Defining the molecular action of in retrograde trafficking pathways

Yi Cui

MSc. Physiology

BSc. Biotechnology

A thesis submitted for the degree of Doctor of Philosophy at

The University of Queensland in 2019

Faculty of Medicine

School of Biomedical Sciences

Abstract

The endo-lysosomal network is a highly dynamic and orchestrated system which serves a vital function in coordinating the communication and the exchange of cargo molecules between intracellular membrane-bound compartments. The endo-lysosomal system is critical for the maintenance of intracellular , and defects on it are often associated with neurological disorders. Retromer is a peripheral membrane complex that coordinates multiple vesicular trafficking events. However, the molecular actions of retromer in the endo-lysosomal system remain unclear and controversial.

This thesis is focused on the characterization of retromer’s molecular action in the retrograde trafficking of transport carriers (ETCs) and dissects the functional association between retromer and its accessory . Firstly, this study demonstrates a selective function of retromer in the retrograde sorting via ETCs. The incorporation of CI-M6PR, via a retromer-dependent method, into ETCs, is restricted to those captured by the trans-Golgi protein GCC88, but not other trans-golgins including golgin-97 or golgin-245. This retromer- dependent retrograde trafficking pathway requires SNX3, but not other retromer-associated cargo binding proteins, such as SNX-BAR proteins and SNX27. Besides, this study shows the requirement of retromer in the maintenance of lysosomal morphology and function. The absence of retromer alters the lysosomal ultrastructure, impairs the autophagic process and the lysosomal .

Secondly, this study demonstrates that GCC88, the tethering factor in retromer-dependent trafficking pathway is required for the endosome-to-TGN retrieval of CI-M6PR and the maintenance of lysosomal proteolytic activity. However, GCC88 deficiency has no impact on the -lysosomal pathway.

Lastly, the final chapter of this thesis demonstrates the Parkinson’s disease-linked retromer variant - Vps35 D620N confers a partial loss of function. The presence of the Vps35 D620N variant rescues the lysosomal defects and the endosomal dissociation of TBC1D5 and the WASH complex subunit FAM21, which are caused by the absence of retromer. However, the Vps35 D620N variant fails to rescue the trafficking defects of retromer-dependent GCC88-captured ETCs.

i 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, financial support 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 higher degree by research 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 and have sought permission from co- authors for any jointly authored works included in the thesis.

ii Publications during candidature

Cui, Y., Yang, Z., Teasdale, R. 2017. The functional roles of retromer in Parkinson's disease. FEBS Lett. 2017 Dec 6. doi: 10.1002/1873-3468.12931

Cui, Y., Carosi, J., Yang, Z., Ariotti, N., Kerr, M., Parton, R., Sargeant, T., Teasdale, R. Retromer has a selective function in cargo sorting via endosome transport carriers. J Cell Biol. 2019 Feb 4. doi: 10.1083/jcb.201806153

Cui, Y., Yang, Z., Teasdale, R. A role of GCC88 in the retrograde transport of CI-M6PR and the maintenance of lysosomal activity. CELL BIOL INT. 2019 Feb 21. doi: 10.1002/cbin.11118

Carosi J, Hattersly K., Cui Y., Yang, Z., Teasdale, R., Sargeant, T. Sargeant1Subcellular fractionation of HeLa cells for enrichment using a continuous Percoll-2 density gradient. Bio-Protocol. 2019 Aug. Accepted.

Conference abstracts

Cui, Y., Yang, Z., Ariotti, N., Flores-Rodriguez N, Parton, R., Teasdale, RD. The selective function of retromer in cargo sorting via endosomal transport carriers. EMBO conference, Cell Polarity and Membrane Dynamics. Sant Feliu de Guixols, Spain, May 26th – 31th 2019 (Poster presentation and flash talk)

Cui, Y., Carosi, J., Yang, Z., Ariotti, N., Kerr, M., Parton, R., Sargeant, T., Teasdale, RD. A selective function of retromer in cargo sorting via endosomal transport carriers. SBMS International Postgraduate Symposium. Brisbane, Australia, Oct 29th - 30th 2018 (Oral presentation).

Cui, Y., Carosi, J., Yang, Z., Kerr, M., Sargeant, T. and Teasdale, R. Retromer is required for the retrograde sorting of cation-independent mannose 6-phosphate into a subset of endosome transport carriers. Combio2018 conference. Sydney, Australia, Sep 23th - 26th 2018 (Oral presentation).

Cui, Y., Yang, Z., Kerr, M., Follett, J. and Teasdale, R. Does Parkinson’s disease associated in retromer impact on the formation of retrograde transport carriers? ASMR Queensland Postgraduate Student Conference, Brisbane, Australia, June 1st, 2016 (Poster presentation).

iii Publications included in this thesis

Cui, Y., Yang, Z., Teasdale, RD. 2017. The functional roles of retromer in Parkinson's disease. FEBS Lett. 2017 Dec 6. doi: 10.1002/1873-3468.12931

Incorporated into Chapter 1

Incorporated into Chapter 1 Contributor Statement of Contribution Yi Cui (candidate) Wrote and edited the manuscript (70%) Zhe Yang Wrote and edited the manuscript (15%) Rohan D. Teasdale Wrote and edited the manuscript (15%)

iv Cui, Y., Carosi, J., Yang, Z., Ariotti, N., Kerr, M., Parton, R., Sargeant, T., Teasdale, RD. Retromer has a selective function in cargo sorting via endosome transport carriers. J Cell Biol. 2019 Feb 4. doi: 10.1083/jcb.201806153

Incorporated as Chapter 2

Incorporated into Chapter 2

Contributor Statement of Contribution

Yi Cui (Candidate) Study conception and design (30%) Acquisition of data (78%) Drafting of manuscript (70%) Editing manuscript (15%)

Julian Carosi Study conception and design (10%) Acquisition of data (10%) Drafting of manuscript (10%) Reviewing and editing manuscript (5%)

Zhe Yang Study conception and design (30%) Acquisition of data (8%) Drafting of manuscript (10%) Reviewing and editing manuscript (20%)

Nicholas Ariotti Acquisition of data (2%)

Markus C Kerr. Acquisition of data (2%)

Robert G. Parton Reviewing and editing manuscript (5%)

Tim Sargeant Reviewing and editing manuscript (5%)

Rohan D. Teasdale Study conception and design (30%) Analysis and interpretation of data (25%) Drafting of manuscript (10%) Reviewing and editing manuscript (50%)

v

Cui, Y., Yang, Z., Teasdale, RD. A role of GCC88 in the retrograde transport of CI-M6PR and the maintenance of lysosomal activity. CELL BIOL INT. 2019 Feb 21. doi: 10.1002/cbin.11118

Incorporated as Chapter 3

Incorporated into Chapter 3

Contributor Statement of Contribution

Yi Cui (Candidate) Study conception and design (60%) Acquisition of data (70%) Drafting of manuscript (70%) Editing manuscript (10%)

Zhe Yang Study conception and design (20%) Acquisition of data (30%) Drafting of manuscript (15%) Reviewing and editing manuscript (40%)

Rohan D. Teasdale Study conception and design (20%) Drafting of manuscript (15%) Analysis and interpretation of data (15%) Reviewing and editing manuscript (50%)

Other publications during candidature

No other publications.

vi Contributions by others to the thesis

Assoc. Prof. Rohan Teasdale and Dr. Zhe Yang formed the Ph.D. supervisory team that was involved in the conception and design of the project, as well as assisting with the editing and review of this thesis.

Mr. Julian Corasi and Dr. Tim Sargeant provided the lysosomal purification study for Chapter 2. Dr. Nicholas Ariotti and Prof. Robert Parton provided the electronic microscopy study for Chapter 2 and Chapter 4. Dr. Markus Kerr provided the Vps35-GFP rescue cell lines used in Chapter 2 and Chapter 4. Mr. Hyun Kim developed the SNX1/2 dKO cells used in Chapter 2. Dr. Xiaying Qi provided the pcDNA3.1-Vps35 and pcDNA3.1-Vps35 D620N plasmid used in Chapter 2 and 4. Dr. Naftali Flores-Rodriguez provided technical guidance in the STED microscopy used in my project and provide the STED microscopy images for analysis performed in Chapter 4. Dr. Jordan Follett provided the preliminary characterization for Vps35 D620N for Chapter 4. Dr. Mie Wong and Prof. Sean Munro provided the plasmids encoding the mitochondria targeting golgins used in thesis.

vii Statement of parts of the thesis submitted to qualify for the award of another degree

No works submitted towards another degree have been include in this thesis.

Research Involving or Animal Subjects

No animal or human subjects were involved in this research.

viii Acknowledgements

First and the foremost, I want to express my deepest gratitude to my supervisor Assoc. Prof. Rohan Teasdale. I am thankful to you for giving me the opportunity to carry out my research in this group, and for providing your patient guidance, unreserved support, and exceptional scientific inputs throughout my Ph.D. More than this, I am grateful to you for mentoring me and proving me the existence of integrity and honesty in science. You have a tremendous influence on me, of course for the better, as a scientist.

I would like to extend my thanks to Dr. Zhe Yang. Thank you for your constant encouragement, support and insightful discussion in my project over the last three and a half years. I would also want to thank Dr. Markus Kerr, Dr. Jordan Follett and all our collaborators for your ideas and interest in my work.

To my Ph.D. thesis committee members, Prof. Rob Parton, Prof. Elizabeth Coulson, Dr. Victor Anggono and Dr. Julia Pagan, thank you for your time and valuable comments, but also the questions that incented me to widen my research from variable perspectives.

I would like to thank all the remainders of the Teasdale Laboratory members, both former and current for your support both in and out of the lab.

A special thank you to Prof. Suzanne Elliott. Thank you for your career advice, constant encouragement, and support in my last year of Ph.D.

Last but not least, I wish to thank my family and all my dear friends, who have been there to give me encouragement whenever I needed it.

ix Financial support

This research was supported by scholarship from China scholarship Council and funding form the Australian Research Council (DP160101573 to RDT).

Keywords retromer, endosome, retrograde trafficking, endosome transport carriers, Vps35 D620N, Parkinson’s disease, GCC88, lysosome, autophagy

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code: 060108, Protein Trafficking, 80%

ANZSRC code: 060110, Receptor and Membrane Biology, 10%

ANZSRC code: 060104, Cell , 10%

Fields of Research (FoR) Classification

FoR code: 0601, Biochemistry and Cell Biology, 100%

x

Dedication

This thesis is dedicated to the memory my grandpa, a kind and smart gentleman who I miss every day.

xi Table of Contents Abstract ...... i

Declaration by author ...... ii

Publications during candidature ...... iii

Conference abstracts ...... iii

Publications included in this thesis ...... iv

Other publications during candidature ...... vi

Contributions by others to the thesis ...... vii

Statement of parts of the thesis submitted to qualify for the award of another degree ...... viii

Research Involving Human or Animal Subjects ...... viii

Acknowledgements ...... ix

Financial support ...... x

Keywords ...... x

Australian and New Zealand Standard Research Classifications (ANZSRC) ...... x

Fields of Research (FoR) Classification ...... x

Dedication ...... xi

Chapter 1. Literature review ...... 1

1.1 Overview of the endosomal system...... 1 1.1.1 Endocytic recycling pathway ...... 2 1.1.2 The endo-lysosomal degradation system ...... 3 1.1.3 Endosome-to-TGN retrograde transport ...... 6

1.2 The Retromer complex ...... 11 1.2.1 Mechanisms for the endosomal recruitment of retromer...... 13 1.2.2 Cargo Recognition ...... 15 1.2.3 Morphogenesis of Tubular Endosomal Transport Carriers ...... 17 1.2.4 Fission of tubulovesicular transport carriers ...... 18

1.3 Retromer and neuronal health ...... 18 1.3.1 Retromer in Alzheimer’s disease ...... 20 1.3.2 Retromer in Parkinson’s disease ...... 21

1.4 Retromer in Parkinson's disease ...... 21 1.4.1 Functional insights into retromer deficiency and PD-linked mutations ...... 22

xii 1.4.2 Connectivity of Retromer with other PD-associated proteins ...... 28

1.5 Aims and Hypotheses ...... 32

Chapter 2. Retromer has a selective function in cargo sorting via endosome transport carriers ...... 34

2.1 Introduction ...... 34

2.2 Materials and methods ...... 36 2.2.1 Chemicals, DNA constructs and antibodies ...... 36 2.2.2 Cell culture and Transfection ...... 37 2.2.3 Generation of CRISPR/Cas9 knock-out cell lines...... 37 2.2.4 Cell Treatment Procedure ...... 38 2.2.5 Purification of lysosomal fractions...... 38 2.2.6 activity assay ...... 38 2.2.7 Secretion Assay ...... 39 2.2.8 SDS-PAGE and western immunoblotting ...... 39 2.2.9 Internalization assay ...... 39 2.2.10 DQ™ Red BSA Assay ...... 40 2.2.11 Magic red cathepsin-B assay...... 40 2.2.12 Indirect Immunofluorescence and co-localization analysis ...... 40 2.2.13 Electron Microscopy ...... 41 2.2.14 Statistics ...... 41

2.3 Results ...... 42 2.3.1 Ultrastructural alteration of lysosomal structures and elevated autophagy upon retromer deficiency ...... 42 2.3.2 Retromer deficiency affects lysosomal activity ...... 45 2.3.3 Retromer deficiency causes defects in CI-M6PR trafficking and its downstream cathepsin-D processing ...... 48 2.3.4 A subset of CI-M6PR-containing endosome derived transport vesicles that are tethered by GCC88 depend on retromer for their generation...... 51 2.3.5 SNX3 associates with retromer to coordinate the trafficking of GCC88-tethered CI-M6PR containing ETC ...... 54

2.4 Discussion ...... 58

Chapter 3. A role of GCC88 in the retrograde transport of CI-M6PR and the maintenance of lysosomal activity ...... 68

3.1 Introduction ...... 68

3.2 Materials and methods ...... 70 3.2.1 DNA constructs, antibodies, and chemicals ...... 70 3.2.2 Cell culture and Transfection ...... 70

xiii 3.2.3 Generation of CRISPR/Cas9 knock-out GCC88 cell line ...... 71 3.2.4 Cell Treatment Procedures...... 71 3.2.5 Secretion Assay ...... 71 3.2.6 SDS-PAGE and western immunoblotting ...... 71 3.2.7 Indirect Immunofluorescence and colocalization analysis ...... 71 3.2.8 Statistics ...... 72

3.3 Results ...... 73 3.3.1 GCC88 is required for the maintenance of the TGN structure ...... 73 3.3.2 GCC88 deficiency impairs the retrograde trafficking of CI-M6PR ...... 74 3.3.3 GCC88 deficiency affects the lysosomal proteolytic capacity ...... 77 3.3.4 The autophagy-lysosomal pathway is not affected in GCC88 KO cells ...... 79

3.4 Discussion ...... 82

Chapter 4. The Parkinson disease-linked Vps35 D620N variant disrupts WASH complex association and impairs the selective function of retromer in cargo sorting ...... 85

4.1 Introduction ...... 85

4.2 Materials and methods ...... 87 4.2.1 DNA constructs ...... 87 4.2.2 Antibodies ...... 87 4.2.3 Cell culture and Transfection ...... 88 4.2.4 Cellular assay ...... 88 4.2.5 Indirect Immunofluorescence ...... 88 4.2.6 Image processing ...... 88 4.2.7 Membrane fractionation...... 89 4.2.8 SDS-PAGE and western immunoblotting ...... 89 4.2.9 Immunoprecipitation ...... 89 4.2.10 Electron Microscopy ...... 89 4.2.11 Statistics ...... 90

4.3 Results ...... 91 4.3.1 The endosomal association of retromer accessory proteins is affected in the presence of Vps35 D620N ...... 91 4.3.2 The presence of Vps35 D620N variant decreased the affinity of FAM21 with retromer and altered endosome morphology ...... 95 4.3.3 Absence of retromer-dependent CI-M6PR ETCs in Vps35 D620N rescue cells...... 98 4.3.4 The presence of Vps35 D620N rescues the lysosomal defects caused by retromer depletion .. 100 4.3.5 Unaffected autophagy pathways in the presence of Vps35 D620N ...... 101

4.4 Discussion ...... 105

Chapter 5. General Discussion ...... 111

xiv 5.1 Molecular actions of retromer in the retrograde pathway ...... 111

5.2 Retromer and the maintenance of lysosomal function ...... 113

5.3 Retromer and the Parkinson’s disease ...... 115

5.4 Future perspectives ...... 117

Chapter 6. References...... 119

xv List of Figures:

Figure 1.1 Endosomal sorting and trafficking pathways in the mammalian endocytic network...... 2

Figure 1.2. Retromer in the endosomal network...... 14

Figure 1.3. An overview of retromer’s roles in PD...... 23

Figure 2.1. Ultrastructural alteration of lysosomal structures and elevated autophagy in Vps35 KO cells...... 45

Figure 2.2. Deficiency of retromer causes reduced lysosomal activity...... 47

Figure 2.3. Vps35 is required for efficient CI-M6PR trafficking and cathepsin-D processing...... 51

Figure 2.4 ETCs containing CI-M6PR, that are tethered by GCC88, are absent in Vps35 KO cells...... 54

Figure 2.5 SNX3 is required for the retrograde transport of CI-M6PR GCC88-tethered ETCs...... 57

Figure 2.S1. Full scan of western blots in Figure 2.3A...... 62

Figure 2.S2. Additional data for Figure 4 and Figure 5...... 64

Figure 2.S3. Absence of retromer has no impact on the trafficking of CD-M6PR ETCs captured by golgin-97 or golgin-245...... 65

Figure 2.S4. The retrograde transport of CI-M6PR ETCs captured by GCC88 and golign- 245 requires the coordination of SNX3 and SNX-BAR proteins, respectively...... 66

Figure 2.S5. The retrograde transport of CD-M6PR ETCs captured by GCC88, golgin-97 or golgin-245 is not affected by the absence of SNX-BAR, SNX3 and SNX27 proteins...... 67

Figure 3.1. GCC88 is required for the maintenance of the TGN structure...... 74

Figure 3.2. GCC88 deficiency affects the retrograde transport of CI-M6PR...... 77

Figure 3.3. GCC88 is required for the maintenance of lysosomal proteolytic activity...... 78

Figure 3.4. The autophagy-lysosomal pathway is not affected by GCC88 deficiency ...... 81

Figure 3.S1. The retrograde trafficking of CD-M6PR is not altered in GCC88 KO cells. ... 84

Figure 4.1. The endosomal association of retromer accessory proteins in Vps35 D620N rescue cells...... 94

xvi Figure 4.2. The Vps35 D620N variant affects the endosomal recruitment of the WASH complex...... 97

Figure 4.3. The Vps35 D620N variant perturbs the retrograde trafficking of CI-M6PR by affecting the formation of CI-M6PR ETCs that are tethered by GCC88...... 100

Figure 4.4. The expression of Vps35 D620N variant rescues the lysosomal proteolytic defects caused by the absence of retromer...... 103

Figure 4.5. The expression of Vps35 D620N variant rescues the altered autophagy flux caused by the absence of retromer...... 104

Figure 4.S1. Supplementary data for Figure 4.1...... 109

Figure 4.S2. Supplementary data for Figure 4.3...... 109

Figure 4.S3. Supplementary data for Figure 4.5...... 110

Figure 5.1. Evidence for two independent types of ETC responsible for the endosome-to- TGN transport of CI-M6PR in mammalian cells...... 114

List of Tables:

Table 1.1. TGN golgins and associated binding partners...... 10

Table 1.2 Summary of retromer variants linked to Parkinson's disease ...... 31

xvii List of Abbreviations AA AD Alzheimer’s disease AMPAR α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor AP-1 Adaptor protein complex 1 APEX Ascorbate peroxidase APP β-amyloid precursor protein ATG9A Autophagy-associated protein Aβ β-amyloid peptide BACE-1 β-secretase BAR Bin-Amphiphysin-Rvs BCA Bicinchoninic acid BDNF -derived neurotrophic factor CD-M6PR Cation-dependent MPR CI-M6PR Cation-independent MPR CLN5 Ceroid lipofuscinosis neuronal protein-5 CMA Chaperone-mediated autophagy COG Conserved oligometric Golgi Crb Crumb DA DMTI-II Divalent metal transporter II DRD1 Dopamine receptor D1 Drp1 -related protein 1 ECL Enhanced chemiluminescent EE Early endosome EHD1 Eps15 homology domain-containing protein-1 ER ESCRT Endosomal sorting complex required for transport ETCs Endosome transport carriers FACS Fluorescence activated cell sorting FANSHY Phe-ala-asn-ser-his-tyr FREM 4.1/ezrin/radixin/moesin GARP/VFT Golgi associated retrograde transport/Vps fifty three GEF Guanine nucleotide exchange factor GLUT1 Glucose transporter 1 gRNA guide RNA

xviii HOPS Homotypic fusion and protein sorting KI Knock-in Kir3 G protein gated inward rectifying potassium channels KO Knock-out Lamp2a Lysosome-associated membrane glycoprotein 2a Mitochondria-targeting transmembrane domain of monoamine MAO oxidase MAO Monoamine oxidase MAPL Mitochondria-anchored protein ligase MDVs Mitochondria-derived vesicles Mfns Mitofusins MPP+ Mitochondria toxicant 1-methyl-4-phenypyridinium MPRs Mannose 6-phosphate receptors mTORC1 Mammalian target of rapamycin complex 1 MUL1 Mitochondria E3 ligase-1 MVB Multivesicular body NSF N-ethylmaleimide-sensitive factor PACS1 Phosphofurin acidic cluster sorting protein 1 PD Parkinson’s disease PDZ PSD-95/discs large/zona occludens PI(3,5)P2 -3,5-bisphosphate PI(3)K P150/hvps34 phosphatidylinositol 3-kinase PI(3)P Phosphoinositide ptdins(3)P PI(4,5)P2 Phosphatidylinositol 4,5-bisphosphate PIPs Phosphatidylinositol PtdIns(3)P Phosphatidylinositol 3-monophosphate PTHR G protein-coupled parathyroid hormone receptor PX Phox homology RE Recycling endosome RME-8 Receptor-mediated 8 ROI Regions of interest SNAPs Soluble NSF-attachment proteins SNX SORLA Sorting-related receptor with A-type repeats t-SNARE Target SNARE TBC1D23 TBC1 domain family member 23 TBC1D5 TBC1 domain family member 5

xix TfR Transferrin receptor TGN Trans-Golgi network TRAP-II Transport particle-II TrkB Tropomyosin-related kinase B v-SNARE Vesicle SNARE Vtl1a Vesicle transport through interaction with t-snares homolog 1A W/F-L-M/V Trp/phe-leu-met/val WASH WASP and SCAR homologue WGA Wheat germ agglutinin Wls Wntless WT Wild type YLL Tyr-leu-leu β2AR β2 adrenergic receptor

xx Chapter 1. Literature review

1.1 Overview of the endosomal system

The cellular internalization of extracellular biomolecules is of fundamental importance for the maintenance of cell homeostasis, signal transduction, nutrient ingestion and several biological processes. Typically, specific macromolecules, in the form of ligands, bind to receptors concentrated in specialized cell surface regions and trigger the process called endocytosis. During endocytosis, extracellular materials including extracellular fluid, , nutrients, and growth factors, as well as transmembrane proteins such as signal receptors, iron channel proteins, and transporter proteins, are taken up into transport vesicles, which then carry those proteins or cargoes to the sub-compartments of the endosomal system for specific functions.

The endosomal network is a dynamic system which allows for the trafficking of cargoes between distinct membrane-bound compartments and plays a central role in cargo sorting, reutilisation and degradation, thus regulates a series of fundamental cellular processes. Endosomal compartments generally comprise the early endosome (EE), the recycling endosome (RE), the late endosome/multivesicular body (MVB) and the lysosome according to their unique identities. Internalized cargoes are initially delivered to the early endosome, which is defined as a general sorting station. From the early endosome, cargoes are destined for three different intracellular destinations (Figure 1.1). Cargoes can be recycled back to the plasma membrane through the recycling endosome, or routed to the late endosome and lysosome for degradation, or sorted into the retrograde pathway to the trans- Golgi network (TGN).

The regulation of multi-directional trafficking pathways is achieved through spatial and temporal control of endosomal identity. Although each endosomal compartment has its unique identity, such as acidification, phosphatidylinositol phospholipids (PIPs) composition and other molecule markers, the identities are fluid and replaceable, due to the dynamic interchange of molecular cargoes. The change of luminal pH, the alterations in PIPs, and the recruitment and activation of different GTPases are all linked to the dynamic remodelling of the endosomal system and the direction of trafficking flow.

1

Figure 1.1 Endosomal sorting and trafficking pathways in the mammalian endocytic network.

The endosomal network is composed of multiple membrane-bounded compartments, including the early endosome (EE), the recycling endosome (RE), the late endosome/multivesicular body (MVB) and the lysosome. The EE acts as the major sorting station in the endocytic pathway. Cargoes internalized are initially targeted to the EE, where cargoes are sorted and transported towards different destinations. Cargoes can be recycled back to the plasma membrane directly or through the RE or transported to the trans-Golgi network (TGN) via endosomal transport vesicles. Besides, a selected subset of cargo molecules are ferried to the late endosome and lysosome along the endosomal maturation process and degraded at the lysosomal compartment.

1.1.1 Endocytic recycling pathway

Endocytic recycling plays a crucial role in the maintenance of membrane compositions and contributes to various cellular processes due to the diversity of recycling cargoes. Upon internalized into cells, receptors/transporters are delivered to the early endosome, then

2 sorted into at least two recycling pathways - the ‘fast’ recycling pathway, and the ‘slow’ recycling pathway. In the former one, cargoes are recycled from the early endosome to cell surface within a few minutes. The rapid and efficient recycling of cargoes is achieved through the extensive tubulation of early endosome membranes. Cargoes are sorted into newly formed tubules and transported back to the plasma membrane. In this process, small GTP-binding proteins are required. Studies showed that Rab4 functions in an early stage of endocytic sorting and is responsible for the rapid recycling of transferrin receptor (TfR) [1, 2] and glycosphingolipids [3]. Besides, another Rab GTPase, Rab35 is described as a vital element for the rapid endocytic recycling of the septin SEPT2, which is required for the postfurrowing steps of cytokinesis [4].

In contrast to the rapid recycling route, cargoes sorted into the ’slow’ recycling pathway are transported from the early endosome to the plasma membrane via the recycling endosome. The recycling endosome is morphologically defined as a large tubular structure devoid of fluid and is typically localized near the microtubule organizing centre and the TGN. The major molecular marker of the recycling endosome is Rab11. During the endosome maturation, the early endosome elaborates tubules to form the recycling endosome, and simultaneously GTP-bound Rab11 is recruited from the to the membrane of the newly-formed recycling endosome [5]. Rab11 and its effectors are incorporated into several vital processes of the ‘slow’ endocytic recycling mechanism, such as the positioning of recycling compartments, the cargo trafficking from the early endosome to the recycling endosome, and that from the recycling endosome to the cell surface. Depletion of Rab11 gives a rise to the peripheral accumulation of TfR and inhibits the tethering and fusion of recycling vesicles to the plasma membrane [6]. In addition, Rab22a GTPase also participates in the regulation of cargo movement from early to recycling compartments. Studies showed that depletion of Rab22a disrupts the structure of perinuclear recycling compartments, thus impairing the ‘slow’ recycling of TfR [7].

1.1.2 The endo-lysosomal degradation system

Digestion of extracellular proteins taken up by endocytosis is an important aspect of cell regulation. In the endosomal system, endocytic degradation is accompanied by the gradual maturation of endosomes and the endo-lysosomal fusion process.

3 1.1.2.1 Endosome maturation

After the endocytic entry, vesicles carry cargoes to the endosomal system under the direction of Rab5, a major regulator for membrane transport to the early endosome. In the active GTP-bound state, Rab5 mediates the endosomal recruitment of proteins including Rabaptin-5/Rabex-5 and EEA1 for endosomal fusion and cargo sorting at the early endosome [8-10]. Rab5 is also responsible for the recruitment of the p150/hVps34 phosphatidylinositol 3-kinase (PI(3)K), which is required for the transport activity of early endosomes, thereby regulates the production of the phosphoinositide PtdIns(3)P (PI(3)P) at the endosome membrane [11]. The morphology, localization and other features of early endosomes are heterogeneous. However, most of them are localized to the peripheral with a slightly acidic luminal pH (~ 6.5), which allows for the dissociation of receptor cargoes (ligands). Upon delivered to the early endosome, selected cargoes are sorted by various sorting mechanisms into the degradation pathway. For example, ubiquitinated proteins are recognized and sorted into the degradation pathway by the endosomal sorting complex required for transport (ESCRT) in a signal-dependent manner [12-14]. Cargoes destined for degradation are then transformed into the endosomal invaginations through the deformation of the endosomal membrane. During this process, the number of intraluminal vesicles increases, which are highly related to the biogenesis of MVB - the multivesicular appearance of the late endosome.

The endosome maturation is accompanied by the conversion of Rab GTPases, a process in which the loss of Rab5 from the endosome membrane occurs with the acquisition of Rab7. Live imaging studies recorded the dynamic transition in details, demonstrating that cargoes destined for degradation are enriched in progressively fewer and larger endosomes, which migrate rapidly from the peripheral cytoplasm to the cell centre via the loss of Rab5 and the recruitment of Rab7 [15]. This conversion requires the class C VPS/HOPS complex, a guanine nucleotide exchange factor (GEF) of Rab7 [15]. Studies showed that Vps39 subunit of the class C VPS/HOPS complex in has GEF activity for Ypt7p, the ortholog of human Rab7, thus regulating the fusion between vesicles and [16]. Consistently, silencing of Vps39 in mammalian cells causes enlarged endosomes and accumulated internalized cargoes within Rab5-positive vacuoles, which eventually acquire Rab7, but with a remarkable delay, demonstrating the functional role of the class C VPS/HOPS complex in Rab7 recruitment [15].

Additionally, during the endosomal maturation, the conversion of PI(3)P to its derivate, phosphatidylinositol-3,5-bisphosphate (PI(3,5)P2) occurs. The generation of PI(3,5)P2

4 is achieved through the hydrolyzation of PI(3)P by the phosphatidylinositol 3-phosphate 5- kinase FAB1/PIKfyve [17], the activity of which is critical for the cargo degradation. Inhibition of PIKfyve causes delayed degradation of epidermal growth factor receptor (EGFR) and other similar defects, such as the trafficking barriers after sorting ligands into the MVB and the improper acidification of [18, 19]. Functional roles of PI(3,5)P2 includes the regulation in the formation of large vacuoles, the activation of ion channels, and the monitoring of vacuole acidification [20]. In the degradation pathways, PI(3,5)P2 exhibits a functional role in partial deformation of endosome membrane during MVB formation through the interaction with its effector Vps24, an ESCRT protein [21].

1.1.2.2 The endo-lysosomal fusion

The last step for cargo degradation is achieved by the fusion of MVB/late endosomes with lysosomal compartments. Like many other fusion events in endocytic trafficking pathways, late endosomes fuse with lysosomes in a SNARE-dependent process, in which N- ethylmaleimide-sensitive factor (NSF), soluble NSF-attachment proteins (SNAPs), Rab7 GTPase and its effectors are required. Generally, the endosomal fusion comprises three steps: tethering, trans-SNARE (SNAP receptor) complex formation, and membrane fusion. As content mixing is only observed when are in physical contact [22], tethering of two organelles is defined as the initial step to trigger the following fusion process. Although the composition of proteins involved in endosome tethering is not well-established currently, evidence indicates that the homotypic fusion and vacuole protein sorting (HOPS) complex can be a good candidate due to its interaction with Rab7 and -7 [16, 23]. Studies revealed that mammalian HOPS subunit Vps18 and Vps39 are required for the clustering of late endosomes and lysosomes [24, 25]. In addition, Rab7 and its effectors are essential for late endosome clustering. Rab7 dominant negative mutants (Rab7T22N and Rab7N125I) lead to the dispersion of late endocytic structures [26]. Following tethering, the trans-SNARE complex needs to be formed. In the complex, a helix from one SNARE wraps around the similar helix on the other three SNARE proteins, thus forming a parallel bundle composed of totally four helices, which is required for membrane fusion. In the centre of the bundle, an ionic layer has an arginine (R) residue and three glutamine (Q) residues, each of which is supported by different SNARE. Accordingly, four SNAREs are termed R-SNARE and Qa-, Qb-, Qc- SNAREs. Studies in cell-free systems demonstrated that the same Qa-, Qb- and Qc- SNAREs, which are syntaxin-7/syntax-8, vti1b, and VAMP8 respectively, are associated with both homotypic and heterotypic fusion events of late endosomes [27]. Upon the formation of trans-SNARE complex, membranes

5 of late endosomes and lysosomes are ready to fuse. Although trans-SNARE complexes can trigger liposomes fusion, bilayer fusion events between late endosomes and lysosomes may not only rely on this complex. Evidence from cell-free experiments demonstrated that other elements such as Ca2+ and calmodulin also participate in the endo- lysosomal fusion process [28]. The chelation of luminal Ca2+ remarkedly reduces the fusion events [28]. Following the successful fusion between MVB/late endosomes and lysosomes, proteins for degradation are transferred to lysosomes, within which proteases and other digestive finally degrade cargoes.

1.1.3 Endosome-to-TGN retrograde transport

In addition to sorting internalized cargoes into endocytic recycling or degradation pathway, the endosomal system also plays critical roles in sorting cargo molecules into the retrograde pathway, through which cargoes are transported to biosynthetic compartments such as the TGN.

1.1.3.1 Retrograde cargoes

The number of retrograde cargoes is expanding rapidly. Due to the functional diversity of retrograde cargoes, retrograde trafficking has been implicated in multiple cellular processes such as the exchange of membrane components, membrane fusion, and transport of selected proteins to the endoplasmic reticulum (ER) membrane.

Well-studied retrograde cargoes in include mannose 6-phosphate receptors (MPRs) that mediate signal exchanges between the endosomal network and the biosynthetic/secretory system. Currently, two MPRs are identified in human, which are known as cation-independent MPR (CI-M6PR) and cation-dependent MPR (CD-M6PR). Although two MRPs differ in structure, both of them are type-1 integral membrane glycoproteins and have functional roles in the sorting of acid hydrolases that are tagged with mannose-6-phosphate (M6P). Through the M6P-binding site on N-linked oligosaccharides, MPRs recognize and transport associated newly-synthesised hydrolases from the TGN to the endosome. Once delivered to endosomal compartments, hydrolases dissociate from the MPRs due to the acid pH, allowing unoccupied MPRs to recycle back to the TGN via retrograde trafficking pathways for the next round of sorting. In addition to MPRs, other receptor families with functional roles in the endosome-to-TGN trafficking pathway include multiligand type-I receptor family (sortilin, SorLA, and SorCSs), which has homology to yeast Vps10 [29]. One of the Vps10 receptor family - sortilin is described as a mediator in the

6 endosomal transport of sphingolipid activator proteins [30]. Consequently, retrograde transport of sortilin supports another round of sorting of sphingolipid activator proteins.

Moreover, the retrograde transport of some additional receptors could be related to the integrity of regulatory circuits. One example is the Wntless (Wls) receptor — a transmembrane transporter fundamental for the secretion of Wnt proteins, which are conserved acylated and glycosylated secreted proteins with established roles in embryonic development and tissue homeostasis [31, 32]. Wls transports Wnt proteins from the ER through the Golgi towards the plasma membrane for secretion. After the Wls-Wnt complex reaching the plasma membrane, Wls is dissociated from the protein complex and retrieved to the Golgi via retrograde trafficking pathways for the next round of Wnt secretion [33-38]. Defect on Wls trafficking impairs Wnt secretion, which may further affect the downstream Wnt signalling [33-36, 38]. Despite receptors, other cargoes including SNAREs and transmembrane peptidases such as furin and carboxypeptidase D also rely on the retrograde mechanisms for fully regulatory function [39-41].

1.1.3.2 Machineries for retrograde transport

Cargo selection and the formation of endosome transport carriers (ETCs) depend on specific endosome membrane subdomains. This process is regulated by multiple mechanisms, which can be classified into two types — the -dependent mechanism and the clathrin- independent mechanism. The retrograde transport of certain cargoes relies on the coating protein clathrin and its adaptors, which recognize sorting motifs within cargo proteins and package them into transport vesicles. Studies identified that clathrin adaptor phosphofurin acidic cluster sorting protein 1 (PACS1) binds directly to the cytosolic domain of MPRs, furin, and the sorting-related receptor with A-type repeats (SORLA), thereby mediating the endosome-to-TGN retrieval of cargoes [42-45]. Another clathrin adaptor — the adaptor protein complex 1 (AP-1) also participates into the retrograde transport of cargoes [46-49]. Deficits of the μ1A subunit of AP-1 disrupt MPRs retrieval and causes the accumulation of MPRs at early endosomes [48]. Besides, EpsinR adaptor is also required for MPRs retrieval [50]. Moreover, a phosphatidylinositol 4,5-bisphosphate (PI(4,5)P2) 5-phosphatase OCRL1 acts as a regulator in retrograde routes via the interaction with clathrin [51, 52]. Either overexpressing or silencing OCRL1 leads to mistrafficking of MPRs [51].

In addition to the clathrin-dependent mechanism, retrograde transport is also regulated by the mechanism independent of clathrin. Previous studies reported the tail-interacting protein of 47 kDa (TIP47) as a cargo-selection device, which recognizes the cytoplasmic domain of

7 MPRs and sorts cargoes into retrograde ETCs [53, 54]. The activated Rab9 GTPase recruits TIP47 to late endosomes and enhanced the binding affinity of TIP47 with MPRs [55-57]. Another critical clathrin-independent retrograde transport pathway is mediated by the retromer complex. The mammalian retromer comprises a (Vps) trimer and a sorting nexin (SNX) dimer. The Vps trimer participates in cargo selection/recognition, while the SNX dimer drives the formation of tubules. Retromer is critical for the integrity of cellular trafficking circuits. Dysfunction of retromer causes transport disorders of associated cargoes, thus contributes to the progression of various neurodegenerative diseases, such as Alzheimer’s disease (AD) and Parkinson’s disease (PD) [58]. Detailed machinery on retromer function will be reviewed in the following section.

Once retrograde transport carriers are formed, tethering factors capture and direct the transport vesicles to the TGN. One classic tethering factor is the Golgin coiled-coil protein family termed “golgins”. Depending on the localization at the Golgi, golgins are classified into three types: cis-golgins (GM130, p115, and GMAP-210), trans-golgins (GCC88, golgin- 97, golgin-245, and GCC185) and other golgins located in Golgi rims [59]. Trans-golgin proteins are responsible for the tethering and transport of ETCs sorted from endosomes to the TGN [60, 61]. Trans-golgins are targeted to the TGN via the conserved GRIP domain at the C-terminus [62]. Recruitment of golgin-97, golgin-245, and GCC88 is mediated by the TGN-anchored GTPase Arl1, which binds with the GRIP domain. GCC185 binds poorly to Arl1, thus its recruitment could be coordinated by several interacting partners [33, 63, 64]. By replacing the C-terminal GRIP domain with a mitochondria-targeting transmembrane domain of monoamine oxidase (MAO), modified trans-golgins can be relocated to the mitochondria outer membrane. Along with the relocation of trans-golgins, associated retrograde cargoes including CI-M6PR, CD-M6PR, TGN46, and the vesicle transport through interaction with t-SNAREs homolog 1A (Vtl1a) SNARE are transported to the mitochondria rather than the TGN [61]. Relocated GCC185 is not sufficient to redirect cargoes to the mitochondria, suggesting distinct machinery with the other three trans-golgins [61].

Specific trans-golgins capture specific classes of transport carriers. In fact, recent studies showed that although trans-golgins capture incoming transport vesicles via the tethering motif at the N-terminus, golgin-97 and golgin-245 share a closely related tethering motif, which is distinct from that in GCC88 [60]. The capture of transport vesicles by golgin-97 or golgin-245 relies on the regulation of TBC1 domain family member 23 (TBC1D23) — a member of the TBC family of Rab6 GAPs. TBC1D23 serves as a bridging factor by

8 interacting with the FAM21 subunit of the WASH complex and the tethering motif within trans-golgins [65]. The detailed tethering machinery for GCC88 is unclear yet. In consideration of the transport carrier formation process, coating adaptors or Rab GTPases could be the point for dissecting the tethering machinery in the future study. Evidence showed that GCC185 recognizes and captures AP-1 containing vesicles by conferring binding sites for AP-1 and Rab9 [66, 67]. In addition, FIPI/RCP, a Rab11 effector, binds to golgin-97 at a region just upstream of the Rab11 binding site and is required for the retrograde transport of TGN38 [68]. These observations suggest the importance of potential binding sites mapping on trans-golgins. By using the system, numerous GTPase binding sites within trans-golgins are identified (Table 1.1), offering clues for future study on the tethering mechanisms [33, 69].

Apart from TGN-targeting golgins, multisubunit tethering complexes also contribute to the capture of specific transport carriers arriving at the Golgi. Three types of multisubunit protein complexes including conserved oligometric Golgi (COG), transport particle-II (TRAP-II) and Golgi associated retrograde transport/Vps fifty three (GARP/VFT) have been implicated in the tethering of endosome-derived transport vesicles delivered to the Golgi. The COG and TRAP-II complexes function in the intra-Golgi trafficking pathway, whereas the GARP/VFT complex functions in the endosome-to-TGN pathway. The GARP complex is composed of four subunits, Vps51, Vps52, Vps53, and Vps54, and localizes at the TGN. The GARP complex is required for MPRs sorting. Depletion of GARP impairs MPR trafficking and the lysosomal targeting of MPR-dependent acid hydrolase — [70]. In addition, GARP is also required for the retrograde trafficking of other cargoes, such as TGN46 and SNAREs [71].

9 Trans-golgins Binding partners Species Reference(s)

GCC88 Arl1 Human [72, 73] Rab2 Drosophila [69] Rab6 Drosophila [74] Rab19 Drosophila [74] Rab30 Drosophila [69, 74] golgin-97 Arl1 Human [72, 75] Rab6 Human [75] Rab6 Drosophila [69, 74] Rab19 Drosophila [69, 74] Rab30 Drosophila [69, 74] FIP1/RCP Human [68] golgin-245 Arl1 Human [72, 75] Arl3 Human [76] Rab6 Human [75] Rab2 Drosophila [69, 74] Rab30 Drosophila [69, 74] MACF1 Human [77] GCC185 Arl1 Human [72, 73] Arl4A Human [78] Rab1A Human [67] Rab1B Human [67] Rab2A Human [67] Rab2B Human [67] Rab6A Human [67] Rab6B Human [67] Rab9A Human [67, 79] Rab9B Human [67] Rab15 Human [67] Rab27B Human [67] Rab30 Human [67] Rab33B Human [67] Rab35 Human [67] Rab36 Human [67] Rab2 Drosophila [69, 74] Rab6 Drosophila [69] AP-1 Human [66] CLASP Human [66, 80, 81] STX16 Human [82]

Table 1.1. TGN golgins and associated binding partners.

10

After tethering factors transport cargo-containing vesicles close to the TGN membrane, vesicle docking and fusion — the final step in retrograde transport is triggered. Similar to nearly all fusion events, the fusion between vesicle and Golgi is achieved through the SNARE four-helix bundle. Briefly, two types of SNARE including the target SNARE (t- SNARE) and the vesicle SNARE (v-SNARE) are required in this process. Through direct interaction, t-SNAREs and v-SNAREs form the SNARE complex, thereby triggering sequential fusion events. The composition of SNARE complexes varies depending on cargoes. In human, two SNAREs complexes have been implicated in the retrograde transport of Shiga toxin from the endosome to the TGN. One comprises the TGN-localized t-SNARE proteins — syntaxin-6, syntaxin-16, and Vtila, and the early/recycling endosomal v-SNARE VAMP3/VAMP4 [83]. The other complex is composed of three t-SNAREs including syntaxin-5, GS28, and Ykt6, and one v-SNARE GS15 [84]. However, for MPRs- containing vesicles, the fusion event is mediated by another SNARE complex that consists of three t-SNAREs including syntaxin-10, syntaxin-16, and Vtila, and a v-SNARE VAMP3 [82].

1.2 The Retromer complex

Retromer is an essential peripheral complex that functions in endosomal sorting of a variety of cargo molecules. Initially identified in eukaryote , the retromer complex is assembled from two functional modules: a cargo- recognition heterotrimer composed of Vps proteins including Vps29, Vps35, and Vps26, and a membrane-associated dimer of Phox homology (PX) domain-containing proteins Vps5 and Vps17. Components of retromer are relatively conserved in higher eukaryotes. Likewise, mammalian retromer consists the high-affinity Vps29-Vps35-Vps26 trimmer and the SNX dimer of a mammalian ortholog of Vps5, namely SNX1 or SNX2, and a Vps17 ortholog, SNX5 or SNX6 [85, 86]. Structural studies have identified that, of the cargo-recognition heterotrimer, the core Vps35 subunit exhibits binding sites for Vps26 and Vps29, thereby promoting the formation and stabilization of retromer complex [87-89]. Mammalian retromer is believed to serve identical functions with that in the yeast system. There are, however, several differences existing between mammalian retromer and its yeast equivalent. In particular, the Vps26 subunit in mammals is defined to have two paralogs, Vps26A and Vps26B, and thus mammalian retromer has two distinct forms of retromer that may be involved in different functions [90-92]. Another important distinction is the functional link

11 between the Vps trimer and SNX dimer (SNX1/2 and SNX5/6) appears to be more flexible in mammals compared with that in yeast. A number of studies in mammals have reported multiple cargo trafficking processes relying on individual retromer sub-modules, as defined by proteins associated with the Vps29-Vps35-Vps26 retromer heterotrimer [93-95]. Mammalian retromer also requires various accessory proteins, such as Eps15 homology domain-containing protein-1 (EHD1), TBC1 domain family member 5 (TBC1D5) and the WASH complex, most of which are not conserved in yeast [96-99]. Hence, the retromer- mediated sorting pathways are more complicated in mammals than in yeast.

Retromer regulates the retrieval of various cargoes from endosomes by its spatial and temporal interaction with a range of associated proteins (Figure 1.2). One well-studied cargo incorporated into retromer-mediated retrograde pathway is CI-M6PR, a type-I transmembrane transporter that shuffles between the TGN and endosomes for the delivery of newly synthesized hydrolyses to lysosomes [100, 101]. Hydrolase precursors are marked for recognition by CI-M6PR via a two-step post-translational modification while they traverse the Golgi. First step in cis Golgi is carried out by the N-acetyglucosamine-1- phosphotransferase (GlcNAc-1-phosphotransferase) which catalyzes the transfer of GluNAc-1-phosphate to the C-6 hydroxyl group of mannose residues to form a phosphodiester. The second step occurs in the TGN where the GlcNAc-1 phosphodiester α-N-acetylglucosaminidase (uncovering enzyme) removes the GlcNAc residue to expose the mannose 6-phosphate [102-105]. The mannose 6-phosphate binding sites mapped on domain 3, 5, 9, and 15 of CI-M6PR extracellular 15-domain region enable the receptor to recognize a wide range of hydrolase precursors with phosphorylated N-glycans [106, 107]. Upon delivered to endosomes, ligands are dissociated from CI-M6PR due to the acidic pH, and unoccupied CI-M6PR returns to the TGN along retrograde pathways for another round of sorting. In the absence of retromer, the trafficking fate of CI-M6PR is altered through inefficient retrograde transport from endosomes to the TGN, resulting in an increased lysosomal turnover of CI-M6PR and a decreased level of intracellular hydrolyses [100]. In addition to CI-M6PR, retromer also regulates the retrograde trafficking of other Vps10 domain-containing protein family members, such as sortilin, SorLA, and SorCSs, which are particularly important in the modulation of neurotrophic signalling pathways [101, 108, 109]. Moreover, retromer also contributes to the fine-tuning of additional cellular functions such as iron homeostasis, cell polarity, and development, through the regulation of the endosome-to-TGN shipping of cargo molecules including divalent metal transporter II (DMTI-II) — an iron transporter, crumb (Crb) — an apical polarity regulator, and Wntless — the mediator for Wnt secretion [36, 110, 111]. While predominately described as a conductor

12 in endosome-to-TGN retrieval, retromer has been recently demonstrated to coordinate the endosomal recycling of multiple transmembrane proteins to the plasma membrane, in conjunction with SNX27, which is a multi-domain scaffolding protein consisting a PX domain, PSD-95/discs large/zona occludens (PDZ) domain, and 4.1/ezrin/radixin/moesin (FERM)- like domain. Both the PX domain and FERM-like domain possess binding affinity for PtdInsP species, thereby promoting the endosomal targeting of SNX27 [112]. The PDZ domain, however, directly binds to the retromer subunit Vps26, and further directs the intracellular recycling of PDZ domain-containing proteins [113]. Transmembrane cargoes including the glucose transporter 1 (GLUT1), β2 adrenergic receptor (β2AR), and α-amino-3-hydroxy-5- methyl-4-isoxazolepropionic acid receptor (AMPAR) have been shown to undergo retromer- SNX27 mediated recycling [113, 114]. Overall, the retromer complex can act as a major sorting platform coordinating the trafficking fate of cargoes, thereby modulating multiple pathways to maintain cellular homeostasis.

1.2.1 Mechanisms for the endosomal recruitment of retromer

Recruiting retromer to the cargo-enriched endosome membrane is the first step to initiate associated sorting events. No known phospholipid or binding affinity has been described within retromer, so its endosomal targeting is therefore achieved indirectly by protein-protein interactions. Typically, retromer is recruited to endosomes by interaction with members of the PX peripheral membrane protein family which contain high-affinity phosphatidylinositol 3-monophosphate (PtdIns(3)P) binding domains [115]. SNX-BAR family members, which are characterized by the presence of the membrane-associated PX domain and the membrane curvature sensing Bin-Amphiphysin-Rvs (BAR) domain [116]. Based on studies in yeast it was originally proposed that mammalian retromer would assemble with SNX-BAR proteins on endosomal membranes. However, studies based on yeast two-hybrid analysis show a weak interaction between the retromer complex and SNX1/2 [117-119]. This evidence appears to be disputable, as biochemical experiments fail to support this association within cells [95, 117]. Therefore, the SNX-BAR subcomplex itself may not be sufficient enough to recruit retromer to endosomes.

SNX3 has been identified to recruit and stabilize retromer on endosomal membranes by directly binding to retromer subunits Vps26 and Vps35 in multiple interfaces via its N- terminal tail and the PX domain [89, 120]. In fact, loss of SNX3 results in a marked dissociation of retromer from endosomal membranes, which further supports its roles in the recruiting process [89]. Another PX domain-containing protein, SNX12, which shares high

13 homology with SNX3, is thought to have a similar role to SNX3 [121]. Likewise, SNX12 associates with the retromer Vps26 subunit, and its downregulation affects the endosomal localization of retromer [122]. Although two PX domain-containing SNXs are implicated in retromer recruitment, it remains to be determined if they function together or independently.

Figure 1.2. Retromer in the endosomal network.

Cargoes endocytosed into the endosomal system are initially targeted to the early endosome. The core cargo-selective retromer complex (Vps29-Vps35-Vps26) is recruited to the cargo-enriched endosomal membrane via mechanisms mediated by Rab7 and SNX3. Once within the endosomal system, retromer recognizes and retrieves cargoes from the degradation fate for the delivery to the cell surface or the TGN. The WASH complex and SNX27 promotes the recycling of several retromer associated receptors, including GLUT1, AMPAR and β2AR, from endosomes to the cell surface. SNX-BAR proteins and SNX3 mediate the retromer-associated endosome-to-TGN transport of receptors. One classical cargo involved in retromer-mediated retrograde pathways is CI-M6PR, which is responsible for the delivery of lysosomal hydrolases. CI-M6PR binds certain classes of newly synthesized hydrolases at the TGN and carries them to the endosomal system. Hydrolases release from CI-M6PR and are transported to lysosomes where they are activated. Unoccupied CI- M6PR is then retrieved to the TGN through retromer-mediated retrograde pathways for the next round of sorting. Arrows indicate the direction of intracellular trafficking.

Rab GTPases are localized at the surface of intracellular membrane-bounded compartments and function in multiple trafficking events. Hence, Rab proteins specifically associated with endosomes may have functional roles in the endosomal targeting of

14 retromer. Indeed, recent studies demonstrated that Rab7 is engaged into the recruiting event through directly interacting with the retromer Vps35 subunit in a guanine nucleotide- dependent manner [98, 123, 124]. Suppression of Rab7 causes a total shift of the retromer complex from the endosome membrane to the cytosolic fraction [98, 123]. In fact, multiple coordinators are involved in the regulation of this Rab7-mediated recruitment. For example, the ceroid lipofuscinosis neuronal protein-5 (CLN5), a lysosome-anchored protein, has been shown to mediate the activation and endosomal localization of Rab7, thus sequentially promoting retromer recruitment [125]. In addition, recent work has demonstrated that the palmitoylation on Rab7 is essential for the tight association between Rab7 and the core retromer complex, although this modification is not required for the endosomal localization of Rab7 [126]. While the endosomal recruitment of retromer appears to be mediated by distinct mechanisms as described above, it is hypothesized that SNX3 and Rab7 may cooperate to recruit retromer, due to the close proximity of binding sites within Vps35 and the similar defects on retromer recruitment derived by SNX3 or Rab7 deficiency [120].

Recent studies in yeast showed that retromer segregates onto the forming endosome derived tubular carriers together with cargo, whereas Rab7 remains on the endosome vacuole [127]. Only after this fission process, retromer is released into the cytoplasm [127]. The displacement of Rab7 from endosomal tubules is due to a consequence of the assembly between the Vps trimer and SNX-BAR subcomplex during tubule formation [128]. These observations provide insights into the coordinative interplay among retromer, SNX-BAR subcomplex and Rab GTPase in the temporally orchestrated membrane dissociation process. An additional coordinator in this releasing process is TBC1D5, a RabGAP for Rab7, serving as a restraint on retromer function [98]. With binding affinity to LC3, the autophagy protein, as well as the core retromer subcomplex, TBC1D5 is proposed to shuttle between LC3-positive autophagosomes and endosomal membrane, thereby dynamically directing retromer disassociation from the endosomal membrane [98, 129-131]. Hence, the endosomal association/dissociation of retromer is a refined and dynamic process that is achieved by multiple coordinators.

1.2.2 Cargo Recognition

Serving as a trafficking hub, the early endosome contains various cargo proteins at the crossroads ready to be delivered toward different destinations. Retromer shows the capability to recognize and sort cargos within endosomes in a sequence-dependent manner. The core retromer complex has been defined as the cargo-recognition module according to

15 both biochemical and structural evidence. Vps35, the core subunit of the Vps trimer, is thought to represent the major cargo-binding subunit. Evidence demonstrated that Vps35 is able to recognize and bind to a conserved Trp/Phe-Leu-Met/Val (W/F-L-M/V) motif within CI-M6PR cytoplasmic tail [100, 101]. This sorting motif is conserved for sortilin, another Vps10-containing protein family member. The W/F-L-M/V motif is essential for the endosome-to-TGN transport of receptors, as mutation within the motif perturbs receptors’ trafficking and leads them to be degraded rapidly [101]. Besides, another sorting sequence - Tyr-Leu-Leu (YLL) - within the cytoplasmic domain of DMT1-II can also be recognized by the retromer Vps35 subunit [111, 120]. Apart from Vps35, other Vps subunits also participate in the cargo recognition process. For example, Vps26 possesses cargo recognition sites and is able to interact with cargo proteins directly. In fact, a recent study showed the location of YLL motif within DMT1-II at the interface between Vps26 and SNX3, suggesting that the interaction with both SNX3 and Vps26A is required for the cargo sorting [120]. Studies also showed that SorLa, a member of Vps10 domain-containing family, contains a Phe-Ala-Asn- Ser-His-Tyr (FANSHY) motif within its C-terminal cytoplasmic tail by which SorLa interacts with Vps26 [108]. Currently, two Vps26 paralogues (Vps26A and Vps26B) are identified in mammals. Intriguingly, within cells Vps26B-containing retromer is not able to interact with CI-M6PR, indicating that two Vps26 paralogues may not be functionally equivalent in the cargo selection process [90, 91].

The machinery regarding how retromer recognizes and sorts cargoes specifically towards the cell surface has been recently revealed by a global proteomics analysis, in which levels of transporters located on the cell surface were examined in retromer- and SNX27- suppressed cells [113]. A variety of cell surface transporters are identified to recycle in a retromer-SNX27-dependent manner [113]. For example, GLUT1 directly interacts with SNX27 via the PDZ binding motif within its C-terminus. This association is demonstrated to be essential for sorting GLUT1 onto the endosome-to-plasma membrane route [113, 132]. Another receptor that can be recognized by SNX27 is β2AR, a fundamental cellular receptor protein associated with neuronal protection [97, 113, 133]. Similar to GLUT1, recycling of β2AR is dependent on the PDZ motif presented within its cytoplasmic tail, which interacts directly with SNX27 [97, 113]. Other known SNX27 cargoes include AMPA, NDMA, G protein gated inward rectifying potassium channels (Kir3) and G protein-coupled parathyroid hormone receptor (PTHR) [113, 134]. While the list of retromer associated cargoes keeps extending, as yet, a relatively conserved rule regarding how cargoes are destined for specific destinations is not elucidated. The molecular details of retromer-associated sorting events seem too broad and complicated, thus there is still much need to be characterized.

16 1.2.3 Morphogenesis of Tubular Endosomal Transport Carriers

Following the retromer-mediated sorting process, the endosomal subdomains enriched with cargo molecules deform to generate tubulovesicular endosome transport carriers, which sequentially detach from endosomal compartments and transport cargos to destined organelles. The machinery underlying tubulation is mostly related to SNX-BAR proteins, which share a conserved BAR domain that has functions in sensing and binding to membrane curvature. SNX-BAR proteins undergo dimerization and membrane association, and continuously oligomerize on the endosome membrane surface thereby generating regular arrays that form membrane tubules. Although data from in vitro experiments indicated multiple SNX-BAR proteins are able to drive the transition from vacuolar to tubular morphology [135], how these proteins function under physiological conditions is less clear. Along with SNX-BAR proteins, microtubule-based motors have been implicated to mediate the morphogenesis of tubular carriers. Generally, the microtubule plus end-directed movement is mediated by , whereas the minus end-directed movement is achieved via motor proteins. The opposite movement of kinesins and is sufficient to pull tubules from membranes. Motor proteins likely cooperate with SNX-BAR proteins to mediate the formation of retromer-associated tubules. Two studies demonstrated that the p150Glued subunit of dynactin/dynein motor complex interacted directly with SNX6 [136, 137]. Association between p150Glued and SNX6 is required for the generation and stabilization of tubular vesicles that involved in retromer-mediated retrograde trafficking. For example, dysfunction of p150Glued impairs CI-M6PR retrieval from endosomes to TGN [136]. Besides, suppression of p150Glued also impairs SNX1-mediated tubulation process [137], indicating SNX oligomerization per se may not be sufficient enough to drive membrane tubulation. In addition, other machinery may also contribute to the biogenesis of tubular transport carriers. Recent work suggested that receptor-mediated endocytosis 8 (RME-8), a DNAJ subfamily member, is required in SNX1-mediated membrane tubulation process. Loss of RME-8 alters the dynamics of SNX1 on endosome membrane and causes an increase of highly branched retromer cargo-containing tubules [138]. Another protein factor mediating the formation of retromer-associated tubules is EHD1. EHD1 contains binding sites for retromer and is required for the stabilization of tubules especially generated from SNX1. Similar to REM-8, suppression of EHD1 destabilizes SNX1-positive membrane tubules and extensively perturbs the endosomes-to-TGN retrieval of cargoes such as CI-M6PR [99].

17 1.2.4 Fission of tubulovesicular transport carriers

Once tubular transport carries reach targeting sites, membrane fission event occurs to pinch off the vesicles from tubular structures before docking and fusion with the target membranes. Membrane fission is not a simply spontaneous process, whereas it requires various enzymes and factors to break the enclosed membrane. One well-characterized factor that participates in the fission reaction is the dynamin GTPase, which belongs to a GTPase family with diverse functions in membrane remodelling. Recent studies in yeast demonstrated that the Vps1 dynamin GTPase is required for the scission of retromer-coated tubules and eventually allows for the release of retromer into the cytosol [127, 139]. Accordingly, Vps1 localizes to endosomes decorated with SNX-BAR proteins and Mvp1, the yeast homologous to human SNX8. Mvp1 promotes the endosomal association of Vps1 through direct interaction. Deficiency of either Vps1 or Mvp1 inhibits the export of retromer- associated cargo from endosomes [139]. Hence, the cooperation between endosomal sorting and fission mechanisms is required for efficient trafficking mediated by retromer.

In mammalian cells, however, the machinery underlying membrane fission of retromer cargo-containing tubular ETCs is not clear yet. So far, no firm evidence shows whether any dynamin GTPase family member is involved in the fission step of retromer-positive tubular transport carriers in mammals. However, EHD1 may participate in the fission of endosomal tubules, due to its function in the vesiculation of tubular recycling endosomes, which shares similarity to the scission process of endosomal tubules from physiological perspective [140]. In addition, EHD1 depletion destabilizes SNX1-positive tubules, further highlighting the possibility of its function in the fission process [99]. WASH complex is also proposed to mediate the fission event, due to its interaction capacity with dynamin and Arp2/3 complex, a regulator of network [141]. Finally, a dynamin-independent fission mechanism, termed friction-drive scission, is applied to the fission step of the endosomal tubular structure, whereby SNX-BAR proteins create a friction barrier for lipid diffusion to cause the increase local tension until the fission occurs [142].

1.3 Retromer and neuronal health

Neurons are highly polarized cells that possess structurally and functionally distinct assemblies, termed axons and dendrites. Both structures are extended from the soma and are responsible for the signal transmitting throughout the neuronal system. Typically, axons are single long structures that transmit external information via releasing , whereas dendrites are relatively shorter structures that receive input from the external of

18 other axons. During the last few decades, much attention has been drawn to dendritic spines, certain classes of dendrites containing small projections, which serve as the main postsynaptic compartment receiving excitatory input and monitoring synaptic remodelling. The rapid signal exchange on dendritic spines is critically associated with the efficient recycling of synaptic receptors. As a vital endosomal sorting assembly, retromer has recently been implicated in the recycling of multiple synaptic receptors from endosomes to the discrete dendritic surfaces. Retromer expresses throughout the neurons and is enriched in endosomes near dendrites. Recent work identified that these retromer-decorated endosomes are broadly distributed along dendrites in ~2 μm axial intervals, thereby enabling the efficient internalization and insertion of receptors from the dendritic surface to these endosomes. Several receptors including β2AR, AMPARs and NMDARs are incorporated into retromer-enriched endosomes, and sequentially get sorted back to the dendritic surface in a retromer-dependent manner [143]. Studies using hippocampal neurons transfected with small hairpin RNA against Vps35 demonstrated a significant decrease in recycling efficiency of β2AR, as well as a marked reduction of excitatory postsynaptic events mediated by NMDARs and AMPARs [143, 144]. While these data may not completely explain the molecular details regarding how retromer promotes these receptors back to the dendritic surface, it is likely that retromer coordinates the sorting processes together with other accessory proteins such as SNX27 [145, 146] to regulate the homeostatic .

Retromer is required for neuronal morphogenesis and physiology. Evidence for this was supported from recent studies on retromer’s roles in neuronal maturation. In the case when Vps35 expression was suppressed, hippocampal neurons displayed abnormal morphologies described as swollen axons, reduced dendritic spines and shorter apical dendrites [147, 148]. These deficits reflect the faulty trafficking within retromer-depleted neurons. Indeed, both Vps35 and Vps26 heterozygote knock-out mice had the elevated β- amyloid peptide (Aβ), a cleavage product generated from β-amyloid precursor protein (APP) [149, 150]. The accumulation of Aβ is considered to be toxic to neurons. These findings demonstrated the fundamental role of retromer in neuronal homeostasis and may account for why the nervous system is sensitive to the altered retromer function or molecular organization. Therefore, it is conceived that retromer is closely related to the pathological processes in brain disorders. It has been implicated that retromer links directly with the pathogenesis of two types of widespread neurodegenerative diseases, including Alzheimer’s disease (AD) and the Parkinson’s disease (PD).

19 1.3.1 Retromer in Alzheimer’s disease

Alzheimer’s disease is a common neurodegenerative disorder that is characterized by the progressive loss of neurons and symptoms such as cognitive defects. Over the last decades, mounting evidence highlights the link between retromer and AD pathology. Initially, the reduced levels of retromer subunits Vps35 and Vps26 are observed in the of AD patients [151]. Retromer deficiency is subsequently implicated to impair the formation of the , the region that is responsible for memory recognition and storage [149].

One hallmark for AD is an abnormal elevation of Aβ peptide-containing plaques in brain tissues. The neurotoxic Aβ aggregates are generally produced in the case when APP is cleaved improperly. APP is a type-I transmembrane protein that transports through the secretory pathway as well as the endocytic pathway. APP is initially synthesized in the ER and transported to the TGN, and then sorted to the plasma membrane for secretion. In normal neurons, APP predominantly undergoes a non-amyloidogenic proteolytic process by α-secretase and γ-secretase at the cell surface, whereas under pathological conditions, APP is mostly reinternalized into the intracellular environment and cleaved by β-secretase (BACE-1) and γ-secretase to generate Aβ. This amyloidogenic process mainly occurs in endosomal compartments [152, 153]. Hence, deficits in endosomal trafficking may link to the production of Aβ.

Retromer can regulate APP processing via retromer cargo proteins including SorLa, sortilin and SorCS1, members of Vps10 domain-containing transmembrane protein family. Studies showed that SorLa acts as a neuronal sorting factor for APP that keeps the precursor away from the amyloidogenic processing, and loss of SorLa leads to the elevated Aβ levels [154, 155]. Another neuronal APP-sorting regulator is sortilin, which promotes the non- amyloidogenic cleaving of APP-mediated by α-secretase, hereby restraining the generation of Aβ [156]. It appears that sortilin functions in distinct neuronal compartments from SorLa. Evidence from hippocampal neurons studies showed that sortilin interacts with APP in neurites, whereas SorLa mainly colocalizes with APP in soma [156], indicating the complexity of APP metabolism in neurons. The third Vps10-containing protein involved in the modulation of APP/Aβ homeostasis is SorCS1. Through the sorting motif within its cytoplasmic tail, SorCS1 regulates the exit of APP out of endosomal compartments, thus promoting APP trafficking to the TGN and decreasing the generation of Aβ [109, 157]. Additionally, retromer is associated with the endosome-to-TGN retrograde transport of BACE-1, a key enzyme for APP cleavage to generate Aβ. Retromer deficiency results in the

20 accumulation of BACE-1 in early endosomes, thereby increasing the colocalization of BACE-1 and APP, and accelerating the amyloidogenic processing [149, 158]. In contrast, pharmacological stabilization of retromer complex can limit the pathological processing of APP [159]. Taken together, these studies demonstrate the fundamental role of retromer in Aβ generation and AD pathogenesis.

1.3.2 Retromer in Parkinson’s disease

Parkinson’s disease (PD) is the second most common neurodegenerative disease only after AD. In PD cases, patients show a variety of motor symptoms such as bradykinesia and muscle rigidity, and non-motor symptoms including cognitive impairment, depression, and other sensory deficits. These symptoms are generally related to the progressive degeneration of dopaminergic neurons in substantia nigra , a PD pathological hallmark. Dopaminergic neurons are responsible for the release of dopamine (DA), an important neurotransmitter with functions in movement, memory, behaviour, and cognition [160]. Impairment of DA neurotransmission affects the excitation of cortical neurons in the substantia nigra and the sequential signal transduction to globus pallidus, which further influence the motor loop that controls movement [161, 162]. Besides, degeneration of dopaminergic neurons also leads to functional deficits in the frontal lobe, the brain region associated with cognition and problem solving [163-165]. Thus, along with the progressive loss of dopaminergic neurons, PD symptoms begin to gradually progress in patients. Another hallmark for PD is the presence of Lewy bodies (LBs), the subcellular structures in the form of large, perinuclear deposits of aggregated proteins. The most accumulated protein detected within LBs is α-synuclein, but they may also contain other proteins or lipids [166-168]. The formation of LBs is clinically associated with the degenerative process at the presynaptic nerve terminal, and results in the neurotransmitter deficiency, hence links to the decreased neuronal function, lifespan and eventually the pathology of PD [169].

1.4 Retromer mutations in Parkinson's disease

Although most PD cases are sporadic, a subset of PD cases are inheritable and ascribed to alterations at specific loci. In 2011, studies from two independent groups demonstrated the point mutation (p. D620N) within the retromer Vps35 subunit in PD cases with familial aetiology, thus highlighting the genetic association between retromer and the late-onset familial PD [170, 171]. Although with a low frequency, this rare variant has been identified in numbers of individuals and families affected with PD worldwide (Table 1.2), demonstrating

21 its pathogenic roles in PD development [170-182]. Evidence from has also revealed an additional Vps35 variant, P316S, in familial PD cases [171]. However, the genetic evidence for the pathogenicity of this variant is inconclusive, due to the discovery of it in a control subject. Further non-synonymous variants have also been reported in Vps35 subunit (p.R524W, p. H599R, p.M607V, p. I560T, p.L774M, p.G51S, p.R32S), as well as in Vps26 (p.K93E, p.M112V, p.K297X, p.R127H and p.N308D) and Vps29 (p.N72H) (Table 1.2) [170-182]. To date, however, the molecular machinery regarding how these variants directly associate with PD progression remain unclear. One possibility is retromer variants may perturb regulatory processes mediated by other PD-linked , thereby possessing roles in disease pathogenesis. This is supported by recent studies on the functional relationship between retromer and LRRK2, which is defined as the second most common causative gene of PD. Several variants within LRRK2 have been implicated in PD- associated cellular pathologies such as defective lysosomal degradation and macroautophagy, and impaired endosome-to-TGN sorting [183-187]. Overexpression of wild type Vps35, but not Vps35 D620N variant, rescues these defects [183]. Moreover, studies using Drosophila models support the functional association between Vps35 and LRRK2, revealing the capacity of retromer to rescue neuronal deficits caused by LRRK2 variants [183, 188]. Hence, retromer functions together with LRRK2 in neuronal protection, and its mutations may play a negative mode of action in this process. Collectively, although the pathogenicity of various retromer variants remains elusive, these findings do point toward the genetic association between retromer and PD and provide vital clues to retromer function in molecular pathways that are perturbed in PD cases.

1.4.1 Functional insights into retromer deficiency and PD-linked mutations

Multiple studies have examined the function of retromer in cellular pathways that may associate with PD progression (Figure 1.3). While it remains unclear if these mechanisms are directly relevant to PD pathology, studies based on cell or animal models provide insights into PD-linked molecular pathways that may be promoted by retromer deficiency or mutations.

22

Figure 1.3. An overview of retromer’s roles in PD.

Under physiological conditions, retromer retrieves cargoes away from the degradation pathway for recycling toward the TGN or the cell surface. Retromer dysfunction has been implicated in the progression of PD. Impairment of retromer causes perturbations in the endosomal trafficking of numerous cargoes affecting many aspects of downstream cellular functions. Retromer dysfunction perturbs CI-M6PR-mediated lysosomal transport of hydrolases such as cathepsin D, which can affect the activity of lysosomes. Furthermore, retromer deficiency may affect the autophagosome formation and the chaperone-mediated autophagy (CMA) by mistrafficking proteins needed in these processes, ATG9A and LIMP2a. Defects on these trafficking events perturb protein degradation mediated by endo-lysosomal system leading to accumulation of proteins such as alpha-synuclein, the most abundant component within Lewy Bodies in PD. Retromer dysfunction can also contribute to PD pathogenesis by affecting neuronal function and mitochondria homeostasis. Retromer deficiency impairs the recycling of neuronal receptors to the cell surface, which affects neuronal signal transmission and neuronal plasticity. Moreover, impairment of retromer causes dysregulation of proteins controlling mitochondria fission/fusion processes, DLP1 and Mfns, and disrupts the formation of mitochondria-derived vesicles (MDVs), thereby affecting mitochondria dynamics and function. Arrows indicate the direction of intracellular trafficking.

23 1.4.1.1 Impaired endo-lysosomal mediated protein degradation

As initially identified, retromer is an essential regulator in endosomal sorting of cargoes towards the TGN. One classic cargo incorporated into this trafficking route is CI-M6PR, which participates in the delivery of lysosomal hydrolases such as cathepsin D to lysosomes [100]. For example, under physiology conditions, CI-M6PR binds newly synthesized cathepsin D at the TGN; both receptor and cargo are sorting into a transport vesicle, which delivers them to the endosomal network. Once within the endosomal system, cathepsin D is released from CI-M6PR and transported to lysosomes, where it becomes activated. Unoccupied CI-M6PR is retrieved along the retromer-mediated pathway to the TGN for the next round of sorting. As a major lysosomal protease, cathepsin D is demonstrated to facilitate the degradation of long-lived proteins including α-synuclein, which is characterized as the most abundant component within LBs [189, 190]. Therefore, one idea is that retromer malfunction reduces the iterative transport of CI-M6PR, and thus causes a defect in cathepsin D processing, leading to a decreased level of its activated forms in lysosomes. Such deficit would induce lower degradative activity, thus leading to the accumulation of non-degraded α-synuclein and the formation of LBs, a prominent PD neuropathological phenotype. The pathophysiological features of retromer in the proteolytic cleavage of α- synuclein are supported by evidence from human α-synuclein transgenic Drosophila models, in which RNAi-mediated silencing of Vps35 perturbs the maturation process of cathepsin D and induces accumulation of insoluble α-synuclein by preventing the endosome-to-TGN retrieval of CI-M6PR [191].

PD-causing retromer variant Vps35 D620N. This variant is assembled into the cargo- selective complex correctly, however, the presence of Vps35 D620N alters the intracellular distribution of CI-M6PR and leads to an elevation of pro-cathepsin D in various cell types including fibroblast cells isolated from PD patients carrying Vps35 D620N variant [183, 192, 193]. Despite the above perturbations, the detailed mechanism regarding how the PD-linked Vps35 D620N variant perturbs retromer function in endosomal trafficking is not completely elucidated. For example, two studies reported the presence of Vps35 D620N variant disrupts retromer’s association with FAM21, a subunit of WASH complex, and further causes a reduced endosomal association of WASH complex [192, 194]. However, another independent study suggested Vps35 D620N remains a high level of co-localization with WASH complex on endosomes and retains unchanged affinity with FAM21 [193]. Recent work reported an additional variant, Vps35 R524W, is also associated with the sorting defects of CI-M6PR. Whilst the formation of the core complex is unaffected by the presence

24 of R524W variant, the endosomal recruitment of retromer is impacted, which further leads to the mistrafficking of CI-M6PR and cathepsin D [195].

Similar trafficking defects on CI-M6PR have been reported in cell-based models expressing PD-causing retromer variant Vps35 D620N. This variant is assembled into the cargo- selective complex correctly, however, the presence of Vps35 D620N alters the intracellular distribution of CI-M6PR and leads to an elevation of pro-cathepsin D in various cell types including fibroblast cells isolated from PD patients carrying Vps35 D620N variant [183, 192, 193]. Despite the above perturbations, the detailed mechanism regarding how the PD-linked Vps35 D620N variant perturbs retromer function in endosomal trafficking is not completely elucidated. For example, two studies reported the presence of Vps35 D620N variant disrupts retromer’s association with FAM21, a subunit of WASH complex, and further causes a reduced endosomal association of WASH complex [192, 194]. However, another independent study suggested Vps35 D620N remains a high level of co-localization with WASH complex on endosomes and retains unchanged affinity with FAM21 [193]. Recent work reported an additional variant, Vps35 R524W, is also associated with the sorting defects of CI-M6PR. Whilst the formation of the core complex is unaffected by the presence of R524W variant, the endosomal recruitment of retromer is impacted, which further leads to the mistrafficking of CI-M6PR and cathepsin D [195].

1.4.1.2 Impaired trafficking of neuronal receptors

Retromer is important for the normal functioning of neurons and neuronal survival. Dysfunction of retromer in neurons has been recently implicated to induce neuropathological deficits observed in PD case. One study demonstrated suppression of Vps35 in hippocampal neurons causes the degeneration of dendritic spines [147]. Spine loss has been reported in a few studies of striatal tissues from PD patients, suggesting its association with neuronal vulnerability [196]. Although it is unclear how Vps35 regulates dynamics, the degeneration of spines may be due to the aberrant recycling of AMPARs in Vps35-deficient neurons. The retromer Vps35 subunit directly interacts with two AMPAR subunits, GluR1 and GluR2, and regulates their targeting to the dendritic surface [148]. Absence of Vps35 leads to a decreased level of dendritic surface-localized GluR1/GluR2, reduces the glutamate transmission and induces impairments of excitatory spine maturation. The spine deficit is partially restored by the overexpression of GluR2 in Vps35 deficient neurons [148]. These data suggest the functional roles of retromer in neuronal integrity and may underlie the pathogenesis of PD. The PD-linked Vps35 D620N variant is suggested to act as a Vps35 loss-of-function variant in the regulation of AMPAR recycling. Recent work 25 has implicated that Vps35 D620N alters AMPAR recycling and perturbs the synaptic transmission in mouse cortical neurons expressing Vps35 D620N and human dopaminergic neurons generated from induced pluripotent stem cells of PD patient with D620N variant [197]. The perturbations to synaptic function seem to generate chronic neuropathological effect toward neuronal integrity that may lead to neuronal damage and contribute to disorders in PD [197].

Impaired dopamine signalling has recently been demonstrated in PD cases [165]. As an essential neurotransmitter, dopamine mediates various brain physiological functions such as locomotion and behaviour through binding to its receptors and triggering downstream signalling pathways. One key receptor for dopamine is dopamine receptor D1 (DRD1), which undergoes neuronal responses such as modulating the cyclic AMP production [198]. Recent studies revealed a functional connection between retromer and DRD1 [199]. Retromer regulates the cell surface recycling of DRD1 and its downstream signalling pathway via directly binding to DRD1. The PD-linked Vps35 D620N variant does not affect retromer’s affinity with DRD1, however, it fails to rescue the DRD1 recycling or associated dopamine signalling in Vps35 knockdown neurons [199]. Impaired dopaminergic neurotransmission was also observed in the Vps35 D620N knock-in mouse model [200], suggesting the molecular action of retromer in PD pathogenesis.

Additional neuronal transports may also be involved in retromer-promoted PD pathology. Post-mortem studies of PD patients reveal a reduction of brain-derived neurotrophic factor (BDNF) in the vulnerable region substantia nigra pars compacta, suggesting a deficiency in BDNF/tropomyosin-related kinase B (TrkB) signalling [201]. Accordingly, inhibition of BDNF or its downstream receptor, TrkB, increase the sensory of dopaminergic neurons to cytotoxic injury, induce the selective loss of dopaminergic neurons and may contribute to PD progression [202]. Recent work has implicated sortilin, a retromer-associated cargo, in BDNF/TrkB signalling pathway by regulating the anterograde trafficking of Trk receptors from soma to [203]. Whilst there is, as yet, no direct evidence linking defects on neurotrophin signalling with retromer-promoted PD pathology, these data still provide us insights into potential directions to explore PD-pathogenesis in future studies.

1.4.1.3 Disrupted mitochondria dynamics and function

Mitochondria homeostasis is critical for neuronal function and survival. Aberrant mitochondria function has been proposed to contribute to the development of PD. Retromer has been suggested to be associated with mitochondria function by mediating vesicle

26 transport between mitochondria and . Vps35 and Vps26 are indicated to have a functional interaction with mitochondria-anchored protein ligase (MAPL), a small ubiquitin- like modifier E3 ligase enriched in mitochondria-derived vesicles (MDVs). Retromer deficiency leads to defects in the formation of MAPL-positive MDVs and thus causes a reduction in the delivery of MAPL to peroxisomes [204]. Retromer has been implicated to play an additional role in mitochondria homeostasis, by protecting against the mitochondria toxicant 1-methyl-4-phenypyridinium (MPP+) [205]. MPP+, the oxidised form of neurotoxin MPTP, causes oxidative stress and mitochondria impairments and has highly selective toxicity to dopaminergic neurons. Enhanced Vps35 expression protects dopaminergic neurons against MPP+ cytotoxicity; however, this neuroprotection function is partially compromised by the presence of Vps35 D620N mutation [205].

Impaired fission/fusion dynamics of mitochondria is likely associated with the progressive degeneration of DA neurons and is proposed to be important for PD pathogenesis. Recent work has provided evidence linking retromer deficiency/mutation to the impairments of mitochondria dynamics and neuronal damage. The fission/fusion processes are generally regulated by two factors - dynamin-related protein 1 (Drp1) and mitofusins (Mfns). Drp1 is required for the fission process by forming a large complex to divide mitochondria; Mfns including MFN1 and MFN2 are required for the fusion processes. Retromer has been demonstrated to regulate the removal of Drp1 complexes from mitochondria via a Vp35- Drp1 interaction and transport these complexes through MDV-dependent routes to lysosomes for degradation. The Vps35 D620N variant displays an enhanced interaction with Drp1 and promotes the turnover of mitochondria Drp1 complexes, which seem likely lead to mitochondria fragmentation and loss of neurons [206]. Another study demonstrated the participation of retromer in the regulation of mitochondria fusion processes. Retromer promotes the degradation of mitochondria E3 -1 (MUL1), a critical E3 ubiquitin ligase of MFN2, and thus suppresses MUL1-mediated MFN2 degradation [207]. In Vps35 deficient dopaminergic neurons, MUL1 levels are elevated, which results in MFN2 reduction, mitochondria fragmentation, and neuronal death. Notably, these phenotypes can be restored by the expression of wild-type Vps35, but not Vps35 D620N variant, which is consistent with the mitochondria deficits in PD [207]. Whilst most molecular details regarding how retromer regulates the mitochondria homeostasis are unknown, these findings suggest an additional layer of the intersection between retromer deregulation and PD pathology, beyond its role in endosomal trafficking pathways.

27 1.4.2 Connectivity of Retromer with other PD-associated proteins

In the last decades, a number of risk genes associated with PD are discovered and implicated in cellular pathways that may contribute to PD pathogenesis. As the list grows, it appears to be important to define whether these genes function independently or interplay with each other. Recent investigations on inter-relationship between retromer and other PD- linked gene products highlight the possibilities that retromer may coordinate common cellular events together with other PD-linked factors, providing novel insights into the pathogenic roles of retromer in PD.

α-synuclein is a presynaptic neuronal protein central to PD. Genetic studies demonstrated missense and multiplication variants within α-synuclein cause the manifestation of PD [208- 211]. Aberrant forms of α-synuclein are neurotoxic and aggregated in LBs, one of the pathological hallmarks of PD. To date, the molecular mechanism regarding how α-synuclein aggregation is formed remains unclear, however, retromer has been recently demonstrated to contribute to α-synuclein catabolism through coordinating the of cathepsin D, one of the main lysosomal enzymes responsible for the degradation of long- lived proteins. Lack of retromer perturbs the maturation of cathepsin D in parallel with the accumulation of α-synuclein in lysosomes [191]. This is supported by observations made in Vps35-knockdown Drosophila models, demonstrating that loss of retromer Vps35 subunit not only induces the accumulation of insoluble α-synuclein in the brain but also causes locomotor impairments and the mild compound eye disorganization [191]. Similar defects on cathepsin D processing and α-synuclein degradation are observed in cellular models overexpressing the Vps35 D620N or Vps35 R524W variants [193, 195]. Hence, it is likely that retromer functional deficiency contributes to α-synuclein aggregation in an indirect manner, thereby further leading to PD progression. Consistent with the above observations, Vps35 loss-of-function enhances α-synuclein toxicity in yeast and C. elegans [212]. In contrast, overexpressing of wild-type Vps35 in α-synuclein transgenic mouse models almost completely rescues the induced by α-synuclein [212]. However, the presence of PD-linked variant Vps35 D620N or Vps35 P316S or knockdown of endogenous Vps35 causes more neuronal loss in the hippocampus from the α-synuclein transgenic mouse, suggesting the functional role of retromer in antagonizing α-synuclein associated neurodegeneration [212]. The connection between retromer and α-synuclein is further supported by evidence from recent proteomic studies [213]. By using the ascorbate peroxidase (APEX)-labelling biotinylation combined with mass spectrometry, α-synuclein has been shown to directly interact with Vps29 as well as SNX1 [214]. Overall, these findings

28 collectively suggest that retromer may modulate α-synuclein trafficking and biogenesis either in an indirect manner with cathepsin D, or in a direct manner by direct interaction.

Another PD- linked gene possessing functional connection with retromer is LRRK2, a multi- domain protein with GTPase and kinase activity. Variants within LRRK2 are defined as one of the most common genetic causes of the late-onset PD. To date, at least five variants in LRRK2 are shown unambiguously to associate with PD pathogenesis. These include the N1437H, R1441C/G/H, Y1699C, I2020T and G2019S variants, among which G2019S presented most frequently [215]. LRRK2 has been shown to coordinate protein sorting in conjunction with Rab7L1, a gene within the PARK16 PD-associated loci, and defects in the LRRK2-Rab7L1 pathway are implicated in deficits of endosomal trafficking, endo-lysosomal protein degradation and macroautophagy [183, 216-219]. Given the intracellular pathways inhibited by retromer deficiency, it is conceivable that LRRK2-Rab7L1 and retromer may converge in common cellular processes. Indeed, the expression of LRRK2 G2019S variant or silencing of Rab7L1 alters the endosomal sorting of CI-M6PR, similar to Vps35 silencing or Vps35 D620N expression, which can be rescued by the overexpression of the wild-type Vps35 [183]. Furthermore, retromer Vps35 overexpression also rescues neuronal deficits such as perturbed neurite processes and neuronal loss, which are induced by LRRK2 G2019S variant or Rab7L1 knockdown, in mammalian and Drosophila neuronal models [183]. The rescue phenotype is further confirmed by another study in which the relevance is extended to another two LRRK2 variants, Y1699C and I2020T [188]. Although it remains unclear how these three regulators interplay and impact on PD pathogenesis, these observations raise the possibility that retromer and LRRK2 might operate in a common pathway. One recent study showed that Vps35 D620N variant activates LRRK2-mediated of Rab10 to a ignificantly greater extent than LRRK2 G2019S or LRRK2 R1441C. LRRK2-mediated phosphorylation of additional Rab GTPases including Rab8a and Rab12 are also elevated upon the presence of Vps35 D620N variant, suggesting that Vps35 D620N might contribute to PD progression by hyperactivation of the LRRK2 kinase [220, 221]. Co-immunoprecipitation studies in SHSY5Y cells and LRRK2 transgenic mouse brain suggested the direct interaction between LRRK2 and the retromer Vps35 subunit, however, another study in human Vps35-transgenic Drosophila models showed opposite results demonstrating no direct binding between the two regulators [183, 222]. Therefore, LRRK2 may indirectly associate with the retromer complex via other protein scaffolds.

The third PD-linked factor functionally associates with retromer is . Parkin is a RING between RING E3 ubiquitin ligase that plays neuroprotective roles in various cellular

29 processes such as mitochondria quality control and stress protection through ubiquitination of targeting substrates. Variants within the PARK2 gene are associated with the early-onset PD. A strong genetic interaction between Parkin (PARK2) and retromer is first demonstrated by a study in Drosophila models [223]. The heterozygote of Vps35 or Parkin display no phenotype compared with the control, however, Vps35 and Parkin double heterozygotes displayed several PD-linked phenotypes such as enhanced climbing defects, loss of dopaminergic neurons and sensitivity to oxidative stressor [223]. Indeed, Overexpression of Vps35 is able to rescue several phenotypes induced by Parkin variants, suggesting the genetic interaction between Parkin (PARK2) and retromer [223]. As an E3 ubiquitination ligase, PARK2 directly targets to lysine residues of several endosomal proteins for ubiquitination. In fact, Vps35 can be ubiquitinated by PARK2 [224]. In addition, Parkin modulates the endo-lysosomal pathways through Rab7 ubiquitination. Loss of Parkin function perturbs the endosomal recruitment of retromer, thereby further impacting retromer- mediated CI-M6PR trafficking [225].

Collectively, studies on the functional association between retromer and other PD-linked proteins suggest a functional network constructed by multiple risk factors in PD and thus reveal the complicated mechanisms underlying PD progression. Further investigation of the detailed molecular action of retromer in this network will help us to gain a better understanding of its pathogenic roles in PD.

30

Sample size Frequency in Found in Gene Variant Location Country/Region Control Reference PD cases cases Controls subjects Austria 1 family 0 1/1 No Austria 486 1568 2/486 No [170] Germany 376 1014 0/376 No United States 1600 FPD 965 1/1600 No Tunisia 199 FPD 362 1/199 No Yemenite Jews 120 FPD 0 2/120 Unknown Canada 574 FPD 320 0/574 No [171] Norway 688 FPD 600 0/688 No Poland 362 FPD 346 0/362 No Ireland 377 FPD 368 0/377 No Taiwan 406 FPD 348 0/406 No United Kingdom 96 FPD 0 1/96 Unknown [180]

300 FPD, 433 3/300 FPD; 1/433 D620N 15 Japan 579 No [172] SPD SPD

Germany 692 0 1/692 Unknown Serbia 418 0 0/418 Unknown [177] United States 441 0 0/441 Unknown Chile 223 0 0/223 Unknown Vps35 France 246 245 3/246 No [178] Italy 475 FPD 0 0/475 Unknown [175]

101 FPD, 411 0/101 FPD; 0/411 China 371 No [182] SPD SPD

54 FPD, 251 0/54 FPD; 0/251 India 100 No [181] SPD SPD

South African 418 528 0/418 No [174] Greece 202 SPD 131 0/202 No [226] R524W Exon 13 Austria, Germany 860 1014 1/860 No [170] United States 106 FPD 3309 1/106 Yes [171] P316S Exon 9 Unknown 218 188 1/218 No [227] Flanders H599R Exon 14 520 800 1/520 No [228] (Belgian) Flanders M607V Exon 14 520 800 1/520 No [228] (Belgian) Flanders I560T Exon 14 520 800 1/520 No [228] (Belgian) L774M Exon 17 19 countries 8870 6513 6/8870 Yes [229] 19 countries 8870 6513 3/8870 Yes [229] G51S Exon 3 Korea, Norway 396 4794 2/396 Yes [176] R32S Exon 2 Spain 134 0 1/134 Unknown [173] Canada 396 4794 1/396 No [176] K93E Exon 4 United States 1906 2053 1/1906 No [179] M112V Exon 4 Korea 396 4794 1/396 No [176] Vps26A K297X Exon 9 Guam 396 4794 1/396 No [176] R127H Exon 4 Norway, Canada 396 4794 2/396 Yes [176] N308D Exon 9 Canada 396 4794 2/396 Yes [176] Vps29 N72H Exon 4 United States 1906 2053 1/1906 Yes [179] FPD, familial Parkinson's disease; SPD, sporadic Parkinson's disease.

Table 1.2 Summary of retromer variants linked to Parkinson's disease

31 1.5 Aims and Hypotheses

The precise delivery of cargo molecules from the endosome to destined organelles is critical for the maintenance of the endo-lysosomal function. While the retromer complex has been identified to have a central role in the spatial and temporal coordination of the endosomal trafficking, its molecular action remains to be uncovered. Firstly, given the established functions of retrograde tethering factors, it is likely that cargo-loaded transport carriers sorted via the retromer-dependent mechanism can be selectively captured by the specific tethering factors. Different types of retromer complexes, such as SNX3-retromer, SNX27-retromer, and SNX-BAR-retromer are hypothesised to serves distinct functions in this process. Secondly, defects in the endosomal trafficking are often linked to perturbations of the endo- lysosomal functions. While retromer-associated lysosomal defects have been previously described, the question regarding how retromer assists the maintenance of lysosomal function is not fully answered yet. Thirdly, retromer dysfunction has been shown to associate with the Parkinson’s disease (PD). However, how a defect in the endosomal system caused by retromer deficiency contributes to the PD progression remains to be characterized. Here, this thesis will examine these hypotheses via following aims:

1) To determine the molecular action of retromer in the retrograde transport of ETCs.

Recently, Wong and Munro reported a novel experimental strategy to relocate trans-golgins to the mitochondria outer membrane, enabling modified trans-golgins to capture and redirect associated transport vesicles to the mitochondria without fusion [61]. This assay allows the analysis of cargo contents within ETCs and can be utilized to examine the trafficking of ETCs that captured by individual trans-golgins. Initially, the retromer knock-out (KO) cells that are depleted of the core retromer subunit Vps35, will be generated via the CRISPR-Cas9 strategy and utilized for the ETC assay to determine whether retromer has a selective function in sorting associated cargoes into a specific class of ETCs, which can be tethered by specific tethering factors. Cells depleted of retromer accessory SNX proteins including SNX3, SNX27, and SNX-BAR will also be utilized for the ETC assay, to examine the role of different types of retromer complexes in the retrograde trafficking processes.

2) To determine how retromer associated trafficking defects affect the endo- lysosomal function.

Given the functional role in protein trafficking, retromer is expected to contribute to additional cellular processes involved in the endosomal pathways. The endosome-to-TGN retrieval mediated by retromer is more directly linked to the endosomal maturation and lysosomal

32 proteolytic processes. To further understand the function of retromer in the endo-lysosomal pathways, the retromer KO cell lines and rescue cell lines will be applied with a number of assays to examine i) the morphological change of lysosomal compartments; ii) the targeting and processing of lysosomal hydrolases and the lysosomal proteolytic activity; iii) the lysosome-autophagosome fusion and autophagic flux. In addition, the endo-lysosomal functions will also be examined in cells that depleted of the individual trans-golgins associated with the retromer-dependent retrograde trafficking pathways.

3) To fully characterize the partial loss-of-function effects caused by PD-linked retromer variant in the endosomal processes.

The PD-linked retromer variant — Vps35 D620N has been identified to have a partial loss- of-function in the endosomal pathways. Although several previous publications including the work from our group have characterized the role of Vps35 D620N retromer, distinct observations are presented [192-194, 230]. To fully characterize the effects of this PD-linked variant, the rescue cell lines stably expressing Vps35 D620N generated in the Vps35 KO background will be used to examine how a defect the D620N variant will cause in i) the retrograde transport of ETCs; ii) the lysosomal degradation and autophagy pathways; iii) the recruitment of retromer accessory molecules.

By doing so, we will dissect the direct molecular action of retromer in endosomal trafficking and provide insights into the link between retromer deregulation and PD pathogenesis.

33 Chapter 2. Retromer has a selective function in cargo sorting via endosome transport carriers

2.1 Introduction

Lysosomes are dynamic organelles primarily associated with the degradation of macromolecules from the endocytic and autophagic pathway [231, 232]. The proteolytic activity of lysosomes requires the continuous delivery of newly synthesized acid hydrolases, which is achieved through multiple trafficking pathways. One of these is the mannose 6- phosphate receptor (M6PR)-dependent pathway, in which newly synthesized soluble acid hydrolase precursors acquire the mannose 6-phosphate sorting signal and are recognized by M6PR at the TGN [231]. Upon delivery to the endosome, the M6PR-hydrolase complexes dissociate due to the acidic pH, and hydrolases are released to the lumen while unoccupied M6PRs are retrieved to the TGN via the retrograde trafficking pathways. Defects in the M6PR trafficking itinerary lead to the inappropriate sorting and secretion of hydrolase precursors, and therefore impairs the degradative capacity of lysosomes [233].

The endosome-to-TGN retrieval of cargo molecules via endosome transport carriers (ETCs) is spatially and temporally coordinated by multiple protein regulators. One of these is retromer, a protein complex assembled from a high-affinity heterotrimeric core complex composed of Vps35, Vps29 and one of two Vps26 subunits, Vps26A or Vps26B [87, 88, 91, 92, 234]. Retromer functions in exporting endosomal cargoes through molecular interactions with a range of associated proteins [235]. Several studies have shown that the retrograde transport of CI-M6PR relies on the coordination of retromer, and reduced levels of the retromer Vps35 or Vps26A subunits result in CI-M6PR mistrafficking [85, 100, 101, 191, 192, 236-243]. Bugarcic et al (2011) first observed that distinct subtypes of retromer, as defined by the Vps26 subunit incorporated, showed distinct capacity to interact with and facilitate retrograde trafficking of CI-M6PR [91]. It is well established that numerous other non- retromer associated proteins also contribute to the retrograde trafficking of CI-M6PR through direct or indirect mechanisms [85, 94, 95, 244-246]. It remains controversial as to how many independent types of retrograde ETCs are formed from endosomes and what their relative contribution to CI-M6PR trafficking is. Recent studies have even suggested that retromer does not contribute to this retrograde trafficking at all [94, 95].

Also required for the delivery of vesicles through the retrograde transport pathway are the TGN-located tethering proteins, which capture and direct the incoming cargo-loaded ETCs

34 toward the TGN. A range of proteins including multi-subunit protein complexes and trans- Golgi-anchored long coiled-coil proteins have been shown to coordinate tethering process [59, 247]. Recent studies reported that three GRIP domain-containing trans-golgins including GCC88, golgin-97 and golgin-245 are able to selectively capture a specific class of ETCs loaded with CI-M6PR and other retrograde cargoes [61]. The vesicular tethering domains within golgin-97 and golgin-245 are closely related, whereas, that in GCC88 is distinct, suggesting that trans-golgins capture different classes of ETCs [248].

In this study, we reveal that retromer is required for the maintenance of endo-lysosomal dynamics. Loss of retromer Vps35 subunit induces enlarged lysosomes at the ultrastructural level and leads to perturbations in autophagy and lysosomal proteolytic processes. The targeting and processing of M6PR-dependent hydrolases are impaired in the absence of retromer, which is consistent with CI-M6PR mis-trafficking detected in retromer depleted cells. Using the previously established mitochondria targeting assay [61], we further show that the retrograde sorting of CI-M6PR-loaded ETCs mediated by retromer is selectively tethered by GCC88, but not golgin-97 or golgin-245. This trafficking pathway requires SNX3- retromer association and is independent of the SNX27 or the SNX-BAR proteins.

35 2.2 Materials and methods

2.2.1 Chemicals, DNA constructs and antibodies

Magic Red Cathepsin B Kit (938) was from ImmunoChemistry Technologies. DQ™ Red BSA (D12051) was purchased from Thermo Fisher Scientific. Cycloheximide (66-81-9) and AZD8055 (16978) were purchased from Cayman chemical. Chloroquine (C6628) and HEPES (H3375-250G) were purchased from Sigma-Aldrich. Trichloroacetic acid (TCA; T6399) was obtained from Sigma-Aldrich. 4-Methylumbelliferone (4-MU) standard (M333100) was obtained from Toronto Research Chemicals. 4-MU-acid phosphatase substrate (4MU-phosphate; M8883), 4-MU-b-hexosaminidase substrate (4-MU-N-acetyl-b- glucosaminide; M2133), 4-MU-b-galactosidase substrate (4-MU-b-D-galactopyranoside; M1633) and 4-MU-b-glucocerebrosidase substrate (4-MU-b-glucopyranoside; M3633) were obtained from Sigma-Aldrich (USA). Ponceau S solution (P7170) was purchased from Sigma-Aldrich (USA). Ethylenediaminetetraacetic acid (EDTA; 180-500G) and sucrose (530-2KG) were from UNIVAR. Protease inhibitor cocktail EDTA-free tablets (04 639 132 001) were obtained from Roche. Percollä (17-0891-01) was sourced from GE Healthcare. Glycine (010119.0500) was sourced from AnalaR.

Mitochondria targeting golgins including trans-golgin: GCC88ΔC-term-hemagglutinin (HA)- monoamine oxidase A (MAO), golgin-97ΔC-term-HA-MAO, golgin-245ΔC-term-HA-MAO, and cis-golgin: GM130ΔC-term-HA-MAO were obtained from S. Munro [61]. The pCMU- CD8a-CI-M6PR construct was described previously [249]. The Vps35 CRISPR/Cas9 KO plasmid was purchased from Santa Cruz. The SNX3 CRISPR guide RNA (gRNA) plasmid (gRNA targeting sequence: CGGCCGACCCCCACCGTTTG), SNX1 CRISPR gRNA plasmid (gRNA targeting sequence: AAATCATCCTACCATGTTAC) and the SNX2 CRISPR gRNA plasmid (gRNA targeting sequence: TGATGGCATGAATGCCTATA) were synthesized by Genscript, USA. The pEGFP-N1-Vps35 plasmid has been described previously [193]. To generate untagged wild type (WT) Vps35 plasmid, full length human Vps35 was amplified from the pEGFP-N1-Vps35 construct as a template using a 5’ primer CCCACCCGGTACCATGCCTACAACACAGCAG and a 3’ primer CCCACCCCTCGA- GTTAAAGGATGAGACCTTCAT, and subcloned into pcDNA3.1 (+) vector using KpnI and Xhol multicloning sites. Resulted construct was confirmed by DNA sequencing.

Mouse monoclonal anti-CI-M6PR (clone 2G11; ab2733), rabbit monoclonal anti-CI-M6PR (clone EPR6599; ab124767), mouse monoclonal anti-SNX27 (clone 1C6; ab77799), rabbit monoclonal anti-b-hexosaminidase/HEXB (clone EPR7978; ab140649), rabbit monoclonal 36 anti-b-galactosidase-1 (clone EPR8250; ab128993), rabbit polyclonal anti-acid phosphatase 2 (ab84896), rabbit polyclonal anti-b-glucocerebrosidase/GBA antibody (ab92997), rabbit polyclonal anti-SNX3 (ab56078), rabbit polyclonal anti-LC3A/B (ab128025) and rabbit polyclonal anti-Vps26 (ab23892) were purchased from Abcam. Goat polyclonal anti-Vps35 (NB100-1397) was purchased from Novus Biologicals. Mouse monoclonal anti-CD-M6PR (22d4) was purchased from the Developmental Studies Hybridoma Bank. Mouse monoclonal anti-HA (clone 16B12; 901513) was purchased from BioLegend. Rabbit monoclonal anti-HA (clone C29F4; 3724), rabbit monoclonal anti-golgin-97 (clone D8P2K; 13192), rabbit monoclonal anti-Rab5 (clone C8B1; 3547), rabbit monoclonal anti-LC3B/LC3- II (clone D11; 3868) and rabbit monoclonal anti-mTOR (clone 7C10; 2983) were purchased from Cell Signaling Technology. Mouse monoclonal anti-CD8a (OKT8; 14-0086) was purchased from eBioscience. Mouse monoclonal anti-EEA1 (clone 14/EEA1; 610457), mouse monoclonal anti-p230 (clone 15/p230; 611280) and mouse monoclonal anti-LAMP1 (clone H4A3; 555798) were purchased from BD Biosciences. Mouse monoclonal anti- transferrin receptor (clone H68.4; 13-6800) was purchased from Life Technologies. Goat polyclonal anti-cathepsin-D (AF1014) was purchased from R&D systems. Mouse monoclonal anti-alpha-tubulin (clone DM1A; T9026) and mouse monoclonal anti-GAPDH (clone GAPDH-71.1; G8795) were purchased from Sigma-Aldrich. Secondary donkey anti- mouse IgG Alexa FluorTM 488 (A21202), donkey anti-mouse IgG Alexa Fluor TM 555 (A31570), donkey anti-rabbit IgG Alexa FluorTM 488 (A21026) and donkey anti-rabbit IgG Alexa FluorTM 555 (A31572) were purchased from Thermo Fisher Scientific.

2.2.2 Cell culture and Transfection

HeLa cells (ATCC CCL-2) were maintained in high glucose Dulbecco’s Modified Eagle Medium (DMEM; Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (FBS), 2mM L-glutamine (Thermo Fisher Scientific) and 5 mg/ml penicillin and streptomycin

(Thermo Fisher Scientific) in a humidified 37°C incubator with 5% CO2. Transfection was performed using Lipofectamine 2000 (Thermo Fisher Scientific) according to manufacturer’s instructions.

2.2.3 Generation of CRISPR/Cas9 knock-out cell lines

CRISPR/Cas9 knock-out cell lines were generated as previously described [250]. These cell lines were authenticated and confirmed to by mycoplasma free (CellBank Australia). Briefly, HeLa cells stably expressing mCherry-Cas9 were transiently transfected with the CRISPR

37 gRNA plasmids targeting to Vps35, SNX1, SNX2 or SNX3 in a 60 mm dish. Forty-eight hours post-transfection, cells were collected for fluorescence activated cell sorting (FACS). GFP-positive cells were collected and allowed to grow until confluent. Cells were then diluted as one cell per well into 96-well plates until single colonies formed. CRISPR/Cas9 mediated knock-out of targeted genes was confirmed by indirect immunofluorescence assay and immunoblot analysis. HeLa SNX1 KO cell line was transfected with CRISPR gRNA plasmid targeting to SNX2, and then subjected to second round of single cell sorting to generate SNX1/SNX2 double knock-out (SNX1/2 dKO) cell line. The SNX27 CRISPR/Cas9 KO cell line was previously generated [251].

2.2.4 Cell Treatment Procedure

For lysosomal degradation inhibition experiments, cells were treated with 50 µM chloroquine for 6 hrs. Whereas to inhibit mTORC1 activation, cells were treated with AZD8055 – a mTORC1 inhibitor for 25 hrs.

To induce mTORC1 activation by essential amino acid (AA), cells were starved in amino acid-free DMEM medium (D9800-13, USBiological) for 2 hrs before stimulated by MEM essential AA solution (final concentration: 2 x; 11130051, Thermo Fisher Scientific) for 30 min.

2.2.5 Purification of lysosomal fractions

For lysosomal fraction purification, confluent cells were pelleted and resuspended in sucrose buffer (250 mM sucrose, 10 mM HEPES and 1 mM EDTA at pH 7) supplemented with protease inhibitor cocktail). Cell lysates were homogenized and subjected to centrifugation at 1,000 rpm for 10 min at 4 °C to separate the heavy membrane fraction (plasma and nuclear membranes) from the light membrane (endosomes, lysosomes and endoplasmic reticulum etc.) and cytosolic fractions. The light membrane and cytosolic fractions were then collected and layered onto an 18% (v/v) Percollä/sucrose buffer cushion, which were further subjected to centrifugation at 20,000 rpm for 1 h at 4 °C. Sixteen fractions were collected and assayed for the activity of various lysosomal enzymes to confirm the enrichment of lysosomal vesicles.

2.2.6 Enzyme activity assay

The enzyme activity assay has been described previously [252, 253]. Briefly, each subcellular fraction was incubated with the respective 4-MU-linked substrate at 37 °C for 30 38 min or 1 h. 4-MU-linked substrate only was used as a blank control. Reactions were stopped by 200 mM glycine buffer, and enzyme activity was assessed by fluorometric analysis (355 nm/460 nm for 1.0 s) using the VICTOR3 1420 multi-label counter (PerkinElmer, USA) and the Wallac 1420 workstation software (PerkinElmer, USA). The enzyme activity within each sub-fraction was normalised to input protein of the light membrane and cytosolic fraction.

2.2.7 Secretion Assay

The secretion assay was performed as previously described [193]. HeLa cells were seeded onto 6-well plate 24 h prior to use. Cells were incubated with serum-free DMEM containing cycloheximide (100 µg/ml; Sigma-Aldrich) at 37 °C for indicated time points. Medium samples were collected, concentrated by using 10% trichloroacetic acid (Sigma-Aldrich) at 4 °C overnight, and analysed using western blotting.

2.2.8 SDS-PAGE and western immunoblotting

SDS-PAGE and western blotting was performed as previously described [254]. In brief, cell lysates were harvested in a lysis buffer containing protease inhibitor cocktail. Protein concentrations were quantified by Pierce bicinchoninic acid (BCA) assay (Thermo Fisher Scientific). Equal amounts of protein samples were resolved on SDS-PAGE and transferred onto the PVDF membrane (Millipore) according to manufacturer’s instructions. Western blotting was developed either by enhanced chemiluminescent (ECL) system using Clarity ECL substrate (Bio-rad) or Odyssey® infrared imaging system (Li-cor).

2.2.9 Internalization assay

HeLa cells grown on coverslips were co-transfected with the CD8a-M6PR reporter and individual mitochondria-targeting golgins for 24 h. Transfected cells were washed with ice- cold PBS for twice and incubated in fresh serum free medium containing anti-CD8a antibody (5 μg/ml) on ice for 30 min. After washing with ice-cold PBS and low-pH solution containing 0.1M glycine, 0.15M NaCl (pH 3.0) to remove unbounded antibodies, cells were then chased in serum free medium for indicated time points at 37 °C to allow the internalization of bounded antibodies. At indicated time points, cells were fixed with 4% PFA and subjected to indirect immunofluorescence.

39

2.2.10 DQ™ Red BSA Assay

HeLa cells grown on coverslips were incubated with culture medium containing 10 µg/ml of DQ™ Red BSA (Thermo Fisher Scientific) at 37 °C overnight. Cells were washed with PBS for three times, fixed and subjected to indirect immunofluorescence.

2.2.11 Magic red cathepsin-B assay

HeLa cells were seeded onto 96-well imaging plate at 30,000 cells per well 24 h prior to use. To measure lysosomal cathepsin B activity, cells were incubated with Magic Red Cathepsin- B Kit (ImmunoChemistry Technologies) according to the manufacturer’s instructions, and live cell time-lapse imaging was recorded using the Nikon Ti-E inverted microscope equipped with a 40x Plan Apochromatic objective.

2.2.12 Indirect Immunofluorescence and co-localization analysis

Cells on coverslips were routinely fixed with 4% PFA in PBS for 15 min then permeabilised with 0.1% Triton X100 in PBS for 10 min. Alternatively for LC3 and mTORC1 staining, coverslips were fixed in ice-cold methanol for 5 min at -20 °C. Fixed and permeabilised cell monolayers were blocked with 2% BSA in PBS for 30 min to reduce non-specific binding. Cells were then labelled with diluted primary antibodies and corresponding secondary antibodies sequentially for 1 hr at room temperature. Coverslips were mounted on glass microscope slides using Fluorescent Mounting Medium (Dako), and the images were taken at room temperature using the Zeiss LSM 710 Meta confocal laser scanning microscope or the Leica DMi8 SP8 Inverted confocal equipped with 63x Plan Apochromatic objectives. For quantification, images were taken from multiple random positions for each sample.

Images were processed using ImageJ software. Co-localization analysis was performed using the ImageJ JACoP plugin [255]. Transfected cells were segregated from fields of view containing both transfected and non-transfected cells by generating regions of interest (ROI). The selected ROI was cropped, split into separated channels and applied for threshold processing. Co-localization analysis was conducted on 3 independent experiments. Co- localization values were exported to GraphPad Prism 7 software and tabulated accordingly.

40 2.2.13 Electron Microscopy

Electron microscopy was performed as previously described [250]. Briefly, cells seeded onto 35 mm dishes were fixed in 2.5% glutaraldehyde in PBS and processed for flat embedded in resin. Upon curing, the resin containing cells was broken away from the plastic dish. Ultrathin sections (60 nm) were cut and imaged using a Jeol 1011 transmission electron microscopy at 80 kV fitted with a Morada Soft Imaging camera (Olympus) at 2-fold binning. The volume density of lysosomal compartments (defined by size and electron dense/multivesicular content) relative to the cytoplasmic volume was determined by point counting using a double lattice grid as described previously [256].

2.2.14 Statistics

All statistical analyses were completed using GraphPad Prism software 7 and described in the appropriate figure legends. Error bars on graphs were represented as the standard error of the mean (± SEM). P values were calculated using two-tailed Student’s t test. P < 0.05 was considered as significant.

41 2.3 Results

2.3.1 Ultrastructural alteration of lysosomal structures and elevated autophagy upon retromer deficiency

Aberrant lysosomal morphology and function were previously reported upon Vps35 knockdown in Drosophila larva fat body cells [257]. Using CRISPR/Cas9-mediated gene editing technology, clonal Vps35 KO HeLa cells were generated, in which the core retromer subunit Vps35 was depleted, as determined by immunoblotting (Figure 2.1A). As expected, the absence of Vps35 prevents the formation of the trimeric retromer complex, resulting in reduced levels of the other two subunits - Vps26A and Vps29 (Figure 2.1A). Stable rescue cell lines were generated by the expression of GFP tagged wild-type Vps35 in the Vps35 KO cell lines. By immunoblotting with Vps35, Vps26A, and Vps29 antibodies, we failed to detect any cell clones demonstrating full rescue of the expression for the retromer complex subunits (Figure 2.1A). Therefore, we selected the highest expressing clone for comparison. To determine whether the deficiency of retromer affects lysosomal compartments, we analyzed control HeLa cells and Vps35 KO cells using electron microscopy. Late endosomal/lysosomal structures, defined as large circular and electron-dense organelles, occupied approximately 3 % of the total cytoplasmic volume in the control HeLa cells (Figure 2.1B). In comparison, the lysosomes of Vps35 KO cells showed an almost 3-fold increase in volume (Figure 2.1B), which is similar to that described in Drosophila models [257]. Rescue of the Vps35 KO cells by expression of Vps35-GFP was able to fully restore this phenotype (Figure 2.1B). HeLa and Vps35 KO cells were next treated with Lysotracker-Red, a fluorescent dye labeling acidic endo-lysosomal compartments. Flow cytometry revealed comparable levels of fluorescence intensities for Lysotracker-Red between HeLa and Vps35 KO cells (Figure 2.1C), which indicates similar levels of acidic endo-lysosomal compartments and further suggests a proportion of the observed enlarged and electron- dense lysosomal compartments within Vps35 KO cells are not acidic and might represent lysosomes that are functionally impaired.

Functional lysosomes are required to fuse with autophagosomes for the final stages of autophagy. During this step, membrane bound LC3-II that exists on luminal membrane of the autophagosome is degraded in the new autolysosome environment. To determine if retromer deficiency affects this process, the subcellular distribution of LC3-II was examined. HeLa, Vps35 KO and Vps35-GFP rescue cells were co-immunolabelled with antibodies against the endogenous LC3 and LAMP1 – a late-endosome/lysosome marker. As indicated

42 by confocal microscopy, Vps35 KO cells showed more membrane-bound LC3-II, with an increased co-localization with LAMP1, relative to HeLa and Vps35-GFP rescue cells (RHeLa

= 0.4983; RVps35 KO = 0.7217; RVps35-GFP rescue = 0.3133; Figure 2.1D), suggesting reduced degradation of autophagic cargo caused by retromer deficiency. Changes in the basal levels of autophagy may be due to alterations in the kinetics of autophagy initiation (i.e. autophagosome formation) or the maturation of autophagosomes by fusion with lysosomes to generate functional degradation compartments, referred to as autophagic flux. These two processes can be uncoupled by chloroquine treatment, which selectively blocks the autophagic degradation function by preventing autophagosome-lysosome fusion [258]. HeLa and Vps35 KO cells were treated with chloroquine for 6 h, and cell lysates were collected for immunoblotting. As expected, with chloroquine treatment, an increased level of LC3-II (~ 5 fold) was detected in HeLa cells (Figure 2.1E), consistent with a blockage of autophagic degradation. Vps35 KO cells exhibited higher basal levels of LC3-II (Figure 2.1E), which is in agreement with the immunofluorescent observations. Upon treatment with chloroquine, the LC3-II level increased slightly (~ 1.5 fold) in Vps35 KO cells, which was still higher than that in treated HeLa cells (Figure 2.1E). These observations indicate that autophagosomes formed in Vps35 KO cells still undergo autophagic flux which is inhibited by chloroquine treatment of these cells.

To further determine if the induction of autophagy is affected by retromer deficiency, we investigated the intracellular activity of mTORC1 signalling pathway, which is antagonistic to the induction of autophagy. As revealed by confocal microscopy, under amino acid starved conditions, mTORC1 displayed a diffuse, cytoplasmic distribution in HeLa cells, whereas stimulation with essential amino acids promoted the recruitment of mTORC1 from the cytosol to lysosomal compartments, as demonstrated with colocalization with LAMP1 (Figure 2.1F). In Vps35 KO cells, while essential amino acid stimulation still resulted in the lysosomal recruitment of mTORC1, co-localization analysis indicated a significantly decreased recruitment when compared to HeLa control, suggesting decreased mTORC1 signalling activity. The decreased mTORC1 recruitment to lysosome by amino acid stimulation in Vps35 KO cells was rescued in Vps35-GFP rescue cells (HeLa: R-AA = 0.1967,

R+AA = 0.4933; Vps35 KO: R-AA = 0.1333, R+AA = 0.3000; Vps35-GFP rescue: R-AA = 0.1467,

R+AA = 0.4417; Figure 2.1F). To further confirm this finding, treatment of AZD8055, a pan- mTOR inhibitor, was employed to examine the rate of autophagy initiation. HeLa and Vps35 KO cells were treated with AZD8055 for 25 h, and cell lysates were collected for immunoblotting. As expected, AZD8055-treated HeLa cells demonstrated increased LC3-II levels, consistent with an elevation in autophagy induction (Figure 2.1G). In contrast,

43 mTORC1 inhibition in Vps35 KO cells did not increase the LC3-II level further when compared to untreated cells, indicating little response in autophagy induction upon mTORC1 inhibition (Figure 2.1G). Taken together, these observations indicate that retromer deficiency induces dysfunctional autophagy by causing an elevated autophagy initiation and potentially altering the autophagic flux.

(* See figure legend on the next page.)

44 Figure 2.1. Ultrastructural alteration of lysosomal structures and elevated autophagy in Vps35 KO cells.

(A) Generation of CRISPR/Cas9 mediated Vps35 KO HeLa cells and Vps35-GFP rescue cells. Equal amounts of cell lysates from HeLa, Vps35 KO and Vps35-GFP rescue cells were subjected to SDS- PAGE and immunoblotted with antibodies against Vps35, Vps26A, Vps29 and tubulin. (B) Electron micrographs of HeLa, Vps35 KO and Vps35-GFP rescue cells. Enlarged circular structures are indicated as late endosomal/lysosomal structures. Scale bars represent 2000 nm, in zoomed images represent 500 nm. Graph represents the percentage of cytoplasmic space occupied by lysosomal compartments in HeLa, Vps35 KO and Vps35-GFP rescue cells (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance. **, p < 0.01, ***, p < 0.001. N = two independent experiments with ten images each. (C) Flow cytometric analysis of cellular acidification based on Lysotracker fluorescence in HeLa and Vps35 KO cells. Graph represents the mean fluorescent intensity (MFI) within HeLa and Vps35 KO cells (means ± SEM). Two-tailed student’s t test was utilized to determine the statistical significance (n = 3). ns, not significant. (D) HeLa, Vps35 KO and Vps35-GFP rescue cells were fixed, co-immunolabelled with antibodies against LC3-II and LAMP1, followed by Alexa Fluor conjugated fluorescent secondary antibodies. Scale bars, 5 μm. The co-localization between LC3-II and LAMP1 was quantified by the Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance among HeLa, Vps35 KO and Vps35-GFP rescue cells upon amino acid stimulation (n = 3). ***, p < 0.001, ****, p < 0.0001. (E) HeLa and Vps35 KO cells were treated with chloroquine (CQ, 50 μM) for 6 h. Cells were harvested, and equal amounts of protein samples were used for SDS-PAGE and immunoblotting with antibodies against LC3-II, Vps35 and GAPDH. Graph represents the level of LC3-II normalized to GAPDH (mean ± SEM). Two-tailed Student’s t test was used to determine the statistical significance (n = 3). *, p < 0.05, **, p < 0.01. (F) Amino acid starved HeLa, Vps35 KO and Vps35-GFP rescue cells were treated with 2x essential amino acid (AA) solution for 30 min, fixed with ice-cold methanol and co-immunolabelled with antibodies against mTORC1 and LAMP1, followed by Alexa Fluor conjugated fluorescent secondary antibodies (means ± SEM). Scale bars, 5 μm. The co-localization of mTORC1 with LAMP1 was quantified by Pearson’s correlation coefficient. Two-tailed Student’s t test indicates the difference between HeLa and Vps35 KO cells upon amino acid stimulation (n = 3). ***, p < 0.001, ****, p < 0.0001. (G) HeLa and Vps35 KO cells were treated with AZD8055 (1 μM) for 25 h before subjected to SDS-PAGE and immunoblotted with antibodies against LC3-II, Vps35 and GAPDH. Graph represents the expression level of LC3-II normalized to GAPDH (mean ± SEM). Two-tailed Student’s t test was used to determine the statistical significance (n = 3). *, p < 0.05.

2.3.2 Retromer deficiency affects lysosomal activity

As lysosomes play important roles in protein degradation, we compared intracellular proteolytic kinetics using DQ-BSA, a self-quenched dye conjugated with BSA, which enters into the endosomal system through endocytosis and generates strong fluorescent signal upon proteolytic cleavage within lysosomal compartments. Initially, cells were treated with Alexa fluor 647 conjugated BSA for 1 h at 37°C to measure the delivery efficiency of endocytosed material to endosomes. Fluorescence imaging showed no difference in fluorescent BSA conjugate staining between HeLa and Vps35 KO cells (Figure 2.2A), indicating that the comparable equilibrium rates of endocytosis/exocytosis for BSA trafficking. HeLa, Vps35 KO and Vps35-GFP rescue cells were treated with DQ-BSA overnight, fixed and immunolabelled with antibodies against the lysosome marker, LAMP1.

45 Confocal microscopy and fluorescence intensity analysis revealed strong DQ-BSA fluorescence signal in both HeLa and Vps35-GFP rescue cell lines, which was largely overlapping with LAMP1 staining, indicating efficient proteolysis of DQ-BSA within lysosomes (Figure 2.2B). In contrast, Vps35 KO cells had a significantly lower DQ-BSA fluorescence within LAMP1-positive lysosomes, indicating impaired lysosomal proteolysis (Figure 2.2B).

(* See figure legend on the next page.)

46 Figure 2.2. Deficiency of retromer causes reduced lysosomal activity.

(A) HeLa and Vps35 KO cells were treated with AF647 conjugated BSA (200 µg/ml) in complete medium at 37°C for 1 h. Cells were then fixed and subjected to fluorescence microscopy. Graph represents the fluorescent intensity of endocytosed BSA conjugate (A.U, arbitrary units) within HeLa and Vps35 KO cells (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ns, not significant. (B) HeLa, Vps35 KO and Vps35-GFP rescue cells were treated with DQ-BSA Red (10 µg/ml) in complete medium at 37°C for overnight. Cells were fixed and immunolabelled with antibodies against LAMP1, followed by Alexa Fluor conjugated fluorescent secondary antibodies. Scale bars, 10 μm. Graph represents the fluorescent intensity of DQ-BSA Red (A.U, arbitrary units) within HeLa, Vps35 KO and Vps35-GFP rescue cells (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). *, p < 0.05, ***, p < 0.001. (C) Lysosomal sub-fraction from HeLa and Vps35 KO cells were prepared and the protein levels of lysosome enzymes b-hexosaminidase, acid phosphatase 2, b-galactosidase and b-glucocerebrosidase were determined by immunoblotting. Ponceau S staining of the relevant area of the blot was included as a loading control. Graph represents the level of b-galactosidase within the lysosome enriched fraction of HeLa and Vps35 KO cells (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). *, p < 0.05. (D) Lysosomal Enzyme activities from lysosome enriched fractions were analyzed by fluorometric analysis using a 4MU-detection system and were calculated in nmol/min/mg of input protein. Graph represents the fold difference of enzyme activities between HeLa and Vps35 KO cells (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). *, p < 0.05. (E) HeLa, Vps35 KO and Vps35-GFP rescue cells were treated with Magic Red Cathepsin-B in complete medium at 37 °C and imaged by a time-lapsed video microscopy performed with the inverted Nikon Ti-E microscopy with Hamamatsu Flash 4.0 sCMOS camera. Graph represents the Magic Red Cathepsin B fluorescent intensity (A.U, arbitrary units) within HeLa, Vps35 KO and Vps35-GFP rescue cells at the indicated time (means ± SEM). N = two independent experiments with eight random points each. *, p < 0.05, **, p < 0.01, ****, p < 0.0001, ns, not significant.

Efficient lysosomal proteolysis is achieved through the action of various lysosomal enzymes. To determine if lysosomal enzyme activities are affected by retromer deficiency, lysosomal sub-fractions from HeLa and Vps35 KO cells were purified and utilized to investigate the levels of four common enzymes b-hexosaminidase, acid phosphatase 2, b-galactosidase, and b-glucocerebrosidase. As indicated by immunoblotting, the mature forms of all four enzymes were detectable in lysosomal sub-fractions from control cells. In comparison, lysosomal sub-fractions from Vps35 KO cells demonstrated unchanged levels of b- glucocerebrosidase and acid phosphatase 2 (Figure 2.2C), both of which are targeted to lysosomes independent of M6PR [259]. b-galactosidase and b-hexosaminidase depend on M6PR for delivery to lysosomes [259], but only b-galactosidase displayed a moderate decrease in lysosome protein levels in Vps35 KO cells when compared to HeLa control (Figure 2.2C). In addition, fluorometric analysis was employed to examine the hydrolytic activity of the four enzymes in lysosomal sub-fractions from HeLa and Vps35 KO cells. In comparison to control cells, lysosomes purified from Vps35 KO cells demonstrated the same levels of enzyme activity for b-hexosaminidase, acid phosphatase and b- glucocerebrosidase, but significantly reduced activity of b-galactosidase (Figure 2.2D),

47 consistent with the observed protein levels. Collectively, these data suggest the absence of retromer selectively can reduce the lysosomal targeting of M6PR-dependent enzymes, thereby affecting their functional levels within lysosomes. We next monitored the changes in enzyme targeting and activities within live intact cells using Magic Red cathepsin-B assay. HeLa and Vps35 KO cells were treated with Magic Red cathepsin-B, a cell permeant fluorogenic substrate, which becomes fluorescent when cleaved by cathepsin enzymes in lysosomes. Live cell time-lapse imaging and fluorescent intensity analysis revealed Magic Red fluorescence in HeLa cells increased rapidly during the first 30 min, then maintained at relatively stable levels after that (Figure 2.2E). In contrast, fluorescence intensities for Magic Red in Vps35 KO cells were approximately 40% lower than that in control cells, suggesting reduced cathepsin enzyme activity in the absence of retromer (Figure 2.2E). In contrast to other results, the Vps35-GFP rescue cells showed only a partial rescue of this phenotype suggesting this aspect of endosome function is more sensitive to the level of retromer present in these cell models.

2.3.3 Retromer deficiency causes defects in CI-M6PR trafficking and its downstream cathepsin-D processing

As the lysosomal activities of M6PR-dependent hydrolases were impacted upon retromer KO, we next employed cycloheximide chase experiments to investigate the processing and secretion of newly-synthesized cathepsin-D, a well-characterized hydrolase relying on CI- M6PR for lysosomal delivery [91, 123, 193]. HeLa and Vps35 KO cells were treated with cycloheximide, and medium samples were collected from cell monolayers containing equal number of cells at 0, 3, 6 and 9 h post-chase. As indicated by immunoblotting, the secreted cathepsin-D precursor was detected at 3 h post-chase in HeLa cells (Figure 2.3A and Figure 2.S1A). In comparison, the secretion level of cathepsin-D precursor was significantly higher in the media from Vps35 KO cells at 3 h post-chase, and further increased at 6 and 9 h post- chase (Figure 2.3A and Figure 2.S1A). Immunoblotting of cell lysates to quantify tubulin levels confirmed that the secreted material was collected from equivalent numbers of cells (Figure 2.S1A). Analysis of the cathepsin-D precursor in the untreated cell monolayer lysates indicates a higher level of precursor within Vps35 KO cells (Figure 2.S1A and S1B). The ratio of secreted cathepsin-D precursor to the basal amount in the cell lysates showed that more than half of the precursor was secreted to the medium at 6h post-chase in Vps35 KO cells, whereas only around 30% precursor was secreted in HeLa cells (Figure. 2.S1C). Collectively this data indicates that retromer deficiency perturbs the intracellular processing

48 of the cathepsin-D precursor, leading to its increased secretion rather than delivery to endosomes.

Impairment of M6PR-dependent lysosomal enzymes is most likely due to improper intracellular trafficking of CI-M6PR. Dysregulation of retromer has been previously shown to perturb the endosome-to-TGN delivery of CI-M6PR [85, 91, 100, 191, 236, 238-242]. To examine if the retrograde transport of CI-M6PR is perturbed in the Vps35 KO cells, we performed a series of co-localization experiments using TGN and endosome markers. HeLa and Vps35 KO cells were co-immunolabelled with antibodies against endogenous CI-M6PR together with individual organelle markers. As expected, CI-M6PR was observed to tightly localize to perinuclear regions in control cells at steady state and displayed high levels of co-localization with p230-positive TGN, EEA1-positive early endosomes and transferrin receptor (TfR)-positive recycling endosomes, but with minor or no overlap with LAMP1- positive lysosomes (Figure 2.3B). However, its intracellular distribution was clearly altered in Vps35 KO cells, displaying a relatively dispersed peripheral pattern (Figure 2.3B), in agreement with previously observations [85, 91, 100, 191, 236, 238-242]. In fact, co- localization analysis revealed a reduced overlap of CI-M6PR with p230, but a mildly increased overlap with EEA1 and TfR in Vps35 KO cells (p230: RHeLa = 0.4005, RVps35 KO =

0.3264; EEA1: RHeLa = 0.4220, RVps35 KO = 0.5219; TfR: RHeLa = 0.4229, RVps35 KO = 0.5007; Figure 2.3B), suggesting impaired CI-M6PR retrograde transport in the absence of retromer. In order to further confirm the disrupted trafficking itinerary caused by retromer deficiency, we next employed a CD8a internalization assay to investigate the intracellular trafficking of CI-M6PR. HeLa and Vps35 KO cells transiently transfected with the CD8a-CI-M6PR reporter were incubated with antibodies against CD8a on ice for 30 min, washed twice with acid solutions, CD8a internalization was then chased at 37 °C for up to 60 min, fixed and immunolabelled with antibodies against golgin-97, a TGN marker, or Rab5, an early endosome marker. As revealed by confocal microscopy and co-localization analysis, antibodies bound to CD8a-CI-M6PR were internalized following endocytic pathways and delivered to the golgin-97 positive TGN compartment at 60 min post-chase in control cells, suggesting the efficient retrograde transport of the reporter (Figure 2.3C). In contrast, reduced CD8a-CI-M6PR delivery to the TGN was observed in Vps35 KO cells, which was correlated with a significantly increased co-localization with early endosomal compartments at 60 min post-chase (Golgin-97: RHeLa = 0.3908, RVps35 KO = 0.2340; Rab5: RHeLa = 0.1960,

RVps35 KO = 0.2453; Figure 2.3C). This data indicates that retromer is required for the efficient

49 retrograde transport of CI-M6PR, and its deficiency induces increased receptor retained in endosomal compartments.

(* See figure legend on the next page.)

50 Figure 2.3. Vps35 is required for efficient CI-M6PR trafficking and cathepsin-D processing.

(A) HeLa and Vps35 KO cells were pulsed with 100 μg/ml of cycloheximide in serum free medium containing 2 mM glutamine for 0, 3, 6 and 9 h. The medium samples were collected from cell monolayers containing equal number of cells and subjected to immunoblotting with antibodies against cathepsin-D. Corresponding cell lysates were collected and Tubulin was utilized as the loading control. (B) The distribution of CI-M6PR in HeLa and Vps35 KO cells was determined by the indirect immunofluorescence using antibodies against endogenous CI-M6PR, p230, EEA1, TfR and LAMP1. Scale bars, 10 μm. The co-localization between CI-M6PR and different organelle markers was quantified by the Pearson’s correlation coefficient and represented in graphs (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). *, p < 0.05, **, p < 0.01, ns, not significant. (C) The internalization assay was performed in HeLa and Vps35 KO cells transiently transfected with the CD8a-CI-M6PR construct, using antibodies against CD8a. The delivery of CD8a-CI-M6PR complex to the TGN and early endosome after uptake for 60 min at 37 °C was determined by the indirect immunofluorescence using antibodies against golgin-97 and Rab5, followed by Alexa Fluor conjugated fluorescent secondary antibodies. Scale bars, 10 μm. The co- localization between CD8a-CI-M6PR and indicated molecular markers was quantified by the Pearson’s correlation coefficient and represented in graphs (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ****, p < 0.0001.

2.3.4 A subset of CI-M6PR-containing endosome derived transport vesicles that are tethered by GCC88 depend on retromer for their generation

Retrograde transport of CI-M6PR is achieved by multiple intracellular pathways engaging a range of intracellular machinery. For example, GRIP domain-containing golgin proteins including GCC88, golgin-97 and golgin-245 have all been implicated as the tethering proteins for ETCs incorporating CI-M6PR [61, 249]. We have applied a recently developed rerouting and capture assay, by which mitochondrial targeted golgins capture and relocate ETCs loaded with specific classes of cargoes to the mitochondria instead of the TGN [61]. HeLa and Vps35 KO cells were transiently transfected with individual HA-tagged golgin- MAO constructs, in which the golgin’s C-terminal Golgi-targeting domain was engineered to be replaced by the mitochondria-targeting transmembrane domain of monoamine oxidase (MAO) [61]. Transfected cells were fixed and co-immunolabelled against HA and endogenous CI-M6PR. Confocal microscopy revealed a filamentous, mitochondrial pattern for CI-M6PR in control cells when expressing GCC88-MAO, golgin-97-MAO or golgin-245- MAO construct, whereas cells expressing the cis-golgin, GM130-MAO, demonstrated little relocation of CI-M6PR to the mitochondria (Figure 2.4A), consistent with previous observations [61]. Indeed, co-localization analysis indicated significant overlap of CI-M6PR with HA-labelled golgin-MAO in HeLa cells expressing GCC88-MAO, golgin-97-MAO or golgin-245-MAO (HeLa: RGCC88-MAO = 0.3911, Rgolgin-97-MAO = 0.6186 and Rgolgin-245-MAO =

51 0.3564), compared to that of cells expressing GM130-MAO (HeLa: RGM130-MAO = 0.1516; Figure 2.4B). Retromer deficiency did not affect the relocation of CI-M6PR-loaded ETCs captured by golgin-97-MAO and golgin-245-MAO, but it did perturb the relocation of CI- M6PR ETCs tethered by GCC88 (Figure 2.4A). In support of this, co-localization analysis revealed a high level of overlap of CI-M6PR with golgin-97-MAO or golgin-245-MAO in

Vps35 KO cells (Vps35 KO: Rgolgin-97-MAO = 0.5890 and Rgolgin-245-MAO = 0.2906), which was similar to control HeLa, but a significant reduction in the co-localization of CI-M6PR with

GCC88-MAO (Vps35 KO: RGCC88-MAO = 0.1772; Figure 2.4B). Taken together, this data indicates that retromer is selectively required for the incorporation of CI-M6PR into a subset of ETCs defined by their capacity to be tethered by GCC88. To determine if retromer can be retained on the released ETCs and potentially function as part of the GCC88 tethers, confocal microscopy and colocalization analysis revealed little overlap of Vps35 with GCC88-MAO or the other mitochondrial targeted golgins (Figure 2.S2A), suggesting that retromer is not associated with the ETCs released from the endosomes, which is consistent with previous reports [127].

To further test the specificity for GCC88 tethered retromer-sorted ETCs, we performed the re-routing assay to investigate the trafficking of CD-M6PR, proposed to be a retromer- independent retrograde cargo [100, 238]. HeLa and Vps35 KO cells were transiently transfected with individual HA-tagged golgin-MAO constructs, fixed and immunolabelled with antibodies against HA and endogenous CD-M6PR. Confocal microscopy revealed relocation of CD-M6PR loaded ETCs to the mitochondria in both control HeLa and Vps35 KO cells transfected with the GCC88-MAO, golgin-97-MAO or golgin-245-MAO construct (Figure 2.4C; Figure 2.S3). This observation was further supported by the co-localization analysis (HeLa: RGCC88-MAO = 0.3973, Rgolgin-97-MAO = 0.6798, Rgolgin-245-MAO = 0.5045 and

RGM130-MAO = 0.1183; Vps35 KO: RGCC88-MAO = 0.2783, Rgolgin-97-MAO = 0.7056, Rgolgin-245-MAO =

0.5537 and RGM130-MAO = 0.0796; Figure 2.4C). Collectively, this data suggests that the absence of retromer has no impact on GCC88-captured ETCs loaded with cargoes that traffic in a retromer-independent manner. These observations are consistent with retromer functions involving a direct action in ETC formation and protein trafficking, through binding with its associated cargoes, including CI-M6PR.

To confirm this phenotype, two additional cell models were generated. Firstly, a stable wild- type Vps35 rescue cell line was generated by the expression of untagged wild-type Vps35 in the Vps35 KO cell lines (Figure 2.S2B). Secondly, a second retromer subunit, Vps26A KO cell line was independently generated using the same approach. These cell lines were

52 utilized for the re-routing assay to investigate the trafficking of CI-M6PR vesicles (Figure 2.4D and Figure 2.S2B). Vps35 KO, Vps35-GFP rescue, Vps35-wt rescue and Vps26A KO cells were transiently transfected with GCC88-MAO or GM130-MAO, fixed and co- immunolabelled with antibodies against HA and endogenous CI-M6PR. As expected, CI- M6PR staining in both Vps35 rescue cell lines showed a significantly increased co- localization with HA-GCC88-MAO compared to Vps35 KO cell line (Vps35-GFP rescue:

RGCC88-MAO = 0.4363; Vps35-wt rescue: RGCC88-MAO = 0.3634; Vps35 KO: RGCC88-MAO = 0.1548; Figure 2.4D). The Vps26A KO cells showed the same phenotype as the Vps35 KO cells with a reduced co-localization of CI-M6PR with HA-GCC88-MAO, when compared to control cells (HeLa: RGCC88-MAO = 0.4258; Vps26A KO: RGCC88-MAO = 0.1789; Figure 2.S2B). These results further verify the essential action of retromer in the retrograde transport of GCC88- tethered CI-M6PR vesicles.

(* See figure legend on the next page.) 53 Figure 2.4 ETCs containing CI-M6PR, that are tethered by GCC88, are absent in Vps35 KO cells.

(A) HeLa and Vps35 KO cells were transiently transfected with individual HA-tagged mitochondria- targeting golgin constructs: GCC88-MAO, Golgin-97-MAO, Golgin-245-MAO and GM130-MAO. Cells were fixed, co-immunolabelled with antibodies against HA and endogenous CI-M6PR, followed by Alexa Fluor conjugated fluorescence secondary antibodies. Scale bars, 10 μm. The intensity plots of the fluorescent intensity (y-axis) against distance (x-axis) represent the overlap between channels. (B) The co-localization between CI-M6PR and HA-tagged golgin-mito proteins was quantified by Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ****, p < 0.0001, ns, not significant. (C) HeLa and Vps35 KO cells were transiently transfected with HA-tagged mitochondria-targeting golgin constructs: GCC88-MAO, Golgin-97-MAO, Golgin-245-MAO and GM130-MAO, fixed and co-immunolabelled with antibodies against HA and endogenous CD-M6PR, followed by Alexa Fluor conjugated fluorescence secondary antibodies. Scale bars, 10 μm. The intensity plots of the fluorescent intensity (y-axis) against distance (x-axis) represent the overlap between channels. The co-localization between CD-M6PR and HA-tagged golgin-mito proteins was quantified by Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t was utilized to determine the statistical (n = 3). ns, not significant. (D) Vps35 KO, Vps35-GFP rescue and the untagged wild-type Vps35 rescue cells were transiently transfected with HA-tagged GCC88-MAO or GM130-MAO construct, fixed and co- immunolabelled with antibodies against HA and endogenous CI-M6PR, followed by Alexa Fluor conjugated fluorescence secondary antibodies. The co-localization between CI-M6PR and HA- tagged golgin-mito protein GCC88 and GM130 was quantified by Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ****, p < 0.0001, ns, not significant.

2.3.5 SNX3 associates with retromer to coordinate the trafficking of GCC88-tethered CI-M6PR containing ETC

Current models propose that within endosomes the retromer contributes to the sorting and transport of a range of cargoes via interactions with distinct cargo binding proteins. To determine which of these are contributing to retromer dependent trafficking of CI-M6PR we generated a series of KO cells depleted for the following cargo binding retromer complexes, SNX3-Retromer, SNX-BAR-Retromer and SNX27-Retromer. Initially, we employed the CRISPR/Cas9-based targeting plasmids to generate clonal SNX1/2 double knock-out (dKO), SNX3 KO and SNX27 KO cell lines. HeLa SNX27 KO was generated and described previously [251]. Immunoblotting revealed the absence of SNX1 and SNX2 in SNX1/2 dKO cell line (Figure 2.5A), further resulting in the loss of SNX5 and SNX6 expression whose stability and endosome recruitment is dependent on the formation of the functional SNX- BAR dimers [260]. On the other hand, the depletion of SNX1/2, SNX3 or SNX27 had no effect on the levels of Vps35 and Vps26A (Figure 2.5A). The rerouting assay was then performed using these KO cell lines to investigate the trafficking of GCC88-tethered CI- M6PR vesicles. HeLa, SNX1/2 dKO, SNX3 KO and SNX27 KO cells were transiently transfected with HA-tagged GCC88-MAO or GM130-MAO, fixed and co-immunolabeled with

54 antibodies against HA and endogenous CI-M6PR. Co-localization analysis was performed to quantify the relocation of CI-M6PR to the golgin-positive mitochondria. CI-M6PR is redistributed to the mitochondria in GCC88-MAO transfected HeLa cells (HeLa: RGCC88-MAO = 0.3804; Figure 2.5B and Figure 2.S4A). Similar levels of overlap between CI-M6PR and GCC88-MAO were observed in both SNX1/2 dKO cells and SNX27 KO cells, when compared to the control cells (SNX1/2 dKO: RGCC88-MAO = 0.4340; SNX27 KO: RGCC88-MAO = 0.3950; Figure 2.5B and Figure 2.S4A). In contrast, a significantly decreased co-localization of CI-M6PR with GCC88-MAO was detected in SNX3 KO cells (SNX3 KO: RGCC88-MAO = 0.1556; Figure 2.5B and Figure 2.S4A), which was similar to that observed in Vps35 KO cells. This data indicates that apart from the retromer, the endosomal transport of the subset of CI-M6PR vesicles captured by GCC88 also depends on the coordination of SNX3, but not SNX-BAR proteins and SNX27. Like retromer, SNX3 was not redistributed to the mitochondria in HeLa cells expressing GCC88-MAO, indicating it is not associated with these mature ETC which are recognized and tethered by GCC88 (Figure 2.S2C). The possibility that SNX proteins might directly coordinate CI-M6PR retrieval independent of retromer has been recently suggested [94, 95, 261]. Given this, we further examined the trafficking of CI-M6PR ETCs tethered by golgin-97 or golgin-245 in SNX KO cell lines. As indicated by the confocal microscopy and colocalization analysis, SNX1/2 dKO, SNX3 KO and SNX27 KO cells all showed comparable levels of colocalization between CI-M6PR and golgin-97-MAO, when compared to the HeLa control cells (HeLa: Rgolgin-97-MAO = 0.7330;

SNX1/2 dKO: Rgolgin-97-MAO = 0.7562; SNX3 KO: Rgolgin-97-MAO = 0.7431; SNX27 KO: Rgolgin-97-

MAO = 0.7323; Figure 2.5B and Figure 2.S4B). In contrast, SNX1/2 dKO cells demonstrated a decreased colocalization level between CI-M6PR and golgin-245-MAO, whereas SNX3

KO and SNX27 KO cells remained similar colocalization level with HeLa cells (HeLa: Rgolgin-

245-MAO = 0.4796; SNX1/2 dKO: Rgolgin-245-MAO = 0.2580; SNX3 KO: Rgolgin-245-MAO = 0.5747;

SNX27 KO: Rgolgin-245-MAO = 0.4660; Figure 2.5B and Figure 2.S4C), indicating that SNX-BAR proteins contribute to the trafficking of the subset of CI-M6PR ETCs that are captured by golgin-245.

We next examined the transport of CD-M6PR, a retromer-independent cargo, in the SNX KO cell lines, through the rerouting assay. As indicated by the co-localization analysis, SNX1/2 dKO, SNX3 KO and SNX27 KO cells all demonstrated considerable overlap of CD-

M6PR with GCC88-MAO that was similar to control cells (HeLa: RGCC88-MAO = 0.3285;

SNX1/2 dKO: RGCC88-MAO = 0.3383; SNX3 KO: RGCC88-MAO = 0.3530; SNX27 KO: RGCC88-MAO = 0.4306; Figure 2.5C and Figure 2.S5A). In addition, the redirection of CD-M6PR ETCs captured by golgin-97 or golgin-245 was not affected in SNX3 KO and SNX27 KO cells, as

55 well as in SNX1/2 dKO cells (HeLa: Rgolgin-97-MAO = 0.6899 and Rgolgin-245-MAO = 0.4899;

SNX1/2 dKO: Rgolgin-97-MAO = 0.6536 and Rgolgin-245-MAO = 0.4262; SNX3 KO: Rgolgin-97-MAO =

0.7145 and Rgolgin-245-MAO = 0.4928; SNX27 KO: Rgolgin-97-MAO = 0.6505 and Rgolgin-245-MAO = 0.4973; Figure 2.5C, Figure 2.S5B and S5C). Therefore, this data suggests that the absence of SNX3, SNX1/2 and SNX27 has no essential role on the protein trafficking of CD-M6PR within ETCs captured by GCC88, golgin-97 or golgin-245.

(* See figure legend on the next page.) 56 Figure 2.5 SNX3 is required for the retrograde transport of CI-M6PR GCC88-tethered ETCs.

(A) Equal amounts of cell lysates from HeLa, SNX1/2 dKO, SNX3 KO and SNX27 KO cells were subjected to SDS-PAGE and immunoblotted with antibodies against Vps35, Vps26A, SNX1, SNX2, SNX5, SNX6, SNX27, SNX3 and tubulin. (B and C) HeLa, SNX3 KO, SNX1/2 dKO and SNX27 cells were transiently transfected with HA-tagged mitochondria-targeting golgin constructs: GCC88-MAO, Golgin-97-MAO, Golgin-245-MAO or GM130-MAO, fixed and co-immunolabelled with antibodies against HA and endogenous CI-M6PR (B) or CD-M6PR (C), followed by Alexa Fluor conjugated secondary antibodies. The intensity plots of the fluorescent intensity (y-axis) against distance (x-axis) represent the overlap between channels. The co-localization of CI-M6PR (B) or CD-M6PR (C) with HA-tagged golgin-mito proteins was quantified by Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ****, p < 0.0001, ns, not significant.

57 2.4 Discussion

The endo-lysosomal system represents a dynamic set of intracellular compartments that plays a critical role in the protein trafficking and degradation of transmembrane and soluble cargoes within the cell. The maintenance of the endo-lysosomal homeostasis is achieved via the temporally and spatially coordination of multiple protein machineries. Enlarged lysosomal structures were observed in the absence of Vps35 which is consistent with previous reports in retromer-depleted Drosophila photoreceptor cells and larva fat body cells [257, 262], as well as mammalian cells [100]. We propose that the enlarged lysosomal structures observed in the absence of Vps35 is due to the improper trafficking of newly synthesized lysosomal enzymes by CI-M6PR, which decreases the lysosomal degradative activities and therefore results in an accumulation of undegraded materials within lysosomal compartments. It remains to be determined if the absence of retromer directly impacts upon the recently observed lysosome maturation cycle [263]. Changes of lysosomal functions upon depletion of Vps35 are in agreement with the roles of lysosome in the regulation of autophagy and other signalling cascades. The levels of LC3-II, an adaptor and cargo for the autophagosome, was greatly increased upon the depletion of Vps35. This is partially due to the decreased lysosomal proteolytic activity, but also reflected by the contribution of the interaction between retromer and TBC1D5 in the controlling of LC3 shuttling [129]. Lysosomal compartments, in contrast, appear to act as a central scaffold in the lysosomal recruitment of mTORC1 and the subsequent mTORC1 activation through various mechanisms [264]. Consistent with this concept, mTORC1 recruitment to lysosome in response to amino acid stimulation in Vps35 KO cells was greatly diminished, indicating that impaired lysosomal function impairs mTORC1 recruitment and activation.

Since being initially identified [100, 238], retromer has been considered to function in the coordination of the endosomal protein sorting and trafficking of a variety of cargo molecules. It is well established that the endosome-to-TGN retrieval of CI-M6PR is mediated by retromer. Previous studies demonstrate an impaired trafficking itinerary of CI-M6PR upon the depletion of Vps35 [85, 191, 239, 242] or Vps26A [85, 100, 238, 240] in mammalian cells, indicating a requirement for the retromer in the retrograde pathway. Our observations reconfirm the functional role of the retromer in the endosome-to-TGN retrieval of CI-M6PR. In cells with a complete knock-out of Vps35, we observed redistribution of CI-M6PR from the TGN to peripheral endosomal structures, causing defective lysosomal targeting and processing of M6P-dependent enzymes such as cathepsin-D and β-galactosidase. The lysosomal targeting of M6P-independent enzymes including acid 2 phosphatase and β-

58 glucocerebrosidase are not affected [265]. We found that the impaired CI-M6PR trafficking doesn’t affect the lysosomal targeting of β-hexosaminidase, which has also been shown as a M6P-dependent enzyme [265], suggesting that the trafficking of β-hexosaminidase could majorly rely upon CD-M6PR rather than CI-M6PR. This could be due to the different structure of extracellular regions of two MPRs and the distinct N-glycan (i.e. size, number of sialic acids, and general structures) of lysosomal enzymes [107, 266].

Intriguingly, we found that only a subset of CI-M6PR containing ETCs sorted from endosomes to the TGN depend upon retromer. Using the recently developed re-routing assay, which enables dissection of retrograde ETCs [61], we showed that the absence of Vps35 only impairs the transport of CI-M6PR ETCs that are captured by the trans-Golgi- anchored tethering factor GCC88, rather than other tethering factors such as golgin-97 or golgin-245. Our observations suggest a more limited but defined role of retromer in the CI- M6PR retrograde transport to the TGN, in support of the hypothesis that multiple types of machinery are involved in this retrograde trafficking [44, 48, 50, 54]. This is consistent with the observations that Vps26B-containing retromer does not co-precipitate CI-M6PR [91] which indicates that not all retromer complexes engage with the CI-M6PR. Likewise, a range of other endosome associated proteins, including many associated with retromer, have been implicated in CI-M6PR retrograde trafficking. We observed that the CI-M6PR protein levels at the steady state were unchanged in the absence of Vps35 (data not shown), consistent with other studies using the Vps35-depleted cells [239, 242], indicating that retromer independent retrograde trafficking of CI-M6PR still occurs. Recently, retromer having any role in the CI-M6PR retrograde has been questioned, based on observations that CI-M6PR was not re-distributed in retromer KO cells lines [94, 95] and that its trafficking was only dependent on SNX-BAR proteins. With respect to retromer, the observations in these manuscripts directly conflict with the data presented in this thesis. Likewise, these studies conflict with their own earlier work in which efficient RNAi depletion identified that SNX5 and SNX6 alter CI-M6PR trafficking in the same way that suppression of known retromer subunits did [85]. This difference with what we presume involves a consistent methodology was not discussed in their recent publications [94, 95]. One explanation for the failure to observe a difference in CI-M6PR trafficking is that both manuscripts [94, 95] determined changes in CI-M6PR trafficking based on correlation/co-localisation relative to TGN46 only. Prior to these studies it was already well established that TGN46 is incorporated in the ETCs [61] and that its retrograde trafficking was dependent on retromer [267] and SNX-BAR proteins [85]. Therefore, both CI-M6PR and TGN46 would be expected to change cellular distribution in the absence of the retromer. Furthermore, these studies with respect to

59 retromer would be considered incomplete as neither of these studies includes the standard assay used for CI-M6PR trafficking which is the miss-trafficking of newly synthesized lysosomal hydrolases.

We also demonstrated that the retrograde transport of CI-M6PR through ETCs tethered by GCC88, was perturbed in the absence of SNX3. This observation is consistent with previous reports that RNAi knock-down of SNX3 alters CI-M6PR retrograde transport [122, 268]. SNX3 interacts with the retromer subunits Vps35 and Vps26 via multiple interfaces through its N-terminal tail and this molecular interaction enables the endosomal recruitment of retromer via the SNX3 PX domain [36, 89, 120, 269, 270]. Recent structural studies determined that at the interface between Vps26 and SNX3, there is a binding site for a canonical sorting signal, ØX(L/M) consensus motif (where Ø can be any hydrophobic amino acid) for cargo recognition by retromer [89, 111]. The recycling motif of DMT1-II, a retromer cargo that cycles between the endosome to the cell surface, has been shown to fit in this binding mode, demonstrating the role of SNX3-retromer interaction in cargo sorting and membrane recruitment [89]. CI-M6PR has a related sorting motif with the structural properties that would enable it to interact with SNX3-retromer [89, 101], which supports this as the molecular mechanism by which SNX3 contributes to retromer-dependent CI-M6PR retrograde transport.

Apart from SNX3, we also examined other cargo-binding retromer associated proteins including SNX-BAR and SNX27. Intriguingly, our data demonstrate that the double knock- out of SNX-BAR proteins, SNX1 and SNX2, which results in the endosomal dissociation of SNX5 and SNX6, has no impact on the intracellular trafficking of GCC88-tethered CI-M6PR vesicles, suggesting that the SNX-BAR dimer is not required for this endosomal sorting process. We, however, did observe that the trafficking of golgin-245-tethered CI-M6PR vesicles was affected in SNX1/2 dKO cells, raising the possibility that the SNX-BAR dimer may function independently of retromer in the retrograde transport of CI-M6PR vesicles. These observations are consistent with previous reports demonstrating that the deficiency of SNX-BAR dimer alters CI-M6PR retrieval [85, 136, 245, 246], and is supported by recent studies suggesting the direct interaction between SNX5 and CI-M6PR [94, 95, 261]. It has been identified that golgin-97 and golgin-245 capture ETCs that are decorated with the WASH complex through a vesicle-golgin adaptor protein TBC1D23 [65]. Given the established association between the WASH complex and retromer [271-273], it is difficult to reconcile why the knock-out of retromer does not disrupt the trafficking of CI-M6PR vesicles captured by golgin-97 and golgin-245. However, as proposed by Shin et al (2017) its

60 plausible that redundant tethering mechanism is acting in parallel which may not be dissected by our experimental approaches [65]. Recent reported proteomics data, indicates that SNX27, another retromer accessory protein which functions in the recycling of cargo from endosomes to the plasma membrane, also associates with the CI-M6PR [94]. However, we show that SNX27 are not required for the sorting and generation of retromer-mediated CI-M6PR ETCs. Hence, given that retromer function is restricted to CI-M6PR incorporation into ETCs tethered by GCC88, and that other endosomal proteins clearly also modulate CI- M6PR retrograde trafficking independent of the retromer, it is critical in future studies to determine how CI-M6PR is temporally and spatially organized on endosomes relative to these various proteins associated with distinct retrograde transport pathways.

Overall, we demonstrate the requirement of retromer in the retrograde trafficking of a subset of CI-M6PR-containing ETCs, which are defined by their capability to be tethered by the trans-Golgin, GCC88. The sorting and generation of retromer-dependent CI-M6PR ETCs requires SNX3, which is consistent with the structural features of SNX3-retromer complex in cargo signal recognition and membrane recruitment [89]. In conclusion, we have identified that retromer is involved in the sorting and generation of retrograde ETCs and this pathway is critical for the efficient targeting of lysosomal enzymes, which contributes to the maintenance of lysosomal function.

61

Figure 2.S1. Full scan of western blots in Figure 2.3A.

(A) Immunoblots of indicated proteins in medium samples and corresponding cell lysates collected from HeLa and Vps35 KO cells pulsed with cycloheximide (100 μg/ml) in serum-free medium with 2 mM glutamine at 0, 3, 6 and 9 h post-chase. Tubulin was utilized as the loading control. Expression levels of cathepsin-D precursor within the media of HeLa and Vps35 KO cells were quantified and represented in graphs (means ± SEM). Two-tailed Student’s t test was used to determine the statistical significance (n = 3). *, p < 0.05, **, p < 0.01, ****, p < 0.0001. A.U, arbitrary units. (B) Expression levels of cathepsin-D precursor within lysates of HeLa and Vps35 KO cells collected at 0 h were quantified and represented in the graph (means ± SEM). Two-tailed Student’s t test was used to determine the statistical significance (n = 3). **, p < 0.01. A.U, arbitrary units. (C) Ratio of secreted cathepsin-D precursor to the basal amount of precursor in the cell lysates of HeLa and Vps35 KO cells was quantified and represented in the graph (means ± SEM). Two-tailed Student’s t test was used to determine the statistical significance (n = 3). *, p < 0.05.

62

(* See figure legend on the next page.)

63 Figure 2.S2. Additional data for Figure 4 and Figure 5.

(A) HeLa cells were transiently transfected with individual HA-tagged mitochondria-targeting golgin constructs: GCC88-MAO, Golgin-97-MAO, Golgin-245-MAO and GM130-MAO. Cells were fixed, co- immunolabelled with antibodies against HA and endogenous Vps35, followed by Alexa Fluor conjugated fluorescence secondary antibodies. Scale bars, 10 μm. The co-localization between Vps35 and HA-tagged golgin-mito proteins was quantified by Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). (B) HeLa and Vps26A KO cells were transiently transfected with individual HA-tagged mitochondria-targeting golgin constructs: GCC88-MAO and GM130-MAO. Cells were fixed, co- immunolabelled with antibodies against HA and endogenous CI-M6PR, followed by Alexa Fluor conjugated fluorescence secondary antibodies. Scale bars, 10 μm. The intensity plots of the fluorescent intensity (y-axis) against distance (x-axis) represent the overlap between channels. The co-localization between CI-M6PR and HA-tagged golgin-mito proteins was quantified by Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ****, p < 0.0001, ns, not significant. (C) HeLa cells were transiently transfected with individual HA-tagged mitochondria-targeting golgin constructs: GCC88-MAO, Golgin-97-MAO, Golgin-245-MAO and GM130-MAO. Cells were fixed, co-immunolabelled with antibodies against HA and endogenous SNX3, followed by Alexa Fluor conjugated fluorescence secondary antibodies. Scale bars, 10 μm. The co-localization between SNX3 and HA-tagged golgin- mito proteins was quantified by Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3).

64

Figure 2.S3. Absence of retromer has no impact on the trafficking of CD-M6PR ETCs captured by golgin-97 or golgin-245.

HeLa and Vps35 KO cells were transiently transfected with individual HA-tagged mitochondria- targeting golgin constructs: Golgin-97-MAO, Golgin-245-MAO and GM130-MAO. Cells were fixed, co-immunolabelled with antibodies against HA and endogenous CD-M6PR, followed by Alexa Fluor conjugated fluorescence secondary antibodies. Scale bars, 10 μm. The intensity plots of the fluorescent intensity (y-axis) against distance (x-axis) represent the overlap between channels.

65

Figure 2.S4. The retrograde transport of CI-M6PR ETCs captured by GCC88 and golign-245 requires the coordination of SNX3 and SNX-BAR proteins, respectively.

(A - D) HeLa, SNX1/2 dKO, SNX3 KO and SNX27 KO cells were transiently transfected with individual HA-tagged mitochondria-targeting golgin constructs: GCC88-MAO (A), Golgin-97-MAO (B), Golgin-245-MAO (C) and GM130-MAO (D). Cells were fixed, co-immunolabelled with antibodies against HA and endogenous CI-M6PR, followed by Alexa Fluor conjugated fluorescence secondary antibodies. Scale bars, 10 μm. The intensity plots of the fluorescent intensity (y-axis) against distance (x-axis) represent the overlap between channels.

66

Figure 2.S5. The retrograde transport of CD-M6PR ETCs captured by GCC88, golgin- 97 or golgin-245 is not affected by the absence of SNX-BAR, SNX3 and SNX27 proteins.

(A - D) HeLa, SNX1/2 dKO, SNX3 KO and SNX27 KO cells were transiently transfected with individual HA-tagged mitochondria-targeting golgin constructs: GCC88-MAO (A), Golgin-97-MAO (B), Golgin-245-MAO (C) and GM130-MAO (D). Cells were fixed, co-immunolabelled with antibodies against HA and endogenous CD-M6PR, followed by Alexa Fluor conjugated fluorescence secondary antibodies. Scale bars, 10 μm. The intensity plots of the fluorescent intensity (y-axis) against distance (x-axis) represent the overlap between channels.

67 Chapter 3. A role of GCC88 in the retrograde transport of CI-M6PR and the maintenance of lysosomal activity

3.1 Introduction

The trans-Golgi network (TGN) is a dynamic compartment, which not only sorts secretory proteins into distinct transport carriers for export to endosomes or to the plasma membrane but also coordinates the retrograde traffic from endocytic pathways. The fidelity of cargo transports in the TGN export/import routes is crucial for the maintenance of key cellular functions. The organization and function of the TGN are achieved by the coordination of multiple protein regulators anchored at the trans-Golgi. Membrane tethering factors, such as multi-subunit protein complexes and long coiled-coil Golgin proteins, are essential parts to receive incoming cargo-loaded transport carriers originating from the early endosome via the retrograde pathways [248, 274].

GCC88 is a peripheral membrane Golgi protein comprising a coiled-coil region, a C-terminal conserved GRIP domain, and an N-terminal vesicle binding region. GCC88 is recruited to specific trans-Golgi subdomains via the interaction between the GRIP domain and trans- Golgi-localized GTPases and captures endosomal transport carriers (ETCs) by the vesicle binding region at the N-terminus [60-62, 73]. The cation-independent mannose-6-phosphate receptor (CI-M6PR), a type-I transmembrane protein that is responsible for the lysosomal delivery of newly synthesized hydrolases is an established cargo protein incorporated into GCC88-captured ETCs [60, 61, 275]. CI-M6PR recognizes hydrolases containing the mannose-6-phosphate (M6P) sorting signal at the TGN and transports them to the early endosomes. Once arriving at the early endosome, the hydrolases are released and remained within the endosome as they mature into late endosomes and lysosomes. Unoccupied CI-M6PR is incorporated into ETCs for retrograde transport to the TGN enabling reuse of the receptor.

One of the major protein regulators responsible for the endosome-to-TGN retrograde transport of CI-M6PR is retromer, which is a peripheral membrane protein complex composed of Vps35, Vps29, and Vps26 subunits [87, 88, 91, 92, 234]. Retromer serves a vital role in the retrograde sorting of CI-M6PR into a specific subset of endosomal transport carriers (ETCs), and retromer dysregulation causes CI-M6PR mistrafficking and impairs

68 lysosomal proteolysis and autophagy pathways [85, 100, 101, 191, 194, 276]. The data in Chapter 2 showed that retromer-associated CI-M6PR vesicles are selectively tethered by GCC88, but not other tethering factors, such as golgin-97 and golgin-245/p230, suggesting the specific association of GCC88 and retromer-mediated retrograde trafficking pathways.

In this study, we reveal that GCC88 is required for the maintenance of the TGN structure and the efficient endosome-to-TGN trafficking of CI-M6PR. Knock-out of GCC88 partially phenocopies a loss of retromer, which alters CI-M6PR retrieval, impairs lysosomal enzymes processing and causes aberrant lysosomal proteolysis. However, unlike the depletion of retromer, the absence of GCC88 doesn’t affect the formation of CI-M6PR ETCs and the autophagy pathways.

69 3.2 Materials and methods

3.2.1 DNA constructs, antibodies, and chemicals

The pCMU-CD8a-CI-M6PR construct was described previously [249]. Mitochondria targeting golgins including trans-golgin: GCC88ΔC-term-hemagglutinin (HA)-monoamine oxidase A (MAO) and cis-golgin: GM130ΔC-term-HA-MAO were obtained from S. Munro [61]. The GCC88 CRISPR guide RNA (gRNA) plasmid (gRNA targeting sequence: CAACTGGCCTCTTCGGACTT) was synthesized by Genscript, USA.

Mouse monoclonal anti-CI-M6PR (clone 2G11; ab2733) and rabbit monoclonal anti-CI- M6PR (clone EPR6599; ab124767) were purchased from Abcam. Mouse monoclonal anti- CD-M6PR (22d4) was purchased from the Development Studies Hybridoma Bank. Rabbit monoclonal anti-golgin-97 (clone D8P2K; 13192), rabbit monoclonal anti-Rab5 (clone C8B1; 3547), rabbit monoclonal anti-LC3B/LC3-II (clone D11; 3868) and rabbit monoclonal anti- mTOR (clone 7C10; 2983) were purchased from Cell Signaling Technology. Mouse monoclonal anti-CD8a (OKT8; 14-0086) was purchased from eBioscience. Mouse monoclonal anti-EEA1 (clone 14/EEA1; 610457), mouse monoclonal anti-p230 (clone 15/p230; 611280) and mouse monoclonal anti-LAMP1 (clone H4A3; 555798) were purchased from BD Biosciences. Mouse monoclonal anti-transferrin receptor (clone H68.4; 13-6800) was purchased from Life Technologies. Goat polyclonal anti-cathepsin-D (AF1014) was purchased from R&D systems. Mouse monoclonal anti-alpha-tubulin (clone DM1A; T9026) was purchased from Sigma-Aldrich. Secondary donkey anti-mouse IgG Alexa FluorTM 488 (A21202), donkey anti-mouse IgG Alexa Fluor TM 555 (A31570), donkey anti- rabbit IgG Alexa FluorTM 488 (A21026) and donkey anti-rabbit IgG Alexa FluorTM 555 (A31572) were purchased from Thermo Fisher Scientific.

Magic Red Cathepsin B Kit (938) was from ImmunoChemistry Technologies. DQ™ Red BSA (D12051) was purchased from Thermo Fisher Scientific. Cycloheximide (66-81-9) and AZD8055 (16978) were purchased from Cayman chemical. Trichloroacetic acid (TCA; T6399) was obtained from Sigma-Aldrich.

3.2.2 Cell culture and Transfection

HeLa cells (ATCC CCL-2) were cultured and transfected as described in Chapter 2.

70 3.2.3 Generation of CRISPR/Cas9 knock-out GCC88 cell line

CRISPR/Cas9 knock-out cell lines were generated as described in Chapter 2. These cell lines were authenticated and confirmed to by mycoplasma free (CellBank Australia). Briefly, HeLa cells stably expressing mCherry-Cas9 were transiently transfected with the CRISPR gRNA PX552 plasmids [277] targeting to GCC88 which also will co-express GFP in a 60 mm dish. Forty-eight hours post-transfection, cells were collected for fluorescence-activated cell sorting (FACS). GFP positive cells were collected and allowed to grow until confluent. Cells were then diluted as one cell per well into 96-well plates until single colonies formed. CRISPR/Cas9 mediated knock-out of targeted genes was confirmed by indirect immunofluorescence assay and immunoblot analysis.

3.2.4 Cell Treatment Procedures

For starvation and lysosomal degradation inhibition experiments, cells were treated with Hank’s Balanced Salt Solution (HBSS, Thermo Fisher Scientific) with or without bafilomycin A1 (final concentration: 200 nM; 88899-55-2, Cayman Chemical) for 6 hrs.

The mTORC1 activation assay, antibody internalization assay, DQ™ Red BSA Assay, and Magic red cathepsin-B assay were performed as described in Chapter 2.

3.2.5 Secretion Assay

The protein secretion assay was performed as described in Chapter 2.

3.2.6 SDS-PAGE and western immunoblotting

SDS-PAGE and western blotting was performed as described in Chapter 2.

3.2.7 Indirect Immunofluorescence and colocalization analysis

The indirect immunofluorescence was performed as described in Chapter 2.

Images were processed using ImageJ software. Colocalization analysis was performed using the ImageJ JACoP plugin [255]. Colocalization analysis was conducted on 3 independent experiments. Colocalization values were exported to GraphPad Prism 7 software and tabulated accordingly.

71 3.2.8 Statistics

All statistical analyses were completed using GraphPad Prism software 7 and described in the appropriate figure legends. Error bars on graphs were represented as the standard error of the mean (± SEM). P values were calculated using two-tailed Student's t test. P < 0.05 was considered as significant.

72 3.3 Results

3.3.1 GCC88 is required for the maintenance of the TGN structure

To dissect the role of GCC88 in the endosomal system, the clonal GCC88 KO cells were generated using CRISPR/Cas9-medicated gene editing technology (Figure 3.1A). As indicated by immunoblotting, the absence of GCC88 has no effect on the level of the other two TGN-anchored GRIP domain-containing trans-Golgi proteins – golgin-97 and golgin- 245/p230 (Figure 3.1A), consistent with previous study [249]. The distribution of these two TGN markers was next examined in HeLa and GCC88 KO cells, to determine if the absence of GCC88 affects the TGN morphology. As revealed by immunofluorescence and the quantification analysis, around 40% of GCC88 KO cells showed elongated TGN structure positive with both golgin-97 and golgin-245/p230, whereas only 8% of HeLa control cells displayed the dispersed TGN staining pattern (Figure 3.1B). These observations indicate that depletion of GCC88 didn’t affect the intracellular level of the other two TGN Golgin tethers, but altered the morphology of the TGN.

(* See figure legend on the next page.)

73 Figure 3.1. GCC88 is required for the maintenance of the TGN structure.

(A) Generation of CRISPR/Cas9 mediated GCC88 KO HeLa cells. Equal amounts of cell lysates from HeLa and GCC88 KO cells were subjected to SDS-PAGE and immunoblotted with antibodies against GCC88, golgin-97, golgin-245/p230, and tubulin. (B) The distribution of golgin-97 and golgin- 245/p230 in HeLa and GCC88 KO cells was determined by the indirect immunofluorescence using antibodies against endogenous golgin-97 and golgin-245/p230, followed by Alexa Fluor-conjugated fluorescent secondary antibodies. Scale bars, 10 μm. The proportion of cells with elongated TGN was represented in the graph (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ****, p < 0.0001.

3.3.2 GCC88 deficiency impairs the retrograde trafficking of CI-M6PR

Chapter 2 has demonstrated that GCC88 is capable to selectively capture a subset of CI- M6PR transport vesicles mediated by retromer. Therefore, the absence of GCC88 is likely to perturb the retrograde trafficking of this subset of CI-M6PR vesicles. To determine this, we performed a series of colocalization experiments using TGN and endosomal markers. HeLa and GCC88 KO cells were co-immunolabelled with antibodies against endogenous CI-M6PR together with individual organelle markers. Consistent with observations in Chapter 2, in HeLa cells, CI-M6PR was distributed predominantly to the perinuclear region, showing high co-localization with the TGN marker – golgin-245/p230, the early endosome marker - EEA1 and the recycling marker - TfR, but low or minor co-localization with the late endosome / lysosome marker - LAMP1 (Figure 3.2A). However, its distribution was clearly altered in GCC88 KO cells, in which CI-M6PR displayed dispersed cytoplasmic staining patterns. Initially, calculation of the Pearson’s coefficient was used to compare the overlap in fluorescent signal, GCC88 KO cells demonstrated comparable overlap with EEA1, TfR and LAMP1, but a reduced overlap of CI-M6PR with golgin-245/p230, when compared to

HeLa control (golgin-245/p230: RHeLa = 0.401, RGCC88 KO = 0.320; EEA1: RHeLa = 0.422,

RGCC88 KO = 0.401; TfR: RHeLa = 0.423, RGCC88 KO = 0.446; LAMP1: RHeLa = 0.241, RGCC88 KO = 0.265; Figure 3.2A), suggesting impaired CI-M6PR retrieval in the absence of GCC88. Consistently, the analysis using the Manders’ coefficient indicated a reduced overlap of CI- M6PR with golgin-245/p230 in GCC88 KO cells, which is approximately half of that in HeLa cells (golgin-245/p230: RHeLa = 0.661, RGCC88 KO = 0.333). We found that the distribution of TfR-labeled recycling endosomes and LAMP1-labeled late endosomes / lysosomes changed mildly in GCC88 KO cells, which could be caused by a general re-organisation of the intracellular organelles like that observed for the TGN. Reduced level of CI-M6PR at the whole cell level was observed in GCC88 KO cells. Indeed, immunoblotting indicated that CI- M6PR was reduced in GCC88 KO cells by approximately 30%, when comparing to HeLa cells (Figure 3.2B). The incorporation of the other mannose-6-phosphate receptor — cation-

74 dependent mannose-6-phosphate receptor (CD-M6PR) in GCC88-capture transport vesicles have been reported. To determine whether the absence of GCC88 causes similar defects in CD-M6PR trafficking, we examined the intracellular distribution of CD-M6PR using TGN and endosomal markers. HeLa and GCC88 KO cells were co-immunolabelled with antibodies against endogenous CD-M6PR together with the TGN marker — golgin-97 or the early endosome marker — Rab5. Confocal microscopy revealed high colocalization of CD-M6PR with the TGN labelled with golgin-97 and lower colocalization with the endosomal compartments labelled with Rab5 in HeLa cells (Figure 3.1S). Unlike CI-M6PR, GCC88 KO cells showed unchanged CD-M6PR distribution, when compared to HeLa cells (Figure 3.1S). The Pearson’s coefficient analysis further confirmed this observation demonstrating comparable co-localization of CD-M6PR with golgin-97 or Rab5 in HeLa and

GCC88 KO cells (golgin-97: RHeLa = 0.418, RGCC88 KO = 0.425; Rab5: RHeLa = 0.318, RGCC88

KO = 0.278; Figure 3.1S). In support, the Manders’ coefficient showed similar co-localization level of CD-M6PR with individual molecular markers in HeLa and GCC88 KO cells (golgin-

97: RHeLa = 0.610, RGCC88 KO = 0.602; Rab5: RHeLa = 0.348, RGCC88 KO = 0.364; Figure 3.1S). Therefore, these data suggest that the absence of GCC88 alters the retrograde trafficking of CI-M6PR, but with minor or no effect to CD-M6PR trafficking.

To further confirm the disrupted CI-M6PR trafficking in GCC88 KO cells, we employed the CD8α internalization assay to examine the intracellular trafficking of CI-M6PR. HeLa and GCC88 KO cells were transiently transfected with the CD8α-CI-M6PR construct, incubated with antibodies against CD8α on ice for 30 min, washed twice with acidic solutions, chased at 37 °C for up to 60 min and immunolabelled with antibodies against golgin-97, a TGN marker, or Rab5, an early endosome marker. Confocal microscopy revealed that CD8α- bounded CI-M6PR was internalized via the endocytic pathway and delivered to the TGN at 60 min post-chase in HeLa cells, suggesting the efficient endosome-to-TGN retrograde trafficking of CI-M6PR (Figure 3.2C). In contrast, more of CD8α-CI-M6PR was observed to be retained at endosomal structures and less was delivered to the TGN in GCC88 KO cells (Figure 3.2C), suggesting the requirement of GCC88 for the efficient retrograde transport of CI-M6PR. To investigate whether the absence of GCC88 affects the formation of its associated ETCs, we next employed the redirecting assay to examine the trafficking of CI- M6PR vesicles [61]. HeLa and GCC88 KO cells were transiently transfected with HA-tagged mitochondria targeting construct, GCC88-MAO or GM130-MAO. Transfected cells were fixed and co-immunolabelled with antibodies against HA and endogenous CI-M6PR. Confocal microscopy revealed that CI-M6PR displayed a filamentous, mitochondria staining pattern in HeLa cells expressing GCC88-MAO, whereas little redistribution of CI-M6PR was

75 observed in HeLa cells expressing the cis-Golgi protein, GM130-MAO (Figure 3.2D), consistent with previous studies [61]. The relocation of CI-M6PR vesicles that were captured by GCC88 was not affected in GCC88 KO cells. In fact, the co-localization analysis revealed a high level of overlap of CI-M6PR with GCC88-MAO in GCC88 KO cells, which was similar to HeLa cells (GCC88-MAO: RHeLa = 0.407, RGCC88 KO = 0.492; Figure 3.2D). Therefore, the absence of GCC88 does not prevent the formation of these ETCs.

(* See figure legend on the next page.) 76 Figure 3.2. GCC88 deficiency affects the retrograde transport of CI-M6PR.

(A) The distribution of CI-M6PR in HeLa and GCC88 KO cells was determined by the indirect immunofluorescence using antibodies against endogenous CI-M6PR, golgin-245/p230, EEA1, TfR and LAMP1, followed by Alexa Fluor-conjugated fluorescent secondary antibodies. Scale bars, 10 μm. The colocalization between CI-M6PR and different organelle markers was quantified by the Pearson’s correlation coefficient or Manders’ correlation coefficient and represented in graphs (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ****, p < 0.0001, **p < 0.01, ns, not significant. (B) Equal amounts of cell lysates from HeLa and GCC88 KO cells were subjected to SDS-PAGE and immunoblotted with antibodies against CI-M6PR and tubulin. The level of endogenous CI-M6PR in HeLa and GCC88 KO cells was quantified and represented in the graph (means ± SEM). Two-tailed Student’s t test was used to determine the statistical significance (n = 3). *, p < 0.05. (C) The internalization assay was performed in HeLa and GCC88 KO cells transiently transfected with the CD8a-CI-M6PR construct, using antibodies against CD8a. The delivery of CD8a-CI-M6PR complex to the TGN and early endosome after uptake for 60 min at 37 °C was determined by the indirect immunofluorescence using antibodies against golgin- 97 and Rab5, followed by Alexa Fluor-conjugated fluorescent secondary antibodies. Scale bars, 10 μm. The colocalization between CD8a-CI-M6PR and indicated molecular markers was quantified by the Pearson’s coefficient and represented in graphs (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). *, p < 0.05, ****, p < 0.0001. (D) HeLa and GCC88 KO cells were transiently transfected with the GCC88-MAO or GM130-MAO construct, fixed and co-immunolabelled with antibodies against HA and endogenous CI-M6PR, followed by Alexa Fluor-conjugated fluorescent secondary antibodies. Scale bars, 10 μm. The intensity plots of the fluorescent intensity (y-axis) against distance (x-axis) represent the overlap between channels. The co-localization between CI-M6PR and HA-tagged golgin-mito proteins was quantified by Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ns, not significant.

3.3.3 GCC88 deficiency affects the lysosomal proteolytic capacity

CI-M6PR is responsible for the lysosomal delivery of its associated newly synthesized hydrolases. Therefore, the processing of CI-M6PR-dependent hydrolases is potentially perturbed, as a consequence of CI-M6PR trafficking defect caused by GCC88 deficiency. To examine this, we employed cycloheximide chase experiments to investigate the processing and secretion of newly synthesized cathepsin-D, a well-characterized hydrolase relying on CI-M6PR for lysosomal delivery. HeLa and GCC88 KO cells were treated with cycloheximide, and medium samples were collected at 0, 3, 6 and 9 h post-chase. As indicated by immunoblotting, the secreted cathepsin-D precursor was detected at 3 h post- chase in HeLa cells (Figure 3.3A). In comparison, GCC88 KO cells demonstrated a higher level of the cathepsin-D precursor at 3 h post-chase with an increase trend observed at 6 and 9 h post-chase (Figure 3.3A). Immunoblotting of cell lysates to quantify tubulin levels confirmed that the secreted material was collected from equivalent numbers of cells (Figure 3.3A). Collectively these data indicate that GCC88 deficiency perturbs the intracellular processing of the cathepsin-D precursor, leading to its increased secretion rather than delivery to endosomes.

77

Figure 3.3. GCC88 is required for the maintenance of lysosomal proteolytic activity.

(A) HeLa and GCC88 KO cells were pulsed with 100 μg/ml of cycloheximide in serum-free medium containing 2 mM glutamine for 0, 3, 6 and 9 h. Medium and cell lysate samples were collected for immunoblotting with indicated proteins. The expression level of cathepsin-D precursor within the medium was quantified and represented in graphs (means ± SEM). All data was normalized to tubulin levels. Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). **p < 0.01, ns, not significant. (B) HeLa and GCC88 KO cells were treated with Magic Red Cathepsin- B in complete medium at 37 °C and imaged by a time-lapsed video microscopy performed with the inverted Nikon Ti-E microscopy with Hamamatsu Flash 4.0 sCMOS camera. The graph represents the Magic Red Cathepsin B fluorescent intensity (A.U, arbitrary units) within HeLa and GCC88 KO cells at the indicated time (means ± SEM). N = two independent experiments with eight random points each. **, p < 0.01, ****, p < 0.0001, ns, not significant. (C) HeLa and GCC88 KO cells were treated with DQ-BSA Red in complete medium at 37°C for overnight. Cells were fixed and immunolabelled with antibodies against LAMP1, followed by Alexa Fluor-conjugated fluorescent secondary antibodies. Scale bars, 10 μm. The graph represents the fluorescence intensity of DQ- BSA Red (A.U, arbitrary units) within HeLa and GCC88 KO cells (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). *, p < 0.05.

78 The impaired enzyme processing is often associated with an alteration of the lysosomal proteolytic activity. To determine whether GCC88 deficiency affects the lysosomal activity, we monitored the changes in hydrolytic activity within intact cells using the Magic Red cathepsin-B assay. HeLa and GCC88 KO cells were treated with Magic Red cathepsin-B, a cell-permeant fluorogenic substrate, which becomes fluorescent when cleaved by cathepsin enzymes in lysosomes. As indicated by live cell time-lapse imaging and fluorescent intensity analysis, Magic Red fluorescence in HeLa cells increased rapidly during the first 30 min, then maintained at relatively stable levels after that. In contrast, fluorescence intensities for Magic Red in GCC88 KO cells maintained at relatively stable levels, which were notably lower than HeLa cells, suggesting reduced cathepsin enzyme activity in the absence of GCC88 (Figure 3.3B). We next compared the intracellular proteolytic kinetics using DQ-BSA, a self-quenched dye conjugated with BSA, which enters into the endosomal system through endocytosis and generates a strong fluorescent signal upon proteolytic cleavage within lysosomal compartments. HeLa and GCC88 KO cells were treated with DQ-BSA overnight, fixed and immunolabelled with antibodies against the lysosome marker LAMP1. Confocal microscopy and fluorescent intensity analysis revealed strong DQ-BSA fluorescence signal in HeLa cells, which was largely overlap with LAMP1, indicating efficient proteolysis of DQ- BSA within lysosomes (Figure 3.3C). However, GCC88 demonstrated a significantly lower DQ-BSA fluorescence within LAMP1-positive lysosomes (Figure 3.3C), indicating impaired lysosomal proteolytic processes.

3.3.4 The autophagy-lysosomal pathway is not affected in GCC88 KO cells

Apart from mediating lysosomal proteolysis, another important function of lysosomes is to fuse with autophagosomes for the final stages of autophagy. During this step, the autophagosome marker LC3-II is degraded in the new autolysosomal environment. To determine if GCC88 deficiency affects autophagy-lysosomal pathways, the subcellular distribution of the membrane-bound LC3-II was examined. HeLa and GCC88 KO cells were co-immunolabelled with antibodies against endogenous LC3-II and LAMP1, a late endosome/lysosome marker. As revealed by the confocal microscopy and the colocalization assay, GCC88 KO cells showed comparable membrane-bound LC3-II, with an increased colocalization with LAMP1, when comparing to HeLa cells (Figure 3.4A), suggesting unaffected autophagy levels. To further confirm autophagy is unaffected in the absence of GCC88 KO, levels of LC3-II were examined in HBSS-treated starvation condition with or without bafilomycin A1, an autophagy inhibitor which blocks autophagy flux by inhibiting 79 autolysosome acidification and lysosome-autophagosome fusion. Cell lysates of treated HeLa and GCC88 KO cells were collected and subjected to immunoblotting. As shown in Figure 3.4B, upon HBSS-treated starvation, the level of LC3-II within GCC88 KO cells was comparable with that in HeLa cells (Figure 3.4B). Upon bafilomycin A1 treatment, LC3-II increased in both GCC88 KO cells and HeLa cells to similar levels (Figure 3.4B). These data indicate similar changes of the autophagy processes upon starvation or inhibition of autophagic degradation in the absence of GCC88, consistent with our previous observation. To determine if the induction of autophagy is affected by GCC88 deficiency, we next examined the intracellular activity of the mTORC1 signalling pathway, which is antagonistic to the induction of autophagy. HeLa and GCC88 KO cells were either starved of amino acids or stimulated with essential amino acids for the intracellular recruitment of mTORC1. As revealed by confocal microscopy, under the amino acid-starved condition, mTORC1 displayed a diffuse, cytoplasmic distribution in both HeLa and GCC88 KO cells (Figure 3.4C). Stimulation with essential amino acids promoted comparable levels of mTORC1 recruited from the cytosol to lysosomal compartments in HeLa and GCC88 KO cells, as demonstrated with colocalization with LAMP1 (Figure 3.4C), suggesting mTOR signalling in the absence of GCC88 is not affected.

80

Figure 3.4. The autophagy-lysosomal pathway is not affected by GCC88 deficiency

(A) HeLa and GCC88 KO cells were fixed, co-immunolabelled with antibodies against LC3-II and LAMP1, followed by Alexa Fluor-conjugated fluorescent secondary antibodies. Scale bars, 10 μm. The colocalization between LC3-II and LAMP1 was quantified by the Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ns, not significant. (B) HeLa and GCC88 KO cells starved with HBSS were treated with or without bafilomycin A1. Cells were harvested, and equal amounts of protein samples were used for SDS-PAGE and immunoblotting with antibodies against LC3-II and tubulin. (C) Amino acid starved HeLa and GCC88 KO cells were treated with 2x essential amino acid (AA) solution for 30 min, fixed with ice-cold methanol and co-immunolabelled with antibodies against mTORC1 and LAMP1, followed by Alexa Fluor-conjugated fluorescent secondary antibodies (means ± SEM). Scale bars, 10 μm. The colocalization of mTORC1 with LAMP1 was quantified by Pearson’s correlation coefficient. Two-tailed Student’s t test indicates the difference between HeLa and GCC88 KO cells upon amino acid stimulation (n = 3). ****, p < 0.0001, ns, not significant.

81 3.4 Discussion

The precise delivery of cargo molecules through multiple trafficking pathways is critical for the maintenance of the endo-lysosomal homeostasis. Serving as the essential compartment in the retrograde transport pathway, the TGN captures the incoming retrograde vesicles via various TGN-anchored tethering factors. Among them, the GRIP domain-containing peripheral coiled-coil membrane proteins have been proposed to play a role in maintaining the TGN dynamics and defining the molecular characteristics of the TGN [248, 278]. Our work has revealed a role of GCC88 in keeping the normal morphology and function of TGN, which further contributes to the maintenance of the endo-lysosomal homeostasis. The complete knock-out of GCC88 doesn’t affect the cellular level of other TGN golgins, which is consistent with the previous report based on the RNAi-mediated suppression of GCC88 [275]. Using fluorescent microscopy we readily observed elongated TGN in approximately a half of the GCC88 depleted cells, indicating the perturbed TGN structure. These observations are consistent with a recent study that demonstrated that knock-down of GCC88 causes elongated Golgi ribbon at the ultrastructural level [279]. Furthermore, overexpression of GCC88 also causes a change in the TGN morphology [62, 73, 279]. Likewise, the silencing or overexpression of other TGN golgins has been implicated in the maintenance of TGN morphology [79, 275, 280]. Hence, it may indicate that the individual TGN golgins are incorporated into the dynamic organization of the Golgi network, and the alteration of golgin levels is likely to contribute to the morphological change of the TGN. Structural alteration of the Golgi network has been recognized as a pathological feature of various neurodegenerative diseases [281].

Previous studies demonstrate GCC88 serves as a tethering factor that captures the incoming retrograde transport vesicles by its N-terminus tethering motif [60, 61]. Chapter 2 has demonstrated a specific functional association of GCC88 with retromer, by showing the capability of GCC88 to selectively capture the subset of CI-M6PR vesicles sorted into the retromer-mediated retrograde transport pathway. Serving as a cargo sorting protein, retromer is required for the endosome-to-TGN transport of its associated cargo, and deficiency of retromer impairs the CI-M6PR retrieval by causing a redistribution of the receptor from the TGN to the peripheral endosomal structures [85, 100, 101, 238]. Therefore, functioning as the tethering factor in this trafficking pathway, GCC88 deficiency is likely to result in similar trafficking defects with that caused by retromer deficiency. Our observations confirm the requirement of GCC88 in CI-M6PR retrieval. In cells with a complete knock-out of GCC88, we observed reduced CI-M6PR distributed at the TGN, consistent with previous

82 studies based on the RNAi-mediated suppression [275]. Unlike being redistributed to the endosomal structures in retromer Vps35 depleted cells shown in Chapter 2, CI-M6PR is degraded rapidly in the absence of GCC88. Moreover, we observed that the CI-M6PR ETCs were still formed in cells absent of GCC88. Therefore, our work indicates that the subset of CI-M6PR ETCs mediated by retromer can still be formed in the absence of GCC88. CI- M6PR ETCs formed by distinct machinery can utilize distinct tethering mechanisms. In fact, another Golgin GCC185 has also been implicated in the tethering of a subset of CI-M6PR vesicles mediated by Rab9 GTPase [79, 275]. Cells transiently depleted of GCC185 also showed decreased CI-M6PR levels [79, 275]. The fate of any ETCs formed in GCC88 KO cells is unknown but they may undergo inefficient fusion with other membranes, associated with other tethering proteins or be degraded within the cytoplasm. The trafficking defect caused by GCC88 deficiency further leads to the defective processing of CI-M6PR- dependent lysosomal enzyme, cathepsin-D, and the impairment of the lysosomal proteolytic activity, which are similar to the phenotypes caused by retromer deficiency. Apart from the lysosomal proteolysis, another critical intracellular process is the lysosome-autophagosome fusion for the final stage of autophagy. Despite a recent study demonstrates that the overexpression of GCC88 causes increased autophagy and compromised mTOR signaling [279], we show that a complete knock-out of GCC88 has no effect on the autophagy and the mTOR signaling pathway, suggesting that the depletion of GCC88 does not contribute to the autophagy pathway directly or that these cells have adapted with respect to this critical cellular process.

Overall, combined with a previous analysis from the retromer knock-out study, our findings demonstrate a role of GCC88 in the regulation of CI-M6PR retrieval and the maintenance of lysosomal proteolytic activity. However, deficiency of GCC88 does not directly impact on the autophagy and mTOR signaling network.

83

Figure 3.S1. The retrograde trafficking of CD-M6PR is not altered in GCC88 KO cells.

The distribution of CD-M6PR in HeLa and GCC88 KO cells was determined by the indirect immunofluorescence using antibodies against the endogenous CI-M6PR, golgin-97, and Rab5, followed by Alexa Fluor-conjugated fluorescent secondary antibodies. Scale bars, 10 μm. The colocalization between CD-M6PR and organelle markers was quantified by the Pearson’s correlation coefficient or Manders’ correlation coefficient and represented in graphs (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ns, not significant.

84 Chapter 4. The Parkinson disease-linked Vps35 D620N variant disrupts WASH complex association and impairs the selective function of retromer in cargo sorting

4.1 Introduction

Retromer is a conserved peripheral membrane protein complex assembled from a core high- affinity heterotrimeric complex composed of Vps35, Vps29 and one of two Vps26 subunits, Vps26A or Vps26B [88, 91, 92, 99, 282]. Retromer coordinates endosomal processes spatially and temporally via interactions with its accessory proteins [235]. One functional role of retromer in the endosomal system is to orchestrate the sorting and trafficking of transmembrane receptors. One well-established retromer-associated cargo protein is the cation-independent mannose-6-phosphate receptor (CI-M6PR), a type-I transmembrane protein that shuffles between the endosome and the trans-Golgi network (TGN) for the delivery of newly-synthesized acid hydrolases. Depletion or silencing retromer subunit Vps35 or Vps26A affects CI-M6PR trafficking itinerary [85, 100, 101, 191, 192, 236-243]. Chapter 2 has suggested that the CI-M6PR-loaded endosome transport carriers (ETCs) sorted via retromer-mediated mechanism can be captured by a specific downstream tethering factor — GCC88, a coil-coil protein anchored at the TGN. In addition to retrograde trafficking, retromer, when in the association with VARP, SNX27, or the WASH complex, has been implicated in the endosome-to-plasma membrane recycling of cargoes, such as GLUT1 and G-protein coupled receptors (GPCRs) [113, 132, 143, 283]. Besides roles in the regulation of cargo trafficking, the involvement of retromer in the maintenance of lysosomal function has also been proposed [257]. Studies showed that loss of retromer subunit Vps35 reduces the hydrolytic activity within lysosomes, and affects the lysosomal degradative processes [257].

Impairments on the endo-lysosomal system caused by retromer deficiency have been proposed to contribute to the progression of Parkinson’s disease (PD), one most common neurological disease characterized by the loss of dopaminergic (DA) neurons, and the presence of insoluble, cytoplasmic inclusion termed Lewy bodies. Recent studies report a heterozygous point mutation (p.D620N) within the retromer subunit Vps35 in PD cases with familial aetiology, highlighting the genetic association of retromer with the manifestation of

85 the late-onset PD [170, 171]. Structure and biochemical studies demonstrate that Vps35 D620N folds correctly and possesses similar binding affinity with the other retromer subunits as the wild-type Vps35 [88, 193, 220]. However, given that the D620N point mutation is located at the surface of the convex face of an α-solenoid domain, the Vps35 D620N variant has been proposed to impact the interaction with uncovered effectors, thus affecting the interactome of retromer [220]. To date, multiple mechanisms have been postulated the effects of Vps35 D620N on cellular functions such as receptor trafficking, lysosomal degradation, and mitochondria dynamics (reviewed in Chapter 1), which are likely to contribute to the development of PD phenotypes. Recent studies using the Vps35 D620N knock-in (KI) mouse model report that a heterozygous or homozygous D620N mutation alters the dopaminergic system, causing progressive degeneration of DA neutrons, demonstrating the neuropathological consequences in PD development [200, 284, 285]. Nevertheless, the specific role of how Vps35 D620N induces these PD-linked defects is not fully uncovered yet.

In this chapter, I characterized the intracellular effect of Vps35 D620N variant in detail and confirm its partial loss-of-function by using stable Vps35 D620N rescue cell lines generated in Vps35 knock-out (KO) background. We show that Vps35 D620N-containing retromer selectively affect the endosomal association of SNX3 and the WASH complex subunit FAM21 by the disruption of the binding between Retromer and FAM21 of WASH complex. By rebuilding the 3D structure of single endosome, we show that the presence of Vps35 D620N alters the endosomal morphology, which is hypothesized to reflect an impaired endosomal deformation function. Although Vps35 D620N impairs the retrograde trafficking of retromer-mediated CI-M6PR-loaded ETCs that are tethered by GCC88, the lysosomal proteolysis and autophagy pathways are not affected.

86 4.2 Materials and methods

4.2.1 DNA constructs

Mitochondria targeting golgins including GCC88ΔC-term-hemagglutinin (HA)-monoamine oxidase A (MAO) and GM130ΔC-term-HA-MAO were obtained from S. Munro [61]. The untagged Vps35 wildtype construct, the GFP-tagged Vps35 wildtype construct, and the GFP-tagged Vps35 D620N construct were described in Chapter 2. To generate untagged Vps35 D620N plasmid, full length human Vps35 D620N was amplified from pEGFP-N1- Vps35 D620N construct as a template [193], using a 5’ primer CCCACCC GGTACC ATG CCT ACA ACA CAG CAG and a 3’ primer CCCACCC CTCGAG TTA AAG GAT GAG ACC TTC AT, and subcloned into pcDNA3.1 (+) vector using KpnI and Xhol multicloning sites. Resulted construct was validated by DNA sequencing.

4.2.2 Antibodies

Mouse monoclonal anti-CI-M6PR (clone 2G11; ab2733), rabbit monoclonal anti-CI-M6PR (clone EPR6599; ab124767), mouse monoclonal anti-SNX27 (clone 1C6; ab77799), rabbit polyclonal anti-SNX3 (ab56078), and rabbit polyclonal anti-Vps26A (ab23892) were purchased from Abcam. Goat polyclonal anti-Vps35 (NB100-1397) was purchased from Novus Biologicals. Rabbit polyclonal anti-VARP (A302-998A) was purchased from Bethyl Laboratories. Mouse monoclonal anti-HA (clone 16B12; 901513) was purchased from BioLegend. Rabbit monoclonal anti-HA (clone C29F4; 3724), rabbit monoclonal anti-Rab7 (clone D95F2; 9367), rabbit monoclonal anti-LC3B/LC3-II (clone D11; 3868) and rabbit monoclonal anti-mTOR (clone 7C10; 2983) were purchased from Cell Signaling Technology. Mouse monoclonal anti-SNX2 (clone 13/SNX2; 611308), mouse monoclonal anti-EEA1 (clone 14/EEA1; 610457), mouse monoclonal anti-p230 (clone 15/p230; 611280) and mouse monoclonal anti-LAMP1 (clone H4A3; 555798) were purchased from BD Biosciences. Mouse monoclonal anti-transferrin receptor (clone H68.4; 13-6800) was purchased from Life Technologies. Mouse monoclonal anti-alpha-tubulin (clone DM1A; T9026) was purchased from Sigma-Aldrich. Rabbit polyclonal anti-WASH complex subunit FAM21C (ABT79) was purchased from Millipore. Goat polyclonal anti-TBC1D5 (sc-99661) was purchased from Santa Cruz Biotechnology. Wheat germ agglutinin Alexa FluorTM 647 conjugate (W32466) was purchased from Thermo Fisher Scientific. Secondary donkey anti- mouse IgG Alexa FluorTM 488 (A21202), donkey anti-mouse IgG Alexa Fluor TM 555 (A31570), donkey anti-rabbit IgG Alexa FluorTM 488 (A21026), donkey anti-rabbit IgG Alexa

87 FluorTM 555 (A31572), donkey anti-rabbit IgG Alexa FluorTM 594 (A21207), and donkey anti- goat IgG IgG Alexa FluorTM 488 (A11055) were purchased from Thermo Fisher Scientific. Magic Red Cathepsin B Kit (938) was from ImmunoChemistry Technologies.

4.2.3 Cell culture and Transfection

HeLa cells were cultured and transfected as described in Chapter 2.

4.2.4 Cellular assay

The mTORC1 activation assay, DQ™ Red BSA assay, Magic red cathepsin-B assay were performed as described in Chapter 2.

4.2.5 Indirect Immunofluorescence

The indirect immunofluorescence was performed as described in Chapter 2.

For WGA staining, cells were incubated in culture medium containing 5 μg/ml WGA conjugates for 15 min at 37 °C, followed by three-times washes in PBS. Cells were then fixed, permeabilized, and labelled as described in Chapter 2. For dual-labelling of FAM21 and Vps26A, the Vps26A polyclonal antibody was conjugated with Alexa Fluor 488 using Zenon Rabbit IgG Labelling Kit (Z25302) according to the manufacturer’s instructions. Coverslips were mounted on glass microscope slides using the Fluorescent Mounting Medium (Dako; S3023) or the ProlongTM Diamond Antifade Mountant (Thermo Fisher Scientific; P36965). Images were taken at room temperature using the Leica DMi8 SP8 Inverted confocal equipped with 63x Plan Apochromatic objective or the Leica STED 3x Super Resolution microscope equipped with 93x Plan Apochromatic objective. For quantification, images were taken from multiple random positions for each sample.

4.2.6 Image processing

Images were processed using ImageJ/Fiji software. Co-localization analysis was performed using the ImageJ/Fiji JACoP plugin [255]. Transfected cells were segregated from fields of view containing both transfected and non-transfected cells by generating regions of interest (ROI). The selected ROI was cropped, split into separated channels and applied for threshold processing. Co-localization analysis was conducted on 3 independent experiments. Co-localization values were exported to GraphPad Prism 7 software and tabulated accordingly.

88 Single endosome analysis based on super-resolution images was performed with ImageJ/Fiji software. Endosome regions (>50 nm) were segmented by the WGA staining and converted to binary mask. FAM21 staining pattern or Vps26A staining was masked to binary and merged with the WGA-labelled endosome regions. Color threshold was performed to find the endosomes positive with FAM21 or Vps26A, followed with the ROI comparing assay to finally determine the exact region of endosomes positive with selected protein. The endosomes positive with both FAM21 and Vps26A were identified by re- performing the ROI comparing assay as described above. Single endosomes segmented were then utilized for co-localization assay, size measurement, and morphological descriptor measurement. 3D structure of single endosomes was rebuilt with Imaris.

4.2.7 Membrane fractionation

Membrane fractionation was performed as described previously [254]. Cells were collected and homogenized in the HES buffer containing phosphatase inhibitors and protease cocktails (Sigma-Aldrich). Cell lysates were subjected to spin at 500 g for 10 min, followed by 17,200 g for 20 min. Supernatants were then subjected to spin at 175,000 g for 75 min to pellet the fraction. The resulting supernatant is the cytosol fraction.

4.2.8 SDS-PAGE and western immunoblotting

SDS-PAGE and western immunoblotting was performed as previously described in Chapter 2.

4.2.9 Immunoprecipitation

Immunoprecipitation was performed as described previously [286]. Cell monolayers were washed with ice-cold PBS and lysed in lysis buffer containing protease inhibitor cocktail. Cell lysates were subjected to spin at 17,000 g. The resulting supernatants were collected and incubated with GFP-NanoTrap beads for 2 hr at 4°C. Immunoprecipitated proteins were washed with lysis buffer for three times.

4.2.10 Electron Microscopy

Electron microscopy was performed as described in Chapter 2.

89 4.2.11 Statistics

All statistical analyses were completed using GraphPad Prism software 7 and described in the appropriate figure legends. Error bars on graphs were represented as the standard error of the mean (± SEM). P values were calculated using the two-tailed Student's t-test. P < 0.05 was considered as significant.

90 4.3 Results

4.3.1 The endosomal association of retromer accessory proteins is affected in the presence of Vps35 D620N

Retromer core subcomplex functions as a critical endosomal scaffold to recruit its interacting partners to the endosomal membrane. The endosomal dissociation of accessory proteins has been shown to impact the retromer-dependent endosomal processes [98, 121, 123, 194, 287-289]. In order to examine the capability of Vps35 D620N variant in the endosomal recruitment of its interactors, we established rescue cell clones stably expressing the GFP- tagged Vps35 D620N or the untagged Vps35 D620N in the Vps35 KO background. As indicated by immunoblotting with Vps35 and Vps26A, Vps35 D620N rescue cell clones demonstrated comparable rescue efficiency of the retromer subunits to the wild-type Vps35 rescue cells as described in Chapter 2 (Figure 4.1A). To investigate the effect of Vps35 D620N variant in the endosomal recruitment of accessory proteins, membrane fractionation was employed to determine the relative proportion of retromer accessory proteins associated with membranes. HeLa, Vps35 KO, Vps35-GFP rescue and Vps35 D620N-GFP rescue cells were lysed in the sucrose-containing buffer, and lysates were subjected to differential ultracentrifugation to isolate the microsome and cytosol fractions. As expected, retromer accessory proteins including SNX2, SNX5, SNX3, the WASH complex subunit FAM21, TBC1D5, VARP and Rab7 were detected in both the microsome and cytosol fractions of HeLa and Vps35-GFP rescue cells. In comparison, Vps35 KO cells demonstrated comparable levels of SNX2, SNX5, and Rab7 in the microsome fraction (Figure 4.1B), suggesting unaffected endosomal recruitment of these accessory proteins upon the absence of retromer subunit Vps35, consistent with previous reports [94, 290, 291]. Unaffected recruitment was also detected in from Vps35 D620N-GFP rescue cells, suggesting that the presence of Vps35 D620N variant has no effect on the endosomal recruitment of SNX-BAR proteins and Rab7. Decreased SNX3 level was detected in the microsomes of Vps35 KO cells, which was slightly increased upon the expression of wild- type Vps35 or Vps35 D620N (Figure 4.1B). Immunofluorescence and colocalization analysis were performed to validate the phenotype indicated by the biochemical subcellular fractionation. As revealed by confocal microscopy, SNX3 displayed dispersed punctate distribution in HeLa control cells at the steady state and showed high level of co-localization with EEA1-positive early endosomes (Figure 4.1C). The intracellular distribution of SNX3 was altered in Vps35 KO cells, showing reduced endosomal staining but increased

91 cytoplasmic distribution (Figure 4.1C). Indeed, Pearson's and Manders’ coefficient both indicated a significant decrease in the overlapping level of SNX3 with EEA1 in Vps35 KO cells when compared to HeLa control cells, consistent with the previous observation. Vps35 rescue cells showed high co-localization of SNX3 with EEA1-labeled early endosomes, which is comparable with HeLa control cells, whereas Vps35 D620N rescue cells showed low level of co-localization that was similar to that in Vps35 KO cells (Pearson : RHeLa =

0.4450, RVps35 KO = 0.3248, RVps35 rescue = 0.3843, RVps35 D620N rescue = 0.2761; Manders: RHeLa

= 0.5569, RVps35 KO = 0.3912, RVps35 rescue = 0.5703, RVps35 D620N rescue = 0.3304; Figure 4.1C). Therefore, these data suggest that the Vps35 D620N variant is uncapable to rescue the endosomal dissociation of SNX3 caused by Vps35 depletion.

Apart from SNX3, decreased level of FAM21 was detected in the microsome fraction of Vps35 KO cells (Figure 4.1B), consistent with a role of retromer in the endosomal recruitment of FAM21 [272, 273]. Increased level of FAM21 was observed in the microsome fraction of Vps35 rescue cells (Figure 4.1B). Expression of the Vps35 D620N variant also rescued FAM21 level in the microsome fraction, however, at a lower rate than the wild-type Vps35 (Figure 4.1B). The intracellular distribution of FAM21 was next examined with the confocal microscopy and colocalization analysis. As indicated in Figure 4.1D, FAM21 highly overlapped with EEA1-labeled early endosome in HeLa control cells (Figure 4.1D). The depletion of Vps35 caused reduced intracellular fluorescence intensity for FAM21 and decreased co-localization of FAM21 with EEA1 (Figure 4.1D and Figure 4.S1), consistent with the membrane fractionation data. The dissociated FAM21 observed in Vps35 KO cells was recruited to endosomal membranes in Vps35 rescue cells, as well as Vps35 D620N rescue cells (Pearson : RHeLa = 0.5312, RVps35 KO = 0.363, RVps35 rescue = 0.4499, RVps35 D620N rescue = 0.4569; Manders: RHeLa = 0.4849, RVps35 KO = 0.2122, RVps35 rescue = 0.4163, RVps35

D620N rescue = 0.5361; Figure 4.1D and Figure 4.S1). These data indicate that the Vps35 D620N variant still possess the capability to recruit FAM21 to the endosomal membrane, however, it appears to be unable to fully rescue the reduction of FAM21 observed on purified microsomes.

92

(* See figure legend on the next page.)

93 Figure 4.1. The endosomal association of retromer accessory proteins in Vps35 D620N rescue cells.

(A) Generation of untagged Vps35 rescue, untagged Vps35 D620N rescue, Vps35-GFP rescue and Vps35 D620N-GFP rescue cell lines. Equal amounts of cell lysates form indicated cells were subjected to SDS-PAGE and immunoblotted with antibodies against Vps35, Vps26A, and Tubulin. Graphs represent the expression level of retromer subunit Vps35 and Vps26A within indicated cell clones (means ± SD). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ns, not significant. (B) HeLa, Vps35 KO, Vps35-GFP rescue and Vps35 D620N-GFP rescue cells were harvested and subjected to subcellular fractionation to generated purified endosomal membrane and cytosolic fractions. Equal amount of whole cell lysates and 20 μg of protein samples from each fraction were subjected to SDS-PAGE, and immunolabelled with antibodies against SNX2, SNX5, SNX3, FAM21, TBC1D5, VARP and Rab7. Representative example is shown from three independent experiments. (C) The distribution of SNX3 in HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells was determined by the indirect immunofluorescence using antibodies against endogenous SNX3, EEA1 and Vps35. Scale bars, 10 μm. The co-localization between SNX3 and the early endosome marker EEA1 within indicated cells was quantified by the Pearson’s coefficient and the Manders’ coefficient, respectively (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). *, p < 0.05, **, p < 0.01, ***, p < 0.001, ****, p < 0.0001, ns, not significant. (D) The distribution of the WASH complex subunit FAM21 in HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells was determined by the indirect immunofluorescence using antibodies against endogenous FAM21 and EEA1. Scale bars, 10 μm. Graph represents the percentage of EEA1-labelled endosomes positive with FAM21 (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). *, p < 0.05, ****, p < 0.0001, ns, not significant. (E) The distribution of TBC1D5 in HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells was determined by the indirect immunofluorescence using antibodies against endogenous TBC1D5 and the early endosome marker, SNX1. Scale bars, 10 μm. The co- localization between TBC1D5 and SNX1 within indicated cells was quantified by the Pearson’s correlation coefficient or the Manders’ coefficient and represented in the graph (means ± SEM). Two- tailed Student’s t test was utilized to determine the statistical significance (n = 3). ****, p < 0.0001, ns, not significant.

The decreased level of TBC1D5 was also detected in the microsome fractions from Vps35 KO cells, which could be rescued by the expression of wild-type Vps35 or Vps35 D620N (Fig.1B). This observation was further confirmed by the immunofluorescence and colocalization analysis. In fact, TBC1D5 displayed decreased endosomal distribution in Vps35 KO cells, which was observed to be rescued in both Vps35 rescue cells and Vps35 D620N cells. Co-localization analysis indicated comparable co-localization levels of

TBC1D5 and SNX1, the early endosome marker, in two rescue cell lines (Pearson : RHeLa =

0.3980, RVps35 KO = 0.0311, RVps35 rescue = 0.2966, RVps35 D620N rescue = 0.2984; Manders: RHeLa

= 0.3768, RVps35 KO = 0.0892, RVps35 rescue = 0.3892, RVps35 D620N rescue = 0.3762; Figure 4.1E). These data indicate that Vps35 D620N is capable to rescue the endosomal dissociation of TBC1D5 caused by Vps35 knock-out. Taken together, these data suggest that Vps35 D620N partially rescues the endosomal dissociation of SNX3 and FAM21 caused by retromer depletion, but fully rescues the dissociation of TBC1D5.

94 4.3.2 The presence of Vps35 D620N variant decreased the affinity of FAM21 with retromer and altered endosome morphology

Retromer has been shown to interact directly with the WASH complex subunit FAM21 [272, 273]. Therefore, the impaired binding affinity of retromer to FAM21 could be a cause of endosomal dissociation of FAM21. To investigate whether the interaction is affected upon the presence of Vps35 D620N, Vps35-GFP rescue and Vps35 D620N rescue cells were harvested for co-immunoprecipitation using GFPTrap agarose beads. As shown in Figure 4.2A, the WASH complex subunit FAM21 was detected in association with retromer in Vps35-GFP rescue cells, as well as in Vps35 D620N-GFP rescue cells. However, the level of FAM21 that co-precipitated in Vps35 D620N-GFP rescue cells was significantly lower than that in Vps35-GFP rescue cells (Figure 4.2A), suggesting that the interaction is weaken in the presence of Vps35 D620N. Functional FAM21-retromer complexes are anticipated to be located within the same specific endosomal subdomains. Given the weak binding affinity of Vps35 D620N-containing retromer with FAM21/WASH complex, it is likely that the endosomal recruitment of FAM21 still happens, however, the recruitment efficiency is much lower or the stability of the retromer-FAM21 interaction upon membranes is reduced in the presence of Vps35 D620N. To determine this, STED super-resolution microscopy was employed to investigate the localization of FAM21 and retromer on individual endosomes. Vps35 D620N rescue and control cells were incubated with the wheat germ agglutinin (WGA) that is utilized to label the membrane of endosomal structures, fixed, and immunolabeled with FAM21 antibodies and conjugated Vps26A antibodies. STED microscopy revealed a large overlap of FAM21 and Vps26A in both HeLa and Vps35 D620N rescue cells (Figure 4.2B), suggesting that the WASH complex can still be recruited to the retromer-positive endosomes in the presence of Vps35 D620N, consistent with our previous obversions. Quantification of different types of endosomes revealed that around 95 % WGA-labeled endosomes were positive with both FAM21 and retromer, and only less than 5% of WGA- labeled endosomes were only positive with either FAM21 or retromer (Figure 4.2C). The proportion of endosomes positive with both FAM21 and retromer decreased by approximately 15% in Vps35 D620N rescue cells, compared to control cells. The proportion of endosomes positive with retromer only increased to over 10%, whereas the proportion of endosomes positive with FAM21 only increased slightly to be around 5% in Vps35 D620N rescue cells (Figure 4.2C). This change in the proportion of the different subsets of endosomes containing either FAM21 alone or retromer alone in the presence of Vps35 D620N variant, indicating that the endosomal recruitment of FAM21 is altered. In support,

95 both Pearson’s coefficient and Manders’ coefficient indicated a significant reduction in the co-localization of FAM21 with Vps26A in Vps35 D620N rescue cells at the super-resolution level, when comparing to HeLa control cells (Pearson : RHeLa = 0.5830, RVps35 D620N rescue =

0.4149; Manders: RHeLa = 0.7553, RVps35 D620N rescue = 0.5876; Figure 4.2D), which is unlike the almost unchanged colonization level of FAM21 and retromer showed by analysis at the confocal microscopy resolution (Figure 4.1D). In addition, the STED microscopy demonstrated decreased fluorescence intensity of FAM21 at endosomes in Vps35 D620N rescue cells (Figure 4.2E), unlike the comparable fluorescence intensity shown by confocal microscopy (Figure 4.1S). Taken together, these data indicate a redistribution of FAM21 in the presence of Vps35 D620N variant on individual endosomes, which is consistent with the previous observation demonstrating a reduced interaction of FAM21 with the Vps35 D620N variant.

Morphology features including size, circularity and aspect ratio were next quantified to investigate whether Vps35 D620N retromer impacts the endosomal morphology. As indicated by quantification, endosomes in Vps35 D620N rescue cells were smaller in size, with higher circularity score and lower aspect ratio score, compared to control cells (Figure 4.2F). This data demonstrates that the presence of Vps35 D620N impacts the endosomal morphology, resulting in smaller and rounder endosomes. The morphological change of endosomes is likely to reflect an alteration of endosomal maturation, deformation or endosome-derived tubule formation. To determine this, we rebuilt the 3D structure of individual endosomes in Vps35 D620N rescue cells and HeLa control cells, using images obtained with the super-resolution microscopy. Consistent with our previous observations, HeLa control cells demonstrated larger endosomes, most of which with extending tubular structure, whereas Vps35 D620N rescue cells showed smaller and rounder endosome, with less tubular structures observed (Figure 4.2G). In addition, we observed that Vps26A- positive regions at the individual endosome were almost fully occupied by FAM21 in control cells. However, reduced FAM21 overlapped with Vps26A-positive structure was detected in Vps35 D620N rescue cells (Figure 4.2G), suggesting partially affected endosomal association of FAM21 upon the presence of Vps35 D620N. Collectively, these data suggest that the Vps35 D620N-containing retromer is still capable of interacting and recruiting FAM21 to endosomes, however its capacity of maintaining the retromer-FAM21 complex after recruitment is partially affected, which will likely impact the endosomal morphology and tubule formation.

96

Figure 4.2. The Vps35 D620N variant affects the endosomal recruitment of the WASH complex.

(A) Cell lysates of HeLa, Vps35 KO, Vps35-GFP rescue, and Vps35 D620N-GFP rescue cells were collected and subjected to immunoprecipitation by GFP-NanoTrap beads. The binding of the WASH complex was determined by immunoblotting. (B) HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells were incubated with WGA-containing culture medium at 37 °C for 15 min, fixed, and co- stained with antibodies against Vps26A and FAM21, followed by Alexa Fluor conjugated fluorescent secondary antibodies, and imaged by the STED super-resolution microscopy. Scale bars, 0.5 μm. (C) Graph represents the percentage of distinct types of endosomes within HeLa and Vps35 D620N rescue cells. (D) The co-localization between FAM21 and Vps26A on endosomes within HeLa and Vps35 D620N rescue cells was quantified by the Pearson’s correlation coefficient or Manders’ correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance. *, p < 0.05, **, p < 0.01. (E) Graph represents the fluorescence intensity (means ± SEM) of FAM21 on endosomes co-positive with both FAM21 and Vps26A, within HeLa and Vps35 D620N rescue cells. N > 200 individual endosomes. ****, p < 0.0001. (F) Graphs represent the morphological features -- endosome size, circularity score, aspect ratio of the WGA+WASH+Vps26A+ endosomes within HeLa and Vps35 D620N rescue cells. N > 200 individual endosomes. **, p < 0.01, ****, p < 0.0001. (G) 3D reconstruction of the Vps26A, FAM21 and WGA staining in the individual WGA+WASH+Vps26A+ endosome within HeLa and Vps35 D620N rescue cells. Scale bars, 0.2 μm.

97 4.3.3 Absence of retromer-dependent CI-M6PR ETCs in Vps35 D620N rescue cells

Chapter 2 showed that retromer-mediated CI-M6PR ETCs are specifically captured by a specific trans-golgin protein, GCC88. To determine whether the Vps35 D620N variant impacts the retromer-dependent retrograde pathway, the ETC assay was performed to examine the trafficking of CI-M6PR-loaded ETCs. HeLa, Vps35 KO, Vps35 rescue, and Vps35 D620N rescue cells were transiently transfected with HA-tagged GCC88-MAO or GM130-MAO construct, fixed, and immunolabeled with antibodies against HA and CI-M6PR. As revealed by confocal microscopy, CI-M6PR vesicles were redistributed to the mitochondria in HeLa cells and Vps35 rescue cells expressing GCC88-MAO, but not in Vps35 KO cells expressing GCC88-MAO (Figure 4.3A), consistent with the previous study. However, the relocation of CI-M6PR was not observed in Vps35 D620N rescue cells (Figure 4.3A), suggesting that the expression of Vps35 D620N variant fails to rescue the trafficking defect of CI-M6PR-loaded ETCs captured by GCC88 caused by the absence of retromer.

This observation was confirmed by the Pearson’s coefficient (GCC88: RHeLa = 0.3398, RVps35

KO = 0.2097, RVps35 rescue = 0.2831, RVps35 D620N rescue = 0.1990; GM130: RHeLa = 0.2368, RVps35

KO = 0.1939, RVps35 rescue = 0.2218, RVps35 D620N rescue = 0.1748; Figure 4.3B). Similar defect of CI-M6PR ETCs trafficking was also observed in Vps35 D620N-GFP rescue cells (Figure 4.S2A and S2B). To determine if the presence of Vps35 D620N retromer causes similar retrograde trafficking defect, we next performed a series of colocalization analysis to examine the intracellular trafficking of CI-M6PR. HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells were fixed and co-immunolabeled with antibodies against CI-M6PR and individual molecular markers. As expected, in HeLa control cells, CI-M6PR displayed a perinuclear staining pattern with high overlap with p230-labeled TGN and EEA1-labeled early endosomes, but with minor overlap with LAMP1-labeled late endosomes/lysosomes (Figure 4.3C). Vps35 KO cells demonstrated altered CI-M6PR distribution which displayed a dispersed peripheral pattern (Figure 4.3C), consistent with previous reports [85, 91, 100, 191, 236, 238-242]. The aberrant CI-M6PR distribution was rescued by the expression of wild-type Vps35, but not fully rescued by Vps35 D620N (Figure 4.3C). In fact, colocalization analysis revealed a comparable reduction in the overlap of CI-M6PR with p230 in Vps35 KO and Vps35 D620N rescue cells, when comparing to HeLa control and Vps35 rescue cells (Figure 4.3D). The overlapping level of CI-M6PR with EEA1 was increased in Vps35 KO cells and was rescued to a normal level comparable to the control cells in Vps35 rescue cells (Figure 4.3D). Its colocalization with EEA1 was also rescued in Vps35 D620N rescue

98 cells, however, unlike Vps35 rescue cells, the level was lower than that in control cells (Figure 4.3D). These data suggest that Vps35 D620N retromer causes a reduction of receptors at endosomes and is not capable to fully rescue the retrograde trafficking defect of CI-M6PR caused by the absence of retromer.

(* See figure legend on the next page.)

99 Figure 4.3. The Vps35 D620N variant perturbs the retrograde trafficking of CI-M6PR by affecting the formation of CI-M6PR ETCs that are tethered by GCC88.

(A) HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells were transiently transfected with the HA-tagged mitochondria-targeting golgin construct GCC88-MAO or GM130-MAO. Cells were fixed, co-immunolabelled with antibodies against HA and endogenous CI-M6PR, followed by Alexa Fluor conjugated fluorescence secondary antibodies. Scale bars, 10 μm. The intensity plots of the fluorescent intensity (y-axis) against distance (x-axis) represent the overlap between channels. (B) The co-localization between CI-M6PR and HA-tagged golgin-mito proteins was quantified by Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). **, p < 0.01, ****, p < 0.0001, ns, not significant. (C) The distribution of CI-M6PR in HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells was determined by the indirect immunofluorescence using antibodies against endogenous CI-M6PR, p230, EEA1 and LAMP1. Scale bars, 10 μm. (D) The co-localization between CI-M6PR and different molecular markers was quantified by the Pearson’s correlation coefficient and represented in graph (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). *, p < 0.05, **, p < 0.01, ***, p < 0.001, ****, p < 0.0001, ns, not significant. (E) Equal amounts of cell lysates form HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells were subjected to SDS-PAGE and immunoblotted with antibodies against CI-M6PR, Vps35, and Tubulin.

4.3.4 The presence of Vps35 D620N rescues the lysosomal defects caused by retromer depletion

Retromer has been shown to play an important role in the maintenance of the endo- lysosomal system. To determine the effect of Vps35 D620N variant on the endo-lysosomal system, the lysosomal compartments were analyzed using electron microscopy (EM). Lysosomes defined as large, circular and electron-dense organelles occupied approximately 6% of the total cytoplasmic space in Vps35 KO cells, which was around two folds of that in HeLa control cells (Figure 4.4A), consistent with the data presented in Chapter 1. The volume made up by lysosomal compartments was rescued to normal level in both Vps35- GFP rescue and Vps35 D620N rescue cells (Figure 4.4A), suggesting that the presence of Vps35 D620N variant has no effect on lysosomal structures at the ultrastructure level. To determine if the Vps35 D620N variant affects the lysosomal function, the intracellular proteolytic kinetics was examined using DQ-BSA, a fluorogenic substrate for proteases, which enters into the intracellular environment via endocytosis and generates a strong fluorescence product when proteolyzed in lysosomes. HeLa, Vps35 KO, Vps35 rescue, and Vps35 D620N rescue were treated with DQ-BSA overnight, fixed, and immunolabeled with antibodies against endogenous LAMP1, the late endosome/lysosome marker. Consistent with the previous report, Vps35 KO cells showed low DQ-BSA fluorescence intensity in LAMP1-labeled lysosomes, when comparing to HeLa control cells (Figure 4.4B), suggesting impaired lysosomal proteolysis. Vps35 rescue and Vps35 D620N rescue cells demonstrated strong DQ-BSA fluorescence, which was comparable with that in control cells, within 100 LAMP1-positive lysosomes (Figure 4.4B), suggesting that the Vps35 D620N variant possess the capability to fully rescue the impaired lysosomal proteolysis caused by the absence of retromer. We next performed Magic Red cathepsin-B assay to monitor the changes in hydrolytic activity within intact cells to determine the effect of Vps35 D620N variant on lysosomal activity. HeLa, Vps35 KO, Vps35 rescue, and Vps35 D620N rescue cells were treated with Magic Red cathepsin-B, a cell-permeant fluorogenic substrate which generates fluorescence signal upon cleavage by cathepsin enzymes with lysosomal compartments. As indicated by live cell time-lapse imaging and fluorescence intensity analysis, Magic Red fluorescence increased gradually to a relatively stable level in HeLa control cells (Figure 4.4C). In comparison, the Magic Red fluorescence intensity maintained at an obviously lower level in Vps35 KO cells, suggesting reduced lysosomal enzyme activity (Figure 4.4C). Rescue phenotypes were observed in both Vps35 rescue and Vps35 D620N rescue cells, both of which demonstrated comparable levels of Magic Red fluorescence intensity with that in HeLa control cells. Therefore, the presence of Vps35 D620N variant rescues the decreased hydrolytic enzyme activity caused by the absence of retromer. Taken together, these data suggest that Vps35 D620N retromer has no impact on lysosomal morphology and proteolytic function.

4.3.5 Unaffected autophagy pathways in the presence of Vps35 D620N

Apart from lysosomal proteolysis, retromer deficiency has been shown to affect lysosome- autophagosome fusion, which is required for the final stages of autophagy. To determine if the Vps35 D620N variant affects the autophagy pathway, we examined the subcellular distribution of LC3-II, which is localized in the autophagosome membrane. HeLa, Vps35 KO, Vps35 rescue, Vps35 D620N rescue cells were fixed, co-immunolabeled with antibodies against endogenous LC3-II and lysosome marker LAMP1. As indicated by confocal microscopy and Pearson’s coefficient, Vps35 KO showed elevated membrane-bound LC3- II, with an increased colocalization with LAMP1, when comparing to HeLa control cells (Figure 4.5A), suggesting altered autophagy in the absence of retromer. Vps35 D620N rescue cells demonstrated comparable levels of LC3-II at LAMP1-positive lysosomes, relative to Vps35 rescue and HeLa cells (Figure 4.5A), suggesting that the presence of Vps35 D620N is capable to rescue the altered autophagy caused by retromer absence. To determine that if the autophagy induction is affected by the Vps35 D620N variant, the activity of the mTORC1 signaling pathway was investigated. As indicated by confocal microscopy, mTORC1 displayed a diffuse, cytoplasmic distribution in HeLa cells under amino acid starvation. Upon stimulated with amino acids, mTORC1 was recruited to LAMP1-positive

101 lysosomal compartments (Figure 4.5B). Vps35 KO demonstrated a reduction on the lysosomal recruitment of mTORC1 upon stimulation (Figure 4.5B), consistent with the previous report. In comparison, Vps35 D620N rescue cells rescued the defect observed in Vps35 KO cells, and showed comparable mTORC1 levels at lysosomal compartments, when comparing to Vps35 rescue cells and HeLa cells (Figure 4.5B). Similar phenotypes were observed in Vps35 D620N-GFP cells, which further supported the observations (Figure 4.S3A and S3B). Taken together, these data suggest that the presence of Vps35 D620N variant is capable to rescue the impaired autophagy pathways caused by retromer depletion.

(* See figure legend on the next page.)

102 Figure 4.4. The expression of Vps35 D620N variant rescues the lysosomal proteolytic defects caused by the absence of retromer.

(A) Electron micrographs of HeLa, Vps35 KO, Vps35-GFP rescue, and Vps35 D620N-GFP rescue cells. Enlarged circular structures are indicated as late endosomal/lysosomal structures. Scale bars represent 2000 nm, in zoomed images represent 500 nm. Graph represents the percentage volume density of lysosomal compartments relative to the cytoplasm in HeLa, Vps35 KO and Vps35-GFP rescue cells (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance. **, p < 0.01, ***, p < 0.001. N = two independent experiments with ten images each. (B) HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells were treated with DQ-BSA Red in complete medium at 37°C for overnight. Cells were fixed and immunolabelled with antibodies against LAMP1, followed by Alexa Fluor conjugated fluorescent secondary antibodies. Scale bars, 10 μm. Graph represents the fluorescent intensity of DQ-BSA Red (A.U, arbitrary units) within HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). *, p < 0.05, **, p < 0.01, ns, not significant. (C) HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells were treated with Magic Red Cathepsin-B in complete medium at 37 °C and imaged by a time-lapsed video microscopy performed with the inverted Nikon Ti-E microscopy with Hamamatsu Flash 4.0 sCMOS camera. Graph represents the Magic Red Cathepsin B fluorescent intensity (A.U, arbitrary units) within HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells at indicated time (means ± SEM). N = two independent experiments with eight random points each. *, p < 0.05, **, p < 0.01, ***, p < 0.001, ****, p < 0.0001.

103

Figure 4.5. The expression of Vps35 D620N variant rescues the altered autophagy flux caused by the absence of retromer.

(A) HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells were fixed, co-immunolabelled with antibodies against LC3-II and LAMP1, followed by Alexa Fluor conjugated fluorescent secondary antibodies. Scale bars, 10 μm. The co-localization between LC3-II and LAMP1 was quantified by the Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ***, p < 0.001, ****, p < 0.0001. (B) Amino acid starved HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells were treated with 2x essential amino acid (AA) solution for 30 min, fixed with ice-cold methanol and co-immunolabelled with antibodies against mTORC1 and LAMP1, followed by Alexa Fluor conjugated fluorescent secondary antibodies (means ± SEM). Scale bars, 10 μm. The co-localization of mTORC1 with LAMP1 was quantified by Pearson’s correlation coefficient. Two-tailed Student’s t test indicates the difference upon amino acid stimulation (n = 3). ****, p < 0.0001.

104 4.4 Discussion

In the last decades, a growing list of genetic variants in a range of endosomal trafficking proteins have been identified as risk factors in PD cases (reviewed in Chapter 1). Defined as the second most common cause of familial PD, mutations in the core retromer complex have been proposed to contribute to PD progression by affecting the endo-lysosomal pathways. In this study, we detailed characterize the effect of the point mutation D620N within retromer Vps35 subunit on its peripheral association with retromer associated interactome and on the endosomal dynamics and functions. Through evaluating and comparing the rescue efficiency between Vps35 D620N and wild-type Vps35 on reported deficits caused by retromer depletion, we conclude a partial loss function of Vps35 D620N- containing retromer in the endo-lysosomal system. We demonstrate that the depletion of retromer Vps35 results in the endosomal dissociation of a range of retromer interacting proteins including the WASH complex subunit FAM21, SNX3, TBC1D5, and VARP, consistent with a role of the core retromer complex as the endosomal recruitment hub for selected accessory proteins [98, 131, 192, 268, 272, 289, 292, 293]. The endosomal dissociation of FAM21 and SNX3 is only partially rescued by the presence of Vps35 D620N compared to the wild-type Vps35, the endosomal dissociation defect on TBC1D5 and VARP, however, is fully rescued. Consistent with previous studies [141, 192, 194, 284], we show that Vps35 D620N confers a profoundly reduced level of interaction with the FAM21 subunit of WASH complex, by which it further affects the endosomal association of this accessory protein complex. Moreover, we demonstrate that Vps35 D620N causes the presence of small and circular endosomes with less tubule-like branches, reflecting altered endosomal dynamics. The major proportion of endosomes exhibiting the morphological changes are co-positive with the core retromer complex and the WASH complex, suggesting that the retromer-mediated endosome functions could be impaired. Indeed, the presence of Vps35 D620N results in reduced endosome-to-TGN sorting of CI-M6PR, as demonstrated by the absence of GCC88-captured CI-M6PR-loaded ETCs and the altered distribution of CI-M6PR. Finally, we demonstrate that the lysosome functions including lysosomal proteolysis and autophagy at the whole cell level are not affected by the Vps35 D620N variant.

Since being identified by two independent groups in familial PD cases in 2011 [170, 171], the D620N point mutation within Vps35 has been proposed to contribute to the disease progression by impacting retromer’s function. While Vps35 D620N binds with the Vps26A and Vps29 subunit with high affinity, the point mutation is likely to affect the molecular interaction with retromer interacting effectors due to the positioning of D620N residue within

105 the Vps35 protein structure [88, 193, 220]. Evidence implicates that the core retromer complex interacts with accessory proteins to form distinct functional modules to regulate the cargo trafficking within the endosomal system [235]. One important endosomal assembly that associates with the core retromer complex is the WASH complex, a major actin nucleator fine-tuning the actin polymerization at the endosomal membrane. WASH is a multimeric protein complex composed of five subunits. Among them, one critical component is the FAM21 subunit, which confers binding sites for multiple retromer Vps35 subunits at its tail domain [268, 272, 293]. Consistent with a role of retromer in the endosomal recruitment of FAM21 [268, 272, 293], the membrane dissociation of this accessory protein was observed in cells with a knock-out of retromer Vps35 subunit. The presence of Vps35 D620N variant in the knock-out system only partially rescues the FAM21 dissociation, which could be caused by the reduced binding affinity of Vps35 D620N and FAM21, in support of previous studies [192, 194, 284]. Rather than completely abolish the membrane association, we show that Vps35 D620N is still capable of recruiting FAM21 to the endosomes. However, stimulated emission depletion (STED) microscopy revealed that the proportion of individual retromer positive endosomes which also recruited FAM21 was reduced in Vps35 D620N rescue cells. One recent structural study proposes that the Vps retromer core complex may function in the form of an arch-like structure by a homodimeric interaction of Vps35 subunits [294]. The D620N point mutation is localized adjacent to the homodimerization interface of Vps35, raising a possibility that the PD-linked mutation may affect the dimerization of Vps35, thus affecting the overall assembly and the function of the retromer complex [294]. Therefore, the Vps35 D620N variant may cause an alternation in the functional conformation of the core retromer complex, thereby affects its binding stability for FAM21 and leads to dissociation of this accessory protein. In addition, the protein post-translational modifications such as the ubiquitination, on Vps35 may also associate with the capability of retromer to bind and recruit FAM21. Recently, six lysine residues that can be targeted for the ubiquitination have been identified within the C-terminal region of Vps35 adjacent to the D620N position [295]. TRIM27, an RING E3 ubiquitin ligase, together with its enhancer – melanoma antigen (MAGE)-L2, has been shown to directly bind with the core retromer complex to promote WASH ubiquitination and regulate the WASH activity [296]. Although it is unknown that whether TRIM27-MAGE-L1 directly promotes Vps35 ubiquitination, Vps35 D620N variant may interfere its binding with the TRIM27, thereby affecting the ability of TRIM27 on the WASH/FAM21 ubiquitination and its activity.

Our previous study as well as the studies from other groups show that the depletion of SNX3 alters retrieval of retromer-mediated CI-M6PR vesicles, suggesting the incorporation of

106 SNX3 in the retromer-mediated retrograde pathways [122, 268]. Defined as a sorting nexin that only contains the PX domain, SNX3 functions in the endosomal recruiting of the core retromer complex by binding with the phophosinostide-3-phosphate (PI3P) enriched in early endosomes via its PX domain and the retromer subunit Vps35 and Vps26A [36, 120, 270, 289]. Intriguingly, we show that apart from FAM21, the Vps35 D620N also causes reduced endosome association of SNX3. We speculate that this defect could be caused by a mechanism distinct from the WASH dissociation, since the SNX3 binding region within Vps35 is distal from the position of the D620N point mutation [88, 289]. Despite the exact molecular mechanisms regarding how Vps35 D620N reduces the endosomal association of the WASH complex and SNX3 are still unclear, yet the mis-trafficking of CI-M6PR in the presence of Vps35 D620N variant is established consistent with previous reports. CI-M6PR is responsible for the lysosomal delivery of newly-synthesized hydrolases, such as cathepsin-D. Overexpression of the PD-linked Vps35 D620N has been shown to impact the processing of cathepsin-D [193]. Unlike what we expected, the lysosomal morphology and the lysosomal proteolytic activity at the whole cell level are unaffected by Vps35 D620N, suggesting that in addition to CI-M6PR mistrafficking, other molecular mechanisms could also contribute to the lysosomal defects observed in cells absent of retromer. Moreover, the Vps35 D620N also fully rescues the autophagy dysfunction caused by the absence of retromer Vps35 (Chapter 2), which is demonstrated by the comparable level of LC3-II, an adaptor for the autophagosomes. This observation is consistent with the observation in Vps35 D620N cells demonstrating unaffected endosomal association of TBC1D5, which controls the LC3 shuttling [129]. The observed defects on the lysosomal recruitment of mTORC1 and the subsequent activation in the absence of Vps35, are also rescued to comparable level by Vps35 D620N with the wild-type Vps35. While no effect on the autophagy pathways observed in our cell models, previous study demonstrates inhibited autophagy in cells stably over-expressing Vps35 D620N [194], suggesting that the amount of abnormal retromer within intracellular environment may associate with the autophagy processes. Furthermore, it has previously been reported that the trafficking of Lamp2a, a key component for chaperon-mediated autophagy, can be affected in the dopamine neurons deficient with Vps35 or expressing of D620N variant [276], suggesting that retromer Vps35 D620N variant may have a very specific function in this sub-type of autophagy pathway.

In conclusion, my data demonstrated that the PD-linked Vps35 D620N variant partially perturbs the endosomal association of the WASH complex subunit FAM21 by impairing the molecular interaction of Vps35 and FAM21. The endosomal morphology of the proportion of endosomes co-positive with retromer and WASH complex is altered at the super-resolution

107 level, suggesting the altered endosomal dynamics and function in the presence of Vps35 D620N variant. Besides, the peripheral association of SNX3 is also affected by Vps35 D620N, which partially contributes to CI-M6PR mistrafficking. However, the lysosomal morphology and functions are unaffected by the Vps35 D620N variant.

108

Figure 4.S1. Supplementary data for Figure 4.1.

Graph represents the total fluorescent intensity of FAM21 (A.U, arbitrary units) within HeLa, Vps35 KO, Vps35 rescue and Vps35 D620N rescue cells at steady state (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ****, p < 0.0001.

Figure 4.S2. Supplementary data for Figure 4.3.

(A) Vps35-GFP rescue and Vps35 D620N-GFP rescue cells were transiently transfected with the HA-tagged mitochondria-targeting golgin construct GCC88-MAO or GM130-MAO. Cells were fixed, co-immunolabelled with antibodies against HA and endogenous CI-M6PR, followed by Alexa Fluor conjugated fluorescence secondary antibodies. Scale bars, 10 μm. (B) The co-localization between CI-M6PR and HA-tagged golgin-mito proteins was quantified by Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ****, p < 0.0001, ns, not significant.

109

Figure 4.S3. Supplementary data for Figure 4.5.

(A) HeLa, Vps35 KO, Vps35-GFP rescue and Vps35 D620N-GFP rescue cells were fixed, co- immunolabelled with antibodies against LC3-II and LAMP1, followed by Alexa Fluor conjugated fluorescent secondary antibodies. Scale bars, 10 μm. The co-localization between LC3-II and LAMP1 was quantified by the Pearson’s correlation coefficient (means ± SEM). Two-tailed Student’s t test was utilized to determine the statistical significance (n = 3). ****, p < 0.0001. (B) Amino acid starved HeLa, Vps35 KO, Vps35-GFP rescue and Vps35 D620N-GFP rescue cells were treated with 2x essential amino acid (AA) solution for 30 min, fixed with ice-cold methanol and co-immunolabelled with antibodies against mTORC1 and LAMP1, followed by Alexa Fluor conjugated fluorescent secondary antibodies (means ± SEM). Scale bars, 10 μm. The co-localization of mTORC1 with LAMP1 was quantified by Pearson’s correlation coefficient. Two-tailed Student’s t test indicates the difference upon amino acid stimulation (n = 3). ****, p < 0.0001.

110 Chapter 5. General Discussion

While retromer has been implicated as a critical endosomal scaffold in numerous studies over the last decades, its exact molecular action remains to be elucidated. This thesis has updated the current understanding for retromer’s function in the endosomal trafficking, and the maintenance of lysosomal dynamics, and demonstrated how the PD-linked Vps35 D620N variant impacts the function of retromer within the endo-lysosomal system.

5.1 Molecular actions of retromer in the retrograde pathway

Since initially characterized by two independent groups in 2004 [100, 238], the functioning of the mammalian retromer in the endosome-to-TGN trafficking of a variety of cargo receptors has been detailed by multiple studies (reviewed in Chapter 1). One working model that has been widely accepted for a long period is that the core retromer complex, comprising Vps35, Vps29, and Vps26 subunits, mediates the retrograde transport of the CI- M6PR via a direct interaction between the retromer Vps35 subunit and the Trp-Leu-Met (WLM) motif within the cytoplasmic tail of CI-M6PR [101]. The ability of the core retromer complex to bind CI-M6PR is dependent on the retromer-SNX3 association. Structural studies demonstrate that a binding site for a canonical sorting signal, ØX(L/M) consensus motif (where Ø can be any hydrophobic amino acid), which is located at the interface between SNX3 and the retromer Vps26 subunit, is involved in the cargo recognition of retromer [89, 111]. The requirement of the core retromer complex in CI-M6PR retrieval has been demonstrated by a series of CI-M6PR trafficking assay performed in cells silenced for Vps35 or Vps26A by RNA interference [85, 100, 101, 191, 192, 236-243]. The established model by earlier studies has recently been challenged. Reports from three independent labs suggest that the endosome-to-TGN retrieval of CI-M6PR doesn’t require the core retromer complex, by showing unaffected CI-M6PR distribution in cells depleted with Vps35 [94, 95, 261]. Instead, a direct interaction between the CI-M6PR tail and SNX5 or SNX6 has been revealed, suggesting a role of the SNX-BAR dimer, composed of SNX1 or SNX2 with either SNX5 or SNX6, in the cargo recognition and endosomal sorting [94, 95]. There is no definitive answer to the question of which the best way is to measure the endosome-to-TGN trafficking of CI-M6PR. While the recent studies [94, 95, 261] have also applied an array of microscopy-based assays (i.e. CI-M6PR dispersal assay, CD8α-CI-M6PR internalization assay) which are commonly used in earlier studies [85, 100, 101, 191, 192, 236-243], the effects of CI-M6PR to its downstream pathway, such as CI-M6PR-dependent enzyme trafficking, should also be examined. The contradictory observations demonstrated by the

111 earlier studies and the recent studies have raised the question on the exact function of the core retromer complex in the endosomal system [237]. Regardless of the variation of experimental conditions used by different research groups, the distinct observations suggest the requirement of further experimentation and call us to re-examine the current working model for the core retromer complex.

This thesis updates the current understanding of retromer’s functional roles in the endosomal trafficking by proposing the working model outlined the action of retromer in at least two independent pathways (Fig. 5.1). By using the recently developed ETC assay [61], the core retromer complex sorts CI-M6PR into a subset of retrograde transport vesicles that can be selectively captured by GCC88. CI-M6PR sorted by the SNX-BAR proteins are likely to be packaged into a distinct type of transport carriers captured by golgin-245. Thus, this study, on one hand, demonstrates the requirement of the core retromer complex in CI-M6PR retrieval, consistent with the earlier studies [85, 100, 101, 191, 192, 236-243], and on the other hand supports the SNX-BAR functioning in sorting of CI-M6PR that is suggested by recent studies. However, it is still unclear whether the sorting of CI-M6PR retrograde trafficking mediated by SNX-BAR proteins also requires retromer core complex, given that SNX-BAR can function independent of retromer core complex. Previous studies on retromer interactome provide insights for the molecular mechanisms underlying how the full function of the core retromer complex is achieved (see Chapter 1). By investigating the functional association of the core retromer complex and its accessory proteins, I revealed that SNX3 was required for the retrograde transport of CI-M6PR vesicles sorted by the core retromer complex, highlighting an emerging role of SNX3 in the endosomal trafficking.

By combining the previous studies and the data shown in the Chapter 2, the updated working model is that the retromer is firstly recruited to endosomal membranes through a direct interaction with SNX3, which associates with early endosomes via its PI(3)P binding PX domain [89, 122, 270]. While other regulators involved in the recruitment of retromer to the endosomes have been identified including Rab7 [98, 123] and SNX12 [122], a direct requirement of this retromer-dependent trafficking pathway needs to be determined, given that retromer is required to be spatially and temporally recruited for each of its multiple membrane transport pathways. For instance, SNX12, similar to SNX3, is a SNX containing PX-domain only. However, its relative expression in HeLa is lower compared to SNX3 expression [122]. Therefore, it is possible that SNX3-containing retromer and SNX12- containing retromer can mediate retromer-dependent cargo retrograde trafficking in a cell type-dependent manner. Once recruited to the cargo-enriched endosomes, SNX3-retromer

112 appears selectively interact with a consensus XfL/M motif within cytoplasmic domains of transmembrane cargoes including CI-M6PR [89, 101]. Retromer’s direct role in the formation of ETCs is less clear, as the formation of GCC88-tethered ETCs which contain CD-M6PR can still be observed in the absence of retromer. Retromer function maybe restricted to the recruitment of cargo into the forming ETC or alternatively, the ETC containing CD-M6PR observed are a distinct type of ETC that are also tethered by GCC88. After released from endosomes, the ETCs are recognised by tethering factors anchored at the TGN which capture the specific classes of ETCs [61]. The retromer dependent subset of CI-M6PR ETCs are specifically recognised by GCC88, rather than golgin-97 or golgin- 245. This specificity is consistent with observations that the tethering domains within golgin- 97 and golgin-245 are related and quite different from that in GCC88, [60] and that the tethering of ETCs by GCC88 is independent of the bridging factor TBC1D23 which connects ETC associated WASH with golgin-245 [65]. Moreover, the study in Chapter 3 further demonstrates the requirement of GCC88 in the endosome-to-TGN retrieval of CI-M6PR, consistent with the proposed functional association of GCC88 and the core retromer complex in Chapter 2.

5.2 Retromer and the maintenance of lysosomal function

Lysosomes are dynamic organelles that mediate the degradation of macromolecules from the endocytic and autophagic pathway [231, 232]. Functional lysosome ensures the efficient digestion of intracellular obsolete components, and materials endocytosed from the extracellular environment and therefore is essential for the maintenance of cellular homeostasis and cell survival. Efficient lysosomal proteolysis is achieved by the activity of acid hydrolases within lysosomes. Previous studies demonstrate that CI-M6PR is responsible for the lysosomal targeting of one subset of hydrolases with the M6P binding signal. Mistrafficking of CI-M6PR caused by retromer depletion has been shown to affect the processing of CI-M6PR-associated hydrolases such as cathepsin-D [100]. Therefore, the possibility is raised that the functional retromer could contribute to the maintenance of the normal lysosomal function. The study in Chapter 2 supports this hypothesis by demonstrating the decreased enzyme activity within lysosomes and the impaired proteolysis upon the depletion of the retromer Vps35 subunit. In addition, retromer depletion induces enlarged lysosomes at the ultrastructural level, suggesting the accumulation of undegraded materials, consistent with the affected enzyme delivery. Moreover, these lysosomes with morphological alteration are suggested to be non-acidic and may represent the lysosomes with functional defects. Consistent with this concept, the study in Chapter 3 demonstrates

113 similar defects on lysosomal enzyme activity and proteolysis upon the depletion of GCC88, a tethering factor for retromer-mediated retrograde vesicles.

Figure 5.1. Evidence for two independent types of ETC responsible for the endosome- to-TGN transport of CI-M6PR in mammalian cells.

In mammalian cells, CI-M6PR can be sorted into ETCs which depend on retromer/SNX3 and are tethered by GCC88 at the TGN. Alternatively, CI-M6PR can be sorted into ETCs which are independent of retromer but depend on SNX-BAR proteins and are tethered by golgin-245 at the TGN. The recently identified bridging protein TBC1D23 [65] is not required for tethering of the GCC88 dependent ETCs but is required for the tethering of ETC by golgin-245.

Apart from mediating the degradation via proteolysis, functional lysosomes are also required to fuse with autophagosomes for the final stage of autophagy. By investigating the intracellular level of the membrane-bounded LC3-II, an adaptor and cargo for autophagy, I reveal the altered autophagy upon the depletion of the core retromer complex by showing an increased level of membrane-bounded LC3-II at lysosomes. Increased induction of autophagy in retromer depleted cells can be due to the dissociation of TBC1D5 from endosome, leading to the increased cycling of LC3. In addition, the dissociation of TBC1D5 in retromer Vps35 KO cells also can result in the miss-trafficking of ATG9 to induce the autophagy [194]. The altered autophagy is further supported by a reduction of mTORC1 signalling activity, which is antagonistic to the induction of autophagy, upon stimulation in cells depleted with retromer. Depletion of GCC88 doesn’t affect the mTORC1 signalling

114 pathway, suggesting that there is no direct association of retromer-GCC88-mediated retrograde trafficking with the regulation of autophagy pathways.

5.3 Retromer and the Parkinson’s disease

Ongoing studies continue to revealed a variety of point mutations within a number of genes associated with familial PD cases, highlighting the association of genetic variation with the pathogenesis of PD. Retromer variants are the second common genetic cause for the late- onset familial PD cases, just after LRRK2 variants. Our group has previously characterized the PD-linked Vps35 variants including Vps35 D620N, Vps35 R524W, and Vps35 P316S using cell models overexpressing individual variants [193]. While all three variants are still able to assemble the core Vps trimeric complex, Vps35 P316S has a minor effect on retromer function, whereas Vps35 D620N and R524W perturbs CI-M6PR retrieval mediated by the core retromer complex. Although both Vps35 D620N and R524W show a primary effect on retromer function, distinct molecular mechanisms regarding how these PD-linked variants contribute to the disease-related defects have been demonstrated. Previous studies demonstrated that the presence of Vps35 R524W variant impairs the endosomal association of the core retromer complex [193], which could be due to the reduced association between retromer and its interacting partners, including TBC1D5, Rab7, and the WASH complex subunit FAM21. Unlike the Vps35 R524W variant, the endosomal association of the core retromer complex is not affected by the presence of Vps35 D620N, highlighting a more sophisticate mechanism underlying how Vps35 D620N contributes to the PD progression.

Since initially discovered by two independent groups in 2011 [170, 171], several studies have suggested the effect of Vps35 D620N in the endo-lysosomal system [192-194]. Previous studies based on mouse neuron models have evaluated the pathogenetic defects of the Vps35 D620N variant on the transport of neuronal receptors, the trafficking of cargoes related to autophagy pathways including CMA and macroautophagy, and mitochondria dynamics (reviewed in Chapter 1). However, the precise biochemical mechanism of the action of Vps35 D620N remains to be elucidated. Particularly, previous studies done in cell models mostly used the approach by over-expressing Vps35 D620N variant [192-194]. Given that cells having both the endogenous wild-type retromer and the over-expressed Vps35 D620N-containg retromer, it is difficult to reconcile the exact machinery of how Vps35 D620N variant contributes to defects in the endosomal system. The study in Chapter 4 characterizes the detailed effects of Vps35 D620N on retromer’s functions including the recruitment of accessory proteins, the endosomal trafficking, and the maintenance of

115 lysosomal functions. The approach utilized in this study is to stably-expressing either the wild-type Vps35 or Vps35 D620N variant in Vps35 KO cell line to avoid the effect from endogenous retromer. The requirement of the core retromer complex in the endosomal recruitment of SNX3, the WASH complex subunit FAM21, TBC1D5, and VARP is demonstrated, consistent with the proposed role of the core retromer complex as the recruitment hub for selected accessory effectors [98, 131, 192, 268, 272, 289, 292, 293]. Distinct rescue efficiency on the endosomal dissociation of different accessory proteins is revealed in Vps35 D620N rescue cells. In fact, Vps35 D620N rescues the recruitment defect of TBC1D5 and VARP caused by Vps35 depletion, whereas only partially rescues that of FAM21 and SNX3. This study further demonstrates that Vps35 D620N affects the binding affinity of retromer with FAM21, but not with other retromer interacting partners, such as TBC1D5 (data not shown), consistent with previous reports [192, 194, 295]. Further demonstrated by the super-resolution microscopy, Vps35 D620N causes a reduced level of FAM21 at individual endosomes positive with the core retromer complex, suggesting a partial loss-of-function. Chapter 2 has demonstrated the incorporation of SNX3 in the retrograde transport of CI-M6PR sorted via the retromer-mediated mechanism, thus the altered SNX3 distribution caused by Vps35 D620N is proposed to impact the CI-M6PR retrieval. Indeed, the absence of GCC88-captured CI-M6PR and mistrafficking of CI-M6PR are observed in the presence of Vps35 D620N, suggesting that the Vps35 D620N affects the formation of the CI-M6PR-loaded ETCs that are sorted via a retromer-SNX3 mediated mechanism. Besides, I speculate that the impaired retromer-mediated cargo retrieval in presence of Vps35 D620N could also be due to the morphological alteration of endosomes co-positive with retromer and FAM21. In fact, at super-resolution levels, the endosome in Vps35 D620N rescue cells display a morphology of smaller, more vesicular and with less tubule-like branches. Changes on the endosomal morphology may reflect a defect on the formation and fission of endosome-derived tubules, suggesting the altered endosomal function.

Initial studies proposed that the core retromer complex contributes to the maintenance of lysosomal functioning by mediating the endosomal trafficking of CI-M6PR, an acid hydrolase transporter, as demonstrated by the defects on lysosomal functions in Vps35 depleted cells. Thus, the impaired CI-M6PR retrieval caused by Vps35 D620N could also cause lysosomal defects similar to Vps35 depletion. However, here, my study shows that Vps35 D620N is capable of rescuing the defects on lysosomal proteolysis and autophagy observed in cells without retromer. These observations are consistent with a partial loss-of-function of Vps35 D620N and suggest that the impairment on trafficking of retromer-associated cargo via

116 GCC88 tethered ETCs is not the only cause for the lysosomal defects observed in cells depleted of retromer Vps35 subunit, highlighting the involvement of other retromer- associated mechanisms in the maintenance of lysosomal function. In fact, Vps35 D620N rescues the endosomal dissociation of TBC1D5 caused by Vps35 depletion. TBC1D5 regulates LC3 cycling between the endosome and the cytosol [129], thus the retromer- TBC1D5 mechanism could partially contribute to the maintenance of lysosomal function. While showing no effect for Vps35 D620N on the lysosomal function, it doesn’t mean that this PD-linked variant has no contribution to the disease-associated defects on the endo- lysosomal system. Studies on mouse neuron models expressing Vps35 D620N have suggested the effect of the variant on the transport of Lamp2a, a key regulator for the CMA pathway, suggesting a specific action of Vps35 D620N in the regulation of autophagy [276].

5.4 Future perspectives

While this thesis extends the current understanding of the action of the core retromer complex in the endosomal system, more molecular details remain to be elucidated. The core retromer complex selectively sorts CI-M6PR into a subset of endosomal transport vesicles that can be captured by GCC88, however, the mechanism regarding how GCC88 recognizes and captures this specific subset of ETCs is unknown. As it has been shown that the core retromer complex is released from the forming endosomal transport vesicles once after the fission step [127], the GCC88-mediated tethering for retromer-associated ETCs is unlikely to be regulated via a direct mechanism. Recent studies have identified TBC1D23 as a bridging factor between golgin-97, golgin-245 and transport vesicles using mass spectrometry of proteins biotinylated by the mitochondria targeting trans-golgin protein fused with the biotin ligase [65]. The tethering process mediated by GCC88 is independent of TBC1D23, which is consistent with the distinct vesicle-tethering domain of GCC88 from that of golgin-97 and golgin-245 [60, 65]. However, these studies suggest that the adaptor proteins and bridge factors, which connects the retromer-mediated ETCs with the N-terminal vesicle-tethering domain in GCC88, may be required for GCC88 to selectively tether ETCs sorted via the retromer-mediated mechanism. The identification of these adaptor proteins or bridge factors is critical for the fully understanding of how retromer/GCC88 mediates the retrograde pathway. Besides, in addition to CI-M6PR, it is unclear if there is any other retromer-associated retrograde cargoes can also be loaded into the subset of ETCs captured by GCC88. It is possible that cargo trafficking mediated by the retromer/GCC88 mechanism is disturbed under certain disease conditions, such as PD. Hence, investigation of the additional cargo proteins incorporated into the retromer/GCC88 pathway will improve

117 our understanding for the physiological context. Apart from CI-M6PR ETCs sorted by the core retromer complex, GCC88 is also capable to capture vesicles loaded with CD-M6PR, a retromer-independent cargo receptor. This observation raises the question of whether different cargoes sorted via the mechanism dependent or independent of retromer are packaged into the same vesicles or distinct type of vesicles. Electronic microscopy studies show that the cargo vesicles redirected to the mitochondria outer membrane are circular and uniform in size (~50 nm in diameter) [61]. Therefore, the individual vesicles can be investigated by the STED super-resolution microscopy which is able to detect fluorescent particles at the resolution of 20 nm.

Retromer dysfunction has been shown to contribute to the progression of neurodegenerative diseases. Here, my study has demonstrated that the PD-linked Vps35 D620N has a partial loss-of-function in the endo-lysosomal system using cell models stably expressing Vps35 D620N in the Vps35 knock-out background. While consistent with previous reports suggesting that Vps35 D620N causes defects on the endosomal trafficking [192, 193], my study shows that this variant causes no obvious effects on the autophagy pathways, which is inconsistent with the previously reported cell models stably expressing Vps35 D620N. Such difference could be due to the amount of abnormal retromer, raising the possibility that the intracellular level of normal or abnormal retromer may result in different cellular defects. In fact, recent structure studies demonstrate that the core retromer complex functions via forming homodimer by the interactions on Vps35 subunits [294]. The point mutation of D620N is located within the binding region of Vps35 subunits, suggesting that this PD-linked point mutation may affect the formation of functioning retromer homodimer, thus further impairing retromer's function. However, how the change in the ratio of normal retromer and variant could affect the functioning structure of retromer is unclear yet.

Apart from retromer, a growing list of proteins (i.e. α-synuclein, LRRK2, and Parkin) have been reported as risk factors in familial PD cases (reviewed in Chapter 1). The demonstrated functional association between retromer and other PD-linked proteins highlights the complicated mechanism underlying PD progression. However, the detailed molecular action of the PD-linked Vps35 D620N variant in this functional network remains to be elucidated. This could be investigated by using the recently reported heterozygous and homozygous Vps35 D620N knock-in mouse models, which both display PD-linked pathological defects [200, 284, 285].

118 Chapter 6. References

1. van der Sluijs, P., Hull, M., Webster, P., Male, P., Goud, B. & Mellman, I. (1992) The small GTP- binding protein rab4 controls an early sorting event on the endocytic pathway, Cell. 70, 729-40. 2. Sheff, D. R., Daro, E. A., Hull, M. & Mellman, I. (1999) The receptor recycling pathway contains two distinct populations of early endosomes with different sorting functions, J Cell Biol. 145, 123-39. 3. Choudhury, A., Sharma, D. K., Marks, D. L. & Pagano, R. E. (2004) Elevated endosomal levels in Niemann-Pick cells inhibit rab4 and perturb membrane recycling, Mol Biol Cell. 15, 4500-11. 4. Kouranti, I., Sachse, M., Arouche, N., Goud, B. & Echard, A. (2006) Rab35 regulates an endocytic recycling pathway essential for the terminal steps of cytokinesis, Curr Biol. 16, 1719-25. 5. Sonnichsen, B., De Renzis, S., Nielsen, E., Rietdorf, J. & Zerial, M. (2000) Distinct membrane domains on endosomes in the recycling pathway visualized by multicolor imaging of Rab4, Rab5, and Rab11, J Cell Biol. 149, 901-14. 6. Takahashi, S., Kubo, K., Waguri, S., Yabashi, A., Shin, H. W., Katoh, Y. & Nakayama, K. (2012) Rab11 regulates exocytosis of recycling vesicles at the plasma membrane, J Cell Sci. 125, 4049-57. 7. Magadan, J. G., Barbieri, M. A., Mesa, R., Stahl, P. D. & Mayorga, L. S. (2006) Rab22a regulates the sorting of transferrin to recycling endosomes, Mol Cell Biol. 26, 2595-614. 8. Simonsen, A., Lippe, R., Christoforidis, S., Gaullier, J. M., Brech, A., Callaghan, J., Toh, B. H., Murphy, C., Zerial, M. & Stenmark, H. (1998) EEA1 links PI(3)K function to Rab5 regulation of endosome fusion, Nature. 394, 494-8. 9. Zhu, H., Zhu, G., Liu, J., Liang, Z., Zhang, X. C. & Li, G. (2007) Rabaptin-5-independent membrane targeting and Rab5 activation by Rabex-5 in the cell, Mol Biol Cell. 18, 4119-28. 10. Aikawa, Y. & Lee, S. (2013) Role of Rabex-5 in the sorting of ubiquitinated cargo at an early stage in the endocytic pathway, Commun Integr Biol. 6, e24463. 11. Murray, J. T., Panaretou, C., Stenmark, H., Miaczynska, M. & Backer, J. M. (2002) Role of Rab5 in the recruitment of hVps34/p150 to the early endosome, Traffic. 3, 416-427. 12. Katzmann, D. J., Babst, M. & Emr, S. D. (2001) Ubiquitin-dependent sorting into the multivesicular body pathway requires the function of a conserved endosomal protein sorting complex, ESCRT-I, Cell. 106, 145-55. 13. Teo, H., Gill, D. J., Sun, J., Perisic, O., Veprintsev, D. B., Vallis, Y., Emr, S. D. & Williams, R. L. (2006) ESCRT-I core and ESCRT-II GLUE domain structures reveal role for GLUE in linking to ESCRT-I and membranes, Cell. 125, 99-111. 14. Raiborg, C. & Stenmark, H. (2009) The ESCRT machinery in endosomal sorting of ubiquitylated membrane proteins, Nature. 458, 445-52. 15. Rink, J., Ghigo, E., Kalaidzidis, Y. & Zerial, M. (2005) Rab conversion as a mechanism of progression from early to late endosomes, Cell. 122, 735-49. 16. Wurmser, A. E., Sato, T. K. & Emr, S. D. (2000) 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. 151, 551-62. 17. Gary, J. D., Wurmser, A. E., Bonangelino, C. J., Weisman, L. S. & Emr, S. D. (1998) Fab1p is essential for PtdIns(3)P 5-kinase activity and the maintenance of vacuolar size and membrane homeostasis, J Cell Biol. 143, 65-79. 18. Er, E. E., Mendoza, M. C., Mackey, A. M., Rameh, L. E. & Blenis, J. (2013) AKT facilitates EGFR trafficking and degradation by phosphorylating and activating PIKfyve, Sci Signal. 6, ra45. 19. Jefferies, H. B., Cooke, F. T., Jat, P., Boucheron, C., Koizumi, T., Hayakawa, M., Kaizawa, H., Ohishi, T., Workman, P., Waterfield, M. D. & Parker, P. J. (2008) A selective PIKfyve inhibitor blocks

119 PtdIns(3,5)P(2) production and disrupts endomembrane transport and retroviral budding, EMBO Rep. 9, 164-70. 20. McCartney, A. J., Zhang, Y. & Weisman, L. S. (2014) Phosphatidylinositol 3,5-bisphosphate: low abundance, high significance, Bioessays. 36, 52-64. 21. Whitley, P., Reaves, B. J., Hashimoto, M., Riley, A. M., Potter, B. V. & Holman, G. D. (2003) Identification of mammalian Vps24p as an effector of phosphatidylinositol 3,5-bisphosphate- dependent endosome compartmentalization, J Biol Chem. 278, 38786-95. 22. Bright, N. A., Gratian, M. J. & Luzio, J. P. (2005) Endocytic delivery to lysosomes mediated by concurrent fusion and kissing events in living cells, Curr Biol. 15, 360-5. 23. Kim, B. Y., Kramer, H., Yamamoto, A., Kominami, E., Kohsaka, S. & Akazawa, C. (2001) Molecular characterization of mammalian homologues of class C Vps proteins that interact with syntaxin-7, J Biol Chem. 276, 29393-402. 24. Caplan, S., Hartnell, L. M., Aguilar, R. C., Naslavsky, N. & Bonifacino, J. S. (2001) Human Vam6p promotes lysosome clustering and fusion in vivo, J Cell Biol. 154, 109-22. 25. Poupon, V., Stewart, A., Gray, S. R., Piper, R. C. & Luzio, J. P. (2003) The role of mVps18p in clustering, fusion, and intracellular localization of late endocytic organelles, Mol Biol Cell. 14, 4015- 27. 26. Bucci, C., Thomsen, P., Nicoziani, P., McCarthy, J. & van Deurs, B. (2000) Rab7: a key to lysosome biogenesis, Mol Biol Cell. 11, 467-80. 27. Pryor, P. R., Mullock, B. M., Bright, N. A., Lindsay, M. R., Gray, S. R., Richardson, S. C., Stewart, A., James, D. E., Piper, R. C. & Luzio, J. P. (2004) Combinatorial SNARE complexes with VAMP7 or VAMP8 define different late endocytic fusion events, EMBO Rep. 5, 590-5. 28. Pryor, P. R., Mullock, B. M., Bright, N. A., Gray, S. R. & Luzio, J. P. (2000) The role of intraorganellar Ca(2+) in late endosome-lysosome heterotypic fusion and in the reformation of lysosomes from hybrid organelles, J Cell Biol. 149, 1053-62. 29. Mari, M., Bujny, M. V., Zeuschner, D., Geerts, W. J., Griffith, J., Petersen, C. M., Cullen, P. J., Klumperman, J. & Geuze, H. J. (2008) SNX1 defines an early endosomal recycling exit for sortilin and mannose 6-phosphate receptors, Traffic. 9, 380-93. 30. Lefrancois, S., Zeng, J., Hassan, A. J., Canuel, M. & Morales, C. R. (2003) The lysosomal trafficking of sphingolipid activator proteins (SAPs) is mediated by sortilin, EMBO J. 22, 6430-7. 31. van Amerongen, R. & Nusse, R. (2009) Towards an integrated view of Wnt signaling in development, Development. 136, 3205-14. 32. Das, S., Yu, S., Sakamori, R., Stypulkowski, E. & Gao, N. (2012) Wntless in Wnt secretion: molecular, cellular and genetic aspects, Front Biol (Beijing). 7, 587-593. 33. Burguete, A. S., Fenn, T. D., Brunger, A. T. & Pfeffer, S. R. (2008) Rab and Arl GTPase family members cooperate in the localization of the golgin GCC185, Cell. 132, 286-98. 34. Yang, P. T., Lorenowicz, M. J., Silhankova, M., Coudreuse, D. Y., Betist, M. C. & Korswagen, H. C. (2008) Wnt signaling requires retromer-dependent recycling of MIG-14/Wntless in Wnt- producing cells, Developmental cell. 14, 140-7. 35. Gasnereau, I., Herr, P., Chia, P. Z., Basler, K. & Gleeson, P. A. (2011) Identification of an endocytosis motif in an intracellular loop of Wntless protein, essential for its recycling and the control of Wnt protein signaling, The Journal of biological chemistry. 286, 43324-33. 36. Harterink, M., Port, F., Lorenowicz, M. J., McGough, I. J., Silhankova, M., Betist, M. C., van Weering, J. R., van Heesbeen, R. G., Middelkoop, T. C., Basler, K., Cullen, P. J. & Korswagen, H. C. (2011) A SNX3-dependent retromer pathway mediates retrograde transport of the Wnt sorting receptor Wntless and is required for Wnt secretion, Nature cell biology. 13, 914-23. 37. Yu, J., Chia, J., Canning, C. A., Jones, C. M., Bard, F. A. & Virshup, D. M. (2014) WLS retrograde transport to the endoplasmic reticulum during Wnt secretion, Developmental cell. 29, 277-91.

120 38. McGough, I. J., de Groot, R. E. A., Jellett, A. P., Betist, M. C., Varandas, K. C., Danson, C. M., Heesom, K. J., Korswagen, H. C. & Cullen, P. J. (2018) SNX3-retromer requires an evolutionary conserved MON2:DOPEY2:ATP9A complex to mediate Wntless sorting and Wnt secretion, Nat Commun. 9, 3737. 39. Varlamov, O. & Fricker, L. D. (1998) Intracellular trafficking of metallocarboxypeptidase D in AtT-20 cells: localization to the trans-Golgi network and recycling from the cell surface, J Cell Sci. 111 ( Pt 7), 877-85. 40. Johannes, L. & Popoff, V. (2008) Tracing the retrograde route in protein trafficking, Cell. 135, 1175-87. 41. Chia, P. Z., Gasnereau, I., Lieu, Z. Z. & Gleeson, P. A. (2011) Rab9-dependent retrograde transport and endosomal sorting of the endopeptidase furin, J Cell Sci. 124, 2401-13. 42. Wan, L., Molloy, S. S., Thomas, L., Liu, G., Xiang, Y., Rybak, S. L. & Thomas, G. (1998) PACS- 1 defines a novel gene family of cytosolic sorting proteins required for trans-Golgi network localization, Cell. 94, 205-16. 43. Crump, C. M., Xiang, Y., Thomas, L., Gu, F., Austin, C., Tooze, S. A. & Thomas, G. (2001) PACS-1 binding to adaptors is required for acidic cluster motif-mediated protein traffic, EMBO J. 20, 2191-201. 44. Scott, G. K., Fei, H., Thomas, L., Medigeshi, G. R. & Thomas, G. (2006) A PACS-1, GGA3 and CK2 complex regulates CI-MPR trafficking, EMBO J. 25, 4423-35. 45. Burgert, T., Schmidt, V., Caglayan, S., Lin, F., Fuchtbauer, A., Fuchtbauer, E. M., Nykjaer, A., Carlo, A. S. & Willnow, T. E. (2013) SORLA-dependent and -independent functions for PACS1 in control of amyloidogenic processes, Molecular and cellular biology. 33, 4308-20. 46. Dittie, A. S., Thomas, L., Thomas, G. & Tooze, S. A. (1997) Interaction of furin in immature secretory granules from neuroendocrine cells with the AP-1 adaptor complex is modulated by casein kinase II phosphorylation, EMBO J. 16, 4859-70. 47. Teuchert, M., Schafer, W., Berghofer, S., Hoflack, B., Klenk, H. D. & Garten, W. (1999) Sorting of furin at the trans-Golgi network. Interaction of the cytoplasmic tail sorting signals with AP-1 Golgi- specific assembly proteins, J Biol Chem. 274, 8199-207. 48. Meyer, C., Zizioli, D., Lausmann, S., Eskelinen, E. L., Hamann, J., Saftig, P., von Figura, K. & Schu, P. (2000) mu1A-adaptin-deficient mice: lethality, loss of AP-1 binding and rerouting of mannose 6-phosphate receptors, EMBO J. 19, 2193-203. 49. Nakagawa, T., Setou, M., Seog, D., Ogasawara, K., Dohmae, N., Takio, K. & Hirokawa, N. (2000) A novel motor, KIF13A, transports mannose-6-phosphate receptor to plasma membrane through direct interaction with AP-1 complex, Cell. 103, 569-81. 50. Saint-Pol, A., Yelamos, B., Amessou, M., Mills, I. G., Dugast, M., Tenza, D., Schu, P., Antony, C., McMahon, H. T., Lamaze, C. & Johannes, L. (2004) Clathrin adaptor epsinR is required for retrograde sorting on early endosomal membranes, Dev Cell. 6, 525-38. 51. Choudhury, R., Diao, A., Zhang, F., Eisenberg, E., Saint-Pol, A., Williams, C., Konstantakopoulos, A., Lucocq, J., Johannes, L., Rabouille, C., Greene, L. E. & Lowe, M. (2005) Lowe syndrome protein OCRL1 interacts with clathrin and regulates protein trafficking between endosomes and the trans-Golgi network, Molecular biology of the cell. 16, 3467-79. 52. van Rahden, V. A., Brand, K., Najm, J., Heeren, J., Pfeffer, S. R., Braulke, T. & Kutsche, K. (2012) The 5-phosphatase OCRL mediates retrograde transport of the mannose 6-phosphate receptor by regulating a Rac1-cofilin signalling module, Hum Mol Genet. 21, 5019-38. 53. Orsel, J. G., Sincock, P. M., Krise, J. P. & Pfeffer, S. R. (2000) Recognition of the 300-kDa mannose 6-phosphate receptor cytoplasmic domain by 47-kDa tail-interacting protein, Proc Natl Acad Sci U S A. 97, 9047-51. 54. Diaz, E. & Pfeffer, S. R. (1998) TIP47: a cargo selection device for mannose 6-phosphate receptor trafficking, Cell. 93, 433-43.

121 55. Burguete, A. S., Sivars, U. & Pfeffer, S. (2005) Purification and analysis of TIP47 function in Rab9-dependent mannose 6-phosphate receptor trafficking, Methods Enzymol. 403, 357-66. 56. Hanna, J., Carroll, K. & Pfeffer, S. R. (2002) Identification of residues in TIP47 essential for Rab9 binding, Proc Natl Acad Sci U S A. 99, 7450-4. 57. Carroll, K. S., Hanna, J., Simon, I., Krise, J., Barbero, P. & Pfeffer, S. R. (2001) Role of Rab9 GTPase in facilitating receptor recruitment by TIP47, Science. 292, 1373-6. 58. Reitz, C. (2018) Retromer Dysfunction and Neurodegenerative Disease, Curr Genomics. 19, 279-288. 59. Munro, S. (2011) The golgin coiled-coil proteins of the , Cold Spring Harb Perspect Biol. 3. 60. Wong, M., Gillingham, A. K. & Munro, S. (2017) The golgin coiled-coil proteins capture different types of transport carriers via distinct N-terminal motifs, BMC Biol. 15, 3. 61. Wong, M. & Munro, S. (2014) Membrane trafficking. The specificity of vesicle traffic to the Golgi is encoded in the golgin coiled-coil proteins, Science. 346, 1256898. 62. Derby, M. C., van Vliet, C., Brown, D., Luke, M. R., Lu, L., Hong, W., Stow, J. L. & Gleeson, P. A. (2004) Mammalian GRIP domain proteins differ in their membrane binding properties and are recruited to distinct domains of the TGN, Journal of cell science. 117, 5865-74. 63. Houghton, F. J., Chew, P. L., Lodeho, S., Goud, B. & Gleeson, P. A. (2009) The localization of the Golgin GCC185 is independent of Rab6A/A' and Arl1, Cell. 138, 787-94. 64. Torres, I. L., Rosa-Ferreira, C. & Munro, S. (2014) The Arf family G protein Arl1 is required for secretory granule biogenesis in Drosophila, J Cell Sci. 127, 2151-60. 65. Shin, J. J. H., Gillingham, A. K., Begum, F., Chadwick, J. & Munro, S. (2017) TBC1D23 is a bridging factor for endosomal vesicle capture by golgins at the trans-Golgi, Nature cell biology. 66. Brown, F. C., Schindelhaim, C. H. & Pfeffer, S. R. (2011) GCC185 plays independent roles in Golgi structure maintenance and AP-1-mediated vesicle tethering, J Cell Biol. 194, 779-87. 67. Hayes, G. L., Brown, F. C., Haas, A. K., Nottingham, R. M., Barr, F. A. & Pfeffer, S. R. (2009) Multiple Rab GTPase binding sites in GCC185 suggest a model for vesicle tethering at the trans- Golgi, Mol Biol Cell. 20, 209-17. 68. Jing, J., Junutula, J. R., Wu, C., Burden, J., Matern, H., Peden, A. A. & Prekeris, R. (2010) FIP1/RCP binding to Golgin-97 regulates retrograde transport from recycling endosomes to the trans-Golgi network, Mol Biol Cell. 21, 3041-53. 69. Gillingham, A. K., Sinka, R., Torres, I. L., Lilley, K. S. & Munro, S. (2014) Toward a comprehensive map of the effectors of rab GTPases, Developmental cell. 31, 358-73. 70. Perez-Victoria, F. J., Mardones, G. A. & Bonifacino, J. S. (2008) Requirement of the human GARP complex for mannose 6-phosphate-receptor-dependent sorting of cathepsin D to lysosomes, Mol Biol Cell. 19, 2350-62. 71. Perez-Victoria, F. J. & Bonifacino, J. S. (2009) Dual roles of the mammalian GARP complex in tethering and SNARE complex assembly at the trans-golgi network, Molecular and cellular biology. 29, 5251-63. 72. Panic, B., Perisic, O., Veprintsev, D. B., Williams, R. L. & Munro, S. (2003) Structural basis for Arl1-dependent targeting of homodimeric GRIP domains to the Golgi apparatus, Mol Cell. 12, 863- 74. 73. Luke, M. R., Kjer-Nielsen, L., Brown, D. L., Stow, J. L. & Gleeson, P. A. (2003) GRIP domain- mediated targeting of two new coiled-coil proteins, GCC88 and GCC185, to subcompartments of the trans-Golgi network, The Journal of biological chemistry. 278, 4216-26. 74. Sinka, R., Gillingham, A. K., Kondylis, V. & Munro, S. (2008) Golgi coiled-coil proteins contain multiple binding sites for Rab family G proteins, J Cell Biol. 183, 607-15.

122 75. Barr, F. A. (1999) A novel Rab6-interacting domain defines a family of Golgi-targeted coiled-coil proteins, Curr Biol. 9, 381-4. 76. Van Valkenburgh, H., Shern, J. F., Sharer, J. D., Zhu, X. & Kahn, R. A. (2001) ADP-ribosylation factors (ARFs) and ARF-like 1 (ARL1) have both specific and shared effectors: characterizing ARL1- binding proteins, J Biol Chem. 276, 22826-37. 77. Kakinuma, T., Ichikawa, H., Tsukada, Y., Nakamura, T. & Toh, B. H. (2004) Interaction between p230 and MACF1 is associated with transport of a glycosyl phosphatidyl inositol-anchored protein from the Golgi to the cell periphery, Exp Cell Res. 298, 388-98. 78. Lin, Y. C., Chiang, T. C., Liu, Y. T., Tsai, Y. T., Jang, L. T. & Lee, F. J. (2011) ARL4A acts with GCC185 to modulate Golgi complex organization, J Cell Sci. 124, 4014-26. 79. Reddy, J. V., Burguete, A. S., Sridevi, K., Ganley, I. G., Nottingham, R. M. & Pfeffer, S. R. (2006) A functional role for the GCC185 golgin in mannose 6-phosphate receptor recycling, Mol Biol Cell. 17, 4353-63. 80. Yu, N., Signorile, L., Basu, S., Ottema, S., Lebbink, J. H., Leslie, K., Smal, I., Dekkers, D., Demmers, J. & Galjart, N. (2016) Isolation of Functional Tubulin Dimers and of Tubulin-Associated Proteins from Mammalian Cells, Curr Biol. 26, 1728-36. 81. Efimov, A., Kharitonov, A., Efimova, N., Loncarek, J., Miller, P. M., Andreyeva, N., Gleeson, P., Galjart, N., Maia, A. R., McLeod, I. X., Yates, J. R., 3rd, Maiato, H., Khodjakov, A., Akhmanova, A. & Kaverina, I. (2007) Asymmetric CLASP-dependent nucleation of noncentrosomal microtubules at the trans-Golgi network, Dev Cell. 12, 917-30. 82. Ganley, I. G., Espinosa, E. & Pfeffer, S. R. (2008) A syntaxin 10-SNARE complex distinguishes two distinct transport routes from endosomes to the trans-Golgi in human cells, J Cell Biol. 180, 159- 72. 83. Mallard, F., Tang, B. L., Galli, T., Tenza, D., Saint-Pol, A., Yue, X., Antony, C., Hong, W., Goud, B. & Johannes, L. (2002) Early/recycling endosomes-to-TGN transport involves two SNARE complexes and a Rab6 isoform, J Cell Biol. 156, 653-64. 84. Tai, G., Lu, L., Wang, T. L., Tang, B. L., Goud, B., Johannes, L. & Hong, W. (2004) Participation of the syntaxin 5/Ykt6/GS28/GS15 SNARE complex in transport from the early/recycling endosome to the trans-Golgi network, Mol Biol Cell. 15, 4011-22. 85. Wassmer, T., Attar, N., Bujny, M. V., Oakley, J., Traer, C. J. & Cullen, P. J. (2007) A loss-of- function screen reveals SNX5 and SNX6 as potential components of the mammalian retromer, Journal of cell science. 120, 45-54. 86. Griffin, C. T., Trejo, J. & Magnuson, T. (2005) Genetic evidence for a mammalian retromer complex containing sorting nexins 1 and 2, Proc Natl Acad Sci U S A. 102, 15173-7. 87. Gokool, S., Tattersall, D., Reddy, J. V. & Seaman, M. N. (2007) Identification of a conserved motif required for Vps35p/Vps26p interaction and assembly of the retromer complex, The Biochemical journal. 408, 287-95. 88. Hierro, A., Rojas, A. L., Rojas, R., Murthy, N., Effantin, G., Kajava, A. V., Steven, A. C., Bonifacino, J. S. & Hurley, J. H. (2007) Functional architecture of the retromer cargo-recognition complex, Nature. 449, 1063-7. 89. Lucas, M., Gershlick, D. C., Vidaurrazaga, A., Rojas, A. L., Bonifacino, J. S. & Hierro, A. (2016) Structural Mechanism for Cargo Recognition by the Retromer Complex, Cell. 167, 1623-1635 e14. 90. Bugarcic, A., Vetter, I., Chalmers, S., Kinna, G., Collins, B. M. & Teasdale, R. D. (2015) Vps26B- retromer negatively regulates plasma membrane resensitization of PAR-2, Cell Biol Int. 39, 1299- 306. 91. Bugarcic, A., Zhe, Y., Kerr, M. C., Griffin, J., Collins, B. M. & Teasdale, R. D. (2011) Vps26A and Vps26B subunits define distinct retromer complexes, Traffic. 12, 1759-73. 92. Kerr, M. C., Bennetts, J. S., Simpson, F., Thomas, E. C., Flegg, C., Gleeson, P. A., Wicking, C. & Teasdale, R. D. (2005) A novel mammalian retromer component, Vps26B, Traffic. 6, 991-1001.

123 93. Nisar, S., Kelly, E., Cullen, P. J. & Mundell, S. J. (2010) Regulation of P2Y1 receptor traffic by sorting Nexin 1 is retromer independent, Traffic. 11, 508-19. 94. Kvainickas, A., Jimenez-Orgaz, A., Nagele, H., Hu, Z., Dengjel, J. & Steinberg, F. (2017) Cargo- selective SNX-BAR proteins mediate retromer trimer independent retrograde transport, J Cell Biol. 95. Simonetti, B., Danson, C. M., Heesom, K. J. & Cullen, P. J. (2017) Sequence-dependent cargo recognition by SNX-BARs mediates retromer-independent transport of CI-MPR, J Cell Biol. 96. Seaman, M. N., Gautreau, A. & Billadeau, D. D. (2013) Retromer-mediated endosomal protein sorting: all WASHed up!, Trends Cell Biol. 23, 522-8. 97. Temkin, P., Lauffer, B., Jager, S., Cimermancic, P., Krogan, N. J. & von Zastrow, M. (2011) SNX27 mediates retromer tubule entry and endosome-to-plasma membrane trafficking of signalling receptors, Nat Cell Biol. 13, 715-21. 98. Seaman, M. N., Harbour, M. E., Tattersall, D., Read, E. & Bright, N. (2009) 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. 122, 2371-82. 99. Gokool, S., Tattersall, D. & Seaman, M. N. (2007) EHD1 interacts with retromer to stabilize SNX1 tubules and facilitate endosome-to-Golgi retrieval, Traffic. 8, 1873-86. 100. Arighi, C. N., Hartnell, L. M., Aguilar, R. C., Haft, C. R. & Bonifacino, J. S. (2004) Role of the mammalian retromer in sorting of the cation-independent mannose 6-phosphate receptor, The Journal of cell biology. 165, 123-33. 101. Seaman, M. N. (2007) Identification of a novel conserved sorting motif required for retromer- mediated endosome-to-TGN retrieval, Journal of cell science. 120, 2378-89. 102. Bao, M., Booth, J. L., Elmendorf, B. J. & Canfield, W. M. (1996) Bovine UDP-N- acetylglucosamine:lysosomal-enzyme N-acetylglucosamine-1-phosphotransferase. I. Purification and subunit structure, J Biol Chem. 271, 31437-45. 103. Hasilik, A., Klein, U., Waheed, A., Strecker, G. & von Figura, K. (1980) Phosphorylated oligosaccharides in lysosomal enzymes: identification of alpha-N- acetylglucosamine(1)phospho(6)mannose diester groups, Proc Natl Acad Sci U S A. 77, 7074-8. 104. Reitman, M. L. & Kornfeld, S. (1981) UDP-N-acetylglucosamine:glycoprotein N- acetylglucosamine-1-phosphotransferase. Proposed enzyme for the phosphorylation of the high mannose oligosaccharide units of lysosomal enzymes, J Biol Chem. 256, 4275-81. 105. Varki, A. & Kornfeld, S. (1980) Structural studies of phosphorylated high mannose-type oligosaccharides, J Biol Chem. 255, 10847-58. 106. Olson, L. J., Castonguay, A. C., Lasanajak, Y., Peterson, F. C., Cummings, R. D., Smith, D. F. & Dahms, N. M. (2015) Identification of a fourth mannose 6-phosphate binding site in the cation- independent mannose 6-phosphate receptor, Glycobiology. 25, 591-606. 107. Bohnsack, R. N., Song, X., Olson, L. J., Kudo, M., Gotschall, R. R., Canfield, W. M., Cummings, R. D., Smith, D. F. & Dahms, N. M. (2009) Cation-independent mannose 6-phosphate receptor: a composite of distinct phosphomannosyl binding sites, J Biol Chem. 284, 35215-26. 108. Fjorback, A. W., Seaman, M., Gustafsen, C., Mehmedbasic, A., Gokool, S., Wu, C., Militz, D., Schmidt, V., Madsen, P., Nyengaard, J. R., Willnow, T. E., Christensen, E. I., Mobley, W. B., Nykjaer, A. & Andersen, O. M. (2012) Retromer binds the FANSHY sorting motif in SorLA to regulate amyloid precursor protein sorting and processing, J Neurosci. 32, 1467-80. 109. Lane, R. F., Raines, S. M., Steele, J. W., Ehrlich, M. E., Lah, J. A., Small, S. A., Tanzi, R. E., Attie, A. D. & Gandy, S. (2010) Diabetes-associated SorCS1 regulates Alzheimer's amyloid-beta metabolism: evidence for involvement of SorL1 and the retromer complex, J Neurosci. 30, 13110-5. 110. Pocha, S. M., Wassmer, T., Niehage, C., Hoflack, B. & Knust, E. (2011) Retromer controls epithelial cell polarity by trafficking the apical determinant Crumbs, Curr Biol. 21, 1111-7.

124 111. Tabuchi, M., Yanatori, I., Kawai, Y. & Kishi, F. (2010) Retromer-mediated direct sorting is required for proper endosomal recycling of the mammalian iron transporter DMT1, J Cell Sci. 123, 756-66. 112. Ghai, R., Tello-Lafoz, M., Norwood, S. J., Yang, Z., Clairfeuille, T., Teasdale, R. D., Merida, I. & Collins, B. M. (2015) Phosphoinositide binding by the SNX27 FERM domain regulates its localization at the immune synapse of activated T-cells, J Cell Sci. 128, 553-65. 113. Steinberg, F., Gallon, M., Winfield, M., Thomas, E. C., Bell, A. J., Heesom, K. J., Tavare, J. M. & Cullen, P. J. (2013) A global analysis of SNX27-retromer assembly and cargo specificity reveals a function in glucose and metal ion transport, Nat Cell Biol. 15, 461-71. 114. Hussain, N. K., Diering, G. H., Sole, J., Anggono, V. & Huganir, R. L. (2014) Sorting Nexin 27 regulates basal and activity-dependent trafficking of AMPARs, Proc Natl Acad Sci U S A. 111, 11840- 5. 115. Teasdale, R. D. & Collins, B. M. (2012) Insights into the PX (phox-homology) domain and SNX (sorting nexin) protein families: structures, functions and roles in disease, The Biochemical journal. 441, 39-59. 116. Horazdovsky, B. F., Davies, B. A., Seaman, M. N., McLaughlin, S. A., Yoon, S. & Emr, S. D. (1997) A sorting nexin-1 homologue, Vps5p, forms a complex with Vps17p and is required for recycling the vacuolar protein-sorting receptor, Mol Biol Cell. 8, 1529-41. 117. Rojas, R., Kametaka, S., Haft, C. R. & Bonifacino, J. S. (2007) 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. 27, 1112-24. 118. Haft, C. R., de la Luz Sierra, M., Bafford, R., Lesniak, M. A., Barr, V. A. & Taylor, S. I. (2000) Human orthologs of yeast vacuolar protein sorting proteins Vps26, 29, and 35: assembly into multimeric complexes, Mol Biol Cell. 11, 4105-16. 119. Collins, B. M., Skinner, C. F., Watson, P. J., Seaman, M. N. & Owen, D. J. (2005) Vps29 has a phosphoesterase fold that acts as a protein interaction scaffold for retromer assembly, Nat Struct Mol Biol. 12, 594-602. 120. Harrison, M. S., Hung, C. S., Liu, T. T., Christiano, R., Walther, T. C. & Burd, C. G. (2014) A mechanism for retromer endosomal coat complex assembly with cargo, Proceedings of the National Academy of Sciences of the United States of America. 111, 267-72. 121. Pons, V., Ustunel, C., Rolland, C., Torti, E., Parton, R. G. & Gruenberg, J. (2012) SNX12 role in endosome membrane transport, PloS one. 7, e38949. 122. Priya, A., Sugatha, J., Parveen, S., Lacas-Gervais, S., Raj, P., Gilleron, J. & Datta, S. (2017) Essential and selective role of SNX12 in transport of endocytic and retrograde cargo, Journal of cell science. 130, 2707-2721. 123. Rojas, R., van Vlijmen, T., Mardones, G. A., Prabhu, Y., Rojas, A. L., Mohammed, S., Heck, A. J., Raposo, G., van der Sluijs, P. & Bonifacino, J. S. (2008) Regulation of retromer recruitment to endosomes by sequential action of Rab5 and Rab7, J Cell Biol. 183, 513-26. 124. Amulya, K., Reddy, M. V. & Mohan, S. V. (2014) Acidogenic spent wash valorization through polyhydroxyalkanoate (PHA) synthesis coupled with fermentative biohydrogen production, Bioresour Technol. 158, 336-42. 125. Mamo, A., Jules, F., Dumaresq-Doiron, K., Costantino, S. & Lefrancois, S. (2012) The role of ceroid lipofuscinosis neuronal protein 5 (CLN5) in endosomal sorting, Mol Cell Biol. 32, 1855-66. 126. Modica, G., Skorobogata, O., Sauvageau, E., Vissa, A., Yip, C. M., Kim, P. K., Wurtele, H. & Lefrancois, S. (2017) Rab7 palmitoylation is required for efficient endosome-to-TGN trafficking, J Cell Sci. 130, 2579-2590. 127. Arlt, H., Reggiori, F. & Ungermann, C. (2015) Retromer and the dynamin Vps1 cooperate in the retrieval of transmembrane proteins from vacuoles, J Cell Sci. 128, 645-55.

125 128. Purushothaman, L. K., Arlt, H., Kuhlee, A., Raunser, S. & Ungermann, C. (2017) Retromer- driven membrane tubulation separates endosomal recycling from Rab7/Ypt7-dependent fusion, Mol Biol Cell. 28, 783-791. 129. Roy, S., Leidal, A. M., Ye, J., Ronen, S. M. & Debnath, J. (2017) Autophagy-Dependent Shuttling of TBC1D5 Controls Plasma Membrane Translocation of GLUT1 and Glucose Uptake, Mol Cell. 67, 84-95 e5. 130. Popovic, D., Akutsu, M., Novak, I., Harper, J. W., Behrends, C. & Dikic, I. (2012) Rab GTPase- activating proteins in autophagy: regulation of endocytic and autophagy pathways by direct binding to human ATG8 modifiers, Mol Cell Biol. 32, 1733-44. 131. Jia, D., Zhang, J. S., Li, F., Wang, J., Deng, Z., White, M. A., Osborne, D. G., Phillips-Krawczak, C., Gomez, T. S., Li, H., Singla, A., Burstein, E., Billadeau, D. D. & Rosen, M. K. (2016) Structural and mechanistic insights into regulation of the retromer coat by TBC1d5, Nat Commun. 7, 13305. 132. Gallon, M., Clairfeuille, T., Steinberg, F., Mas, C., Ghai, R., Sessions, R. B., Teasdale, R. D., Collins, B. M. & Cullen, P. J. (2014) A unique PDZ domain and arrestin-like fold interaction reveals mechanistic details of endocytic recycling by SNX27-retromer, Proceedings of the National Academy of Sciences of the United States of America. 111, E3604-13. 133. Peterson, L., Ismond, K. P., Chapman, E. & Flood, P. (2014) Potential benefits of therapeutic use of beta2-adrenergic receptor agonists in neuroprotection and Parkinsonmus disease, J Immunol Res. 2014, 103780. 134. Chan, A. S., Clairfeuille, T., Landao-Bassonga, E., Kinna, G., Ng, P. Y., Loo, L. S., Cheng, T. S., Zheng, M., Hong, W., Teasdale, R. D., Collins, B. M. & Pavlos, N. J. (2016) Sorting nexin 27 couples PTHR trafficking to retromer for signal regulation in osteoblasts during bone growth, Mol Biol Cell. 27, 1367-82. 135. van Weering, J. R., Sessions, R. B., Traer, C. J., Kloer, D. P., Bhatia, V. K., Stamou, D., Carlsson, S. R., Hurley, J. H. & Cullen, P. J. (2012) Molecular basis for SNX-BAR-mediated assembly of distinct endosomal sorting tubules, EMBO J. 31, 4466-80. 136. Hong, Z., Yang, Y., Zhang, C., Niu, Y., Li, K., Zhao, X. & Liu, J. J. (2009) The retromer component SNX6 interacts with dynactin p150(Glued) and mediates endosome-to-TGN transport, Cell Res. 19, 1334-49. 137. Wassmer, T., Attar, N., Harterink, M., van Weering, J. R., Traer, C. J., Oakley, J., Goud, B., Stephens, D. J., Verkade, P., Korswagen, H. C. & Cullen, P. J. (2009) The retromer coat complex coordinates endosomal sorting and dynein-mediated transport, with carrier recognition by the trans- Golgi network, Dev Cell. 17, 110-22. 138. Freeman, C. L., Hesketh, G. & Seaman, M. N. (2014) RME-8 coordinates the activity of the WASH complex with the function of the retromer SNX dimer to control endosomal tubulation, J Cell Sci. 127, 2053-70. 139. Chi, R. J., Liu, J., West, M., Wang, J., Odorizzi, G. & Burd, C. G. (2014) Fission of SNX-BAR- coated endosomal retrograde transport carriers is promoted by the dynamin-related protein Vps1, J Cell Biol. 204, 793-806. 140. Cai, B., Xie, S., Caplan, S. & Naslavsky, N. (2014) GRAF1 forms a complex with MICAL-L1 and EHD1 to cooperate in tubular recycling endosome vesiculation, Front Cell Dev Biol. 2, 22. 141. Derivery, E., Sousa, C., Gautier, J. J., Lombard, B., Loew, D. & Gautreau, A. (2009) The Arp2/3 activator WASH controls the fission of endosomes through a large multiprotein complex, Developmental cell. 17, 712-23. 142. Simunovic, M., Manneville, J. B., Renard, H. F., Evergren, E., Raghunathan, K., Bhatia, D., Kenworthy, A. K., Voth, G. A., Prost, J., McMahon, H. T., Johannes, L., Bassereau, P. & Callan- Jones, A. (2017) Friction Mediates Scission of Tubular Membranes Scaffolded by BAR Proteins, Cell. 170, 172-184 e11.

126 143. Choy, R. W., Park, M., Temkin, P., Herring, B. E., Marley, A., Nicoll, R. A. & von Zastrow, M. (2014) Retromer mediates a discrete route of local membrane delivery to dendrites, Neuron. 82, 55- 62. 144. Temkin, P., Morishita, W., Goswami, D., Arendt, K., Chen, L. & Malenka, R. (2017) The Retromer Supports AMPA Receptor Trafficking During LTP, Neuron. 94, 74-82 e5. 145. Loo, L. S., Tang, N., Al-Haddawi, M., Dawe, G. S. & Hong, W. (2014) A role for sorting nexin 27 in AMPA receptor trafficking, Nat Commun. 5, 3176. 146. Damseh, N., Danson, C. M., Al-Ashhab, M., Abu-Libdeh, B., Gallon, M., Sharma, K., Yaacov, B., Coulthard, E., Caldwell, M. A., Edvardson, S., Cullen, P. J. & Elpeleg, O. (2015) A defect in the retromer accessory protein, SNX27, manifests by infantile myoclonic epilepsy and neurodegeneration, Neurogenetics. 16, 215-221. 147. Wang, C. L., Tang, F. L., Peng, Y., Shen, C. Y., Mei, L. & Xiong, W. C. (2012) VPS35 regulates developing mouse hippocampal neuronal morphogenesis by promoting retrograde trafficking of BACE1, Biol Open. 1, 1248-57. 148. Tian, Y., Tang, F. L., Sun, X., Wen, L., Mei, L., Tang, B. S. & Xiong, W. C. (2015) VPS35- deficiency results in an impaired AMPA receptor trafficking and decreased dendritic spine maturation, Mol Brain. 8, 70. 149. Muhammad, A., Flores, I., Zhang, H., Yu, R., Staniszewski, A., Planel, E., Herman, M., Ho, L., Kreber, R., Honig, L. S., Ganetzky, B., Duff, K., Arancio, O. & Small, S. A. (2008) Retromer deficiency observed in Alzheimer's disease causes hippocampal dysfunction, neurodegeneration, and Abeta accumulation, Proc Natl Acad Sci U S A. 105, 7327-32. 150. Wen, L., Tang, F. L., Hong, Y., Luo, S. W., Wang, C. L., He, W., Shen, C., Jung, J. U., Xiong, F., Lee, D. H., Zhang, Q. G., Brann, D., Kim, T. W., Yan, R., Mei, L. & Xiong, W. C. (2011) VPS35 haploinsufficiency increases Alzheimer's disease neuropathology, J Cell Biol. 195, 765-79. 151. Small, S. A., Kent, K., Pierce, A., Leung, C., Kang, M. S., Okada, H., Honig, L., Vonsattel, J. P. & Kim, T. W. (2005) Model-guided microarray implicates the retromer complex in Alzheimer's disease, Ann Neurol. 58, 909-19. 152. Huse, J. T., Pijak, D. S., Leslie, G. J., Lee, V. M. & Doms, R. W. (2000) Maturation and endosomal targeting of beta-site amyloid precursor protein-cleaving enzyme. The Alzheimer's disease beta-secretase, J Biol Chem. 275, 33729-37. 153. Kinoshita, A., Fukumoto, H., Shah, T., Whelan, C. M., Irizarry, M. C. & Hyman, B. T. (2003) Demonstration by FRET of BACE interaction with the amyloid precursor protein at the cell surface and in early endosomes, J Cell Sci. 116, 3339-46. 154. Andersen, O. M., Reiche, J., Schmidt, V., Gotthardt, M., Spoelgen, R., Behlke, J., von Arnim, C. A., Breiderhoff, T., Jansen, P., Wu, X., Bales, K. R., Cappai, R., Masters, C. L., Gliemann, J., Mufson, E. J., Hyman, B. T., Paul, S. M., Nykjaer, A. & Willnow, T. E. (2005) Neuronal sorting protein- related receptor sorLA/LR11 regulates processing of the amyloid precursor protein, Proc Natl Acad Sci U S A. 102, 13461-6. 155. Rogaeva, E., Meng, Y., Lee, J. H., Gu, Y., Kawarai, T., Zou, F., Katayama, T., Baldwin, C. T., Cheng, R., Hasegawa, H., Chen, F., Shibata, N., Lunetta, K. L., Pardossi-Piquard, R., Bohm, C., Wakutani, Y., Cupples, L. A., Cuenco, K. T., Green, R. C., Pinessi, L., Rainero, I., Sorbi, S., Bruni, A., Duara, R., Friedland, R. P., Inzelberg, R., Hampe, W., Bujo, H., Song, Y. Q., Andersen, O. M., Willnow, T. E., Graff-Radford, N., Petersen, R. C., Dickson, D., Der, S. D., Fraser, P. E., Schmitt- Ulms, G., Younkin, S., Mayeux, R., Farrer, L. A. & St George-Hyslop, P. (2007) The neuronal sortilin- related receptor SORL1 is genetically associated with Alzheimer disease, Nat Genet. 39, 168-77. 156. Gustafsen, C., Glerup, S., Pallesen, L. T., Olsen, D., Andersen, O. M., Nykjaer, A., Madsen, P. & Petersen, C. M. (2013) Sortilin and SorLA display distinct roles in processing and trafficking of amyloid precursor protein, J Neurosci. 33, 64-71.

127 157. Lane, R. F., Steele, J. W., Cai, D., Ehrlich, M. E., Attie, A. D. & Gandy, S. (2013) Protein sorting motifs in the cytoplasmic tail of SorCS1 control generation of Alzheimer's amyloid-beta peptide, J Neurosci. 33, 7099-107. 158. He, X., Li, F., Chang, W. P. & Tang, J. (2005) GGA proteins mediate the recycling pathway of memapsin 2 (BACE), J Biol Chem. 280, 11696-703. 159. Mecozzi, V. J., Berman, D. E., Simoes, S., Vetanovetz, C., Awal, M. R., Patel, V. M., Schneider, R. T., Petsko, G. A., Ringe, D. & Small, S. A. (2014) Pharmacological chaperones stabilize retromer to limit APP processing, Nat Chem Biol. 10, 443-9. 160. Dragicevic, E., Schiemann, J. & Liss, B. (2015) Dopamine midbrain neurons in health and Parkinson's disease: emerging roles of voltage-gated calcium channels and ATP-sensitive potassium channels, Neuroscience. 284, 798-814. 161. Nieoullon, A. & Kerkerian-Le Goff, L. (1992) Cellular interactions in the involving neuronal systems using "classical" : possible functional implications, Mov Disord. 7, 311-25. 162. Robertson, R. G., Clarke, C. A., Boyce, S., Sambrook, M. A. & Crossman, A. R. (1990) The role of striatopallidal neurones utilizing gamma-aminobutyric acid in the pathophysiology of MPTP- induced in the primate: evidence from [3H]flunitrazepam autoradiography, Brain Res. 531, 95-104. 163. Lange, K. W., Robbins, T. W., Marsden, C. D., James, M., Owen, A. M. & Paul, G. M. (1992) L-dopa withdrawal in Parkinson's disease selectively impairs cognitive performance in tests sensitive to frontal lobe dysfunction, Psychopharmacology (Berl). 107, 394-404. 164. Lange, K. W., Paul, G. M., Naumann, M. & Gsell, W. (1995) Dopaminergic effects on cognitive performance in patients with Parkinson's disease, J Neural Transm Suppl. 46, 423-32. 165. Narayanan, N. S., Rodnitzky, R. L. & Uc, E. Y. (2013) Prefrontal dopamine signaling and cognitive symptoms of Parkinson's disease, Rev Neurosci. 24, 267-78. 166. Wakabayashi, K., Tanji, K., Mori, F. & Takahashi, H. (2007) The in Parkinson's disease: molecules implicated in the formation and degradation of alpha-synuclein aggregates, Neuropathology. 27, 494-506. 167. Schulz-Schaeffer, W. J. (2010) The synaptic pathology of alpha-synuclein aggregation in dementia with Lewy bodies, Parkinson's disease and Parkinson's disease dementia, Acta Neuropathol. 120, 131-43. 168. Baba, M., Nakajo, S., Tu, P. H., Tomita, T., Nakaya, K., Lee, V. M., Trojanowski, J. Q. & Iwatsubo, T. (1998) Aggregation of alpha-synuclein in Lewy bodies of sporadic Parkinson's disease and dementia with Lewy bodies, Am J Pathol. 152, 879-84. 169. Xu, J., Kao, S. Y., Lee, F. J., Song, W., Jin, L. W. & Yankner, B. A. (2002) Dopamine-dependent neurotoxicity of alpha-synuclein: a mechanism for selective neurodegeneration in Parkinson disease, Nat Med. 8, 600-6. 170. Zimprich, A., Benet-Pages, A., Struhal, W., Graf, E., Eck, S. H., Offman, M. N., Haubenberger, D., Spielberger, S., Schulte, E. C., Lichtner, P., Rossle, S. C., Klopp, N., Wolf, E., Seppi, K., Pirker, W., Presslauer, S., Mollenhauer, B., Katzenschlager, R., Foki, T., Hotzy, C., Reinthaler, E., Harutyunyan, A., Kralovics, R., Peters, A., Zimprich, F., Brucke, T., Poewe, W., Auff, E., Trenkwalder, C., Rost, B., Ransmayr, G., Winkelmann, J., Meitinger, T. & Strom, T. M. (2011) A mutation in VPS35, encoding a subunit of the retromer complex, causes late-onset Parkinson disease, Am J Hum Genet. 89, 168-75. 171. Vilarino-Guell, C., Wider, C., Ross, O. A., Dachsel, J. C., Kachergus, J. M., Lincoln, S. J., Soto- Ortolaza, A. I., Cobb, S. A., Wilhoite, G. J., Bacon, J. A., Behrouz, B., Melrose, H. L., Hentati, E., Puschmann, A., Evans, D. M., Conibear, E., Wasserman, W. W., Aasly, J. O., Burkhard, P. R., Djaldetti, R., Ghika, J., Hentati, F., Krygowska-Wajs, A., Lynch, T., Melamed, E., Rajput, A., Rajput, A. H., Solida, A., Wu, R. M., Uitti, R. J., Wszolek, Z. K., Vingerhoets, F. & Farrer, M. J. (2011) VPS35 mutations in Parkinson disease, Am J Hum Genet. 89, 162-7.

128 172. 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., Yoshino, H., Hatano, T., Tomiyama, H. & Hattori, N. (2012) VPS35 mutation in Japanese patients with typical Parkinson's disease, Mov Disord. 27, 1413-7. 173. Bandres-Ciga, S., Mencacci, N. E., Duran, R., Barrero, F. J., Escamilla-Sevilla, F., Morgan, S., Hehir, J., Vives, F., Hardy, J. & Pittman, A. M. (2016) Analysis of the genetic variability in Parkinson's disease from Southern Spain, Neurobiol Aging. 37, 210 e1-210 e5. 174. Blanckenberg, J., Ntsapi, C., Carr, J. A. & Bardien, S. (2014) EIF4G1 R1205H and VPS35 D620N mutations are rare in Parkinson's disease from South Africa, Neurobiol Aging. 35, 445 e1-3. 175. Guella, I., Solda, G., Cilia, R., Pezzoli, G., Asselta, R., Duga, S. & Goldwurm, S. (2012) The Asp620asn mutation in VPS35 is not a common cause of familial Parkinson's disease, Mov Disord. 27, 800-1. 176. Gustavsson, E. K., Guella, I., Trinh, J., Szu-Tu, C., Rajput, A., Rajput, A. H., Steele, J. C., McKeown, M., Jeon, B. S., Aasly, J. O. & Farrer, M. J. (2015) Genetic variability of the retromer cargo recognition complex in parkinsonism, Mov Disord. 30, 580-4. 177. Kumar, K. R., Weissbach, A., Heldmann, M., Kasten, M., Tunc, S., Sue, C. M., Svetel, M., Kostic, V. S., Segura-Aguilar, J., Ramirez, A., Simon, D. K., Vieregge, P., Munte, T. F., Hagenah, J., Klein, C. & Lohmann, K. (2012) Frequency of the D620N mutation in VPS35 in Parkinson disease, Arch Neurol. 69, 1360-4. 178. Lesage, S., Condroyer, C., Klebe, S., Honore, A., Tison, F., Brefel-Courbon, C., Durr, A., Brice, A. & French Parkinson's Disease Genetics Study, G. (2012) Identification of VPS35 mutations replicated in French families with Parkinson disease, Neurology. 78, 1449-50. 179. Shannon, B., Soto-Ortolaza, A., Rayaprolu, S., Cannon, H. D., Labbe, C., Benitez, B. A., Choi, J., Lynch, T., Boczarska-Jedynak, M., Opala, G., Krygowska-Wajs, A., Barcikowska, M., Van Gerpen, J. A., Uitti, R. J., Springer, W., Cruchaga, C., Wszolek, Z. K. & Ross, O. A. (2014) Genetic variation of the retromer subunits VPS26A/B-VPS29 in Parkinson's disease, Neurobiol Aging. 35, 1958 e1-2. 180. Sheerin, U. M., Charlesworth, G., Bras, J., Guerreiro, R., Bhatia, K., Foltynie, T., Limousin, P., Silveira-Moriyama, L., Lees, A. & Wood, N. (2012) Screening for VPS35 mutations in Parkinson's disease, Neurobiol Aging. 33, 838 e1-5. 181. Sudhaman, S., Behari, M., Govindappa, S. T., Muthane, U. B., Juyal, R. C. & Thelma, B. K. (2013) VPS35 and EIF4G1 mutations are rare in Parkinson's disease among Indians, Neurobiol Aging. 34, 2442 e1-3. 182. Zhang, Y., Chen, S., Xiao, Q., Cao, L., Liu, J., Rong, T. Y., Ma, J. F., Wang, G., Wang, Y. & Chen, S. D. (2012) Vacuolar protein sorting 35 Asp620Asn mutation is rare in the ethnic Chinese population with Parkinson's disease, Parkinsonism Relat Disord. 18, 638-40. 183. MacLeod, D. A., Rhinn, H., Kuwahara, T., Zolin, A., Di Paolo, G., McCabe, B. D., Marder, K. S., Honig, L. S., Clark, L. N., Small, S. A. & Abeliovich, A. (2013) RAB7L1 interacts with LRRK2 to modify intraneuronal protein sorting and Parkinson's disease risk, Neuron. 77, 425-39. 184. Dodson, M. W., Zhang, T., Jiang, C., Chen, S. & Guo, M. (2012) Roles of the Drosophila LRRK2 homolog in Rab7-dependent lysosomal positioning, Hum Mol Genet. 21, 1350-63. 185. Schapansky, J., Nardozzi, J. D., Felizia, F. & LaVoie, M. J. (2014) Membrane recruitment of endogenous LRRK2 precedes its potent regulation of autophagy, Hum Mol Genet. 23, 4201-14. 186. Heo, H. Y., Kim, K. S. & Seol, W. (2010) Coordinate Regulation of Neurite Outgrowth by LRRK2 and Its Interactor, Rab5, Exp Neurobiol. 19, 97-105. 187. Beilina, A., Rudenko, I. N., Kaganovich, A., Civiero, L., Chau, H., Kalia, S. K., Kalia, L. V., Lobbestael, E., Chia, R., Ndukwe, K., Ding, J., Nalls, M. A., International Parkinson's Disease Genomics, C., North American Brain Expression, C., Olszewski, M., Hauser, D. N., Kumaran, R., Lozano, A. M., Baekelandt, V., Greene, L. E., Taymans, J. M., Greggio, E. & Cookson, M. R. (2014) Unbiased screen for interactors of leucine-rich repeat kinase 2 supports a common pathway for sporadic and familial Parkinson disease, Proc Natl Acad Sci U S A. 111, 2626-31.

129 188. Linhart, R., Wong, S. A., Cao, J., Tran, M., Huynh, A., Ardrey, C., Park, J. M., Hsu, C., Taha, S., Peterson, R., Shea, S., Kurian, J. & Venderova, K. (2014) 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. 9, 23. 189. Cullen, V., Lindfors, M., Ng, J., Paetau, A., Swinton, E., Kolodziej, P., Boston, H., Saftig, P., Woulfe, J., Feany, M. B., Myllykangas, L., Schlossmacher, M. G. & Tyynela, J. (2009) Cathepsin D expression level affects alpha-synuclein processing, aggregation, and toxicity in vivo, Mol Brain. 2, 5. 190. Sevlever, D., Jiang, P. & Yen, S. H. (2008) Cathepsin D is the main lysosomal enzyme involved in the degradation of alpha-synuclein and generation of its carboxy-terminally truncated species, Biochemistry. 47, 9678-87. 191. Miura, E., Hasegawa, T., Konno, M., Suzuki, M., Sugeno, N., Fujikake, N., Geisler, S., Tabuchi, M., Oshima, R., Kikuchi, A., Baba, T., Wada, K., Nagai, Y., Takeda, A. & Aoki, M. (2014) VPS35 dysfunction impairs lysosomal degradation of alpha-synuclein and exacerbates neurotoxicity in a Drosophila model of Parkinson's disease, Neurobiol Dis. 71, 1-13. 192. McGough, I. J., Steinberg, F., Jia, D., Barbuti, P. A., McMillan, K. J., Heesom, K. J., Whone, A. L., Caldwell, M. A., Billadeau, D. D., Rosen, M. K. & Cullen, P. J. (2014) Retromer binding to FAM21 and the WASH complex is perturbed by the Parkinson disease-linked VPS35(D620N) mutation, Curr Biol. 24, 1670-6. 193. Follett, J., Norwood, S. J., Hamilton, N. A., Mohan, M., Kovtun, O., Tay, S., Zhe, Y., Wood, S. A., Mellick, G. D., Silburn, P. A., Collins, B. M., Bugarcic, A. & Teasdale, R. D. (2014) The Vps35 D620N mutation linked to Parkinson's disease disrupts the cargo sorting function of retromer, Traffic. 15, 230-44. 194. Zavodszky, E., Seaman, M. N., Moreau, K., Jimenez-Sanchez, M., Breusegem, S. Y., Harbour, M. E. & Rubinsztein, D. C. (2014) Mutation in VPS35 associated with Parkinson's disease impairs WASH complex association and inhibits autophagy, Nat Commun. 5, 3828. 195. Follett, J., Bugarcic, A., Yang, Z., Ariotti, N., Norwood, S. J., Collins, B. M., Parton, R. G. & Teasdale, R. D. (2016) Parkinson Disease-linked Vps35 R524W Mutation Impairs the Endosomal Association of Retromer and Induces alpha-Synuclein Aggregation, J Biol Chem. 291, 18283-98. 196. Villalba, R. M. & Smith, Y. (2013) Differential striatal spine pathology in Parkinson's disease and cocaine addiction: a key role of dopamine?, Neuroscience. 251, 2-20. 197. Munsie, L. N., Milnerwood, A. J., Seibler, P., Beccano-Kelly, D. A., Tatarnikov, I., Khinda, J., Volta, M., Kadgien, C., Cao, L. P., Tapia, L., Klein, C. & Farrer, M. J. (2015) Retromer-dependent neurotransmitter receptor trafficking to is altered by the Parkinson's disease VPS35 mutation p.D620N, Hum Mol Genet. 24, 1691-703. 198. Hasbi, A., O'Dowd, B. F. & George, S. R. (2011) Dopamine D1-D2 receptor heteromer signaling pathway in the brain: emerging physiological relevance, Mol Brain. 4, 26. 199. Wang, C., Niu, M., Zhou, Z., Zheng, X., Zhang, L., Tian, Y., Yu, X., Bu, G., Xu, H., Ma, Q. & Zhang, Y. W. (2016) VPS35 regulates cell surface recycling and signaling of dopamine receptor D1, Neurobiol Aging. 46, 22-31. 200. Ishizu, N., Yui, D., Hebisawa, A., Aizawa, H., Cui, W., Fujita, Y., Hashimoto, K., Ajioka, I., Mizusawa, H., Yokota, T. & Watase, K. (2016) Impaired striatal dopamine release in homozygous Vps35 D620N knock-in mice, Hum Mol Genet. 25, 4507-4517. 201. Howells, D. W., Porritt, M. J., Wong, J. Y., Batchelor, P. E., Kalnins, R., Hughes, A. J. & Donnan, G. A. (2000) Reduced BDNF mRNA expression in the Parkinson's disease substantia nigra, Exp Neurol. 166, 127-35. 202. Tejeda, G. S. & Diaz-Guerra, M. (2017) Integral Characterization of Defective BDNF/TrkB Signalling in Neurological and Psychiatric Disorders Leads the Way to New Therapies, Int J Mol Sci. 18.

130 203. Vaegter, C. B., Jansen, P., Fjorback, A. W., Glerup, S., Skeldal, S., Kjolby, M., Richner, M., Erdmann, B., Nyengaard, J. R., Tessarollo, L., Lewin, G. R., Willnow, T. E., Chao, M. V. & Nykjaer, A. (2011) Sortilin associates with Trk receptors to enhance anterograde transport and neurotrophin signaling, Nat Neurosci. 14, 54-61. 204. Braschi, E., Goyon, V., Zunino, R., Mohanty, A., Xu, L. & McBride, H. M. (2010) Vps35 mediates vesicle transport between the mitochondria and peroxisomes, Curr Biol. 20, 1310-5. 205. Bi, F., Li, F., Huang, C. & Zhou, H. (2013) Pathogenic mutation in VPS35 impairs its protection against MPP(+) cytotoxicity, Int J Biol Sci. 9, 149-55. 206. Wang, W., Wang, X., Fujioka, H., Hoppel, C., Whone, A. L., Caldwell, M. A., Cullen, P. J., Liu, J. & Zhu, X. (2016) Parkinson's disease-associated mutant VPS35 causes mitochondrial dysfunction by recycling DLP1 complexes, Nat Med. 22, 54-63. 207. Tang, F. L., Liu, W., Hu, J. X., Erion, J. R., Ye, J., Mei, L. & Xiong, W. C. (2015) VPS35 Deficiency or Mutation Causes Dopaminergic Neuronal Loss by Impairing and Function, Cell Rep. 12, 1631-43. 208. Singleton, A. B., Farrer, M., Johnson, J., Singleton, A., Hague, S., Kachergus, J., Hulihan, M., Peuralinna, T., Dutra, A., Nussbaum, R., Lincoln, S., Crawley, A., Hanson, M., Maraganore, D., Adler, C., Cookson, M. R., Muenter, M., Baptista, M., Miller, D., Blancato, J., Hardy, J. & Gwinn-Hardy, K. (2003) alpha-Synuclein triplication causes Parkinson's disease, Science. 302, 841. 209. Polymeropoulos, M. H., Lavedan, C., Leroy, E., Ide, S. E., Dehejia, A., Dutra, A., Pike, B., Root, H., Rubenstein, J., Boyer, R., Stenroos, E. S., Chandrasekharappa, S., Athanassiadou, A., Papapetropoulos, T., Johnson, W. G., Lazzarini, A. M., Duvoisin, R. C., Di Iorio, G., Golbe, L. I. & Nussbaum, R. L. (1997) Mutation in the alpha-synuclein gene identified in families with Parkinson's disease, Science. 276, 2045-7. 210. Kruger, R., Kuhn, W., Muller, T., Woitalla, D., Graeber, M., Kosel, S., Przuntek, H., Epplen, J. T., Schols, L. & Riess, O. (1998) Ala30Pro mutation in the gene encoding alpha-synuclein in Parkinson's disease, Nat Genet. 18, 106-8. 211. Zarranz, J. J., Alegre, J., Gomez-Esteban, J. C., Lezcano, E., Ros, R., Ampuero, I., Vidal, L., Hoenicka, J., Rodriguez, O., Atares, B., Llorens, V., Gomez Tortosa, E., del Ser, T., Munoz, D. G. & de Yebenes, J. G. (2004) The new mutation, E46K, of alpha-synuclein causes Parkinson and Lewy body dementia, Ann Neurol. 55, 164-73. 212. Dhungel, N., Eleuteri, S., Li, L. B., Kramer, N. J., Chartron, J. W., Spencer, B., Kosberg, K., Fields, J. A., Stafa, K., Adame, A., Lashuel, H., Frydman, J., Shen, K., Masliah, E. & Gitler, A. D. (2015) Parkinson's disease genes VPS35 and EIF4G1 interact genetically and converge on alpha- synuclein, Neuron. 85, 76-87. 213. Khurana, V., Peng, J., Chung, C. Y., Auluck, P. K., Fanning, S., Tardiff, D. F., Bartels, T., Koeva, M., Eichhorn, S. W., Benyamini, H., Lou, Y., Nutter-Upham, A., Baru, V., Freyzon, Y., Tuncbag, N., Costanzo, M., San Luis, B. J., Schondorf, D. C., Barrasa, M. I., Ehsani, S., Sanjana, N., Zhong, Q., Gasser, T., Bartel, D. P., Vidal, M., Deleidi, M., Boone, C., Fraenkel, E., Berger, B. & Lindquist, S. (2017) Genome-Scale Networks Link Neurodegenerative Disease Genes to alpha- Synuclein through Specific Molecular Pathways, Cell Syst. 4, 157-170 e14. 214. Chung, C. Y., Khurana, V., Yi, S., Sahni, N., Loh, K. H., Auluck, P. K., Baru, V., Udeshi, N. D., Freyzon, Y., Carr, S. A., Hill, D. E., Vidal, M., Ting, A. Y. & Lindquist, S. (2017) In Situ Peroxidase Labeling and Mass-Spectrometry Connects Alpha-Synuclein Directly to Endocytic Trafficking and mRNA Metabolism in Neurons, Cell Syst. 4, 242-250 e4. 215. Paisan-Ruiz, C., Lewis, P. A. & Singleton, A. B. (2013) LRRK2: cause, risk, and mechanism, J Parkinsons Dis. 3, 85-103. 216. Tong, Y., Yamaguchi, H., Giaime, E., Boyle, S., Kopan, R., Kelleher, R. J., 3rd & Shen, J. (2010) Loss of leucine-rich repeat kinase 2 causes impairment of protein degradation pathways, accumulation of alpha-synuclein, and apoptotic cell death in aged mice, Proc Natl Acad Sci U S A. 107, 9879-84.

131 217. Xiong, Y., Coombes, C. E., Kilaru, A., Li, X., Gitler, A. D., Bowers, W. J., Dawson, V. L., Dawson, T. M. & Moore, D. J. (2010) GTPase activity plays a key role in the pathobiology of LRRK2, PLoS Genet. 6, e1000902. 218. Orenstein, S. J., Kuo, S. H., Tasset, I., Arias, E., Koga, H., Fernandez-Carasa, I., Cortes, E., Honig, L. S., Dauer, W., Consiglio, A., Raya, A., Sulzer, D. & Cuervo, A. M. (2013) Interplay of LRRK2 with chaperone-mediated autophagy, Nat Neurosci. 16, 394-406. 219. Dodson, M. W., Leung, L. K., Lone, M., Lizzio, M. A. & Guo, M. (2014) Novel ethyl methanesulfonate (EMS)-induced null alleles of the Drosophila homolog of LRRK2 reveal a crucial role in endolysosomal functions and autophagy in vivo, Dis Model Mech. 7, 1351-63. 220. Mir, R., Tonelli, F., Lis, P., Macartney, T., Polinski, N. K., Martinez, T. N., Chou, M. Y., Howden, A. J. M., Konig, T., Hotzy, C., Milenkovic, I., Brucke, T., Zimprich, A., Sammler, E. & Alessi, D. R. (2018) The Parkinson's disease VPS35[D620N] mutation enhances LRRK2-mediated Rab in mouse and human, The Biochemical journal. 475, 1861-1883. 221. Kiral, F. R., Kohrs, F. E., Jin, E. J. & Hiesinger, P. R. (2018) Rab GTPases and Membrane Trafficking in Neurodegeneration, Curr Biol. 28, R471-R486. 222. Inoshita, T., Arano, T., Hosaka, Y., Meng, H., Umezaki, Y., Kosugi, S., Morimoto, T., Koike, M., Chang, H. Y., Imai, Y. & Hattori, N. (2017) Vps35 in cooperation with LRRK2 regulates endocytosis through the endosomal pathway in Drosophila, Hum Mol Genet. 26, 2933-2948. 223. Malik, B. R., Godena, V. K. & Whitworth, A. J. (2015) VPS35 pathogenic mutations confer no dominant toxicity but partial loss of function in Drosophila and genetically interact with parkin, Hum Mol Genet. 24, 6106-17. 224. Martinez, A., Lectez, B., Ramirez, J., Popp, O., Sutherland, J. D., Urbe, S., Dittmar, G., Clague, M. J. & Mayor, U. (2017) Quantitative proteomic analysis of Parkin substrates in Drosophila neurons, Mol Neurodegener. 12, 29. 225. Song, P., Trajkovic, K., Tsunemi, T. & Krainc, D. (2016) Parkin Modulates Endosomal Organization and Function of the Endo-Lysosomal Pathway, J Neurosci. 36, 2425-37. 226. Kalinderi, K., Bostantjopoulou, S., Katsarou, Z., Dimikiotou, M. & Fidani, L. (2015) D620N mutation in the VPS35 gene and R1205H mutation in the EIF4G1 gene are uncommon in the Greek population, Neurosci Lett. 606, 113-6. 227. Nuytemans, K., Bademci, G., Inchausti, V., Dressen, A., Kinnamon, D. D., Mehta, A., Wang, L., Zuchner, S., Beecham, G. W., Martin, E. R., Scott, W. K. & Vance, J. M. (2013) Whole exome sequencing of rare variants in EIF4G1 and VPS35 in Parkinson disease, Neurology. 80, 982-9. 228. Verstraeten, A., Wauters, E., Crosiers, D., Meeus, B., Corsmit, E., Elinck, E., Mattheijssens, M., Peeters, K., Cras, P., Pickut, B., Vandenberghe, R., Engelborghs, S., De Deyn, P. P., Van Broeckhoven, C. & Theuns, J. (2012) Contribution of VPS35 genetic variability to LBD in the Flanders-Belgian population, Neurobiol Aging. 33, 1844 e11-3. 229. Sharma, M., Ioannidis, J. P., Aasly, J. O., Annesi, G., Brice, A., Bertram, L., Bozi, M., Barcikowska, M., Crosiers, D., Clarke, C. E., Facheris, M. F., Farrer, M., Garraux, G., Gispert, S., Auburger, G., Vilarino-Guell, C., Hadjigeorgiou, G. M., Hicks, A. A., Hattori, N., Jeon, B. S., Jamrozik, Z., Krygowska-Wajs, A., Lesage, S., Lill, C. M., Lin, J. J., Lynch, T., Lichtner, P., Lang, A. E., Libioulle, C., Murata, M., Mok, V., Jasinska-Myga, B., Mellick, G. D., Morrison, K. E., Meitnger, T., Zimprich, A., Opala, G., Pramstaller, P. P., Pichler, I., Park, S. S., Quattrone, A., Rogaeva, E., Ross, O. A., Stefanis, L., Stockton, J. D., Satake, W., Silburn, P. A., Strom, T. M., Theuns, J., Tan, E. K., Toda, T., Tomiyama, H., Uitti, R. J., Van Broeckhoven, C., Wirdefeldt, K., Wszolek, Z., Xiromerisiou, G., Yomono, H. S., Yueh, K. C., Zhao, Y., Gasser, T., Maraganore, D., Kruger, R. & consortium, G. (2012) A multi-centre clinico-genetic analysis of the VPS35 gene in Parkinson disease indicates reduced penetrance for disease-associated variants, J Med Genet. 49, 721-6. 230. Seaman, M. & Freeman, C. L. (2014) Analysis of the Retromer complex-WASH complex interaction illuminates new avenues to explore in Parkinson disease, Commun Integr Biol. 7, e29483.

132 231. Saftig, P. & Klumperman, J. (2009) Lysosome biogenesis and lysosomal membrane proteins: trafficking meets function, Nat Rev Mol Cell Biol. 10, 623-35. 232. Settembre, C., Fraldi, A., Medina, D. L. & Ballabio, A. (2013) Signals from the lysosome: a control centre for cellular clearance and energy metabolism, Nat Rev Mol Cell Biol. 14, 283-96. 233. Ghosh, P., Dahms, N. M. & Kornfeld, S. (2003) Mannose 6-phosphate receptors: new twists in the tale, Nat Rev Mol Cell Biol. 4, 202-12. 234. Collins, B. M., Norwood, S. J., Kerr, M. C., Mahony, D., Seaman, M. N., Teasdale, R. D. & Owen, D. J. (2008) Structure of Vps26B and mapping of its interaction with the retromer protein complex, Traffic. 9, 366-79. 235. Seaman, M. N. (2012) The retromer complex - endosomal protein recycling and beyond, Journal of cell science. 125, 4693-702. 236. Tammineni, P., Jeong, Y. Y., Feng, T., Aikal, D. & Cai, Q. (2017) Impaired axonal retrograde trafficking of the retromer complex augments lysosomal deficits in Alzheimer's disease neurons, Hum Mol Genet. 26, 4352-4366. 237. Seaman, M. N. J. (2018) Retromer and the cation-independent mannose 6-phosphate receptor-Time for a trial separation?, Traffic. 19, 150-152. 238. Seaman, M. N. (2004) Cargo-selective endosomal sorting for retrieval to the Golgi requires retromer, The Journal of cell biology. 165, 111-22. 239. Osborne, D. G., Phillips-Krawczak, C. A. & Billadeau, D. D. (2015) Monitoring receptor trafficking following retromer and WASH deregulation, Methods Cell Biol. 130, 199-213. 240. McKenzie, J. E., Raisley, B., Zhou, X., Naslavsky, N., Taguchi, T., Caplan, S. & Sheff, D. (2012) Retromer guides STxB and CD8-M6PR from early to recycling endosomes, EHD1 guides STxB from recycling endosome to Golgi, Traffic. 13, 1140-59. 241. Kingston, D., Chang, H., Ensser, A., Lee, H. R., Lee, J., Lee, S. H., Jung, J. U. & Cho, N. H. (2011) Inhibition of retromer activity by herpesvirus saimiri tip leads to CD4 downregulation and efficient T cell transformation, J Virol. 85, 10627-38. 242. Hirst, J., Itzhak, D. N., Antrobus, R., Borner, G. H. H. & Robinson, M. S. (2018) Role of the AP- 5 adaptor protein complex in late endosome-to-Golgi retrieval, PLoS Biol. 16, e2004411. 243. Bulankina, A. V., Deggerich, A., Wenzel, D., Mutenda, K., Wittmann, J. G., Rudolph, M. G., Burger, K. N. & Honing, S. (2009) TIP47 functions in the biogenesis of lipid droplets, J Cell Biol. 185, 641-55. 244. Breusegem, S. Y. & Seaman, M. N. (2014) Genome-wide RNAi screen reveals a role for multipass membrane proteins in endosome-to-golgi retrieval, Cell Rep. 9, 1931-45. 245. Carlton, J., Bujny, M., Peter, B. J., Oorschot, V. M., Rutherford, A., Mellor, H., Klumperman, J., McMahon, H. T. & Cullen, P. J. (2004) Sorting nexin-1 mediates tubular endosome-to-TGN transport through coincidence sensing of high- curvature membranes and 3-phosphoinositides, Curr Biol. 14, 1791-800. 246. Hara, S., Kiyokawa, E., Iemura, S., Natsume, T., Wassmer, T., Cullen, P. J., Hiai, H. & Matsuda, M. (2008) The DHR1 domain of DOCK180 binds to SNX5 and regulates cation-independent mannose 6-phosphate receptor transport, Mol Biol Cell. 19, 3823-35. 247. Brocker, C., Engelbrecht-Vandre, S. & Ungermann, C. (2010) Multisubunit tethering complexes and their role in membrane fusion, Curr Biol. 20, R943-52. 248. Gillingham, A. K. & Munro, S. (2016) Finding the Golgi: Golgin Coiled-Coil Proteins Show the Way, Trends in cell biology. 26, 399-408. 249. Lieu, Z. Z., Derby, M. C., Teasdale, R. D., Hart, C., Gunn, P. & Gleeson, P. A. (2007) The golgin GCC88 is required for efficient retrograde transport of cargo from the early endosomes to the trans-Golgi network, Mol Biol Cell. 18, 4979-91.

133 250. Kerr, M. C., Gomez, G. A., Ferguson, C., Tanzer, M. C., Murphy, J. M., Yap, A. S., Parton, R. G., Huston, W. M. & Teasdale, R. D. (2017) Laser-mediated rupture of chlamydial inclusions triggers pathogen egress and host cell necrosis, Nat Commun. 8, 14729. 251. Clairfeuille, T., Mas, C., Chan, A. S., Yang, Z., Tello-Lafoz, M., Chandra, M., Widagdo, J., Kerr, M. C., Paul, B., Merida, I., Teasdale, R. D., Pavlos, N. J., Anggono, V. & Collins, B. M. (2016) A molecular code for endosomal recycling of phosphorylated cargos by the SNX27-retromer complex, Nat Struct Mol Biol. 23, 921-932. 252. Leaback, D. H. & Walker, P. G. (1961) Studies on glucosaminidase. 4. The fluorimetric assay of N-acetyl-beta-glucosaminidase, The Biochemical journal. 78, 151-6. 253. Kolodny, E. H. & Mumford, R. A. (1976) Human leukocyte acid hydrolases: characterization of eleven lysosomal enzymes and study of reaction conditions for their automated analysis, Clin Chim Acta. 70, 247-57. 254. Yang, Z., Hong, L. K., Follett, J., Wabitsch, M., Hamilton, N. A., Collins, B. M., Bugarcic, A. & Teasdale, R. D. (2015) Functional characterization of retromer in GLUT4 storage vesicle formation and adipocyte differentiation, FASEB journal : official publication of the Federation of American Societies for Experimental Biology. 255. Bolte, S. & Cordelieres, F. P. (2006) A guided tour into subcellular colocalization analysis in light microscopy, J Microsc. 224, 213-32. 256. Parton, R. G., Prydz, K., Bomsel, M., Simons, K. & Griffiths, G. (1989) Meeting of the apical and basolateral endocytic pathways of the Madin-Darby canine cell in late endosomes, J Cell Biol. 109, 3259-72. 257. Maruzs, T., Lorincz, P., Szatmari, Z., Szeplaki, S., Sandor, Z., Lakatos, Z., Puska, G., Juhasz, G. & Sass, M. (2015) Retromer Ensures the Degradation of Autophagic Cargo by Maintaining Lysosome Function in Drosophila, Traffic. 16, 1088-107. 258. Mizushima, N., Yoshimori, T. & Levine, B. (2010) Methods in mammalian autophagy research, Cell. 140, 313-26. 259. Markmann, S., Thelen, M., Cornils, K., Schweizer, M., Brocke-Ahmadinejad, N., Willnow, T., Heeren, J., Gieselmann, V., Braulke, T. & Kollmann, K. (2015) Lrp1/LDL Receptor Play Critical Roles in Mannose 6-Phosphate-Independent Lysosomal Enzyme Targeting, Traffic. 16, 743-59. 260. Kerr, M. C., Lindsay, M. R., Luetterforst, R., Hamilton, N., Simpson, F., Parton, R. G., Gleeson, P. A. & Teasdale, R. D. (2006) Visualisation of macropinosome maturation by the recruitment of sorting nexins, J Cell Sci. 119, 3967-80. 261. Elwell, C. A., Czudnochowski, N., von Dollen, J., Johnson, J. R., Nakagawa, R., Mirrashidi, K., Krogan, N. J., Engel, J. N. & Rosenberg, O. S. (2017) Chlamydia interfere with an interaction between the mannose-6-phosphate receptor and sorting nexins to counteract host restriction, Elife. 6. 262. Wang, S., Tan, K. L., Agosto, M. A., Xiong, B., Yamamoto, S., Sandoval, H., Jaiswal, M., Bayat, V., Zhang, K., Charng, W. L., David, G., Duraine, L., Venkatachalam, K., Wensel, T. G. & Bellen, H. J. (2014) The retromer complex is required for rhodopsin recycling and its loss leads to photoreceptor degeneration, PLoS Biol. 12, e1001847. 263. Bright, N. A., Davis, L. J. & Luzio, J. P. (2016) Endolysosomes Are the Principal Intracellular Sites of Acid Hydrolase Activity, Curr Biol. 26, 2233-45. 264. Goberdhan, D. C., Wilson, C. & Harris, A. L. (2016) Amino Acid Sensing by mTORC1: Intracellular Transporters Mark the Spot, Cell Metab. 23, 580-9. 265. Kollmann, K., Damme, M., Markmann, S., Morelle, W., Schweizer, M., Hermans-Borgmeyer, I., Rochert, A. K., Pohl, S., Lubke, T., Michalski, J. C., Kakela, R., Walkley, S. U. & Braulke, T. (2012) Lysosomal dysfunction causes neurodegeneration in mucolipidosis II 'knock-in' mice, Brain. 135, 2661-75. 266. Dahms, N. M., Olson, L. J. & Kim, J. J. (2008) Strategies for carbohydrate recognition by the mannose 6-phosphate receptors, Glycobiology. 18, 664-78. 134 267. Lieu, Z. Z. & Gleeson, P. A. (2010) Identification of different itineraries and retromer components for endosome-to-Golgi transport of TGN38 and Shiga toxin, Eur J Cell Biol. 89, 379-93. 268. Harbour, M. E., Breusegem, S. Y., Antrobus, R., Freeman, C., Reid, E. & Seaman, M. N. (2010) The cargo-selective retromer complex is a recruiting hub for protein complexes that regulate endosomal tubule dynamics, Journal of cell science. 123, 3703-17. 269. Vardarajan, B. N., Bruesegem, S. Y., Harbour, M. E., Inzelberg, R., Friedland, R., St George- Hyslop, P., Seaman, M. N. & Farrer, L. A. (2012) Identification of Alzheimer disease-associated variants in genes that regulate retromer function, Neurobiol Aging. 33, 2231 e15-2231 e30. 270. Lenoir, M., Ustunel, C., Rajesh, S., Kaur, J., Moreau, D., Gruenberg, J. & Overduin, M. (2018) Phosphorylation of conserved phosphoinositide binding pocket regulates sorting nexin membrane targeting, Nat Commun. 9, 993. 271. Gomez, T. S. & Billadeau, D. D. (2009) A FAM21-containing WASH complex regulates retromer-dependent sorting, Dev Cell. 17, 699-711. 272. Harbour, M. E., Breusegem, S. Y. & Seaman, M. N. (2012) Recruitment of the endosomal WASH complex is mediated by the extended 'tail' of Fam21 binding to the retromer protein Vps35, The Biochemical journal. 442, 209-20. 273. Jia, D., Gomez, T. S., Billadeau, D. D. & Rosen, M. K. (2012) Multiple repeat elements within the FAM21 tail link the WASH actin regulatory complex to the retromer, Molecular biology of the cell. 23, 2352-61. 274. Witkos, T. M. & Lowe, M. (2017) Recognition and tethering of transport vesicles at the Golgi apparatus, Current opinion in cell biology. 47, 16-23. 275. Derby, M. C., Lieu, Z. Z., Brown, D., Stow, J. L., Goud, B. & Gleeson, P. A. (2007) The trans- Golgi network golgin, GCC185, is required for endosome-to-Golgi transport and maintenance of Golgi structure, Traffic. 8, 758-73. 276. Tang, F. L., Erion, J. R., Tian, Y., Liu, W., Yin, D. M., Ye, J., Tang, B., Mei, L. & Xiong, W. C. (2015) VPS35 in Dopamine Neurons Is Required for Endosome-to-Golgi Retrieval of Lamp2a, a Receptor of Chaperone-Mediated Autophagy That Is Critical for alpha-Synuclein Degradation and Prevention of Pathogenesis of Parkinson's Disease, J Neurosci. 35, 10613-28. 277. Swiech, L., Heidenreich, M., Banerjee, A., Habib, N., Li, Y., Trombetta, J., Sur, M. & Zhang, F. (2015) In vivo interrogation of gene function in the mammalian brain using CRISPR-Cas9, Nat Biotechnol. 33, 102-6. 278. Goud, B. & Gleeson, P. A. (2010) TGN golgins, Rabs and cytoskeleton: regulating the Golgi trafficking highways, Trends in cell biology. 20, 329-36. 279. Gosavi, P., Houghton, F. J., McMillan, P. J., Hanssen, E. & Gleeson, P. A. (2018) The Golgi ribbon in mammalian cells negatively regulates autophagy by modulating mTOR activity, J Cell Sci. 131. 280. Yoshino, A., Setty, S. R., Poynton, C., Whiteman, E. L., Saint-Pol, A., Burd, C. G., Johannes, L., Holzbaur, E. L., Koval, M., McCaffery, J. M. & Marks, M. S. (2005) tGolgin-1 (p230, golgin-245) modulates Shiga-toxin transport to the Golgi and Golgi motility towards the microtubule-organizing centre, J Cell Sci. 118, 2279-93. 281. Rabouille, C. & Haase, G. (2015) Editorial: Golgi Pathology in Neurodegenerative Diseases, Front Neurosci. 9, 489. 282. Collins, B. M. (2008) The structure and function of the retromer protein complex, Traffic. 9, 1811-22. 283. Kvainickas, A., Orgaz, A. J., Nagele, H., Diedrich, B., Heesom, K. J., Dengjel, J., Cullen, P. J. & Steinberg, F. (2017) Retromer- and WASH-dependent sorting of nutrient transporters requires a multivalent interaction network with ANKRD50, Journal of cell science. 130, 382-395.

135 284. Chen, X., Kordich, J. K., Williams, E. T., Levine, N., Cole-Strauss, A., Marshall, L., Labrie, V., Ma, J., Lipton, J. W. & Moore, D. J. (2019) Parkinson's disease-linked D620N VPS35 knockin mice manifest tau neuropathology and dopaminergic neurodegeneration, Proc Natl Acad Sci U S A. 285. Cataldi, S., Follett, J., Fox, J. D., Tatarnikov, I., Kadgien, C., Gustavsson, E. K., Khinda, J., Milnerwood, A. J. & Farrer, M. J. (2018) Altered dopamine release and monoamine transporters in Vps35 p.D620N knock-in mice, NPJ Parkinsons Dis. 4, 27. 286. Cui, Y., Yang, Z. & Teasdale, R. D. (2018) The functional roles of retromer in Parkinson's disease, FEBS Lett. 592, 1096-1112. 287. Seaman, M. N. J., Mukadam, A. S. & Breusegem, S. Y. (2018) Inhibition of TBC1D5 activates Rab7a and can enhance the function of the retromer cargo-selective complex, Journal of cell science. 131. 288. Jimenez-Orgaz, A., Kvainickas, A., Nagele, H., Denner, J., Eimer, S., Dengjel, J. & Steinberg, F. (2018) Control of RAB7 activity and localization through the retromer-TBC1D5 complex enables RAB7-dependent mitophagy, EMBO J. 37, 235-254. 289. Lucas, M., Gershlick, David C., Vidaurrazaga, A., Rojas, Adriana L., Bonifacino, Juan S. & Hierro, A. (2016) Structural Mechanism for Cargo Recognition by the Retromer Complex, Cell. 290. Simonetti, B., Danson, C. M., Heesom, K. J. & Cullen, P. J. (2017) Sequence-dependent cargo recognition by SNX-BARs mediates retromer-independent transport of CI-MPR, The Journal of cell biology. 216, 3695-3712. 291. Fuse, A., Furuya, N., Kakuta, S., Inose, A., Sato, M., Koike, M., Saiki, S. & Hattori, N. (2015) VPS29-VPS35 intermediate of retromer is stable and may be involved in the retromer complex assembly process, FEBS Lett. 589, 1430-6. 292. Fukuda, M. (2016) Multiple Roles of VARP in Endosomal Trafficking: Rabs, Retromer Components and R-SNARE VAMP7 Meet on VARP, Traffic. 17, 709-19. 293. Helfer, E., Harbour, M. E., Henriot, V., Lakisic, G., Sousa-Blin, C., Volceanov, L., Seaman, M. N. & Gautreau, A. (2013) Endosomal recruitment of the WASH complex: active sequences and mutations impairing interaction with the retromer, Biol Cell. 105, 191-207. 294. Kovtun, O., Leneva, N., Bykov, Y. S., Ariotti, N., Teasdale, R. D., Schaffer, M., Engel, B. D., Owen, D. J., Briggs, J. A. G. & Collins, B. M. (2018) Structure of the membrane-assembled retromer coat determined by cryo-electron tomography, Nature. 561, 561-564. 295. Williams, E. T., Glauser, L., Tsika, E., Jiang, H., Islam, S. & Moore, D. J. (2018) Parkin mediates the ubiquitination of VPS35 and modulates retromer-dependent endosomal sorting, Hum Mol Genet. 27, 3189-3205. 296. Hao, Y. H., Doyle, J. M., Ramanathan, S., Gomez, T. S., Jia, D., Xu, M., Chen, Z. J., Billadeau, D. D., Rosen, M. K. & Potts, P. R. (2013) Regulation of WASH-dependent actin polymerization and protein trafficking by ubiquitination, Cell. 152, 1051-64.

136