The Role of Retromer in Parkinson’s disease

Jordan G. Follett B.BiomedSc (Hons)

A thesis submitted for the degree of Doctor of Philosophy at The University of Queensland in 2015

Institute for Molecular Biosciences

i

Abstract: Parkinson’s disease (PD) is an irreversible neurological disorder characterized by protein aggregate accumulation, endo-lysosomal dysfunction and neuronal cell death. Recently, several novel point mutations linked to late onset PD were identified in the retromer subunit; Vps35. Retromer is a highly conserved protein sorting complex that governs the endosome-to-Trans Golgi Network (TGN) trafficking pathway (hereafter referred to as retrograde sorting). While a great deal of evidence currently exists detailing the mechanisms of retromer in the retrograde sorting process, very little information conclusively highlights how sorting is misguided by retromer in PD.

Firstly, this thesis investigated the importance of retromer in regulating overall lysosomal function and clearance of α-synuclein. This study demonstrated the consistent accumulation of aggregated α-synuclein in Rab7A positive compartments following stable depletion of the retromer complex. Upon use of exogenous stimuli, a large percentage of retromer depleted cells displayed aggregates positive for α-synuclein, when compared to non-silenced cells. Further, this accumulation in the late endocytic network negatively regulated the ability for material derived from the TGN to be appropriately processed in the late endosomal network, further perturbing the ability for aggregated α-synuclein protein to be cleared and likely contributing to the death of the cell.

Secondly, I identified several novel shortfalls of retromer in the various point mutations that are linked to PD and additionally, expanded on how retromer plays a fundamental role in preservation of the endo-lysosomal network. This thesis demonstrates that Vps35 D620N, Vps35 P316S or Vps35 R524W do not disrupt the overall formation of the retromer complex, but functionally demonstrates incorrect sorting of specific cargo to the TGN. The failed sorting observed by Vps35 D620N containing retromer results in vacuolization of the late endosome, and large disruptions to the overall localization of endocytic compartments. As a result, the lysosomal hydrolase, Cathepsin D, an enzyme needed for the degradation of α-synuclein, is not transported to the late endocytic network and can be detected in the extracellular medium. Subsequent investigation into additional point mutations (Vps35 P316S and Vps35 R524W) highlighted the inability for Vps35 R524W containing retromer to correctly associate with the endosome compartment. While Vps35 P316S containing retromer appears to function identical to that of the wild type retromer, Vps35 R524W negatively regulates retromer-dependent machinery at the endosome surface, leads to the mislocalization of receptors and delays the ability of α- synuclein to be cleared from the cell. Collectively, it is clear that the findings presented here ii strongly implicate the PD-linked Vps35 variants as loss-of-function mutations and lead to the presence of cellular phenotypes witnessed in PD.

Lastly, I investigated the use of a pharmacological agent to enhance retromer’s function that consequently, aided in significant clearance of α-synuclein inclusions that were observed following the depletion of Vps35 or expression of PD linked mutations Vps35 P316S and Vps35 R524W. These findings support the notion that therapeutic intervention against retromer may aid in delaying the cellular phenotypes observed in PD tissue and subsequently, bridge the gap between the mammalian retromer complex, the lysosome and human disease.

iii

Declaration by author:

This thesis is composed of my original work, and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my research higher degree candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and, subject to the policy and procedures of The University of Queensland, the thesis be made available for research and study in accordance with the Copyright Act 1968 unless a period of embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis.

iv

Publications during candidature: The introduction is primarily based on the following invited review article: Retromer’s Role in Endosomal Trafficking and Impaired function in Neurodegenerative Diseases

Authored by:

Jordan Follett1, Andrea Bugarcic1, Brett M. Collins1 and Rohan D. Teasdale1*

Review was written for a 2015 special edition review series titled “Proteomic alterations by mutations involved in Parkinson’s disease and related disorders” by Current Peptide and Protein Science.

Chapter 4: Chapter is based solely on the manuscript: The Vps35 D620N mutation linked to Parkinson’s disease Disrupts the Cargo Sorting Function of Retromer Authored by: Jordan Follett, Suzanne J. Norwood, Nicholas A. Hamilton, Megha Mohan, Oleksiy Kovtun, Stephanie Tay, Yang Zhe, Stephen A. Wood, George D. Mellick, Peter A. Silburn, Brett M. Collins, Andrea Bugarcic, and Rohan D. Teasdale

Manuscript was published in: Traffic, 2014 Feb; 15 (2):230-44

Chapter 5: Chapter will be solely based on the manuscript: Parkinson’s disease linked mutation in Vps35 impairs endosome association of retromer and impairs degradation of α-synuclein

Authored by: Jordan Follett, Andrea Bugarcic, Zhe Yang, Suzanne J. Norwood, Brett M. Collins & Rohan D. Teasdale

Manuscript is currently under review at Human Molecular Genetics.

v

Presentations and conference attendance: 2015 Hunter Cell Biology Meeting, Hunter Valley, Australia: Poster title: “Parkinson’s disease linked mutation in Vps35 impairs endosome association of retromer” 9th Alzheimer’s + Parkinson’s disease (A+PD) Symposium, The University of Queensland, Australia: Poster title: “Parkinson’s disease linked mutation in Vps35 impairs endosome association of retromer” Molecular Cell Biology Division Forum, IMB

2014 ComBio Meeting, Canberra, Australia: Oral title: “PARKINSON’S DISEASE LINKED MUTATION ALTERS THE FUNCTION OF RETROMER” Gordon Research Conference: Neurobiology of Brains Disorders, Girona, Spain. Poster title: The role of the Retromer complex in Parksinon’s Disease Focus on Microscopy Conference, The University of Sydney, NSW, Australia.

2013 Molecular Cell Biology Division Forum 7th Annual Alzheimer’s and Parkinson’s disease Symposium, University of Queensland, Australia: Poster Title: “The Vps35 D620N mutation linked to Parkinson's disease disrupts the cargo sorting function of retromer”

2012 Molecular Cell Biology Division Forum

vi

Publications included in this thesis:

Publication citation – incorporated as Chapter 1. Retromer’s Role in Endosomal Trafficking and Impaired function in Neurodegenerative Diseases.

Authored by: Jordan Follett, Andrea Bugarcic, Brett M. Collins and Rohan D. Teasdale.

Review was written for a 2015 special edition review series titled “Proteomic alterations by mutations involved in Parkinson’s disease and related disorders” by Current Peptide and Protein Science and is being reviewed.

Contributor Statement of contribution

Jordan Follett (Candidate) Wrote and edited drafts of the paper (70%)

Andrea Bugarcic Wrote and edited drafts of the paper (15%)

Brett. M. Collins Wrote and edited drafts of the paper (5%) Rohan D. Teasdale Edited drafts of the paper

(10%)

Publication citation – incorporated as Chapter 4 and attached as Appendix

The Vps35 D620N mutation linked to Parkinson’s disease Disrupts the Cargo Sorting Function of Retromer

Authored by: Jordan Follett, Suzanne J. Norwood, Nicholas A. Hamilton, Megha Mohan, Oleksiy Kovtun, Stephanie Tay, Yang Zhe, Stephen A. Wood, George D. Mellick, Peter A. Silburn, Brett M. Collins, Andrea Bugarcic, and Rohan D. Teasdale

Manuscript was published in Traffic, 2014 Feb; 15 (2):230-44. See next page for author contributions.

vii

Contributor Statement of contribution Jordan Follett Generated and validated DNA constructs used in (Candidate) study, designed and conducted experiments with the exception of Figure 4.1 A-C (70%). Wrote the paper

(70%)

Suzanne J. Protein purification, bacterial expression construct

Norwood design, ITC and CD Spectra used in Figures 4.1 A-C

(40%) Wrote the paper (0%) Nicholas A. Designed analysis tool used in Figure 4.2C and 4.5C Hamilton (100%) Wrote the paper (0%) Megha Mohan Supplied reagent used in Figures 4.5 & 4.6 Oleksiy Kovtun, Protein purification, bacterial expression construct

design, ITC and CD Spectra used in Figures 4.1 A-C (10%) Wrote the paper (0%)

Stephanie Tay Protein purification, bacterial expression construct

design, ITC and CD Spectra used in Figures 4.1 A-C (10%) Wrote the paper (0%)

Zhe Yang Designed and conducted experiments with the exception of Figure 4.1 A-C (2%). Wrote the paper (0%)

Stephen A. Supplied reagent used in Figures 4.5 & 4.6 Wood, George D. Supplied reagent used in Figures 4.5 & 4.6 Mellick, Peter A. Silburn Supplied reagent used in Figures 4.5 & 4.6 Brett. M.Collins Protein purification, bacterial expression construct design, ITC and CD Spectra used in Figures 4.1 A-C (40%) Wrote and edited paper (5%)

Andrea Designed and conducted experiments with the Bugarcic exception of Figure 4.1 A-C (10%).

Wrote and edited paper (15%) Rohan D. Designed experiments (18%)

Teasdale Wrote and edited paper (10%)

viii

Publication citation – incorporated as Chapter 5.

Parkinson’s disease linked mutation in Vps35 impairs endosome association of retromer and impairs degradation of α-synuclein

Authored by: Jordan Follett, Andrea Bugarcic, Zhe Yang, Suzanne J. Norwood, Brett M. Collins & Rohan D. Teasdale

Manuscript is currently under review at Human Molecular Genetics.

Contributor Statement of contribution Jordan Follett Generated and validated DNA constructs used in study, designed (Candidate) and conducted experiments with the exception of Figure 5.1A, 5.6E and 5.7 E-F (85%) Wrote the paper (70%) Andrea Bugarcic Designed and conducted experiments used in Figure 5.6E (100%) Wrote and edited paper (15%) Yang Zhe Designed and conducted experiments used in Figure 5.7E-F (100%) Wrote the paper (0%) Suzanne J. Protein purification, bacterial expression construct designed and ITC Norwood used in Figures 5.1A (50%) Wrote the paper (0%) Brett. M.Collins Protein purification, bacterial expression construct designed and ITC used in Figures 5.1A (50%) Wrote and edited paper (5%) Rohan D. Designed experiments (15%) Teasdale Wrote and edited paper (10%)

Contributions by others to the thesis Rohan Teasdale and Andrea Bugarcic formed the PhD supervisory team that was involved in experimental design and research direction guidance, as well as assisting with the editing and review of this thesis.

Statement of parts of the thesis submitted to qualify for the award of another degree None

ix

Acknowledgements I wish to thank my supervisor, Dr. Rohan Teasdale and co-supervisor, Dr. Andrea Bugarcic. Thank you for your supervision, support, enthusiasm, encouragement, freedom and understanding throughout my PhD project. Your help has been a constant source of support and has made my PhD project a success and an overall pleasure to work on.

To Drs Markus Kerr, Zhe Yang and Genevieve Kinna, thank you for your constant encouragement, ideas, support and interest in my project over the last 4 years. Your help, both intellectually and physically, has provided a great network of knowledge to help enhance my project and my skills as a scientist.

I would like to thank my PhD committee members Professor Robert G. Parton, Professor Elizabeth Coulsin and Dr. Brett M. Collins for investing time in my PhD project, providing constant feedback on my work and showing overall general interest in the biology detailed in this thesis.

To the remainder of the Teasdale Laboratory members, both former and current, thank you for creating an enjoyable and positive environment to work in and for your constant support both in and outside the lab.

Lastly, I wish to thank my loving family and amazing partner, Marlee. They have provided me with undivided support, encouragement, reassurance and have worked tirelessly to provide for me. I will forever be in their debt for giving me the opportunity to follow my passion for science and for allowing me the time to complete my PhD.

x

Keywords:

Parkinson’s disease, α-synuclein, Retromer, Vacuolar protein sorting, endosome, Cathepsin D

Australian and New Zealand Standard Research Classifications (ANZSRC):

060108 Protein Trafficking, 90%

060199 Biochemistry and Cell Biology not elsewhere classified, 10%

Fields of Research (FoR) Classification

0601 Biochemistry and Cell Biology, 70%

1109 Neurosciences, 30%

xi

TABLE OF CONTENTS: Thesis Abstract …………...... i Declaration by author ……………………………….………………………...... ii Publications during candidature ...……...... iV Presentations and conference attendance...... V Publications included in this thesis:……...... Vi Contributions by others to the thesis…………………………………………………...... Viii Statement of parts of the thesis submitted to qualify for the award of another degree……………………………………………………………………………………………..Viii Acknowledgements………………………………………………………………………………..iX Keywords…………………………………………………………………………………………...X Australian and New Zealand Standard Research Classifications (ANZSRC)………….……X Fields of Research (FoR) Classification………………………………………………………....X Table of contents………...... Xi List of figures…………...... XV List of tables……...... XVi List of abbreviations...... XVii

Chapter 1: Introduction 1.0 Introduction and literature review………………...... 1 1.1 The Endosomal Network………...... 1 1.1.2 Endosomal Maturation………...... 1 1.1.3 Endo-Lysosomal fusion………………………...... 4 1.1.4 Lysosomes and Autophagy……….………………………………………………………..5 1.1.5 Motility of Endosomes……….…...... 7 1.2 The Retromer Complex……...……...... 8 1.2.1 Retromer Mediated Trafficking from Endosome to TGN…………………...... 9 1.2.2 Retromer Mediated Recycling from Endosome to Cell surface……….…..……...... 10 1.2.3 Recruitment of Retromer to the Endosomal Membrane………………………...... 12 1.2.4 Cargo Recognition by the Retromer Complex……………………………...... 14 1.2.5 Scission of Retromer-positive Tubulovesicular transport carriers …………………....18 1.2.6 Docking of Retromer Positive Tubules at the TGN……………………...……..………19 1.2.7 Dissociation of Retromer from the Endosomal Membrane …………………..……….20 1.3 Retromer Cargo Trafficking in Neurons………………………………………....…………21 1.3.1 Trafficking and Neurodegeneration……………………………………………….……...22 xii

1.3.2 Retromer in Alzheimer’s disease……………………………………………………..…..23 1.3.3 Parkinson’s disease...... ……....25 1.3.3.1 Structure and localization of α-synuclein…………………………………………...... 26 1.3.3.2 α-synuclein pathogenesis………………………………………………………….…....28 1.3.3.3 Proteasomal Degradation of α-synuclein……………………………………………...29 1.3.3.4 Degradation of α-synuclein within the Lysosomal System…………………...……..30 1.3.3.5 Cathepsin D….…………………………………………………………………………...32 1.4 Retromer and Parkinson’s disease…………..…………………………………………….33 1.4.1 Parkinson’s disease mutations in Relation to Retromer’s Structure…………..……...35 1.4.2 α-synuclein degradation, Cathepsin D and the Retromer complex…………...... 36 1.5 Aims and Hypotheses………………………………………………………………………..38

Chapter 2: Materials and Methodology 2.1 Materials…..……………………………………………………………….………………….40 2.2 DNA Constructs……………………………………………………………….……………...40 2.2.1 Protein Purification and Isothermal Titration Calorimetry and Circular Dichroism Spectroscopy………………………………………………………………………….…………..40 2.2.2 Lentivirus Production and Generation of stable knockdown cell lines……………….41 2.2.3 Antibodies….………………………………………………………………………….…….41 2.2.4 Cell Culture and Transfection………………………………………………………….….42 2.2.5 α-synuclein aggregation………..…...…………………………………………………….42 2.2.6 KCl Induced Membrane Depolarization……..….……….…………………………...... 43 2.2.7 Quantification of α-synuclein Aggregation……………….………………………….…..43 2.2.8 Immunoprecipitations……………………………………………………………………...43 2.2.9 Microsomal Fractionation…………...………………………………………………….…43 2.2.10 Western blotting……….………………………………………………………………….44 2.2.11 Antibody Uptake assay……...…………………………………………………………..44 2.2.12 Cell surface Biotinylation………….……………………………………………….…….44 2.2.13 Glucose uptake Assay……..……………………………………………………….……44 2.2.14 Patient data…..……………………………………………………………………...……45 2.2.15 Establishment of Skin biopsy cultures……………...…………………………………..45 2.2.16 Microscopy of Live cells………...………………………………………………….…….45 2.2.17 Immunofluorescence.………….………………………………………….…………..….46 xiii

2.2.18 Nocadazole Treatment……….…………………………………………………………..46 2.2.19 Secretion Assay…….………………………………………………………………...…..46 2.2.20 Quantification of the Intracellular Distribution of Endosomes…………………….….46 2.2.21 Colocalization analysis………...………………………………………………………...47

Chapter 3: Disruption of the Mammalian Retromer complex induces α-synuclein Inclusion formation 3.0 Introduction…...... 48 3.1 Results….……………………………………………………………………………………..50 3.1.1 Loss of Retromer Induces α-synuclein aggregation………………….…….……….….50 3.1.2 α-synuclein accumulates in Rab7A positive comparments…………………...………52 3.1.3 Formation of α-synuclein aggregates impair Cathepsin D processing…….…………54 3.1.4 Presence of α-synuclein aggregates do not influence levels of Retromer...... 56 3.2 Discussion….………………………………………………………..…………………….….58 3.2.1 Retromer Associates with α-synuclein Inclusions……………………………………....58 3.2.2 Loss of Cathepsin D impairs Degradation of α-synuclein inlcusions…………………59 3.3.3 Retromer, TBC1D5 and Clearance of α-synuclein……………………………………..61

Chapter 4: The Vps35 D620N Mutation linked to Parkinson’s disease Disrupts the Cargo Sorting Function of Retromer 4.0 Introduction………..…….………………………………………………...... 64 4.1 Results………………………………………………………………………………………...66 4.1.1 Vps35 D620N binds Vps29 and Vps26A with same affinity as Vps35 WT ……...….66 4.1.2 Expression of Vps35 D620N mutant causes redistribution of endosomes ………....66 4.1.3 Identification of the redistributed endosome population ……………………….…..….70 4.1.4 Vps35 D620N mutant causes a defect in cathepsin D trafficking ……………………71 4.1.5 Vps35 D620N is associated with redistributed endosomes in PD patient fibroblasts ……………………………………………..……………………………………………………….73 4.2 Discussion….………………………………………………………………………………....78

Chapter 5: Parkinson’s disease linked Vps35 R524W mutation impairs retromer’s endosomal association and induces α-synuclein inclusion formation 5.0 Introduction…………………………………………………………………………………....82 5.1 Results….……………………………………………………………………………………..82 5.1.1 Vps35 P316S and Vps35 R524W are incorporated into retromer complexes ……...82 xiv

5.1.2 Vps35 R524W-containing retromer has impaired endosome recruitment..….……...85 5.1.3 The association of Vps35 R524W with regulators of the retromer complex is impaired ………………………………………………………………………………………………….…..87 5.1.4 Retromer deficiency induces α-synuclein aggregation ………………...……..……....90 5.1.5 Vps35 R524W expression induces α-synuclein aggregation………………...……….92 5.1.6 Retrograde sorting is delayed in the presence of Vps35 R524W ……..……….…….94 5.1.7 SNX27-retromer dependent recycling of GLUT1 is unaffected in the presence of Vps35 R524W ……………………………………..……………………………………….…….97 5.2 Discussion….………………………………………………………………..……………....100

Chapter 6: 6.0 General Discussion………………………………………………………………………....104 6.1 Retromer as a regulator of disease…………...…………………………………….…….104 6.1.1 Regulator of mechanisms involved in PD manifestation……………………………..105 6.2 Lysosomes: a readout for disease …………………………….…………………………107 6.3 Retromer interacting proteins and disease ………………………………..……….……110 6.4 Retromer mediated Neuroprotection.…………….….………….…………………..……111 6.5 Pharmacological modulators of Retromer ……….…………………..………………….113 6.6 Future perspectives ………………………..…………………………………….………...114

Chapter 7: References…….………………………………...………………………………………………117

Appendix……………………………………………………………..……………….…………133

xv

List of Figures: Figure 1.1: Multi-directional sorting pathways within the mammalian endocytic network...... 2 Figure 1.2: Current known mammalian retromer complexes…………………………………9 Figure 1.3: Sorting pathways mediated by the mammalian retromer complex…………....12 Figure 1.4: Working models of retromer recruitment to the endosomal membrane…...…13 Figure 1.5: Working model for retromer dissociation from the endosome…………...... …21 Figure 1.6 Domain architecture of monomeric α-synuclein…………………………….……27 Figure 1.7: Step-wise processing and sub-cellular localization of the Cathepsin D……...33 Figure 3.1: Loss of retromer induces α-synuclein inclusions…………………………………51 Figure 3.2: α-synuclein inclusions reside in the Rab7A positive compartments in cells depleted of retromer……………………………………………………………………………...53 Figure 3.3: Disruption in cathepsin D processing leads to accumulation of α-synuclein inclusions……………………………………………………………………………………….….56 Figure 3.4: Western blot analysis of endosomal machinery during aggregation time course……………………………………………………………………………………………...57 Figure 4.1. Vps35 mutant D620N binds to Vps29 and Vps26A in vitro and in vivo…………67 Figure 4.2: Ectopic expression of Vps35 D620N alters endosome morphology and distribution………………………………………………………………………………………....69 Figure 4.3: Ectopically expressed Vps35 D620N positive endosomes contain both early and late endosome markers…………………………………………………………………………..70 Figure 4.4: Ectopically expressed Vps35 D620N interacts with CI-M6PR but alters the receptors’ capacity to transport cathepsin D……………………………………………………72 Figure 4.5: Parkinson’s disease Vps35 D620N patient fibroblasts have redistributed endosomes………………………………………………………………………………………...74 Figure 4.6: Cathepsin D processing is impaired in PD Vps35 D620N patient fibroblasts………………………………………………………………………………………….76 Figure 5.1: Parkinson’s disease linked Vps35 mutants P316S and R524W do not disrupt trimer formation……………………………………………………………………………….…..84 Figure 5.2:Vps35 R524W disrupts recruitment of retromer to the endosomal membrane…………………………………………………………………………………………86

Figure 5.3:Vps35 R524W has diminished association with regulators of retromer…….…88

Figure 5.4: Loss of retromer induces α-synuclein aggregates that can be rescued using R55, a pharmacological chaperone…………………………………………………………….91 xvi

Figure 5.5: Parkinson’s disease linked Vps35 point mutation increase the production of α- synuclein inclusions………………………………………………………………………………93

Figure 5.6: Expression of Vps35 R524W perturbs localization and sorting of CI- M6PR………………………………………………………………………………………………96 Figure 5.7: Vps35 R524W does not disrupt SNX27-dependent cargo recycling……...... 98

Supplementary Figure 1: Quantification of peri-nuclear and cytoplasmic protein……....133

Supplementary figure 2: PD-linked Vps35 mutants P316S and R524W do not impact Golgi morphology……………………………………………………………………………………....135 Supplementary figure 3: PD-linked Vps35 mutants P316S and R524W do not disrupt the localization of SNX27………………………………………………………………………..….136

List of Tables: Table 1: Transmembrane cargo shown to interact with the retromer complex………...……17 Table 2: Retromer mutations linked to Parkinson’s disease………………………………….34

xvii

List of Abbreviations:

Aβ Amyloid-β-peptide

AMPAR α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor

Arp 2/3 Actin related protein 2/3

APP Amyloid Precursor Protein

ATG Autophagy-related genes

BAR Bin/Amphiphysin/Rvs

CCV Clathrin-coated Vesicle

CMA Chaperone Mediated Autophagy

DA Dopamine

DAPI 4',6-diamidino-2-phenylindole

DMEM Dulbecco’s Modified Eagle Media

DNA Deoxyribonucleic Acid

EE Early endosome

EEA1 Early Endosomal Antigen 1

EHD1 Eps15 homology domain-containing protein-1

FCS Foetal Calf Serum

FERM 4.1/ezrin/radixin/moesin

FYVE Fab1, YOTB, Vac1, and EEA1

GAP GTPase Activating Protein

GDP Guanosine Diphosphate

GEF Guanine Nucleotide Exchange Factor

GFP Green Fluorescent Protein

GLUT1 Glucose Transporter-1

GTP Guanosine Triphosphate

xviii

GTPase Guanosine Triphosphatase

HOPS Homotypic vacuolar Protein Sorting kDa (kilo)Dalton

KD Knockdown

LAMP1 Lysosomal-associated Membrane Protein-1

LAMP2 Lysosomal associated membrane protein-2

LB Lewy Body

LE Late endosome

LRRK2 Leucine Rich Repeat Kinase-2

M6PR Mannose-6-Phosphate Receptor

MTOC Microtubule Organization Center

MVB Multi-Vesicular Body

NMDAR N-methyl-D-aspartate receptor

NSF N-ethylmaleimide-Sensitive Factor

PAGE Polyacrylamide Gel Electrophoresis

PBS Phosphate Buffered Saline

PFA Paraformaldehyde

PD Parkinson’s disease

PDZ PSD95, Dlg1, zo-1

PI Phosphatidylinositol

PI(3)K Phosphatidylinositol 3 kinase

PIKFYVE phosphoinositide kinase, FYVE finger containing

PX Phox Homology

Rab Ras related protein in the Brain

RNAi RNA interference

RPMI Roswell Park Memorial Institute medium

xix

SDS Sodium Dodecyl Sulphate shRNA short hairpin RNA siRNA small/short Interfering RNA

SNARES N-ethylmaleimide-sensitive factor attachment protein receptor

SNX Sorting Nexin

TBC1D5 Tre2, Bub2, and Cdc16-Domain containing protein 5

TfnR Transferrin Receptor

TIP47 Tail interaction protein-47

TGN trans-Golgi Network

TH Tyrosine Hydroxylase

Ub Ubiquitin

UPS Ubiquitin-Proteasome system

UVRAG UV radiation resistance-associated gene protein

VARP VPS9-ankyrin-repeat protein

VTI1B Vps10 tail interactor-1B

VPS Vacuolar Protein Sorting

WASH Wiskott–Aldrich Syndrome protein Homology-1

1

Chapter 1: Literature review

1.1 The endosomal network Despite being one of several thousand different pathways characterized in the mammalian cell, the endosomal network is one of tight regulation and organization. Typically, an extracellular stimulus, in the form of a ligand, is bound by its receptor at the plasma membrane and triggers the processes known as endocytosis. Endocytosis is the broad term given to the internalization of extracellular fluid, solutes, and plasma membrane components by specialized invaginations, which are delivered to internal compartments termed endosomes.

The compartments of the endosomal system which include: the recycling endosome; the early endosome; the MVB/late endosome; the lysosome and the autophagosome provide the basis for a dynamic network of membrane exchange. Despite classically being described as single entities within a much large network of membrane trafficking, collectively endosomes act in concert to sort and segregate cargo molecules endocytosed at the plasma membrane into one of 3 distinct pathways (see figure 1.1). Cargo may be trafficked from the early endosome through the endosomal network to the lysosome for degradation (EGFR), from the early endosome to the plasma membrane via the recycling endosome (APP, TfnR, GLUT1), or concentrated into the retrograde pathway from the early endosome to the TGN (CI-M6PR). Although the endosome can be viewed as a single pre-formed, stable organelle that shuttles cargo molecules, a multitude of step-wise modifications occur to ensure correct progression of cargo sorting and maturation occurs.

1.1.2 Endosomal maturation Following endocytosis, cargo enters the early or sorting endosome (EE) (see Figure 1.1). Irrespective of cargo selectivity, the newly formed endosomal compartment will undergo several rounds of modification to aid in the formation of a mature organelle destined for fusion with the lysosome. In order to do so, several modifications of the endosome periphery must first take place in a temporal manner, one of which is the generation of vital lipid components derived from phosphoinositides (PIs). As the endosome starts to mature, a type III PI3K (hVps34/p150) complex selectively generates phosphatidylinositol-3- phosphate (PI3P) from phosphatidylinositol [1, 2]. Generation of PI(3)P and its derivative,

PI(3,5)P2, act as molecular signalling platforms and initiate recruitment of several membrane 2 targeting coat proteins such as Sorting nexins (SNX) and FYVE (Fab1, YOTB, Vac1, and EEA1)-domain containing proteins.

Figure 1.1: Multi-directional sorting and endosomal maturation within the mammalian endocytic network

3

Initial recruitment of PI3K/Vps34 to the endosome is mediated by direct interaction between the p150 subunit of the PI3K complex and the small GTPase Rab5, the latter of which is regulated by the GDP/GTP exchange factor (GEF), Rabex-5 [3-5]. The exchange of GDP for GTP on Rab5 acts as a positive feedback loop and mediates further recruitment of Rab5- GTP to the cytosolic face of the endosome. In turn, this leads to rapid accumulation of endosomal PI3K, PI3P synthesis and PI3P targeting proteins, allowing for recruitment of additional sorting machinery to the maturing endosome [6]. Furthermore, membrane associated Rab5 mediates homotypic fusion events between early endosomes, a vital step in ensuring condensation of endosomal numbers and preparation of the endosome for cargo degradation. This observation of homotypic fusion is exemplified by the extensive vacuolization witnessed in the presence of GTP-locked Rab5, which leads to hybrid organelles rich in both early and late endosomal markers [7-9]. As little hydrolytic activity is present within the lumen of the compartment at this stage of the maturation, the organelle is primarily used as a central platform rich in nascent tubules decorated with cargo proteins [10].

For the now established endosomal compartment to begin the next stage of maturation, the Rab5 positive feedback loop must be down-regulated, replaced by Rab7 and conversion of PI(3)P to its derivate, PI (3,5)P2, must occur. For this to take place, endosomal co-factors, SAND-1/Monz1 and Ccz1, are recruited from the cytosol to engage GTP bound Rab5, promote recruitment of Rab7 and the HOPS (homotypic fusion and protein sorting) complex and displace Rabex-5 from the cytosolic face of the compartment, thus terminating the Rab5 positive feedback loop [11]. The functional role for the HOPS complex in mammalian endosomal biology relies on data derived from yeast demonstrating that the Vps39p subunit of the HOPS complex has GEF (Guanine/GDP/GTP exchange factor) activity towards Rab7-GDP [12], however this does not appear to be conserved in high order organisms [13]. One plausible explanation might be, that instead of the HOPS complex, the SAND-1/Monz1/Ccz1 complex possess GEF activity toward Rab7 [14] and aids in its recruitment to the membrane or, additionally displaces Rab7-GDP of its GDI (GDP dissociation inhibitor), a factor that locks Rab proteins in their inactive state [10]. However, regulation of the HOPS complex may in fact be via the Vps34 kinase complex, namely the UVRAG subunit, which is negatively regulated by the Rab7 effector, Rubicon [15]. Activation of Rab7-GDP leads to increased interaction with Rubicon, freeing UVRAG and subsequently increased positive regulation of the HOPS complex, thus increasing fusion between late 4 endocytic compartments and possibly finalizing the Rab5/7 switch [15]. Despite being a late endocytic compartment, large amounts of evidence support the recruitment of membrane coat proteins/cargo sorting machinery (SNX proteins, retromer complex, Rab GTPases). For example, SNX1 has been shown to directly bind both PI(3)P and PI(3,5)P2 [16, 17], which are found on early and late endosomes, respectively, whereas association of retromer with the late endosome is mediate by the aforementioned GTPase, Rab7A [18]. Thus, suggesting that multiple rounds of cargo selectivity occur and likely depend on a temporal platform where cargo destined to be sorted back to the cell surface is done so immediately following endocytosis and not retained in the maturing endosome.

In conjunction with recruitment of endosomal machinery needed to facilitate cargo selection and fusion events, maturation is accompanied by generation of PI(3,5)P2 from PI(3)P. This conversion is achieved by the phosphatidylinositol 3-phosphate 5-kinase, PIKfyve [19], and is initiated by the PI3P targeting FYVE domain of PIKfyve itself [20]. Thus, inferring that the synthesis of PI (3,5)P2 is limited to the amount of existing membranes rich in PI(3)P. Interestingly, PIKfyve forms a complex with ArPIKfyve/Vac14p and the PI(3,5)P2 phosphatase, Sac3, allowing for internal positive and negative regulation of function, respectively [21]. Furthermore, PIKfyve is reported to play a regulator role in cargo sorting, as exemplified by kinase activity against p40, a Rab9 effector, needed to cycle the Cation- Independent Mannose-6-Phosphate receptor (CI-M6PR) back to the TGN [22]. Additionally, silencing of PIKfyve grossly perturbs the retrograde sorting of the CI-M6PR from the EE, consistent with its role in PI(3)P binding [23] and endosomal maturation. However, this misorting is likely a product of impaired recruitment of endosomal machinery following the generation of PI(3,5)P2 rather than a direct consequence of PIKfyve suppression.

1.1.3 Endo-lysosomal Fusion The formation of a mature late endosome/MVB is accompanied by the fusion with lysosomal compartments. Until recently, direct fusion between LE and Lysosomes was controversial due to the strong juxtanuclear localization of these compartments nearing the Microtubule Organization Center (MTOC). However, recently using live cell microscopy and fluorescently labelled dextran, several things were noted. Content mixing between lysosomes and endosomes was witnessed only when, (1) organelles were in physical contact with each other, (2) organelles undergo transient or permanent fusion or (3) nascent tubules connect to neighboring compartments [24]. Together, these observations 5 demonstrate that material is not exchanged in a bidirectional manner and secondly, fusion with lysosomal compartments is a point of no return, an argument exemplified by the fact that lysosomes are typically devoid of cargos [25, 26]. However, it must be noted that the transient period in which a hybrid organelle forms following fusion between the LE and lysosome, allows for a final recover of specific receptors (CI-M6PR) and cellular machinery (SNAREs) prior to degradation, however this mechanism is not well characterized [26].

Consistent with other parts of the endocytic network, fusion between the late endosome and lysosome organelle requires tethering molecules, however the composition of the tethers is not entirely established. It is likely that the aforementioned HOPS complex and small GTPase Rab7A play a role in endo-lysosomal fusion as over-expression of HOPS complex subunit, Vps18, induces endo-lysosomal clustering at the perinuclear space [27, 28]. Following successful tethering and subsequent fusion between compartments, SNARE complex formation is initiated by subunits Syntaxin-7, VTI1B (Vps10 tail interactor-1B) and Syntaxin-8. Furthermore, it is worth noting that heterotypic endo-lysosomal fusion requires the addition of the R-SNARE, VAMP7, whereas late endosomal homotypic fusion requires VAMP8 [29-31]. Following immediate fusion, it is logical to assume that the product is a hybrid organelle that possess both MPRs and lysosomal enzymes and may represent the final opportunity to retrieve cargo prior to degradation [26]. However, the endocytic machinery that facilitates cargo sorting at this stage in the pathway is still unknown. Interestingly, the small GTPase Rab7A demonstrates strong localization with lysosomal compartments, whereas retromer is absent from this subset of hybrid compartments [32]. One possibility is the GTPase Rab9 in conjunction with TIP47 and p40 which rescues the CI-M6PR to the TGN independent of the retromer complex [22] and thus, requires Rab7 for priming prior to Rab9 association.

1.1.4 Lysosomes and Autophagy Successful endosomal maturation and fusion with lysosomal compartments is followed by heterotypic fusion with autophagosomes to facilitate degradation of endocytosed product. Autophagy is commonly considered a non-selective process that allows for overall housekeeping and cellular homeostasis, however autophagy is implicated in the pathogenesis underlying neurological diseases such as Parkinson’s and Huntington’s disease, suggesting specific roles in clearance of pathogenic proteins. Generation of the double membrane autophagosome compartment is a multi-step process that, similar to the 6 endocytic system, requires a plethora of effector machinery (ULK1/2, FIP200, ATG13, ATG14L, Vps34, Beclin1 and p150 to name a few) and donor membrane derived from early, recycling, MVB and Golgi organelles which, when complete, fuses with a lysosomal compartment to form a autolysosome [33]. To date, over 37 core autophagy related proteins have been implicated in the generation of the double membrane organelle in all eukaryotes [34]. However, for autophagy to occur, the generation of an isolation membrane must take place [33], however, how this membrane is positioned correctly for autophagy initiation is not understood. Following this initiation stage, the ULK1 (UNC51-like kinase) complex composed of ATG13, FIP200, ATG101 and the autophagy specific Beclin-1 containing PI3K complex is recruited to initiate extension/nucleation of the existing membrane [33, 35-39]. Following recruitment and activation of the ULK1 complex, nucleation specific ATG proteins are recruited to the newly generated pools of PI(3)P to facilitate continual extension of the newly formed membrane. Subsequent to this, expansion of the membrane is achieved via the ATG16L1 complex (ATG5-ATG12 and ATG16L1) to allow engagement of the cargo adaptor molecule, LC3, to the expanding membrane [40-43]. Ligation of lipidated LC3 to the newly formed membrane initiates engagement of the organelle/aggregate destined for degradation, and subsequent closure of the autophagosome compartment [44].

Engulfment and closure of the autophagosome compartment is accompanied by fusion with the acidified lysosomal compartment, allowing for membrane turnover and degradation of damaged cellular production. How newly formed autophagosomes undergo fusion with lysosomal compartments is ill defined. The HOPS complex along with Rab7, Syntaxin-7 and VTI1B, form the machinery implicated in late endosomal-lysosomal fusion, and have also all been associated with autophagosome-lysosomal fusion [45-47]. Although this does not directly implicate these tethering molecules in the molecular process, it is likely that they play a role in general membrane tethering complexes. Furthermore, depletion of lysosomal associated membrane protein-1 and 2 (LAMP1/2) has also been shown to reduce fusion with the autophagosome [48], inferring that lysosomal function, maturity and luminal content may also regulate fusion.

Despite the complex nature of autophagy within the mammalian system, significant amounts of research have established how autophagy may be altered in neurodegenerative diseases. One such example that is discussed in detail later is the notion that molecular chaperones transport misfolded/damaged α-synuclein from the cytosol to the lysosome 7 membrane in order to be degraded, a process termed Chaperone-Mediated autophagy [49]. Interesting, the familial mutation in α-synuclein (A53T) is transported to the lysosome but fails to translocate into the lumen of the lysosome and as such, perturbs degradation of the protein via autophagy [49]. Moreover, as initiation of autophagy may derive its membrane from the endoplasmic reticulum, this secretory pathway is an obvious route for implication in autophagy. Interestingly, the over-expression of wild-type α-synuclein perturbs trafficking of this route, blocks secretion and leads to fragmentation of the Golgi [50]. Moreover, the small GTPase Rab1a that governs ER to Golgi trafficking can rescue the α-synuclein induced ER to Golgi block in trafficking. As this particular pathway falls prior to the formation of the autophagosome it is likely that several other ATG proteins are implicated by perturbing this trafficking pathway and thus, compromise formation of the autophagosome.

Autophagy is implicated in other neurological diseases such as Alzheimer’s and Huntington’s. Interestingly, there is a strong mechanistic relationship between the total amount of mutant Huntington (mHTT) and the activation of the autophagy pathway. For example, the presence of mHTT contributes to an increase in basal autophagosome formation by sequestering and inactivating mTOR [51-53], the main kinase complex responsible for autophagy induction [54]. However, despite the reported increase in the number of autophagosomes present, they are in fact devoid of canonical machinery needed for cargo recognition and degradation [53]. Although the presence of mHTT appears to alter direct regulators of autophagy, several other aspects of the endo-lysosomal system appear compromised. For example, the expression of mHTT perturbs the dynamics of the autophagosome and consequently, inhibits fusion with the lysosome, thus blocking degradation [55]. Second to this, mHTT has decreased interactions with the autophagy receptor, Optineurin, and also results in altered phagosome kinetics [56]. Overall, it is clear that autophagy plays a fundamental role in maintaining cellular homeostasis and ensuring potentially pathogenic proteins are degraded to prevent disease.

1.1.5 Motility of Endosomes Although endocytosis largely depends on the continuous generation of lipid membranes, the kinetics of endocytic compartments relies on the microtubule network and the machinery linking it to organelles. Large numbers of small, highly mobile, peripheral endosomes are, over time, replaced by fewer large, often dense compartments as they near the MTOC. EE compartments are subjected to slow, short range oscillating movements that, 8 in conjunction with overall conversion into a late endosome, undergo rapid long-range movements with a net trajectory toward the perinuclear region of the cell. This process is mediated by the by the microtubule motor proteins, Kinesin and Dynein, that facilitate plus end and minus end movement, respectively [57]. For example, movement of EEs in a plus- end dependent manner can be mediated the kinesin motor protein, KIF16B, which via its PX domain, binds PI(3)P and facilitates endosomal movement. Consequently, KIF16B strongly influences receptor degradation kinetics as siRNA mediated suppression impairs recycling of cell surface cargo proteins and subsequently increases the rate of receptor turnover [58].

In contrast to this, movement of the LE is dependent on the dynein dependent minus- end transport. Interestingly, the interaction between dynein and maturing/late endosomes is reportedly reliant on the aforementioned small GTPase Rab7a which once GTP bound, allows for dynein to engage the Dynactin subunit, p150 (glued) and facilitates minus-end transport [59, 60]. Moreover, appropriate spatio-temporal organization of endosomes and their contents is not simply by chance and as such, sorting of receptors away from endosome compartments requires inter-connection between endosomes, motor proteins and cargo sorting machinery. As an example, a member of the PX-containing Sorting Nexin family, SNX6, which has been implicated in sorting of cargo to the TGN, binds both SNX1 and the Dynein activating protein, p150 (glued) [61], thus inferring that the movement of endosomes along the microtubule network is intrinsically linked to recruitment of sorting machinery such as SNX proteins Rab GTPases and the retromer complex.

1.2 The Retromer Complex The sorting of transmembrane receptors (cargo) requires a tightly regulated process within the endosomal system to enable a spatio-temporal organization within the cell. Retromer is an essential cytoplasmic, high affinity protein complex that facilitates the endosomal trafficking of many different receptor cargos. Initially characterized in S. cerevisiae, retromer is a heteropentameric protein complex composed of the core Vacuolar protein sorting (Vps) gene products - Vps26, Vps29 and Vps35 - in association with a membrane targeting dimer of Sorting Nexin (SNX) orthologues, Vps5 and Vps17 [62, 63]. The retromer complex is highly conserved in all eukaryotes [63, 64], and mammalian retromer likewise consists of a high affinity Vps29-Vps35-Vps26 cargo recognition trimer [65], which can interact with a range of proteins including multiple members of the SNX protein family (reviewed in [66]). Despite maintaining a high level of evolutionary 9 conservation, several functional differences exist between the yeast retromer and its mammalian equivalent; possibly in adaptation to a more complex protein sorting pathways found in multicellular organisms. For example, in higher eukaryotes retromer exists as two distinct types defined by different Vps26 paralogs, Vps26A or Vps26B [67-69].

Figure 1.2: Current known mammalian retromer complexes.

Vps35 (green) forms a stable complex with Vps26 (blue) and Vps29 (red) and associates with SNX27 (top) to mediated endosome-to-plasma membrane recycling of receptors or a SNX-BAR-retromer (middle) or SNX3- retromer complex (bottom) to facilitate endosome-to-Golgi trafficking.

Together, retromer and SNX complexes associate with the cytosolic face of multiple endosomal compartments to facilitate the retrieval of transmembrane cargo molecules to various membrane compartments, including the TGN [70, 71] - a hub for sorting of newly synthesized proteins - and the plasma membrane [72]. While retromer was originally defined as a regulatory complex needed to sort acid hydrolases to the endo-lysosomal network via the cation independent mannose 6-phosphate (CI-M6PR) receptor pathway [18, 70], recent studies have identified a variety of cargo molecules such as mammalian iron transporter (DMT1-II) [73], cell polarity regulator, Crb [74], and amyloid precursor protein (APP) adaptor, SorLA [75, 76], that require retromer for correct sorting.

1.2.1 Retromer-mediated trafficking from endosomes to the TGN Successful delivery of retromer cargo from the endosome to the TGN is still not fully characterized at the molecular level. Efficient retrieval of cargo from the endosome requires an endosomal sub-domain to be generated via the coordinated recruitment of a range of proteins which form a molecular assembly complex on both the donor and acceptor 10 membranes, some of which may possess the ability to aid in membrane remodelling. The most important molecular complex which not only interacts directly with cargo but is capable of organizing these molecular assemblies is the retromer complex. Following recruitment to the endosomal sub-domains, retromer and associated cargo are concentrated in nascent membrane tubules generated by dimers of SNX proteins containing a bin/amphiphysin/rvs (BAR) domain, likely involving heterodimers composed of SNX1/SNX2 or SNX5/SNX6 [61, 77]. Cargo is then concentrated at the end of the tubules capable of forming retromer- positive vesicular compartments by the process of tubule scission, which are transported toward the TGN via the microtubule network.

Studies using Caenorhabditis elegans (C. elegans) described the requirement for retromer in maintaining long-range signaling of the Wnt ortholog EGL-20 [78]. Moreover, recent data demonstrates that retromer is required for retrieval of Wntless/MIG-14, a receptor essential for secretion of the Wnt proteins, and that loss of retromer appears to result in degradation of the receptor and failed signal transduction [74, 79-81]. The interplay between retromer and Wntless/MIG-14 appears to be a SNX3-dependent retrieval process, remaining distinct from the previously described SNX-BAR/retromer complex [82, 83]. Further, using C. elegans, SNX3-retromer was shown to mediate retrieval of a part of the yeast iron transporter machinery, Ftr1, which cycles to and from the Golgi apparatus under periods to stress [84].

Besides maintaining developmental gradients, retromer has also been demonstrated to play a role in aiding the clearance of apoptotic cells. Following internalization, the CED-1 receptor dissociates from its ligand, an apoptotic cell, and is stored in the Golgi network until it is transported to the cell surface to undergo further rounds of cell clearance [85]. Though this process has been described in C. elegans it is possible that similar clearance processes are achieved with the help of retromer in higher order eukaryotes during development of the embryo. In line with that, retromer has been reported to play a role in the establishment of epithelial cell polarity through the sorting of Crumbs (Crb), further suggesting a role for retromer in development [86, 87].

1.2.2 Retromer-mediated recycling from endosomes to the cell surface While retromer’s predominant role within endosomes is the sorting of membrane cargo to the TGN, recently an additional role in trafficking of transmembrane proteins from 11 the endosome to the plasma membrane has been described [72]. SNX27 has been implicated in the recycling of many different types of transmembrane cargos including solute carriers, glutamate receptors and potassium channels [72, 88-92]. Recycling of cargos back to the cell surface requires retromer to form a complex with SNX27, which acts as an adaptor capable of interacting directly with cytoplasmic domains of cargo [72, 93]. SNX27 contains an N-terminal PSD95, Dlg1, zo-1 (PDZ) domain and a C-terminal 4.1/ezrin/radixin/moesin (FERM) domain. Consequently, SNX27-retromer has been shown to mediate the recycling of cargo molecules such as the Glucose Transporter (GLUT1), which contains a C-terminal PDZ biding motif (PDZbm) via its PDZ domain, while engaging cargo containing an FxNPxY motif via its FERM domain [72, 94].

Recent findings have demonstrated how retromer is able to interact with cargo molecules and adaptor proteins simultaneously [95, 96]. The crystal structure of the SNX27 PDZ domain in complex with the retromer subunit Vps26 revealed an exposed SNX27 β hairpin that is responsible for engaging a conserved groove in Vps26. Interestingly, association of SNX27 PDZ with retromer increases the affinity for PDZbms, suggesting that cargo sorting is allosterically coupled to the formation of the SNX27-retromer assembly [95]. Despite this, siRNA mediated suppression of either SNX27 or Vps35 appears to have no direct effect on the steady state level of the other protein. Further, suppression of either Vps35 or SNX27 modulates the recycling of a partially overlapping set of cargo proteins to the plasma membrane. Therefore, arguing that engagement of cargo for entry into the recycling pathway can depend on either protein [72].

Multi-pass transporters such as GLUT1, β2 adrenergic receptor (β2AR), and AMPAR (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor) have been shown to undergo SNX27-retromer mediate recycling, however whether additional endosomal proteins, such as the WASH complex, or post translation modifications are required for the interaction with individual cargos is currently not known [72, 89]. One such molecule that may regulate differing SNX27 sub-populations is the retromer-interacting protein VARP (VPS9-ankyrin-repeat protein) [97]. VARP was recently reported to associate with a complex comprising retromer and an endo-lysosomal R-snare, VAMP7, where together they mediate delivery of GLUT1 to the cell surface [97]. However, others reported that silencing of VARP failed to disrupt recycling of GLUT1, despite confirming the interaction with the retromer complex [98]. Currently, these discrepancies surrounding the role of VARP in SNX27- retromer trafficking of GLUT1 remain unresolved. 12

Figure 1.3: Sorting pathways mediated by the mammalian retromer complex

Simplified sorting of single and multi-pass transmembrane receptors back to the cell surface via a SNX27-retromer dependent pathway (top) or, to the TGN via either a SNX-BAR- retromer or SNX3/Rab7- retromer complex (bottom).

1.2.3 Recruitment of retromer to the endosomal membrane Retromer has no known lipid binding capacity so its recruitment to the endosome occurs via a number of protein-protein interactions, generally with proteins that are membrane-associated via a phosphatidylinositol-3-phosphate (PtdIns3P)-dependent mechanism. Amongst these regulatory proteins are the membrane tubulating SNX-BAR proteins, the PX-domain only protein SNX3 [82], and the small GTPase Rab7A [18, 99-102]. 13

Engagement of SNX-BAR dimers with the endosomal membrane generates nascent tubules incorporating retromer-associated cargo, acting in part with various other molecules to generate a stable sorting platform. The precise recruitment of the SNX dimer requires binding via their conserved phox-homology (PX) domains [103], to the classical early endosome component, PtdIns3P [16], which is generated by the phosphoinositide 3-kinase (PI3K) [2]. In addition to this, several groups have demonstrated the ability of SNX1 and

SNX2 to bind PtdIns(3,5)P2 [16], a membrane phospholipid generated from PtdIins3P by PIKfyve [104], which is found on late or maturing endosomes [105, 106]. This difference in lipid biding ability gives rise to the idea of selective, spatio-temporal sorting of retromer cargos in conjunction with endosome maturation.

Figure 1.4: Working models of retromer recruitment to the endosomal membrane

Left: SNX3 and Rab7 are recruited to cytosolic retromer and mediate translocation to the endosomal compartment. Right: Endosomal association of SNX3 precedes Rab7-dependent translocation of retromer to the endosome to act as a recruitment domain.

SNX3 lacks a BAR domain and is therefore unlikely to aid in the formation of new tubules by itself. Recent publications demonstrated the ability for SNX3 to interact with the Vps35 subunit via a three-stranded β-sheet that becomes exposed following localization to 14 the membrane [82, 99]. Rab7a is itself recruited through a PI3K-dependent mechanism [102], and typically associated with late or maturing endosomes. This mechanism for recruitment of the retromer complex appears to facilitate retrieval of cargo from endosomes prior to its fusion with a lysosomal compartment; an organelle typically void of retromer. Although there is room for a temporal division of SNX3/Rab7a-mediated recruitment of retromer to the endosome, recent evidence suggests that this may not be the case. Suppression of SNX3 or Rab7A in cell culture systems results in impaired recruitment of the retromer complex to the endosome. Interestingly, silencing of SNX3 followed by over- expression of Rab7a fails to rescue the loss of retromer from the endosomal membrane, suggesting a distinct role of SNX3 and Rab7A in recruiting retromer or that additional modifications to the maturating endosomal compartment following SNX3 recruitment is required to enable Rab7a to act [107].

1.2.4 Cargo recognition by the retromer complex Newly formed endosomes will contain cargo required to be sorted into the recycling pathway or the retrograde pathway or alternatively, retained for degradation as the endosome matures into a late-endosome/lysosomal compartment. What defines the selection of differing cargos within this compartment is not fully understood. Retromer’s capacity to sort different cargos into distinct nascent tubular-vesicular carriers may lie with accessory molecules and the ability for local signaling or sorting events to recruit appropriate machinery based on the cargo type. Despite growing knowledge on how cargo is identified, engaged and positioned optimally within the endosome, the mechanism of cargo recognition remains a persistent question in retromer biology.

Initial yeast two-hybrid studies revealed that Vps35 amino acids 500-693 are involved in binding the CI-M6PR [108-110]. Interestingly, Vps26 was also described as a cargo- interacting retromer subunit, binding to the cytoplasmic tail of SorLA, a known co-receptor of the amyloid precursor protein (APP) [111]. While this finding is important as it recognizes the possibility of multiple retromer-cargo engagement points, the field still tends to favour the idea of Vps35 as a cargo-recognition subunit, while Vps26 and Vps29 subunits contribute to its function by interacting with retromer-associated proteins (Table 1).

15

Over the last decade various interaction techniques both in vitro (e.g. yeast two- hybrid and GST-pull down using cell or tissue lysates) and in vivo (immunoprecipitations of tagged or endogenous receptors) have defined a number of retromer cargos (see table 1 for specifics). The complementary cellular depletion of retromer using siRNA and shRNA methodologies used to determine subcellular localization of the newly identified cargo, where the depletion leads to a receptors inability to be sorted into the retrograde pathway and is, as such, trapped in the early endosome and eventually degraded in the lysosome [67, 70, 112]. These techniques enabled the identification of several single transmembrane receptors, including CI-M6PR, SorLA and Transforming growth factor-β type II receptor (TβRII), as well as multi-membrane span GPCRs, including PTHR and Vasopressin Receptor Type 2, as retromer cargo (see table 1 for specific references).

While using the aforementioned method of identification for retromer cargo is important, it uses a targeted approach that does not lend itself to generalization. Therefore, driven by studies that defined receptor motifs recognized by retromer-interacting sorting machinery (for example, recognition of NPxY motifs on receptors intracellular tails [113]), the retromer field has been asking the following question – can characterization of a specific sorting motif in the cytoplasmic region of a receptor aid in identification of further retromer cargo? Using some of the receptors identified by the targeted approach, several studies performed analysis of the cargo-binding motif and identified a conserved [F/W] -L- [M/V] sorting sequence [71, 73, 109, 114]. For example, CI-M6PR contains a WLM sequence within its cytoplasmic tail that has been shown to be essential for interaction between CI- M6PR and Vps35 [115]. Another version of this sorting motif (FLV) is present in sortilin, a receptor for neurotrophins and cathepsins [114]. Despite the identification of these conserved sorting motifs, mining of receptor sequences for further potential retromer cargo has been largely unsuccessful. One potential explanation for the low rates of cargo discovery may lie in the fact that the binding of retromer to its cargo molecule is in fact predominately transient in nature. Additionally, second to being a transient interaction between retromer and its cargo, it is likely that these protein-protein interactions display weak binding properties. Moreover, this may also be because retromer is capable of recognizing various sorting motifs in potential cargo molecules via the additional subunits present in the core retromer complex, and thus does not lend to a generalized sorting motif. For example, the SorLA receptor possesses an [F/W]L[M/V] sequence, however its interaction with retromer subunit Vps26 is via the FANSHY sorting motif [111]. Furthermore, 16 a C-terminal di-leucine (LL) motif was recently demonstrated to be important in the interaction between -secretase and DMT1, two completely unrelated receptors (table 1), with retromer mediated immunoprecipitation and in vitro techniques respectively [73, 99, 109].

The targeted approach described above has been successful in defining the retromer cargo sorted into the retrograde pathway. On the other hand, to define the cargo associated with the retromer-mediated recycling pathway, studies have taken a more global indirect approach, in which the levels of different receptors at the plasma membrane of mammalian cells have been analysed in the context of retromer (Vps35 or Vps26) depletion [72, 116]. The first receptor identified using this approach was the β-adrenergic receptor, when it was demonstrated that its interaction with SNX27 mediates sorting onto retromer-positive tubules destined for plasma membrane [93, 116]. In the neuronal context, this process of SNX27-retromer synergy was proposed to be important in supporting an endosomal network capable of rapid and local delivery of specific receptors to the plasma membrane of synapses [117]. These receptors include β-adrenergic receptors, but also other known SNX27 cargo such as AMPA and NDMA receptors [88-90, 92]. Following from this finding, Cullen and colleagues devised a methodology combining SILAC and MS/MS to identify a range of receptors that are lost from the cell surface of HeLa cells upon knockdown of Vps35 or SNX27 [72]. The extensive list of potential cargo of either retromer, SNX27 or the combination of both, included single transmembrane receptors, such as LRP10 and PDGFRβ, as well as multi membrane span receptors, including GLUT1 and GRP180, involved in a range or cellular and physiological processes. While the list of potential retromer cargo obtained using this approach will be important to verify using more targeted approaches (e.g. immunoprecipitations), this global analysis can also aid in identification and classification of specific sorting motifs engaged by retromer. Overall, while the retromer field has progressed enormously in identifying and classifying the retromer cargo molecules, the molecular details of these interactions still appear to be too broad and poorly defined, suggesting multiple, characterized and uncharacterized mechanisms underlying retromer- cargo interactions.

17

Cargo Context Physiological process Reference(s) mediated Vps10 receptor Vps10p Saccharomyces Sorting of lysosomal hydrolases [64] cerevisiae CI-M6PR HeLa cells Sorting of lysosomal hydrolases [115, 118] Sortilin HeLa cells Sorting of lysosomal hydrolases [114] [71] GlVps Giardia trophozoite Sorting of lysosomal hydrolases [119] SorLA HeLa cells Sorting of amyloid precursor [76, 111] protein SorCS Neuroglioma cell line Sorting of amyloid precursor [120] protein GPCR GPR107 Hek293 cells co-receptor responsible for [121] trafficking of receptors to plasma membrane Vasopressin Hek293 cells receptor involved in balance of [122] Receptor Type 2 water and sodium ions across (V2R) membranes Parathyroid Hek293 cells Hormone receptor involved in [123] hormone bone development and mineral receptor-1 ion homeostasis (PTHR) Divalent metal Saccharomyces Iron uptake across apical [73, 99] transporter 1 cerevisiae membrane of brush border of (DMT1-II) and in vitro duodenal enterocytes Glycine receptor Rat brain Inhibitory glycine receptors found [124] beta (GlyRβ) in the postsynaptic membrane

18

Single trans-membrane receptors Bone morphogenic Caenorhabditis Member of the Ser/Thr protein [125] protein type I receptor elegans kinase family and regulates cell (SMA-6) adhesion Type II TGF-β receptor MDCK cells Member of the Ser/Thr protein [126] (TβRII) kinase family and regulates cell proliferation pIgAR Rat brain Sorting of polymeric IgA [127] MDCK cells Neuroligin-2 Mouse brain Inhibitory synapse specific cell [128] surface protein that functions in cell adhesion and synaptic organization Crumbs Drosophila Cell polarity [86, melanogaster 87] Mouse brain HeLa cells γ-secretase (BACE) MDCK cells APP processing enzyme [109] CED-1 Caenorhabditis Clearance of apoptotic cells [85] elegans; in vitro Table 1: Transmembrane cargo shown to interact with the retromer complex

1.2.5 Scission of retromer-positive tubulovesiculular transport carriers Until recently, the molecular details of how retromer-positive tubules containing the specific cargo are able to detach from the endosomal compartment had not been described. In 2010, Seaman and colleagues first identified the WASH complex; which includes Wiskott- Aldrich syndrome homologue 1 (Wash1) and associated proteins Strumpellin, FAM21, KIAA1033 (SWIP), and CCDC53 [129]; as a retromer-interacting complex [130]. The WASH complex is recruited to endosomal structures through the elongated C-terminal tail of FAM21, which harbours multiple binding sites for the Vps35 subunit [131, 132]. shRNA depletion of Wash1 leads to dispersed CI-M6PR, suggestive of a fundamental role for the WASH complex in retrograde trafficking, but not on endocytosis as a whole [133]. This role 19 for the WASH complex in retrograde trafficking is likely dependent on the activation of Arp2/3, a well-defined F-actin nucleation branching factor that stabilizes nascent membrane tubules that present cargo for selection [134]. Further, the WASH complex is reported to interact with Dynamin II, a known membrane scission protein [135]. Thus, suppression of the WASH complex may perturb recruitment of molecules needed for the scission of the membrane into distinct vesicles. Though whether Dynamin II acts to specifically regulate budding retromer vesicles via the WASH complex is still undetermined.

Besides the SNX-BAR proteins and the WASH complex, other retromer-associated proteins have also been described as having a role in generation and stabilization of membrane tubules. Eps15 homology domain-containing protein-1 (EHD1) appears to stabilize SNX1 positive tubules, likely via its interaction with the retromer complex [136]. Similarly to the previously described WASH complex, suppression of EHD1 leads to the generation of extensive SNX1-positive tubules, suggestive of a failure to recruit the machinery required to cut newly forming vesicles from the tubule. While suppression of EHD1 was shown to lead to extensive perturbations in delivery of CI-M6PR to the TGN, the loss of EHD1 may also impact trafficking pathways independent of retromer [136]. For example, studies focusing on the recycling of transferrin receptor in EHD1 knockout mice show major disruption in receptor repopulation at the cell surface, suggestive of multiple roles for EHD1 in cellular trafficking independent of retromer [137]. Therefore, while it is unlikely that WASH and EHD1 interact or depend on one another, they assume similar functions within the endocytic network and both appear to aid in the scission of retromer- positive vesicles from the endosomal membrane.

1.2.6 Docking of retromer tubules to the TGN It is the sorting nexins that mediate transport of retromer-positive vesicles to the acceptor membranes via their interaction with microtubule motor proteins [138]. Evidence for this came from the observation that SNX5 and SNX6 interact with the C-terminal tail of p150glued, a subunit of the Dynactin complex needed for the minus-end microtubule dependent movements mediated by the motor protein, Dynein [61]. Additionally, suppression of p150glued by siRNA retained CI-M6PR in dispersed SNX1-positive tubules, indicative of hindered retromer sorting. Though the interaction with p150glued appears to be specific for SNX6, suppression of either SNX1 or SNX2 leads to missorting of CI-M6PR away from the TGN [103, 139], potentially as a result of ablating SNX heterodimerization. 20

Mallard and co-workers reported that retrieval of the Shiga toxin, a retromer dependent cargo, to the TGN requires interaction between the TGN localized t-SNARE complex syntaxin 6, syntaxin 16, Vti1a and endosomal localized v-SNAREs, VAMP3 and 4, which may form a complex to allow membrane docking and delivery of cargo [140]. Interestingly, it was also reported that inhibition of Golgi-localized Rab6 demonstrated a marked reduction in delivery of Shiga toxin to the TGN, implicating a potential role of this GTPase in Golgi docking of retromer-positive vesicles. This was supported by the findings that retromer- associated SNX1 interacts with Rab6 interacting protein-1 (Rab6ip1), and that siRNA depletion resulted in dispersed CI-M6PR positive endosomes [61]. Together, these studies demonstrated that this multi-step process needed for the retrieval, recognition and fusion of retromer-positive vesicles to the TGN membrane, has large dependence on retromer itself.

Despite the ever-growing understanding of how retromer regulates cargo sorting to the TGN, we currently have very little understanding if this is translatable in the context of endosome to plasma membrane transport. Given the established difference in adaptor proteins that operate with retromer [72, 97, 99, 130], it is likely that other proteins (e.g. VARP) and/or families of proteins are required that do not play a role in retromer-mediated retrograde trafficking.

1.2.7 Dissociation of retromer from the membrane Efficient and accurate recruitment of retromer to the membrane to engage and retrieve cargo is a futile process without a controlled mechanism to dissociate retromer from the compartment. One molecule implicated in the dissociation of retromer from the endosomal membrane is the RabGAP TBC1D5. RabGAPs promote the hydrolysis of Rab- bound GTP to GDP, thus negatively regulating Rab proteins. Recent findings highlighted the ability for TBC1D5 to bind the Vps29 subunit of the trimer, an interaction that appears to negatively regulate the membrane association of retromer [130, 141]. Given the role of Rab7a in retromer membrane association, it is possible that TBC1D5 influences retromer detachment through deactivation of Rab7a or another as of yet unidentified Rab protein. The over-expression of TBC1D5 does not influence levels of SNX1 on the endosome suggesting selectivity in recruitment of the trimer to the endosome and not a broad inhibition [18]. Given the complexity of receptor trafficking in higher order eukaryotes, it is likely that several other membrane-associated molecules play a pivotal role in cycling retromer on and off the endosome; potentially by acting on retromer-associated proteins and not directly on 21 the complex. For example, recently identified retromer interacting protein VARP binds to GTPase Rab32. siRNA suppression of Rab32 appeared to slightly decrease the total level of retromer subunit Vps29, suggestive of a membrane associated relationship between retromer, VARP and Rab32 [97].

Figure 1.5: Working model for retromer dissociation from the endosome

Left: Retromer engages cargo molecules (purple) and is bound by the RabGAP TBC1D5. Middle: Association of TBC1D5 with the retromer complex negatively regulates Rab7- retromer interaction. Right: Exchange of GTP for GDP on Rab7 allows for retromer to cycle off the endosomal membrane sort cargo molecules.

1.3 Retromer Cargo Trafficking in Neurons

Neurons are highly polarized cells comprised of specialized nerve terminals, which concentrate pools of transmembrane proteins; long axonal tracks rich in trafficking machinery and endosomes [142]; and a relatively small soma. Despite the spatial difference between nerve terminal, axon and soma, receptors are efficiently sorted into anterograde or retrograde compartments to mediate rapid changes in neuronal biology. Retromer is highly expressed in neurons and has recently been demonstrated to play a fundamental role in the 22 trafficking of β2 adrenergic receptor (β2AR) [116, 117]. In neuronal cultures expressing β2AR that were treated with the agonist isoproterenol for periods of up to 30 minutes, the internalized β2AR was observed in early endosomal (EE) compartments and not the TGN; while the depletion of endogenous Vps35 using shRNA reduced the presence of β2AR at the dendritic surface, further arguing the importance of retromer in neuronal protein sorting [117]. Recently, a second synaptic receptor, Neuroligin-2, was reported to interact with and be dependent on retromer for trafficking [128]. Retromer and Neuroligin-2 were observed to have a substantial amount of co-localization in cortical neuronal cultures where they localized to both the dendrite and soma. Together, this data demonstrates the requirement of retromer-mediated receptor delivery to discrete locations throughout the dendrite surface. Therefore, retromer involvement in the trafficking of other synaptic receptors - such as the ionotropic glutamate receptors NMDAR (N-methyl-D-aspartate receptor) and AMPAR, that underpin learning, memory and transmission of action potentials - is not unexpected [143]. Recent work using hippocampal slices bearing neurons transfected with shRNA against Vps35 demonstrated marked reductions in synaptic AMPA and NDMA excitatory post- synaptic currents, whereas GABA inhibitory post-synaptic currents were unchanged [117]. Although it is unlikely that these data completely explain the mechanism underlying trafficking of receptors at the synapse, it is likely that retromer plays a fundamental role in conjunction with other membrane recycling proteins such as SNX27 and PICK1 [89, 144], to regulate synaptic function, physiology and disease [145]. Evidence of retromer-regulated synaptic health and neuronal physiology was supported when Vps26 heterozygote knockout mice demonstrated hippocampal memory problems, impaired synaptic function and elevated production of endogenous proteolytic forms of processed APP, generally considered neurotoxic [146]. Earlier reports using hippocampal neurons expressing shRNA against Vps35 demonstrated a reduction in the frequency of APP positive endosomes [147], consistent with the notion that retromer may regulate recycling of certain transmembrane receptors at the synapse. These findings support the notion that retromer maintains the repopulation of certain receptors at the synapse while ensuring long-range retrograde transport is facilitated at an efficient manner.

1.3.1 Trafficking and Neurodegeneration Neurodegenerative diseases are progressive, irreversible brain disorders characterized by a decline in motor and cognitive functions. Histologically, these pathologies are often accompanied by protein aggregation and selective loss of neurons. However, why 23 certain areas of the brain appear more susceptible than others is still poorly understood. Over the past decade though, compounding evidence supports a direct link between endosome trafficking and the onset of neurodegenerative diseases, including AD and PD [107, 147, 148].

1.3.2 Retromer in Alzheimer’s disease Amyloid precursor protein (APP) is a type 1 transmembrane protein that has been reported to play a fundamental role in multiple cellular events; including neurite outgrowth, cell adhesion, protein transport, synaptogenesis and cell signalling [149]. Following synthesis in the endoplasmic reticulum, APP is transported to the TGN and then sorted to the plasma membrane, the site of which proteolytic cleavage mediated by various secretase enzymes resulting in differing APP fragments may occur, or eventual internalization as full length APP [150, 151]. Following internalization, APP can be recycled to the cell surface, transported to the TGN, or sorted to the lysosome for degradation [151-153]. Like many other transmembrane proteins in mammalian systems, a large amount of synthesized APP protein is stored in the TGN and not at the plasma membrane [151].

Under pathogenic conditions in AD, APP is cleaved by β-secretase BACE1 (β-site APP cleaving enzyme 1) through the ectodomain, releasing soluble APP-β (sAPPβ) into the extracellular space [154-156]. The remaining β C-terminal anchor fragment (β-CTF) is then cleaved by γ-secretase complex, leading to the production of the Amyloid-β (Aβ) peptide and an APP intracellular domain (AICD). Alternatively, under non-pathogenic conditions, APP is cleavage by α-secretase through the motif needed for β-secretase, leading to the production of soluble APP-α. From here, the membrane-anchored α C-terminal fragment (α- CTF) is cleaved by γ-secretase resulting in AICD and p3 peptide production, inhibiting the production of any toxic protein species (for extensive reviews on APP processing see [149, 157-159]). During periods where APP is retained within the endosomal system, growing evidence supports an increase in processing via the amyloidogenic pathway, producing toxic Aβ [111, 146, 160]. In contrast to this, accumulating evidence supports a spatial distinction between APP and β-secretase suggesting the need for fusion of specific compartments or additional endosome co-factors needed to drive amyloidogenic processing. One such protein is ARF6, the endosome localized GTPase ribosylation factor 6. Alterations to either the functionality or expression of ARF6 results in impaired internalization of BACE1 and changes in the processing of APP to Aβ, highlighting the 24 importance of additional endosome proteins in maintaining processing of APP away from the pathogenic state [161]. On the other hand, SNX12 has been demonstrated to interact with BACE1 and influence trafficking between the endosome and cell surface [162].

Furthermore, several lines of evidence support the involvement of retromer in AD pathology. These include observations of reduced levels of retromer subunits in the brains of AD patients [146], retromer’s involvement in recognition and trafficking of APP receptors and regulators [111, 163], as well as retromer’s interaction with BACE1, an APP processing enzyme [163]. Post mortem studies have demonstrated striking declines in the retromer subunit, Vps35, in patients suffering late-onset Alzheimer’s disease (LOAD). This loss appears to be confined to the hippocampal regions, a site demonstrated to be imperative for memory formation and storage [146, 148, 164]. Despite APP shuttling from the endosome to TGN, retromer does not appear to directly interact with APP itself. Successful retrieval of APP to the TGN is mediated via retromer cargo molecules SorLA and sortilin, members of the Vps10p-domain containing APOE binding family of receptors [165, 166]. However, it appears that the interaction of either SorLA or sortilin with APP may have distinct roles, or at least function in distinct pathways. Sortilin and APP exhibit a high level of overlap throughout the neurite of the neuron, whereas SorLA and APP show a heavy localization to the soma, suggesting distinct spatio-temporal pathways [167]. As retromer is considered a regulator of APP trafficking and processing, it is of note that SorLA, a well-defined retromer cargo, is reported to be heavily down regulated in the frontal cortex of AD patients and has a strong correlation with impaired cognitive decline [165, 168]. Interestingly, recent work identified a small molecular chaperone (termed R55) that decreases pathogenic processing of APP [160], via an increase in retromer’s overall stability. Treatment of hippocampal neurons with R55 increased retromer levels and shifted SorLA and its receptor, APP, away from endosomal localization and therefore had decreased pathogenic APP processing, demonstrating a neuroprotective mechanism. Together these findings demonstrate a pivotal role for retromer in the regulation of APP trafficking and identify it as a potential beneficial therapeutic target. Although retromer is promising candidate for therapeutic target in AD, one limitation of this study is the relevance of studying viable Hippocampal neurons derived from non-transgenic mice with pharmacological agents that will be employed in symptomatic AD patients. As previous studies have demonstrated the loss of retromer transcript expression in brain tissue isolated from AD patients [148], and little evidence exists supporting loss-of-function mutations for 25 retromer in AD, this begs the question of the need to target the up regulation of retromer mRNA levels instead of targeting protein stability/function to halt progression of AD.

1.3.3 Parkinson’s disease Parkinson’s disease (PD) is an incurable neurological condition that is estimated to affect 2% of the population aged above 60. Patients show signs of memory deficiency, confusion, and impaired movement [169]. PD primarily affects the dopaminergic neurons of the Substantia nigra pars compacta; a small anatomical region located within the mid-brain [170]. The dopaminergic neurons of the Substantia nigra are responsible for release of the dopamine (DA), a neurotransmitter shown to regulate movement, coordination and addiction [171, 172]. Primary symptoms of PD such as shuffling gate, rigidity, confusion and bradykinesia are not witnessed until approximately 50% of the dopaminergic neurons are lost, a point at which the disease continues to progressively worsen [171]. Although current understanding of why these particular neurons undergo selective loss is minimal, the pathway associated with motor planning and control and how it is impacted in PD is well characterized [171]. At rest, neurons within the globus pallidus are spontaneously active and thus are constantly inhibiting the Ventrolateral (VL) nucleus of the thalamus, blocking activation of innervated motor areas involved in motor planning. Following input from the cortex, the neurons of the globus pallidus are inhibited, freeing up the VL and subsequently allowing positive signals to be sent to the supplementary motor area (SMA), which is used in movement planning. Moreover, neurons from the Substantia nigra actively signal to the globus pallidus and control unwanted movement initiated by the motor loop. Following the loss of the dopaminergic neurons, inability to keep the motor loop inhibited is lost, resulting in unwanted movement witnessed in patients suffering PD [171]. Furthermore, dopaminergic neurons also project to the frontal lobe of the brain, the section of the cerebral cortex associated with rational thought, behavior and problem solving [172, 173]. As the dopaminergic neurons begin to become less viable, PD patients begin to exhibit signs of cognitive impairment and other neurological symptoms (altered sleep-wake cycles, anxiety and depression) [171].

Histologically, PD is characterized by the formation of large, perinuclear accumulations of insoluble proteins termed Lewy Bodies (LB), which are immunoreactive for the protein α-synuclein [174-176]. Interestingly, LBs have been reported in neurons of the brain stem and cerebral cortex and demonstrate differing properties depending on their 26 location. LBs found within the brainstem characteristically appear with a dense core and surrounding halo-like glow, whereas cortical LBs present with more of a homogeneous organization, inferring potential differing mechanisms underlying their formation and composition [177, 178]. Overall, the formation of the LBs is thought to be detrimental to the health of the neuron, as marked by changes in cell function, morphology and lifespan, and ultimately the generation of PD in patients [179-183]. However, despite the presence of α- synuclein positive LBs being a diagnostic hallmark of PD, alternative evidence also supports the presence of α-synuclein and LBs having a neuroprotective mechanism to sequester protein into an insoluble mass and prolong neuron survival [184-186], similar that observed in Huntington disease [187]. Why these differences exist is currently not clear, however it lends to the hypothesis that PD is multi-factorial and is influenced by secondary non-α- synuclein based mechanisms (signaling pathways, lipid oxidation, pro/anti-apoptotic stimuli and nutrient deprivation) [186, 188].

1.3.3.1 Structure and localization of α-synuclein α-synuclein, or PARK1, is a small, unstructured, monomeric protein reported to play a primary role in the pathogenesis of PD [179-181, 189]. Encoded by the gene SNCA (4q21.3-q22), α-synuclein is highly expressed throughout the brain, primarily in neurons, but is also expressed in other tissues [190]. Despite this broad tissue expression, the exact role α-synuclein has in the human body, and more importantly in the neuron, still remains unclear.

Structurally, α-synuclein contains 3 individual domains – N-terminal, central and C- terminal domain (see figure 1.6). The N-terminal domain of α-synuclein (1-60AA) is characterized by the presence of the loosely repeated consensus sequence (KTKEGV), which is believed to introduce hydrophobicity into the protein. Presence of 11 KTKEGV N- terminal repeats gives α-synuclein characteristics similar to those found in helical structures of apolipoproteins, an observation further supported by secondary structure prediction [191, 192]. The central region of the α-synuclein (61-95AA) comprises a domain which contains amino acids prone to self-aggregation and is called the Non-Amyloid beta Component (NAC) domain [189]. Structural evidence of recombinant NAC domain (57-102AA) demonstrated that it may possess helical conformation that allows direct interaction with the micelle surface [193]. In contrast, the Carboxy terminal does not have an known secondary structure and overall, possesses a strong negative charge due to the strong prevalence of 27 aspartic and glutamic acid residues [191, 194]. Interestingly, the C-terminal tail domain holds a calcium (II) binding sequence that has been directly implicated in α-synuclein aggregation [195, 196] as truncation studies using α-synuclein-ΔC do not form calcium-induced aggregates.

Figure 1.6: Domain architecture of monomeric α-synuclein

Representation of unstructured α-synuclein highlighting N-terminal lipid binding domain (green), common familial point mutations (yellow), central hydrophobic domain (purple) with identified sequences prone to self-aggregation (red) and C-terminal, calcium binding domain (blue).

Complementary to these domain architecture studies is the discovery that monomeric α-synuclein interacts with synthetic acidic phospholipid containing vesicles [197, 198], via its N-terminal domain [199]. Using Nuclear magnetic resonance (NMR) of membrane- mimetic sodium dodecyl sulfate (SDS) micelles, α-synuclein was shown to be composed of two curved, antiparallel α-helices joined by a flexible linker at residues 38-44 [181, 200, 201] and proposed to adopt an overall “horseshoe” appearance [202]. This “horseshoe” conformation of α-synuclein, which is facilitated by its interaction with the lipid membrane, is thought to aid in its structural stabilization and decreased exposure of the NAC domain to the cytosol, potentially acting as a protective mechanism to limit aggregate formation [203].

28

Initial subcellular localization studies using anti-α-synuclein antibodies in rat primary neurons7demonstrated strong labeling of neuronal synapses [192, 204], a phenotype drastically different to patients suffering Parkinson’s disease. Why α-synuclein is localized to the synaptic junction is not well defined, however, several lines of evidence suggest a role in regulation of synaptic vesicle dynamics [205, 206] and endosome recycling kinetics [207]. Confirmation of α-synucleins functional role in synaptic vesicle release was demonstrated by its direct interaction with synaptobrevin-2/vesicle-associated membrane protein 2 (VAMP2) of the soluble N-ethylmaleimide–sensitive factor attachment protein receptor (SNARE) complex. Interaction between VAMP2 and α-synuclein was shown to facilitate the assembly and disassembly of the SNARE complex that directly functions in docking and release of pre-synaptic vesicles, signifying a direct function of a-synuclein in high-frequency pre-synaptic vesicle release [208]. Overall, this observation supported earlier work describing strong localization of α-synuclein to the pre-synaptic junction in primary neurons and may explain the dysregulation of DA release shown to occur in mouse models and patients of PD [192].

1.3.3.2 α-synuclein pathogenesis Given the anchorage of α-synuclein to the lipid membrane, it is likely that remodeling of vesicle membranes, oxidative modification of lipids via depolarization or dysregulation of metal ion homeostasis, all of which may occur at the neuronal synapse, may be directly responsible for initiating aggregation [208-210].

Disrupting the ability for α-synuclein to engage the lipid membrane reportedly increases cytosolic aggregation and neurotoxicity [211], supporting the notion of α-synuclein adopting a α-helical confirmation when bound to membranes by limiting exposure of the NAC to the cytoplasm [200, 202, 203]. Under pathological conditions where α-synuclein dissociates from the membrane, it is possible that it undergoes self-association and forms β-sheet oligomeric structures, an initiation step for fibril formation [211, 212]. The formation of fibrils is an essential building block in the generation of aggregated α-synuclein species. This is achieved by manipulating proto-fibril structures into a partial secondary structure rich in β-sheets [213]. Using X-ray crystallography and X-ray diffraction technology, independent studies demonstrated that α-synuclein fibrils are composed of proto-filaments in which β- strands are organized in a parallel manner [214, 215]. The side chains of amino acids protruding from each individual β-sheet are organized in a highly systematized, 29 complementary manner, which results in the production of steric zippers and excludes water from reaching the interface between the two sheets [213, 216-219]. At a molecular level, the fibrils can be classified as either straight or twisted ribbons [214, 215] that are divided by the arrangement of the proto-fibrils. For example, a straight fibril is composed of 4-5 β-sheets aligned in a unidirectional manner, whereas in the ribbon structure, two proto-fibrils twist around each other and then further twist around another ribbon proto-fibril to form a fibril, similar to organization of DNA [214, 215]. Here, the mature fibrils amorphously organize into a cross-linked meshwork and further into sphere-like masses, termed Lewy Bodies [220].

1.3.3.3 Proteasomal degradation of α-synuclein Despite the LB being a mass of aggregated protein, several studies have elucidated that both the Ubiquitin-Proteasome system (UPS) and the Auto-lysosomal Pathway (ALP) control the turnover of α-synuclein and may underpin the onset and progression of aggregate formation. Initial hints that the UPS plays a fundamental role in the clearance of α-synuclein arose from the observation that post-mortem LBs isolated from PD patients were enriched for ubiquitin, subunits of the proteasome and α-synuclein phosphorylated at Serine129 (pS129) [221-224]. More recently, a number of studies further supported this hypothesis by demonstrating that under non-pathogenic conditions, monomeric α-synuclein is targeted by the UPS [225-227], suggesting that post-translation modifications of α- synuclein may regulate its degradation. In contrast to these observations, oligomeric forms of α-synuclein preferentially bind to and inhibit function of the proteasome while selectively being redirected to the ALP (see below), supporting potential cross talk between the UPS and ALP systems [228]. For example, Tenreiro et al 2014 reported that phosphorylation of α-synuclein at serine129 leads to increased proteasomal degradation, as assayed by appearance of cytoplasmic inclusions and reduced cellular toxicity, when compared to phosphorylation deficient mutant S129A that exhibited increased production of inclusions and toxicity. Additionally, cells expressing α-synuclein-S129A demonstrated a reduced level of autophagy induction when compared to that observed in cells expressing α-synuclein-WT [229], suggesting that phosphorylation may regulate cross talk between the two degradation systems and ultimately contribute to the formation of LBs in PD patients.

PD is not the only neurological disorder which has links with abnormal functionality of the proteasome. For example, Wang and colleagues demonstrated the impairment of the synaptic proteasome in neurons isolated from HD mice expressing mutant Htt, which 30 consequently lead to the build-up of aggregated mHTT [230]. Interestingly though, this aligns with and builds upon an earlier report demonstrating that impairment of the UPS occurs prior to the formation of visible aggregates, suggesting that the malfunction in degradation is a symptom of mutant protein or early stage toxicity [231]. Additionally, as the proteasome is able to degrade soluble monomeric protein species, it begs the question of whether expanded poly-glutamine containing proteins are efficiently degraded via the proteasome. Work to support this hypothesis came independently from Holmberg et al and Venkatraman et al, who demonstrated that poly-glutamine containing proteins are incompletely degraded via the proteasome and must be released from the proteasome to undergo further rounds of proteolysis [232, 233]. Therefore, it is clear that both α-synuclein and Huntington may selectively modulate the function of the proteasome in order to limit proteolysis and potential cellular toxicity.

1.3.3.4 Degradation of α-synuclein within the lysosomal system Although α-synuclein aggregation has been largely implicated in the pathogenesis of PD, several other factors have been shown to contribute to the progressive decline in general neuronal function, one of which is the lysosomal system. The lysosome itself is a small, highly acidic, membrane-enclosed compartment that degrades cellular proteins using resident proteases and fusion with the autophagosome. Given that the lysosome, and for that matter, all endosomes, are membrane-enclosed structures, several different pathways have been described for entry of proteins into the endosomal compartment. These include (1) endocytosis of receptors [102], (2) fusion with an autophagosome (Macroautophagy) [33] or (3) Chaperone Mediated Autophagy (CMA) [49].

CMA has recently been recognized and described as a pathway used by α-synuclein to selectively gain entry into the lysosomes, as no specific endosomal receptor has been described for α-synuclein, and macroautophagy is a non-selective, bulk degradative pathway [33, 49, 234]. CMA actively delivers cytosolic proteins to the lysosome via Heat shock cognate protein 70 (Hsc70), Lysosomal associated membrane protein 2 (LAMP2) and a pentapeptide consensus sequence found in substrate proteins [49, 235]. Following engagement of Hsc70 within the substrate protein in the cytosol, direct chaperone activity and subsequent interaction between the substrate protein and the LAMP2A protein occurs [236]. One of the many substrate proteins containing a CMA recognition sequence is α- synuclein [49, 236]. Interestingly, the previously described α-synuclein familial PD point 31 mutations A53T and A30P have been shown to inhibit the chaperone-mediated autophagy (CMA) pathway by impeding translocation into the lysosome [49]. Successful delivery of α- synuclein to the lysosomal membrane by Hsc70 allows for a direct interaction between α- synuclein and monomeric LAMP2A receptor. As a result of this interaction, a 700 kDa multimeric pore complex composed of several LAMP2A receptors is constructed and allows substrate proteins to transverse the lysosomal membrane [237]. Interestingly, reorganization of monomeric LAMP2A into a multimer is stabilized by luminal Heat Shock Protein-90 (Hsp90) and not the chaperone Hsc70 [237]. Simultaneous to organization of a stable multimeric LAMP2A structure, Hsc70 likely facilitates the unfolding of substrate proteins to allow for entry into the lumen of the lysosome. Whether cytosolic Hsc70 plays a role in the physical translocation of unfolded substrate proteins into the lumen is not known. However, luminal Hsc70 (lys-Hsc) is thought to complete the translocation by pulling the substrate protein through the pore and impair it reentering the cytosol [238]. Interestingly, active delivery and the formation of multimeric LAMP2A complexes appears to decline with age [239, 240] and likely contributes to neurological diseases such as PD, further supporting a decline in lysosomal function observed with age.

Similarly, several reports highlight the importance of the lyso-autophagy system in other neurological diseases such as Huntington’s [53, 241]. Interestingly, although the presence of mHtt appears to positively influence the autophagy pathway, the ability for it to be cleared is perturbed [53, 241]. For example, Koga et al demonstrated the increased localization of LAMP2 receptor, along with the cytosolic chaperon hsc70, to the lysosomal membrane. Using these same cells lines, it was observed that the basal level of CMA was significantly unregulated whereas the turnover of LAMP2 in HD cells was delayed when compared to wild type cells. Consistent with these findings is the observation that HD cells expressing 111-QHtt repeats fail to turnover autophagic substrates by impaired cargo recognition. Martinez-Vicente et al report that HD cells derived from both mouse models and humans with HD have fully formed autophagic vesicles, but fail to trap substrate molecules in the lumen, a vital step to ensure turnover of potentially toxic proteins such as Htt and α- synuclein.

Despite there being strong evidence of a dysfunctional autophagy system in HD, there appears to be little to no impact on the function of the lysosome, or the degradation of Htt between WT and HD cells [241, 242]. Interestingly though, several reports highlight significant changes in the sub-cellular localization of LAMP1/LAMP2 positive compartments 32 from a uniform cellular distribution to a condensed, perinuclear redistribution. However, despite these strong phenotypical changes, the overall function and manipulation via the use of protease inhibitors appears to be identical to that witnessed in non-HD cells [242]. Second to this, processing of Cathepsin D into its mature form, a process that relies heavily on the presence of an acidic lumen, is reportedly normal in HEK293 cells expressing full- length Htt protein with a short polyQ repeat (23QHtt) or mHtt protein with a long polyQ repeat (145QmHtt), suggesting that trafficking of Golgi-derived compartments and the lysosomal system is unaffected [243].

1.3.3.5 Cathepsin D Once inside the lumen of the lysosome, substrate proteins are irreversibly cleaved by lysosomal proteases. For example, α-synuclein is readily cleaved by the aspartic protease cathepsin D [244]. Mature cathepsin D is an aspartic protease localized to endosome and lysosome compartments, where it facilitates the degradation of delivered proteins [245, 246]. Cathepsin D, like many other enzymes, is synthesized as an inactive precursor that undergoes several post-translational modifications yielding an active peptide.

The 52kDa inactive precursor is synthesized on the rough endoplasmic reticulum (RER) as a pre-pro enzyme which is subsequently imported into the Golgi network, where the N-terminal signal peptide is cleaved to generate a 50 kDa pro-cathepsin D protein. Further, several mannose residues on pro-cathepsin D are phosphorylated [247] and act as a signal for engagement with one of the two Mannose-6-Phosphate Receptors, the Cation Independent (CI-MPR) and the Cation Dependent (CD-M6PR). Interaction between M6PR and Cathepsin D results in export of the receptor-ligand from the Golgi via Clathrin Coated vesicles (CCVs), through direct interaction of the CI-M6PR C-terminal tail and the AP-1 adaptor protein [248].

Delivery of CI-M6PR/Cathepsin D to the maturing endosome compartment permits disengagement of the complex and further allows for enzymatic processing of pro-Cathepsin D by cysteine and aspartic proteases [249, 250] and return of CI-MPR to the Golgi apparatus via Rab9 [251, 252], its effector p40 [253] and Tail Interacting Protein of 47 kDa (TIP47) [254, 255] dependent pathway. Successful cleavage of Pro- cathepsin D cleavage results in two mature peptides – 14 kDa and 34 kDa, of which 14 kDa Cathepsin D regulates degradation of proteins delivered to the lysosome [245, 246]. Therefore, the proteolytic 33 activity of Cathepsin D within the endosomal system is indeed dependent on its correct sorting from the TGN and, subsequent processing within the endosomal system, both cellular function that are regulated by CI-M6PR and the retromer complex.

Figure 1.7: Step-wise processing and sub-cellular localization of the Cathepsin D. Following synthesis, Cathepsin D is exported from the ER and subjected to several rounds of enzymatic processing before yielding the lysosomal-localized, light chain peptide (red).

1.4 Retromer and Parkinson’s disease A number of genetic studies have identified several familial point mutations within the retromer complex that may be causative in late onset PD. Initial Genome wide association studies (GWAS) identified mutations within the Vps35 (PARK17) subunit of the retromer complex [256-258]. Recent studies have identified further point mutations that may be linked to PD in subunits Vps29 and Vps26 [259, 260] [260] (see Table 2), however it is not clear if the identified mutations are causative of PD.

To date, little information exists surrounding how these identified point mutations may play a role in the biology of PD. However, recent in vivo data of Drosophila expressing human Vps35 D620N mutations revealed the loss of Tyrosine Hydroxylase (TH)-positive Dopaminergic neurons, decreased lifespan and heightened sensitivity to the mitochondrial toxin, rotenone [261]. Expression of Vps35 P316S, which was identified in both patients and controls [258] [262], phenocopied the Vps35 D620N induced cell loss to a lesser extent, 34 whereas expression of Vps35 L774M demonstrated no signs of cellular toxicity [261]. Although this does support the notion that expression of Vps35 mutations linked to PD may be toxic at the whole organism level, a recent publication demonstrates contradictory evidence and suggests that expression of Vps35 Drosophila orthologos confer no dominant toxicity [263].

Gene Substitution Location Found in Controls Reference(s)

Vps35 G51S Exon 3 Y [256] Vps35 M57I Exon 3 Y [256] Vps35 T82R Exon 4 Y [256] Vps35 I241M Exon 7 N [256] Vps35 P316S Exon 9 Y [258, 262] Vps35 R524W Exon 13 N [256, 258] Vps35 I560T Exon 14 Y [264] Vps35 H599R Exon 14 N [264] Vps35 M607V Exon 14 N [264] Vps35 D620N Exon 15 N [256-258, 262, 265] Vps35 L774M Exon 17 Y [256] Vps35 G51S Exon 3 N [260]

Vps26A M112I Exon 4 N [260, 266] Vps26A K93E Exon 4 N [259, 260] Vps26A R127H Exon 4 Y [260] Vps26A K297X Exon 9 N [260] Vps26A N308D Exon 9 Y [260] Vps26A P316S Exon 9 N [266]

Vps29 N72H Exon 5 Y [259] Table 2: Retromer mutations linked to Parkinson’s disease

One potential explanation underlying the toxicity witnessed following expression of the Vps35 D620N is supported by recent publications reporting a loss of function mutation leading to impaired retrograde sorting by the retromer complex [267, 268]. Consistent with these initial cell culture studies was the observation that over-expression of Vps35 WT, but 35 not Vps35 D620N, successfully rescued LRRK2 induced missorting of the CI-M6PR, further supporting a direct impact of Vps35 D620N on the retrograde pathway [269]. These studies collectively demonstrate that retromer may function downstream of LRRK2 to facilitate endosomal sorting [269] and also highlight the possibility of various mutations identified in controls possessing a role in the molecular mechanisms of disease.

Whether other retromer-associated pathways are also impacted by the presence of D620N is controversial. Investigation into recycling or SNX27-retromer dependent pathway has produced conflicting results. Using GLUT1 as a marker for receptor recycling, one study reported a loss of cell surface staining with an increase in intracellular puncta staining, reminiscent of failed recycling [270]. On the other hand, a second study reported plasma membrane biotinylation levels of GLUT1 are identical between Vps35WT and Vps35 D620N-expressing cells, suggesting no impact on the recycling pathway. Moreover, expression of Vps35 D620N reportedly rescued GLUT1 missorting to the lysosome following knockdown of endogenous Vps35, highlighting no strong impact on the recycling pathway, or at least on the GLUT1 transporter [268, 270]. Despite these differences, and while confirming the SNX27-Vps35 D620N interaction, neither study reported investigating the sub-cellular localization of SNX27 following expression of the D620N mutant, a protein reported to be fundamental in the endosome to cell surface recycling [268].

Given the reported impact on the retrograde pathway and its link to PD, it is not unexpected that expression of D620N will exhibit a level of cellular toxicity. Expression of D620N, and to a lesser extent, Vps35 WT, in cortical neurons markedly reduced cell viability as indicated by TUNEL staining; a phenotype that was further exaggerated following challenge with either rotenone or hydrogen peroxide in comparison to controls [271]. Interestingly, these findings suggest greater susceptibility to cellular stress and a heightened level of apoptosis in the presence of D620N-containing retromer, consistent with recent work published using Drosophila [261] . Whether these results are a secondary consequence of receptor mis-trafficking or impact on other pathways directly is not known.

1.4.1 PD mutations in relation to retromer structure As mentioned above, retromer is comprised of a stable heterotrimer of Vps35, Vps26 and Vps29. Crystal structures are now available for Vps26A and Vps26B [68, 272], Vps29 [273-275], and Vps29 in complex with a C-terminal fragment of Vps35 [276]. Vps26 belongs 36 to the arrestin-family of proteins. Vps29 resembles metallophosphoesterase enzymes with similar N- and C-terminal sub-domains, composed of -sandwich folds, but alterations in active site residues render it catalytically inactive. Vps35 is composed of 17 -helical HEAT- like repeats. Electron microscopy and small angle X-ray scattering studies of retromer have led to a model of the holo-complex whereby Vps35 acts as an elongated structure bound by Vps26 and Vps29 at independent N- and C-terminal regions [276, 277].

Overall, the retromer complex provides a highly extended scaffold for engaging accessory proteins and cargo molecules. The structure-function relationship of the various mutations of retromer remains unclear. While the D620N mutation is adjacent to the Vps29 binding interface, it does not affect Vps29 (or Vps26) interactions [267]. The causal explanation for the subtle trafficking defects observed for D620N mutant and how the reduced association with both the WASH complex and FKBP15 occurs is not clear at the structural level [268, 270]. Other mutations in Vps35 reside within the same structural region of Vps35 (near the Vps29 binding site), while mutations in the N-terminus of Vps35 could potentially influence interactions with a number of proteins including Vps26, SNX3 and SNX1 - although none of these mutants have been characterised to date. The K93E mutation of Vps26 will alter the surface charge of the protein, but does not reside near any known binding sites; including surfaces for Vps35 [272, 277] and SNX27 [95]. Perhaps more interestingly, the N72H mutation in Vps29 resides directly at the interface with Vps35 and could therefore affect complex integrity [276]. Overall, it is clear that further characterization of the various retromer mutations, combined with high resolution structures of retromer in complex with different accessory and cargo molecules will be required to reveal the mechanistic impacts of the mutants on retromer function.

1.4.2 α-synuclein degradation, Cathepsin D and the Retromer complex The past two decades of researching the role of Cathepsin D in α-synuclein protein homeostasis has provided valuable information regarding potential mechanisms underlying an inability of neuronal cells to clear lysosomal α-synuclein. In an initial study published in 2001, recombinant α-synuclein was digested with α-chymotrypsin or cathepsin D in an attempt to map the central hydrophobic domain [278, 279]. Interestingly, post-digestion mass-spectral analysis of α-synuclein treated with cathepsin D revealed intact central hydrophobic and N-terminal domains, whereas the C-terminal domain was found to be digested at multiple cleavage sites [244]. This report was the first to demonstrate a direct 37 link between processing of α-synuclein and a lysosomal protease in vitro. Further, cell-based studies using shRNA against cathepsin D, demonstrated marked increase in total α- synuclein levels in purified lysosomes [280]. To compliment these observations, mice devoid of cathepsin D (Ctsd-/-) were found to have a maximum postnatal lifespan of 26 days and isolated cortical regions demonstrated a strong shift in the formation of higher molecular weight α-synuclein species, both soluble and insoluble fractions, while minimal changes in monomeric species or total mRNA levels were observed [281]. Further, use of a C. elegans model to monitor α-synuclein aggregation in the presence of full-length wild-type Cathepsin D, Cathepsin B or Cathepsin L revealed that Cathepsins B and L did not reduce levels of α- synuclein toxicity when compared to the presence of full-length Cathepsin D, suggesting a highly specific, evolutionarily conserved mechanism and overall protective role for Cathepsin D resulting in cleavage of α-synuclein [282]. Taken as a whole, the in vitro and in vivo data demonstrate the importance and specificity of Cathepsin D in proteosomal degradation of monomeric α-synuclein, which leads to neuroprotection by decreasing the levels of Lewy Body formation and resulting α-synuclein toxicity.

Retromer is an evolutionarily conserved protein sorting complex that regulates the sub-cellular localization of various receptors (See table 1 of introduction), one of which is the previously described CI-M6PR (see section 4.2.5). Following endocytosis, receptors are sorted into early endosomes, concentrated in nascent tubules rich in retromer sorting machinery and sorted into newly formed vesicular compartments. In order to maintain lysosomal homeostasis, retromer-rich endosomes will deliver endocytosed receptors, such as CI-M6PR, to the TGN where it is able to engage multiple ligands, including Cathepsin D. As detailed in the introduction (see section 1.4), loss of retromer function, either by RNAi or mutations linked to PD [70, 102, 267], strongly impacts the sub-cellular localization of endocytosed receptors destined for the TGN and subsequently leads to large missorting of Cathepsin D from the TGN. As such, the overall hypothesis for this thesis is: loss of the retromer complex function and subsequent missorting of CI-M6PR and Cathepsin D will result in the accumulation of α-synuclein in the late endosomal network and lead to formation of α-synuclein positive inclusions. 38

1.5 Aims and Hypotheses

Currently, our understanding of how retromer may regulate the biological changes linked with PD is limited. It can be hypothesized that the loss of retromer expression, or alteration to retromer’s function through familial linked mutations, will negatively impact the endo- lysosomal system, delay the sorting of transmembrane cargo proteins within the cell, and result in impaired lysosomal degradation of potentially toxic protein species associated in PD, such as α-synuclein. Here, this thesis will examine this hypothesis via the following aims:

Investigating the relationship between retromer and α-synuclein To understand the function of Retromer in regulating the endo-lysosomal system and degradation of α-synuclein, characterization of the Dopaminergic cell line (SH-SY5Y) will occur. Cells stably knocked down for Vps35 will be generated and the sub-cellular localization of α-synuclein, alongside the size and frequency of any observed inclusions will be investigated. Second to this, characterization of the endo-lysosomal system using known organelle markers will be employed to determine morphological changes in the absence of retromer and the potential localization of retromer to α-synuclein inclusions.

Characterization the Vps35 D620N mutation linked to Parkinson’s disease I will generate GFP fusion constructs harboring the Asp620Asn mutation to test the impact of the mutation on, A) the formation of the retromer complex, B) the sub-cellular localization of the Vps35 D620N containing complex within the endosomal system, and C) the cargo sorting properties of retromer in the presence of the Vps35 D620N. In addition to these ectopic expression studies, I will use a Human Dermal Fibroblast line derived from a PD patient genotyped for the Asp620Asn variant and complement any findings from the ectopic model.

Investigating the relationship between PD-linked Vps35 mutations, retromer function, α-synuclein aggregation and pharmacological modulators of retromer. For this study, I will initially investigate the impact of the PD-linked mutations Vps35 P316S and Vps35 R524W on, A) the formation of the retromer complex, B) sub-cellular localization of both mutants C), the ability of Vps35 P316S and Vps35 R524W to disrupt interactions 39 with retromer interacting proteins and D) the cargo sorting properties in the presence of each mutation. Following this, I will investigate the ability for Vps35 D620N, Vps35 P316S and Vps35 R524W expression to induced α-synuclein positive inclusions. Lastly, I will employ a molecular chaperone that stabilizes the retromer complex in an attempt to modulate the production of α-synuclein positive inclusions.

The completion of these aims will allow the detailed understanding of the molecular mechanism of retromer in regulating the endolysosmal system and turnover of α-synuclein, as well as the characterization of each described Vps35 mutation within a cellular system. This will provide a basis for our understanding of the relationship between the retromer complex and PD, an area that is currently ill defined. 40

Chapter 2: Materials and Methodology

2.1 Materials: Thiophene-2,5-diylbis(methylene) dicarbamimidothioate dihydrochloride (TPT-260,R55) was purchased from MedKoo Biosciences (North Carolina, USA) and re- suspended in DMSO at 10mM.

2.2 DNA Constructs: Full length human wild-type Vps35 was amplified using a 5′ primer CCCCTCGAGATGCCTACAACACAGCAG and 3′ primer CCCGGATCCAGAAGGATGAGACCTTCATA and sub-cloned into pEGFP-N1 using BamH1 and XhoI. The D620N, P316S and R524W point mutation were generated using a Quikchange mutagenesis kit (Stratagene) using the following PCR mutagenic primer pairs: Vps35D620N; 5′-GAAGATGAAATCAGTAATTCTAAAGCACAGCTG and 3′- CAGCTGTGCTTTCGAATTACTGATTTCATCTTC. Vps35 P316S; 5’ GCTCACCGTGAAGATGGATCCGGAATCCCAGCGGAT and 3’- ATCCGCTGGGATTCCAGGACCATCTTCACGGTGAGC and Vps35 R524W; 5’ GCTGGTGGAAATCAGTGGATTCGCTTCACACTG and 3’ CAGTGTGAAGCGAATCCACTGATTTCCACCAGC, respectively. The pCMU-CD8/CI-M6PR construct was described previously [67]. Constructs for expression in Escherichia coli of mouse Vps35, Vps29 and Vps26A were described previously [277]. The D620N, P316S and R524W point mutations within Vps35 were inserted into pGEX4T-2 (GE Healthcare) was generated as described above for the human construct.

2.2.1 Protein purification and isothermal titration calorimetry (ITC) and circular dichroism (CD) spectroscopy: Recombinant proteins used for the ITC experiments were prepared as previously described [277]. All proteins were further purified by gel filtration chromatography using 20 mM Tris (pH 8.0), 200 mM NaCl, 1 mM DTT (ITC buffer). Isothermal titration calorimetry was carried out at 283 K using a MicroCal iTC200 (GE Healthcare), with 16 x 2.5 µl injections of 100 M Vps29 into 10 M Vps35, or 50 M Vps26A into 5 M Vps35. Integration of the titration curves was performed using the ORIGIN software (OriginLab) to extract thermodynamic parameters, stoichiometry N, equilibrium

-1 association constant Ka (=Kd ) and the binding enthalpy H. The Gibbs free energy of binding G was calculated from the relation G = -RTln(Ka) and the binding entropy S was 41 deduced from the equation (G = H - TS). CD spectra were recorded on Jasco-810 spectropolarimeter (Jasco GmbH) at 0.01 cm-1 optical path, 0.5 nm interval, 1 nm bandwidth and 50 nm/min scanning speed. Protein samples were set at 3 mg/mL in ITC buffer and their polarization spectra were recorded three times followed by averaging and background subtraction. The HT voltage was below 500 V over the entire range of recording (200-250 nm).

2.2.2 Lentivirus production and generation of stable knockdown cell lines: Generation of SH-SY5Y stable knockdown cell lines were produced using a lentivirus small hairpin RNA (shRNA) system (Thermo Scientific, USA). Generation of lentivirus particles was achieved by co-transfection of HEK293T cells with pGIPZ shRNA plasmids (non-silencing control shRNA, RHS4346; human Vps35 shRNA, V2LHS_156301) and Trans-lentiviral packaging plasmids (Thermo Scientific) using calcium phosphate transfection protocol. pGIPZ shRNA plasmids and Trans-lentiviral packaging plasmids were diluted at a ratio of 1:3 in 2x HEPES- buffered saline (HBS) (280 mM NaCl, 10 mM KCl, 1.5 mM Na2HPO4, 12 mM dextrose and

50 mM HEPES, pH 7.05) before addition of CaCl2 to final concentration of 125 mM and set to incubate for 30 min at room temperature. Plasmid DNA complexes were then transfected for 24h hrs. Lentivirus particles were harvested at 48 and 72 hrs post transfection and concentrated using Lenti-X concentrator (Clontech, USA).

2.2.3 Antibodies: The following primary antibodies were purchased as indicated: Mouse monoclonal against human α-synuclein (ab27766), CIM6PR (ab2733), SNX27 (C16) (ab77799), goat polyclonal anti-Vps35 (ab10099), rabbit polyclonal against Vps26A (ab23892) and GLUT1 (ab15309) were purchased from Abcam. Mouse monoclonal anti-

GFP (11814460001) and rabbit polyclonal anti-GFP (A-6455) were from Roche applied science and ThermoFisher Scientific, respectively; Mouse monoclonal anti-human EEA1

(610457), p230 (611280) and LAMP1 ( 555798) (BD Transduction Laboratories); Rabbit monoclonal anti Rab7a (D95F2) XP (#9367) (Cell signaling); Mouse monoclonal anti-β-tubulin (T9026) (Sigma Aldrich); Sheep polyclonal anti TGN-46 (AHP500GT) (AbD Serotec); Goat polyclonal anti-TBC1D5 (C-14) (sc-99661) (Santa Cruz Biotechnology); Rabbit polyclonal anti-cathepsin D (06-114) and anti-FAM21C (ABT79) (Merck Millipore). Anti-CD8 and Anti-Vps29 antibodies were 42

described previously [67]. Phalloidin-fluorescent conjugates (T-7471), were purchased from Molecular Probes (Life technologies). Secondary donkey anti-rabbit IgG Alexa Fluor488 (A21026), donkey anti-mouse IgG Alexa Fluor546 (A10036), goat anti-rabbit IgG Alexa Fluor546 (A11035), donkey anti-goat IgG Alexa Fluor647 (A21447) and goat anti-mouse IgG Alexa Fluor647 (57932A) were purchased from Life Technologies. Horse-radish peroxidase-conjugated swine anti-rabbit antibody was purchased from Dako. Dextran (MW 10,000 Da) conjugated to Alexa-647 was purchased from Life Technologies.

2.2.4 Cell Culture and Transfection: SH-SY5Y cells (Sigma-Aldrich) were maintained in Roswell Park Memorial Institute (RPMI) medium supplemented with 10% fetal bovine serum (FBS, Gibco) and 2 mM L-glutamine (Life Technologies) and cultured in a humidified 37°C incubator with 5% CO2. A431, HEK293 and HeLa cells were cultured in a humidified 37°C incubator with 5% CO2 and maintained in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% foetal bovine serum (FBS) and 2 mM L-glutamine (Life Technologies). Fibroblast cells were cultured in DMEM (Life Technologies) supplemented with 10% FBS and 2 mM L-glutamine. Cultures were maintained in T25 culture flasks and split at 1:4 throughout experimentation. Mammalian constructs were transfected into cells using LipofectAMINE 2000 (Life Technologies) according to manufacturer's instructions.

To generate stable knockdown cell lines, SH-SY5Y cells were transduced with lentivirus particles (MOI=15) in serum-free RPMI containing polybrene (8 µg/mL) at 37°C in 5% CO2 for 4 hrs when virus-containing medium was replaced by normal medium for additional 48 hrs. Transduced cells were then subjected to puromycin (Sigma Aldrich) selection (2µg/mL) to generate a stable knockdown population.

2.2.5 α-synuclein aggregation: SH-SY5Y cells grown on glass coverslips were incubated with 50mM KCl (Sigma) for 60 minutes at 37°C. Following the treatment, cells were either fixed (0 hour time point) or allowed to recover in complete growth medium for up to 48 hours, when they were fixed and immunofluorescence microscopy performed using anti-α- synuclein antibody.

43

2.2.6 KCl induced Membrane depolarization: Potassium Chloride (Sigma Aldrich) was dissolved in Dulbecco's phosphate-buffered saline (DPBS) to a final concentration of 500mM. Pre-warmed complete RPMI medium was supplemented with KCl solution to a final concentration of 50mM. SH-SY5Y cells were incubated with RPMI supplemented medium for 60 minutes to induce membrane depolarization (time 0 hrs). Cells were then incubated in complete media for a further 48 hrs.

2.2.7 Quantification of α-synuclein aggregation: SH-SY5Y cells were scored as α- synuclein aggregate positive on the basis of containing bright, concentrated α-synuclein immunofluorescence distinct from the diffuse surrounding cytoplasmic staining. Cells were excluded from counting if nuclear or cell morphology displayed evidence of condensing, fragmenting, rounding, blebbing or any other artifact. Inclusions were quantified by firstly counting the number of nuclei, as marker by the counterstain, DAPI, in a field of view and then subsequently counting the number of corresponding α-synuclein inclusions. Cells were excluded from the count if 50% or more of the cell staining was out of the field of view or overlapped with neighbouring cells. All results were tabulated in GraphPad software version 6 and presented as a percentage out of 100.

2.2.8 Immunoprecipitation: Cell monolayers were washed in ice-cold phosphate buffered saline (PBS) and lysed in 50 mM HEPES, 150 mM NaCl, 1% Triton X-100, 10 mM Na4P2O7,

30 mM NaF, 2 mM Na3VO4, 10 mM EDTA, 0.5 mM AEBSF and Complete Mini protease inhibitor cocktail for 10 min on ice. Cell lysates were centrifuged at 17,000 x g and supernatant incubated with GFP-NanoTrap beads [283] for 2 h at 4oC under constant rotation. Complexes were removed from the GFP-NanoTrap beads by boiling after three consecutive washes with TK lysis buffer.

2.2.9 Microsomal Fractionation: HeLa cells seeded in 15cm2 culture dishes were washed twice on ice in PBS, collected and homogenized with 20 passages through a 22G ¾ needle in buffer containing 20 mM HEPES, pH 7.4, 250 mM sucrose, 1mM EDTA, 2mM Na3VO4, 10mM EDTA and 0.5mM AEBSF. Lysates were spun at 500xg for 5 min, and a portion of the resulting supernatant was collected and stored as the whole cell lysate. The resulting supernatant was further centrifuged at 17,200xg for 20 min to isolate the crude plasma membrane fraction followed by 175,000xg for 75 min to isolate the microsome fraction from the remaining cytosolic fraction. Pelleted microsomes were re-suspended in 44 homogenization buffer (20 mM HEPES, pH 7.4, 250 mM sucrose, 1mM EDTA , 2mM

Na3VO4, 10mM EDTA and 0.5mM AEBSF).

2.2.10 Western Blotting: Protein lysates were resolved on SDS-PAGE and transferred onto a PVDF membrane (Immobilon-P and Immobilon-FL; Millipore) according to the manufacturer's instructions. Western blotting using ECL and Odyssey infrared imaging system (LI-COR Biosciences) were performed as described previously [67].

2.2.11 Antibody uptake assay: HeLa cells seeded on glass coverslips were co-transfected with the previously described GFP/CD8-CI-M6PR fusion constructs for 16hrs. Transfected cells were serum starved in DMEM supplemented with 2 mM L-glutamine for 4 hours prior to being chased at 37°C with anti-CD8 antibody. Post chase coverslips were fixed in 4% PFA and subjected to indirect immunofluorescence.

2.2.12 Cell surface biotinylation: 16 -18 hrs post transfection, transfected HeLa cells in 15cm2 cell culture plates were washed 3 times of ice-cold PBS, followed by incubation with 0.25 mg/mL of EZ-link Sulfo-NHS-LC-LC Biotin (Thermo Fisher Scientific) for 10 min on ice, when reactions were quenched by 50 mM Glycine for 10 min. Cells were harvested in TK lysis buffer and protein concentrations were determined by BCA protein quantitation assay. Equal amounts of protein lysates were incubated with Neutroavidine Agarose beads (Thermo Fisher Scientific) for 1 hour at 4oC. After washing the beads 3 times with TK lysis buffer, beads were boiled in 2 x SDS sample loading dye, and pull-down samples were subjected to SDS-PAGE/immunoblotting.

2.2.13 Glucose Uptake Assay: Transfected HeLa cells in a 12-well plated were washed with warmed Kreb-Ringer’s Phosphate (KRP) buffer containing 10 mM HEPES, pH 7.4, 136 mM NaCl, 4.7 mM KCl, 1.25 mM CaCl2, 1.25 mM MgSO4, 0.6 mM Na2HPO4, 0.4 mM

NaH2PO4. Cell monolayers were incubated for 10 min in the presence of 50 µM 2-deoxy-D- glucose and 1 µCi/mL 2-deoxy-H3-glucose. The amount of glucose transported into the cells was determined by harvesting the cells in 1% Triton X-100 and the associated radioactivity quantified using a MicroBeta liquid scintillation counter (Perkin Elmer Biosciences). The counts per min (cpm) were normalized to total protein concentrations.

45

2.2.14 Patient Data: An isolated skin biopsy was taken from a 73 y/o European Male Parkinson’s disease patient previously genotyped for the Vps35 variant, D620N. The patient had a diagnosis of Parkinson’s disease according to the UK Brain Bank clinical criteria [284] and a self-reported age of symptom onset of 46 years. The single heterozygote point mutation in Exon 15 of the Vps35 gene leading to the D620N amino acid change was confirmed by Sanger sequencing of DNA isolated from the cell line. Control human fibroblasts were isolated from a healthy, neurologically normal 44 year old male subject on no medication and with no family history of PD.

2.2.15 Establishment of Skin Biopsy Cultures: Isolated biopsy samples were placed in 50 mL conical flasks filled with cell culture medium (DMEM: F12 supplemented with 10% FBS and stored on ice until tissue dissociation. Samples were kept on ice for no longer than 24 h. Cell culture medium was aspirated and biopsy samples were washed a total of three times with 10 mL of room temperature PBS. Tissue were transferred into 6 cm sterile culture dishes containing pre-warmed culture medium and divided into pinhead-sized explants using a scalpel blade. Using a 1 mL tuberculin syringe with a 23-G needle attached, one pinhead sized explant was placed into a 25 cm flask. Culture flasks were then placed in a

37 ºC, 5% CO2 humidified circulating incubator for 5-10 min allowing sufficient time for the explants to attach to the flask before adding pre-warmed culture medium. 5 mL of culture medium was added to the adhered cultures and returned to the 37 ºC, 5% CO2 incubator. Cultures were left to grow for a minimum of 72 h before being handled. On the 4th day, cultures were checked and those who displayed evidence of radial outgrowth from initial explants were supplemented with fresh medium and left for another 14-21 days, with periodic medium changes, before subculturing.

2.2.16 Microscopy of live cells: A431 cells were seeded in 35 mm glass bottom dishes (MatTek Corporation) 48 h prior to use and transfected with indicated GFP fusion Vps35 constructs. Dextran loading was performed as previously described [285]. Briefly, dextran conjugated to Alexa-647 (MW 10,000 Da) (Life Technologies) at a final concentration of 100 µg/mL was loaded into the A431 cells by incubation at 37°C for 1 h in complete media. Cells were then washed to remove excess dextran and imaged in CO2-independent media supplemented with 10% FBS (Life Technologies). Time-lapse microscopy was performed by capturing 1 Airy Unit (1AU) z-slices using 63x objective on Zeiss LSM 710 FCS Inverted 46

Scanning Laser confocal microscope. Movies were edited using Image J 1.47f and still capture frames from the movies were edited using Adobe Photoshop.

2.2.17 Immunofluorescence: A431, HeLa or SH-SY5Y cells seeded on coverslips were transiently transfected with mammalian constructs, fixed and stained with indicated antibodies as described previously [67]. Coverslips were mounted using Fluorescent Mounting Medium (Dako) and imaging was performed using 63x objective on a Ziess LSM 710 Upright Scanning Laser confocal fluorescent microscope. Images represent a 1AU z- plane single slice. All images were analyzed using Zeiss LSM 5.0 and Adobe Photoshop software.

2.2.18 Nocadazole Treatment: Cell monolayers were treated with 2 µM Nocadazole

(Sigma-Aldrich) for 60 min in a humidified 37°C incubator with 5% CO2. Medium was aspirated and cells were washed three times in PBS prior to PFA fixation.

2.2.19 Secretion Assay: HEK293 or HeLa cells were plated in 6-well dishes 48 h prior to use and transfected with GFP fusion constructs for 24 h. Cyclohexamide (final concentration 100 µg/mL, Sigma Aldrich) was diluted in serum-free DMEM supplemented with 2 mM L- Glutamine (Life Technologies) and used throughout experimentation. Medium samples were collected at indicated times and analysed using Western blotting.

2.2.20 Quantification of the intracellular distribution of endosomes: To quantify perinuclear/cytoplasmic localization, a macro was created in ImageJ* (version 1.47i). Two channels were utilized from each image: a DAPI channel to define the nuclear regions; and the protein of interest (POI) channel. A typical image contains 5 or 6 nuclei in the field of view. First, the nuclear image was converted to a binary mask to define the nuclear regions using an auto-threshold algorithm. Nuclei touching the edge of the image were excluded from further analysis. A perinuclear region around the nuclei was then defined as follows. The nuclear mask regions selected were expanded by a distance of 40 pixels (corresponding to a distance of 1.5 µm). Removing the nuclear region from this expanded region selection then gave an annular region around, but not including, each nucleus. The average intensity in the POI channel of all the perinuclear annuli regions was then recorded. To select a region more distant from the nuclei, a similar process was used: the nuclear mask regions were expanded by 80 pixels (corresponding to 3.0 µm) and a region obtained 47 by expanding the nuclear regions by 40 pixels subtracted from it. This was used to generate an area adjacent to the perinuclear annuli but more distant from the nuclei, i.e. a cytoplasmic region. In the POI channel, the average intensity across the union of these 'cytoplasmic' regions was then recorded for the image. Finally, for comparison between images and experiments, the ratio of the so calculated average perinuclear intensity to the average cytoplasmic region intensity was recorded. The script was fully automated with the only user interaction required being the selection of the directory containing the images to be quantified. The macro is available in supplementary material (Supplementary Figure 1).

2.2.21 Colocalization analysis: Immunoflourescence images (1AU slices) were captured using a Ziess LSM 710 Upright Scanning Laser confocal fluorescent microscope under 63x magnification using identical laser power, light pathways and band passes. Captured images were analyzed using Image J Version J 1.47f. Multi-channel images were split into corresponding grey-scale format and threshold settings were applied to each individual channel. Under these conditions colocalization was quantified using the Image J plug-in, Colocalization finder (U.S. National Institutes of Health, Bethesda, MD, USA [286]. http://rsb.info.nih.gov/ij/plugins/colocalization-finder.html. Raw data representing colocalization co-efficient was analyzed using GraphPad Prism 5 software version 5.03. Graphs represent a global average of data collected.

48

Chapter 3: Disruption of mammalian retromer complex induces α-synuclein aggregation

Introduction

Parkinson’s disease (PD) primarily affects the dopaminergic neurons of the substantia nigra pars compacta; a small anatomical region located within the midbrain that regulates movement. On a cellular level, PD is a complex disease characterized by the subcellular formation of large, perinuclear accumulations of insoluble proteins, called Lewy Bodies (LBs). The most abundant protein found in LBs is the pre-synaptic plasma membrane-associated protein called α-synuclein, but may also contain other proteins and lipids [182, 222, 287]. The formation of LBs, which individual neurons are unable to clear, has been directly linked to activation of apoptotic pathways in the dopaminergic neurons [182].

Recently, several point mutations in the protein sorting complex, retromer, were implicated in late onset familial PD. Retromer is highly conserved protein sorting complex needed for the retrieval of endosomal receptors and trafficking of endo-lysosomal proteases, such as cathepsin D, which is needed for degradation of α-synuclein. For example, retromer is responsible for the trafficking and sub-cellular localization of the CI-M6PR [67, 70]. In the absence of retromer, the lysosomal enzyme and ligand of the CI-M6PR, cathepsin D, is retained in the TGN and eventually, secreted from the cell [102] and the degradation of toxic protein species, such as α-synuclein, may be impaired.

Here, perturbations in retromer by RNAi led to the accumulation of α-synuclein in the late endosomal network in SH-SY5Y cells. Consistent with previous reports, α-synuclein inclusions demonstrated immunoreactivity for the small GTPase and late endosomal maker, Rab7a, but not Rab5, supporting accumulation within the late endosomes. Unexpectedly, inclusions positive for α-synuclein also demonstrated a strong immunoreactivity for retromer subunit, Vps35, an observation not currently reported. Additionally, the accumulation of 49 endosomal localized α-synuclein in shRNA Vps35 cells, and to a lesser extent in the non- silencing cells, disrupted the localization and processing of the Cathepsin D, an aspartic protease and ligand of CI-M6PR, a retromer-dependent cargo. Overall, the significance of the work presented in this chapter highlights the importance of the retromer complex in maintaining the homeostasis of the endo-lysosomal pathway, one of the identified routes needed for degradation of α-synuclein. These results provide insight into the relationship between endo-lysosomal homeostasis, protein degradation, the retromer complex and their roles in PD.

50

3.1 Results 3.1.1 Loss of the retromer complex induces α-synuclein inclusions α-synuclein aggregates are bright, often spherical structures, typically found juxtaposed to the nucleus in cultured cells that vary widely in size [210, 288]. To investigate the relationship between the retromer and α-synuclein, SH-SY5Y cells with stable knockdown for retromer subunit Vps35 were generated (see section 2.2.2). Currently, the literature supports a role for retromer in trafficking of mitochondrial and also late endosome/lysosomal proteins. Common techniques used to induce α-synuclein positive inclusions include Rotenone, a mitochondrial targeting reagent, or Bafilomycin to block fusion between the lysosome and autophagosome [289]. As such, long term use of these reagents may result in global changes to the endo-lysosomal membranes within a cell and thus alter the function of the retromer complex. Use of KCl as a method to induce the formation of α-synuclein positive inclusions is previously published [210, 290, 291] and was chosen as a minimally invasive technique that mimics the physiological process of membrane depolarization in neuronal cells that results in the transient influx of calcium.

Non-silencing and shVps35 SH-SY5Y cells grown on coverslips were subjected to 60 minutes KCl treatment, fixed at 0hrs or left to recover for 48 hrs in recovery medium, and analyzed by indirect immunofluorescence using antibodies against α-synuclein and retromer subunit, Vps35 (see section 2.2.3 & 2.2.17). Unexpectedly, retromer subunit, Vps35, was found to be associated with the perimeter of the α-synuclein inclusion (Figure 3.1B), supporting the possibility of α-synuclein being membrane enclosed. Moreover, it was observed that following stable depletion of retromer (Figure 3.1A) an average of 28.3% (+/- 4.3) of cells were scored positive for α-synuclein inclusions, in contrast to only 8% (+/- 2.8) of non-silencing control cells (Figure 3.1B & 3.11C). 48hrs after KCl treatment, 25% (+/- 3.5) of the non-silencing and 44.6% (+/- 5.2) of shVps35 cells displayed α-synuclein inclusions (Figure 3.1B & 3.1C), supporting the dependence of retromer in the clearance of α-synuclein via the lysosomal pathway.

Interestingly, a size difference of α-synuclein inclusions was also observed between non-silencing and knockdown populations 48hrs after membrane depolarization was induced (Figure 3.1B and 3.1D). Surprisingly, minimal difference was observed between inclusion size in non-silencing ( = 0.71µM) and knockdown ( = 0.91µM) cells at basal 51 level, or after treatment with PBS control ( = 0.76µM; knockdown ( = 0.81µM); Figure 3.1D). However, following treatment of non-silencing cells with KCl, a noticeable difference in inclusion size ( = 1.45µM) was observed when compared to non-silencing cells treated with PBS for 48 hrs. This trend was also witnessed in shVps35 cells 48hrs after treatment with KCl, with inclusion appearing even larger than that of a control ( = 2.48µM; Figure 3.1D).

52

Figure 3.1: Loss of retromer induces α-synuclein inclusions (A) Western blot analysis of knockdown efficiency in SH-SY5Y cells stably expressing non- silencing or silencing shRNA against Vps35. Membranes were probed with anitbodies against Vps35 and β-tubulin. (B) Representative confocal images of SH-SY5Y cells transduced with non-silencing or silencing shRNA against retromer subunit, Vps35, immunolabeled with antibodies against α-synuclein (green) and Vps35 (red) following treatment with 50mM KCl for 60 minutes and fixed immediately or left to recover for 48hrs. Images are representative of three independent experiments with 10 images captured experiment for each time point and condition per cell line. (C) Quantitation of non-silencing and shRNA Vps35 cells bearing α-synuclein inclusions at 0 and 48hrs post depolarization. Graph represents two independent experiments with 10-15 images per experiment captured in a Z-stack with 10-20 cells per field of view, for each stated time point and respective treatment. (Error bars displayed as +/- SEM; *p<0.05;**p<0.01; ***p<0.001; ****p<0.0001 by ANOVA followed by Bonferroni multiple comparisons test). (D) Average aggregate size (diameter) observed in control or knockdown cells at 0 and 48hrs post treatment with 50mM KCl or PBS. Error bars displayed as +/- SEM; *p<0.05;**p<0.01; ***p<0.001 by ANOVA followed by Bonferroni multiple comparisons test

3.1.2 α-synuclein inclusions are Rab7 positive Previous reports demonstrate localization of α-synuclein to the late endosome as a pathway of proteasome-independent degradation [292]. As such, it was investigated whether the loss of retromer influenced accumulation of α-synuclein in the late endosomal system. To do so, fixed cells were immunolabeled with antibodies against the small GTPases Rab7a and Rab5 to mark the late and early endosomes, respectively.

In non-silencing cells fixed 0hrs post KCl treatment, immunolabeling of Rab7a demonstrated localization to discrete puncta structures throughout the cell, which were negative for α-synuclein (Figure 3.2A). In contrast, shVps35 cells showed the accumulation of α-synuclein in the lumen of Rab7a positive vesicles (Figure 3.2A). Consistent with this observation and the literature, α-synuclein inclusions observed in non-silencing and shVps35 cells 48hrs after KCl treatment were also positive for the late endosomal marker Rab7a.

Furthermore, it was observed that the α-synuclein inclusions in shVps35 cells displayed stronger Rab7a immunoreactivity than inclusions seen in non-silencing cells concurrent to a simultaneous loss of Rab7a puncta structures from the cytoplasm (Figure 3.2A). In contrast to these findings, α-synuclein inclusions observed in non-silencing and shVps35 cells at 48hrs post treatment with KCl were devoid of Rab5 immunoreactivity and the production of α-synuclein inclusions did not influence the overall sub-cellular localization 53

of Rab5-positive puncta (Figure 3.2B), supporting accumulation of α-synuclein in the late endocytic network following depletion of retromer.

54

Figure 3.2: α-synuclein inclusions reside in the Rab7A positive compartments in cells depleted of retromer (A) Representative confocal images of SH-SY5Y cells transduced with non-silencing or silencing shRNA against Vps35 immunolabeled with antibodies against α-synuclein (red) and Rab7a (green) or (B) with antibodies against Rab5 (green) at time points described in Figure 3.1. Images are representative of three independent experiments with 10 images captured experimental condition. Scale bar: 5µm.

3.1.3 Impaired Localization and processing of Cathepsin D leads to generation of α- synuclein inclusions Previous reports show recruitment of the retromer complex to the cytosolic face of the endosome is mediated by the small GTPase Rab7A [18]. Furthermore, both Rab7a and the retromer complex have been implicated in the correct localization and processing of cathepsin D into its mature state [70, 102]. As α-synuclein is a known substrate for cathepsin D [244, 281, 282] and both retromer and Rab7a were found to localize to aggregated structures, the localization and processing of cathepsin D was investigated. Non-silencing cells treated with KCl at 0 hours were fixed and immunolabeled with antibodies against total endogenous Cathepsin D demonstrated strong localization of the enzyme to the endosomal puncta and to a lesser extent, membranes consistent with the Golgi apparatus, both of which were negative for α-synuclein (Figure 3.3A). In contrast, loss of retromer expression resulted in observable redistribution of Cathepsin D from dispersed puncta to the perinuclear space, reminiscent of the Golgi apparatus, consistent with impaired trafficking. However, the population of shVps35 cells positive for α-synuclein inclusions at 0hrs post KCl treatment demonstrated some evidence of Cathepsin D localization to the inclusions. Interestingly, in non-silencing cells at 48hrs post KCl treatment, distinct localization of cathepsin D to the α- synuclein inclusion site was observed alongside a modest accumulation at the perinuclear space when compared to 0hrs post treatment. In contrast, α-synuclein inclusions observed in shVps35 cells 48hrs post KCL treatment displayed no evidence of Cathepsin D immunoreactivity and a strong accumulation at the perinuclear space (Figure 3.3A), suggestive of impaired trafficking or processing from the TGN.

In order to assay the change in Cathepsin D processing, non-silencing and shVps35 cells treated with KCl or PBS were harvested at 0, 9, 24 and 48hrs post treatment and subjected to Western blotting (Figure 3.3B). Using antibodies against cathepsin D, non- silencing and shVps35 cells treated with PBS demonstrated no decrease in the levels of 55 mature cathepsin D at later stages of treatment when compared to the levels observed at 0 hrs. In contrast, a slight reduction in the total level of mature Cathepsin D protein at 48hrs post-treatment with KCl was observed in non-silencing cells, suggesting impaired delivery or processing within the late endosomal system.

56

Figure 3.3: Disruption in cathepsin D processing leads to accumulation of α- synuclein inclusions. (A) Representative confocal images of SH-SY5Y cells transduced with non-silencing or silencing shRNA against retromer subunit, Vps35, immunolabeled with antibodies against α-synuclein (red) or Cathepsin D (green) at listed time points. Images are representative of three independent experiments with 10 images captured experiment for each time point and condition per cell line. Scale bar: 5µm. (B) Representative Western blots of stably transduced SH-SY5Y lysates collected at 0, 9, 24 and 48hrs post 50mM KCl treatment. Membranes were probed with antibodie against Cathespin D and β-tubulin. Blots are representative of three independent experiments.

While the total amount of Cathepsin D showed an overall decrease in shVps35 cells when compared to that of NS cells, the amount of the mature Cathepsin D was visibly lower at 48 hours than that observed at 0, 9 and 24hrs post KCl treatment (Figure 3.3B). Taken together, these findings indicate that the loss of retromer interferes with the processing and localization of immature cathepsin D into its mature form and may consequently lead to the production of α-synuclein inclusions.

3.1.4 Presence of α-synuclein inclusions do not influence levels of retromer Next, Western blotting was employed to investigate changes in the total levels of specific proteins following KCl treatment. Selection of proteins subjected to Western blot investigation was based on following criteria – protein is a known component of the retromer complex (Vps35), retromer interacting protein (TBC1D5 and Rab7) [18, 102] or a marker of the endocytic network (Beclin-1, Rab5, Rab7 and LAMP1) (Figure 3.4).

Total levels of Vps35, TBC1D5, Rab5 and Rab7A showed no change following the production of α-synuclein inclusions in both non-silencing and shVps35 cells at 0 and 48hrs post treatment both in presence of absence of 50mM KCl treatment. Consistent with previous reports [293], the stable loss of retromer in the cell system used here also resulted in an increased turnover of TBC1D5 (figure 3.4). In contrast, autophagy-linked protein Beclin1 was found to increase 48hrs after treatment with 50mM KCl in non-silencing and shVps35 cells when compared to levels observed at 0hrs time point, whereas changes in LAMP1 levels were only observed in shVps35 cells.

57

Figure 3.4: Western blot analysis of endosomal machinery during aggregation time course (A) SH-SY5Y cells stably transduced with non-silencing or shRNA against Vps35 were collected at 0 and 48hrs post treatment with 50mM KCl or PBS and subjected to Western blot analysis (n=2). Membranes were probed with listed antibodies and are representative of two individual experiments.

58

3.2 Discussion This chapter has examined the relationship between the endosomal trafficking complex, retromer, and the α-synuclein inclusion. Following the stable depletion of the retromer complex in SH-SY5Y cells, α-synuclein positive inclusions were present in approximately 28% of cells in contrast to only 8% in non-silencing cells. The use of KCl to induce calcium influx led to an increase in both size and frequency of α-synuclein inclusions in shVps35 cells when compared to non-silencing control cells. Interestingly, the increase in both non-silencing and shVps35 cells was approximately 2-fold higher than the basal level 48hrs after KCl. Moreover, the α-synuclein inclusions were observed to localize in compartments positive for Rab7a and Vps35 positive, inferring that the inclusion may reside in endocytic membranes and appeared to further impair the trafficking of the lysosomal protease, cathepsin D. Overall, the use of SH-SY5Y cell line with a stable depletion of the retromer complex has provided understanding into the relationship between the endosomal network, retromer-mediated receptor trafficking and α-synuclein inclusion formation in the current chapter.

3.2.1 Retromer associates with the α-synuclein inclusions To date, little evidence suggests that retromer or its subunits are likely to associate with α-synuclein inclusions. Interestingly however, a study by Xia and colleagues identified retromer subunit Vps35, along with other known proteins (Ubiquitin, α-synuclein, Heat Shock Protein-70), as a component of the LB isolated from PD post-mortem tissue [294]. Although this does not provide evidence of a physical protein-protein interaction between Vps35 and α-synuclein, it does support the notion that retromer is present in LBs. Consistent with these findings, evidence presented here observes immunoreactivity for Vps35 within the α- synuclein inclusion whereas diffuse, cytosolic α-synuclein displayed no evidence of Vps35 immunoreactivity (see Figure 3.1B). Furthermore, recruitment of retromer to the endosome is mediated by the small GTPase Rab7A [18, 102], which has also been previously observed to be associated with structures encompassing α-synuclein inclusions [292, 295] and is additionally reported in this thesis. However, further experimentation and use of a range of methodologies, including GST-pull down, co-immunoprecipitations, knockdown of Rab7A and electron microscopy, is needed to ascertain if retromer has a direct association with aggregated α-synuclein and provide insight into any Rab7-dependent association of retromer with the inclusion. Moreover, to determine if the loss of retromer expression plays 59 a role in the pathogenesis of PD via induction of toxic LB formation, antibodies raised against the phosphorylated form of α-synuclein would also provide great insight and strengthen the link of retromer as casual role in PD.

Although there is no current rationale for localization of retromer to the α-synuclein inclusion, it is not likely the observed staining is an artifact (see Figure 3.1B). Characterizing the sub-cellular localization of the α-synuclein positive inclusions using antibodies against the small GTPase Rab5 to mark the early endosome was employed. It was observed that Rab5 positive compartments were to be spatially distinct from structures positive for α- synuclein (Figure 3.2B). If the production of α-synuclein inclusions, or treatment with KCl, was leading to the exposure of N-terminal α-synuclein residues previously embedded in the plasma membrane that all antibodies used in this study were cross reacting with, one would assume that all antibodies would give a strong cross-reaction and, therefore, localization to the inclusions. Consistent with the observations reported for Rab5, α-synuclein positive inclusions observed in the shVps35 cells are devoid of Cathepsin D immunostaining (Figure 3.3A) inferring that the localization of endogenous retromer with α-synuclein is authentic. Taken as a whole, the findings presented in this thesis demonstrate that the localization of retromer subunit, Vps35, to α-synuclein positive inclusions in the SH-SY5Y cell model further support and implicate retromer in PD biology in human tissue.

3.2.2 Loss of Cathepsin D impairs the degradation of α-synuclein inclusions In order to maintain the flow of Cathepsin D from the TGN to the early/late endosome, retromer mediates the sorting of the CI-M6PR [102]. Following endocytosis, retromer engages the CI-M6PR at the endosome and retrieves it to the TGN. Once delivered to the TGN, the CI-M6PR engages its ligand, Cathepsin D, and facilitates its delivery to the late endosomal network where it undergoes processing to yield an active (mature) enzyme [245, 246], while CI-M6PR is trafficked back to the TGN by engaging the retromer complex. Therefore, following the loss of retromer the CI-M6PR is retained in endosomes and eventually degraded via normal lysosomal maturation pathway [102]. Consequently, the level of Golgi-localized cathepsin D heavily outweighs that of the lysosome-localized cathepsin D, a direct result of impaired delivery of CI-M6PR to the TGN. In turn, this results in a distinct decline in the delivery of cathepsin D to the late endocytic network, decreased processing of the enzyme into the mature form, secretion of pre-pro-Cathepsin D, and ultimately, increased buildup of non-degraded material within the late endosome/lysosome. 60

In conjunction with this retromer-mediated process, several other retromer interacting proteins have also been shown to regulate the sorting of the CI-M6PR to the TGN.

Two direct consequences observed here following the loss of retromer are the decrease in levels of mature Cathepsin D and the consistent accumulation of α-synuclein positive inclusions in late endocytic compartments (see Figure 3.1, 3.2 & 3.3). As α- synuclein is a substrate for the lysosomal protease cathepsin D [244], the phenotype observed here in the absence of retromer may be a direct result from the impaired trafficking of receptors and lysosomal proteases, and not a result of proteasome dysregulation. Explanation for this observed phenotype lies in the impaired transport of cathepsin D from the TGN to the endocytic network. Support for this hypothesis lies in the observation that following the loss of retromer, cathepsin D was found clustered at the perinuclear space and not in dispersed punctate structures, as witnessed in non-silencing cells (see Figure 3.3). Moreover, the production of α-synuclein inclusions as a result of retromer deficiency displayed minimal evidence of cathepsin D immunoreactivity when compared to non- silencing cells, consistent with their localization to the late endosome. Lastly, levels of mature Cathepsin D in shVps35 cells are markedly decreased, compared with non-silencing controls (Figure 3.4), likely underpinning the accumulation of α-synuclein. Indeed, the importance of Cathepsin D and α-synuclein aggregate formation is well established by the literature as cathepsin D (-/-) knockout mice present with significant amounts of insoluble high molecular weight α-synuclein when compared to wild type littermates [281]. Together, data presented here infers that the loss or incorrect localization of mature cathepsin D in the absence of retromer leads to the generation of α-synuclein inclusions in the late endosomal system.

Moreover, siRNA mediated suppression of retromer interacting protein, Rab7A, has previously been shown to decrease the retrieval of CI-M6PR to the TGN and subsequently impair the delivery of Cathepsin D to the lysosome [102]. However, Western blot analysis of Rab7A protein levels 48hrs after KCl treatment were identical to those witnessed at both 0hrs post-depolarization and also 48hrs after PBS control treatment. Consistent with the literature, Rab7A was observed on structures enclosing inclusions [296], suggesting that the production of α-synuclein inclusions following the loss of retromer does not influence the levels of interacting machinery needed for efficient sorting of receptors, such as CI-M6PR, 61 but may induce changes in the sub-cellular localization, and function, of particular proteins. Although there is no current data to suggest that loss of retromer interfers with the localization/function of Rab7a, it is possible that the redistribution of sorting machinery may mimic a loss of function/knockdown system as witnessed with the heavy localization of Rab7A and Vps35 to the α-synuclein inclusion second to a loss of retromer (see Figure 3.1, 3.2 & 3.3).

Taken together, it is clear that retromer plays a vital role in the sorting of receptors and their ligands that are required to clear lysosomal proteins, such as α-synuclein. This is evident by the formation of endo-lysosomal α-synuclein protein inclusions forming following the loss of retromer that further disrupt the sub-cellular localization and likely function, of additional receptor sorting machinery.

3.3.3 Retromer, TBC1D5 and clearance of α-synuclein While retromer plays a fundamental role in the recycling of receptors and turn-over of α-synuclein, several of its interacting proteins are required for localization and function of retromer itself. For example, the RabGAP TBC1D5 is needed to disassociate retromer from the endosomal membrane following engagement of cargo [18]. Interestingly, TBC1D5 also interacts with autophagy regulator protein LC3, and appears to positively regulate autophagic flux in cell culture systems, suggesting it may act as molecular switch between cargo recycling and degradation [293, 297]. However, this does not appear to be true for retromer, as RNAi against the Vps35 subunit has minimal impact on the autophagic flux [270]. Importantly, it is well established that the autophagic system plays an important role in the clearance of α-synuclein, but whether the autophagy pathway can clear aggregated α-synuclein is still controversial [298]. Recent reports demonstrate strong recruitment of autophagic markers (p62, LC3, Beclin1) to aggregated α-synuclein and significant increases in total protein levels when compared to controls in cell culture systems and murine DA neurons at 3 weeks post midbrain injection with α-synuclein [298, 299], indicating the ability for the cell to recognize and assemble autophagic machinery when increased amounts of α-synuclein are present.

Interestingly, as observed in Figure 3.4, and consistent with previous reports [293], the loss of retromer negatively impacts the total protein level of TBC1D5. A loss of TBC1D5 witnessed following stable depletion of retromer subunit Vps35 negatively influences the 62 level of autophagy within the system and ultimately, may contribute to the inability of the cell to clear the α-synuclein inclusion. Given that increased α-synuclein inclusions were observed at basal conditions in shVps35 cells, the loss of TBC1D5 likely contributes to the cells’ inability to clear these inclusions via dysregulation of the autophagy pathway. In conjunction with this, given that following the loss of retromer α-synuclein accumulates in the late endosomal system as non-degraded protein species, this would impede fusion between the lysosome and the autophagosome, and further delay the cells’ ability to clear the aggregated protein.

Exaggerated consequences of an inability to turnover lysosomal compartments by fusion with the autophagosome is observed here by the increase in LAMP1 levels, a marker of lysosomes, and Beclin-1, an initiator of autophagy, 48hrs after treatment with KCl. An increase in LAMP1 in shVps35 cells would suggest a cumulative impact of lysosomal dysfunction arising from impaired cathepsin D trafficking and subsequent increased pre- disposition to form large, insoluble α-synuclein inclusions alongside inhibited autophagic flux. This is in contrast to non-silencing cells which have the capacity to clear α-synuclein product, as the delivery of Cathepsin D to the late endosome appears uninterrupted (see Figure 3.3) and machinery needed for induction of autophagy is not impaired (see Figure 3.4). However, it must be noted that these observations were only witnessed 48hrs after KCl treatment and not at 0hr time points, inferring that this phenotype may also arise partially from cellular stress induced by the treatment itself. The use of KCl to induce membrane depolarization and calcium influx promotes the formation of α-synuclein inclusions through engagement of the α-synuclein C-terminal calcium-binding domain [195, 196]. Secondary to this, it is well established that increased free cytosolic calcium can activate pro-apoptotic pathways [300, 301]. However, a marked increase in Beclin-1 and no change in LAMP1 levels in non-silencing cells 48hrs after KCl may signify a response to promote cell viability through increasing autophagy, and thus clearing LAMP1 positive compartments, a mechanism that may be compromised in shVps35 cells as witnessed by the lower levels of Beclin-1 and marked increase in LAMP1.

Collectively, it is clear that the loss of both retromer and TBC1D5 negatively influence different pathways that block the ability for accumulated α-synuclein to be cleared by the lysosome. As observed in this chapter, the stable depletion of retromer leads to a decrease in levels of endogenous TBC1D5, and approximately 28% of cells were found positive for α- 63 synuclein inclusions when compared to that of non-silencing control cells. Primarily it can be hypothesized that the removal of retromer impairs the sorting of the CI-M6PR to the TGN and subsequently decreases the delivery of cathepsin D the lysosome, which in turn promotes the formation of α-synuclein inclusions. However, secondary to this, loss of TBC1D5 and lysosomal dysfunction also lead to the observed increase in autophagy markers that compromises overall cell health, likely balancing between autophagy and apoptosis. As the number of mutations identified in endocytic machinery that are linked to PD is increasing, thorough understanding of the relationship between endocytic sorting, the auto-lysosomal pathway, α-synuclein aggregation and PD manifestation is needed to establish the key components that are causal in dysregulation of this pathway.

64

Chapter 4: The Vps35 D620N mutation linked to Parkinson’s disease disrupts the cargo sorting function of retromer

Introduction

Mammalian retromer is a protein complex composed of three proteins Vps26, Vps29 and Vps35, with an important role in the sorting and trafficking of transmembrane receptors within the endosome. The retromer forms a stable trimer, which associates with a range of other proteins, including sorting nexins (SNXs), which modulate its function (reviewed in [66]. Sorting of cargo from endosomes to the TGN is a tightly organized process, and retromer is a key player in mediating trafficking of several receptors on this route, including CI-M6PR [71]. Interaction between retromer and CI-M6PR is mediated through a tripeptide sequence within the C-terminal cytoplasmic domain of CI-M6PR [71] and amino acids 500- 693 of the Vps35 subunit [70]. Disruption of the retromer-cargo interaction, either by downregulation of retromer levels [70] or modification of cargo cytoplasmic domains [71] leads to accumulation of receptors in early endosomes and ultimately to their degradation [67, 102]. Within the TGN, CI-M6PR binds enzymes such as newly synthesized cathepsin D, and delivers them to the late endosome network where they undergo processing to yield enzymatically active forms responsible for protein degradation within the lysosome. Upon dissociation of CI-M6PR from its ligand, CI-M6PR is trafficked back to the TGN by interacting with the retromer complex within the endosomal membrane [70, 112]. Any failure within this CI-M6PR cycling has been shown to result in increased turnover of the receptor as well as improper intracellular processing and secretion of the cathepsin D into the surrounding medium [67, 70, 102]. Successful delivery of cathepsin D to the endosomal network results in processing of the “pro” form into a mature, active form of this enzyme [246]. As such, the availability of CI-M6PR at the TGN, via its association with retromer, is crucial for sorting of cathepsin D from TGN to endosomes.

A number of recent reports determined a link between several point mutations of Vps35, a retromer subunit, and manifestation of late-onset Parkinson’s disease [256-258]. 65

Reports implicating Vps35 in PD demonstrate that a single point mutation of highly conserved amino acid 620 in Swiss, Austrian and German families leads to an autosomal dominant, high penetrance mode of PD inheritance [302]. Using direct sequencing of samples from PD patients, two individual groups further confirmed the presence and same inheritance mode of the D620N variant in French and Japanese ethnic groups [257, 258]. A recent large multi-center study used more than 15,000 subjects worldwide to screen for all known PD variants and confirmed the presence of D620N in 5 familial cases and two seemingly sporadic cases, emphasizing the importance of this particular point mutation in PD patient cohorts worldwide [303]. Collectively, genetic evidence suggests that the pathogenic D620N Vps35 variant is a rare cause of familial as well as idiopathic forms of PD, and points to endosomal trafficking as a critical process in the disease.

To understand the underlying molecular mechanism of the Parkinson’s disease causing Vps35 D620N mutation we examined if an established retromer function, namely regulation of CI-M6PR trafficking, was disrupted by the expression of this variant. Here, we investigated details underlying the molecular mechanism by which the D620N mutation in Vps35 modulates retromer function, including examination of retromer’s subcellular localization and its capacity to sort cargo. We show that expression of the PD-linked Vps35 D620N mutant redistributes retromer-positive endosomes to a perinuclear subcellular localization and that these endosomes are enlarged in both model cell lines and fibroblasts isolated from a PD patient. Vps35 D620N is correctly folded and binds Vps29 and Vps26A with the same affinity as wild-type Vps35. While PD-linked point mutant Vps35 D620N interacts with the cation-independent mannose-6-phosphate receptor (CI-M6PR), a known retromer cargo, we find that its expression disrupts the trafficking of cathepsin D, a CI-M6PR ligand and protease responsible for degradation of α-synuclein, a causative agent of Parkinson’s disease. In summary, we find that the expression of Vps35 D620N leads to endosomal alterations and trafficking defects that may partly explain its action in PD. The following chapter was published in Traffic. 2014 Feb;15(2):230-44. doi: 10.1111/tra.12136 and a full copy of the submitted document can be found in the Appendix.

66

4.1 Results 4.1.1 Vps35 D620N binds Vps29 and Vps26A with same affinity as Vps35 WT To assess the stability and functionality of the Vps35 D620N mutant in vitro, recombinant proteins were expressed and purified by affinity chromatography and gel filtration. Vps35 mutant D620N displayed similar elution profiles by gel filtration to the wild- type (WT) protein, and additionally CD spectra exhibited similar -helical contents consistent with the Vps35 -solenoid structure [276] (Figure 4.1B). We conclude therefore that the Vps35 D620N mutation does not lead to major misfolding of Vps35. We next tested the impact of the D620N mutation on the ability of Vps35 to form high affinity complexes with Vps26A and Vps29 in vitro (Figure 1C) [276]. Structural data for Vps35 only exists for the C-terminal segment from residues 476-780, bound to the small retromer subunit Vps29 [276]. The Asn 620 residue is present in an exposed loop between -helices of the repeating pairs of HEAT-like -helical repeats, and does not contribute to the Vps35 interface with Vps29 (Figure 4.1A). As shown by ITC, we find that the D620N mutant binds to Vps26A and Vps29 with affinities and thermodynamic parameters indistinguishable from the wild type

Vps35 molecule (Figure 4.1C). Vps26A binds Vps35 WT and D620N with affinities (Kds) of 1.1 and 0.5 nM respectively, while Vps29 binds WT and D620N with affinities of 170 nM and 180 nM respectively. Furthermore, in vivo co-immunoprecipitation experiments detected no differences in levels of endogenous Vps26A and Vps29 subunits associated with Vps35 WT-GFP or the Vps35 D620N-GFP mutant (Figure 4.1D). Overall we conclude that the presence of the D620N mutation in Vps35 does not directly affect global VPS35 folding and does not prevent the formation of a stable trimeric retromer protein complex.

4.1.2 Expression of Vps35 D620N mutant causes redistribution of endosomes To determine if the presence of the D620N mutation in Vps35 altered the subcellular localization of retromer we first performed live cell imaging of A431 cells transiently transfected with either Vps35 WT-GFP or Vps35 D620N-GFP pulsed with a fluid phase marker, Dextran-647. The Vps35 WT-GFP dextran positive endosomes were found throughout the cell cytoplasm and displayed mobility typical of early endosomes including fusion between Vps35-GFP positive endosomes (Supplemental Movie 1). In contrast, the Vps35 D620N-GFP positive endosomes were enlarged and redistributed to a tight, perinuclear localization (Figure 4.2A).

67

Figure 4.1. Vps35 mutant D620N binds to Vps29 and Vps26A in vitro and in vivo. (A) Structure of the Vps35 (483-780) complex with Vps29 [276]. Vps35 is shown in grey ribbons with transparent surface and Vps29 is shown in green ribbons. The D620N mutation is indicated in red. (B) CD spectra of Vps35 WT and Vps35 D620N proteins. (C) Isothermal titration calorimetry of purified recombinant Vps35 WT and Vps35 D620N protein to retromer subunits, Vps26A (right panel) and Vps29 (left panel) based on 16 x 2.5 mL injections of 100 µM Vps29 into 10 µM Vps35, or 50 µM Vps26A into 5 µM Vps35 D620N. Data shown represents integrated values of titration binding curves (n=3). (D) Sub-confluent HEK293 cells were transiently transfected with GFP, Vps35 WT-GFP or Vps35 D620N-GFP for 16 h washed with ice-cold PBS and lysed on ice. Protein complexes were isolated using GFP- Nanotrap beads, resolved by SDS-PAGE and transferred to PVDF membranes. Membranes were incubated with primary antibodies against GFP, Vps26A or Vps29 followed by IRDye 680/800 secondary antibodies and imaged using the Odyssey infrared imaging system.

68

Time-lapse microscopy demonstrated these enlarged endosomes contained dextran within their lumens supporting their capacity to receive endocytosed materials. Enlarged Vps35 D620N-GFP positive endosomes were observed to arise from smaller, dispersed endosomes, likely to be early endosome structures undergoing fusion during endosome maturation [304] (Supplemental Movie 2).

To determine if the retromer complex was also present on these enlarged redistributed endosomes induced in the presence of the Vps35 D620N mutant, we used indirect immunofluorescence on A431 cells transiently transfected with Vps35 WT-GFP or Vps35 D620N-GFP and immuno-labeled against endogenous Vps26A. Imaging of fixed cells overexpressing Vps35 WT-GFP show Vps35 WT-GFP/Vps26A positive endosomes with morphologies indistinguishable from Vps26A positive endosomes in neighboring untransfected cells. However, in cells overexpressing Vps35 D620N-GFP we observed endosomes positive for endogenous Vps26A redistributed to a perinuclear localization, dramatically different to surrounding untransfected cells showing Vps26A positive endosomes (Figure 4.2B). This analysis allowed us to confirm the presence of the retromer subunit, Vps26A, on the redistributed Vps35 D620N positive endosomes.

Quantification of the endosome redistribution in cells overexpressing Vps35 WT-GFP or Vps35 D620N-GFP was performed by measuring the distance of each Vps35-positive endosome from the middle of the nucleus. By designing a distance-based algorithm which measured fluorescent units by distinguishing perinuclear from non-perinuclear structures, we show that perinuclear fluorescence, representing endosomes, is increased in cells over- expressing Vps35 D620N-GFP when compared to Vps35 WT-GFP positive cells (Figure 4.2C). To determine if the localization of the modified endosomes induced by Vps35 D620N expression was microtubule-dependent and associated with the microtubule organization center we treated cells with a microtubule depolymerization agent, nocadazole. As shown in Figure 4.2D, cells overexpressing Vps35 D620N-GFP treated with nocadazole displayed a dissociation of the tightly localized endosomes to a dispersed cytoplasmic localization. Therefore the localization of the induced endosomes is not a product of aggregation.

69

Figure 4.2: Ectopic expression of Vps35 D620N alters endosome morphology and distribution. (A) Live cell time lapse confocal microscopy was performed on A431 cells expressing Vps35 WT-GFP or Vps35 D620N-GFP pulsed with 200 µg/mL Dextran-647 for 90 min using a Ziess LSM 710 FCS scanning confocal microscope. See supplementary data of publication for movies displayed in Figure 4.2A. Image represents single frame of 30 min movie captured. Scale bar 10 µm. (B) A431 cells expressing Vps35 WT-GFP or Vps35 D620N-GFP were fixed and indirect immunofluorescence was performed using antibodies against Vps26A. (C) Quantification of the intracellular endosomal distribution based on perinuclear intensity ratio of A431 cells over-expressing Vps35 WT-GFP or VPS35 D620N-GFP. Graph represents the mean of two independent experiments with fifteen images each (n=2;**p<0.01, Error bars +/- SEM). (D) A431 cells transfected with construct expressing Vps35 D620N-GFP were treated with 2 µM nocadazole for 60 min at 37°C, fixed and indirect immunofluorescence was performed using antibodies against anti-β-tubulin.

70

4.1.3 Identification of the redistributed endosome population To further investigate the properties of the enlarged redistributed endosomes induced in the presence of the Vps35 D620N mutant we performed a series of colocalisation experiments with endosome and TGN marker proteins. A431 cells were transiently transfected with Vps35 WT-GFP or Vps35 D620N-GFP for 24 h and indirect immunofluorescence was performed to compare the subcellular localization of endogenous EEA1, LAMP1 and p230 (Figure 4.3A). In cells overexpressing Vps35 WT-GFP the distribution of EEA1 positive endosomes was indistinguishable from that observed in neighboring untransfected cells. In contrast, Vps35 D620N expression caused a shift in EEA1-positive endosomes to the perinuclear region that was obvious when compared to neighboring untransfected cells (Figure 4.3A and Figure 4.3B).

71

Figure 4.3: Ectopically expressed Vps35 D620N positive endosomes contain both early and late endosome markers. (A) A431 cells transfected with constructs expressing Vps35 WT-GFP or Vps35 D620N- GFP were fixed and indirect immunofluorescence performed using monoclonal antibodies against EEA1, p230 or LAMP1 and counterstained with DAPI. All images represent a 1AU single slice captured using a Ziess LSM 710 Upright Scanning Laser confocal microscope at 63x magnification. Scale bar 5 µm. (B) Quantification of colocalization between Vps35 WT-GFP or Vps35 D620N-GFP and EEA1, p230 or LAMP1. Graph represents the mean of three independent experiments with ten images each (n=3;**p<0.01, Error bars represent +/- SEM).

Confocal analysis revealed a low, but consistent colocalization between Vps35 WT- GFP and EEA1 (colocalization coefficient (R) = 0.137) while Vps35 D620N-GFP showed an increase in colocalization with EEA1 (R = 0.207) (Figure 4.3B).While Vps35 WT-GFP showed low colocalization with LAMP1 (R = 0.101) a late endosome marker, and identical subcellular localization of LAMP1-positive compartments compared to neighbouring untransfected cells, Vps35 D620N-positive endosomes showed an increase in colocalisation with this marker (R = 0.158) on the large, redistributed endosomes (Figure 4.3A and Figure 4.3B).

We next investigated the possibility of an association between the Vps35 D620N- GFP and the TGN using colocalisation with the p230 marker. However, we observed no evidence of association between Vps35 D620N-GFP and p230 positive membranes (Figure 4.3A and Figure 4.3B) and no distinct morphological differences in p230-positive Golgi structures between Vps35 WT-GFP and Vps35 D620N expressing cells was observed (Figure 4.3A). These data suggest increased association of Vps35 D620N-positive endosomes with both early and late endosome markers in an induced tight perinuclear subcellular location.

4.1.4 Vps35 D620N mutant causes a defect in cathepsin D trafficking Retromer directly binds the CI-M6PR cytoplasmic domain via a region of Vps35 that overlaps with the position of the D620N mutation [70]. To examine the CI-M6PR interaction with Vps35 we co-transfected the CD8-CI-M6PR fusion construct and either Vps35 WT-GFP or Vps35 D620N-GFP and performed GFP-NanoTrap co-immunoprecipitations [67]. As seen in Figure 4.4A, Vps35 D620N-GFP can co-precipitate CD8-CI-M6PR at levels comparable to Vps35 WT-GFP. We conclude that the D620N mutation in Vps35 does not disrupt retromer’s capacity to interact with this cargo. 72

Next we examined if the Vps35 D620N mutant is able to alter the subcellular localization of CI-M6PR by disrupting its normal trafficking itinerary. We transiently transfected A431 cells with Vps35 WT-GFP or Vps35 D620N-GFP and determined the subcellular distribution of endogenous CI-M6PR. Consistent with previous reports [70], a proportion of endogenous CI-M6PR was localized to Vps35 WT-GFP-positive endosomes. Interestingly, the proportion of CI-M6PR/retromer-positive endosomal structures appeared to increase in cells overexpressing Vps35 D620N-GFP mutant (Figure 4.4B), with endogenous CI-M6PR also showing redistribution into tight, punctate perinuclear localization. To determine if the redistribution of the receptor impacted on delivery of CI- M6PR cargo, we investigated processing of cathepsin D in the presence of Vps35 D620N. Firstly we characterized the processing of cathepsin D from pro- into a mature form within the cells and secondly we assessed the levels of pro-cathepsin D secreted into the media of the transfected cells.

73

Figure 4.4: Ectopically expressed Vps35 D620N interacts with CI-M6PR but alters the receptors’ capacity to transport cathepsin D. (A) HEK293 cells co-transfected with Vps35 WT-GFP/CD8-CI-M6PR or Vps35 D620N- GFP/CD8-CI-M6PR were lysed and co-immunoprecipitation was performed using GFP- NanoTrap beads. Total cell lysates (input; 50 µg per lane) and isolated interacting proteins were analysed by Western immunoblotting using antibodies to CD8. (B) Sub-confluent A431 cells grown on coverslips were transiently transfected with Vps35 WT-GFP or Vps35 D620N-GFP for 16 h, fixed and indirect immunofluorescence performed using a mouse monoclonal against endogenous CI-M6PR. Images represent a 1AU single slice captured using a Ziess LSM 710 Upright Scanning Laser confocal microscope at 63x magnification. Scale bar 5 µm. (C) Sub-confluent A431 cells transiently expressing Vps35-GFP or Vps35D620N-GFP for 24 h were pulsed with 100 µg/mL of cycloheximide in serum-free medium. Medium and corresponding cell lysates were collected at 0, 3 and 7 h post treatment. Media samples (500 µl), precipitated on ice using 10% TCA, and total cell lysates (50 µg per lane) were analysed for levels of cathepsin D and β-tubulin by Western immunoblotting.

HEK293 cells were transiently transfected with either Vps35 WT-GFP or Vps35 D620N-GFP for 24 h, incubated with cyclohexamide for up to 7 h and cell lysate and medium samples collected at 0, 3 and 7 h post chase [67]. We observed that cells expressing the Vps35 D620N-GFP protein secreted the 50 kDa pro-cathepsin D into the media at 7 h post chase (Figure 4.4C). This result is consistent with impaired trafficking of CI-M6PR resulting in a deficiency of receptors at the TGN to interact with the pro-cathepsin D [70]. Additionally, cells overexpressing the Vps35 D620N-GFP mutant, when compared to Vps35 WT-GFP protein, showed decreased levels of mature 20 kDa cathepsin D in cell lysates, further demonstrating impaired delivery of pro-cathepsin D to the late-endosome/lysosome for processing into the mature 20 kDa form. Taken together, these results show that even though the Vps35 D620N mutant retains its ability to bind CI-M6PR, cathepsin D, the soluble ligand of this receptor, is not processed efficiently resulting in secretion of the immature pro- cathepsin D from the cells.

4.1.5 Vps35 D620N is associated with redistributed endosomes in PD patient fibroblasts Human dermal fibroblasts from a PD patient genotyped for the Vps35 D620N heterozygote point mutation were isolated and the subcellular localization of Vps35-positive endosomes and morphology of endocytic compartments was examined. By using indirect immunofluorescence to detect endogenous retromer subunits we detected a strong colocalisation of Vps35 with Vps26A in both control and patient fibroblasts (Figure 4.5B). In 74

PD patient fibroblasts we also observed a shift of Vps35-positive endosomes to a perinuclear localization, in contrast to control fibroblasts where Vps35-positive endosomes showed a broader distribution throughout the cytoplasm (Figure 4.5A).

75

Figure 4.5: Parkinson’s disease Vps35 D620N patient fibroblasts have redistributed endosomes. (A) The colocalisation of endogenous Vps35 in sub-confluent control and Parkinson’s disease patient fibroblasts using antibodies against Vps26A, EEA1, p230 or LAMP1 was determined using indirect immunofluorescence. Cell monolayers were counterstained with DAPI. All images represent a 1AU single slice captured using a Ziess LSM 710 Upright Scanning Laser confocal microscope at 63x magnification. Scale bar 5µm. (B) Quantification of colocalization analysis performed in (A). Graph represents the mean of two independent experiments with ten images each (n=2;*p<0.05, Errors bars represent +/- SEM). (C) The intracellular distribution of Vps35 positive endosomes quantified using perinuclear intensity ratio in control and Parkinson’s disease patient fibroblasts. Graph represents the mean of two independent experiments with ten images each (n=2;*p<0.05, Error bars represent +/- SEM).

Quantification of this phenotype using the distance-based methodology described above again showed a significant increase in the perinuclear intensity ratio in patient fibroblasts compared to control fibroblasts (Figure 4.5C). Subcellular localisation of the Vps35-positive endosomes in fibroblasts was performed using endogenous markers as described above. Colocalisation studies with EEA1, an early endosome marker, show an increased proportion of Vps35-positive punctate structures positive for the EEA1 in patient fibroblasts (R = 0.304) compared to the control (R = 0.209) (Figure 4.5B). Further colocalisation of Vps35 with LAMP1, a late endosome marker, showed a higher colocalisation coefficient between Vps35 and LAMP1 in patient fibroblasts (R =0.274), compared to control (R = 0.197) (Figure 4.5B). Interestingly, using both EEA1 and LAMP1 markers we also observed consistent evidence of enlarged redistributed endosomes, with LAMP1-positive endosomes showing consistent enlargement of the endosomal lumen (Figure 4.5A, LAMP1 panels). Investigation of p230, a TGN marker, in control and patient fibroblasts showed no change in colocalization between Vps35 and p230 positive structures (Figure 4.5B) and no morphological difference in p230 positive structures (Figure 4.5A, p230 panels). Taken together, these results confirm increased localization of Vps35 D620N mutant to both early and late endosomes and corroborate the ectopic studies above.

To determine if the presence of the Vps35 D620N mutation in patient fibroblasts impacted on the retromer mediated trafficking of CI-M6PR we initially performed colocalisation studies between endogenous Vps35 and CI-M6PR in control and patient fibroblasts. Within PD patient fibroblasts expressing Vps35 D620N we observed an increase in the colocalization of CI-M6PR with Vps35 (R = 0.257) compared to that found in control 76 cells (R = 0.203) (Figure 4.6B), as well as a clear shift in CI-M6PR-positive endosomes into a more perinuclear subcellular localization (Figure 4.6A).

Figure 4.6: Cathepsin D processing is impaired in Parkinson’s disease Vps35 D620N patient fibroblasts. (A) Indirect immunofluorescence on sub-confluent control and Parkinson’s disease patient fibroblasts was performed using antibodies against Vps35, CI-M6PR and counterstained with DAPI. Images represent a 1AU single slice captured using a Ziess LSM 710 Upright Scanning Laser confocal microscope. Scale bar 5 µm. (B) Quantification of colocalization analysis performed in (A). Graph represents the mean of two independent experiments with ten images each (n=2;*p<0.05, Error bars represent +/- SEM). (C) Monolayers of control and Parkinson’s disease patient fibroblasts were pulsed with 100 µg/mL of cyclohexamide for 0, 3 and 7 h, when cells were lysed and analyzed by Western immunoblotting using α- cathepsin D antibody.

We further examined the processing of CI-M6PR ligand, cathepsin D, using the cyclohexamide assay described previously. The fibroblasts from the patient displayed higher levels of mature cathepsin D at steady state when compared to the normal patient fibroblasts. After the inhibition of protein synthesis for 7 h we observed that the level of mature 20 kDa cathepsin D in patient fibroblasts decreased when compared to the fibroblasts from a control sample (Figure 4.6C). This is consistent with a reduction in the efficiency to transport newly synthesized pro-cathepsin D to the lysosome. However, we 77 were unable to observe secretion of the pro-cathepsin D into the media presumably due to the lower level of cathepsin D expressed by these cells (data not shown). These results demonstrate a defect in trafficking of CI-M6PR in patient cells, resulting in processing disruption of cathepsin D into a mature form of this enzyme. These observations are consistent with those observed in the ectopic expression experiments performed above.

78

4.2 Discussion Here we identified that the Parkinson’s disease-linked Vps35 D620N mutation does not interfere with Vps35’s capacity to form high affinity interactions with Vps26A and Vps29, as evidenced by in vitro and cell-based assays. Ectopic expression of the Vps35 D620N protein in mammalian cells, as well as examination of fibroblasts isolated from a Parkinson’s disease patient containing the D620N mutation in Vps35, demonstrated retromer association with enlarged endosomes that were concentrated to a perinuclear subcellular localization that also displayed evidence of retention of early and late endosome markers. Additionally, we demonstrate that even though Vps35 D620N-retromer interacts with its receptor cargo, CI-M6PR, its expression caused the receptors subcellular distribution and function to be modified. In both the overexpression model and Parkinson’s disease patient fibroblasts, the processing of cathepsin D, a CI-M6PR ligand, into the mature 20 kDa active form was markedly decreased in the presence of Vps35 D620N protein with resulting pro- cathepsin D secreted from the cells. Therefore the expression of retromer incorporating the mutant Vps35 D620N protein induces a disruption of the normal trafficking itinerary of CI- M6PR.

To date, little information exists surrounding how these identified point mutations may play a role in the biology of PD. Exploitation of over-expression systems and fibroblasts isolated from PD patients positive for the D620N substitution has given rise to differing hypotheses surrounding the defective pathways. Nonetheless, recent publications demonstrate an impact on the retrograde pathway [267, 268]. Immunofluorescence microscopy revealed endosome swelling, primarily of the late endosome; a direct indication of impaired trafficking from the endosome compartment. Under normal conditions, retromer retrieves CI-M6PR from the endosome and delivers it to the TGN [70]. Here, CI-M6PR binds cathepsin D and is trafficked from the TGN to the endosomal network where receptor and ligand dissociate. Cathepsin D is further processed in the lumen of the endosome to yield an active peptide which degrades luminal proteins delivered to the lysosome. Examination of the cathepsin D in the presence of Vps35 D620N demonstrated marked reductions in total mature (enzymatically active) levels found within the endosomal network, further supporting a direct impact on retrieval of the CI-M6PR to the TGN in the presence of this mutation [267]. Expression of the D620N results in increased endosomal localization of the CI-M6PR, when compared to its steady state localization, which is concentrated at the TGN 79

[267, 268]. Expression of Vps35 D620N impacted the localization of the CI-M6PR and subsequently disrupted the trafficking of its ligand within the endosomal network. It is clear this disruption in trafficking is not a result of reduced interaction between Vps35 D620N and CI-M6PR [267]. These findings demonstrate the expression of D620N having a negative impact on the retrograde pathway. These initial findings have been confined to cell culture systems, as stereotactic injection of rats with AVV2/6 expression vectors carrying D620N failed to impact levels of cathepsin D at 12 weeks post-delivery [271]. Though unexpected, it may simply be representative of short experimental time and require extended investigation to produce early stage pathogenesis in animal models.

Though these findings demonstrate an impact on Vps35 D620N to correctly sort cargo away from the endosome, it does not provide an underlying mechanism. One proposed explanation for failed sorting observed in the presence of Vps35 D620N, is reported to be a result of decreased interaction with WASH complex subunit, FAM21 [268, 270], with no change in endosomal sub-cellular localization. Interestingly, siRNA suppression of WASH subunit, Wash1, is reported to redistribute retromer-positive endosomes to the perinuclear space and perturb the redistribution of ATG9A from the perinuclear space, leading to impaired autophagy, a phenotype that was also described following stable expression of Vps35 D620N [270]. While the redistribution of retromer- positive endosomes to the perinuclear space following expression of D620N is consistent with a loss of WASH functionality, others have detected no change in the interaction of the retromer complex with the FAM21 subunit using co-immunoprecipitation or GST pull downs [267]. Though contradictory, it is possible these differences arise from differing sensitivity of applied methodologies.

The redistributed, altered endosome structures positive for Vps35 D620N-GFP, represent a common phenotype observed with endosome dysfunction and neurodegeneration [306]. Live cell imaging of these enlarged and redistributed endosomes demonstrated they are motile entities rather than static aggregates, and the delivery of dextran via endocytosis supports their capacity to still interact with other endosomal compartments. The redistributed Vps35 D620N positive endosomes in both overexpression and fibroblast models had increased colocalisation with EEA1 and LAMP1 relative to wild- type Vps35. The retention of these markers is consistent with a defect in the endosomal maturation process [7], where the conversion between early and late endosome 80 compartments may be underpinned by a failure of Vps35 D620N endosomes to mature at a constant rate, leading to the dilated endosome phenotype. The dysregulation of retromer itself and also a number of proteins that associate with retromer are known to cause a similar change in endosomes. For example, recruitment of retromer to the endosomal membranes has been reported to require the two RabGTPases, Rab5 and Rab7, acting in concert [18, 102]. The failure of Rab5 to dissociate from the membrane [7], overexpression of dominant- negative Rab5 and Rab7 [47] or depletion of the retromer cargo recognition complex have all been reported to result in increased endosome size [70], like the phenotype we observed when Vps35 D620N protein is expressed. In addition, recent studies demonstrate the importance of the WASH complex in maintaining endosome integrity and lumen size while linking vesicles and F-actin networks [32]. The role of the WASH complex is largely focused on early or pre-endocytic compartments, however it has also been described to associate with late endocytic structures [135, 307]. As retromer has been shown to interact directly with the WASH complex via the Fam21 subunit [72, 131, 132], it is plausible that the dilated endosomes are a result of an inability of the Vps35 D620N to activate the WASH complex with high affinity resulting in a disruption of F-actin dependent cellular mechanisms including the formation of endosome derived tubular-vesicular transport vesicles. However, we find that Vps35 D620N retains a high level of colocalization (>85%) with endogenous FAM21 and does not perturb the described interaction between retromer and the WASH complex using in cell immunoprecipitation and GST pull down assays (data not presented in manuscript).

The D620N mutation is within the region of Vps35 identified to bind CI-M6PR [70, 276]; however our study shows that Vps35 D620N-containing retromer still has the capacity to directly interact with CI-M6PR. Therefore the detection of the immature form of pro- cathepsin D in the medium of cells overexpressing Vps35 D620N mutant and decreased amount of mature cathepsin D in the over-expression and human fibroblast models is not due to an inability of the Vps35 D620N protein to bind CI-M6PR. CI-M6PR was present in the altered retromer positive endosomes of cells expressing the Vps35 D620N mutation, suggesting the inability of the receptor to be efficiently transported from endosomes to the TGN. Therefore, the expression of the Vps35 D620N mutant may result in a reduced ability of this mutant to create transport vesicles from the endosome, leading to retention of retromer and cargo on the altered endosomes, not from an inability of Vps35 D620N retromer to engage its cargo. 81

Interestingly, the defect in CI-M6PR trafficking that manifests in missorting of cathepsin D in the presence of Vps35 D620N mutation may have a direct correlation to the development and progression of Parkinson’s disease as α-synuclein, a protein implicated in Parkinson’s disease pathogenesis, has been identified as a cathepsin D target protein within the lysosomes [244, 280]. For example, overexpression of cathepsin D in dopaminergic cell cultures increases proteolysis of endogenous α-synuclein, while cathepsin D knockout mice show improper processing of α-synuclein in dopaminergic neurons that ultimately leads to α-synuclein cellular toxicity [281]. Additionally, enzymatic inactivation of cathepsin D leads to signs of early onset, progressively fatal neurodegenerative disease in humans [308-310]. This further supports the notion that the regulation of lysosomal proteolysis of α-synuclein is an important contributing factor in PD pathogenesis. Therefore, it is plausible that the molecular mechanisms that underpin the Vps35 D620N disease manifestation include altered trafficking of CI-M6PR and its cargo, cathepsin D, ultimately leading to impaired degradation of proteins delivered to the lysosome, including α-synuclein and resulting in formation of Lewy bodies, a hallmark of Parkinson’s disease.

Overall, we have shown that Parkinson’s disease linked Vps35 D620N mutant causes a deficit in retromer-dependent trafficking of CI-M6PR, and its ligand cathepsin D, likely arising from the generation and redistribution of enlarged endosome compartments that retain retromer. The ability for this mutation to contribute to the pathogenesis of Parkinson’s disease is likely secondary to these reported trafficking defects by reducing the breakdown of disease-associated proteins such as α-synuclein, in the late endosomal network.

82

Chapter 5: Parkinson’s disease linked Vps35 R524W mutation impairs retromer’s endosomal association and induces α-synuclein inclusion formation

Introduction

Retromer is a high-affinity heterotrimeric complex composed of Vps29, Vps35 and one of the two Vps26 subunits, Vps26A or Vps26B [65, 67, 69]. Retromer has a central role in the sorting of receptor cargo within endosomal membranes, which is essential to coordinate the specific spatio-temporal localisation of individual receptors enabling them to perform their varied functions [70, 112]. One such cargo is the cation-independent mannose-6-phosphate receptor (CI-M6PR) that controls the sorting of lysosomal enzymes such as cathepsin D which is required for protein turnover [70, 112, 115]. Retromer serves as a multi-functional scaffold forming an interaction hub for a wide array of endosome-associated proteins, collectively termed the retromer interactome. These interactions aid in the formation of the cargo-containing tubulovesicular membrane carriers destined for other compartments such as the Golgi apparatus and plasma membrane. The diverse proteins that associate with retromer include regulatory molecules, proteins required for membrane recruitment, and protein complexes that control membrane tubulation and scission (reviewed in [66]). It is the spatial and temporal coordination of the interactions between the retromer and the retromer interactome that enables it to coordinate multiple endosome-derived trafficking pathways.

Rare familial mutations in a range of proteins have been identified and have provided significant insight into the molecular pathways involved in the manifestation of PD (reviewed in [311]). Recently, a number of point mutations (D620N, P316S, R524W, L774M) within the retromer subunit Vps35 were associated with late-onset PD and analysis of 144 individuals confirmed that overall levels of Vps35 mRNA in the substantia nigra were significantly decreased in PD-affected patients [256, 258, 269].

83

Here, we report the cellular characterization of two familial PD-linked Vps35 variants, Vps35 P316S and Vps35 R524W. While the P316S variant appears to have little impact on retromer assembly or function, Vps35 R524W is poorly recruited to endosomes and impairs the recruitment of retromer-dependent interacting proteins and the trafficking of CI-M6PR. Expression of Vps35 R524W induced higher levels of α-synuclein aggregates, which can be decreased by stabilization of retromer using the recently described pharmacological agent, R55 [160]. These findings provide insight into the underlying molecular mechanism of PD- linked Vps35 R524W, its involvement in the molecular pathways associated with PD and the ability of the newly described retromer stabilizing agent to influence the formation of LBs, a hallmark of PD pathology.

84

5.1 Results

5.1.1 Vps35 P316S and Vps35 R524W are incorporated into retromer complexes To determine if the formation of the heterotrimeric retromer complex in the presence of Vps35 variants P316S and R524W is altered, in vitro isothermal titration calorimetry (ITC) and in vivo co-immunoprecipitation were employed using full-length GST fusion recombinant protein and GFP fusion constructs, respectively. The arginine 524 residue is present in an exposed loop between -helices of the reiterating pairs of HEAT-like -helical repeats, and contributes to the VPS35 interface with VPS29 as shown in the co-crystal structure [276]. The proline 316 residue is predicted to be within an intervening loop, so its substitution is unlikely to disrupt the overall Vps35 structure. As demonstrated by ITC, Vps35 P316S retained interaction with retromer subunits Vps26A and Vps29 (Kd = 1.5nM and 170nM, respectively) at thermodynamic profiles similar to that of wild-type Vps35 (Kd = 1.1nM and 250nM, respectively). The Vps35 R524W mutant also demonstrated a similar binding affinity to that of the wild-type Vps35 for Vps29 (Kd = 303nM), and Vps26A (Kd = 1.4nM) (Figure 5.1A). In support of these in vitro experiments, in vivo co-immunoprecipitation from HeLa cells transiently expressing GFP fusion constructs demonstrated that Vps35 P316S and Vps35 R524W interact with retromer subunits Vps26A and Vps29 at levels similar to that of Vps35 WT-GFP (Figure 5.1B) and their expression does not alter steady state protein levels of either subunit. This data demonstrates that both PD-associated mutations do not appreciably perturb the formation of a stable retromer trimer.

85

Figure 5.1: Parkinson’s disease linked Vps35 mutants P316S and R524W do not disrupt trimer formation (A) Isothermal Titration Calorimetry of Vps35 point mutations with retromer subunits, Vps29 (left panel) and Vps26A (right panel). For a summary of ITC binding data see Supplementary Table S1. (B) Representative (n=3) GFP-NanoTrap immunoprecipitations of HeLa cells transiently expressing GFP, Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP followed by immunoblotting analysis of precipitated complexes with anti-GFP, anti-Vps26A and anti-Vps29 antibodies.

5.1.2 Vps35 R524W-containing retromer has impaired endosome recruitment The subcellular localization of Vps35 P316S and Vps35 R524W was determined in HeLa cells transiently expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W- GFP. Consistent with the sub-cellular localization of Vps35 WT-GFP, Vps35 P316S-GFP demonstrated a high level of endosomal recruitment and displayed a high level of co- localization with Vps26A on these endosomes (Figure 5.2A). In contrast to this, while Vps35 R524W-GFP showed partial endosomal recruitment where it co-localised with Vps26A, a significant fraction of the protein also showed a diffuse localization to the cytosol (Figure 5.2A).

Membrane fractionation was employed to determine the relative proportion of retromer associated with membranes. HeLa cells transiently transfected with GFP fusion constructs were lysed in sucrose-containing buffer, and cytosolic and microsome fractions were isolated. Loading controls and purity of the fractions were monitored using late endosome marker LAMP1 (microsome) and β-tubulin (cytoplasmic). Consistent with the previously described data, wild-type retromer proteins, including Vps35 WT-GFP, were observed associated with membranes and within the cytosol. Vps35 P316S-GFP showed a similar distribution to Vps35 WT-GFP whereas Vps35 R524W-GFP was predominantly distributed in the cytosol with a proportional decrease in the protein associated with membranes. Further, Western blotting of endogenous Vps26A, Vps35 and Vps29 demonstrate no significant shift in total levels of retromer from the microsomal fraction following expression of the GFP fusion constructs (Figure 5.2B). The biochemical and immunofluorescence findings together suggest impaired endosome association of Vps35 R524W containing retromer.

86

We employed confocal microscopy to further examine the endosome morphology and retromer localization in the presence of Vps35 P316S and Vps35 R524W (Figure 5.2C). HeLa cells transiently expressing GFP fusion constructs were immunolabeled with antibodies against early endosome marker, EEA1 or late endosomal marker, LAMP1.

87

Figure 5.2. Vps35 R524W disrupts recruitment of retromer to the endosomal membrane (A) Confocal analysis of HeLa cells transiently expressing Vps35 WT-GFP, Vps35 P316S- GFP or Vps35 R524W-GFP immunolabeled with anti Vps26A (n=3, 10 images per group). Scale bar: 5µm. (B) Representative immunoblotting (n=4) of HeLa cells transiently expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W subjected to fractionation, SDS-PAGE and immuno-labelling with antibodies against the listed proteins. (C) Representative immunofluorescence images of HeLa cells transiently expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP labelled with EEA1 or LAMP1 followed by counterstaining with DAPI. Scale bar: 5µm. (D) Analysis of co-localization from (c) represented by Pearson’s correlation co-efficient. Graphs representative of three independent experiments with 10 images per group with 5-7 transfected cells per field of view (n=3, Error bars represent +/- SEM; *p<0.01; One-way ANOVA followed by Bonferroni Post hoc test).

Consistent with the expression of Vps35 WT-GFP, confocal microscopy revealed recruitment of Vps35 P316S-GFP to EEA1 positive endosomes (RWt = 0.2130, RP316S = 0.2013, Figure 5.2D), while cells expressing Vps35 R524W-GFP demonstrated reduced overlap with EEA1 positive compartments (RR524W = 0.1551, Figure 2D). However, no differences in the morphology of EEA1 positive endosomes, when compared to neighbouring untransfected cells, was observed in transfected cells expressing any of the Vps35 proteins (Figure 5.2C). Immunolabeling of HeLa cells with antibodies against LAMP1 revealed no gross morphological changes of late endosome compartments in cells expressing Vps35 WT-GFP. In addition, expression of Vps35 P316S or Vps35 R524W do not impact Golgi morphology as marked by p230 immuno-staining (supplementary figure 2).

5.1.3 The association of Vps35 R524W with regulators of the retromer complex is impaired Retromer does not bind directly to membranes and its recruitment to endosomes is controlled by its capacity to coordinate interactions with a number of membrane-associated proteins. Association of the retromer complex with the endosomal membrane is positively regulated by the small GTPase Rab7a and negatively regulated by the RabGAP TBC1D5 [18]. Following recruitment to the endosome, additional retromer dependent sorting machinery such as the Arp2/3 activating complex, WASH, is recruited to aid in cargo sorting [130-132]. We examined the sub-cellular localization of WASH complex subunit FAM21, Rab7a and TBC1D5 and their ability to interact with the retromer complex

88

89

Figure 5.3. Vps35 R524W has diminished association with regulators of retromer (A) Immunofluorescence staining of FAM21 in HeLa cells transiently expressing Vps35 WT- GFP, Vps35 P316S-GFP and Vps35 R524W-GFP followed by DAPI counterstaining. (B) Analysis of co-localization from (A) represented by Pearson’s co-efficient. Graphs representative of three independent experiments with 10 images per group with 5-7 transfected cells per field of view (n=3, Error bars represent +/- SEM; *p<0.01; One-way ANOVA followed by Bonferroni Post hoc test). (C) Representative co-immunoprecipiation of GFP, Vps35 WT-GFP, Vps35 P316S-GFP and Vps35 R524W-GFP with endogenous FAM21 from HeLa cells. (D) Representative immunofluorescence images of HeLa cells transiently expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP labelled with late endosome marker, Rab7. (E) Analysis of co-localization from (D) represented by Pearson’s co-efficient. Graphs representative of three independent experiments with 10 images per group with 5-7 transfected cells per field of view (n=3, Error bars represent +/- SEM; *p<0.01; One-way ANOVA followed by Bonferroni Post hoc test). (F) Representative co-immunoprecipiation of GFP, Vps35 WT-GFP, Vps35 P316S-GFP and Vps35 R524W- GFP with endogenous Rab7 from HeLa cells. (G) Localization of TBC1D5 in HeLa cells expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP counterstained with DAPI. (H) Analysis of co-localization from (G) represented by Pearson’s co-efficient. Graphs representative of three independent experiments with 10 images per group with 5-7 transfected cells per field of view (n=3, Error bars represent +/- SEM; *p<0.01; One-way ANOVA followed by Bonferroni Post hoc test). (I) Representative Western blots of TBC1D5 co-immunoprecipitation with Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP containing retromer from HeLa cells. via co-immunoprecipiation following expression of Vps35 P316S-GFP and Vps35 R524W- GFP. HeLa cells transiently expressing GFP fusion constructs were fixed and immunolabeled with antibodies against endogenous FAM21 (Figure 5.3A), Rab7a (Figure 5.3D) and TBC1D5 (Figure 5.3G). Consistent with previous findings, Vps35 WT-GFP or

Vps35 P316S-GFP demonstrated high level of co-localization (RWt = 0.8900; RP316S = 0.9035 (Figure 5.3B)) in contrast to cells expressing Vps35 R524W-GFP where a significant decrease in co- localization with FAM21 (RR524W = 0.8011, Figure 5.3B) was observed. Similar to this, HeLa cells expressing Vps35 WT-GFP or Vps35 P316S-GFP immunolabeled with antibodies against Rab7a or TBC1D5 displayed a moderate, but consistent, amount of co-localization (Rab7a: RWt = 0.2828; RP316S = 0.2856; TBC1D5: RWT = 0.2967; RP316S = 0.2820, Figure 5.3E and Figure 5.3H, respectively). In contrast to these observations, expression of Vps35 R524W revealed a marked decrease in overlap with both Rab7a and

TBC1D5 (RR524W = 0.2166 and RR524W = 0.2056, Figure 5.3E and Figure 5.3H, respectively).

Further, consistent with the previously described LAMP1 immunolabeling (Figure 5.2D), Rab7a-positive endosomes demonstrated swelling following expression of Vps35 R524W, but not Vps35 P316S or Vps35 WT, whereas endosomes positive for FAM21 or 90

TBC1D5 displayed no evidence of gross morphological changes or differences in subcellular localization when compared to Vps35 WT-GFP or Vps35 P316S-GFP expression. Co- immunoprecipitation was employed to investigate the ability for Vps35 P316S and Vps35 R524W-containing retromer to interact with, FAM21, Rab7a and TBC1D5 (Figure 5.3). HeLa cells expressing GFP fusion constructs were lysed and protein complexes were immunoprecipitated using GFP NanoTrap, resolved by SDS-PAGE and identified using Western blotting technique. Using antibodies against GFP and endogenous FAM21, Rab7A and TBC1D5 no differences in the ability for Vps35 P316S-GFP-containing retromer to interact with FAM21, Rab7A or TBC1D5 was observed, as determined by comparison to the levels of co-precipitated protein using Vps35 WT-GFP (Figure 5.3C, 5.3F and 5.3I, respectively). In contrast, Vps35 R524W-GFP demonstrated a clear reduction in its capacity to co-precipitate FAM21, Rab7a and TBC1D5, despite the total levels of all proteins being similar. Overall, Vps35 R524W consistently shows a decreased level of recruitment to endosomes, which is reflected in a lower level of interaction with proteins known to function in its recruitment to membranes.

5.1.4 Retromer deficiency induces α-synuclein aggregation α-synuclein is the major component of LBs, a prominent phenotype in PD pathogenesis. Although the underlying cause of LB formation is not fully understood it appears to be caused by the perturbation of several distinct cellular homeostasis processes, including defects in endosomal degradation pathways [49, 281]. To address if retromer plays a role in the accumulation of aggregated α-synuclein, we used the SH-SY5Y neuroblastoma cell model which endogenously expresses α-synuclein, possesses machinery needed for dopamine uptake and metabolism, and responds to external stimuli, including depolarisation using KCl [210]. SH-SY5Y cells were transduced with a control or silencing shRNA hairpin against Vps35 and selected to generate a stable knockdown population. Confocal microscopy was employed to quantify the proportion of cells that were positive for endogenous α-synuclein inclusions following treatment with 50 mM KCl. Untreated control cells were found to have a small percentage (8.3%) of cells containing α-synuclein inclusions while depletion of retromer resulted in α-synuclein inclusions observed in 21% of cells (Fig 5.4A & B). SH-SY5Y cells 48 hr after treatment with 50 mM KCl had a higher proportion of cells with α-synuclein inclusions in both control (28%) and shRNA Vps35 cells (43%) (Fig 5.4A & B). The depletion of retromer therefore results in an increased probability of SH- 91

SY5Y cells forming inclusions, indicating a fundamental role for retromer and endosomal trafficking in induction or clearance of α-synuclein inclusions.

Figure 5.4. Loss of retromer induces α-synuclein inclusions that can be rescued using R55, a pharmacological chaperone (A) Representative confocal images of stably transduced SH-SY5Y cells immunolabled for endogenous α-synuclein 0 and 48hrs post depolarization in recovery medium supplemented with 5µM R55 or DMSO. (B) Quantitation of control and shRNA Vps35 cells bearing α- synuclein aggregates at 0 and 48hrs post depolarization following incubation with R55 or DMSO. Graph represents two independent experiments with 10-15 images per experiment captured in a Z-stack with 10-20 cells per field of view, for each stated time point and respective treatment. Error bars displayed as +/- SEM; *p<0.05;**p<0.01; ***p<0.001; ****p<0.0001 by ANOVA followed by Bonferroni multiple comparisons test). (c) Representative Western blot analysis of retromer subunit, Vps35, in stably transduced SH- SY5Y cells treated with 5µM R55 or DMSO for 48hrs (n=4, repeated and represented in triplicate).

Recently, a retromer stabilizing agent R55 that increases the total level of retromer in cells, was shown to decrease the formation of the Aβ1-42 pathogenic form of amyloid  (A) peptide associated with Alzheimer’s disease pathology [160]. To determine if R55 can provide the same protective effect on formation of α-synuclein positive inclusions, SH-SY5Y cells were treated with 50 mM KCl and allowed to recover for 48 h in medium supplemented 92 with either DMSO or 5 µM R55. Consistent with previous reports [160], incubation of cells with R55 resulted in increased total Vps35 levels after 48 h compared to DMSO control (Fig 5.4C). R55 treatment restored the levels of Vps35 protein in shRNA depleted cells to that observed normally for the endogenous protein. Furthermore, incubation with 5 µM R55 reduced the numbers of control cells bearing α-synuclein positive inclusions from 28% to 18%, while the number of shRNA Vps35 cells with inclusions was reduced from 43% to less than 30% (Fig 5.4A & B). Taken together, this data suggests that use of R55 agent can increase levels of the retromer complex in SH-SY5Y cell model which can reduce the level of LB formation.

5.1.5 Vps35 R524W expression induces α-synuclein aggregation The reduced recruitment of Vps35 R524W to endosomes would impair retromer’s function and therefore the expression of this Vps35 PD-associated mutant may increase α- synuclein aggregation. Using the described methodology, SH-SY5Y cells expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP were incubated with 50 mM KCl for 60 min and allowed to recover in complete growth medium for 48 hours, immunolabeled for endogenous α-synuclein and transfected cells were scored as positive or negative LBs. 22% of SH-SY5Y cells expressing Vps35 WT-GFP were found to bear α–synuclein inclusions, whereas 30% of Vps35 P316S-GFP and 43% of Vps35 R524W-GFP transfected cells were positive for α-synuclein inclusions (Figure 5.5A & B). Therefore, the expression of these PD- associated Vps35 mutants can directly influence the pathogenesis of PD by modulating retromer’s function, which ultimately results in the formation of α-synuclein inclusions. To determine if this increased formation of inclusions in cells expressing PD retromer mutants can be reversed by the R55 retromer stabilizing agent, cells were incubated with 5 µM R55 for 48 h. The total number of transfected cells bearing inclusions were reduced to 17, 20 and 30% for cells expressing Vps35 WT-GFP, Vps35 P316S-GFP and Vps35 R524W-GFP, respectively (Figure 5.5A & B). Based on these results and previous observations that R55 acts as a retromer stabilizing agent [160], we investigated if the presence of R55 can increase total and endosomal levels of Vps35 R524W-GFP. For this, we employed both immunoflourescence and biochemical methodologies described above. HeLa cells transiently expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP were incubated with 5 µM R55 or DMSO control for 48 hours, fixed and immuno-labelled with the early endosomal marker, EEA1. 93

94

Figure 5.5 Parkinson’s disease linked Vps35 point mutation increase the production of α-synuclein inclusions (A) Confocal immunofluorescence images of SH-SY5Y cells expressing GFP fusion constructs, immunolabeled with anti-α-synuclein and counterstained with DAPI following treatment with DMSO or 5µM R55. (B) Graph representing number of transfected SH-SY5Y cells bearing α-synuclein aggregates following incubation with DMSO or R55 for 48hrs post depolarization. Graph represents two independent experiments with 10-15 images per experiment captured in a Z-stack with 10-20 cells per field of view, for each stated time point and respective treatment. Error bars displayed as +/- SEM; *p<0.05 by by ANOVA followed by Bonferroni multiple comparisons test. (C) Confocal images of HeLa cells transiently expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP incubated with DMSO or 5μM R55 immunolabeled with EEA1 and counterstained with DAPI. (D) Graphical representation of co-localization observed in (C) from two independent experiments with 10 images per construct and time point. Error bars displayed as +/- SEM; * p<0.05; by ANOVA followed by Bonferroni multiple comparisons test. (E) Representative Western blots of HeLa cells transiently expressing GFP, Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W- GFP treated with 5μM R55 or DMSO for 48hrs (n=4). Membranes were also analysed using anti-LAMP1 for loading control.

As previously observed (Figure 2D), in cells incubated with DMSO vehicle control co- localisation analysis between GFP and EEA1 showed a decrease in Vps35 R524W-GFP at the early endosome (RWT =0.214, RP316S = 0.213, RR524W =0.158; Figure 5.5C & 5.5D). Interestingly, treatment of cells transfected with Vps35 R524W-GFP and treated with R55 agent demonstrated an increased endosomal localization to the levels demonstrated for

Vps35 WT-GFP DMSO control (Figure 5.5C & 5.5E, RWT =0.269, RP316S = 0.271, RR524W =0.211). In agreement with this observation, biochemical fractionation methodology of cells transfected with GFP, Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP showed increase in the total level of Vps35 WT-GFP and Vps35 P316S-GFP and a marked shift of Vps35 R524W-GFP from the cytosolic to microsome fraction (Figure 5.5E). However, the presence of R55 did not completely rescue the loss of Vps35 R524W-GFP in the microsome fraction to a level equivalent to that observed for Vps35 WT-GFP or Vps35 P316S-GFP (Figure 5.5E).

5.1.6 Retrograde sorting is delayed in the presence of Vps35 R524W To test the functional impact the expression of Vps35 R524W has on retrograde endosomal cargo sorting, we investigated the recycling of the well-characterized retromer cargo, Cation-Independent Mannose-6-Phosphate Receptor (CI-M6PR). Previous reports mapped the interaction between retromer and the cytosolic tail of the CI-M6PR to amino acids 500-693 of the Vps35 subunit [70]. Given the location of the point mutations on the 95

Vps35 in relation to the Vps29 and the cargo-interacting domain within Vps35, the interaction of retromer with CI-M6PR, modulation of soluble lysosomal enzyme delivery, relative co- localisation and kinetics of retrograde trafficking were analysed as previously described [67, 70, 267]. Vps35 P316S or Vps35 R524W containing retromer were found to co-precipitate the M6PR at levels identical to that of the wild-type retromer (Figure 5.6A). Expression of Vps35 P316S-GFP, like Vps35 WT-GFP, did not disrupt the lysosomal delivery of cathepsin D, which is dependent on mannosylation and binding to the CI-M6PR for the transport of the Golgi-processed precursor into the endosomal system. In contrast, expression of Vps35 R524W-GFP clearly disrupted the trafficking itinerary of the CI-M6PR as secretion of pre- cathepsin into the extracellular media was observed after protein synthesis was inhibited for 7 hours (Figure 5.6B).

Next, the sub-cellular distribution of the endogenous CI-M6PR relative to retromer was examined. The CI-M6PR was localised to punctate and perinuclear organelles in cells transfected with Vps35 WT-GFP and Vps35 P316S-GFP and demonstrated a small amount of overlap with retromer (RWT = 0.210 and RP316S = 0.204; Figure 5.6C & 5.6D). In comparison, expression of Vps35 R524W-GFP was found to alter the sub-cellular distribution of the CI-M6PR from predominantly perinuclear staining to dispersed puncta (Figure 5.6C). These dispersed CI-M6PR-positive puncta demonstrated increased co- localisation with the Vps35 R524W-GFP endosomal retromer (RR524W =0.261) relative to controls (Figure 5.6D).

Next, HeLa cells were co-transfected with CD8-CI-M6PR to monitor the retrograde delivery of internalised antibodies to the TGN, using a well-established antibody uptake assay [67]. Cells were incubated on ice with anti-CD8 antibody and chased at 37°C for up to 30 min. In cells expressing CD8-CIM6PR and GFP only, strong CD8 immunostaining was observed on membranes positive for p230, demonstrating efficient delivery of the reporter to the TGN network following retromer-mediated transport through the endosome. Consistent with this, strong CD8 immunostaining was observed to overlap with p230- positive membranes in cells expressing either Vps35 WT-GFP or Vps35 P316S-GFP at 30 minutes post-chase (Figure 5.6E, arrows). 96

However, in cells expressing Vps35 R524W-GFP, the majority of the p230-positive membranes did not display evidence of CD8 staining at 30 minutes post-chase and showed strong localization to dispersed punctate structures, reminiscent of endosomes, indicative of strong delays in the retrograde sorting pathway (Figure 5.6E). Based on these multiple assays, the expression of the PD-associated Vps35 R524W mutant retromer protein impacts significantly on the retrograde trafficking pathway and accumulates CI-M6PR in endosomes due to an inefficiency to transport proteins from endosomes to the TGN. 97

Figure 5.6. Expression of Vps35 R524W perturbs localization and sorting of CI-M6PR (A) Representative co-immunoprecipitation of GFP fusion constructs with CD8-M6PR reporter resolved by SDS-PAGE. Membranes were immunolabelled with anti-CD8. (B) Representative immunoblots of culture medium and cell lysates from HeLa cells transiently transfected with Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP and incubated with 100µg/mL of cyclohexamide for up to 7hrs (n=3). Membranes were probed with antibodies raised against β-tubulin and Cathepsin D. (C) Confocal analysis of HeLa cells transiently expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W immunolabeled with antibodies against endogenous CI-M6PR and counterstained with DAPI. (D) Graphical representation of co-localization observed in (C). Graph represents three independent experiments with 10 images per group with 5-7 transfected cells per field of view (n=3, Error bars displayed as +/- SEM; *p<0.01 by by ANOVA followed by Bonferroni multiple comparisons test). (E) Representative images of HeLa cells expressing CD8- CIM6PR and GFP fusion constructs at 30 minutes post chase with anti-CD8 and immunolabeled with anti-p230 (red) and anti-CD8 (green) antibodies.

5.1.7 SNX27-retromer dependent recycling of GLUT1 is unaffected in the presence of Vps35 R524W In addition to retrograde trafficking of cargo, retromer also functions to recycle proteins from endosomes to the plasma membrane by forming a complex with SNX27 [72, 116]. To determine if Vps35 P316S-GFP or Vps35 R524W-GFP expression impacts the retromer-mediated recycling pathway, we investigated the sub-cellular localization of endogenous SNX27 in transfected HeLa cells. Co-localization analysis revealed a moderate overlap between both Vps35 WT-GFP and SNX27 (R=0.3562) and Vps35 P316S-GFP with SNX27 (R = 0.3851), whereas analysis of Vps35 R524W-GFP with SNX27 demonstrated a marked reduction in co-localization (R=0.2659; Figure 5.7B). Membrane fractionation showed unchanged levels of SNX27 recruited to membranes (Supplementary Figure 3) and the sub-cellular localization and overall morphology of SNX27-positive endosomes in the presence of Vps35 R524W-GFP was not altered (Figure 5.7A). SNX27 co- immunoprecipitated with Vps35 WT-GFP or Vps35 P316S-GFP to similar levels, while less SNX27 was co- immunoprecipitated with Vps35 R524W-GFP (Figure 5.7C).

To determine if this decrease in SNX27 binding to Vps35 R524W retromer had a dominant negative effect on the recycling pathway as it does for the retrograde pathway we analysed GLUT1 as a representative and well-established SNX27-retromer dependent cargo [72, 95]. HeLa cells expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP were immunolabeled with antibodies raised against endogenous GLUT1 and LAMP1 (Figure 5.7D). 98

Confocal microscopy revealed localization of GLUT1 to both the cell surface and punctate structures in cells expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP, reminiscent of the neighbouring untransfected cells staining. Additionally, intracellular localization of GLUT1 demonstrated minimal overlap with late endosome marker, LAMP1, supporting uninterrupted recycling in the presence of Vps35 P316S-GFP or Vps35 R524W-GFP (Figure 7D). 99

Figure 5.7. Vps35 R524W does not disrupt SNX27-dependent cargo recycling (A) Immunofluorescence of HeLa cells transiently expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP immunolabelled with anti-SNX27 and co-localization quantification is represented as Pearson’s co-efficient. (B) Graphs representative of three independent experiments with 10 images per group with 5-7 transfected cells per field of view (n=3, Error bars displayed as +/- SEM; *p<0.05 by by ANOVA followed by Bonferroni multiple comparisons test). (C) Co-immunoprecipiation using GFP NanoTrap of GFP, Vps35 WT-GFP, Vps35 P316S-GFP and Vps35 R524W-GFP with endogenous SNX27. (D) Representative immunofluorescence images of HeLa cells expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP followed by co- immunolabeling with anti-GLUT1 and anti-LAMP1. (E) Representative Western blots (n=3) of HeLa cells expressing GFP fusion constructs showing biotinylation of cell surface and total GLUT1 levels. (F) Graphical demonstration of glucose uptake in HeLa cells transiently transfected with GFP fusion constructs. Data representative of 3 independent experiments conducted in triplicate. Error bars represent +/- SEM.

To further confirm these results, we labeled the plasma membrane of transfected cells with biotin and used streptavidin-based precipitation to assess the cell surface levels of GLUT1 and TfnR. Biotinolyation revealed similar cell surface levels of SNX27-retromer dependent cargo, GLUT1 and retromer independent receptor, TfnR, in cells expressing GFP, Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP. We also find Vps35 D620N expression does not impact recycling of GLUT1 (Figure 5.7E), consistent with previous reports [268]. In support to this finding, HeLa cells expressing all GFP fusion constructs demonstrated identical levels of glucose uptake compared to controls and was consistent with GLUT1 biotinylation observations (Figure 5.7F). Therefore, the reduced levels of Vps35 R524W-GFP from SNX27-positive endosomes does not appear to negatively impact the recycling function of retromer.

100

5.2 Discussion

While the PD linked point mutations Vps35 P316S and Vps35 R524W do not destabilize the retromer’s high affinity trimer structure, the presence of the R524W mutation does modulate retromer’s cellular functions and induces α-synuclein inclusion formation, which is an underlying hallmark of PD. Despite genetic evidence linking this mutation to late onset PD, expression of Vps35 P316S did not display any distinct cellular phenotypes when compared to expression of Vps35 WT. Vps35 R524W demonstrated poor recruitment to the endosome and negatively impacted the interaction of retromer with several well-established retromer interacting proteins. This mislocalization of Vps35 R524W-containing retromer resulted in reduced endosome-to-Golgi but not endosome-to-plasma membrane receptor sorting. This was demonstrated by several CI-M6PR trafficking assays where we show the redistribution of endogenous CI-M6PR from the perinuclear space to dispersed puncta, late endosome swelling and detectable levels of pre-cathepsin D in the culture medium which was mirrored by the decrease in cathepsin D processing in the cells. Finally, we demonstrate that loss of endogenous retromer or expression of the R524W mutation leads to increased levels of α-synuclein inclusions, a phenotype which can be modulated by the addition of the previously described retromer stabilizing agent, R55 [160].

Previous studies demonstrate the recruitment of additional endosomal machinery such as the WASH complex, TBC1D5 and members of the SNX-BAR to the endosome to assist retromer in mediating sorting of its cargo to the TGN [133, 138, 297]. However, it is the combination of the small GTPase Rab7a and PX-containing protein SNX3 that is thought to initiate this cascade by recruiting retromer to the membrane [18, 99]. The current view of retromer mediated transport is that the Vps35-Vps26-Vps29 proteins form a stable core trimer that is a hub for associating with regulatory and cargo proteins [130]. Our data shows that although the core complex remains unaffected by the P316S and R524W mutations, the association with regulatory molecules is significantly altered in the presence of Vps35 R524W mutation. Previously, Vps35-Rab7a-SNX3 interaction interface studies demonstrated that binding of Rab7a and SNX3 is within the first 300 amino acids of Vps35 and immediately adjacent to the Vps26-Vps35 interface [101]. Despite this, we showed that R524W but not the P316S mutation within Vps35 severely impacts the retromer-Rab7a interaction and Vps35 R524W-retromer membrane association, coupled with reduced interactions with the WASH complex subunit FAM21, the RabGAP TBC1D5 and PDZ motif 101 cargo adaptor SNX27. While Vps35 R524W-retromer displayed impaired binding to multiple retromer-associated proteins this is not likely due to the mutation directly impacting on each of the binding surfaces for these molecules. Interestingly, the use of R55, a recently described retromer stabilizing agent [160], showed a rescue phenotype in membrane association of the Vps35 R524W-retromer, possibly by stabilizing a conformational change needed for retromer-Rab7a interaction required for membrane recruitment. Therefore, these findings suggest that the primary defect in Vps35 R524W arises from its diminished interaction with machinery needed to recruit retromer to the membrane and not from any differences in the initial formation of the trimer.

Given the described role of CI-M6PR in trafficking of cathepsin D, a protease needed for clearance of α-synuclein [244, 280], and the importance of retromer in correct localization of the CI-M6PR and sorting of lysosomal cathepsin D [70, 71, 312], it is not surprising that a stable loss of retromer or expression of Vps35 R524W leads to the increased presence of α-synuclein positive inclusions. Also, studies show a complete knockout of cathepsin D leads to extensive levels of high molecular weight α-synuclein species and increase in Lewy Body numbers in brains of knockout animals [281]. Together, these findings support the concept that regulation of the lysosomal pathway and its luminal content plays a fundamental role in turnover of proteins, a pathway that is already highly susceptible to errors during ageing [239]. In order to examine the role of retromer in α-synuclein aggregate production and clearance, we used a non-chemical based stimuli (KCl) rather than other reagents (Rotenone, Bafilomycin A1) which target mitochondrial and lysosomal function, respectively, and may indirectly impact retromer’s cellular function. Using this assay, our findings indicate that the expression of Vps35 R524W leads to CI-M6PR and cathepsin D mistrafficking, suggesting a direct correlation between the loss of retromer, the Vps35 R524W point mutation and the production of α-synuclein inclusions in a cellular system. Interestingly, another previously described PD linked Vps35 mutation (Vps35 D620N) was also reported to impair the degradation of α-synuclein, a product of diminished endo- lysosomal functionality, further supporting a relationship between receptor trafficking and α- synuclein aggregate production [270]. Several lines of evidence also demonstrate the requirement of both Rab7 and TBC1D5 in maintaining constant autophagic flux, retromer- dependent trafficking and optimal function of the lysosomal compartment [47, 102, 281, 297]. Although these studies do not conclusively demonstrate a direct link between the retromer complex, autophagic flux itself and formation of α-synuclein aggregates, they do 102 support the emerging notion that retromer-dependent machinery is required for functionality of the endolysosomal-autophagy clearance pathways.

The interaction between SNX27 PDZ domain and Vps26 subunit of the retromer complex has been characterized to be important in regulation of the recycling pathway [95], but their recruitment to the endosomal membrane appears to be independent of each other [72]. Here we confirm SNX27’s membrane association is not impaired in the presence of Vps35 R524W expression and also demonstrate recycling of GLUT1, a known PDZ motif- containing cargo (see supplementary figure 3), was identical to that observed in the presence of Vps35 WT-GFP expression. This may be due to the interaction of the endogenous SNX27 with the more abundant, endogenous retromer over the poorly recruited Vps35 R524W-retromer leading to the uninterrupted recycling of SNX27-dependent cargo observed here. Therefore, the PD-linked R524W mutation does not influence the recycling trafficking pathway in our experimental conditions, like the PD-linked Vps35 D620N mutation which appears to be similarly functional in GLUT1 recycling.

Overall, we have demonstrated that Vps35 R524W containing retromer poorly interacts with known endosomal machinery and consequently impacts the retrograde cargo sorting properties of the retromer complex and not its role in recycling SNX27-dependent cargo. Expression of Vps35 R524W is similar to the sorting defect witnessed for the Vps35 D620N mutation linked to PD in which, the CI-M6PR is not efficiently delivered to the TGN [267-269]. However, Vps35 D620N-containing retromer reportedly interacts with accessory proteins TBC1D5 and SNX27 identical to that witnessed for wild-type retromer but has minor reductions in its ability to interact with the WASH complex [268, 270], whereas Vps35 R524W has reduced interactions with TBC1D5, SNX27 and the WASH complex. Vps35 D620N broadly modifies endosome morphology [267], whereas Vps35 R524W expression alters late endosome morphology with minimal enrichment at the endosome, supporting very distinct changes to the molecular and cellular properties of the retromer sorting pathways in the presence of the two mutations. Despite these described differences between the D620N and R524W Vps35 PD linked mutations, it is clear that the retrograde sorting properties of retromer, and not its role in endosome to plasma membrane trafficking, is indeed defective in PD. Although this does not rule out Vps35 P316S as being causative in PD, it infers that the primary defect leading to the disease lies outside the immediate 103 retromer interactome described here, and may lead to disease through a currently undefined retromer-dependent mechanism.

We demonstrated that the implementation of a retromer stabilization drug, R55, was able to decrease the production of α-synuclein inclusions observed through the stable loss of retromer or by the expression of PD associated Vps35 R524W mutant. These preliminary observations indicate the potential to therapeutically modulate retromer function in PD, as has recently been proposed for Alzheimer’s disease [160]. Given the emerging evidence for the contribution of endosomes to the pathology of PD, it remains to be determined if other causes of PD can be modulated by increasing the stability of retromer and enhancing the cell’s natural clearance mechanisms.

104

Chapter 6: General discussion

This thesis has demonstrated how the previously described PD-linked Vps35 mutations (D620N, P316S and R524W) influence the sorting functions of retromer, modify the endocytic sorting and lead to the accumulation of non-degraded α-synuclein within the late endosomal network. Here, it will be argued that retromer and some of its interacting proteins play a key role in the onset of PD, how the intracellular sorting of receptors needs to be maintained for clearance of toxic proteins, and the emerging notion that retromer possesses a potential neuroprotective role.

6.1 Retromer as a regulator of disease Recent years has provided great insight into the role of retromer in maintaining a healthy cellular environment. For example, the retromer complex is essential for the transport of APP, the protein implicated in AD, away from the pathogenic processing pathway, and for correct localization of the AMPA receptor, one of many receptors needed for synaptic transmission [111, 117]. Given these retromer-driven regulatory processes, it is of little surprise that over the last decade several publications have also linked retromer to various pathophysiological conditions that have strong prevalence in the modern world.

In 2005, using a microarray analysis, Small et al reported loss of Vps35 and Vps26 gene expression in post mortem tissue harvested from patients with AD, the first indication that retromer may be a regulator of AD onset and progression [148]. Complementary to these observations was the use of an RNAi system in cell culture that demonstrated a profound effect on pathogenic Aβ production following modulation of retromer expression [148, 313]. For example, Bhalla et al demonstrated that the loss of retromer selectively increases the pathogenic processing of Amyloid pre-cursor protein (APP) into its disease causing isoform, Aβ42, establishing that retromer is fundamental to the trafficking of APP [313]. Interestingly, despite this strong link between the retromer complex and APP trafficking, APP is not a direct retromer cargo. Instead, APP intracellular trafficking requires SorLA, an established retromer- cargo and APP adaptor protein [111]. Currently two distinct routes have been proposed for the enzymatic processing of APP. Under non-pathogenic conditions, APP is first cleaved at the cell surface by α-secretase, giving rise to soluble APP and a C-terminal fragment-α (CTFα), the latter of which is endocytosed. Following internalization and sorting through the endosomal network, the CTFα is then targeted by the 105

γ-secretase complex for further processing and ultimately degraded via the lysosome. Under periods of pathogenesis, or perturbed protein sorting, increased amounts of full length APP are endocytosed and retained within the endocytic network. As α-secretase is concentrated at the cell surface, this allows for the endosomal localized β-secretase to cleave APP, producing a CTFβ. From here, the CTFβ is processed by Υ-secretase, giving rise to the pathogenic Aβ42 peptide [158, 159].

In the absence of retromer, SorLA is incorrectly localized within the cell, indirectly leading to increased endosomal localization of APP, which in turn, promotes the pathogenic processing by the β-secretase [111]. Support for this pathway following the loss of retromer lies in the observation that Vps26 heterozygote (+/-) knockout mice display elevated production of toxic Aβ42 and severe impairment in memory maze experiments used to test function of hippocampus, the anatomical region of the brain susceptible to AD [146]. Together, these observations further support the idea that retromer plays a protective role in disease cellular homeostasis and may regulate the onset of disease.

6.1.1 Regulator of mechanisms involved in PD manifestation Within recent years several point mutations (see section 1.4 and 1.4.1) in the Vps35 subunit of retromer found to be linked to PD have been described, along with subsequent mutations in Vps26A and Vps29 subunits [256-259, 262, 264]. Overall, it is clear that retromer plays an important role in maintaining the flow of receptors within a cell. Disruption to this trafficking either by Vps35 mutations linked to PD or loss of retromer expression gives rise to cumulative phenotypes that are witnessed in tissue isolated from PD patients, such as LB formation and lysosomal abnormalities, supporting the pathogenicity underpinned by Vps35 D620N and Vps35 R524W described in this thesis using a cellular system.

Here, only the Vps35 D620N and R524W mutations have been characterized and the primary defect witnessed in these disease causing mutation arises from perturbed cargo sorting, as described in Chapter 4 and 5, and not as a direct influence on the formation of the retromer complex itself, as previously predicted [256]. Surprisingly, ITC data confirmed that the presence of Vps35 D620N, P316S or R524W mutations do not influence the formation of the high affinity trimer, supporting a functional loss as opposed to a structural deficit (see Chapters 4 and 5). In line with the findings presented in Chapter 4 of this thesis is the observation that over-expression of Vps35 D620N in primary cortical neurons strongly hindered endosome motility when compared to Vps35 WT expression [314], suggesting a 106 possible underlying mechanism for impaired sorting of cargo. Consistent with the findings of Follett et al, the authors also demonstrate increased localization of Vps35 D620N-containing retromer with early endosomes. Prolonged association of retromer with the endosome would delay, or potentially block, the sorting of internalized cargo to various cellular organelles, further supporting impaired cargo sorting as a means of pathogenicity in PD. In contrast, expression of Vps35 R524W was localized primarily to the cytosol and likely obstructs the recruitment of retromer-dependent machinery needed for the retrieval of receptors.

While characterization of Vps35 D620N and R524W-containing retromer demonstrated impaired cargo sorting properties as marked by mislocalization of CI-M6PR, endosomal abnormalities and secretion of cathepsin D, exactly how these mutations impair receptor sorting is yet to be demonstrated.

As the retromer complex lacks any currently identified membrane-targeting domains, its ability to associate with and disassociate from endosomal membranes is governed by several interacting proteins. The small GTPase Rab7A and the PX-containing SNX3 mediated recruitment of retromer to the endosome, whereas disassociation from the membrane is achieved via the RabGAP, TBC1D5 [18]. As Vps35 D620N-containing retromer is readily recruited to the endosomal membrane it is likely that interactions with positive regulators, SNX3 and Rab7A, are maintained, but this is yet to be investigated. Moreover, interactions between Vps35 D620N-containing retromer that are dependent on retromer’s endosomal localization (SNX27 and TBC1D5) are reportedly sustained [268], though diminished interactions have been observed with FAM21 [268]. Vps35 P316S- retromer is readily recruited to early and late endocytic compartments and has preserved interactions with Rab7A, TBC1D5, FAM21 and SNX27 (see Chapter 5), potentially inferring that pathogenicity of this mutation is not from a loss of membrane association mediated by these protein-protein interactions. Conversely, Vps35 R524W-containing retromer has diminished interactions with Rab7A, SNX27, TBC1D5 and FAM21 subunit of the WASH complex, consequently impairing recruitment of retromer to the endosome. Despite that D620N and R524W Vps35 mutations display very distinct cellular phenotypes, functional consequences appear consistent, including impaired retrograde cargo sorting, impaired sorting and localization of Cathepsin D, lysosomal swelling. As endosomal swelling is a broad indication of impaired sorting/impaired degradation and several aspects of trafficking may contribute to this phenotype, the specific loss of Cathepsin D trafficking is underpinned by a failure of Vps35 D620N- and Vps35 R524W-containing retromer to retrieve the CI- 107

M6PR to the TGN (see Chapters 4 and 5). Taken together it is apparent that despite the described mutations not impacting the formation of the retromer complex itself, the described sorting deficits and α-synuclein inclusion formation arise from an impaired endosomal pathways, phenotypes that are mediated by the protein-protein interactions between retromer and its regulators.

6.2 Lysosomes: a readout for disease Given the lysosome represents an endpoint in the endocytic network, perturbations in the cellular trafficking system often indirectly culminate with its gross morphological changes (enlarged luminal space) and compromised functionality (decreased enzyme activity, accumulation of non-degraded/aggregated product). Consistent with this notion, several lysosomal storage diseases arise from mutations in lysosomal enzymes or lysosomal associated proteins. For example, lysosomal dysfunction diseases such as Niemann-Pick disease or mucopolysaccharidosis type IIIB develop from mutations in genes that give rise to lysosomal localized enzymes which, over time, lead to a build-up of non- degraded product accumulating in the lumen of the lysosome [315, 316]. In Niemann-Pick disease A and B, Sphingomyelin phosphodiesterase-1, the enzyme responsible for the metabolism of Sphingomylein is deficient, leading to the accumulation of toxic fat waste in the lysosome [317]. Similarly, in Niemann-Pick Disease type 3 mutations in proteins needed for the transport of cholesterol from the endosomal compartments are defective. In healthy cells, low density lipoproteins are delivered to endosomes where they are metabolized, yielding free cholesterol. The newly generated free cholesterol is exported from the endocytic compartment and delivered to the plasma membrane. In NPC, cholesterol fails to exit the endosome and accumulates within the lysosome [318-320]. Consequence of this is exemplified by a reduction in the motility of cholesterol rich late endosomes and the accumulation of endocytosed receptors (e.g. CI-M6PR) in the lumen, signifying global impairment of exportation from endosomes [321-324].

Likewise, Mucopolysaccharidosis type IIIB (MPSIIIB), a lysosomal storage disease underpinned by the inability to clear heparan sulfate (HS), results in the accumulation of undigested oligosaccharides and onset of neurodegeneration. Using a mouse model of MPSIIIB, Vitry and colleagues demonstrated the strong accumulation of swollen LAMP1 positive compartments throughout mouse cortical neurons, which failed to receive endocytosed material [325]. Consistent with this finding, electron microscopy revealed 108 highly heterogeneous, distended membrane enclosed compartments which contained luminal content ranging from protein aggregates to other organelles [325]. Taken together, it is clear that the lysosome, and the functionality of its luminal content, plays an important role in not only in the continuous clearance of potential toxic protein species but the prevention and onset of many neurological pathologies.

Interestingly, these described changes observed in lysosome organelle morphology and function are frequently reported in other neurological pathologies such as AD and PD [326]. One example of cellular dysfunction in the pathogenesis of PD is the previously discussed link between the loss of retromer and lysosomal function [70, 102]. Under basal conditions, the lysosome is often spatially distinct from compartments positive for the retromer complex as is the majority of its cargo [32, 327]. However, silencing of not only retromer (Vps35), but other endocytic proteins (e.g. Rab7A, Rab7L1), or expression of PD- linked LRRK2 G2019S mutant, leads to changes in lysosomal morphology and content alongside notable trafficking defects, as marked by the absence of receptors localized to the Golgi-apparatus [269].Similarly, findings presented in this thesis detail how the impaired cargo sorting properties of retromer alter lysosomal morphology, sub-cellular localization and function following expression of either Vps35 D620N or Vps35 R524W. Interestingly, a recent paper by Hockey and colleagues reported significant changes in sub-cellular localization and morphology of lysosomes in fibroblasts obtained from PD patients with the familial LRRK2 G2019S mutation [328]. When compared to control samples, lysosomal compartments were approximately 2-fold greater in area and appeared clustered around the perinuclear space, consistent with observations following expression of Vps35 D620N [267].

Evidence of impaired lysosomal function and an inability to breakdown delivered product was witnessed following expression of either Vps35 D620N or Vps35 R524W in SH- SY5Y cells as marked by the generation of α-synuclein inclusions (Vps35 D620N data not presented in manuscript of results chapter 5). Here, we hypothesize that the formation of α- synuclein inclusions following the loss of retromer or expression of PD-causing Vps35 mutations arises from a failure in CI-M6PR trafficking from the endosome to the TGN. Consequence of this is the failed sorting of the aspartic protease cathepsin D to the late endosomal network. As α-synuclein is a substrate of cathepsin D, a loss of retromer sorting impairs the turnover of α-synuclein, leading to its accumulation in inclusions. This phenotype is consistent with the literature as cathepsin D localizes to lysosomal compartments, and 109 following its depletion mice, large amounts of insoluble high molecular weight α-synuclein are observed when compared to wild-type littermates [281, 282]. Interestingly, the observed disruption to M6PR/Cathepsin D trafficking following the expression of Vps35 D620N displays strong cellular phenotypes consistent with those reported for Inclusion-cell disease (I-cell disease). I-cell disease is a lysosomal storage disorder underpinned by the failure for a Golgi-localized phosphatase (N-acetylglucosamine-1-phosphate transferase) to conjugate phosphate moieties to mannose residues on specific protein [329-332]. Similar to the aforementioned modification on the Cathepsin D, tagging of Golgi-localized proteins with phosphate targets them for export to the lysosome. In the absence of a successful phosphorylation event, N-linked gylcoproteins are rerouted through the default secretory pathway and consequently found in the extracellular space or in blood samples isolated from patients [333]. The magnitude of impaired trafficking and the absence of functional lysosomal enzymes is the formation of large cytoplasmic inclusions enriched in fats, lipids and carbohydrates and eventual death of the patient.

The late endosomal accumulation of α-synuclein following retromer dysfunction is further supported by Sugeno and colleagues study that demonstrated that incubating cultured cells expressing either Rab5, Rab7 or Rab11 with fluorescently conjugated α- synuclein leads to its accumulation in Rab7-positive compartments, inferring that following entry into the endosomal lumen it is destined for degradation [292]. Moreover, the authors demonstrate that internalization of exogenous α-synuclein induces swelling of Rab7-positive compartments and not those marked by Rab5 or Rab11. Although no investigation into retromer’s localization is presented by Sugeno and colleagues, it is likely to also be partially affiliated with endosomes marked with Rab7A, as previously demonstrated [102].

Overall, despite α-synuclein not having a defined subcellular localization, it is clear that its entry into the lumen of the endosome will define its pathway into the lysosome and eventual degradation by specific lysosomal protease(s). Interestingly, as exemplified by lysosomal storage diseases, it is clear that unperturbed lysosomal function is fundamental to the continuous clearance of cellular proteins. Despite the fact that upstream blockages in receptor sorting or deficiencies in lysosomal enzymes lead to the accumulation of non- degraded protein species, it is clear that this product negatively influences general function and health of a neuron, ultimately giving rise to a broad range of neurodegenerative diseases. 110

6.3 Retromer interacting proteins and disease Over the last decade it has become evident that not only retromer, but also retromer interacting proteins contribute to the onset of neurological disease pathology. For example, study by Vardarajan et al, used genetic testing of 8,000 cases to demonstrate significant association between 50% of SNPs in both SNX3 and Rab7A and AD [107]. Although no functional analysis was presented in the paper, SNPs in the molecular machinery needed to drive the association of retromer with the endosomal membrane may act as a loss-of- function mutation and hinder receptor sorting. Despite studies demonstrating the segregation of genes that produce several retromer interacting proteins with AD, currently no data elucidates the potential for retromer interacting proteins in PD.

Interestingly, SNX3 and Rab7A are not alone as retromer interacting proteins that may underpin the onset of disease. Recent work by Wang et al demonstrated that retromer interacting protein, SNX27, has reduced expression in patients with Down’s syndrome [92]. As SNX27 is important in brain development, receptor recycling and axonal branching [92], a reduction in SNX27-retromer interaction, as witnessed in this thesis with the Vps35 R524W mutation, may contribute to trafficking changes and onset of PD in patients positive for the mutation. However, as reported in this thesis, expression of Vps35 D620N, Vps35 R524W or Vps35 P316S does not impact the recycling of GLUT1, an identified SNX27-dependent cargo [72]. In contrast, Munsie and colleagues recently reported that expression of Vps35 D620N impairs synaptic recycling of the SNX27-dependent cargo, AMPAR, in cortical neurons [314]. These differences in recycling properties may arise from the selectivity of receptors or the context in which they were studied. For example, MacLeod and colleagues reported reduced axonal length following expression of Vps35 D620N in primary rat cortical neurons [269]. Reductions in axonal length would limit connections made between pre and post-synaptic neurons and, as AMPA receptors are important for mediating synaptic transmission/communication, reduced neuronal connections may lead to impaired or down- regulated recycling of specific receptors under extended periods of synaptic depression or in a disease state. Although this does demonstrate that GLUT1 and AMPAR are cargo of the retromer complex, it may indicate that several other factors determine the function of retromer, such as cell viability, under periods of toxicity/pathogenicity, as is witnessed following expression of Vps35 D620N [271].

111

Moreover, LRRK2, another retromer interacting protein, is one of the major contributors to Familial PD through various mutations, possibly acting in a toxic loss-of- function manner [334]. Recently, MacLeod and colleagues reported a novel interaction between the Vps35 subunit of retromer and LRRK2. Indeed this relationship appears to be concentrated at the endosomal surface as the expression of PD-linked kinase mutant, LRRK2-G2019S, blocked the sorting of the CI-M6PR to the TGN and induced lysosome swelling [269]. Interestingly, co-expression of LRRK2-G2019S with Vps35-WT was capable of rescuing the aforementioned retrograde sorting defect, suggesting that LRRK2 may potentially regulate the degradation and or function of the retromer complex. However, this observation appears to be between LRRK2 and retromer and not a result of a particular disease-causing mutant as both LRRK2-WT transgenic mice or R1441C mutants display significant loss of retromer expression. It is clear that SNX3, Rab7A and TBC1D5 regulate retromer’s function by governing its membrane association. If LRRK2 were also capable of regulating retromer’s membrane association, the CI-M6PR would be degraded at a greater rate than the controls but instead, displays increase localization to cell surface when compared to the controls [269]. Therefore, it is likely that despite LRRK2 being a potential retromer interacting protein, this interaction may in fact occur independent of membrane localization and thus not directly regulate the sorting function of retromer, or be an indirect association.

Although it is clear that the retromer complex plays a fundamental role in the correct localization of internalized receptors, it is evident that the precise function of retromer is governed by different interacting proteins that are capable of orchestrating a multitude of functions carried out by retromer itself. Therefore, these interactions ultimately determine the function of retromer and may provide novel insight into how the mislocalization of receptors occurs in PD as demonstrated in this thesis.

6.4 Retromer mediated Neuroprotection As retromer is strongly implicated in several neurodegenerative diseases through loss-of-function mutations or a decrease in gene expression, it is interesting to consider that retromer may have a neuroprotective role within the brain. For example, using a PD transgenic Drosophila models expressing LRRK2 mutations, Linhart and colleagues demonstrate that the over-expression of the wild-type Vps35 in DA neurons significantly increased the lifespan of the flies [335]. Furthermore, the introduction of exogenous Vps35 112 or Vps26 on a PD-linked LRRK2 mutant background improved Drosophila lifespan following treatment with the mitochondrial toxin Rotenone, inferring that retromer can rescue phenotypes induced by expression of PD-linked LRRK2 mutations and may also play an important role in PD caused by exposure to neurotoxic agents [335]. This is further supported by Bi and colleagues where expression of wild-type Vps35 but not the pathogenic D620N mutation in primary rat neurons isolated from the midbrain significantly increased cell viability following exposure to mitochondrial toxin MPP+ [336]. Whether the loss of neuronal viability witnessed with the D620N mutation is also observed in other Vps35 mutations, such as R524W, is currently unknown, however recent work by Wang et al, 2014 suggests that expression of Vps35 P316S modestly impacts DA neuron viability in Drosophila [261]. Although neuronal viability was not tested in this thesis, consistent with these findings is the loss of mature Cathepsin D and an increase in α-synuclein inclusions following expression of Vps35 P316S (see Chapter 5) compared to Vps35 WT expression, consistent with idea that this mutation may have a weak level of pathogenicity and compromise neuronal viability. Taken together, these studies show loss of retromer or expression of PD-causing Vps35 mutations leads to decrease of the overall neuronal health and, therefore, contributes to the onset of disease pathogenicity.

Another aspect of compromised neuronal health may arise from mitochondria- derived vesicles (MDVs), small vesicles that shuttle mitochondrial proteins to various organelles such as the lysosome and peroxisome. Interestingly, recent work from Braschi et al demonstrated a role for retromer in transporting MDVs to the peroxisome, facilitating the trafficking of a mitochondrial protein ligase [337]. It is unlikely that this pathway is lost in the presence of the Vps35 mutations but instead, that overexpression of pathogenic mutations delay the turnover of material derived from mitochondrial membranes, such as hydrogen peroxide, which is reduced following fusion with the peroxisome [338]. Additionally, as expression of Vps35 D620N disrupted endosomal localization and induced lysosome swelling, it is possible the mitochondrial-derived vesicles are mislocalized and fail to fuse with the peroxisome. Moreover, as expression of Vps35 D620N and Vps35 R524W perturb lysosome function, it is also possible that turnover of damaged mitochondria (mitophagy) via the auto-lysosome pathway is delayed, thus decreasing the turnover of damaged organelles resulting in increased cellular stress and compromised neuronal viability. 113

As a whole, it is clear that the health and viability of a neuron determines the lifespan of an organism. The molecular pathway controlled by retromer in its potential function as a positive regulator of cell viability and overall health of an organism, is currently unknown. However it is possible that by elevating the total amount of retromer protein, the cell is capable of enduring greater amounts of exogenous stress, organelle damage, and ultimately increased cellular lifespan.

6.5 Pharmacological modulators of Retromer As growing evidence supports the PD-linked Vps35 mutations act as a loss-of- function mutation, strong interest into pharmacologically rescuing these perturbations has emerged. As discussed in the introduction to this thesis and section 5.1.4 of chapter 5, a chaperone, called R55, capable of increasing the total protein level of retromer by stabilization of the complex was recently described [160]. Through this action, R55 is capable of enhancing the cargo sorting capacity of retromer and consequently decreasing the pathogenic processing of APP [160]. Consistent with this study, this thesis shows that use of R55 increased the total Vps35 protein level and limited the formation of α-synuclein inclusions, even following the stable loss of retromer using a shRNA system (See section 3.1.1 and Figure 3.1). Furthermore, R55 treatment increased endosomal levels of Vps35 R524W and the total number of cells positive for α-synuclein inclusions was reduced, inferring increased stability of Vps35 R524W-containing retromer. Therefore, it is clear that the loss of retromer function impairs the cells’ ability to clear α-synuclein inclusions that arise as a result of impaired protease sorting and that this phenotype can be partially reversed by increasing expression level of retromer. However, whether this observation can be translated into model organisms or human patients is currently unknown.

Though the underlying mechanism as to why Vps35 R524W does not interact with most of the known interacting partners was not described in this thesis, the loss of Vps35 R524W-retromer’s membrane association may be explained by the previous in vitro study which characterized a conformational shift within the complex at the initial stages of the trimer formation [277]. Given that the formation of the retromer trimer is a prerequisite for its interaction with both Rab7A and the endosomal membrane [101, 276], one explanation for the observations reported here could be influenced by the capacity of the Vps35 R524W- retromer to undergo this conformational change, rather than impacting on the binding surface of a particular molecular interaction. For example, if following the formation of the 114 retromer complex, the Vps35-Vps29 interface undergoes a conformational shift to allow for an “open/active confirmation”, exposing previously hidden residues, this complex is then recruited to the endosome via Rab7A. Use of the described R55 drug may potentially stabilize retromer in its “open” confirmation by engaging the now exposed residues and thus promote increased endosomal localization. In contrast, following a shift into an “open/active state”, the presence of a Tryptophan at residue 524 may cause local instability of the complex which results in a shift back into a “closed” confirmation. This consequently leading to decreased interaction with Rab7A/Fam21/TBC1D5 and thus predominately cytosolic localization. Docking of R55 with Vps35 R524W may promote partial stabilization once an “open” state is initially achieved, giving the ability to partially rescue the mislocalization of Vps35 R524W-GFP reported in this thesis.

Additionally, whether the trafficking defect observed following the expression of Vps35 D620N could be rescued following incubation with R55 is currently unknown. However, as Vps35 D620N-containing retromer is recruited to endosomes [267], it is likely that employment of R55 would further promote endosomal association, which may in turn compensate for the loss-of-function phenotype described in chapter 3 of this thesis.

Though the use of pharmaceutical agents to target the function of retromer appears promising as a potential mechanism to reverse neurodegenerative phenotypes, it is unlikely that employment of molecular chaperones in a whole organism will give a desirable and symptom free result. As demonstrated in this thesis and discussion, minor adjustments to the function, localization and expression of retromer can lead to detrimental phenotypes observed in PD and AD. As such, use of a brain specific delivery system to either A) correct the loss of retromer witnessed in AD or B) rescue the mislocalization of Vps35 R524W- containing retromer in PD (See section 5.1.5 and figure 5.5) may grant partial longevity to patients of these disease. However, as demonstrated in chapter 1, production of α-synuclein inclusions following the loss of retromer are only partially reversible following employment of R55, and as such, it is unlikely that the onset of pathology phenotypes (protein aggregation/neuron health) will be completely rescued in a patient.

6.6 Future perspectives There is now intense interest in understanding the role of retromer dysfunction in PD. Genetic screening of PD patients will determine if additional disease causing mutations exist in the retromer complex or the retromer-associated proteins. As the molecular detail of the 115 impact of these mutations is elucidated, the membrane trafficking or cellular pathways impacted will also need to be determined. Association of these cellular properties with the clinical and pathological properties of individuals diagnosed with PD should enable sub- classification of these retromer-associated defects into the various forms of PD and other synucleinopathies. Second to investigating the gross morphological properties classically witnessed in PD patients, it still remains unknown whether these individuals harbour additional point mutations in retromer-associated proteins or other PD-associated genes. For example, two individual families from Sweden diagnosed with early onset AD were found to harbour two point mutations in the APP gene, now collectively referred to as APPSwe, supporting the notion of co-segregation of disease causing mutations [339]. Interestingly, this type of screen has previously been reported using AD patients; genotyping known retromer interacting proteins. Significant mutations were reported in retromer interacting proteins, such as SNX3 and Rab7A, demonstrating additional risk genes in the onset on AD [107]. Whether this is the same for individuals with PD harbouring Vps35 mutations is currently unknown.

For now, however, a fundamental question that needs to be addressed is whether expression of Vps35 D620N increases the rate, size and pathogenicity of α-synuclein inclusion formation. This is commonly investigated using cell culture systems over- expressing α-synuclein followed by treatment with various cellular stresses or inducers of aggregation, such as KCl, H202, ionomycin, rotenone or bafilomycin [210]. Unfortunately this may not accurately represent the biology of a neuron, the environment or the time frame of disease progression. It is doubtful that retromer traffics α-synuclein by direct protein-protein interaction, but more likely influences its degradation by controlling normal lysosomal and autophagic clearance mechanisms. Previous proteomic screening analyses using LB isolated from cortical tissue of post-mortem PD patients identified retromer subunit, Vps35, as a component of the aggregate [294]. Though this does not definitively demonstrate retromer as a component of LB, it does support the potential for trafficking proteins being redirected to local sites of protein accumulation or signalling events. Currently, if retromer associates with LB structures found within in the substantia nigra (not just cortical LBs), or if levels of retromer decline with age is currently unknown. Analysis of retromer in brain samples from human patients with either idiopathic PD or PD-linked G2019S LRRK2 mutation revealed reduced levels of Vps35 [271]. This suggests that the loss of retromer is likely to be secondary to broader defects in endosomal trafficking. 116

Recent years have seen a boom in the process of genetically transdifferentiating fibroblasts derived from patients into neurons. While such cell lines provide excellent models for further study, genetically modifying either rat or mouse genomes to express Vps35 variants, such as D620N, will be more relevant models to assess the impact of the Vps35 variants on cellular trafficking and neurodegeneration. They will open new avenues to examine Vps35-linked pathogenesis in the context of a whole organism - similarly to a recent publication using Drosophila, which reported reduced life-span of the flies expressing human Vps35 D620N [261].

Current PD therapeutic approaches attempt to increase total levels of the neurotransmitter dopamine, which is effective only for relatively short periods. One potential route towards targeting retromer-mediated transport in PD (and AD) was recently suggested by a study that demonstrated a small molecule, R55, that stabilizes retromer is able to increase retromer levels in cells and decrease processing of APP to the toxic amyloid peptide [160]. Given that cathepsin D is a lysosome protease that can cleave α-synuclein, it is conceivable that treatment with R55 will facilitate the perturbed membrane transport pathways, ultimately facilitating increased degradation of α-synuclein. Although treatment with such ‘molecular chaperone’ compounds will be limited until the causative link between retromer and α-synuclein aggregation is fully understood, it provides hope that retromer- targeted therapies may be possible in the future.

117

Chapter 7: References

1. Schu, P., et al., Phosphatidylinositol 3-kinase encoded by yeast VPS34 gene essential for protein sorting. Science, 1993. 260(5104): p. 88-91. 2. Stack, J.H., et al., A membrane-associated complex containing the Vps15 protein kinase and the Vps34 PI 3-kinase is essential for protein sorting to the yeast lysosome-like vacuole. Embo J, 1993. 12(5): p. 2195-204. 3. Christoforidis, S., et al., Phosphatidylinositol-3-OH kinases are Rab5 effectors. Nat Cell Biol, 1999. 1(4): p. 249-252. 4. Murray, J.T., et al., Role of Rab5 in the Recruitment of hVps34/p150 to the Early Endosome. Traffic, 2002. 3(6): p. 416-427. 5. Horiuchi, H., et al., A Novel Rab5 GDP/GTP Exchange Factor Complexed to Rabaptin-5 Links Nucleotide Exchange to Effector Recruitment and Function. Cell, 1997. 90(6): p. 1149- 1159. 6. Lippe, R., et al., Functional synergy between Rab5 effector Rabaptin-5 and exchange factor Rabex-5 when physically associated in a complex. Mol Biol Cell, 2001. 12(7): p. 2219-28. 7. Rink, J., et al., Rab conversion as a mechanism of progression from early to late endosomes. Cell, 2005. 122(5): p. 735-49. 8. Stenmark, H., et al., Inhibition of rab5 GTPase activity stimulates membrane fusion in endocytosis. Embo j, 1994. 13(6): p. 1287-96. 9. Hirota, Y., et al., A role for Rab5 activity in the biogenesis of endosomal and lysosomal compartments. Biochemical and Biophysical Research Communications, 2007. 364(1): p. 40- 47. 10. Huotari, J. and A. Helenius, Endosome maturation. Embo j, 2011. 30(17): p. 3481-500. 11. Del Conte-Zerial, P., et al., Membrane identity and GTPase cascades regulated by toggle and cut-out switches. Molecular Systems Biology, 2008. 4: p. 206. 12. Wurmser, A.E., T.K. Sato, and S.D. Emr, New component of the vacuolar class C-Vps complex couples nucleotide exchange on the Ypt7 GTPase to SNARE-dependent docking and fusion. J Cell Biol, 2000. 151(3): p. 551-62. 13. Peralta, E.R., B.C. Martin, and A.L. Edinger, Differential effects of TBC1D15 and mammalian Vps39 on Rab7 activation state, lysosomal morphology, and growth factor dependence. J Biol Chem, 2010. 285(22): p. 16814-21. 14. Nordmann, M., et al., The Mon1-Ccz1 complex is the GEF of the late endosomal Rab7 homolog Ypt7. Curr Biol, 2010. 20(18): p. 1654-9. 15. Sun, Q., et al., Rubicon controls endosome maturation as a Rab7 effector. Proc Natl Acad Sci U S A, 2010. 107(45): p. 19338-43. 16. Cozier, G.E., et al., The phox homology (PX) domain-dependent, 3-phosphoinositide- mediated association of sorting nexin-1 with an early sorting endosomal compartment is required for its ability to regulate epidermal growth factor receptor degradation. J Biol Chem, 2002. 277(50): p. 48730-6. 17. Zhong, Q., et al., Endosomal localization and function of sorting nexin 1. Proc Natl Acad Sci U S A, 2002. 99(10): p. 6767-72. 18. Seaman, M.N., et al., Membrane recruitment of the cargo-selective retromer subcomplex is catalysed by the small GTPase Rab7 and inhibited by the Rab-GAP TBC1D5. J Cell Sci, 2009. 122(Pt 14): p. 2371-82. 19. Gary, J.D., et al., Fab1p is essential for PtdIns(3)P 5-kinase activity and the maintenance of vacuolar size and membrane homeostasis. J Cell Biol, 1998. 143(1): p. 65-79. 118

20. Sbrissa, D., O.C. Ikonomov, and A. Shisheva, Phosphatidylinositol 3-phosphate-interacting domains in PIKfyve. Binding specificity and role in PIKfyve. Endomenbrane localization. J Biol Chem, 2002. 277(8): p. 6073-9. 21. Shisheva, A., PIKfyve: PARTNERS, SIGNIFICANCE, DEBATES AND PARADOXES. Cell biology international, 2008. 32(6): p. 591-604. 22. Ikonomov, O.C., et al., Active PIKfyve associates with and promotes the membrane attachment of the late endosome-to-trans-Golgi network transport factor Rab9 effector p40. J Biol Chem, 2003. 278(51): p. 50863-71. 23. Rutherford, A.C., et al., The mammalian phosphatidylinositol 3-phosphate 5-kinase (PIKfyve) regulates endosome-to-TGN retrograde transport. J Cell Sci, 2006. 119(Pt 19): p. 3944-57. 24. Bright, N.A., M.J. Gratian, and J.P. Luzio, Endocytic Delivery to Lysosomes Mediated by Concurrent Fusion and Kissing Events in Living Cells. Current Biology, 2005. 15(4): p. 360- 365. 25. Griffiths, G., et al., The mannose 6-phosphate receptor and the biogenesis of lysosomes. Cell, 1988. 52(3): p. 329-341. 26. Luzio, J.P., P.R. Pryor, and N.A. Bright, Lysosomes: fusion and function. Nat Rev Mol Cell Biol, 2007. 8(8): p. 622-632. 27. Caplan, S., et al., Human Vam6p promotes lysosome clustering and fusion in vivo. J Cell Biol, 2001. 154(1): p. 109-22. 28. Poupon, V., et al., The role of mVps18p in clustering, fusion, and intracellular localization of late endocytic organelles. Mol Biol Cell, 2003. 14(10): p. 4015-27. 29. Antonin, W., et al., A SNARE complex mediating fusion of late endosomes defines conserved properties of SNARE structure and function. Embo j, 2000. 19(23): p. 6453-64. 30. Pryor, P.R., et al., Combinatorial SNARE complexes with VAMP7 or VAMP8 define different late endocytic fusion events. EMBO Rep, 2004. 5(6): p. 590-5. 31. Antonin, W., et al., Crystal structure of the endosomal SNARE complex reveals common structural principles of all SNAREs. Nature structural biology, 2002. 9(2): p. 107-11. 32. Gomez, T.S., et al., Trafficking defects in WASH-knockout fibroblasts originate from collapsed endosomal and lysosomal networks. Mol Biol Cell, 2012. 23(16): p. 3215-28. 33. Lamb, C.A., T. Yoshimori, and S.A. Tooze, The autophagosome: origins unknown, biogenesis complex. Nat Rev Mol Cell Biol, 2013. 14(12): p. 759-774. 34. Monastyrska, I., et al., Multiple roles of the cytoskeleton in autophagy. Biological reviews of the Cambridge Philosophical Society, 2009. 84(3): p. 431-48. 35. Chan, E.Y., et al., Kinase-inactivated ULK proteins inhibit autophagy via their conserved C- terminal domains using an Atg13-independent mechanism. Mol Cell Biol, 2009. 29(1): p. 157-71. 36. Wirth, M., J. Joachim, and S.A. Tooze, Autophagosome formation--the role of ULK1 and Beclin1-PI3KC3 complexes in setting the stage. Seminars in cancer biology, 2013. 23(5): p. 301-9. 37. Furuya, N., et al., The evolutionarily conserved domain of Beclin 1 is required for Vps34 binding, autophagy and tumor suppressor function. Autophagy, 2005. 1(1): p. 46-52. 38. Zeng, X., J.H. Overmeyer, and W.A. Maltese, Functional specificity of the mammalian Beclin-Vps34 PI 3-kinase complex in macroautophagy versus endocytosis and lysosomal enzyme trafficking. J Cell Sci, 2006. 119(Pt 2): p. 259-70. 39. Kim, J., et al., Convergence of multiple autophagy and cytoplasm to vacuole targeting components to a perivacuolar membrane compartment prior to de novo vesicle formation. J Biol Chem, 2002. 277(1): p. 763-73. 40. Mizushima, N., et al., Mouse Apg16L, a novel WD-repeat protein, targets to the autophagic isolation membrane with the Apg12-Apg5 conjugate. J Cell Sci, 2003. 116(Pt 9): p. 1679-88. 119

41. Kabeya, Y., et al., LC3, a mammalian homologue of yeast Apg8p, is localized in autophagosome membranes after processing. Embo j, 2000. 19(21): p. 5720-8. 42. Matsushita, M., et al., Structure of Atg5·Atg16, a Complex Essential for Autophagy. Journal of Biological Chemistry, 2007. 282(9): p. 6763-6772. 43. Mizushima, N., et al., Dissection of autophagosome formation using Apg5-deficient mouse embryonic stem cells. J Cell Biol, 2001. 152(4): p. 657-68. 44. Xie, Z. and D.J. Klionsky, Autophagosome formation: core machinery and adaptations. Nat Cell Biol, 2007. 9(10): p. 1102-9. 45. Darsow, T., S.E. Rieder, and S.D. Emr, A multispecificity syntaxin homologue, Vam3p, essential for autophagic and biosynthetic protein transport to the vacuole. J Cell Biol, 1997. 138(3): p. 517-29. 46. von Mollard, G.F. and T.H. Stevens, The Saccharomyces cerevisiae v-SNARE Vti1p Is Required for Multiple Membrane Transport Pathways to the Vacuole. Mol Biol Cell, 1999. 10(6): p. 1719-32. 47. Jager, S., et al., Role for Rab7 in maturation of late autophagic vacuoles. J Cell Sci, 2004. 117(Pt 20): p. 4837-48. 48. Huynh, K.K., et al., LAMP proteins are required for fusion of lysosomes with phagosomes. Embo j, 2007. 26(2): p. 313-24. 49. Cuervo, A.M., et al., Impaired degradation of mutant alpha-synuclein by chaperone-mediated autophagy. Science, 2004. 305(5688): p. 1292-5. 50. Cooper, A.A., et al., Alpha-synuclein blocks ER-Golgi traffic and Rab1 rescues neuron loss in Parkinson's models. Science (New York, N.Y.), 2006. 313(5785): p. 324-8. 51. Kegel, K.B., et al., Huntingtin expression stimulates endosomal-lysosomal activity, endosome tubulation, and autophagy. J Neurosci, 2000. 20(19): p. 7268-78. 52. Petersen, A., et al., Expanded CAG repeats in exon 1 of the Huntington's disease gene stimulate dopamine-mediated striatal neuron autophagy and degeneration. Hum Mol Genet, 2001. 10(12): p. 1243-54. 53. Martinez-Vicente, M., et al., Cargo recognition failure is responsible for inefficient autophagy in Huntington's disease. Nat Neurosci, 2010. 13(5): p. 567-76. 54. Ravikumar, B., et al., Inhibition of mTOR induces autophagy and reduces toxicity of polyglutamine expansions in fly and mouse models of Huntington disease. Nat Genet, 2004. 36(6): p. 585-95. 55. Wong, Y.C. and E.L. Holzbaur, The regulation of autophagosome dynamics by huntingtin and HAP1 is disrupted by expression of mutant huntingtin, leading to defective cargo degradation. J Neurosci, 2014. 34(4): p. 1293-305. 56. del Toro, D., et al., Mutant Huntingtin Impairs Post-Golgi Trafficking to Lysosomes by Delocalizing Optineurin/Rab8 Complex from the Golgi Apparatus. Mol Biol Cell, 2009. 20(5): p. 1478-92. 57. Loubery, S., et al., Different microtubule motors move early and late endocytic compartments. Traffic, 2008. 9(4): p. 492-509. 58. Hoepfner, S., et al., Modulation of receptor recycling and degradation by the endosomal kinesin KIF16B. Cell, 2005. 121(3): p. 437-50. 59. Jordens, I., et al., The Rab7 effector protein RILP controls lysosomal transport by inducing the recruitment of dynein-dynactin motors. Curr Biol, 2001. 11(21): p. 1680-5. 60. Johansson, M., et al., Activation of endosomal dynein motors by stepwise assembly of Rab7- RILP-p150Glued, ORP1L, and the receptor betalll spectrin. J Cell Biol, 2007. 176(4): p. 459- 71. 61. Wassmer, T., et al., The retromer coat complex coordinates endosomal sorting and dynein- mediated transport, with carrier recognition by the trans-Golgi network. Dev Cell, 2009. 17(1): p. 110-22. 120

62. Horazdovsky, B.F., et al., A sorting nexin-1 homologue, Vps5p, forms a complex with Vps17p and is required for recycling the vacuolar protein-sorting receptor. Molecular Biology of the Cell, 1997. 8(8): p. 1529-41. 63. Seaman, M.N., J.M. McCaffery, and S.D. Emr, A membrane coat complex essential for endosome-to-Golgi retrograde transport in yeast. The Journal of Cell Biology, 1998. 142(3): p. 665-81. 64. Seaman, M.N.J., et al., Endosome to Golgi Retrieval of the Vacuolar Protein Sorting Receptor, Vps10p, Requires the Function of the VPS29, VPS30, and VPS35 Gene Products. The Journal of Cell Biology, 1997. 137(1): p. 79-92. 65. Haft, C.R., et al., Human orthologs of yeast vacuolar protein sorting proteins Vps26, 29, and 35: assembly into multimeric complexes. Mol Biol Cell, 2000. 11(12): p. 4105-16. 66. Seaman, M.N., The retromer complex - endosomal protein recycling and beyond. Journal of cell science, 2012. 125(Pt 20): p. 4693-702. 67. Bugarcic, A., et al., Vps26A and Vps26B subunits define distinct retromer complexes. Traffic, 2011. 12(12): p. 1759-73. 68. Collins, B.M., The structure and function of the retromer protein complex. Traffic, 2008. 9(11): p. 1811-22. 69. Kerr, M.C., et al., A novel mammalian retromer component, Vps26B. Traffic, 2005. 6(11): p. 991-1001. 70. Arighi, C.N., et al., Role of the mammalian retromer in sorting of the cation-independent mannose 6-phosphate receptor. J Cell Biol, 2004. 165(1): p. 123-33. 71. Seaman, M.N., Identification of a novel conserved sorting motif required for retromer- mediated endosome-to-TGN retrieval. J Cell Sci, 2007. 120(Pt 14): p. 2378-89. 72. Steinberg, F., et al., A global analysis of SNX27-retromer assembly and cargo specificity reveals a function in glucose and metal ion transport. Nature cell biology, 2013. 15(5): p. 461-71. 73. Tabuchi, M., et al., Retromer-mediated direct sorting is required for proper endosomal recycling of the mammalian iron transporter DMT1. J Cell Sci, 2010. 123(Pt 5): p. 756-66. 74. Belenkaya, T.Y., et al., The Retromer Complex Influences Wnt Secretion by Recycling Wntless from Endosomes to the Trans-Golgi Network. Developmental Cell, 2008. 14(1): p. 120-131. 75. Fjorback, A.W., et al., Retromer Binds the FANSHY Sorting Motif in SorLA to Regulate Amyloid Precursor Protein Sorting and Processing. The Journal of Neuroscience, 2012. 32(4): p. 1467-1480. 76. Nielsen, M.S., et al., Sorting by the cytoplasmic domain of the amyloid precursor protein binding receptor SorLA. Mol Cell Biol, 2007. 27(19): p. 6842-51. 77. Wassmer, T., et al., A loss-of-function screen reveals SNX5 and SNX6 as potential components of the mammalian retromer. J Cell Sci, 2007. 120(Pt 1): p. 45-54. 78. Coudreuse, D.Y., et al., Wnt gradient formation requires retromer function in Wnt-producing cells. Science, 2006. 312(5775): p. 921-4. 79. Yang, P.-T., et al., Wnt Signaling Requires Retromer-Dependent Recycling of MIG- 14/Wntless in Wnt-Producing Cells. Developmental Cell, 2008. 14(1): p. 140-147. 80. Pan, C.-L., et al., C. elegans AP-2 and Retromer Control Wnt Signaling by Regulating MIG- 14/Wntless. Developmental Cell, 2008. 14(1): p. 132-139. 81. Franch-Marro, X., et al., Wingless secretion requires endosome-to-Golgi retrieval of Wntless/Evi/Sprinter by the retromer complex. Nat Cell Biol, 2008. 10(2): p. 170-177. 82. Harterink, M., et al., A SNX3-dependent retromer pathway mediates retrograde transport of the Wnt sorting receptor Wntless and is required for Wnt secretion. Nat Cell Biol, 2011. 13(8): p. 914-23. 83. Zhang, P., et al., SNX3 controls Wingless/Wnt secretion through regulating retromer- dependent recycling of Wntless. Cell Res, 2011. 21(12): p. 1677-90. 121

84. Strochlic, T.I., et al., Grd19/Snx3p functions as a cargo-specific adapter for retromer- dependent endocytic recycling. J Cell Biol, 2007. 177(1): p. 115-25. 85. Chen, D., et al., Retromer Is Required for Apoptotic Cell Clearance by Phagocytic Receptor Recycling. Science, 2010. 327(5970): p. 1261-1264. 86. Pocha, Shirin M., et al., Retromer Controls Epithelial Cell Polarity by Trafficking the Apical Determinant Crumbs. Current Biology, 2011. 21(13): p. 1111-1117. 87. Zhou, B., Y. Wu, and X. Lin, Retromer regulates apical-basal polarity through recycling Crumbs. Dev Biol, 2011. 360(1): p. 87-95. 88. Cai, L., et al., Deficiency of sorting nexin 27 (SNX27) leads to growth retardation and elevated levels of N-methyl-D-aspartate receptor 2C (NR2C). Mol Cell Biol, 2011. 31(8): p. 1734-47. 89. Hussain, N.K., et al., Sorting Nexin 27 regulates basal and activity-dependent trafficking of AMPARs. Proc Natl Acad Sci U S A, 2014. 111(32): p. 11840-5. 90. Loo, L.S., et al., A role for sorting nexin 27 in AMPA receptor trafficking. Nature communications, 2014. 5: p. 3176. 91. Lunn, M.L., et al., A unique sorting nexin regulates trafficking of potassium channels via a PDZ domain interaction. Nat Neurosci, 2007. 10(10): p. 1249-59. 92. Wang, X., et al., Loss of sorting nexin 27 contributes to excitatory synaptic dysfunction by modulating glutamate receptor recycling in Down's syndrome. Nature medicine, 2013. 19(4): p. 473-80. 93. Lauffer, B.E., et al., SNX27 mediates PDZ-directed sorting from endosomes to the plasma membrane. J Cell Biol, 2010. 190(4): p. 565-74. 94. Ghai, R., et al., Structural basis for endosomal trafficking of diverse transmembrane cargos by PX-FERM proteins. Proceedings of the National Academy of Sciences of the United States of America, 2013. 110(8): p. E643-52. 95. Gallon, M., et al., A unique PDZ domain and arrestin-like fold interaction reveals mechanistic details of endocytic recycling by SNX27-retromer. Proc Natl Acad Sci U S A, 2014. 111(35): p. 18. 96. Balana, B., et al., Mechanism underlying selective regulation of G protein-gated inwardly rectifying potassium channels by the psychostimulant-sensitive sorting nexin 27. Proc Natl Acad Sci U S A, 2011. 108(14): p. 5831-6. 97. Hesketh, G.G., et al., VARP is recruited on to endosomes by direct interaction with retromer, where together they function in export to the cell surface. Dev Cell, 2014. 29(5): p. 591-606. 98. McGough, I.J., et al., Identification of molecular heterogeneity in SNX27-retromer-mediated endosome-to-plasma membrane recycling. J Cell Sci, 2014. 2: p. 156299. 99. Harrison, M.S., et al., A mechanism for retromer endosomal coat complex assembly with cargo. Proc Natl Acad Sci U S A, 2014. 111(1): p. 267-72. 100. Liu, T.T., et al., Rab GTPase regulation of retromer-mediated cargo export during endosome maturation. Mol Biol Cell, 2012. 23(13): p. 2505-15. 101. Priya, A., et al., Molecular insights into Rab7-mediated endosomal recruitment of core retromer: Deciphering the role of Vps26 and Vps35. Traffic, 2014. 102. Rojas, R., et al., Regulation of retromer recruitment to endosomes by sequential action of Rab5 and Rab7. J Cell Biol, 2008. 183(3): p. 513-26. 103. Rojas, R., et al., Interchangeable but essential functions of SNX1 and SNX2 in the association of retromer with endosomes and the trafficking of mannose 6-phosphate receptors. Mol Cell Biol, 2007. 27(3): p. 1112-24. 104. Zolov, S.N., et al., In vivo, Pikfyve generates PI(3,5)P2, which serves as both a signaling lipid and the major precursor for PI5P. Proc Natl Acad Sci U S A, 2012. 109(43): p. 17472-7. 105. Ikonomov, O.C., D. Sbrissa, and A. Shisheva, Mammalian cell morphology and endocytic membrane homeostasis require enzymatically active phosphoinositide 5-kinase PIKfyve. J Biol Chem, 2001. 276(28): p. 26141-7. 122

106. Zhang, Y., et al., Modulation of synaptic function by VAC14, a protein that regulates the phosphoinositides PI(3,5)P(2) and PI(5)P. Embo J, 2012. 31(16): p. 3442-56. 107. Vardarajan, B.N., et al., Identification of Alzheimer disease-associated variants in genes that regulate retromer function. Neurobiol Aging, 2012. 33(9): p. 5. 108. Arighi, C.N., et al., Role of the mammalian retromer in sorting of the cation-independent mannose 6-phosphate receptor. The Journal of Cell Biology, 2004. 165(1): p. 123-133. 109. Cuartero, Y., et al., Retromer regulates postendocytic sorting of beta-secretase in polarized Madin-Darby canine kidney cells. Traffic, 2012. 13(10): p. 1393-410. 110. Nothwehr, S.F. and A.E. Hindes, The yeast VPS5/GRD2 gene encodes a sorting nexin-1-like protein required for localizing membrane proteins to the late Golgi. J Cell Sci, 1997. 110(Pt 9): p. 1063-72. 111. Fjorback, A.W., et al., Retromer binds the FANSHY sorting motif in SorLA to regulate amyloid precursor protein sorting and processing. J Neurosci, 2012. 32(4): p. 1467-80. 112. Seaman, M.N.J., Cargo-selective endosomal sorting for retrieval to the Golgi requires retromer. The Journal of Cell Biology, 2004. 165(1): p. 111-122. 113. Pandey, K.N., Small peptide recognition sequence for intracellular sorting. Curr Opin Biotechnol, 2010. 21(5): p. 611-20. 114. Canuel, M., et al., AP-1 and retromer play opposite roles in the trafficking of sortilin between the Golgi apparatus and the lysosomes. Biochem Biophys Res Commun, 2008. 366(3): p. 724-30. 115. Seaman, M.N.J., Identification of a novel conserved sorting motif required for retromer- mediated endosome-to-TGN retreival. J. Cell Sci., 2007. 120: p. 2378-2389. 116. Temkin, P., et al., SNX27 mediates retromer tubule entry and endosome-to-plasma membrane trafficking of signalling receptors. Nature cell biology, 2011. 13(6): p. 715-21. 117. Choy, R.W., et al., Retromer mediates a discrete route of local membrane delivery to dendrites. Neuron, 2014. 82(1): p. 55-62. 118. Arighi, C.N., et al., Role of the mammalian retromer in sorting of the cation-independent mannose 6-phosphate receptor. J. Cell Biol., 2004. 165(1): p. 123-33. 119. Miras, S.L., et al., The giardial VPS35 retromer subunit is necessary for multimeric complex assembly and interaction with the vacuolar protein sorting receptor. Biochim Biophys Acta, 2013. 12(38): p. 26. 120. Lane, R.F., et al., Diabetes-associated SorCS1 regulates Alzheimer's amyloid-beta metabolism: evidence for involvement of SorL1 and the retromer complex. J Neurosci, 2010. 30(39): p. 13110-5. 121. Zhou, G.L., et al., Deficits in receptor-mediated endocytosis and recycling in cells from mice with Gpr107 locus disruption. J Cell Sci, 2014. 127(Pt 18): p. 3916-27. 122. Feinstein, T.N., et al., Noncanonical control of vasopressin receptor type 2 signaling by retromer and arrestin. J Biol Chem, 2013. 288(39): p. 27849-60. 123. Feinstein, T.N., et al., Retromer terminates the generation of cAMP by internalized PTH receptors. Nat Chem Biol, 2011. 7(5): p. 278-84. 124. del Pino, I., et al., The trafficking proteins Vacuolar Protein Sorting 35 and Neurobeachin interact with the glycine receptor beta-subunit. Biochem Biophys Res Commun, 2011. 412(3): p. 435-40. 125. Gleason, R.J., et al., BMP signaling requires retromer-dependent recycling of the type I receptor. Proceedings of the National Academy of Sciences, 2014. 111(7): p. 2578-2583. 126. Yin, X., et al., Retromer maintains basolateral distribution of the type II TGF-beta receptor via the recycling endosome. Mol Biol Cell, 2013. 24(14): p. 2285-98. 127. Verges, M., et al., The mammalian retromer regulates transcytosis of the polymeric immunoglobulin receptor. Nat Cell Biol, 2004. 6(8): p. 763-9. 123

128. Kang, Y., et al., A combined transgenic proteomic analysis and regulated trafficking of neuroligin-2. J Biol Chem, 2014. 289(42): p. 29350-64. 129. Jia, D., et al., WASH and WAVE actin regulators of the Wiskott-Aldrich syndrome protein (WASP) family are controlled by analogous structurally related complexes. Proc Natl Acad Sci U S A, 2010. 107(23): p. 10442-7. 130. Harbour, M.E., et al., The cargo-selective retromer complex is a recruiting hub for protein complexes that regulate endosomal tubule dynamics. J Cell Sci, 2010. 123(Pt 21): p. 3703- 17. 131. Harbour, M.E., S.Y. Breusegem, and M.N. Seaman, Recruitment of the endosomal WASH complex is mediated by the extended 'tail' of Fam21 binding to the retromer protein Vps35. Biochem J, 2012. 442(1): p. 209-20. 132. Jia, D., et al., Multiple repeat elements within the FAM21 tail link the WASH actin regulatory complex to the retromer. Mol Biol Cell, 2012. 23(12): p. 2352-61. 133. Gomez, T.S. and D.D. Billadeau, A FAM21-Containing WASH Complex Regulates Retromer- Dependent Sorting. Developmental Cell, 2009. 17(5): p. 699-711. 134. Puthenveedu, M.A., et al., Sequence-Dependent Sorting of Recycling Proteins by Actin- Stabilized Endosomal Microdomains. Cell, 2010. 143(5): p. 761-773. 135. Derivery, E., et al., The Arp2/3 activator WASH controls the fission of endosomes through a large multiprotein complex. Dev Cell, 2009. 17(5): p. 712-23. 136. Gokool, S., D. Tattersall, and M.N. Seaman, EHD1 interacts with retromer to stabilize SNX1 tubules and facilitate endosome-to-Golgi retrieval. Traffic, 2007. 8(12): p. 1873-86. 137. Rapaport, D., et al., Recycling to the plasma membrane is delayed in EHD1 knockout mice. Traffic, 2006. 7(1): p. 52-60. 138. Hong, Z., et al., The retromer component SNX6 interacts with dynactin p150(Glued) and mediates endosome-to-TGN transport. Cell Res, 2009. 19(12): p. 1334-49. 139. Carlton, J., et al., Sorting nexin-1 mediates tubular endosome-to-TGN transport through coincidence sensing of high- curvature membranes and 3-phosphoinositides. Curr Biol, 2004. 14(20): p. 1791-800. 140. Mallard, F., et al., Early/recycling endosomes-to-TGN transport involves two SNARE complexes and a Rab6 isoform. J Cell Biol, 2002. 156(4): p. 653-64. 141. Mukhopadhyay, A., et al., An endocytic pathway as a target of tubby for regulation of fat storage. EMBO Rep, 2007. 8(10): p. 931-8. 142. Cooney, J.R., et al., Endosomal compartments serve multiple hippocampal dendritic spines from a widespread rather than a local store of recycling membrane. J Neurosci, 2002. 22(6): p. 2215-24. 143. Anggono, V. and R.L. Huganir, Regulation of AMPA receptor trafficking and synaptic plasticity. Current Opinion in Neurobiology, 2012. 22(3): p. 461-469. 144. Anggono, V., et al., PICK1 interacts with PACSIN to regulate AMPA receptor internalization and cerebellar long-term depression. Proc Natl Acad Sci U S A, 2013. 110(34): p. 13976-81. 145. Mejias, R., et al., Gain-of-function glutamate receptor interacting protein 1 variants alter GluA2 recycling and surface distribution in patients with autism. Proc Natl Acad Sci U S A, 2011. 108(12): p. 4920-5. 146. Muhammad, A., et al., Retromer deficiency observed in Alzheimer's disease causes hippocampal dysfunction, neurodegeneration, and Aβ accumulation. Proceedings of the National Academy of Sciences, 2008. 105(20): p. 7327-7332. 147. Bhalla, A., et al., The location and trafficking routes of the neuronal retromer and its role in amyloid precursor protein transport. Neurobiology of Disease, 2012. 47(1): p. 126-134. 148. Small, S.A., et al., Model-guided microarray implicates the retromer complex in Alzheimer's disease. Ann Neurol, 2005. 58(6): p. 909-19. 124

149. Zheng, H. and E.H. Koo, Biology and pathophysiology of the amyloid precursor protein. Mol Neurodegener, 2011. 6(1): p. 1750-1326. 150. Xu, H., et al., Generation of Alzheimer beta-amyloid protein in the trans-Golgi network in the apparent absence of vesicle formation. Proc Natl Acad Sci U S A, 1997. 94(8): p. 3748-52. 151. Greenfield, J.P., et al., Endoplasmic reticulum and trans-Golgi network generate distinct populations of Alzheimer beta-amyloid peptides. Proc Natl Acad Sci U S A, 1999. 96(2): p. 742-7. 152. Nordstedt, C., et al., Identification of the Alzheimer beta/A4 amyloid precursor protein in clathrin-coated vesicles purified from PC12 cells. J Biol Chem, 1993. 268(1): p. 608-12. 153. Caporaso, G.L., et al., Morphologic and biochemical analysis of the intracellular trafficking of the Alzheimer beta/A4 amyloid precursor protein. J Neurosci, 1994. 14(5 Pt 2): p. 3122- 38. 154. Cai, H., et al., BACE1 is the major beta-secretase for generation of Abeta peptides by neurons. Nat Neurosci, 2001. 4(3): p. 233-4. 155. Roberds, S.L., et al., BACE knockout mice are healthy despite lacking the primary beta- secretase activity in brain: implications for Alzheimer's disease therapeutics. Hum Mol Genet, 2001. 10(12): p. 1317-24. 156. Luo, Y., et al., BACE1 (beta-secretase) knockout mice do not acquire compensatory gene expression changes or develop neural lesions over time. Neurobiol Dis, 2003. 14(1): p. 81-8. 157. Jiang, S., et al., Trafficking regulation of proteins in Alzheimer's disease. Mol Neurodegener, 2014. 9(6): p. 1750-1326. 158. Haass, C., et al., Trafficking and proteolytic processing of APP. Cold Spring Harb Perspect Med, 2012. 2(5). 159. Haass, C., Take five--BACE and the gamma-secretase quartet conduct Alzheimer's amyloid beta-peptide generation. Embo J, 2004. 23(3): p. 483-8. 160. Mecozzi, V.J., et al., Pharmacological chaperones stabilize retromer to limit APP processing. Nat Chem Biol, 2014. 10(6): p. 443-9. 161. Sannerud, R., et al., ADP ribosylation factor 6 (ARF6) controls amyloid precursor protein (APP) processing by mediating the endosomal sorting of BACE1. Proc Natl Acad Sci U S A, 2011. 108(34): p. 8. 162. Zhao, Y., et al., Sorting nexin 12 interacts with BACE1 and regulates BACE1-mediated APP processing. Mol Neurodegener, 2012. 7(30): p. 1750-1326. 163. Wen, L., et al., VPS35 haploinsufficiency increases Alzheimer's disease neuropathology. J Cell Biol, 2011. 195(5): p. 765-79. 164. Small, S.A. and S. Gandy, Sorting through the Cell Biology of Alzheimer's Disease: Intracellular Pathways to Pathogenesis. Neuron, 2006. 52(1): p. 15-31. 165. Andersen, O.M., et al., Neuronal sorting protein-related receptor sorLA/LR11 regulates processing of the amyloid precursor protein. Proc Natl Acad Sci U S A, 2005. 102(38): p. 13461-6. 166. Spoelgen, R., et al., Interaction of the cytosolic domains of sorLA/LR11 with the amyloid precursor protein (APP) and beta-secretase beta-site APP-cleaving enzyme. J Neurosci, 2006. 26(2): p. 418-28. 167. Gustafsen, C., et al., Sortilin and SorLA display distinct roles in processing and trafficking of amyloid precursor protein. J Neurosci, 2013. 33(1): p. 64-71. 168. Sager, K.L., et al., Neuronal LR11/sorLA expression is reduced in mild cognitive impairment. Ann Neurol, 2007. 62(6): p. 640-7. 169. Parkinson, J., An essay on the shaking palsy. 1817. The Journal of neuropsychiatry and clinical neurosciences, 2002. 14(2): p. 223-36; discussion 222. 170. Schultz, W., Activity of dopamine neurons in the behaving primate. Seminars in Neuroscience, 1992. 4(2): p. 129-138. 125

171. Bear, M.F., B.W. Connors, and M.A. Paradiso, Neuroscience : exploring the brain. 2007, Philadelphia, PA: Lippincott Williams & Wilkins. 172. Schultz, W., Multiple reward signals in the brain. Nat Rev Neurosci, 2000. 1(3): p. 199-207. 173. Francois, C., et al., Dopaminergic cell group A8 in the monkey: anatomical organization and projections to the striatum. The Journal of comparative neurology, 1999. 414(3): p. 334-47. 174. Galpern, W.R. and A.E. Lang, Interface between tauopathies and synucleinopathies: a tale of two proteins. Ann Neurol, 2006. 59(3): p. 449-58. 175. Baba, M., et al., Aggregation of alpha-synuclein in Lewy bodies of sporadic Parkinson's disease and dementia with Lewy bodies. The American journal of pathology, 1998. 152(4): p. 879-84. 176. Schulz-Schaeffer, W.J., The synaptic pathology of α-synuclein aggregation in dementia with Lewy bodies, Parkinson’s disease and Parkinson’s disease dementia. Acta Neuropathologica, 2010. 120(2): p. 131-43. 177. Mrak, R.E. and W.S.T. Griffin, Dementia with Lewy bodies: Definition, diagnosis, and pathogenic relationship to Alzheimer’s disease. Neuropsychiatric Disease and Treatment, 2007. 3(5): p. 619-25. 178. Brion, J.P. and A.M. Couck, Cortical and brainstem-type Lewy bodies are immunoreactive for the cyclin-dependent kinase 5. The American journal of pathology, 1995. 147(5): p. 1465- 76. 179. Spillantini, M.G., et al., [alpha]-Synuclein in Lewy bodies. Nature, 1997. 388(6645): p. 839- 840. 180. Mezey, E., et al., Alpha synuclein is present in Lewy bodies in sporadic Parkinson's disease. Molecular psychiatry, 1998. 3(6): p. 493-9. 181. Ulmer, T.S., et al., Structure and dynamics of micelle-bound human alpha-synuclein. J Biol Chem, 2005. 280(10): p. 9595-603. 182. Xu, J., et al., Dopamine-dependent neurotoxicity of alpha-synuclein: a mechanism for selective neurodegeneration in Parkinson disease. Nat Med, 2002. 8(6): p. 600-6. 183. Kirik, D., et al., Parkinson-like neurodegeneration induced by targeted overexpression of alpha-synuclein in the nigrostriatal system. J Neurosci, 2002. 22(7): p. 2780-91. 184. da Costa, C.A., K. Ancolio, and F. Checler, Wild-type but not Parkinson's disease-related ala-53 --> Thr mutant alpha -synuclein protects neuronal cells from apoptotic stimuli. J Biol Chem, 2000. 275(31): p. 24065-9. 185. Hashimoto, M., et al., alpha-Synuclein protects against oxidative stress via inactivation of the c-Jun N-terminal kinase stress-signaling pathway in neuronal cells. J Biol Chem, 2002. 277(13): p. 11465-72. 186. Seo, J.H., et al., Alpha-synuclein regulates neuronal survival via Bcl-2 family expression and PI3/Akt kinase pathway. Faseb j, 2002. 16(13): p. 1826-8. 187. Arrasate, M., et al., Inclusion body formation reduces levels of mutant huntingtin and the risk of neuronal death. Nature, 2004. 431(7010): p. 805-810. 188. Sidhu, A., et al., The role of alpha-synuclein in both neuroprotection and neurodegeneration. Annals of the New York Academy of Sciences, 2004. 1035: p. 250-70. 189. Uéda, K., et al., Molecular cloning of cDNA encoding an unrecognized component of amyloid in Alzheimer disease. Proc Natl Acad Sci U S A, 1993. 90(23): p. 11282-6. 190. Ltic, S., et al., Alpha-synuclein is expressed in different tissues during human fetal development. Journal of molecular neuroscience : MN, 2004. 22(3): p. 199-204. 191. George, J.M., et al., Characterization of a novel protein regulated during the critical period for song learning in the zebra finch. Neuron, 1995. 15(2): p. 361-72. 192. Maroteaux, L., J.T. Campanelli, and R.H. Scheller, Synuclein: a neuron-specific protein localized to the nucleus and presynaptic nerve terminal. J Neurosci, 1988. 8(8): p. 2804-15. 126

193. Bisaglia, M., et al., Structure and topology of the non-amyloid-β component fragment of human α-synuclein bound to micelles: Implications for the aggregation process. Protein Science, 2006. 15(6): p. 1408-1416. 194. McClendon, S., C.C. Rospigliosi, and D. Eliezer, Charge neutralization and collapse of the C-terminal tail of alpha-synuclein at low pH. Protein science : a publication of the Protein Society, 2009. 18(7): p. 1531-40. 195. Nielsen, M.S., et al., Ca2+ binding to alpha-synuclein regulates ligand binding and oligomerization. J Biol Chem, 2001. 276(25): p. 22680-4. 196. Nath, S., et al., Raised calcium promotes alpha-synuclein aggregate formation. Mol Cell Neurosci, 2011. 46(2): p. 516-26. 197. Davidson, W.S., et al., Stabilization of alpha-synuclein secondary structure upon binding to synthetic membranes. J Biol Chem, 1998. 273(16): p. 9443-9. 198. Zhu, M., et al., Alpha-synuclein can function as an antioxidant preventing oxidation of unsaturated lipid in vesicles. Biochemistry, 2006. 45(26): p. 8135-42. 199. Eliezer, D., et al., Conformational properties of alpha-synuclein in its free and lipid- associated states. Journal of molecular biology, 2001. 307(4): p. 1061-73. 200. Bussell, R., Jr. and D. Eliezer, A structural and functional role for 11-mer repeats in alpha- synuclein and other exchangeable lipid binding proteins. Journal of molecular biology, 2003. 329(4): p. 763-78. 201. Chandra, S., et al., A broken alpha -helix in folded alpha -Synuclein. J Biol Chem, 2003. 278(17): p. 15313-8. 202. Robotta, M., et al., Direct Evidence of Coexisting Horseshoe and Extended Helix Conformations of Membrane-Bound Alpha-Synuclein. ChemPhysChem, 2011. 12(2): p. 267- 269. 203. Narayanan, V. and S. Scarlata, Membrane binding and self-association of alpha-synucleins. Biochemistry, 2001. 40(33): p. 9927-34. 204. Clayton, D.F. and J.M. George, Synucleins in synaptic plasticity and neurodegenerative disorders. Journal of neuroscience research, 1999. 58(1): p. 120-9. 205. Lee, S.J., H. Jeon, and K.V. Kandror, Alpha-synuclein is localized in a subpopulation of rat brain synaptic vesicles. Acta Neurobiol Exp, 2008. 68(4): p. 509-15. 206. Zhang, L., et al., Semi-quantitative analysis of alpha-synuclein in subcellular pools of rat brain neurons: an immunogold electron microscopic study using a C-terminal specific monoclonal antibody. Brain Res, 2008. 9: p. 40-52. 207. Cabin, D.E., et al., Synaptic vesicle depletion correlates with attenuated synaptic responses to prolonged repetitive stimulation in mice lacking alpha-synuclein. J Neurosci, 2002. 22(20): p. 8797-807. 208. Burre, J., et al., Alpha-synuclein promotes SNARE-complex assembly in vivo and in vitro. Science, 2010. 329(5999): p. 1663-7. 209. Ostrerova-Golts, N., et al., The A53T alpha-synuclein mutation increases iron-dependent aggregation and toxicity. J Neurosci, 2000. 20(16): p. 6048-54. 210. Follett, J., et al., Potassium depolarization and raised calcium induces alpha-synuclein aggregates. Neurotox Res, 2013. 23(4): p. 378-92. 211. Burre, J., M. Sharma, and T.C. Sudhof, Definition of a molecular pathway mediating alpha- synuclein neurotoxicity. J Neurosci, 2015. 35(13): p. 5221-32. 212. Uversky, V.N., et al., Stabilization of partially folded conformation during alpha-synuclein oligomerization in both purified and cytosolic preparations. J Biol Chem, 2001. 276(47): p. 43495-8. 213. Serpell, L.C., et al., Fiber diffraction of synthetic alpha-synuclein filaments shows amyloid- like cross-beta conformation. Proc Natl Acad Sci U S A, 2000. 97(9): p. 4897-902. 127

214. Heise, H., et al., Molecular-level secondary structure, polymorphism, and dynamics of full- length alpha-synuclein fibrils studied by solid-state NMR. Proc Natl Acad Sci U S A, 2005. 102(44): p. 15871-6. 215. Vilar, M., et al., The fold of alpha-synuclein fibrils. Proc Natl Acad Sci U S A, 2008. 105(25): p. 8637-42. 216. Nelson, R., et al., Structure of the cross-beta spine of amyloid-like fibrils. Nature, 2005. 435(7043): p. 773-8. 217. Sawaya, M.R., et al., Atomic structures of amyloid cross-beta spines reveal varied steric zippers. Nature, 2007. 447(7143): p. 453-7. 218. Matthes, D., V. Gapsys, and B.L. de Groot, Driving forces and structural determinants of steric zipper peptide oligomer formation elucidated by atomistic simulations. Journal of molecular biology, 2012. 421(2-3): p. 390-416. 219. Conway, K.A., J.D. Harper, and P.T. Lansbury, Jr., Fibrils formed in vitro from alpha- synuclein and two mutant forms linked to Parkinson's disease are typical amyloid. Biochemistry, 2000. 39(10): p. 2552-63. 220. Crowther, R.A., S.E. Daniel, and M. Goedert, Characterisation of isolated alpha-synuclein filaments from substantia nigra of Parkinson's disease brain. Neuroscience letters, 2000. 292(2): p. 128-30. 221. Tofaris, G.K., et al., Ubiquitination of alpha-synuclein in Lewy bodies is a pathological event not associated with impairment of proteasome function. J Biol Chem, 2003. 278(45): p. 44405-11. 222. Emmanouilidou, E., L. Stefanis, and K. Vekrellis, Cell-produced alpha-synuclein oligomers are targeted to, and impair, the 26S proteasome. Neurobiol Aging, 2010. 31(6): p. 953-68. 223. Anderson, J.P., et al., Phosphorylation of Ser-129 is the dominant pathological modification of alpha-synuclein in familial and sporadic Lewy body disease. J Biol Chem, 2006. 281(40): p. 29739-52. 224. Engelender, S., Ubiquitination of alpha-synuclein and autophagy in Parkinson's disease. Autophagy, 2008. 4(3): p. 372-4. 225. Bennett, M.C., et al., Degradation of alpha-synuclein by proteasome. J Biol Chem, 1999. 274(48): p. 33855-8. 226. Webb, J.L., et al., Alpha-Synuclein is degraded by both autophagy and the proteasome. J Biol Chem, 2003. 278(27): p. 25009-13. 227. Ebrahimi-Fakhari, D., et al., Distinct roles in vivo for the ubiquitin-proteasome system and the autophagy-lysosomal pathway in the degradation of alpha-synuclein. J Neurosci, 2011. 31(41): p. 14508-20. 228. Snyder, H., et al., Aggregated and monomeric alpha-synuclein bind to the S6' proteasomal protein and inhibit proteasomal function. J Biol Chem, 2003. 278(14): p. 11753-9. 229. Tenreiro, S., et al., Phosphorylation modulates clearance of alpha-synuclein inclusions in a yeast model of Parkinson's disease. PLoS genetics, 2014. 10(5): p. e1004302. 230. Wang, J., et al., Impaired ubiquitin-proteasome system activity in the synapses of Huntington's disease mice. J Cell Biol, 2008. 180(6): p. 1177-89. 231. Bennett, E.J., et al., Global impairment of the ubiquitin-proteasome system by nuclear or cytoplasmic protein aggregates precedes inclusion body formation. Molecular cell, 2005. 17(3): p. 351-65. 232. Holmberg, C.I., et al., Inefficient degradation of truncated polyglutamine proteins by the proteasome. Embo j, 2004. 23(21): p. 4307-18. 233. Venkatraman, P., et al., Eukaryotic Proteasomes Cannot Digest Polyglutamine Sequences and Release Them during Degradation of Polyglutamine-Containing Proteins. Molecular cell, 2004. 14(1): p. 95-104. 128

234. Cuervo, A.M. and E. Wong, Chaperone-mediated autophagy: roles in disease and aging. Cell Res, 2014. 24(1): p. 92-104. 235. Chiang, H.L., et al., A role for a 70-kilodalton heat shock protein in lysosomal degradation of intracellular proteins. Science, 1989. 246(4928): p. 382-5. 236. Cuervo, A.M. and J.F. Dice, A receptor for the selective uptake and degradation of proteins by lysosomes. Science, 1996. 273(5274): p. 501-3. 237. Bandyopadhyay, U., et al., Identification of regulators of chaperone-mediated autophagy. Molecular cell, 2010. 39(4): p. 535-47. 238. Agarraberes, F.A., S.R. Terlecky, and J.F. Dice, An intralysosomal hsp70 is required for a selective pathway of lysosomal protein degradation. J Cell Biol, 1997. 137(4): p. 825-34. 239. Cuervo, A.M. and J.F. Dice, Age-related decline in chaperone-mediated autophagy. J Biol Chem, 2000. 275(40): p. 31505-13. 240. Zhang, C. and A.M. Cuervo, Restoration of chaperone-mediated autophagy in aging liver improves cellular maintenance and hepatic function. Nature medicine, 2008. 14(9): p. 959- 965. 241. Koga, H., et al., Constitutive upregulation of chaperone-mediated autophagy in Huntington's disease. J Neurosci, 2011. 31(50): p. 18492-505. 242. Erie, C., et al., Altered lysosomal positioning affects lysosomal functions in a cellular model of Huntington's disease. Eur J Neurosci, 2015. 42(3): p. 1941-51. 243. Liang, Q., et al., Reduction of mutant huntingtin accumulation and toxicity by lysosomal cathepsins D and B in neurons. Mol Neurodegener, 2011. 6: p. 37. 244. Hossain, S., et al., Limited proteolysis of NACP/alpha-synuclein. J Alzheimers Dis, 2001. 3(6): p. 577-584. 245. Gieselmann, V., A. Hasilik, and K. von Figura, Processing of human cathepsin D in lysosomes in vitro. Journal of Biological Chemistry, 1985. 260(5): p. 3215-3220. 246. Laurent-Matha, V., et al., Processing of Human Cathepsin D Is Independent of Its Catalytic Function and Auto-Activation: Involvement of Cathepsins L and B. Journal of biochemistry, 2006. 139(3): p. 363-371. 247. Hasilik, A., et al., Phosphorylated oligosaccharides in lysosomal enzymes: identification of alpha-N-acetylglucosamine(1)phospho(6)mannose diester groups. Proc Natl Acad Sci U S A, 1980. 77(12): p. 7074-8. 248. Höning, S., et al., The 46-kDa Mannose 6-Phosphate Receptor Contains Multiple Binding Sites for Clathrin Adaptors. Journal of Biological Chemistry, 1997. 272(32): p. 19884-19890. 249. Hentze, M., A. Hasilik, and K. Vonfigura, ENHANCED DEGRADATION OF CATHEPSIN- D SYNTHESIZED IN THE PRESENCE OF THE THREONINE ANALOG BETA- HYDROXYNORVALINE. Archives of Biochemistry and Biophysics, 1984. 230(1): p. 375- 382. 250. Samarel, A.M., et al., Effects of cysteine protease inhibitors on rabbit cathepsin D maturation. Am J Physiol, 1989. 257(6 Pt 1): p. C1069-79. 251. Riederer, M.A., et al., Lysosome biogenesis requires Rab9 function and receptor recycling from endosomes to the trans-Golgi network. J Cell Biol, 1994. 125(3): p. 573-82. 252. Lombardi, D., et al., Rab9 functions in transport between late endosomes and the trans Golgi network. Embo J, 1993. 12(2): p. 677-82. 253. Díaz, E., F. Schimmöller, and S.R. Pfeffer, A Novel Rab9 Effector Required for Endosome- to-TGN Transport. The Journal of Cell Biology, 1997. 138(2): p. 283-290. 254. Aivazian, D., R.L. Serrano, and S. Pfeffer, TIP47 is a key effector for Rab9 localization. The Journal of Cell Biology, 2006. 173(6): p. 917-926. 255. Carroll, K.S., et al., Role of Rab9 GTPase in Facilitating Receptor Recruitment by TIP47. Science, 2001. 292(5520): p. 1373-1376. 129

256. Zimprich, A., et al., A mutation in VPS35, encoding a subunit of the retromer complex, causes late-onset Parkinson disease. American journal of human genetics, 2011. 89(1): p. 168-75. 257. Ando, M., et al., VPS35 mutation in Japanese patients with typical Parkinson's disease. Mov Disord, 2012. 27(11): p. 1413-7. 258. Lesage, S., et al., Identification of VPS35 mutations replicated in French families with Parkinson disease. Neurology, 2012. 78(18): p. 1449-50. 259. Shannon, B., et al., Genetic variation of the retromer subunits VPS26A/B-VPS29 in Parkinson's disease. Neurobiol Aging, 2014. 35(8): p. 5. 260. Gustavsson, E.K., et al., Genetic variability of the retromer cargo recognition complex in parkinsonism. Mov Disord, 2014. 5(10): p. 26104. 261. WANG, H., et al., In vivo evidence of pathogenicity of VPS35 Mutations in the Drosophila. Molecular Brain, 2014. 7(1): p. 73. 262. Vilariño-Güell, C., et al., VPS35 Mutations in Parkinson Disease. The American Journal of Human Genetics, 2011. 89(1): p. 162-167. 263. Malik, B.R., V.K. Godena, and A.J. Whitworth, VPS35 pathogenic mutations confer no dominant toxicity but partial loss of function in Drosophila and genetically interact with parkin. Human Molecular Genetics, 2015. 264. Verstraeten, A., et al., Contribution of VPS35 genetic variability to LBD in the Flanders- Belgian population. Neurobiol Aging, 2012. 33(8): p. 14. 265. Sheerin, U.M., et al., Screening for VPS35 mutations in Parkinson's disease. Neurobiol Aging, 2012. 33(4): p. 7. 266. Koschmidder, E., et al., Mutations in VPS26A are not a frequent cause of Parkinson's disease. Neurobiology of Aging, 2014. 35(6): p. 1512.e1-1512.e2. 267. Follett, J., et al., The Vps35 D620N mutation linked to Parkinson's disease disrupts the cargo sorting function of retromer. Traffic, 2014. 15(2): p. 230-44. 268. McGough, I.J., et al., Retromer binding to FAM21 and the WASH complex is perturbed by the Parkinson disease-linked VPS35(D620N) mutation. Current biology : CB, 2014. 24(14): p. 1670-6. 269. MacLeod, D.A., et al., RAB7L1 interacts with LRRK2 to modify intraneuronal protein sorting and Parkinson's disease risk. Neuron, 2013. 77(3): p. 425-39. 270. Zavodszky, E., et al., Mutation in VPS35 associated with Parkinson's disease impairs WASH complex association and inhibits autophagy. Nat Commun, 2014. 5(3828). 271. Tsika, E., et al., Parkinson's disease-linked mutations in VPS35 induce dopaminergic neurodegeneration. Hum Mol Genet, 2014. 23(17): p. 4621-38. 272. Shi, H., et al., The retromer subunit Vps26 has an arrestin fold and binds Vps35 through its C-terminal domain. Nature structural & molecular biology, 2006. 13(6): p. 540-8. 273. Collins, B.M., et al., Vps29 has a phosphoesterase fold that acts as a protein interaction scaffold for retromer assembly. Nature structural & molecular biology, 2005. 12(7): p. 594- 602. 274. Swarbrick, J.D., et al., VPS29 is not an active metallo-phosphatase but is a rigid scaffold required for retromer interaction with accessory proteins. PLoS One, 2011. 6(5): p. e20420. 275. Wang, D., et al., Crystal structure of human vacuolar protein sorting protein 29 reveals a phosphodiesterase/nuclease-like fold and two protein-protein interaction sites. J Biol Chem, 2005. 280(24): p. 22962-7. 276. Hierro, A., et al., Functional architecture of the retromer cargo-recognition complex. Nature, 2007. 449(7165): p. 1063-7. 277. Norwood, S.J., et al., Assembly and Solution Structure of the Core Retromer Protein Complex. Traffic, 2011. 12(1): p. 56-71. 278. Chevallier, N., et al., Cathepsin D displays in vitro beta-secretase-like specificity. Brain Res, 1997. 750(1-2): p. 11-9. 130

279. Thompson, A., et al., Expression and characterization of human beta-secretase candidates metalloendopeptidase MP78 and cathepsin D in beta APP-overexpressing cells. Brain Res Mol Brain Res, 1997. 48(2): p. 206-14. 280. Sevlever, D., P. Jiang, and S.H. Yen, Cathepsin D is the main lysosomal enzyme involved in the degradation of alpha-synuclein and generation of its carboxy-terminally truncated species. Biochemistry, 2008. 47(36): p. 9678-9687. 281. Cullen, V., et al., Cathepsin D expression level affects alpha-synuclein processing, aggregation, and toxicity in vivo. Mol Brain, 2009. 2: p. 5. 282. Qiao, L., et al., Lysosomal enzyme cathepsin D protects against alpha-synuclein aggregation and toxicity. Mol Brain, 2008. 1(17): p. 1756-6606. 283. Kubala, M.H., et al., Structural and thermodynamic analysis of the GFP:GFP-nanobody complex. Protein Science, 2010. 19(12): p. 2389-2401. 284. Gibb, W.R. and A.J. Lees, The relevance of the Lewy body to the pathogenesis of idiopathic Parkinson's disease. Journal of Neurology, Neurosurgery & Psychiatry, 1988. 51(6): p. 745- 752. 285. Kerr, M.C., et al., Inhibition of the PtdIns(5) kinase PIKfyve disrupts intracellular replication of Salmonella. EMBO J, 2010. 29(8): p. 1331-1347. 286. Schneider, C.A., W.S. Rasband, and K.W. Eliceiri, NIH Image to ImageJ: 25 years of image analysis. Nat Meth, 2012. 9(7): p. 671-675. 287. Auluck, P.K., et al., Chaperone Suppression of α-Synuclein Toxicity in a Drosophila Model for Parkinson's Disease. Science, 2002. 295(5556): p. 865-868. 288. Wan, O.W. and K.K. Chung, The role of alpha-synuclein oligomerization and aggregation in cellular and animal models of Parkinson's disease. PLoS One, 2012. 7(6): p. e38545. 289. Yoshimori, T., et al., Bafilomycin A1, a specific inhibitor of vacuolar-type H(+)-ATPase, inhibits acidification and protein degradation in lysosomes of cultured cells. J Biol Chem, 1991. 266(26): p. 17707-12. 290. Rcom-H’cheo-Gauthier, A., J. Goodwin, and D.L. Pountney, Interactions between Calcium and Alpha-Synuclein in Neurodegeneration. Biomolecules, 2014. 4(3): p. 795-811. 291. Goodwin, J., et al., Raised calcium and oxidative stress cooperatively promote alpha- synuclein aggregate formation. Neurochem Int, 2013. 62(5): p. 703-11. 292. Sugeno, N., et al., Lys-63-linked ubiquitination by E3 ubiquitin ligase Nedd4-1 facilitates endosomal sequestration of internalized alpha-synuclein. J Biol Chem, 2014. 289(26): p. 18137-51. 293. Popovic, D. and I. Dikic, TBC1D5 and the AP2 complex regulate ATG9 trafficking and initiation of autophagy. EMBO Rep, 2014. 15(4): p. 392-401. 294. Xia, Q., et al., Proteomic identification of novel proteins associated with Lewy bodies. Front Biosci, 2008. 13: p. 3850-6. 295. Hasegawa, T., et al., The AAA-ATPase VPS4 regulates extracellular secretion and lysosomal targeting of alpha-synuclein. PloS one, 2011. 6(12): p. 22. 296. Hasegawa, T., et al., The AAA-ATPase VPS4 Regulates Extracellular Secretion and Lysosomal Targeting of α-Synuclein. PLoS ONE, 2011. 6(12): p. e29460. 297. Popovic, D., et al., Rab GTPase-activating proteins in autophagy: regulation of endocytic and autophagy pathways by direct binding to human ATG8 modifiers. Mol Cell Biol, 2012. 32(9): p. 1733-44. 298. Tanik, S.A., et al., Lewy body-like alpha-synuclein aggregates resist degradation and impair macroautophagy. J Biol Chem, 2013. 288(21): p. 15194-210. 299. Decressac, M., et al., TFEB-mediated autophagy rescues midbrain dopamine neurons from alpha-synuclein toxicity. Proc Natl Acad Sci U S A, 2013. 110(19): p. E1817-26. 300. Orrenius, S., B. Zhivotovsky, and P. Nicotera, Regulation of cell death: the calcium-apoptosis link. Nat Rev Mol Cell Biol, 2003. 4(7): p. 552-565. 131

301. Miyamoto, A., et al., Apoptosis induction-related cytosolic calcium responses revealed by the dual FRET imaging of calcium signals and caspase-3 activation in a single cell. Biochemical and Biophysical Research Communications, 2015. 460(1): p. 82-87. 302. Zhang, Y., et al., Vacuolar protein sorting 35 Asp620Asn mutation is rare in the ethnic Chinese population with Parkinson's disease. Parkinsonism & Related Disorders, 2012. 18(5): p. 638-640. 303. Sharma, M., et al., A multi-centre clinico-genetic analysis of the VPS35 gene in Parkinson disease indicates reduced penetrance for disease-associated variants. Journal of Medical Genetics, 2012. 49(11): p. 721-726. 304. Roberts, R.L., et al., Endosome fusion in living cells overexpressing GFP-rab5. J Cell Sci, 1999. 112(Pt 21): p. 3667-75. 305. Gieselmann, V., A. Hasilik, and K. von Figura, Processing of human cathepsin D in lysosomes in vitro. J Biol Chem, 1985. 260(5): p. 3215-20. 306. Cataldo, A.M., et al., Endocytic pathway abnormalities precede amyloid beta deposition in sporadic Alzheimer's disease and Down syndrome: differential effects of APOE genotype and presenilin mutations. The American journal of pathology, 2000. 157(1): p. 277-86. 307. Zech, T., et al., The Arp2/3 activator WASH regulates alpha5beta1-integrin-mediated invasive migration. J Cell Sci, 2011. 124(Pt 22): p. 3753-9. 308. Siintola, E., et al., Cathepsin D deficiency underlies congenital human neuronal ceroid- lipofuscinosis. Brain, 2006. 129(Pt 6): p. 1438-45. 309. Tyynela, J., et al., A mutation in the ovine cathepsin D gene causes a congenital lysosomal storage disease with profound neurodegeneration. EMBO J, 2000. 19(12): p. 2786-92. 310. Koike, M., et al., Cathepsin D deficiency induces lysosomal storage with ceroid lipofuscin in mouse CNS neurons. J Neurosci, 2000. 20(18): p. 6898-906. 311. Klein, C. and A. Westenberger, Genetics of Parkinson’s Disease. Cold Spring Harbor Perspectives in Medicine, 2012. 2(1). 312. Miura, E., et al., VPS35 dysfunction impairs lysosomal degradation of alpha-synuclein and exacerbates neurotoxicity in a Drosophila model of Parkinson's disease. Neurobiology of disease, 2014. 71: p. 1-13. 313. Bhalla, A., et al., Characterizing the location and trafficking routes of the neuronal retromer and its role in amyloid precursor protein transport. Neurobiol Dis, 2012. 47(1): p. 126-34. 314. Munsie, L.N., et al., Retromer-dependent neurotransmitter receptor trafficking to synapses is altered by the Parkinson's disease VPS35 mutation p.D620N. Human molecular genetics, 2015. 24(6): p. 1691-703. 315. Liscum, L., Niemann–Pick Type C Mutations Cause Lipid Traffic Jam. Traffic, 2000. 1(3): p. 218-225. 316. Ory, D.S., Niemann-Pick type C: a disorder of cellular cholesterol trafficking. Biochim Biophys Acta, 2000. 1529(1-3): p. 331-9. 317. Cruz, J.C., et al., Role of Niemann-Pick Type C1 Protein in Intracellular Trafficking of Low Density Lipoprotein-derived Cholesterol. Journal of Biological Chemistry, 2000. 275(6): p. 4013-4021. 318. Underwood, K.W., et al., Evidence for a cholesterol transport pathway from lysosomes to endoplasmic reticulum that is independent of the plasma membrane. J Biol Chem, 1998. 273(7): p. 4266-74. 319. Wojtanik, K.M. and L. Liscum, The transport of low density lipoprotein-derived cholesterol to the plasma membrane is defective in NPC1 cells. J Biol Chem, 2003. 278(17): p. 14850-6. 320. Sugii, S., et al., Distinct endosomal compartments in early trafficking of low density lipoprotein-derived cholesterol. J Biol Chem, 2003. 278(29): p. 27180-9. 321. Lebrand, C., et al., Late endosome motility depends on lipids via the small GTPase Rab7. The EMBO Journal, 2002. 21(6): p. 1289-1300. 132

322. Ko, D.C., et al., Dynamic movements of organelles containing Niemann-Pick C1 protein: NPC1 involvement in late endocytic events. Mol Biol Cell, 2001. 12(3): p. 601-14. 323. Kobayashi, T., et al., Late endosomal membranes rich in lysobisphosphatidic acid regulate cholesterol transport. Nat Cell Biol, 1999. 1(2): p. 113-8. 324. Ganley, I.G. and S.R. Pfeffer, Cholesterol accumulation sequesters Rab9 and disrupts late endosome function in NPC1-deficient cells. J Biol Chem, 2006. 281(26): p. 17890-9. 325. Vitry, S., et al., Storage Vesicles in Neurons Are Related to Golgi Complex Alterations in Mucopolysaccharidosis IIIB. The American journal of pathology, 2010. 177(6): p. 2984- 2999. 326. Nixon, R.A. and D.S. Yang, Autophagy and neuronal cell death in neurological disorders. Cold Spring Harbor perspectives in biology, 2012. 4(10). 327. Geuze, H.J., et al., Sorting of mannose 6-phosphate receptors and lysosomal membrane proteins in endocytic vesicles. J Cell Biol, 1988. 107(6 Pt 2): p. 2491-501. 328. Hockey, L.N., et al., Dysregulation of lysosomal morphology by pathogenic LRRK2 is corrected by TPC2 inhibition. J Cell Sci, 2015. 128(2): p. 232-8. 329. Spranger, J.W. and H.R. Wiedemann, The genetic mucolipidoses. Diagnosis and differential diagnosis. Humangenetik, 1970. 9(2): p. 113-39. 330. De Pace, R., et al., Mucolipidosis II-related mutations inhibit the exit from the endoplasmic reticulum and proteolytic cleavage of GlcNAc-1-phosphotransferase precursor protein (GNPTAB). Human mutation, 2014. 35(3): p. 368-76. 331. Tiede, S., et al., Mucolipidosis II is caused by mutations in GNPTA encoding the alpha/beta GlcNAc-1-phosphotransferase. Nature medicine, 2005. 11(10): p. 1109-12. 332. Leroy, J.G. and R.I. Demars, Mutant enzymatic and cytological phenotypes in cultured human fibroblasts. Science, 1967. 157(3790): p. 804-6. 333. Wiesmann, U.N., et al., Multiple lysosomal deficiency due to enzyme leakage? The New England journal of medicine, 1971. 284(2): p. 109-10. 334. Kett, L.R. and W.T. Dauer, Leucine-rich repeat kinase 2 for beginners: six key questions. Cold Spring Harb Perspect Med, 2012. 2(3): p. a009407. 335. Linhart, R., et al., Vacuolar protein sorting 35 (Vps35) rescues locomotor deficits and shortened lifespan in Drosophila expressing a Parkinson's disease mutant of Leucine-Rich Repeat Kinase 2 (LRRK2). Mol Neurodegener, 2014. 9: p. 23. 336. Bi, F., et al., Pathogenic Mutation in VPS35 Impairs Its Protection against MPP(+) Cytotoxicity. International Journal of Biological Sciences, 2013. 9(2): p. 149-55. 337. Braschi, E., et al., Vps35 mediates vesicle transport between the mitochondria and peroxisomes. Curr Biol, 2010. 20(14): p. 1310-5. 338. Bonekamp, N.A., et al., Reactive oxygen species and peroxisomes: struggling for balance. BioFactors (Oxford, England), 2009. 35(4): p. 346-55. 339. Mullan, M., et al., A pathogenic mutation for probable Alzheimer's disease in the APP gene at the N-terminus of beta-amyloid. Nat Genet, 1992. 1(5): p. 345-7.

133

Appendix:

Supplementary Movie 1: Vps35 WT-GFP positive endosomes are mobile and found throughout the cytoplasm. Details in Figure legend 4A.

Supplementary Movie 2: Vps35 D620N-GFP positive endosomes are enlarged and redistributed to a perinuclear localization but retain ability to receive endocytosed material. Details in Figure legend 4A.

// peri-nuclear measurer // (c) Dr Nick 2013 [email protected] radius = 40; // only measure area/intensity run("Set Measurements...", "area mean redirect=None decimal=5"); path = getDirectory("Choose Source Directory "); list = getFileList(path); summaryFile = File.open(path+"Results_Summary.xls"); print(summaryFile,"Image\tAvg Peri Intensity\tPeri Area\tAvg non-Peri Intensity\tnon-Peri Area\tIntensity Ratio\n"); for (i=0; i

Supplementary Figure 1: Quantification of peri-nuclear and cytoplasmic protein. To quantify perinuclear/cytoplasmic localization, a macro was created in ImageJ* (version 1.47i). The macro is fully automated with the only user interaction required being the selection of the directory containing the images to be quantified. Two channels are utilized by the script from each image: a DAPI channel to define the nuclear regions (blue channel); and the protein of interest (POI) channel (green channel). For each image, the ratio of total protein found in the perinuclear to cytoplasmic regions (as defined by distance from the nuclei in the images) is then calculated and recorded in an Excel file. See Materials and Methods for details of the quantification process. 135

Supplementary figure 2: PD-linked Vps35 mutants P316S and R524W do not impact Golgi morphology (A) HeLa cells transiently expressing Vps35 WT-GFP, Vps35 P316S-GFP or Vps35 R524W-GFP were fixed and subjected indirect immunofluorescence using antibodies against human p230, as described. Coverslips were counterstained with DAPI nuclear stain. scale bar: 5µM.

136

Supplementary figure 3: PD-linked Vps35 mutants P316S and R524W do not disrupt the localization of SNX27 (A) HeLa cells transiently expressing GFP fusion constructs were collected and fractioned as per described in the methods. Fractions were resolved by SDS-PAGE transferred onto PVDF membrane (Immobilon-P and Immobilon-FL; Millipore) and probed with antibodies raised against SNX27, LAMP1 and ß-tubulin. Samples displayed here match those in Figure 5.2 B. The Vps35 D620N mutation linked to Parkinson’s disease disrupts the cargo sorting function of retromer.

Jordan Follett1, Suzanne J. Norwood1, Nicholas A Hamilton1, Megha Mohan2, Oleksiy Kovtun1, Stephanie Tay1, Stephen A. Wood2, George D. Mellick2, Peter A. Silburn2,3, Brett M. Collins1, Andrea Bugarcic1,*, and Rohan D. Teasdale1

1Institute for Molecular Bioscience, The University of Queensland, St Lucia, Queensland, Australia, 2Eskitis Institute for Drug Discovery, Griffith University, Nathan, Queensland, Australia, 3The University of Queensland Centre for Clinical Research, Herston, Queensland, Australia

*Corresponding Author: Andrea Bugarcic, Institute for Molecular Bioscience, The University of Queensland, St. Lucia, QLD 4072, Australia; e-mail: [email protected]; phone: +61-7-3346-2031

RUNNING TITLE: Parkinson’s disease causing mutation alters retromer’s function

Keywords: Retromer, Vps35 D620N, Parkinson’s disease, endosome, cathepsin D

Abbreviations: PD – Parkinson’s disease, AD – Alzheimer’s disease, Vps – vacuolar protein sorting, SNX – sorting nexin, TGN – Trans Golgi network

1

2

Abstract

The retromer is a trimeric cargo-recognition protein complex composed of Vps26, Vps29 and Vps35 associated with protein trafficking within endosomes. Recently, a pathogenic point mutation within the Vps35 subunit (D620N) was linked to the manifestation of Parkinson’s disease (PD). Here, we investigated details underlying the molecular mechanism by which the D620N mutation in Vps35 modulates retromer function, including examination of retromer’s subcellular localization and its capacity to sort cargo. We show that expression of the PD-linked Vps35 D620N mutant redistributes retromer-positive endosomes to a perinuclear subcellular localization and that these endosomes are enlarged in both model cell lines and fibroblasts isolated from a PD patient. Vps35 D620N is correctly folded and binds Vps29 and Vps26A with the same affinity as wild-type Vps35. While PD-linked point mutant Vps35 D620N interacts with the cation-independent mannose-6-phosphate receptor (CI-M6PR), a known retromer cargo, we find that its expression disrupts the trafficking of cathepsin D, a CI-M6PR ligand and protease responsible for degradation of α-synuclein, a causative agent of Parkinson’s disease. In summary, we find that the expression of Vps35 D620N leads to endosomal alterations and trafficking defects that may partly explain its action in PD.

3

Introduction

Parkinson’s disease (PD) is a neurodegenerative disorder characterized by the formation of large, insoluble, intracellular protein aggregates termed Lewy bodies. Anatomically, the most critical abnormality in PD is the loss of neurons in the substantia nigra and the depletion of dopamine from basal ganglia circuits, resulting in disruption in the brain communication. Even though the mechanism behind Lewy body formation is controversial, immuno-reactivity has identified proteins such as ubiquitin, molecular chaperone proteins αB-crystallin, Hsp70, Hsp40 and pre-synpatic protein α-synuclein as key components of Lewy body structures (1, 2). α-synuclein is a 14 kDa natively unfolded protein with no known function, but the central region of this protein (amino acids 61-95) contains residues prone to self-aggregation (3, 4). This self-aggregation tendency may be a primary driving factor in the formation of higher molecular weight species observed in Lewy bodies, commonly witnessed in the brains of PD patients. In sporadic Parkinson’s disease cases induction of changes in α-synuclein protein structure and subsequent generation of aggregates has been hypothesized to occur via several different pathways. These include the disregulation of mitochondrial function, specifically leading to increased free radical production (5) and free cytosolic calcium levels, as well as the inhibition of protein-degradation pathways (reviewed in (6, 7). Interestingly, recent studies implicated lysosomes and lysosomal enzymes as fundamental regulators of α-synuclein turn over – (i) Cuervo and colleagues reported binding of α-synuclein to lysosomal membranes leading to inhibition of lysosome function (8), (ii) cathepsin D, a lysosomal protease and cation-independent mannose-6-phosphate receptor (CI-M6PR) ligand, is able to degrade α-synuclein in vivo and in vitro (9, 10) , and (iii) the cathepsin D knockout mouse model shows marked increase in aggregated but unchanged monomeric levels of α-synuclein (11).

Mammalian retromer is a protein complex composed of three proteins Vps26, Vps29 and Vps35, with an important role in the sorting and trafficking of transmembrane receptors within the endosome. The retromer forms a stable trimer, which associates with a range of other proteins, including sorting nexins (SNXs), which modulate its function (reviewed in (12). Sorting of cargo from endosomes to the TGN is a tightly organized process, and retromer is a key player in mediating trafficking of several receptors on this route, including CI-M6PR (13). Interaction between retromer and CI-M6PR is mediated through a tripeptide sequence within the C-terminal cytoplasmic domain of CI-M6PR (13) and amino acids 500-

4

693 of the Vps35 subunit (14). Disruption of the retromer-cargo interaction, either by downregulation of retromer levels (14) or modification of cargo cytoplasmic domains (13) leads to accumulation of receptors in early endosomes and ultimately to their degradation (15, 16). Within the TGN CI-M6PR binds enzymes, such as newly synthesized cathepsin D, and delivers them to the late endosome network where they undergo processing to yield enzymatically active forms responsible for protein degradation within the lysosome (17). Upon dissociation of CI-M6PR from its ligand, CI-M6PR is trafficked back to the TGN by interacting with the retromer complex within the endosomal membrane (14, 18). Any failure within this CI-M6PR cycling has been shown to result in increased turnover of the receptor as well as improper intracellular processing and secretion of the cathepsin D into the surrounding medium (14, 16, 19). Successful delivery of cathepsin D to the endosomal network results in processing of the “pro” form into a mature, active form of this enzyme (20). As such, the availability of CI-M6PR at the TGN, via its association with retromer, is crucial for sorting of cathepsin D from TGN to endosomes.

A number of recent reports determined a link between several point mutations of Vps35, a retromer subunit, and manifestation of late-onset Parkinson’s disease (21-23). Reports implicating Vps35 in PD demonstrate that a single point mutation of highly conserved amino acid 620 in Swiss, Austrian and German families leads to an autosomal dominant, high penetrance mode of PD inheritance (24). Using direct sequencing of samples from PD patients, two individual groups further confirmed the presence and same inheritance mode of the D620N variant in French and Japanese ethnic groups (21, 22). A recent large multi- center study used more than 15,000 subjects worldwide to screen for all known PD variants and confirmed the presence of D620N in 5 familial cases and two seemingly sporadic cases, emphasizing the importance of this particular point mutation in PD patient cohorts worldwide (25). Collectively, genetic evidence suggests that the pathogenic D620N Vps35 variant is a rare cause of familial as well as idiopathic forms of PD, and points to endosomal trafficking as a critical process in the disease.

To understand the underlying molecular mechanism of the Parkinson’s disease causing Vps35 D620N mutation we examined if an established retromer function, namely regulation of CI-M6PR trafficking, was disrupted by the expression of this variant. Any disruption of CI- M6PR trafficking, and therefore cathepsin D processing, has the potential to directly lead to downstream modulation of the levels of α-synuclein monomeric and/or aggregated forms. While Vps35 D620N is correctly folded and assembled into trimeric retromer complexes, its 5 overexpression leads to formation of dilated and misplaced endosomes in model cell lines with same phenotype also observed in skin fibroblasts isolated from a Parkinson’s disease patient with Vps35 D620N mutation. Retromer containing the mutant Vps35 D620N incorrectly traffics its cargo, CI-M6PR, resulting in improper processing of cathepsin D, a CI- M6PR ligand, and leads to increased secretion of pro-cathepsin D. Our results therefore show that the dominant expression of Vps35 D620N mutation results in endosomal trafficking alterations that may underpin its role in PD.

Results

Vps35 D620N binds Vps29 and Vps26A with same affinity as Vps35 WT To assess the stability and functionality of the Vps35 D620N mutant in vitro, recombinant proteins were expressed and purified by affinity chromatography and gel filtration. Vps35 mutant D620N displayed similar elution profiles by gel filtration to the wild-type (WT) protein, and additionally CD spectra exhibited similar -helical contents consistent with the Vps35 -solenoid structure (26) (Figure 1B). We conclude therefore that the Vps35 D620N mutation does not lead to major misfolding of Vps35. We next tested the impact of the D620N mutation on the ability of Vps35 to form high affinity complexes with Vps26A and Vps29 in vitro (Figure 1C) (26) (16). Structural data for Vps35 only exists for the C-terminal segment from residues 476-780, bound to the small retromer subunit Vps29 (26). The Asn 620 residue is present in an exposed loop between -helices of the repeating pairs of HEAT- like -helical repeats, and does not contribute to the Vps35 interface with Vps29 (Figure 1A). As shown by ITC, we find that the D620N mutant binds to Vps26A and Vps29 with affinities and thermodynamic parameters indistinguishable from the wild type Vps35 molecule (Figure 1C). Vps26A binds Vps35 WT and D620N with affinities (Kds) of 1.1 and 0.5 nM respectively, while Vps29 binds WT and D620N with affinities of 170 nM and 180 nM respectively. Furthermore, in vivo co-immunoprecipitation experiments detected no differences in levels of endogenous Vps26A and Vps29 subunits associated with Vps35 WT-GFP or the Vps35 D620N-GFP mutant (Figure 1D). Overall we conclude that the presence of the D620N mutation in Vps35 does not directly affect global VPS35 folding and does not prevent the formation of a stable trimeric retromer protein complex.

6

Expression of Vps35 D620N mutant causes redistribution of endosomes To determine if the presence of the D620N mutation in Vps35 altered the subcellular localization of retromer we first performed live cell imaging of A431 cells transiently transfected with either Vps35 WT-GFP or Vps35 D620N-GFP pulsed with a fluid phase marker, Dextran-647. The Vps35 WT-GFP dextran positive endosomes were found throughout the cell cytoplasm and displayed mobility typical of early endosomes including fusion between Vps35-GFP positive endosomes (Supplemental Movie 1). In contrast, the Vps35 D620N-GFP positive endosomes were enlarged and redistributed to a tight, perinuclear localization (Figure 2A). Time-lapse microscopy demonstrated these enlarged endosomes contained dextran within their lumens supporting their capacity to receive endocytosed materials. Enlarged Vps35 D620N-GFP positive endosomes were observed to arise from smaller, dispersed endosomes, likely to be early endosome structures undergoing fusion during endosome maturation (27) (Supplemental Movie 2).

To determine if the retromer complex was also present on these enlarged redistributed endosomes induced in the presence of the Vps35 D620N mutant, we used indirect immunofluorescence on A431 cells transiently transfected with Vps35 WT-GFP or Vps35 D620N-GFP and immuno-labeled against endogenous Vps26A. Imaging of fixed cells overexpressing Vps35 WT-GFP show Vps35 WT-GFP/Vps26A positive endosomes with morphologies indistinguishable from Vps26A positive endosomes in neighboring untransfected cells. However, in cells overexpressing Vps35 D620N-GFP we observed endosomes positive for endogenous Vps26A redistributed to a perinuclear localization, dramatically different to surrounding untransfected cells showing Vps26A positive endosomes (Figure 2B). This analysis allowed us to confirm the presence of the retromer subunit, Vps26A, on the redistributed Vps35 D620N positive endosomes.

Quantification of the endosome redistribution in cells overexpressing Vps35 WT-GFP or Vps35 D620N-GFP was performed by measuring the distance of each Vps35-positive endosome from the middle of the nucleus. By designing a distance-based algorithm which measured fluorescent units by distinguishing perinuclear from non-perinuclear structures, we show that perinuclear fluorescence, representing endosomes, is increased in cells over- expressing Vps35 D620N-GFP when compared to Vps35 WT-GFP positive cells (Figure 2C).

7

To determine if the localization of the modified endosomes induced by Vps35 D620N expression was microtubule-dependent and associated with the microtubule organization center we treated cells with a microtubule depolymerization agent, nocadazole. As shown in Figure 2D, cells overexpressing Vps35 D620N-GFP treated with nocadazole displayed a dissociation of the tightly localized endosomes to a dispersed cytoplasmic localization. Therefore the localization of the induced endosomes is not a product of aggregation.

Identification of the redistributed endosome population To further investigate the properties of the enlarged redistributed endosomes induced in the presence of the Vps35 D620N mutant we performed a series of colocalisation experiments with endosome and TGN marker proteins. A431 cells were transiently transfected with Vps35 WT-GFP or Vps35 D620N-GFP for 24 h and indirect immunofluorescence was performed to compare the subcellular localization of endogenous EEA1, LAMP1 and p230 (Figure 3A). In cells overexpressing Vps35 WT-GFP the distribution of EEA1 positive endosomes was indistinguishable from that observed in neighboring untransfected cells. In contrast, Vps35 D620N expression caused a shift in EEA1-positive endosomes to the perinuclear region that was obvious when compared to neighboring untransfected cells (Figure 3A and Figure 3B). Confocal analysis revealed a low, but consistent colocalization between Vps35 WT-GFP and EEA1 (colocalization coefficient (R) = 0.137) while Vps35 D620N-GFP showed an increase in colocalization with EEA1 (R = 0.207) (Figure 3B).

While Vps35 WT-GFP showed low colocalisation with LAMP1 (R = 0.101) a late endosome marker, and identical subcellular localization of LAMP1-positive compartments compared to neighbouring untransfected cells, Vps35 D620N-positive endosomes showed an increase in colocalisation with this marker (R = 0.158) on the large, redistributed endosomes (Figure 3A and Figure 3B). We next investigated the possibility of an association between the Vps35 D620N-GFP and the TGN using colocalisation with the p230 marker. However, we observed no evidence of association between Vps35 D620N-GFP and p230 positive membranes (Figure 3A and Figure 3B) and no distinct morphological differences in p230-positive Golgi structures between Vps35 WT-GFP and Vps35 D620N expressing cells was observed (Figure 3A). These data suggest increased association of Vps35 D620N-positive endosomes with both early and late endosome markers in an induced tight perinuclear subcellular location.

8

Vps35 D620N mutant causes a defect in cathepsin D trafficking Retromer directly binds the CI-M6PR cytoplasmic domain via a region of Vps35 that overlaps with the position of the D620N mutation (14). To examine the CI-M6PR interaction with Vps35 we co-transfected the CD8-CI-M6PR fusion construct and either Vps35 WT-GFP or Vps35 D620N-GFP and performed GFP-NanoTrap co-immunoprecipitations (16). As seen in Figure 4A, Vps35 D620N-GFP can co-precipitate CD8-CI-M6PR at levels comparable to Vps35 WT-GFP. We conclude that the D620N mutation in Vps35 does not disrupt retromer’s capacity to interact with this cargo.

Next we examined if the Vps35 D620N mutant is able to alter the subcellular localization of CI-M6PR by disrupting its normal trafficking itinerary. We transiently transfected A431 cells with Vps35 WT-GFP or Vps35 D620N-GFP and determined the subcellular distribution of endogenous CI-M6PR. Consistent with previous reports (14), a proportion of endogenous CI-M6PR was localized to Vps35 WT-GFP-positive endosomes. Interestingly, the proportion of CI-M6PR/retromer-positive endosomal structures appeared to increase in cells overexpressing Vps35 D620N-GFP mutant (Figure 4B), with endogenous CI-M6PR also showing redistribution into tight, punctate perinuclear localization.

To determine if the redistribution of the receptor impacted on delivery of CI-M6PR cargo, we investigated processing of cathepsin D in the presence of Vps35 D620N. Firstly we characterized the processing of cathepsin D from pro- into a mature form within the cells and secondly we assessed the levels of pro-cathepsin D secreted into the media of the transfected cells. HEK293 cells were transiently transfected with either Vps35 WT-GFP or Vps35 D620N-GFP for 24 h, incubated with cyclohexamide for up to 7 h and cell lysate and medium samples collected at 0, 3 and 7 h post chase (16). We observed that cells expressing the Vps35 D620N-GFP protein secreted the 50 kDa pro-cathepsin D into the media at 7 h post chase (Figure 4C). This result is consistent with impaired trafficking of CI- M6PR resulting in a deficiency of receptors at the TGN to interact with the pro-cathepsin D (14). Additionally, cells overexpressing the Vps35 D620N-GFP mutant, when compared to Vps35 WT-GFP protein, showed decreased levels of mature 20 kDa cathepsin D in cell lysates, further demonstrating impaired delivery of pro-cathepsin D to the late- endosome/lysosome for processing into the mature 20 kDa form. Taken together, these results show that even though the Vps35 D620N mutant retains its ability to bind CI-M6PR, cathepsin D, the soluble ligand of this receptor, is not processed efficiently resulting in secretion of the immature pro-cathepsin D from the cells. 9

Vps35 D620N is associated with redistributed endosomes in PD patient fibroblasts Human dermal fibroblasts from a PD patient genotyped for the Vps35 D620N heterozygote point mutation were isolated and the subcellular localization of Vps35-positive endosomes and morphology of endocytic compartments was examined. By using indirect immunofluorescence to detect endogenous retromer subunits we detected a strong colocalisation of Vps35 with Vps26A in both control and patient fibroblasts (Figure 5B). In PD patient fibroblasts we also observed a shift of Vps35-positive endosomes to a perinuclear localization, in contrast to control fibroblasts where Vps35-positive endosomes showed a broader distribution throughout the cytoplasm (Figure 5A). Quantification of this phenotype using the distance-based methodology described above again showed a significant increase in the perinuclear intensity ratio in patient fibroblasts compared to control fibroblasts (Figure 5C).

Subcellular localisation of the Vps35-positive endosomes in fibroblasts was performed using endogenous markers as described above. Colocalisation studies with EEA1, an early endosome marker, show an increased proportion of Vps35-positive punctate structures positive for the EEA1 in patient fibroblasts (R = 0.304) compared to the control (R = 0.209) (Figure 5B). Further colocalisation of Vps35 with LAMP1, a late endosome marker, showed a higher colocalisation coefficient between Vps35 and LAMP1 in patient fibroblasts (R =0.274), compared to control (R = 0.197) (Figure 5B). Interestingly, using both EEA1 and LAMP1 markers we also observed consistent evidence of enlarged redistributed endosomes, with LAMP1-positive endosomes showing consistent enlargement of the endosomal lumen (Figure 5A, LAMP1 panels). Investigation of p230, a TGN marker, in control and patient fibroblasts showed no change in colocalization between Vps35 and p230 positive structures (Figure 5B) and no morphological difference in p230 positive structures (Figure 5A, p230 panels). Taken together, these results confirm increased localization of Vps35 D620N mutant to both early and late endosomes and corroborate the ectopic studies above.

To determine if the presence of the Vps35 D620N mutation in patient fibroblasts impacted on the retromer mediated trafficking of CI-M6PR we initially performed colocalisation studies between endogenous Vps35 and CI-M6PR in control and patient fibroblasts. Within PD patient fibroblasts expressing Vps35 D620N we observed an increase in the colocalization of CI-M6PR with Vps35 (R = 0.257) compared to that found in control cells (R = 0.203) 10

(Figure 6B), as well as a clear shift in CI-M6PR-positive endosomes into a more perinuclear subcellular localization (Figure 6A). We further examined the processing of CI-M6PR ligand, cathepsin D, using the cyclohexamide assay described previously. The fibroblasts from the patient displayed higher levels of mature cathepsin D at steady state when compared to the normal patient fibroblasts. After the inhibition of protein synthesis for 7 h we observed that the level of mature 20 kDa cathepsin D in patient fibroblasts decreased when compared to the fibroblasts from a control sample (Figure 6C). This is consistent with a reduction in the efficiency to transport newly synthesized pro-cathepsin D to the lysosome. However, we were unable to observe secretion of the pro-cathepsin D into the media presumably due to the lower level of cathepsin D expressed by these cells (data not shown). These results demonstrate a defect in trafficking of CI-M6PR in patient cells, resulting in processing disruption of cathepsin D into a mature form of this enzyme. These observations are consistent with those observed in the ectopic expression experiments performed above.

Discussion Here we identified that the Parkinson’s disease-linked Vps35 D620N mutation does not interfere with Vps35’s capacity to form high affinity interactions with Vps26A and Vps29, as evidenced by in vitro and cell-based assays. Ectopic expression of the Vps35 D620N protein in mammalian cells, as well as examination of fibroblasts isolated from a Parkinson’s disease patient containing the D620N mutation in Vps35, demonstrated retromer association with enlarged endosomes that were concentrated to a perinuclear subcellular localization that also displayed evidence of retention of early and late endosome markers. Additionally, we demonstrate that even though Vps35 D620N-retromer interacts with its receptor cargo, CI-M6PR, its expression caused the receptors subcellular distribution and function to be modified. In both the overexpression model and Parkinson’s disease patient fibroblasts, the processing of cathepsin D, a CI-M6PR ligand, into the mature 20 kDa active form was markedly decreased in the presence of Vps35 D620N protein with resulting pro- cathepsin D secreted from the cells. Therefore the expression of retromer incorporating the mutant Vps35 D620N protein induces a disruption of the normal trafficking itinerary of CI- M6PR.

The redistributed, altered endosome structures positive for Vps35 D620N-GFP, represent a common phenotype observed with endosome dysfunction and neurodegeneration (28). Live cell imaging of these enlarged and redistributed endosomes demonstrated they are motile 11 entities rather than static aggregates, and the delivery of dextran via endocytosis supports their capacity to still interact with other endosomal compartments. The redistributed Vps35 D620N positive endosomes in both overexpression and fibroblast models had increased colocalisation with EEA1 and LAMP1 relative to wild-type Vps35. The retention of these markers is consistent with a defect in the endosomal maturation process (29), where the conversion between early and late endosome compartments may be underpinned by a failure of Vps35 D620N endosomes to mature at a constant rate, leading to the dilated endosome phenotype. The dysregulation of retromer itself and also a number of proteins that associate with retromer are known to cause a similar change in endosomes. For example, recruitment of retromer to the endosomal membranes has been reported to require the two RabGTPases, Rab5 and Rab7, acting in concert (19, 30). The failure of Rab5 to dissociate from the membrane (29), overexpression of dominant-negative Rab5 and Rab7 (31) or depletion of the retromer cargo recognition complex have all been reported to result in increased endosome size (14), like the phenotype we observed when Vps35 D620N protein is expressed. In addition, recent studies demonstrate the importance of the WASH complex in maintaining endosome integrity and lumen size while linking vesicles and F-actin networks (32, 33). The role of the WASH complex is largely focused on early or pre- endocytic compartments, however it has also been described to associate with late endocytic structures (34, 35). As retromer has been shown to interact directly with the WASH complex via the Fam21 subunit (33, 36-38), it is plausible that the dilated endosomes are a result of an inability of the Vps35 D620N to activate the WASH complex with high affinity resulting in a disruption of F-actin dependent cellular mechanisms including the formation of endosome derived tubular-vesicular transport vesicles. However, we find that Vps35 D620N retains a high level of colocalization (>85%) with endogenous FAM21, a WASH complex subunit, and does not perturb the described interaction between retromer and the WASH complex using immunoprecipitation and GST pull down assays (data not shown).

The D620N mutation is within the region of Vps35 identified to bind CI-M6PR (14, 26); however our study shows that Vps35 D620N-containing retromer still has the capacity to directly interact with CI-M6PR. Therefore the detection of the immature form of pro- cathepsin D in the medium of cells overexpressing Vps35 D620N mutant and decreased amount of mature cathepsin D in the over-expression and human fibroblast models is not due to an inability of the Vps35 D620N protein to bind CI-M6PR. CI-M6PR was present in the altered retromer positive endosomes of cells expressing the Vps35 D620N mutation, 12 suggesting the inability of the receptor to be efficiently transported from endosomes to the TGN. Therefore, the expression of the Vps35 D620N mutant may result in a reduced ability of this mutant to create transport vesicles from the endosome, leading to retention of retromer and cargo on the altered endosomes, not from an inability of Vps35 D620N retromer to engage its cargo.

Interestingly, the defect in CI-M6PR trafficking that manifests in missorting of cathepsin D in the presence of Vps35 D620N mutation may have a direct correlation to the development and progression of Parkinson’s disease as α-synuclein, a protein implicated in Parkinson’s disease pathogenesis, has been identified as a cathepsin D target protein within the lysosomes (9, 10). For example, overexpression of cathepsin D in dopaminergic cell cultures increases proteolysis of endogenous α-synuclein, while cathepsin D knockout mice show improper processing of α-synuclein in dopaminergic neurons that ultimately leads to α- synuclein cellular toxicity (39). Additionally, enzymatic inactivation of cathepsin D leads to signs of early onset, progressively fatal neurodegenerative disease in humans (40-42). This further supports the notion that the regulation of lysosomal proteolysis of α-synuclein is an important contributing factor in PD pathogenesis. Therefore, it is plausible that the molecular mechanisms that underpin the Vps35 D620N disease manifestation include altered trafficking of CI-M6PR and its cargo, cathepsin D, ultimately leading to impaired degradation of proteins delivered to the lysosome, including α-synuclein and resulting in formation of Lewy bodies, a hallmark of Parkinson’s disease.

Overall, we have shown that Parkinson’s disease linked Vps35 D620N mutant causes a deficit in retromer-dependent trafficking of CI-M6PR, and its ligand cathepsin D, likely arising from the generation and redistribution of enlarged endosome compartments that retain retromer. The ability for this mutation to contribute to the pathogenesis of Parkinson’s disease is likely secondary to these reported trafficking defects by reducing the breakdown of disease-associated proteins such as α-synuclein, in the late endosomal network.

13

Methods and Materials DNA Constructs: Full length human wild-type Vps35 was amplified using a 5′ primer CCCCTCGAGATGCCTACAACACAGCAG and 3′ primer CCCGGATCCAGAAGGATGAGACCTTCATA and sub-cloned into pEGFP-N1 using BamH1 and XhoI. The D620N point mutation was generated using a Quikchange mutagenesis kit (Stratagene) using the following PCR mutagenic primer pair: 5′- GAAGATGAAATCAGTAATTCTAAAGCACAGCTG and 3′- CAGCTGTGCTTTCGAATTACTGATTTCATCTTC. Constructs for expression in Escherichia coli of mouse Vps35, Vps29 and Vps26A were described previously (43). The D620N point mutation within Vps35 inserted into pGEX4T-2 (GE Healthcare) was generated as described above for the human construct. The pCMU-CD8/CI-M6PR construct was described previously (16).

Antibodies: Monoclonal mouse antibodies against human p230 and human EEA1 were purchased from BD Transduction Laboratories. Mouse monoclonal anti-CIM6PR, anti- LAMP1, goat polyclonal anti-Vps35 and rabbit polyclonal anti-Vps26A were purchased from Abcam. Mouse monoclonal anti-β-tubulin was purchased from Sigma Aldrich. Rabbit polyclonal anti-GFP was purchased from Life Technologies. Anti-cathepsin D antibody was purchased from Cell Signaling Technology. Anti-CD8 antibody was described previously (16). Phalloidin-fluorescent conjugates were purchased from Molecular Probes (Life technologies). Secondary donkey anti-rabbit IgG Alexa Fluor488, goat anti-mouse IgG Alexa Fluor568 and goat anti-mouse IgG Alexa Fluor647 were purchased from Life Technologies. Horse-radish peroxidase-conjugated swine anti-rabbit antibody was purchased from Dako. Dextran (MW 10,000 Da) conjugated to Alexa-647 was purchased from Life Technologies.

Protein purification and isothermal titration calorimetry (ITC) and circular dichroism (CD) spectroscopy: Recombinant proteins used for the ITC experiments were prepared as previously described (43). All proteins were further purified by gel filtration chromatography using 20 mM Tris (pH 8.0), 200 mM NaCl, 1 mM DTT (ITC buffer). Isothermal titration calorimetry was carried out at 283 K using a MicroCal iTC200 (GE Healthcare), with 16 x 2.5 µl injections of 100 M Vps29 into 10 M Vps35, or 50 M Vps26A into 5 M Vps35. Integration of the titration curves was performed using the ORIGIN software (OriginLab) to extract thermodynamic parameters, stoichiometry N, equilibrium association constant Ka

-1 (=Kd ) and the binding enthalpy H. The Gibbs free energy of binding G was calculated

14 from the relation G = -RTln(Ka) and the binding entropy S was deduced from the equation (G = H - TS). CD spectra were recorded on Jasco-810 spectropolarimeter (Jasco GmbH) at 0.01 cm-1 optical path, 0.5 nm interval, 1 nm bandwidth and 50 nm/min scanning speed. Protein samples were set at 3 mg/ml in ITC buffer and their polarization spectra were recorded three times followed by averaging and background subtraction. The HT voltage was below 500 V over the entire range of recording (200-250 nm).

Cell Culture and Transfection: A431 and HEK293 cells were grown in a humidified 37°C incubator with 5% CO2 and maintained in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% foetal bovine serum (FBS) and 2 mM L-glutamine (Life Technologies). Fibroblast cells were grown in DMEM (Life Technologies) supplemented with 10% FBS and 2 mM L-glutamine. Cultures were maintained in T25 culture flasks and split at 1:4 throughout experimentation. Mammalian constructs were transfected into cells using LipofectAMINE 2000 (Life Technologies) according to manufacturer's instructions.

Immunoprecipitations: Cell monolayers were washed in ice-cold phosphate buffered saline (PBS) and lysed in TK lysis buffer (50 mM HEPES, 150 mM NaCl, 1% Triton X-100, 10 mM

Na4P2O7, 30 mM NaF, 2 mM Na3VO4, 10 mM EDTA, 0.5 mM AEBSF and Complete Mini protease inhibitor cocktail) for 10 min on ice. Cell lysates were centrifuged at 17,000 x g and supernatant incubated with GFP-NanoTrap beads (44) for 2 h at 4oC under constant rotation. Complexes were removed from the GFP-NanoTrap beads by boiling after three consecutive washes with TK lysis buffer.

Western Blotting: Protein lysates were resolved on SDS-PAGE and transferred onto a PVDF membrane (Immobilon-P and Immobilon-FL; Millipore) according to the manufacturer's instructions. Western blotting using ECL and Odyssey infrared imaging system (LI-COR Biosciences) were performed as described previously (16).

Patient Data: An isolated skin biopsy was taken from a 73 y/o European Male Parkinson’s disease patient previously genotyped for the Vps35 variant, D620N. The patient had a diagnosis of Parkinson’s disease according to the UK Brian Bank clinical criteria and a self- reported age of symptom onset of 46 years. The single heterozygote point mutation in Exon 15 of the Vps35 gene leading to the D620N amino acid change was confirmed by Sanger 15 sequencing of DNA isolated from the cell line. Control human fibroblasts were isolated from a healthy, neurologically normal 44 year old male subject on no medication and with no family history of PD.

Establishment of Skin Biopsy Cultures: Isolated biopsy samples were placed in 50 mL conical flasks filled with cell culture medium (DMEM: F12 supplemented with 10% FBS and stored on ice until tissue dissociation. Samples were kept on ice for no longer than 24 h. Cell culture medium was aspirated and biopsy samples were washed a total of three times with 10 mL of room temperature PBS. Tissue were transferred into 6 cm sterile culture dishes containing pre-warmed culture medium and divided into pinhead-sized explants using a scalpel blade. Using a 1 mL tuberculin syringe with a 23-G needle attached, one pinhead sized explant was placed into a 25 cm flask. Culture flasks were then placed in a

37 ºC, 5% CO2 humidified circulating incubator for 5-10 min allowing sufficient time for the explants to attach to the flask before adding pre-warmed culture medium. 5 mL of culture medium was added to the adhered cultures and returned to the 37 ºC, 5% CO2 incubator. Cultures were left to grow for a minimum of 72 h before being handled. On the 4th day, cultures were checked and those who displayed evidence of radial outgrowth from initial explants were supplemented with fresh medium and left for another 14-21 days, with periodic medium changes, before subculturing.

Microscopy of live cells: A431 cells were plated in 35 mm glass bottom dishes (MatTek Corporation) 48 h prior to use and transfected with indicated GFP fusion Vps35 constructs. Dextran loading was performed as previously described (45). Briefly, dextran conjugated to Alexa-647 (MW 10,000 Da) (Life Technologies) at a final concentration of 100 µg/mL was loaded into the A431 cells by incubation at 37°C for 1 h in complete media. Cells were then washed to remove excess dextran and imaged in CO2-independent media supplemented with 10% FBS (Life Technologies). Time-lapse microscopy was performed by capturing 1 Airy Unit (1AU) z-slices using 63x objective on Zeiss LSM 710 FCS Inverted Scanning Laser confocal microscope. Movies were edited using Image J 1.47f and still capture frames from the movies were edited using Adobe Photoshop.

Indirect Immunoflourescence: A431 cells grown on coverslips were transiently transfected with mammalian constructs, fixed and stained with indicated antibodies as described previously (16). Coverslips were mounted using Fluorescent Mounting Medium (Dako) and imaging was performed using 63x objective on a Ziess LSM 710 Upright Scanning Laser 16 confocal fluorescent microscope. Images represent a 1AU z-plane single slice. All images were analyzed using Zeiss LSM 5.0 and Adobe Photoshop software.

Nocadazole Treatment: Cell monolayers were treated with 2 µM Nocadazole (Sigma-

Aldrich) for 60 min in a humidified 37°C incubator with 5% CO2. Medium was aspirated and cells were washed three times in PBS prior to PFA fixation.

Secretion Assay: HEK293 cells were plated in 6-well dishes 48 h prior to use and transfected with GFP fusion constructs for 24 h. Cyclohexamide (final concentration 100 µg/ml, Sigma Aldrich) was diluted in serum-free DMEM supplemented with 2 mM L-Glutamine (Life Technologies) and used throughout experimentation. Medium samples were collected at indicated times and analysed using western blotting.

Quantification of the intracellular distribution of endosomes: To quantify perinuclear/cytoplasmic localization, a macro was created in ImageJ* (version 1.47i). Two channels were utilized from each image: a DAPI channel to define the nuclear regions; and the protein of interest (POI) channel. A typical image contains 5 or 6 nuclei in the field of view. First, the nuclear image was converted to a binary mask to define the nuclear regions using an auto-threshold algorithm. Nuclei touching the edge of the image were excluded from further analysis. A perinuclear region around the nuclei was then defined as follows. The nuclear mask regions selected were expanded by a distance of 40 pixels (corresponding to a distance of 1.5 µm). Removing the nuclear region from this expanded region selection then gave an annular region around, but not including, each nucleus. The average intensity in the POI channel of the union of all the perinuclear annuli regions was then recorded. To select a region more distant from the nuclei, a similar process was used: the nuclear mask regions were expanded by 80 pixels (corresponding to 3.0 µm) and a region obtained by expanding the nuclear regions by 40 pixels subtracted from it. The effect was to generate an annulus adjacent to the perinuclear annuli but more distant from the nuclei, i.e. a cytoplasmic region. In the POI channel, the average intensity across the union of these 'cytoplasmic' regions was then recorded for the image. Finally, for comparison between images and experiments, the ratio of the so calculated average perinuclear intensity to the average cytoplasmic region intensity was recorded. The script was fully automated with the only user interaction required being the selection of the directory containing the images to be quantified. The macro is available in supplementary material (Figure S1). 17

Colocalization analysis: Immunoflourescence images (1AU slices) were captured using a Ziess LSM 710 Upright Scanning Laser confocal fluorescent microscope under 63x magnification using identical laser power, light pathways and band passes. Captured images were analyzed using Image J Version J 1.47f. Multi-channel images were split into corresponding grey-scale format and threshold settings were applied to each individual channel. Under these conditions colocalization was quantified using the Image J plug-in, Colocalization finder (U.S. National Institutes of Health, Bethesda, MD, USA (46). http://rsb.info.nih.gov/ij/plugins/colocalization-finder.html. Raw data representing colocalization co-efficient was analyzed using GraphPad Prism 5 software version 5.03. Graphs represent a global average of data collected.

Acknowledgements

This work was supported by funding from the National Health and Medical Research Council (NHMRC) of Australia (APP511072, APP631584, APP1025538, APP1042082, APP1010225), Australian Research Council (DP120103930), ANZ Trustees National Medical Program Grant from Judith Jane Mason & Harold Stannett Williams Memorial Foundation, and Parkinson’s Queensland (to MM). RDT is supported by NHMRC Senior Research Fellowship (APP1041929) and BMC is supported by Australian Research Council future fellowship (FT100100027). Microscopy was carried out at the Australian Cancer Research Foundation (ACRF)/ Institute for Molecular Bioscience (IMB) Dynamic Imaging Facility for Cancer Biology. The authors thank Seetha Karunaratne, IMB, UQ, for generation of the full length human Vps35 WT used throughout this project.

18

References

1. Pountney DL, Treweek TM, Chataway T, Huang Y, Chegini F, Blumbergs PC, Raftery MJ, Gai WP. αB-Crystallin is a major component of glial cytoplasmic inclusions in multiple system atrophy. neurotox res 2005;7(1-2):77-85.

2. Auluck PK, Chan HYE, Trojanowski JQ, Lee VM-Y, Bonini NM. Chaperone Suppression of α-Synuclein Toxicity in a Drosophila Model for Parkinson's Disease. Science 2002;295(5556):865-868.

3. Uéda K, Fukushima H, Masliah E, Xia Y, Iwai A, Yoshimoto M, Otero DA, Kondo J, Ihara Y, Saitoh T. Molecular cloning of cDNA encoding an unrecognized component of amyloid in Alzheimer disease. Proceedings of the National Academy of Sciences 1993;90(23):11282-11286.

4. Han H, Weinreb PH, Lansbury Jr PT. The core Alzheimer's peptide NAC forms amyloid fibrils which seed and are seeded by β-amyloid: is NAC a common trigger or target in neurodegenerative disease? Chemistry & Biology 1995;2(3):163-169.

5. Langston J, Ballard P, Tetrud J, Irwin I. Chronic Parkinsonism in humans due to a product of meperidine-analog synthesis. Science 1983;219(4587):979-980.

6. Dawson TM, Ko HS, Dawson VL. Genetic Animal Models of Parkinson's Disease. Neuron 2010;66(5):646-661.

7. Exner N, Lutz AK, Haass C, Winklhofer KF. Mitochondrial dysfunction in Parkinson's disease: molecular mechanisms and pathophysiological consequences. EMBO J 2012;31(14):3038-3062.

8. Cuervo AM, Stefanis L, Fredenburg R, Lansbury PT, Sulzer D. Impaired degradation of mutant α-Synuclein by chaperone-mediated autophagy. Science 2004;305(5688):1292- 1295.

9. Hossain S, Alim A, Takeda K, Kaji H, Shinoda T, Ueda K. Limited proteolysis of NACP/alpha-synuclein. J Alzheimers Dis 2001;3(6):577-584.

19

10. Sevlever D, Jiang P, Yen SH. Cathepsin D is the main lysosomal enzyme involved in the degradation of alpha-synuclein and generation of its carboxy-terminally truncated species. Biochemistry 2008;47(36):9678-9687.

11. Qiao L, Hamamichi S, Caldwell K, Caldwell G, Yacoubian T, Wilson S, Xie Z-L, Speake L, Parks R, Crabtree D, Liang Q, Crimmins S, Schneider L, Uchiyama Y, Iwatsubo T, et al. Lysosomal enzyme cathepsin D protects against alpha-synuclein aggregation and toxicity. Molecular Brain 2008;1(1):17.

12. Seaman MN. The retromer complex - endosomal protein recycling and beyond. J Cell Sci 2012;125(Pt 20):4693-4702.

13. Seaman MN. Identification of a novel conserved sorting motif required for retromer- mediated endosome-to-TGN retrieval. J Cell Sci 2007;120(Pt 14):2378-2389.

14. Arighi CN, Hartnell LM, Aguilar RC, Haft CR, Bonifacino JS. Role of the mammalian retromer in sorting of the cation-independent mannose 6-phosphate receptor. J Cell Biol 2004;165(1):123-133.

15. Choy RW-Y, Cheng Z, Schekman R. Amyloid precursor protein (APP) traffics from the cell surface via endosomes for amyloid β (Aβ) production in the trans-Golgi network. Proceedings of the National Academy of Sciences 2012.

16. Bugarcic A, Zhe Y, Kerr MC, Griffin J, Collins BM, Teasdale RD. Vps26A and Vps26B subunits define distinct retromer complexes. Traffic; 2011. p. 1759-1773.

17. Gieselmann V, Hasilik A, von Figura K. Processing of human cathepsin D in lysosomes in vitro. J Biol Chem 1985;260(5):3215-3220.

18. Seaman MNJ. Cargo-selective endosomal sorting for retrieval to the Golgi requires retromer. The Journal of Cell Biology 2004;165(1):111-122.

19. Rojas R, van Vlijmen T, Mardones GA, Prabhu Y, Rojas AL, Mohammed S, Heck AJR, Raposo G, van der Sluijs P, Bonifacino JS. Regulation of retromer recruitment to endosomes by sequential action of Rab5 and Rab7. The Journal of Cell Biology 2008;183(3):513-526.

20

20. Laurent-Matha V, Derocq D, Prébois C, Katunuma N, Liaudet-Coopman E. Processing of Human Cathepsin D Is Independent of Its Catalytic Function and Auto- Activation: Involvement of Cathepsins L and B. Journal of Biochemistry 2006;139(3):363- 371.

21. Lesage S, Condroyer C, Klebe S, Honore A, Tison F, Brefel-Courbon C, Durr A, Brice A. Identification of VPS35 mutations replicated in French families with Parkinson disease. Neurology 2012;78(18):1449-1450.

22. Ando M, Funayama M, Li Y, Kashihara K, Murakami Y, Ishizu N, Toyoda C, Noguchi K, Hashimoto T, Nakano N, Sasaki R, Kokubo Y, Kuzuhara S, Ogaki K, Yamashita C, et al. VPS35 mutation in Japanese patients with typical Parkinson's disease. Movement Disorders 2012;27(11):1413-1417.

23. Zimprich A, Benet-Pagès A, Struhal W, Graf E, Eck Sebastian H, Offman Marc N, Haubenberger D, Spielberger S, Schulte Eva C, Lichtner P, Rossle Shaila C, Klopp N, Wolf E, Seppi K, Pirker W, et al. A mutation in VPS35, encoding a subunit of the retromer complex, causes late-onset Parkinson disease. The American Journal of Human Genetics 2011;89(1):168-175.

24. Zhang Y, Chen S, Xiao Q, Cao L, Liu J, Rong T-Y, Ma J-F, Wang G, Wang Y, Chen S-D. Vacuolar protein sorting 35 Asp620Asn mutation is rare in the ethnic Chinese population with Parkinson's disease. Parkinsonism & Related Disorders 2012;18(5):638- 640.

25. Sharma M, Ioannidis JPA, Aasly JO, Annesi G, Brice A, Bertram L, Bozi M, Barcikowska M, Crosiers D, Clarke CE, Facheris MF, Farrer M, Garraux G, Gispert S, Auburger G, et al. A multi-centre clinico-genetic analysis of the VPS35 gene in Parkinson disease indicates reduced penetrance for disease-associated variants. Journal of Medical Genetics 2012;49(11):721-726.

26. Hierro A, Rojas AL, Rojas R, Murthy N, Effantin G, Kajava AV, Steven AC, Bonifacino JS, Hurley JH. Functional architecture of the retromer cargo-recognition complex. Nature 2007;449(7165):1063-1067.

27. Roberts RL, Barbieri MA, Pryse KM, Chua M, Morisaki JH, Stahl PD. Endosome fusion in living cells overexpressing GFP-rab5. J Cell Sci 1999;112(Pt 21):3667-3675.

21

28. Cataldo AM, Peterhoff CM, Troncoso JC, Gomez-Isla T, Hyman BT, Nixon RA. Endocytic pathway abnormalities precede amyloid beta deposition in sporadic Alzheimer's disease and Down syndrome: differential effects of APOE genotype and presenilin mutations. Am J Pathol 2000;157(1):277-286.

29. Rink J, Ghigo E, Kalaidzidis Y, Zerial M. Rab conversion as a mechanism of progression from early to late endosomes. Cell 2005;122(5):735-749.

30. Seaman MN, Harbour ME, Tattersall D, Read E, Bright N. Membrane recruitment of the cargo-selective retromer subcomplex is catalysed by the small GTPase Rab7 and inhibited by the Rab-GAP TBC1D5. J Cell Sci 2009;122(Pt 14):2371-2382.

31. Jager S, Bucci C, Tanida I, Ueno T, Kominami E, Saftig P, Eskelinen EL. Role for Rab7 in maturation of late autophagic vacuoles. J Cell Sci 2004;117(Pt 20):4837-4848.

32. Gomez TS, Gorman JA, de Narvajas AA, Koenig AO, Billadeau DD. Trafficking defects in WASH-knockout fibroblasts originate from collapsed endosomal and lysosomal networks. Mol Biol Cell 2012;23(16):3215-3228.

33. Gomez TS, Billadeau DD. A FAM21-containing WASH complex regulates retromer- dependent sorting. Dev Cell 2009;17(5):699-711.

34. Derivery E, Sousa C, Gautier JJ, Lombard B, Loew D, Gautreau A. The Arp2/3 activator WASH controls the fission of endosomes through a large multiprotein complex. Dev Cell 2009;17(5):712-723.

35. Zech T, Calaminus SD, Caswell P, Spence HJ, Carnell M, Insall RH, Norman J, Machesky LM. The Arp2/3 activator WASH regulates alpha5beta1-integrin-mediated invasive migration. J Cell Sci 2011;124(Pt 22):3753-3759.

36. Harbour ME, Breusegem SY, Seaman MNJ. Recruitment of the endosomal WASH complex is mediated by the extended ‘tail’ of Fam21 binding to the retromer protein Vps35. Biochemical Journal 2012;442(1):209-220.

37. Jia D, Gomez TS, Metlagel Z, Umetani J, Otwinowski Z, Rosen MK, Billadeau DD. WASH and WAVE actin regulators of the Wiskott–Aldrich syndrome protein (WASP) family are controlled by analogous structurally related complexes. Proceedings of the National Academy of Sciences 2010;107(23):10442-10447. 22

38. Steinberg F, Gallon M, Winfield M, Thomas EC, Bell AJ, Heesom KJ, Tavare JM, Cullen PJ. A global analysis of SNX27-retromer assembly and cargo specificity reveals a function in glucose and metal ion transport. Nat Cell Biol 2013;15(5):461-471.

39. Cullen V, Lindfors M, Ng J, Paetau A, Swinton E, Kolodziej P, Boston H, Saftig P, Woulfe J, Feany MB, Myllykangas L, Schlossmacher MG, Tyynela J. Cathepsin D expression level affects alpha-synuclein processing, aggregation, and toxicity in vivo. Mol Brain 2009;2(5):1756-6606.

40. Siintola E, Partanen S, Stromme P, Haapanen A, Haltia M, Maehlen J, Lehesjoki AE, Tyynela J. Cathepsin D deficiency underlies congenital human neuronal ceroid- lipofuscinosis. Brain 2006;129(Pt 6):1438-1445.

41. Tyynela J, Sohar I, Sleat DE, Gin RM, Donnelly RJ, Baumann M, Haltia M, Lobel P. A mutation in the ovine cathepsin D gene causes a congenital lysosomal storage disease with profound neurodegeneration. EMBO J 2000;19(12):2786-2792.

42. Koike M, Nakanishi H, Saftig P, Ezaki J, Isahara K, Ohsawa Y, Schulz-Schaeffer W, Watanabe T, Waguri S, Kametaka S, Shibata M, Yamamoto K, Kominami E, Peters C, von Figura K, et al. Cathepsin D deficiency induces lysosomal storage with ceroid lipofuscin in mouse CNS neurons. J Neurosci 2000;20(18):6898-6906.

43. Norwood SJ, Shaw DJ, Cowieson NP, Owen DJ, Teasdale RD, Collins BM. Assembly and Solution Structure of the Core Retromer Protein Complex. Traffic 2011;12(1):56-71.

44. Kubala MH, Kovtun O, Alexandrov K, Collins BM. Structural and thermodynamic analysis of the GFP:GFP-nanobody complex. Protein Science 2010;19(12):2389-2401.

45. Kerr MC, Wang JTH, Castro NA, Hamilton NA, Town L, Brown DL, Meunier FA, Brown NF, Stow JL, Teasdale RD. Inhibition of the PtdIns(5) kinase PIKfyve disrupts intracellular replication of Salmonella. EMBO J 2010;29(8):1331-1347.

46. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat Meth 2012;9(7):671-675.

23

Figure legends

Figure 1. Vps35 mutant D620N binds to Vps29 and Vps26A in vitro and in vivo. (A) Structure of the Vps35 (483-780) complex with Vps29 (26). Vps35 is shown in grey ribbons with transparent surface and Vps29 is shown in green ribbons. The D620N mutation is indicated in red. (B) CD spectra of Vps35 WT and Vps35 D620N proteins. (C) Isothermal titration calorimetry of purified recombinant Vps35 WT and Vps35 D620N protein to retromer subunits, Vps26A (right panel) and Vps29 (left panel) based on 16 x 2.5 ml injections of 100 µM Vps29 into 10 µM Vps35, or 50 µM Vps26A into 5 µM Vps35 D620N. Data shown represents integrated values of titration binding curves. (D) Sub-confluent HEK293 cells were transiently transfected with GFP, Vps35 WT-GFP or Vps35 D620N-GFP for 16 h washed with ice-cold PBS and lysed on ice. Protein complexes were isolated using GFP- Nanotrap beads, resolved by SDS-PAGE and transferred to PVDF membranes. Membranes were incubated with primary antibodies against GFP, Vps26A or Vps29 followed by IRDye 680/800 secondary antibodies and imaged using the Odyssey infrared imaging system.

Figure 2: Ectopic expression of Vps35 D620N alters endosome morphology and distribution. (A) Live cell time lapse confocal microscopy was performed on A431 cells expressing Vps35 WT-GFP or Vps35 D620N-GFP pulsed with 200 µg/ml Dextran-647 for 90 min using a Ziess LSM 710 FCS scanning confocal microscope. Image represents single frame of 30 min movie captured. Scale bar 10 µm. (B) A431 cells expressing Vps35 WT-GFP or Vps35 D620N-GFP were fixed and indirect immunofluorescence was performed using antibodies against Vps26A. (C) Quantification of the intracellular endosomal distribution based on perinuclear intensity ratio of A431 cells over-expressing Vps35 WT-GFP or VPS35 D620N- GFP. Graph represents the mean of two independent experiments with fifteen images each

(n=2;**p<0.01, Error bars +/- SEM). (D) A431 cells transfected with construct expressing Vps35 D620N-GFP were treated with 2 µM nocadazole for 60 min at 37°C, fixed and indirect immunofluorescence was performed using antibodies against anti-β-tubulin.

Figure 3: Ectopically expressed Vps35 D620N positive endosomes contain both early and late endosome markers. (A) A431 cells transfected with constructs expressing Vps35 WT-GFP or Vps35 D620N- GFP were fixed and indirect immunofluorescence performed using monoclonal antibodies against EEA1, p230 or LAMP1 and counterstained with DAPI. All images represent a 1AU 24 single slice captured using a Ziess LSM 710 Upright Scanning Laser confocal microscope at 63x magnification. Scale bar 5 µm. (B) Quantification of colocalization between Vps35 WT-GFP or Vps35 D620N-GFP and EEA1, p230 or LAMP1. Graph represents the mean of three independent experiments with ten images each (n=3;**p<0.01, Error bars represent +/- SEM).

Figure 4: Ectopically expressed Vps35 D620N interacts with CI-M6PR but alters the receptors’ capacity to transport cathepsin D. (A) HEK293 cells co-transfected with Vps35 WT-GFP/CD8-CI-M6PR or Vps35 D620N- GFP/CD8-CI-M6PR were lysed and co-immunoprecipitation was performed using GFP- NanoTrap beads. Total cell lysates (input; 50 µg per lane) and isolated interacting proteins were analysed by western immunoblotting using antibodies to CD8. (B) Sub-confluent A431 cells grown on coverslips were transiently transfected with Vps35 WT-GFP or Vps35 D620N-GFP for 16 h, fixed and indirect immunofluorescence performed using a mouse monoclonal against endogenous CI-M6PR. Images represent a 1AU single slice captured using a Ziess LSM 710 Upright Scanning Laser confocal microscope at 63x magnification. Scale bar 5 µm. (C) Sub-confluent A431 cells transiently expressing Vps35-GFP or Vps35D620N-GFP for 24 h were pulsed with 100 µg/mL of cycloheximide in serum-free medium. Medium and corresponding cell lysates were collected at 0, 3 and 7 h post treatment. Media samples (500 µl), precipitated on ice using 10% TCA, and total cell lysates (50 µg per lane) were analysed for levels of cathepsin D and β-tubulin by western immunoblotting.

Figure 5: Parkinson’s disease Vps35 D620N patient fibroblasts have redistributed endosomes. (A) The colocalisation of endogenous Vps35 in sub-confluent control and Parkinson’s disease patient fibroblasts using antibodies against Vps26A, EEA1, p230 or LAMP1 was determined using indirect immunofluorescence. Cell monolayers were counterstained with DAPI. All images represent a 1AU single slice captured using a Ziess LSM 710 Upright Scanning Laser confocal microscope at 63x magnification. Scale bar 5µm. (B) Quantification of colocalization analysis performed in (A). Graph represents the mean of two independent experiments with ten images each (n=2;*p<0.05, Errors bars represent +/- SEM). (C) The intracellular distribution of Vps35 positive endosomes quantified using perinuclear intensity ratio in control and Parkinson’s disease patient fibroblasts. Graph represents the mean of

25 two independent experiments with ten images each (n=2;*p<0.05, Error bars represent +/- SEM).

Figure 6: Cathepsin D processing is impaired in Parkinson’s disease Vps35 D620N patient fibroblasts. (A) Indirect immunofluorescence on sub-confluent control and Parkinson’s disease patient fibroblasts was performed using antibodies against Vps35, CI-M6PR and counterstained with DAPI. Images represent a 1AU single slice captured using a Ziess LSM 710 Upright Scanning Laser confocal microscope. Scale bar 5 µm. (B) Quantification of colocalization analysis performed in (A). Graph represents the mean of two independent experiments with ten images each (n=2;*p<0.05, Error bars represent +/- SEM). (C) Monolayers of control and Parkinson’s disease patient fibroblasts were pulsed with 100 µg/mL of cyclohexamide for 0, 3 and 7 h, when cells were lysed and analyzed by western immunoblotting using α- cathepsin D antibody.

Supplementary Figure 1: Quantification of peri-nuclear and cytoplasmic protein. To quantify perinuclear/cytoplasmic localization, a macro was created in ImageJ* (version 1.47i). The macro is fully automated with the only user interaction required being the selection of the directory containing the images to be quantified. Two channels are utilized by the script from each image: a DAPI channel to define the nuclear regions (blue channel); and the protein of interest (POI) channel (green channel). For each image, the ratio of total protein found in the perinuclear to cytoplasmic regions (as defined by distance from the nuclei in the images) is then calculated and recorded in an Excel file. See Materials and Methods for details of the quantification process.

Supplementary Movie 1: Vps35 WT-GFP positive endosomes are mobile and found throughout the cytoplasm. Details in Figure legend 2A.

Supplementary Movie 2: Vps35 D620N-GFP positive endosomes are enlarged and redistributed to a perinuclear localization but retain ability to receive endocytosed material. Details in Figure legend 2A.

26