. Zondler, Lisa, Marcus Kostka, Patrick Garidel, Udo Heinzelmann, Bastian Hengerer, Benjamin Mayer, Jochen H Weishaupt, Frank Gillardon, and Karin M PloS one Danzer. 2017. 'Proteasome impairment by α-synuclein', 12: e0184040. Zou, Yanan, Zixuan Chen, Brett L Jennings, Guannan Zhao, Qingqing Gu, Anindya Bhattacharya, Yan Cui, Bo Yu, Kafait U Malik, and Junming Scientific reports Yue. 2018. 'Deletion of DGCR8 in

239

VSMCs of adult mice results in loss of vascular reactivity, reduced blood pressure and neointima formation', 8: 1468.

240

Appendix 1 - SHSY5Y transduction with EGFP – flow cytometry - promoter test Untransduced Control

Control MOI 2 MOI 5 CMV 45.7 79.1 CAG 47.3 84.6 EF1 27.1 55.6 PGK 58.0 84.0

CMV MOI 2

241

CMV MOI 5

CAG MOI 2

242

CAG MOI 5

EF1 MOI 2

243

EF1 MOI 5

PGK MOI 2

244

PGK MOI 5

245

Appendix 2 – SHSY5Y promoter test – FACS sorting for EGFP expression CMV – MOI 2

246

CAG MOI 2

247

EF1 MOI 2

248

PGK MOI 2

249

Appendix 3 – SHSY5Y PD model and control cells transduced with CMV-L10a- EGFP lentivirus particles (MOI 2) – FACS sorted for EGFP expression Control 1

250

Control 2

251

Wildtype 1

252

Wildtype 2

253

Wildtype 3

254

A53T 1

255

A53T 2

256

A53T 3

257

Appendix 4 – RNA integrity of affinity purified samples from SHSY5Y PD model

Total 28S/18S 28S/18S rRNA Well RINe Conc. [pg/µl] Sample Description Alert Observations RNA (Height) (Area) Area Area A0 - - - 3550 Ladder - - B1 8.2 1.2 1.4 2430 C1 5.61 1.96 C1 8.0 1.1 1.5 7910 C2 18.29 6.59 D1 7.9 1.2 1.5 10800 C3 24.87 9.16 E1 7.9 1.2 1.5 6630 W1 15.34 5.44 F1 7.9 1.1 1.6 2900 W2 6.72 2.35 G1 7.6 0.8 1.1 10500 W3 24.30 8.01 RNA concentration H1 7.3 0.9 0.8 32600 A1 ! outside 75.46 27.63 recommended range for RINe A2 7.8 1.0 1.2 13200 A2 30.43 10.83 B2 7.5 0.9 1.0 11200 A3 25.93 9.79

258

Appendix 5 – RNA amplification – dsDNA

D1000 Contrast: 0.50

Well Conc. [ng/µl] Sample Description Alert Observations A0 20.3 D1000 Ladder Ladder A1 8.20 MK_C1 B1 7.44 MK_C2 C1 4.53 MK_C3 D1 6.78 MK_W1 E1 6.40 MK_W2 F1 4.60 MK_W3 G1 10.6 MK_A1 H1 6.34 MK_A2 A2 5.44 MK_A3 B2 blank

259

Control 1

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 6.05 - 372 - Marker 42 4.61 - 167 56.15 195 3.60 - 28.3 43.85 Upper 1,500 6.50 6.50 6.67 - Marker

Control 2

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 6.15 - 379 - Marker 43 3.56 - 129 47.85 190 3.88 - 31.3 52.15 Upper 1,500 6.50 6.50 6.67 - Marker

260

Control 3

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 6.16 - 379 - Marker 43 2.90 - 105 63.91 197 1.64 - 12.8 36.09

Wildtype 1

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 6.05 - 372 - Marker 42 4.78 - 173 70.39 194 2.01 - 15.9 29.61 Upper 1,500 6.50 6.50 6.67 - Marker

261

Wildtype 2

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 6.06 - 373 - Marker 42 3.60 - 131 56.20 188 2.80 - 22.9 43.80 Upper 1,500 6.50 6.50 6.67 - Marker

Wildtype 3

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 6.18 - 381 - Marker 42 3.27 - 119 71.09 193 1.33 - 10.6 28.91 Upper 1,500 6.50 6.50 6.67 - Marker

262

A53T 1

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 6.23 - 383 - Marker 42 4.00 - 147 37.64 74 1.19 - 24.8 11.20 182 5.43 - 45.9 51.16 Upper 1,500 6.50 6.50 6.67 - Marker

A53T 2

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 6.14 - 378 - Marker 42 4.54 - 166 71.55 74 0.697 - 14.5 10.99 189 1.11 - 9.00 17.46 Upper 1,500 6.50 6.50 6.67 - Marker

263

A53T 3

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 6.40 - 394 - Marker 42 4.05 - 148 74.52 195 1.39 - 10.9 25.48 Upper 1,500 6.50 6.50 6.67 - Marker

264

Appendix 6 – Universal/index adapters, amplification primers and P5/P7 adapter sequences Universal and index adapter sequences

Index Primer Sample Sequence Sequenc e Universal AATGATACGGCGACCACCGAGATCTACACTCTTTCCCTAC

Adapter ACGACGCTCTTCCGATCT Control /5Phos/GATCGGAAGAGCACACGTCTGAACTCCAGTCACA Index 1 1 TCACGATCTCGTATGCCGTCTTCTGCTTG ATCACG Control /5Phos/GATCGGAAGAGCACACGTCTGAACTCCAGTCACC Index 2 2 GATGTATCTCGTATGCCGTCTTCTGCTTG CGATGT Control /5Phos/GATCGGAAGAGCACACGTCTGAACTCCAGTCACT Index 3 3 TAGGCATCTCGTATGCCGTCTTCTGCTTG TTAGGC Wildtype /5Phos/GATCGGAAGAGCACACGTCTGAACTCCAGTCACT Index 4 1 GACCAATCTCGTATGCCGTCTTCTGCTTG TGACCA Wildtype /5Phos/GATCGGAAGAGCACACGTCTGAACTCCAGTCACA Index 5 2 CAGTGATCTCGTATGCCGTCTTCTGCTTG ACAGTG Wildtype /5Phos/GATCGGAAGAGCACACGTCTGAACTCCAGTCACG Index 6 3 CCAATATCTCGTATGCCGTCTTCTGCTTG GCCAAT Mutant /5Phos/GATCGGAAGAGCACACGTCTGAACTCCAGTCACC Index 7 A53T 1 AGATCATCTCGTATGCCGTCTTCTGCTTG CAGATC Mutant /5Phos/GATCGGAAGAGCACACGTCTGAACTCCAGTCACA Index 8 A53T 2 CTTGAATCTCGTATGCCGTCTTCTGCTTG ACTTGA Mutant /5Phos/GATCGGAAGAGCACACGTCTGAACTCCAGTCACG Index 9 A53T 3 ATCAGATCTCGTATGCCGTCTTCTGCTTG GATCAG

Amplification primers

Primer Sequence Amp1 AATGATACGGCGACCACCGA Amp2 CAAGCAGAAGACGGCATACGA

P5 and P7 primers: from NEBNext library quant kit for illumina platform (also used for sequencing by illumina chemistry)

Primer Sequence P5: 5' AAT GAT ACG GCG ACC ACC GA 3' P7: 5' CAA GCA GAA GAC GGC ATA CGA GAT 3'

265

Appendix 7 – Adapter ligated Amplicon QC

D1000 Contrast: 0.00

Well Conc. [ng/µl] Sample Description Alert Observations A0 20.3 D1000 Ladder Ladder A1 0.848 MK_C1 B1 0.701 MK_C2 C1 1.28 MK_C3 D1 1.01 KM_W1 E1 0.748 KM_W2 F1 0.925 KM_W3 G1 1.00 MK_A1 H1 1.17 MK_A2 A2 1.25 KM_A3rep

266

Control 1

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 7.40 - 455 - Marker 65 0.204 - 4.82 24.02 120 0.0243 - 0.313 2.87 213 0.620 - 4.48 73.11 Upper 1,500 6.50 6.50 6.67 - Marker

Control 2

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 7.51 - 462 - Marker 63 0.238 - 5.80 33.93 115 0.0725 - 0.969 10.34 232 0.391 - 2.59 55.72 Upper 1,500 6.50 6.50 6.67 - Marker

267

Control 3

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 7.07 - 435 - Marker 66 0.179 - 4.20 14.04 116 0.0620 - 0.820 4.86 220 0.802 - 5.60 62.81 313 0.204 - 1.00 15.99 608 0.0293 - 0.0740 2.29 Upper 1,500 6.50 6.50 6.67 - Marker

Wildtype 1

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 6.67 - 410 - Marker 67 0.286 - 6.53 28.27 121 0.0573 - 0.731 5.66 325 0.669 - 3.17 66.07 Upper 1,500 6.50 6.50 6.67 - Marker

268

Wildtype 2

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 7.45 - 459 - Marker 64 0.216 - 5.18 28.90 118 0.0309 - 0.401 4.13 319 0.501 - 2.42 66.97 Upper 1,500 6.50 6.50 6.67 - Marker

Wildtype 3

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 6.78 - 417 - Marker 67 0.383 - 8.83 41.42 118 0.0725 - 0.947 7.84 320 0.469 - 2.26 50.74 Upper 1,500 6.50 6.50 6.67 - Marker

269

A53T 1

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 7.56 - 466 - Marker 65 0.224 - 5.34 22.38 118 0.0601 - 0.786 6.00 320 0.717 - 3.45 71.62 Upper 1,500 6.50 6.50 6.67 - Marker

A53T 2

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 7.31 - 450 - Marker 63 0.318 - 7.83 27.31 117 0.0618 - 0.816 5.30 318 0.786 - 3.80 67.39 Upper 1,500 6.50 6.50 6.67 - Marker

270

A53T 3

Calibrated Conc. Assigned Conc. Peak Molarity % Integrated Size [bp] Peak Comment Observations [ng/µl] [ng/µl] [nmol/l] Area Lower 25 7.39 - 455 - Marker 62 0.173 - 4.28 13.80 112 0.114 - 1.57 9.13 221 0.693 - 4.82 55.33 307 0.272 - 1.36 21.74 Upper 1,500 6.50 6.50 6.67 - Marker

271

272

273

274