Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Using large-scale screens to identify novel regulators of endosomal trafficking in yeast Dalton, Lauren 2016

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2017_february_dalton_lauren.pdf [ 24.33MB ]
Metadata
JSON: 24-1.0340570.json
JSON-LD: 24-1.0340570-ld.json
RDF/XML (Pretty): 24-1.0340570-rdf.xml
RDF/JSON: 24-1.0340570-rdf.json
Turtle: 24-1.0340570-turtle.txt
N-Triples: 24-1.0340570-rdf-ntriples.txt
Original Record: 24-1.0340570-source.json
Full Text
24-1.0340570-fulltext.txt
Citation
24-1.0340570.ris

Full Text

USING LARGE-SCALE SCREENS TO UNCOVER NOVEL REGULATORS OF ENDOSOMAL TRAFFICKING IN YEAST  by  Lauren Dalton  B.Sc., Pacific Lutheran University, 2011  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Biochemistry and Molecular Biology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  December 2016 © Lauren Dalton, 2016 ii  Abstract Cellular protein trafficking, the concerted action of moving proteins to the appropriate cellular location is important for the proper functioning of the cell. Proteins are sorted at hubs such as the endosome, where they are targeted either for degradation or for recycling to the Golgi. Yeast is an excellent model organism to study protein trafficking due to the conservation with higher eukaryotes and the ease of genetic manipulation. In this thesis, two different high-throughput approaches were used to study the protein machinery that controls the yeast endosomal sorting network. First, correlation analysis, which compares genome-wide phenotypic profiles, was used to uncover new aspects of endosomal sorting and two specific examples were chosen for further study. In the first example, we explored positive and negative genetic interaction profile correlations for members of the uncharacterized yeast complex, BLOC, which suggested a role in endosomal sorting. In particular, we found that BLOC was needed for proper downregulation of a manganese transporter similar to ESCRT (Endosomal Sorting Complex Required for Transport, a well-studied endosomal sorting complex. In the second example, we explored the predictive value of negatively correlated genetic interaction profiles for gene mutants within known protein complexes and found that these negative correlations described two types of regulatory interactions between the resulting proteins, direct inhibition and competition for shared subunits. We then showed that the previously uncharacterized VID Associated Factor 1 (Vaf1) was negatively correlated with respect to its genetic interactions and downregulated by the VID complex. In summary, correlation analysis provides a robust tool to identify the functional relationship between proteins. Second, a quantitative genome-wide endosomal sorting screen followed by a secondary high-throughput microscopy screen, uncovered novel endosomal regulators of the flippase, Neo1. Loss of these regulators caused three phenotypic outcomes:  reduced recycling from endosomes, delayed endosomal progression, or reduced colocalization with known binding partners. In particular, a short motif in the N-terminus of Neo1 was found to be necessary for Snx3 dependent sorting, correct sorting of other Snx3 cargos, and full Neo1 function. Overall, this study illuminates the power of high-throughput screens to discover new regulators of endosomal sorting.  iii  Preface Chapter 2 and Chapter 3 reflect work done in collaboration with Shuye Pu, Mohan Babu, James Vlasblom, Andrew Emili, Jack F. Greenblatt, and Shoshana J. Wodak who developed the genetic correlation software used as the basis for this study. Networks based on this analysis appear in Figures 2.2, 2.4, 3.1, 3.2, 3.3, and 3.4 and Table 3.1. I was responsible, with input and direction from Elizabeth Conibear, for experimental design, execution, and analysis for all figures with the following exceptions. Jeffery Tong conducted the microscopy in Figure 2.1C. In addition, Mike Davey performed the invertase overlay experiment for Figure 2.1A and constructed some of the strains used in this study. A paper corroborating much of the work in Chapter 2 was published while the manuscript was in progress. A version of Chapter 3 is being prepared for publication. A version of Chapter 4 is in preparation with coauthors Bjorn Bean, Mike Davey, and Elizabeth Conibear. I was responsible for writing the manuscript as well as experimental design, execution, and analysis in all figures with input from Elizabeth Conibear and with the following exceptions. While I made the final figures, Bjorn Bean developed the in-house image analysis journals. Mike Davey was instrumental in the design and construction of the many of the strains used in this study. Additionally, he created the robot arrays used in the high-throughput microscopy and performed the original invertase overlay assay. iv  Table of Contents  Abstract .................................................................................................................................... ii Preface ..................................................................................................................................... iii Table of Contents .................................................................................................................... iv List of Tables .......................................................................................................................... vii List of Figures ........................................................................................................................ viii List of Abbreviations ............................................................................................................... ix Acknowledgements ................................................................................................................. xii Dedication .............................................................................................................................. xiii Chapter 1: Introduction ........................................................................................................... 1 1.1 Overview of Yeast Membrane Trafficking Pathways................................................... 1 1.2 Compartment Identity ................................................................................................. 3 1.2.1 Rabs and Their Regulators ...................................................................................... 4 1.2.2 SNAREs and Tethers .............................................................................................. 4 1.2.3 Lipids ...................................................................................................................... 5 1.3 Generating Membrane Curvature for Vesicle Formation ............................................. 6 1.3.1 Amphipathic Helices ............................................................................................... 7 1.3.2 Membrane Curvature Scaffolding Proteins .............................................................. 8 1.3.3 Lipid Translocation ................................................................................................. 9 1.4 Yeast Flippases ......................................................................................................... 10 1.4.1 Structure and Lipid Preference .............................................................................. 11 1.4.2 Spatial Separation ................................................................................................. 13 1.4.3 Regulation of Flippases by Other Factors .............................................................. 13 1.4.4 Drs2 Activity is Needed for Specific Vesicle Formation ........................................ 14 1.5 Trafficking at the Endosome ..................................................................................... 15 1.5.1 Multivesicular Body Pathway................................................................................ 15 1.5.1.1 Endosomal Sorting Complex Required for Transport (ESCRT) ..................... 15 1.5.1.2 Other Sorting Complexes That Act with ESCRT. .......................................... 16 1.5.2 Recycling from Endosomes ................................................................................... 17 1.5.2.1 AP-1 Pathway ............................................................................................... 17 1.5.2.2 Rcy1 Pathway ............................................................................................... 17 1.5.2.3 Snx4/41/42 Pathway ...................................................................................... 18 1.5.2.4 Retromer ....................................................................................................... 18 1.6 High-Throughput Techniques to Study Endosomal Sorting ....................................... 19 1.6.1 Genetic-interaction Studies ................................................................................... 20 1.6.2 Protein-interaction Studies .................................................................................... 21 1.6.3 Phenotypic Studies ................................................................................................ 22 v  1.6.4 High-Throughput Microscopy ............................................................................... 22 1.7 Research Objectives .................................................................................................. 23 Chapter 2: Positive Correlations Reveal a Functional Relationship Between ESCRT and the Novel Protein Complex BLOC ..................................................................................24 2.1 Synopsis ................................................................................................................... 24 2.2 Introduction .............................................................................................................. 24 2.3 Results ...................................................................................................................... 27 2.3.1 Yeast BLOC is not Required for Sorting AP-3 Cargo Proteins .............................. 27 2.3.2 BLOC Colocalizes with ESCRT-0 ........................................................................ 28 2.3.3 BLOC is Required for the Endosomal Sorting of Smf1 ......................................... 31 2.4 Discussion ................................................................................................................ 34 2.5 Methods .................................................................................................................... 37 2.5.1 Network Analysis .................................................................................................. 37 2.5.1.1 Data Sources and Preprocessing .................................................................... 37 2.5.1.2 Correlation Between Phenotypic Interaction Profiles ..................................... 38 2.5.2 Strain Construction ............................................................................................... 39 2.5.3 Invertase Overlay Assay ........................................................................................ 40 2.5.4 Invertase Liquid Assay .......................................................................................... 40 2.5.5 Fluorescence Microscopy ...................................................................................... 41 2.6 Acknowledgements ................................................................................................... 41 Chapter 3: Negative Correlation Analysis Reveals an Antagonistic Relationship Between the VID Complex and the Previously Uncharacterized Protein Vaf1 ....................42 3.1 Synopsis ................................................................................................................... 42 3.2 Introduction .............................................................................................................. 43 3.3 Results ...................................................................................................................... 44 3.3.1 Negative Correlations are Somewhat Rare Within Complexes ............................... 44 3.3.2 Negative Correlations within Complexes Reveal Inhibitory Relationships ............. 46 3.3.3 Negative Correlations May Arise from Competition for Shared Subunits .............. 48 3.3.4 The Previously Uncharacterized Protein Vaf1 is Downregulated by the VID Complex ........................................................................................................................... 50 3.4 Discussion ................................................................................................................ 53 3.5 Methods .................................................................................................................... 55 3.5.1 Network Analysis .................................................................................................. 55 3.5.1.1 Data sources and Preprocessing ..................................................................... 55 3.5.1.2 Correlation Between Genetic Interaction Profiles .......................................... 55 3.5.2 Yeast Strains and Plasmids .................................................................................... 56 3.5.3 Co-immunoprecipitation and Western Blot: .......................................................... 57 Chapter 4: Endosomal Regulators of Yeast Neo1 Identified Through Quantitative Enzymatic and Imaging Analysis............................................................................................59 vi  4.1 Synopsis ................................................................................................................... 59 4.2 Introduction .............................................................................................................. 59 4.3 Results ...................................................................................................................... 61 4.3.1 A Quantitative Snc1-based Assay Identifies Potential Regulators of Neo1 ............ 61 4.3.2 Arl1 Promotes Stability of Dop1 in Complex and on Membranes .......................... 65 4.3.3 Vps13 and PI3K Regulate Neo1 Transit through Endosomes ................................ 67 4.3.4 Vps13 Is Required for Sorting at Early and Late Endosomes ................................. 69 4.3.5 Identification of Factors that Regulate Neo1 Recycling from Endosomes .............. 71 4.3.6 Snx3-dependent Sorting of Neo1 Requires a Short Amino Acid Motif in the N-Terminal Tail .................................................................................................................... 73 4.3.7 Neo1 Sorting by Snx3 is Functionally Important ................................................... 76 4.4 Discussion ................................................................................................................ 79 4.4.1 Identification of Novel Endosome Trafficking Regulators ..................................... 79 4.4.2 Extending Our Knowledge of the Known Neo1 Regulator Arl1 ............................ 80 4.4.3 Endosomal Regulation of Neo1 by Vps13 ............................................................. 80 4.4.4 Recycling of Neo1 by Snx3 ................................................................................... 81 4.5 Materials and Methods .............................................................................................. 82 4.5.1 Yeast Strains and Plasmids .................................................................................... 82 4.5.2 Genome-wide Invertase Overlay Assay ................................................................. 85 4.5.3 High-Content Imaging Screen ............................................................................... 85 4.5.4 Quantitative Image Analysis ................................................................................. 86 4.5.5 Small-scale Invertase Overlay Assay ..................................................................... 87 4.5.6 Sequence Alignments ............................................................................................ 88 4.5.7 Co-immunoprecipitation ....................................................................................... 88 4.5.8 Growth Assay ....................................................................................................... 88 4.6 Acknowledgements ................................................................................................... 88 Chapter 5: Discussion and Future Directions ........................................................................89 5.1 Efficacy and Future of Using High-Throughput Studies ............................................ 89 5.2 Mechanistic Relationship between Yeast BLOC and ESCRT .................................... 90 5.3 Identifying the Role of a Vaf1-containing VID Complex .......................................... 93 5.4 Mon2 and Dop1 Regulation of Neo1 ......................................................................... 94 5.5 Role of Neo1 at the Late Endosome .......................................................................... 96 5.6 Conclusion ................................................................................................................ 98 References ................................................................................................................................99 Appendices ............................................................................................................................. 118 Appendix A : Supplementary Material for Chapter 4 .......................................................... 118 vii  List of Tables  Table 2.1: Strains used in BLOC study (Chapter 2) ............................................................. 39 Table 2.2: Plasmids used in BLOC study (Chapter 2) .......................................................... 40 Table 3.1: Complexes identified that contain a negative correlation ..................................... 45 Table 3.2: Strains used in negative correlations study (Chapter 3) ....................................... 57 Table 3.3: Plasmids used in the negative correlations study (Chapter 3) ............................... 57 Table 4.1: Strains used in the Neo1 study (Chapter 4) .......................................................... 83 Table 4.2: Plasmids used in the Neo1 study (Chapter 4) ...................................................... 84 Table A.1: Genome-wide invertase Z Score ....................................................................... 122 Table A.2: Genes identified in the CLIK analysis to have significantly lower GSS cell surface localization. ........................................................................................................... 238 Table A.3: High-content microscopy screen quantitation ................................................... 246  viii  List of Figures  Figure 1.1: Major vesicle trafficking pathways in a yeast cell ................................................3 Figure 1.2: Summary of proteins responsible for lipid translocation ..................................... 10 Figure 1.3: Schematic diagram illustrating the proposed action of flippases in vesicle biogenesis ........................................................................................................................... 12 Figure 1.4: Summary of known endosomal sorting pathways in yeast. ................................. 15 Figure 2.1: Yeast BLOC does not participate in the AP-3 pathway ...................................... 28 Figure 2.2: BLOC is linked to ESCRTs ............................................................................... 30 Figure 2.3: BLOC and Vps27 are independently recruited to the endosome and BLOC recruitment is PI3P independent. ......................................................................................... 31 Figure 2.4: BLOC specifically affects Smf1 trafficking. ...................................................... 33 Figure 2.5: Model of proposed action by yeast BLOC. ........................................................ 34 Figure 3.1: Relatively few complexes contain negative correlations within a complex. ........ 46 Figure 3.2: Examples of complexes whose negatively correlations can be explained by direct inhibition ............................................................................................................................. 48 Figure 3.3: Examples of complexes whose negatively correlations can be explained by a competition model ............................................................................................................... 49 Figure 3.4: VAF1 is negatively correlated with the VID complex with respect to their genetic interaction profiles............................................................................................................... 50 Figure 3.5: Vaf1 binds strongly to and is downregulated by Vid30. ..................................... 52 Figure 4.1: Using a genome-wide invertase assay in conjunction with high content microscopy to identify potential regulators of Neo1 ............................................................. 64 Figure 4.2: Arl1 promotes stability of Dop1 in complex and on membranes. ....................... 66 Figure 4.3: Vps30/38 (PI3K) and Vps13 regulate Neo1 transit through the endosomes. ....... 68 Figure 4.4: Vps13 is a novel regulator of Neo1, which aids trafficking through the endosomal pathway. .............................................................................................................................. 70 Figure 4.5: Known regulators as well as novel potential regulators cause Neo1 to missort to the vacuole .......................................................................................................................... 72 Figure 4.6: The N-terminal tail of Neo1 is sufficient for localization. .................................. 75 Figure 4.7: A small conserved region on the N-terminal tail of Neo1 is needed for sorting by Snx3 and for sorting of other Snx3 cargo proteins. .............................................................. 78 Figure A.1: Some mutants cause a relocalization of Neo1 to the Golgi .............................. 118 Figure A.2: T-Coffee alignment......................................................................................... 119 Figure A.3: The Snx3 sorting signal is contained in the first 195 amino acids of Neo1....... 120 Figure A.4: Growth assay to assess viability of Neo1 N-terminal tail truncations ............... 121  ix  List of Abbreviations ADP-Adenine Diphosphate ANOVA-Analysis of Variance AP-1-Adaptor Protein Complex-1 AP-2-Adaptor Protein Complex-2 AP-3-Adaptor Proteins Complex-3 ATP-Adenine Triphosphate BAR-Bin/Amphiphysin/Rvs BLOC-Biogenesis of Lysosome-Related Organelle Complex BORC-BLOC-1 Related Complex CKM-Cyclin Kinase Module CLIK-Cutoff Linked to Interaction Knowledge COG-Conserved Oligomeric Golgi Complex COPI-Coat Protein 1 COPII- Coat Protein 2 CORVET- Class C Core Vacuole/Endosome Tethering Complex CPY-Carboxy Peptidase Y DMSO-Dimethylsulfoxide ENTH-epsin N-terminal homology  ER-Endoplasmic Reticulum ERMES- ER-mitochondria encounter structure ESCRT-Endosomal Sorting Complex Required for Transport FBPase-Fructos-1,6,-bisphosphatase FYVE-Fab1, YOTB, Vac1, and EEA1  GAP-GTPase Activating Protein GARP- Golgi-Associated Retrograde Protein Complex GATOR-GAP Towards Rags GDP-Guanosine Diphosphate GEF-GTP Exchange Factor GFP-Green Fluorescent Protein x  GGA-Golgi-localized, Gamma ear-containing, Adaptor GNSS (GFP-Nyv1-Snc1-Suc2) GSS-GFP-Snc1-Suc2 GTP-Guanosine Triphosphate KAN-Kanamycin NAT- Nourseothricin NatC-Nα-acetyltransferase complex NVJ-Nuclear-Vacuolar Junction PAGE-Polyacrylamide Gel Electrophoresis PC-Phosphatidylcholine PCR-Polymerase Chain Reaction PE-Phosphatidylethanolamine PH-Pleckstrin Homology  PI(3,5)P2-Phosphotidylinositol 3,5-phosphate PI(4,5)P2-Phosphotidylinositol 4,5-phosphate PI3K- Phosphatidylinositol 3-kinase PI3P-Phosphotidylinositol 3-phosphate PI4P- Phosphatidylinositol 4-phosphate PI-Phosphoinositol PS-Phosphatidylserine  PX-Phox Homology RFP-Red Fluorescent Protein SDS-Sodium Dodecyl Sulfate SD-Synthetic Dextrose SEACAT-SEA Subcomplex Activating TORC1 SEACIT-SEA Subcomplex Inhibiting TORC1 SEA-Seh1-Associated Complex SGA-Synthetic Genetic Array SNARE-Soluble N-ethylmaleimide-sensitive factor Attachment Protein Receptor  TORC1-Target of Rapamycin Complex 1 xi  TRAPP-Transport Protein Particle Complex UPC-Uncentered Pearson Correlation Coefficient VClamp-Vacuolar-Mitochondrial junction VID-Vacuolar Import and Degradation Complex Y2H-Yeast 2 Hybrid YPD-Yeast Peptone Dextrose   xii  Acknowledgements This thesis would not be possible without the amazing support I received throughout my studies.  First and foremost, I would like to thank my advisor Elizabeth Conibear whose creativity, energy, and devotion continue to inspire me. She has tireless patience and continually challenged me to be a better scientist each and every day. Additionally, vital to my success was the numerous scientific collaborations and advice from mentors outside of the lab.  In particular, I would like to thank my committee members Masa Numata and Orson Moritz for their stimulating discussions and advice throughout my studies as well as our collaborators who performed the bioinformatics, providing the foundation for studies in this thesis. I would like to thank the members of the Conibear lab past and present who were more of a family than mere work partners, supporting me emotionally and intellectually and often providing comic relief.  In particular, Mike Davey was very helpful teaching me and giving advice for many yeast techniques as well as his contributions of technical expertise and key strains to my work.  I would like to thank Bjorn Bean and Shawn Whitfield for the friendship and advice on experiments, despite the occasional groan-worthy joke.  Lastly, I would especially like thank Shawn Whitfield for editing large sections of this thesis.  Finally, I want to give a big thank you to my friends and family who have been really supportive in this process.  In particular, my husband, Lane, who has supported me emotionally, ensured I was well fed and despite not taking any formal biology classes has become an excellent yeast technician.  My parents have always encouraged me to pursue my dreams and continuously strive to understand my work.    xiii  Dedication         To my family:  who has always encouraged me to be the very best version of me,  who is proud of my accomplishments and who is even prouder when I pick myself up from failure, and who gave me the freedom to dream and the support to achieve anything my heart desired.  1  Chapter 1: Introduction 1.1 Overview of Yeast Membrane Trafficking Pathways In a eukaryotic cell, membrane-bound compartments allow for specialized processes to occur in designated areas. Cellular protein trafficking moves proteins needed for the specialized process to the appropriate compartment and is important for the proper functioning of the cell. As proteins are moved throughout the cell, they are sorted at hubs such as the endosome, where they are targeted to the appropriate destination using dedicated sorting machinery. Yeast is an excellent model organism to study protein trafficking due to the conservation with higher eukaryotes and the ease of genetic manipulation. Defects in proteins trafficking pathways are implicated in many diseases. For example, the human homologs of some of the proteins detailed in this thesis are implicated in Hermansky Pudlak Syndrome, a lysosomal storage disorder that causes pigmentation and blood clotting defects, Schizophrenia, and neurodegenerative diseases such as Parkinson’s disease and Alzheimer’s disease (Gautam et al., 2006; Ryder and Faundez, 2009; Vilariño-Güell et al., 2011).  This highlights that work on vesicle protein trafficking has the potential to uncover the mechanisms of disease as well as generate targeted therapeutics. Proteins are transported through the cells in membrane-bound compartments called vesicles. Efficient vesicle formation and transport requires sequestration of cargo proteins into a small area and the binding and recruitment of coat proteins, which help to shape the forming vesicle (Kirchhausen, 2000; Bonifacino and Lippincott-Schwartz, 2003). Cargo adaptors and coat proteins confer specificity, only transporting particular cargo proteins or directing them to a particular destination. Once the vesicle is formed and released from the donor compartment, targeting and fusion of these vesicles with the acceptor compartment is directed by Soluble N-ethylmaleimide-sensitive factor Attachment Protein Receptors (SNAREs) and tethering complexes (Dubuke and Munson, 2016).  The major protein trafficking pathways in a yeast cell are outlined in Figure 1.1. Proteins destined for transport through the endomembrane system, both luminal and membrane bound, are recognized by the signal recognition particle, which targets the proteins to the Endoplasmic Reticulum (ER) (Denks et al., 2014). In the ER, proteins are modified through the addition of sugars (Helenius and Aebi, 2004; Xu and Ng, 2015) and 2  lipids (Lam et al., 2006; Orlean and Menon, 2007). The proteins are then transported using the Coat Protein II (COPII) pathway to the Golgi, a major sorting hub where further modifications occur  (Jensen and Schekman, 2011; Szul and Sztul, 2011).  In the Golgi, additional sugar modifications are added (glycosylation) (Brigance et al., 2000; Munro, 2001; Stanley, 2011). From here proteins are either recycled back to the ER (using the COPI pathway)  (Hsu and Yang, 2009; Hsu et al., 2009; Szul and Sztul, 2011), transported to early endosomes (AP-1 pathway) (Nakatsu and Ohno, 2003), late endosomes (GGA pathway) (Costaguta et al., 2001; Nakayama and Wakatsuki, 2003; Bonifacino, 2004; Ghosh and Kornfeld, 2004; De et al., 2013), or directly to the vacuole (AP-3 pathway) (Odorizzi et al., 1998). Finally, proteins can be targeted to the plasma membrane and secreted either by a constitutive secretion pathway or a specialized Exomer pathway (Wang et al., 2006). After reaching these destinations, the localization of proteins may be modulated depending on cellular needs. For instance, amino acid transporters are only needed at the cell surface during nutrient limiting conditions. When nutrients are plentiful, these proteins may be stored in other compartments or degraded. Endocytosis controls internalization from the cellular surface. One mechanism is mediated by ubiquitin addition to the protein, which targets it for subsequent internalization through the concerted action of actin polymerization, clathrin coats and specialized adaptors such as AP-2 and YAP1801/2 (Kaksonen et al., 2005; Burston et al., 2009) although these components are not always necessary. In the endosomal system, proteins can then be targeted to the vacuole for degradation or recycled back to the Golgi for a return to the cell surface if needed. Recycling out of the endosomal system can occur by several pathways from early and late endosomes using machinery such as Snx4/41/42, retromer, or AP-1 (Hettema et al., 2003; Seaman, 2012). Transport to the vacuole utilizes the Endosomal Sorting Complex Required for Transport (ESCRT) machinery, which internalizes ubiquitinated proteins into compartments within the late endosome (Schmidt and Teis, 2012). The endosome fuses with the vacuole, releasing the vesicular contents which are degraded in the lytic environment. In summary, various vesicle transport machinery coordinates the complicated movement of proteins throughout the cell. 3   Figure 1.1: Major vesicle trafficking pathways in a yeast cell. Black lettering indicates organelles and blue lettering indicates the machinery that controls trafficking in those pathways denoted by the arrows. Stars indicate phophoinositides and the numbers indicate which phosphorylated species is prevalent in the labeled compartment.  1.2 Compartment Identity As mentioned above, compartmentalization in cells is important as it allows separation of otherwise competing processes. However, for proteins to get to the correct compartment, these compartments need to contain distinct features that can be recognized by the trafficking machinery. There are many different molecules, both lipids and proteins, which help to create and recognize compartment identity.  4  1.2.1 Rabs and Their Regulators Rabs are soluble GTPases, which function as molecular switches cycling between their active (GTP-bound) and inactive (GDP-bound) states. In the active conformation, Rab proteins disassociate from regulatory proteins and are able to instead associate with the membrane (Stenmark, 2009). The active conformation also reveals sites for effector proteins such as tethers, coats proteins, kinases etc. which helps to promote vesicle trafficking (Stenmark, 2009). Thus, in the active state, Rabs act as a recruitment platform to allow for efficient and localized formation of vesicles for transport.  In order to downregulate Rab activity, the GTPase activity of the Rab is triggered by association with a GTPase activating protein (GAP) which is recruited independently.  Association of an active Rab with its GAP causes GTP hydrolysis into GDP favoring the inactive conformation. Consequently, membrane association is lost (Stenmark, 2009). In this way, membrane recruitment of the Rab proteins can be controlled by the localization of the regulatory GEF and GAP proteins, which associate with one type of compartment. Thus, Rab GTPases are major drivers of compartment identity and direct the specific recruitment of effector proteins, directing vesicle trafficking.  1.2.2 SNAREs and Tethers SNAREs are important for recognition and fusion of vesicles with their target membrane, thereby contributing to compartment identity and the accuracy of protein traffic. To facilitate fusion, there are two structural classes of SNARE proteins: Q-SNARE and R-SNAREs which are either present on the target membrane (t-SNARE) or on the vesicle (v-SNARE) (Burri and Lithgow, 2004; Hong and Lev, 2014). Most R-SNAREs act as v-SNAREs and most Q-SNAREs act as t-SNARES (Hong and Lev, 2014). When in close proximity, the SNARE proteins form a stable complex (one v-SNARE with three t-SNAREs) with sufficient energy to overcome the energy barrier required for membrane fusion (Burri and Lithgow, 2004; Hong and Lev, 2014; Dubuke and Munson, 2016). In general, SNAREs can bind only to a small selection of other SNAREs, allowing their interaction to be specific (Burri and Lithgow, 2004; Jahn and Scheller, 2006). Thus, the localization of t-SNAREs 5  helps to create organelle identity as it is a signal for fusion of specific vesicles destined for that organelle (Jahn and Scheller, 2006; Dubuke and Munson, 2016).   SNARE complex formation is regulated in part through the action of tethering complexes that help to recruit incoming vesicles to the target organelle (Dubuke and Munson, 2016). There are many tethering complexes which associate with different organelles, thus regulating the vesicle traffic at these locations (i.e. GARP, COG, TRAPP, and CORVET) (Dubuke and Munson, 2016). In general, these complexes will bind to lipids, coat proteins, Rabs, and/or SNAREs to bring the vesicle into close proximity with the target membrane, increasing the likelihood of binding of the appropriate SNAREs (Sztul and Lupashin, 2009; Dubuke and Munson, 2016). In this way, tethers are also identity markers for an organelle and aid in appropriate vesicle fusion.   1.2.3 Lipids Glycerophospholipids, which consist of two fatty acid chains attached to glycerol and a phosphate group, make up a large portion of cellular membrane lipids (van Meer et al., 2008). Often attached to the phosphate is one of a number of diverse polar head groups that range in size and charge (i.e. choline, ethanolamine, serine, inositol, etc.). Each head group contributes different properties to a membrane. Thus, the lipid composition can direct protein recruitment, a major factor in compartment identity. For instance, the negatively charged serine head group on phosphatidylserine allows proteins with a positively charged patch to bind (Di Paolo and De Camilli, 2006; van Meer et al., 2008). While one lipid on its own may not change the overall structure of the membrane, patches of similar lipids can recruit proteins required to generate diverse processes such as membrane deformation and vesicle biogenesis.  Phosphatidylinositols (PIs) are phospholipids with an inositol ring head group that can be reversibly phosphorylated (van Meer et al., 2008). While phosphoinositides (phosphorylated phosphatidylinositols) are in fewer numbers compared to other lipids in the membrane, they play a crucial role in compartment identity and recruitment of cytosolic proteins to those compartments. Phosphoinositides can be phosphorylated on any of three different ring positions (3, 4, 5) (Di Paolo and De Camilli, 2006). The phosphorylation status 6  of PI is regulated through a series of kinases and phosphatases (Di Paolo and De Camilli, 2006), creating distinct pools of differentially phosphorylated phosphoinositides (Figure 1.1).  The localization and activity of these kinases and phosphatases are tightly controlled such that the phosphorylation position acts as a compartment identity marker. Four major phosphoinositide species in the cell identify major sites of trafficking. These include PI4P, PI(4,5)P2, PI3P, and PI(3,5)P2, where the number indicates the position that has been phosphorylated. PI4P is the main phosphoinositide species at the Golgi. Absence of the yeast phosphoinositol 4-kinase causes abnormal accumulations of membrane called “Berkeley bodies” at the Golgi, as well as trafficking defects to the plasma membrane (Odorizzi et al., 2000). At the plasma membrane, a phosphatidylinositol 4-phosphate is phosphorylated at the 5 position, creating PI(4,5)P2 (Di Paolo and De Camilli, 2006). In addition to other roles, this doubly-phosphorylated lipid species regulates traffic at the plasma membrane by recruiting proteins necessary for endocytosis and exocytosis (Di Paolo and De Camilli, 2006). Finally, PI3P at the early endosome and PI(3,5)P2 at the vacuole help to promote endosomal traffic (Di Paolo and De Camilli, 2006; Stahelin et al., 2014). The majority of PI3P is generated by the phosphoinositol-3 kinase, Vps34, from PI. PI3P is needed to recruit many of the endosomal sorting complexes such as retromer and ESCRTs. Some of the PI3P is then converted into PI(3,5)P2 which is needed for proper vacuolar morphology (Di Paolo and De Camilli, 2006). Thus, the cell uses particular pools of phosphorylated phosphoinositides to create organelle identity and recruit proteins to these locations. Recruitment of proteins to phosphoinositide-containing membranes is generally achieved through lipid recognition domains such as the Pleckstrin Homology (PH), the Phox Homology (PX), and the Fab1, YOTB, Vac1, and EEA1 (FYVE) domain.  These domains are specific for different phosphoinositide species and enhance the specificity of membrane recruitment (Lemmon, 2008; Stahelin et al., 2014). In all, lipids and lipid recognition domains help to maintain proper protein composition in each compartment.  1.3 Generating Membrane Curvature for Vesicle Formation In order for appropriate protein trafficking to occur, the membrane must be able to curve, generating small vesicles that contain the cargo protein of interest as well as the 7  trafficking machinery. There are several mechanisms to induce or maintain membrane curvature, including reversible insertion of hydrophobic protein motifs (i.e. amphipathic helices), protein scaffolding (i.e. Bin/Amphiphysin/Rvs (BAR) domains and coat proteins), and induction of lipid asymmetry (i.e. flippases).   1.3.1 Amphipathic Helices  Amphipathic helices have two faces on opposite sides of an α-helix, one hydrophobic and one hydrophilic (Drin and Antonny, 2010), which allows them to reversibly associate with the membrane. In general, when the hydrophobic face is exposed, the hydrophobic effect draws the amphipathic helix into a conformation parallel to the membrane where the hydrophobic groups insert between the fatty acyl chains, and the hydrophilic amino acids associate with the polar head groups (Drin and Antonny, 2010).  One way this may induce curvature is through adding bulk (due to increased lipids) on one side of a membrane creating a positive curvature pressure (Drin and Antonny, 2010). Amphipathic helices vary in length, in their affinities for charged lipids, and in the amino acid composition of the hydrophobic face and thus have differential ability to induce curvature (Mcmahon and Boucrot, 2015).  Amphipathic helices are used to generate or maintain membrane curvature for trafficking processes throughout the cell. For example, epsin N-terminal homology (ENTH) domain-containing proteins have an amphipathic helix that is exposed upon binding to PI(4,5)P2 which allows it to generate membrane curvature to start endocytosis (Sen et al., 2012). Additionally, the small GTPase Arf1 is needed to recruit coat/adaptor proteins such as AP-1 and GGA to the Golgi, needed for Golgi to endosome trafficking (Dell’Angelica et al., 2000; Nakatsu and Ohno, 2003). In the active GTP-bound conformation, the N-terminal amphipathic helix anchors the protein and adds bulk to the cytosolic leaflet (Antonny et al., 1997). Thus, trafficking pathways throughout the cell utilize amphipathic helices as one way to recruit proteins to membranes and create membrane curvature needed for additional recruitment of specific trafficking factors.  8  1.3.2  Membrane Curvature Scaffolding Proteins Some proteins adopt a curved conformation that can sense and maintain curvature by acting as scaffolds to counter the tension caused by bending a membrane. Coat proteins such as clathrin, COPI, and COPII are a family of proteins that cause membrane curvature that is particularly important for the formation of budding vesicles (Mcmahon and Boucrot, 2015). The structure of the forming coat polymer helps to stabilize the membrane curvature of the forming vesicle (Mcmahon and Boucrot, 2015). Due to their reliance on adaptors, these coats are able to be specifically recruited for membrane remodeling only after several sorting and small membrane curvature events have taken place, creating a specialized environment for vesicles to form.  Another type of scaffold is the Bin/Amphiphysin/Rvs (BAR) domain containing proteins, a large structurally-similar superfamily. Featuring a three-helix coiled-coil core, these proteins create crescent-shaped dimers which sense and maintain membrane curvature (Mim and Unger, 2012). These BAR domain-containing proteins have been categorized based on the shape of their BAR domains and include the classical BAR domain, extended FCH domain (F-BAR), and Inverse BAR domain (I-BAR) (Mim and Unger, 2012). This section will focus only on proteins with the classical BAR domain. Classical BAR domains are found to have the highest degree of structural bend. In general, BAR domains bind to acidic membrane surfaces through basic residues (Shin et al., 2012) However, BAR domains are not usually very specific on their own and are often coupled with other lipid sensing domains. These lipid sensing domains further divide the family of classical BAR domain containing proteins into three groups; the pleckstrin homology (PH-BAR) which binds to PI4P derivatives, the amphipathic N-terminal helix (N-BAR) that embeds into one leaflet of the membrane, or the phox-homology (PX-BAR or SNX-BAR) that binds PI3P derivatives (Di Paolo and De Camilli, 2006; Mim and Unger, 2012).  While it has been shown that some BAR domain-containing proteins, specifically N-BARs, tubulate membranes in vitro, there is conflicting evidence based on the lipid species used (Mim and Unger, 2012). It is the general consensus that instead, some amount of pre-9  existing membrane curvature is a needed for BAR domain recruitment, maintenance and stimulation of additional membrane curvature (Mim and Unger, 2012).  1.3.3 Lipid Translocation As mentioned above, the lipid composition of the membrane can have profound effects on cellular biology including compartment identity. However, the shape or charge of the lipid can also be imperative for generating membrane curvature, a mechanism vastly different from the membrane deformation events seen above with the scaffolding proteins and amphipathic helices.  When lipids with similar properties are grouped this can influence membrane curvature. For example, phosphatidylethanolamine (PE) has a very small head group causing it to adopt a conical shape and thus induce curvature stress in a membrane, helpful for membrane fission and fusion events (van Meer et al., 2008). While a single PE lipid in a vast sea of other cylindrical lipids will not have a major effect on the membrane, concentrating these lipids in a small space creates increased curvature pressure, thus mechanisms for regulating membrane composition are essential for generating desired membrane curvature.  There are several proteins that regulate lipid asymmetry such as flippases and floppases, which use ATP to move lipids against their concentration gradients in opposite directions, as well as scramblases, which are passive transporters of lipids with the concentration gradient (Graham, 2004) (summarized in Figure 1.2). These all contribute to the regulation of lipid asymmetry within the cell; however, I will focus on the flippases, part of the P4-type ATPase family, as they have been shown to generate membrane curvature needed for vesicle biogenesis.  10   Figure 1.2: Summary of proteins responsible for lipid translocation. This model has been adapted from Graham, 2004 and Clark, 2011. Flippases and Floppases transport aminophospholipids in opposite directions in an ATP dependent manner.  Scramblases are ATP independent proteins that flip lipids with the concentration gradient. Colored lipid is the transported lipid. PS: Phosphatidylserine, PE: Phosphatidylethanolamine, PC: Phosphatidylcholine, SL: Sphingolipid  In general, flippases, or aminophospholipid translocases, preferentially flip aminophospholipids such as phosphatidylserine, phosphatidylcholine, and phosphatidylethanolamine to the cytosolic leaflet of the membrane (Graham, 2004; Shin et al., 2012). Deletions of these proteins cause aberrations in vesicle trafficking, illustrating their importance in this process (Hua et al., 2002; Graham, 2004; Shin et al., 2012). Although it is not known exactly how flippases cause changes in membrane curvature to occur, it is thought, based on the membrane bilayer couple hypothesis, that rapid expansion of one membrane leaflet could cause the other leaflet to contract to maintain the bilayer conformation, thus causing membrane bending (Sheetz and Singer, 1974) (Figure 1.3).  The bent membrane recruits curvature sensing vesicle formation machinery needed to stabilize and further remodel the membrane (Graham, 2004).  1.4 Yeast Flippases Overexpression of phosphatidylserine in the cytoplasmic leaflet is not sufficient to rescue flippase mutants (Takeda et al., 2014). Instead, the directed action of flippases both spatially and temporally may be required for vesicle formation through localized changes in membrane conformation. There are five flippases in yeast: Dnf1, Dnf2, Dnf3, Drs2, and Neo1, where Neo1 is the only essential member (Catty et al., 1997; Halleck et al., 1998; Hua 11  et al., 2002). However, mutation of Dnf1, 2, 3 and Drs2 together is also lethal, highlighting the importance of these proteins in cellular function (Hua et al., 2002). In general, the activity of flippases in vesicle trafficking is thought to be specified by its structure and lipid preference, their localization, and modulation by other proteins.  1.4.1 Structure and Lipid Preference Drs2 and its mammalian homologs ATP8A1/2 are the only flippases shown to have direct in vitro flippase activity and this activity has been shown to be needed for exocytic vesicle formation (Natarajan et al., 2004; Coleman et al., 2009; Zhou and Graham, 2009). However, there is evidence that other flippases affect lipid asymmetry in vivo suggesting that they indeed flip lipids in a similar capacity. For instance, Dnf1 and Dnf2 are thought to flip phosphatidylcholine (PC) and phosphatidylethanolamine (PE) and its lysophopholipid derivative across the plasma membrane (Pomorski et al., 2003; Riekhof and Voelker, 2006; Baldridge and Graham, 2011, 2012; Baldridge et al., 2013). In addition, temperature sensitive mutants of Neo1 cause disruptions of PE and PS asymmetry (Takar et al., 2016). Thus, it is thought that these five flippases act to control lipid asymmetry within the cell. Lacking a high-resolution structure for these flippases, clues about their structure have been deduced from homology modeling based on the structure of other ATPases. These comparisons suggests that flippases have five domains: nucleotide-binding (N) which binds to ATP, phosphorylation (P) which is transiently phosphorylated, actuator (A) which modulates dephosphorylation, membrane domain (M) involved in translocation, and N or C-terminal tails which may be regulatory domains (R) (Sebastian et al., 2012; Andersen et al., 2016). Several features appear conserved in terms of mechanistic action between all P-type ATPases such as phosphorylation-driven conformational changes (Montigny et al., 2015). However, due to the large size of lipids compared to ions, called the “giant substrate” problem, flippases may not transport the substrate in the same way as other known ATPases (Montigny et al., 2015; Andersen et al., 2016). Through work on mammalian and yeast flippases, a few models of translocation have been proposed. The most favored of these models is the “credit card model”, where the phospholipid head group interacts with the protein, sliding along a groove from the exoplasmic to the cytoplasmic side of the membrane 12  while the fatty acid tails remain in the hydrophobic membrane (Figure 1.3) (Baldridge and Graham, 2012; Andersen et al., 2016). Future studies aim to test this model as well as determine substrate specificity mechanisms.   Figure 1.3: Schematic diagram illustrating the proposed action of flippases in vesicle biogenesis. This model has been adapted from Xu et al., 2013; Montigny et al., 2015; and Andersen et al., 2016. Top panel: flippases are thought to translocate lipids by transporting the hydrophilic head through a groove in the protein while allowing the hydrophobic tails to remain in the membrane, called the “credit card” model. The act of translocating lipids to one side generates positive curvature pressure and thus membrane curvature. Bottom section: a proposed model outlining the role of flippases in vesicle biogenesis. The curvature generated by the flippase recruits small GTPases and the coat adaptor complex, which helps to sequester and recruit coat components needed for vesicle biogenesis.  Flippases have specificities for particular lipid head groups and the mechanism outlining how these lipids are recognized is beginning to emerge. Through genetic and biochemical studies, Baldridge and Graham (2011 and 2012) identified residues needed for Drs2 and Dnf1 to recognize their respective substrates. Mutations in Drs2 were found that caused disruptions in PS asymmetry but interestingly not PE asymmetry. In addition, they 13  were able to switch the lipid preference of Dnf1 from PC to PS by a single point mutation which also complemented some of the phenotypes caused by a deletion in Drs2 (Baldridge and Graham, 2011; Xu et al., 2013). Further investigations suggested that residues controlling substrate specificity formed a kind of entry and exit gate for appropriate lipid transport (Baldridge and Graham, 2012; Baldridge et al., 2013). Thus, substrate specificity is sequence-derived and contributes to the specific activity of the flippase.   1.4.2 Spatial Separation Flippases are spatially separated allowing lipid translocation to occur only in specific cellular locations. Studies thus far have elucidated the steady-state compartment for each of the flippases. Dnf1 and Dnf2 are localized primarily to the plasma membrane while Drs2, Dnf3, and Neo1 appear on internal structures like Golgi and endosomes (Hua et al., 2002; Pomorski et al., 2003). However, the trafficking pathways that are used and the defects caused by mislocalization of flippases require further study. The trafficking of Drs2 is the most well-known of the yeast flippases and localizes to both Golgi and endosomes (Liu et al., 2008). When AP-1 is absent, Drs2 is still able to reach the endosomes, however, it transits the plasma membrane rather than through an AP-1 clathrin-coated vesicle from the Golgi to the endosome (Liu et al., 2008). Additionally, AP-1 is required for Drs2 retrieval from early endosomes to the Golgi as AP-1 mutations in combination with mutations disrupting the progression of proteins to the vacuole causes an accumulation of Drs2 in endosomal structures (Liu et al., 2008). The localization of the protein restricts access of the flippase to specific compartments where it acts.  1.4.3 Regulation of Flippases by Other Factors Flippase activity can be modulated by association with specific proteins and lipids. In this way, translocation of lipids can be turned on and off depending on the presence of certain factors. Some examples of this regulation are non-catalytic β-subunits, binding to phosphoinositides, and binding to coat regulators. There are three non-catalytic β-subunits, Cdc50, Lem3, and Crf1 known to bind specifically to Drs2, Dnf1/2, and Dnf3 respectively and this binding is required for both 14  proper localization and function (Saito et al., 2004; Chen et al., 2006; Furuta et al., 2007). The β-subunits are needed for ER exit and thus they represent important trafficking determinants (Saito et al., 2004). They are also needed for flippase activity, as removal of the β-subunit phenocopies the flippase mutant, with slightly less severe outcomes (Furuta et al., 2007). In mammalian flippases, the β-subunit has been shown to be needed for appropriate phosphorylation of its flippase, suggesting a reason for functional regulation by β-subunit in regulating conformation (Lenoir et al., 2009; Bryde et al., 2010; Andersen et al., 2016). Interestingly, Neo1 does not seem to have an obligatory β-subunit (Barbosa et al., 2010). However, it is known to have binding partners that may be needed for function, Dop1 and Mon2 (Efe et al., 2005; Barbosa et al., 2010). Flippases can also be regulated by coat recruitment factors and lipids. Studies have identified regulatory binding sites on Drs2 for the small GTPase Arl1, the ArfGEF Gea2, PI4P, as well as an early endosome recycling mediator, Rcy1(Chantalat et al., 2004; Tsai et al., 2013; Zhou et al., 2013; Hanamatsu et al., 2014). Thus, many different trafficking regulators bind to Drs2 to recruit it to particular trafficking pathways. Specifically, binding of Drs2 to PI4P has been shown to create a conformational change reducing the autoinhibition of Drs2, allowing it to actively flip lipids (Zhou et al., 2013).   1.4.4 Drs2 Activity is Needed for Specific Vesicle Formation Experimentation on Drs2 has generated evidence that flippase activity is required for specific trafficking pathways and may be tied to its own transport, a concept that links many of the areas of regulation describe above.  Drs2 has been shown to be required for AP-1 dependent trafficking  as deletion of Drs2 has been shown to have a decrease in clathrin coated vesicles and Drs2 affects the trafficking of Chs3, an AP-1 cargo (Chen et al., 1999; Liu et al., 2008). Strikingly, Drs2 activity was also required for its own transport in the AP-1 pathway as disruption through a temperature-sensitive allele caused similar trafficking defects as mutation of AP-1 subunits (Liu et al., 2008) Thus, Drs2 is not only trafficked by AP-1, but its activity has been shown to be important for the formation of AP-1 vesicles themselves (Liu et al., 2008). In all, this suggests a general flippase mechanism where trafficking and activity of flippases may be both required for correct vesicle formation. 15   1.5 Trafficking at the Endosome As mentioned above, endosomal sorting is a way to regulate protein fates in a cell, thus maintaining cellular homeostasis. Several pathways are present at the endosome including transport to the vacuole using the ESCRT pathway or recycling from the early or late endosome (Figure 1.4).  Figure 1.4: Summary of known endosomal sorting pathways in yeast. Two pathways have been shown requiring to work upstream of retromer: AP-1 or Snx4/41/42. At the late endosome, ESCRT sequesters ubiquitinated cargo into multivesicular bodies and retromer recycles proteins back to the Golgi. When the late endosome fuses with the vacuole, the proteins in the multivesicular bodies are degraded.  1.5.1 Multivesicular Body Pathway 1.5.1.1 Endosomal Sorting Complex Required for Transport (ESCRT) The major sorting machinery known to aid in the progression of proteins through the endosomal system for degradation in the vacuole is the well-characterized Endosomal Sorting Complex required for Transport (ESCRT). ESCRT is made up of 4 complexes; ESCRT-0, I, II, III, which act in concert with the AAA-ATPase Vps4 (Babst et al., 2000, 2002a, 2002b; Katzmann et al., 2001). The cytosolic components of each complex work in a sequential manner to cause invagination of the membrane in the late endosomes, forming small vesicles inside the late endosome that contain the proteins targeted for degradation. The general mechanism relies on recognition of ubiquitin by the ESCRT complexes and 16  subsequent invagination of the membrane and finally, scission of the vesicle into the late endosome generating a multivesicular body structure (Schmidt and Teis, 2012). Since ESCRT-0, I, and II all bind ubiquitin they must cooperate to sort proteins into intraluminal vesicles, however the mechanism is still unknown. One possibility is that they mediate the transfer of ubiquitinated cargo from one complex to the next. Another possibility is that they all recognize slightly different proteins to create a platform of ubiquitin binding to sequester ubiquitinated cargo efficiently into the forming vesicles (Wernimont and Weissenhorn, 2004; Teis et al., 2008). ESCRT-III is comprised of four protein subunits, which act in a sequential manner to sequester cargo and inwardly deform the membrane, generating intraluminal vesicles (Babst et al., 2002a). The ESCRT-III subunit, Vps20, serves as a nucleation point off of which Snf7 oligomers form, generating the curvature needed to bend the membrane. Vps24 and Vps2 cap the Snf7 oligomers and recruit the AAA-ATPase, Vps4, which mediates scission of the intraluminal vesicle, thereby releasing the ESCRT complex from the membrane (Babst et al., 2002a; Teis et al., 2008).  Ubiquitin is essential for transporting proteins using ESCRTs. However, it is equally important that these ubiquitins are removed from proteins prior to degradation in order to conserve this resource. The deubiquitinase Doa4 associates with the ESCRT-III subunit through binding of Bro1 (Amerik et al., 2006; Richter et al., 2007). This allows the ubiquitin to be cleaved just prior to the release of the intraluminal vesicle inside the late endosome, thereby preserving ubiquitin for further rounds of protein targeting (Shields and Piper, 2011).   1.5.1.2 Other Sorting Complexes That Act with ESCRT.  While ESCRTs are a well-characterized pathway, targeting proteins for degradation in the vacuole, there are some lines of evidence suggesting there may be more endosomal sorting complexes that work independently or with ESCRT to mediate protein degradation. For instance, Bro1 is an alternative yeast ESCRT-0 like protein that mediates cargo sequestration. Additionally, while it has been canonically described that ESCRT-III subunits require activation by upstream ESCRT complexes, recent evidence suggests there may also be an alternative pathway utilizing Bro1 (Tang et al., 2016). Thus, there are known 17  alternatives that work in coordination with ESCRT proteins to enhance the specificity and capacity of endosomal sorting.  1.5.2 Recycling from Endosomes Another important pathway out of the endosome recycles proteins back to the Golgi. There are at many different pathways described for protein retrieval from endosomes, utilized by proteins, such as Snc1 and Vps10.   1.5.2.1 AP-1 Pathway Several proteins are recycled primarily through the early endosome such as Chitin Synthase 3 (Chs3), the v-SNARE Snc1, and the peptidase Ste13; however, the model for trafficking in this pathway is still emerging. One potential early endosome recycling pathway utilizes clathrin and the clathrin adaptor, AP-1. AP-1 is shown to be required for early endosome recycling of Chs3 and Ste13 but not Snc1 and depends on the presence of the flippase, Drs2 (Valdivia et al., 2002; Foote and Nothwehr, 2006; Liu et al., 2008). This suggests that similar to anterograde pathways, flippase activity aids in the recruitment of coat adaptors for vesicle trafficking. However, AP-1 is not the sole pathway for recycling as deletion of AP-1 components does not contribute to a vast reduction in recycling.   1.5.2.2 Rcy1 Pathway Another potential pathway for recycling from endosomes again utilizes the flippase Drs2 in cooperation with the F-box protein Rcy1. Deletion of either of these proteins causes major morphological changes and trafficking defects (Wiederkehr et al., 2000; Furuta et al., 2007), indicating that they contribute to a majority of the recycling at the early endosome. Rcy1 binds to the flippase Drs2 which is required for formation of recycling vesicles carrying Drs2 as well as the cargo protein Snc1 (Furuta et al., 2007; Hanamatsu et al., 2014). However, the coat and mechanisms for vesicle biogenesis in this pathway are not known.   18  1.5.2.3 Snx4/41/42 Pathway Snx4 has been shown to bind to Snc1 and Snx4/41/42 are required for its trafficking from endosomes to the Golgi, independent of retromer (Hettema et al., 2003).The sorting nexin Snx4 is thought to contain a BAR motif, based on its mammalian homolog, and may help deform the membrane with its partners Snx41/42 (Hettema et al., 2003; van Weering et al., 2012). However, the exact mechanism of action for cargo recruitment and vesicle biogenesis at the early endosome is not known.   1.5.2.4 Retromer At the late endosomes, the highly conserved retromer complex recruits and sequesters proteins for transport to the Golgi (Burd and Cullen, 2016). Retromer was first identified as a stable pentameric complex in yeast that affected the trafficking of the carboxypeptidase Y receptor, Vps10 (Seaman et al., 1998) . Since its discovery in yeast, it has been found to be widely conserved and implicated in diseases such as Parkinson’s and Alzheimer’s (Vilariño-Güell et al., 2011; Fjorback et al., 2012; Zavodszky et al., 2014).  Yeast retromer comprises two separate subcomplexes, the cargo-selective complex (Vps29, Vps26, and Vps35) and the structural subcomplex (Vps5 and Vps17) (Hierro et al., 2007). While these form a stable pentameric structure in yeast, they appear to be more transiently associated in mammalian cells (Mcgough and Cullen, 2011). Regardless, in both yeast and mammalian systems, these two subcomplexes work in tandem to bind and recruit cargo proteins from the late endosome into tubules, which become vesicles targeted for the Golgi.  Retromer requires specific recruitment to the late endosome. The two SNX-BAR proteins that form the structural subcomplex belong to a subfamily of sorting nexins that in addition to a phosphoinositide binding region also contains a membrane-stabilizing curved shape (Carlton and Cullen, 2005). Thus, the SNX-BAR proteins Vps5 and Vps17 bind to phosphoinositide 3-phosphate (PI3P) for their recruitment to endosomes (Burda et al., 2002). In contrast, the cargo-selective complex does not contain any known lipid-binding motifs and therefore, the recruitment of the yeast cargo-selective complex is less well understood. The mammalian cargo-selective complex binds to Rab7 for its recruitment. In yeast, while the 19  yeast Rab7 homolog, Ypt7 is not needed for recruitment to the endosome, binding to Ypt7 is needed for appropriate retrograde trafficking (Liu et al., 2012). Recent evidence suggests that at least one of the three Rab Guanine Exchange Factors (GEFs), which activate the yeast endosomal Rab homolog, Vps21, are needed for recruitment of both the structural as well as cargo-selective complex through the establishment of a PI3P enriched endosomal domains (Bean et al., 2015). Thus, in yeast, the structural components Vps5 and Vps17 binds to PI3P and the cargo-selective subcomplex is recruited through binding to the structural complex but requires the Rab7 homolog for appropriate function (Seaman and Williams, 2002; Liu et al., 2012).   Once retromer is recruited, it is able to deform the membrane, creating 50nm diameter tubules (van Weering et al., 2012) elongating from the late endosome. Cargo proteins bind to the complex through the cargo-selective component, Vps35, as well as the cargo adaptor Snx3 (Nothwehr et al., 1999; Strochlic et al., 2007). When the tubule is finally formed containing sequestered cargo proteins, the dynamin-related protein Vps1 in conjunction with the sorting nexin Mvp1 allow the scission of vesicles bound for the Golgi (Chi et al., 2014).  As mentioned above, the cargo-selective complex mediates binding of cargo proteins. Initially, there were several sites identified on Vps35 that were needed for the proper trafficking of Vps10 but did not affect the Ste13 based cargo protein, A-ALP (Nothwehr et al., 1999). Conversely, A-ALP as well as well as the iron transporters Fet3 and Ftr1 have been shown to utilize the cargo adaptor Snx3 to bind to retromer (Strochlic et al., 2007). Since then, other adaptors in mammalian systems have been identified (Temkin et al., 2011). There is not a known consensus sequence for cargo recognition; however, some signals have been defined that contain a bulky hydrophobic amino acid (i.e. Y, and F). A more clearly defined consensus sequence remains to be identified.  1.6 High-Throughput Techniques to Study Endosomal Sorting While there have been great advances in our understanding of endosomal sorting in yeast and mammals, many mechanisms remain only partially understood. Due to the ease of genetic manipulation and its compact genome, yeast is a valuable model organism to tackle 20  the remaining questions. The genome of yeast was the first to be sequenced, identifying ~5800 genes. Numerous technologies were then developed to identify the function of these genes, the first being a large collection of mutants where all non-essential genes (~4800) were systematically disrupted (Giaever et al., 2002). Since then other collections have been made to study protein overexpression, localization, and protein interactions (Boone et al., 2007). Thus, yeast collections can be screened using broad approaches including genetic interaction screens, phenotypic screens, protein interaction screens, and high content imaging.  These approaches complement one another and integrating data from these different sources offers the most comprehensive view of cellular function.  1.6.1 Genetic-interaction Studies Genetic interactions define pairs of genes whose double mutant phenotype departs from what is expected based on the phenotypes of the corresponding single mutants (Beltrao et al., 2010). For example, a double mutant may have similar phenotype such as colony size to that of its single mutant parents (positive genetic interaction). This suggests that the two genes may have roles in the same pathway as no further cellular harm came from the addition of the second mutation (Guarente, 1993; Beltrao et al., 2010).  However, if the double mutant has a different or worse phenotype than the parents such as being lethal or severely sick (negative genetic interaction), these two genes may participate in independent but complementary pathways (Guarente, 1993; Beltrao et al., 2010).  Probing for genetic interactions may help to identify the functional relationship between two genes. Although, genetic interaction data can be riddled with noise and non-specific data due to experimental error or complicated by complex biological interactions. Thus, additional functional analyses are used in conjunction to strengthen the predictions made by genetic interaction. Despite their pitfalls, genetic interactions are amenable to large-scale screening and large genetic interaction networks have been created which has allowed numerous predictions about the function and relatedness of many uncharacterized proteins (Costanzo et al., 2010; Ryan et al., 2012).   21  1.6.2 Protein-interaction Studies Protein interaction networks are an important technique to display physical relationships between proteins.  Proteins interaction networks were initially completed in yeast using yeast two-hybrid (Y2H) assays, where proteins are attached to one of two parts of a transcription factor (Fields and Song, 1989; Legrain and Rain, 2014).  If the proteins interact, the transcription factor as well as the attached proteins are translocated to the nucleus, allowing the transcription of a readout protein (Fields and Song, 1989). One drawback to this technique is the need for the interacting proteins to be translocated into the nucleus and thus membrane bound proteins are difficult to study using this method. New protein interaction assays have been developed to overcome the need for the protein complex to be translocated into the nucleus such as the split ubiquitin Y2H, DHFR (Dihydrofolate Reductase), and OyCD (Optimized yeast Cytosine Deaminase) assays (Shibasaki et al., 2008; Tarassov et al., 2008; Michnick et al., 2010; Rochette et al., 2015; Ear et al., 2016). These work in a similar manner to traditional Y2H without the need for translocation to the nucleus. However, Y2H remains widely used due to the presence of Y2H yeast collections which make it easy to screen genome-wide as well as the ease and versatility of this technique.    Another popular method to identify protein interaction networks is through affinity purification followed by mass spectrometry.  In this process proteins are purified and subjected to quantitative mass spectrometry where all interacting peptides are identified (Kaake et al., 2010; Legrain and Rain, 2014). However, affinity purification mass spectrometry isn’t as sensitive compared to other techniques and requires a fair amount of technical expertise (Legrain and Rain, 2014). Despite the technical hurdles, there have been large-scale purifications to create large-scale protein interaction networks using this technique which have yielded new and known complexes (Gavin et al., 2002, 2006; Babu et al., 2012). Thus, there are many different techniques to assess protein interaction in the cell in a high-throughput manner, each with its advantages and disadvantages.  The combination of many different screening techniques has provided a fairly comprehensive view of protein-protein interaction networks in yeast.  22  1.6.3 Phenotypic Studies Characterization of genes through screens specific for particular pathways (phenotypic screens) has allowed the identification and characterization of many cellular processes including cargo sorting, vesicle budding, and tethering/fusion, emphasizing the power of this approach.  A significant advantage of phenotypic screens compared to other high-throughput techniques lies in the ability of these types of studies to target dysfunction in a particular pathway. Secretion of Carboxy Peptidase Y (CPY), a vacuolar peptidase that normally localizes to the vacuole has been used to identify proteins responsible for endosomal trafficking (Bonangelino et al., 2002). In Chapter 4, a phenotypic study that monitors the plasma membrane levels of a chimeric Snc1 protein is used to identify proteins needed for early endosome recycling. The same reporter was used previously to identify new endocytosis regulators (Burston et al., 2009).  Thus, the specificity of phenotypic screens allows for identification of proteins involved in the targeted pathway.  1.6.4 High-Throughput Microscopy Fluorescence microscopy is a helpful tool to identify in vivo steady-state localization. Due to technical limitations, for both image acquisition and analysis, microscopy has only recently been adapted as a large-scale technique and is currently one of the fastest growing fields.  An initial high-throughput microscopy screen that identified the steady-state localization of all the proteins within the yeast cell (Huh et al., 2003) was used to create the first localization database.  With advances in technology, these imaging systems have become more accessible for many groups.   In addition, there has been an explosion in imaging software which can help to analyze the vast number of images generated by these screens in a quantitative and unbiased way (Bougen-Zhukov et al., 2016). This technique has allowed an alternate approach to study protein function based on disruption of localization that will become more powerful with increasing technological advances.   23  1.7 Research Objectives A primary goal of high-throughput techniques is to generate large datasets from which functional predictions about proteins and/or protein complexes can be made. We hypothesized that screening the yeast genome using high-throughput techniques would generate specific and testable predictions about endosomal sorting regulation. In this study, we incorporated two different approaches towards this goal.  In one approach we used correlation analysis, the statistical measure of the similarity of interaction profiles (i.e. genetic or drug interactions) between two different gene mutants. Highly correlated interaction profiles have been shown to predict genes that have a similar function (Costanzo et al., 2010). Our overall goal was to determine the efficacy of using correlation analysis - in particular mapping of correlation coefficients onto known complexes - to make new predictions about endosomal sorting regulation.  We focused on negative (anti) correlated interaction profiles, where the interaction profiles of two gene mutants are exactly opposite. We predicted these correlations would reveal biologically relevant antagonistic relationships between the respective proteins.   We chose two predictions based on this analysis for experimental validation (Chapter 2 and Chapter 3). A second high-throughput approach we used to make functional predictions relied on high-content microscopy combined with quantitative image analysis. This technique has been substantially improved over recent years to allow for reliable and sensitive detection of changes in steady-state localization (Usaj et al., 2016). We hypothesized that this technique would allow us to identify regulators that affected the localization of our protein of interest, Neo1. In Chapter 4, we identified three classes of Neo1 endosomal regulators using this technique.    24  Chapter 2: Positive Correlations Reveal a Functional Relationship Between ESCRT and the Novel Protein Complex BLOC   2.1 Synopsis In recent years, high-throughput screens have rewarded the yeast community with insights into genetic and physical relationships between proteins. Correlation analysis makes use of these datasets to provide meaningful predictions about the function of proteins by mathematically computing the similarity of a gene mutant’s interaction profile (phenotypic or genetic) to others. In this chapter, we characterize the BLOC complex based on predictions derived from positive and negative phenotypic correlations. BLOC is a six member yeast complex that contains three proteins related to members of the   mammalian BLOC-1 complex, (Hayes et al., 2011). However, the function of this complex is unknown in yeast. Through correlation analysis, we found that gene mutants corresponding to yeast BLOC and ESCRT-0/1 complexes, responsible for formation of multivesicular bodies in the late endosome, had positively correlated chemicogenomic (response to drug challenges) profiles. Thus, we predicted that this BLOC complex might have a role in endosomal sorting similar to ESCRTs. We explored the relationship between BLOC and ESCRTs and found that these complexes localize to the same compartment.   Additionally, our analysis identified a negative correlated phenotypic profiles between members of BLOC and the manganese transporter Smf1, suggesting an inhibitory relationship. We found BLOC mutants caused a delay in Smf1 degradation similar to the ESCRT mutant, mvb12Δ. Thus, the BLOC complex is needed to target specific proteins to the vacuole. In all, we found that using correlation analysis can predict functional relationships of uncharacterized protein complexes by comparing profiles with known complexes.   2.2 Introduction  Most cellular processes are carried out by multiprotein complexes. Thus, identification and characterization of these complexes is important for a global understanding of cellular physiology. The global identification and characterization of proteins has utilized 25  a two-prong approach, 1) using proteomics (identifying protein-protein interactions) and 2) screens for functional relationships (phenotypic or genetic interaction screens). Thus far, high-throughput proteomics studies have allowed the characterization of protein complexes in diverse organisms including the budding yeast (Gavin et al., 2006; Krogan et al., 2006; Collins et al., 2007; Babu et al., 2012), bacteria (Butland et al., 2005; Hu et al., 2009), fruit fly (Veraksa et al., 2005), and mammalian cell systems (Mak et al., 2010). Further, large-scale screens using genetic interactions, which define pairs of genes based on their fitness (growth or other phenotypes) of the corresponding double mutant (Beltrao et al., 2010), or drug interaction studies (Costanzo et al., 2010; Hoepfner et al., 2014) have been used to identify functional relationships.  Correlation analysis, a statistical measure of whether two gene mutants have similar interaction profiles (i.e. genetic or chemicogenomic) (Michaut and Bader, 2012) provides an improvement to genetic interaction or phenotypic interaction datasets. As this method relies on pattern of interactions rather than a single interaction, experimental variation is limited and the functional predictions are more robust. Highly correlated gene mutant profiles have been already shown to indicate genes that have a similar function (Costanzo et al., 2010) and anticorrelated profiles may indicate an antagonistic relationship. We used correlation coefficient values derived from a chemicogenomics (measures the cellular response to different drug challenges) dataset mapped onto known protein complexes and found correlations and anticorrelations that suggest that an uncharacterized complex in yeast, BLOC, may have a role in endosomal sorting.  Sorting of proteins to the vacuole can occur through several different pathways controlled by specific protein complexes. Two well-known examples are the Endosomal Sorting Complex Required for Transport (ESCRT) complex and the Adaptor Protein 3 (AP-3) pathway. ESCRT is responsible for sorting proteins at the endosome into intraluminal vesicles for degradation in the lysosome (Williams and Urbé, 2007). Conversely, the AP-3 cargo adaptor complex targets proteins from the Golgi/early endosome to the limiting membranes of the lysosome or lysosome-related organelles and these proteins avoid internalization into the multivesicular body (Bonifacino and Traub, 2003; Dell’Angelica, 2009). 26  Additional sorting complexes have been identified in metazoans that were not thought to be present in yeast. For example, the three mammalian BLOC complexes (BLOC-1, BLOC-2, and BLOC-3) also make an important contribution to sorting to lysosome-related organelles (Falcón-Pérez et al., 2002; Martina et al., 2003; Nazarian et al., 2003; Di Pietro et al., 2006). Mammalian BLOC-1 is a conserved octameric complex that is required for cargo sorting in early endosomes and is functionally linked to the AP-3 complex as well as to BLOC-2 and BLOC-3 (Raposo and Marks, 2007; Setty et al., 2007; Newell-litwa et al., 2009; Hayes et al., 2011). In humans, mutations in genes that encode subunits of AP-3 and BLOC-1, 2, and 3 have been linked to Hermansky-Pudlak syndrome, an autosomal recessive disorder characterized by its defects in melanosomes, platelet dense granules, and other lysosome-related organelles, underlining their shared function (Dell’Angelica, 2004). BLOC-1 deletions affect the localization of many known AP-3 cargoes (Salazar et al., 2006). Although, there is evidence that some cargoes are BLOC-1 or AP-3 independent, suggesting that BLOC-1 may not always work with AP-3 (Raposo and Marks, 2007; Sitaram et al., 2012).  There is no evidence of BLOC-2 and BLOC-3 complexes in yeast and yeast lack specialized lysosome-related organelles. However, three yeast proteins were recently found to have sequence homology to subunits of the mammalian BLOC-1 complex (Hayes et al., 2011). High-throughput purification experiments identified these three proteins as predicted components of a six-subunit complex now known as BLOC (Krogan et al., 2006; Hayes et al., 2011; Babu et al., 2012), but the functional roles of these proteins have not been characterized in detail. Here, we have identified genetic connections between yeast BLOC and ESCRT, suggesting that they have a similar function. Further we found that these two complexes localize to the same compartment and are both needed for correct downregulation of the manganese transporter, Smf1. Much of this work was corroborated in a paper published while this work was in progress (Peter et al., 2013).  27  2.3 Results 2.3.1 Yeast BLOC is not Required for Sorting AP-3 Cargo Proteins A recent large-scale mass spectrometry study indicated that three yeast proteins related to subunits of mammalian BLOC-1 (Cnl1, Snn1, and Bls1) form a complex that also contains the conserved protein Kxd1 and the yeast-specific components Vab2 and Bli1 (Babu et al., 2012).  The mouse Kxd1 homolog was recently implicated in the biogenesis of lysosome-related organelles including melanosomes and platelet dense granules (Yang et al., 2012), suggesting it could have a conserved function related to BLOC-1.   In mammalian cells, the BLOC-1 complex functions in the adaptor protein 3 (AP-3) trafficking pathway (for review see Dell’Angelica, 2004). Thus, we tested the role of yeast BLOC in the trafficking of a novel AP-3 specific reporter construct, GNSS (GFP-Nyv1-Snc1-Suc2), which is based on the GNS (GFP-Nyv1-Snc1) reporter (Reggiori et al., 2000). The cytosolic domain of the Nyv1 v-SNARE targets this construct to the vacuolar limiting membrane when the AP-3 pathway is functional (Figure 2.1A). However, when AP-3 is not available, the Snc1 domain targets the construct to the cell surface (Reggiori et al., 2000). The localization of the reporter can be monitored using the cytosolic GFP tag, and the surface-exposed invertase (Suc2) activity can be quantitated using a colorimetric approach to determine the extent of missorting to the plasma membrane (Figure 2.1B-D) (Darsow et al., 2000; Burston et al., 2008; Dalton et al., 2015). We found that in BLOC mutants and in wild-type cells there was not any noticeable GNSS reporter on the cell surface based on Suc2 activity, and instead the GNSS reporter was localized to the vacuole. In contrast, the AP-3 mutant (apl6Δ) showed the expected mislocalization of GNSS to the plasma membrane and was the only mutant to create a dark color in the plate based invertase assay or convert sucrose into glucose for the liquid invertase assay, suggesting that only the AP-3 mutant contains cell surface localized GNSS reporter (Fig 2.1B-D). Because these assays did not detect even a subtle degree of missorting in BLOC mutants, we conclude that the yeast BLOC complex is not required for the functioning of the AP-3 pathway.  28    Figure 2.1: Yeast BLOC does not participate in the AP-3 pathway. A) Schematic diagram of the GNSS (GFP-Nyv1-Snc1-Suc2) reporter used to monitor AP-3 function. B) Suc2 (Invertase) activity on the cell surface was measured in a colorometric assay where glucose liberated from sucrose by the invertase enzyme (Suc2) is enzymatically converted to produce a pink color measured by absorbance at 540nM. GNSS will only be present on the cell surface when the AP-3 pathway is defective. Therefore a high absorbance at 540nM indicates more GNSS plasma membrane localization.  Absorbance values were converted to nmoles of glucose based on a standard curve. suc2∆ strains lacks invertase and should not secrete any glucose. Error bars report standard error of the mean; n=3 C) Invertase Overlay Assay. When overlaid with agar/invertase reagents the glucose liberated is monitored by the formation of a dark brown color. A dark color indicates greater invertase activity at the cell surface and thus a defective AP-3 pathway. D) Live-cell fluorescent microscopy using genomically integrated GFP as well as the lipophilic stain FM4-64 which outlines the yeast vacuole. All images were scaled the same. Scale Bar=2µM  2.3.2 BLOC Colocalizes with ESCRT-0  We used correlation analysis to identify other protein complexes that may be functionally associated with BLOC. Several studies have shown that genes with similar phenotypic profiles are often involved in the same biochemical pathways or form multi-29  protein complexes (Boone et al., 2007; Decourty et al., 2008; Ma et al., 2008; Beltrao et al., 2010). Uncentered Pearson’s correlation coefficients were calculated for each gene based on its mutant growth rate in a large-scale drug screen compared to other gene mutants in the dataset (Hoepfner et al., 2014).  These correlation coefficients were then mapped onto known complexes using a Cytoscape plugin (GIPro) developed by our collaborator, Shuye Pu in the Wodak lab. This analysis showed that members of the yeast BLOC complex were highly correlated with members of ESCRT-0, and ESCRT-1 complexes (Figure 2.2A), suggesting BLOC and ESCRT-0 and ESCRT-1 share a functional connection.  To determine if BLOC is functionally linked to ESCRTs, we first tested the colocalization of BLOC with ESCRT. BLOC subunits were C-terminally tagged with green fluorescent protein (GFP) at the chromosomal locus and co-expressed with the plasmid-encoded ESCRT-III subunit Snf7 tagged with red fluorescent protein (RFP). Localization of the BLOC subunits showed a punctate pattern near the vacuole of the yeast cells (Figure 2.2B top). We quantified the percentage of Snf7-RFP puncta that overlapped with the BLOC-GFP puncta (Figure 2.2C) and found fifty to eighty percent of Snf7 puncta also contained a BLOC-GFP signal. This degree of overlap was comparable to that between Snf7-RFP and the ESCRT-0 subunit Vps27-GFP. Thus, yeast BLOC, like ESCRT, is present at the yeast multivesicular body/late endosome. Since yeast BLOC colocalized with ESCRT subunits, we tested whether yeast BLOC subunits are required to recruit yeast ESCRT to the late endosomes. We found the punctate distribution of the Vps27-GFP was unaffected by loss of yeast BLOC subunits, and conversely, the localization of yeast BLOC subunits did not require Vps27 (Figure 2.3A), suggesting that BLOC and ESCRT complexes are independently targeted to the endosomal membrane.  Additionally, Vps27, a subunit of the ESCRT-0 complex, has FYVE domains that bind to phosphoinositol-3-phosphate (PI3P) and allow recruitment to endosomes. We tested whether BLOC subunits tagged with GFP required PI3P to localize to endosomes.  GFP-tagged BLOC subunits showed a punctate pattern in the vps34Δ cells as in wild-type (Figure 2.3B), which indicates that PI3P is not required to target BLOC subunits to the endosome. 30    Figure 2.2: BLOC is linked to ESCRTs. A) Enriched positive correlations between BLOC and ESCRT complexes, p<0.05. Correlations were calculated based on a chemicogenomics dataset (Hoepfner et al., 2014). 31  Protein-protein interactions were derived from a BIOGRID database B) Duo-labeled fluorescence microscopy of cells co-expressing the indicated GFP-tagged proteins and plasmid-encoded pSnf7-RFP. C) Quantification of pSnf7-RFP puncta that co-localize with the indicated GFP-tagged proteins. At least 100 puncta were counted per trial for each indicated strain. Images scaled independently for best visualization. Scale bar equals 2μm   Figure 2.3: BLOC and Vps27 are independently recruited to the endosome and BLOC recruitment is PI3P independent. A) Localization of overexpressed GFP-tagged Vps27 in WT and BLOC mutants. B) Localization of BLOC-GFP subunits in WT and vps27Δ and vps34Δ mutant strains. Scale bar equals 2μm.  2.3.3 BLOC is Required for the Endosomal Sorting of Smf1 To characterize the role of the yeast BLOC complex, we tested if cargo sorting into intraluminal vesicles at the late endosome was affected by the loss of the BLOC subunits. Ste3 and Sna3 are two well-characterized proteins transported to the vacuole by the ESCRT machinery (Macdonald et al., 2012). Ste3, a pheromone receptor, is targeted for 32  degradation in the vacuole via a ubiquitin-dependent pathway, when its ligand is not present (Chen and Davis, 2002). Sna3 is also trafficked from the plasma membrane and constitutively degraded in the vacuole requiring ubiquitination to enter the multivesicular body (Macdonald et al., 2012). In the ESCRT mutant mvb12Δ, the formation of intraluminal vesicles is impaired, causing the mislocalization of Ste3-GFP and Sna3-GFP to a punctate structure adjacent to the vacuole. However, in the BLOC mutants, Ste3-GFP and Sna3-GFP were primarily localized within the vacuole lumen, similar to wild-type cells (Figure 2.4A). Thus, BLOC does not seem to be needed for the trafficking of all proteins through the endosomal system. To identify potential BLOC cargo proteins, we again used correlation analysis using the GIPro Cytoscape plugin. We identified negative (anti) correlations between members of BLOC and Smf1 and positive correlations between members of BLOC and Bsd2 and Tre1, both of which are required for manganese transport (Supek et al., 1996; Jensen et al., 2009) (Figure 2.4B). When manganese concentrations are not limiting, the manganese transporter Smf1 is rapidly endocytosed and targeted to the vacuole where it is degraded (Sullivan et al., 2007). Tre1 and Bsd2 are required for ubiquitination and thus targeting to the vacuole (Jensen et al., 2009). Thus, it appears that the correlations between BLOC and Tre1 and Bsd2 and the negative correlation between BLOC and Smf1 indicate BLOC may also be responsible for Smf1 degradation. We found that GFP-Smf1 was localized to the vacuolar lumen in wild-type cells, but was retained at the cell surface in the absence of the ubiquitin ligase adaptor Bsd2. In cells lacking Mvb12 (ESCRT-1 subunit), Smf1-GFP was largely absent from the vacuole lumen and instead accumulated in a puncta adjacent to the vacuole, characteristic of ESCRT mutants. The BLOC mutants showed a distinct but much weaker phenotype; in approximately 50% of cells, Smf1-GFP was observed in punctate structures next to the vacuole, in addition to the vacuole lumen (Figure 2.4C). Taken together, these results suggest that BLOC contributes to the sorting of Smf1 at endosomes but is not required for the transport of all cargo through the endocytic pathway (Figure 2.5).   33   Figure 2.4: BLOC specifically affects Smf1 trafficking. A) Localization of Ste3-GFP or Sna3-GFP in the indicated wild-type and mutant strains. B) Correlation analysis using the GIPro plugin in Cytoscape. Correlations are derived from the Novartis drug dataset, BIOGRID protein interaction dataset, and protein complexes defined by the CYC2014 dataset. Correlations are shown with p<0.05 C) Effect of BLOC on the endosomal trafficking of GFP-Smf1. Left - localization of GFP-Smf1 in the indicated wild-type and mutant cells. For some GFP images, the intensity was changed compared to the others to allow better visualization of co-localizing structures. For these, an image with consistent scaling is shown in the inset. Scale bar equals 2μm. Right - quantification of GFP-Smf1 localization pattern. n=3 Scale bars report Standard Error of the Mean. 34   Figure 2.5: Proposed model of BLOC action at the endosome to sort Smf1 to the vacuole. Smf1 is internalized from the plasma membrane and BLOC is needed along with ESCRT for efficient incorporation of Smf1 into multivesicular bodies and subsequent delivery to the vacuolar lumen. This illustration was made by our collaborator Mohan Babu.  2.4 Discussion We have identified a functional connection between BLOC and ESCRT in yeast and found that BLOC, like ESCRT, is required to sort cargo proteins at the multivesicular body for transport to the vacuole. BLOC is present at endosomes where it aids in sorting Smf1 but is not required to sort all proteins that enter multivesicular bodies. While this work was in progress Peter et al reported many of the same findings corroborating the results found here (Peter et al., 2013). PI3P production by the PI 3-kinase Vps34 was not required for BLOC to localize to membranes. However, while this work was in preparation, Peter et al reported that BLOC interacts with the active, GTP-bound form of the yeast Rab5 homolog, Vps21, which is needed for appropriate localization of Vps34 and thus can affect endosomal pools of PI3P. The interaction between Vps21 and BLOC was shown to be needed for BLOC recruitment (Peter et al., 2013). Thus, it appears that interaction with Vps21 rather than the role of Vps21 in PI3P production is needed for BLOC recruitment to endosomes.  35  Loss of the BLOC complex did not affect ESCRT-0 or ESCRT-I recruitment to endosomes (data not shown) suggesting that BLOC does not regulate ESCRT assembly. Instead, BLOC may cooperate with ESCRT-0 in cargo recruitment. A number of other proteins have partially redundant functions with ESCRT-0 as cargo adaptors for MVB sorting. For example, the mammalian Alix protein and its yeast homolog Bro1 have a ubiquitin-binding motif that is critical for MVB sorting of ubiquitinated cargo when ubiquitin binding by ESCRT-0 is attenuated (Pashkova et al., 2013). Furthermore, AP-3 has recently been shown to interact with Alix to mediate the ubiquitin-independent sorting of PAR-1 (Dores et al., 2012), bypassing the ESCRT-0 requirement for MVB sorting.  If BLOC is indeed an alternative cargo receptor, it may be required for sorting a specific set of cargo proteins. Only Smf1 was affected by BLOC mutants in our studies, while many other cargoes were unaffected. Carboxypeptidase Y is protease that is recognized at the Golgi by its receptor, Vps10, for transport to the vacuole. Many proteins that affect endosome integrity and trafficking through the late endosomes perturb Vps10 recycling and cause CPY secretion when deleted; however, this is not the case with BLOC proteins (Schluter et al., 2008). Sna3 and Ste3 are receptors present at the plasma membrane and are trafficked to the vacuole for degradation. Likewise, these are unaffected by mutations in BLOC subunits. Peter et al. reported that trafficking of the lysine permease Lyp1 and the arginine permease Can1 were affected by mutations in BLOC subunits causing resistance to thiolysine and canavanine, respectively (Peter et al., 2013). In agreement with our findings, they did not uncover any perturbation in the trafficking of other cargo tested. This supports the model that BLOC acts as a cargo adaptor to transport specific proteins. Smf1, Can1, and Lyp1 are all transporters whose internalization and subsequent trafficking is regulated by arrestin binding and ubiquitination in response to cellular levels of metals and/or amino acids (Nikko et al., 2008). It is possible that BLOC is specifically involved in trafficking of proteins in response to signaling events, and could recognize specific features of a subclass of cargo proteins to enhance MVB sorting.  ESCRT-0 also binds additional factors that influence the ubiquitination state of cargo proteins and thus influences their sorting. ESCRT-0 recruits the ubiquitin ligase, Rsp5, and the deubiquitin ligase, Doa4, to appropriately control ubiquitin at this trafficking step (Ren et 36  al., 2008). When either recruitment is impaired, the sorting of cargoes to the vacuole is slowed. BLOC could similarly be needed to recruit additional factors that enhance dynamic regulation of ubiquitination, or otherwise modulate cargo sorting. Interestingly, Peters et al reported that BLOC recruits the Vps21 GAP protein Msb3 to inactivate Vps21 (Peter et al., 2013). They propose that recruitment of Msb3 to the endosome inactivates Vps21 and thus promotes the Rab GTPase conversion necessary for the progression of early endosomes to late endosomes. While it would seem that a defect in endosomal maturation would cause global sorting defects, msb3Δ cells, like BLOC mutants, do not show defects in trafficking many proteins through endosomes (Lachmann et al., 2012).  The connection to ESCRT and Msb3 are both likely to be relevant for BLOC function as Vps21 is overactive in ESCRT mutants (Russell et al., 2012). Additionally, a number of other sorting complexes have been shown to recruit RabGAPs in higher eukaryotic cells. For instance, Optineurin is a Golgi-localized cargo adaptor protein in mammalian cells that also recruits the RabGAP TBC1D17 to regulate vesicular trafficking (Vaibhava et al., 2012). Also, subunits of the mammalian retromer cargo-recognition subcomplex bind and recruit the Rab7 GAP, TBC1D5, to negatively regulate Rab7 (Seaman et al., 2009). BLOC was recently shown to be needed for appropriate downregulation of Vps21 through ER-relocalization similar to a vacuolar fusion (HOPS) mutant (Rana et al., 2015). BLOC could, therefore, function as an ESCRT-associated cargo adaptor or ubiquitin modifier while also promoting endosomal maturation by recruiting the Vps21 GAP, Msb3. The yeast BLOC complex was initially identified based on its sequence homology to mammalian BLOC-1 (Hayes et al., 2011). In mammals, BLOC-1 works with the AP-3 cargo adaptor to sort proteins to the limiting membrane of multivesicular bodies (MVBs) and lysosome/lysosome-related organelles (Salazar et al., 2006; Dell’Angelica, 2009). We were, therefore, surprised to find that yeast BLOC is not required for the sorting of AP-3 cargo proteins. However, BLOC-1 is not required for sorting of all mammalian AP-3 cargoes and we only tested the chimeric AP-3 cargo GNSS, which leaves the possibility that yeast BLOC also coordinates with AP-3 to sort a specific set of proteins from this pathway. In addition, recent evidence suggests BLOC-1 can also work with ESCRT in mammalian cells. Melanosomal proteins are generally missorted when AP-3 is dysfunctional, however, 37  Tyrosinase related protein 1, Trp1, has been shown to require BLOC-1 and ESCRT-1 for normal trafficking where ESCRT-1 acts downstream of BLOC-1 (Truschel et al., 2009). Furthermore, the AP-3 complex recognizes a signal in PAR-1, a GPCR for thrombin, and binds Alix to promote the ESCRT-dependent but ubiquitin-independent sorting of Par-1 to intraluminal vesicles (Dores et al., 2012). Thus, there is still much to be learned about the functional connections between BLOC-1, ESCRT, and AP-3 in yeast and mammals. Finally, a new complex has been recently identified in mammals called BORC which contains three of the yeast BLOC homologs including Kxd1, which is not thought to be incorporated in mammalian BLOC-1 (Pu et al., 2015).  The function of this complex is not well understood but it may contribute to maintenance and positioning of the lysosome (Pu et al., 2015).  Thus, the positive correlation between BLOC and ESCRT may be explained by a shared general role in endosomal maturation where yeast BLOC affects trafficking indirectly through changes in vacuole placement and or endosomal progression (Langemeyer and Ungermann, 2015). However, since only some cargo proteins are affected, a general vacuolar maintenance role seems unlikely and thus, we cannot rule out a direct role in sorting. In summary, using correlation analysis, we characterized the yeast BLOC complex showing a role in endosomal sorting. BLOC may work under specialized circumstances to enhance ESCRT-mediated sorting of proteins at the multivesicular body. Further work will be needed to understand the apparent differences in yeast and mammalian BLOC complexes, which could provide further insight into the cellular pathways implicated in disorders such as Hermansky-Pudlak syndrome.   2.5 Methods 2.5.1 Network Analysis 2.5.1.1 Data Sources and Preprocessing The gene networks were generated using the GIPRO (http://wodaklab.org/gipro/), a Cytoscape plug-in developed for integrative analysis and visualization of phenotypic correlations, protein-protein interactions, and protein complexes. Phenotypic correlations were derived from the Novartis HIP-HOP chemicogenomics dataset (Hoepfner et al., 2014).. The chemicogenomics dataset was first filtered to remove dubious open reading frames 38  (ORF) based on gene descriptions from Saccharomyces Genome Database. This filtered dataset was subsequently used to compute the uncentered Pearson correlation coefficient. Protein complexes were defined from a manually curated set of protein complexes supported by literature evidence (CYC 2014). Protein-protein interactions were derived from the BIOGRID 2014 database (version 3.2.108).   2.5.1.2 Correlation Between Phenotypic Interaction Profiles The chemicogenomic interaction profile of a gene is a vector of scores quantifying its growth pattern to different drug challenges compared with all other genes in the dataset. To measure the similarity between genetic interaction profiles between each pair of genes, uncentered Pearson correlation (UPC) coefficient was computed using the filtered dataset. The UPC was computed as follows,    where ai and bi are interaction scores of gene a and gene b with gene i, respectively, and n is the total number of genes in the dataset. The choice of UPC over Pearson correlation is based on the ability of UPC to keep track of signs of genetic interactions (positive or negative) so as to favor identification of coherent interactions. For any pair of genes on the same chromosome, if their genomic coordinates are overlapping or the distance between them on the chromosome is less than 1500 base pairs (the maximum length of 3’-untranslated region is 1461 bp in budding yeast (Nagalakshmi et al., 2008), their correlation cannot be trusted, because deletion of one gene may perturb the structure/function of the other gene. Thus, such pairs were excluded from analysis. Finally, the UPC scores of reciprocal genes (AB and BA) were averaged if the sign of the scores was consistent (both were positive or both were negative); otherwise, both pairs were discarded. For UPC scores from multiple alleles of the same gene correlating with another gene, the UPC scores were averaged to give rise to a single score for the particular gene pair. 39  2.5.2 Strain Construction  The yeast strains and plasmids used in this study are described in Table 2.1 and Table 2.2. Standard rich media (YPD), synthetic media plus dextrose (SD) lacking histidine, and synthetic complete (SC) media containing 2% glucose as a carbon source were used for cell growth as required. SpeI-digested pSEC7-EGFPx3 (a gift from B. Glick) was used to integrate a C-terminal GFP tag at the SEC7 locus in a BY4741 strain. BLOC-GFP strains were made by integration of the GFP + cassette to the C-terminus of each BLOC subunit and, where indicated, were also transformed with pSnf7-RFP (pRS315 backbone). GFP-Smf1 strains were made by integration of the N-terminal GFP cassette (ADH promoter) by homologous recombination.  Table 2.1: Strains used in BLOC study (Chapter 2)  Strain Description Source BY4741 MAT a ura3Δ0 leu2Δ0 his3Δ1 met15Δ0  GE Dharmacon, YSC1048 BY4742 MAT α ura3Δ0 leu2Δ0 his3Δ1 lys2Δ0 GE Dharmacon, YSC1049 bli1Δ BY4741 bli1Δ::KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 Yeast Mat a deletion collection bls1Δ BY4741 bls1Δ:: KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 Yeast Mat a deletion collection cnl1Δ BY4741 cnl1Δ:: KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 Yeast Mat a deletion collection kxd1Δ BY4741 kxd11Δ:: KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 Yeast Mat a deletion collection snn1Δ BY4741 snn1Δ:: KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 Yeast Mat a deletion collection vab2Δ BY4741 vab2Δ:: KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 Yeast Mat a deletion collection mvb12Δ BY4741 mvb12Δ:: KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 Yeast Mat a deletion collection LDY12 BY4741 BLI1-GFP+::HIS3 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY13 BY4741 BLS1-GFP+:: HIS3 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY14 BY4741 CNL1-GFP+:: HIS3 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY15 BY4741 KXD1-GFP+:: HIS3 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY16 BY4741 SNN1-GFP+:: HIS3  leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY25 BY4741 VAB2-GFP+:: HIS3 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY26 BY4741 vps27Δ:: KAN Bli1-GFP+::HIS3 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY27 BY4741 vps27Δ:: KAN BLS1-GFP+:: HIS3 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY28 BY4741 vps27Δ:: KAN CNL1-GFP+:: HIS3 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY29 BY4741 vps27Δ:: KAN SNN1-GFP+:: HIS3 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY30 BY4741 vps27Δ:: KAN VAB2-GFP+:: HIS3 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY35 BY4741 vps27Δ:: KAN KXD1-GFP+:: HIS3 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY17 BY4741 ADH1pr-GFP-SMF1::NAT bsd2Δ::KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY18 BY4741 ADH1pr-GFP- SMF1::NAT cnl1Δ::KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY19 BY4741 ADH1pr-GFP- SMF1::NAT bli1Δ::KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) 40  Strain Description Source LDY20 BY4741 ADH1pr-GFP- SMF1::NAT vps55Δ::KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY21 BY4741 ADH1pr-GFP-SMF1::NAT mvb12Δ::KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY22 BY4741 ADH1pr-GFP- SMF1::NAT his3Δ::KAN his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY192 BY4741 SEC7-3XEGFP::URA, his3Δ1 leu2Δ0 met15Δ0 This Study (Conibear lab) LDY40 BY4741 bli1Δ::KAN,  SUC2::ADH1pr-GFP-NYV1(cyt)-SNC1(TM)-SUC2::NAT his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY41 BY4741 bls1Δ::KAN,  SUC2::ADH1pr-GFP-NYV1(cyt)-SNC1(TM)-SUC2::NAT his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY42 BY4741 cnl1Δ:: KAN SUC2::ADH1pr-GFP-NYV1(cyt)-SNC1(TM)-SUC2::NAT his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY43 BY4741 kxd1Δ:: KAN SUC2::ADH1pr-GFP-NYV1(cyt)-SNC1(TM)-SUC2::NAT his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY44 BY4741 snn1Δ:: KAN SUC2::ADH1pr-GFP-NYV1(cyt)-SNC1(TM)-SUC2::NAT his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY45 BY4741 vab2Δ:: KAN SUC2::ADH1pr-GFP-NYV1(cyt)-SNC1(TM)-SUC2::NAT his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY46 BY4741 mvb12Δ:: KAN SUC2::ADH1pr-GFP-NYV1(cyt)-SNC1(TM)-SUC2::NAT his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY47 BY4741 vps27Δ:: KAN SUC2::ADH1pr-GFP-NYV1(cyt)-SNC1(TM)-SUC2::NAT his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab) LDY49 BY4741 apl6Δ:: KAN SUC2::ADH1pr-GFP-NYV1(cyt)-SNC1(TM)-SUC2::NAT his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 This Study (Conibear lab)  Table 2.2: Plasmids used in BLOC study (Chapter 2) Plasmid Description Source pCS10 pSNF7-RFP (Leu) This Study (Conibear Lab) pJLU34 pSTE3-GFP (ura) Urbanowski and Piper (1999) pGFP-Sna3 pGFP-SNA3 (leu) Reggiori & Pelham 2001  2.5.3 Invertase Overlay Assay  Strains expressing GNSS reporters integrated at the SUC2 locus were patched onto YP-fructose (YPF) and grown overnight (16-18h) at 30C. Plate was overlaid with 20mL of assay reagents (0.1 M K2HPO4, pH 7.0, 347.1 U glucose oxidase, 2.6 ng/mL HRP, 102.6 nM N-ethyl-maleimide, 0.15 mg/mL O-dianisidine, 1.4% agar). Plates were allowed to react for 35 minutes before scanning in color with a CanoScan 4400F scanner.   2.5.4 Invertase Liquid Assay  The liquid invertase assay was modified from Darsow et al., 2000 (Dalton et al., 2015). Strains expressing GNSS reporters integrated into the SUC2 locus were grown 41  overnight in 2 mL yeast peptone fructose media (OD600 ≈ 6). After dilution to 3 OD600/mL, 5 µl of each culture was mixed with 55 µl of 0.1 M NaOAc buffer, pH 4.9, in a 96-well microtiter plate. Subsequent steps were done using a multi-channel pipette with staggered times to create the most consistency: 13 µl of freshly prepared 0.5 M ultra-pure sucrose was added to each well, and the plate was incubated for 5 min at room temp. Next, 100 µl of glucostat reagent (0.1 M K2HPO4, pH 7.0, 347.1 U glucose oxidase, 2.6 ng/mL HRP, 102.6 nM N-ethyl-maleimide, and 0.15 mg/mL O-dianisidine) was added to each well, and the plate was incubated for 12 min at room temp. Finally, 100 µl of 6 N HCl was added to each well to stop the reaction. The absorbance at 540 nm was measured using a VICTOR 1420 platereader (PerkinElmer) to determine glucose concentration by comparison to a glucose standard curve (5 to 50 nM glucose). Results are reported as nanomoles of glucose.   2.5.5 Fluorescence Microscopy  Cells were grown to mid-log phase, concentrated then spotted onto a microscope slide for image acquisition. Images were captured with a CoolSNAP camera (Roper Scientific) using MetaMorph software (Molecular Devices) and adjusted using Adobe Photoshop (Adobe Systems). Images were acquired for 3000ms and were then scaled for consistent brightness and contrast. The images were then converted to an 8-bit format. In Photoshop, the images were converted to 300 dpi from the 72 dpi camera output, scaled to 350 pixels by 350 pixels (resampled), changed to RGB, merged if applicable. In some cases, the images were not scaled to the same brightness and contrast. Under these circumstances, the 350x350 pixel view represents the best visualization while the small 110x110 pixel region represents the same brightness and contrast as the other samples.  2.6 Acknowledgements We thank Benjamin Glick for providing pSEC7-EGFPx3. This research was supported by grants from the Canadian Institutes of Health Research to Elizabeth Conibear and Mohan Babu.   42  Chapter 3: Negative Correlation Analysis Reveals an Antagonistic Relationship Between the VID Complex and the Previously Uncharacterized Protein Vaf1   3.1 Synopsis Positive correlations between two gene mutant interaction profiles can predict that their respective proteins have a similar function (Costanzo et al., 2010). However, while rare, it is also possible for gene mutants, whose proteins are present in the same complex, to be negatively (anti) correlated, where the phenotypic or genetic interaction profiles are exactly opposite.   We hypothesized that while rare, robust predictions can be made from negatively correlated relationships of genes, indicating an antagonistic relationship between the two proteins.  Through this analysis, we found known inhibitory relationships that have been supported experimentally as well as new potential inhibitory relationships within complexes. To further understand the predictive power of these negative correlations, we followed up on an negatively correlated genetic interaction profiles between an uncharacterized open reading frame (ORF) YDL176W, we call VID Associated Factor 1 (Vaf1) and members of the Vacuolar Import and Degradation (VID) complex,  The Vacuolar Import and Degradation (VID) complex is a highly conserved multi-subunit E3 ligase whose function in mammalian cells is not well understood. In yeast, this complex has been shown to be involved in both with proteasomal and vacuolar degradation pathways (Brown et al., 2008; Santt et al., 2008; Snowdon and van der Merwe, 2012). Although Vaf1 has previously been found associated with the VID complex in high-throughput proteomics as well as computational interaction studies, its function relationship with the VID complex has not been explored (Pitre et al., 2006; Babu et al., 2012).  We found that Vaf1 binds and is downregulated by the VID complex in a proteasome-dependent manner. However, further work will be needed to determine if Vaf1 is simply a substrate of the VID complex or if Vaf1 activates VID for another role and is degraded as part of a negative feedback loop. In all, negative correlation analysis is a robust tool to identify regulatory relationships within known complexes.  43  3.2 Introduction Correlations are a statistical measure of the similarity of two genes based on interaction profiles (genetic or phenotypic) and are more robust predictors of functional relationships between proteins compared to a single phenotype or genetic interaction (Michaut and Bader, 2012). Positive correlations are reliable predictors of functionally similar proteins (Costanzo et al., 2010). However, while rare, it is also possible for genes within a complex to be negatively (anti) correlated, where the phenotypic profiles are exactly opposite. The predictive power of these anti-correlations has not yet been explored.  One hypothesis is that negative correlations within a complex might reveal inhibitory relationships between their respective proteins.  For example, we might predict that negative regulators such as an inhibitory kinase might be negatively correlated with its target substrate because increased presence or absence of the kinase is in opposition with the functionality of the substrate, creating an opposite pattern of phenotypes.    By identifying and characterizing these negative correlations within a complex, we can uncover new inhibitory relationships within a cell. Uncentered Pearson’s correlation coefficients were calculated for each gene mutant compared to all the other gene mutants using a genome-wide genetic-interaction dataset (Ryan et al., 2012).  These coefficients were mapped onto known (hand curated literature-supported) complexes. Of the negative correlations identified in this analysis, we followed up on the negative correlation between members of Vacuole Import and Degradation (VID) complex and an uncharacterized open reading frame YDL176W, which we call VID Associated Factor 1 (Vaf1).  E3 ubiquitin ligase complexes ubiquitinate and target proteins for either proteasomal or vacuolar degradation, allowing a quick response to changing cellular environments (Varshavsky A, 1997; Piper and Lehner, 2011). The Vacuolar Import and Degradation (VID) complex is a highly conserved multi-subunit E3 ligase whose function in mammalian cells is not well understood. In yeast, the VID complex has been linked to many cellular processes depending on the nutrient status of the cell.  The most well-characterized role of VID in yeast is targeting Fructose 1,6 bisphosphatase (FBPase) for proteasomal degradation, after the restoration of glucose to glucose-starved cells (Brown et al., 2008; Santt et al., 2008). One member of the complex, 44  VID24 activates the rest of the complex for degradation of FBPase and in the process is also targeted by the complex for degradation (Santt et al., 2008).  This creates a negative feedback loop to control VID activity. Other roles for the VID complex include an autophagy-like role when cells are given glucose after an extended period of starvation (Brown et al., 2008).  In this process, FBPase is incorporated into vesicles containing COPI coat proteins (Brown et al., 2008). These vesicles merge with the vacuole to promote degradation of FBPase (Brown et al., 2008). Finally, VID has been linked to the degradation of a hexose transporter during nitrogen starvation (Snowdon et al., 2008; Snowdon and van der Merwe, 2012). Thus, this complex has many roles in both proteasomal as well as vacuolar degradation depending on conditions within the cell. However, it is not known what factors if any direct this complex to its various roles. We explored the negative correlation identified between VID and Vaf1 to understand how this contributes to VID function and also to validate the predictive power of negative correlations within a protein complex.  3.3 Results 3.3.1 Negative Correlations are Somewhat Rare Within Complexes Positive genetic correlations between gene mutant genetic interaction profiles have been used to predict proteins that have a similar function (Costanzo et al., 2010). However, negative genetic correlations are not only rarer, but the predictive value of these correlations is relatively unknown.  Our collaborators developed a Cytoscape plugin called GIPro and we used this to visualize networks containing uncentered pearson’s correlation coefficients, a mathematical representation of the similarity of genetic interaction profiles, based on a recent genetic interaction dataset (Ryan et al., 2012), BIOGRID protein interaction dataset (version 3.2.108), and a hand curated protein complex database, thereby  generating a genome-wide set of correlations mapped onto 443 known protein complexes. We identified 49 complexes that contained at least one negative correlation within a p<0.25, and 22 complexes with a negative correlation with a p<0.05.  Thus, negative correlations within a complex are rare in comparison to complexes containing only positive correlations (Figure 3.1 and Table 3.1).    45  Table 3.1: Complexes identified that contain a negative correlation. Using the GIPro plugin, genetic interaction data were combined with known protein interaction and protein complex datasets.  Complexes represented here have at least negative correlation p<0.25 and bolded at p<0.05.  If more than one negative correlation occurs within the complex, the most negative value is shown. Complex Name Negative Correlation Value  Sis2p/Ppz1p Phosphatase Complex -0.2808 Seh1-Associated Complex (SEA) -0.2802 SAGA Complex -0.2277 Kornberg's Mediator (SRB) Complex -0.226 Cytoplasmic Ribosomal Small Subunit -0.2119 HMC Complex -0.1742 Cytoplasmic Ribosomal Large Subunit -0.1642 Eisosome -0.1591 SLIK (SAGA-like) Complex -0.1561 Pho85p/Pcl2p Complex -0.1527 cAMP-dependent Protein Kinase -0.1492 Alpha-1,6-Mannosyltransferase Complex (Anp1p/Mnn9p) -0.1471 COMPASS complex -0.1375 Cytochrome bc1 Complex -0.1238 Sap185p/Sit4p Complex -0.1103 Phosphatidylinositol (PtdIns) 3-kinase Complex -0.1093 ARGR complex -0.1077 CCR4-NOT Core Complex -0.1075 Sap190p/Sit4p Complex -0.1038 Vacuolar Import and Degradation (VID) Complex -0.0846 Pho85p/Pcl1p Complex -0.0833 Cdc73p/Paf1p Complex -0.073 Rpd3L Complex -0.067 Mitochondrial DNA-directed RNA Polymerase Complex -0.066 Mlh1p/Mlh3p Complex -0.0643 Set3p Complex -0.0639 TRC Complex -0.0585 Meiotic Recombination Initiation Complex -0.0557 Phosphopantothenoylcysteine Decarboxylase Complex -0.0551 Alpha,Alpha-Trehalose-Phosphate Synthase Complex  -0.0525 Polarisome -0.0513 Kel1p/Kel2p/Lte1p Complex -0.0491 Snt2C Complex -0.049 Pho85p/Pcl8p Complex -0.0476 Atg1p signaling Complex -0.0468 Calcineurin Complex -0.0467 Pho85p/Pho80p Complex -0.0444 mRNA Cap Binding Complex -0.0439 Sap155p/Sit4p Complex -0.0426 Synaptonemal Complex (SC) -0.0422 Chs5p/Arf-1 Binding Proteins (ChAPs) -0.0399 eEF1 -0.0391 Signalosome Complex -0.0319 NatA Complex -0.0316 Dcs1p/Dcs2 Heterodimer -0.0273 Fth1p/Fet5p Complex -0.0272 Mitochondrial Pyruvate Dehydrogenase Complex -0.0267 VTC Heterotetrameric Complex -0.026 46    Figure 3.1: Relatively few complexes contain negative correlations within a complex.  Using the GIPro plugin, genetic interaction data was combined with protein interaction and protein complex datasets.  Complexes counted here have at least one correlation either positive and/or negative with p<0.05.    3.3.2 Negative Correlations within Complexes Reveal Inhibitory Relationships  We anticipated that at least some of the observed anti-correlated profiles might link inhibitory subunits to their target proteins.  As expected, our analysis identified a negative correlation between the gene mutants representing the protein phosphatase Ppz1and its known inhibitor phosphopantothenol cysteine decarboxylase Sis2/Hal3 (Figure 3.2) (De Nadal et al., 1998; Clotet et al., 1999; Yenush et al., 2002), supporting our theory that negative correlations can indicate direct inhibitory interactions between enzyme and inhibitor.  We also identified negative correlations that suggest direct inhibitory relationships are not limited to a pair of proteins, but can also identify situations where multiple proteins form an inhibitory subcomplex.  As an example, the yeast vacuolar SEA (Seh1-Associated) complex, homologous to the mammalian GATOR complex, acts as an inhibitor of TORC1, a complex that controls cellular response to many different signals (Neklesa and Davis, 2009; Dokudovskaya et al., 2011).  Amino acid stimulation of TORC1 is regulated by the GTPases, Gtr1, and Gtr2 (Panchaud et al., 2013). The SEA complex contains two functional subcomplexes, SEACIT (SEAC Subcomplex Containing Iml1) containing Npr2/Npr3/Iml1 and SEACAT containing the remaining 5 subunits. SEACIT acts as a GTPase Activating 0 50 100 150 200 250 300 Positive Correlations Only Contain a Negative Correlation # Complexes 47  Protein (GAP), which stimulates the GTPase activity of Gtr1, causing Gtr1 to hydrolyze GTP and return to its inactive, GDP-bound form. In this way, SEACIT negatively regulates TORC1 (Panchaud et al., 2013). Negative correlations are also present between several members of SEACAT and the SEACIT subunits NPR3 and NPR2 (Figure 3.2). SEACAT has been shown to negatively regulate the GAP activity of SEACIT (Panchaud et al., 2013). Thus, SEACIT downregulates TORC1, but SEACAT can downregulate SEACIT which explains the negative correlations identified.  Another example of negative correlations revealing inhibitory subcomplexes is shown in the Mediator complex. Mediator is a large complex that consists of 4 modules: the head (Med6, Med8, Med11, Srb4, Srb5, Rox3, Srb2, and Srb6), middle (Med1, Med4, Nut1, Med7, Cse2, Nut2, Srb7, and Soh1), tail (Med2, Pgd1, Rgr1, Gal11, and Sin4) and cyclin kinase (CKM) modules (Ssn3, Ssn8, Srb8, and Ssn2). We observe that a number of negative correlations link the tail and CKM modules (Figure 3.2C). Complex antagonistic relationships between submodules have been extensively documented, and work in several labs has established that the CKM forms a negative regulatory submodule where the CKM phosphorylates the Pgd1/Med3 tail component, leading to its ubiquitination and turnover (van de Peppel et al., 2005; Poss et al., 2013; Gonzalez et al., 2014).  In this way, the negative correlation between the CKM module and the tail may be explained by the downregulation of the tail by the CKM module. Thus, our analysis found known regulatory relationships within complexes can be explained by direct inhibition: either a single protein with its target or an inhibitory subcomplex acting on the rest of the complex. This inhibition can be achieved through a variety of mechanisms including enzymatic inhibition, activation of GTP hydrolysis, and protein ubiquitination and turnover.  48   Figure 3.2: Examples of complexes whose negatively correlations can be explained by direct inhibition. All are derived using the GIPro plugin in Cytoscape. Green edges represent negative correlations, red edges are positive correlations and blue lines indicate protein-protein interactions. Gray circle outlines indicate that gene is found in another complex as well. Negative correlations are shown at p<0.25 and positive correlations are shown at p<0.05. The thickness of the line within a complex indicates the strength of the correlation.  3.3.3 Negative Correlations May Arise from Competition for Shared Subunits Not all negative correlations we found in this analysis can be explained by inhibitor-target type interactions. Instead, negatively correlated genetic interaction profiles for proteins within a complex may reflect competition for a shared interactor. Removal or modification of one component in a competing process may favor an opposing process, which then explains the negative correlation.  As an example, Sit4 binds to Sap155, Sap185, and Sap190 as three separate subcomplexes, and binding by the SAP proteins to Sit4 is needed for different cellular roles (Luke et al., 1996; Jablonowski et al., 2009). In our analysis, negative correlations were identified with all the Sap proteins and Sit4, likely reflecting the competition for binding to Sit4 (Fig3.3A).  A similar rationale underlies the anti-correlation between the three related catalytic subunits of the cAMP-dependent protein kinase, yeast protein kinase A (PKA), Tpk1, Tpk2, and Tpk3, which each form a complex with the inhibitory subunit Bcy1 (Figure 3.3B). The 49  three PKA catalytic subunits are 75% identical and share an overlapping essential function yet have different levels of activity and distinct transcriptional signatures (Broach, 2012). Thus, removing one of the Tpk proteins might increase the likelihood of the other subunits binding to the inhibitory Bcy1, explaining the negative correlations seen here.  In a final example, anti-correlations are observed between the PI3-kinase catalytic subunit VPS34 and its partners VPS30 and VPS38 (Figure 3.3C). The PI3-kinase catalytic subunit participates in two distinct heterotetrameric complexes: the Vps34/Vps15/Vps30/Atg14 complex (complex I) that regulates autophagy and the Vps34/Vps15/Vps30/Vps38 complex (complex II), required for vacuolar protein sorting (Kihara et al., 2001; Obara et al., 2006). However, common to both complexes, Vps15 and Vps34, have also been implicated in another role where Vps34 binds to Gpa1 to promote PI3P production required for pheromone response (Slessareva et al., 2006).  In this analysis, negative correlations were observed between VPS30/38 and VPS34.  Loss of Vps30 or Vps38 could release the Vps15/Vps34 dimer, enhancing its interaction with Gpa1 to promote signaling, explaining the negative correlation. Each of the above examples features one or more proteins that regulate distinct pathways yet compete for shared subunit(s). If this shared subunit is limiting, loss of one binding partner might favor association with the alternate interacting protein, thus enhancing the activity of the opposing pathway.   Figure 3.3: Examples of complexes whose negatively correlations can be explained by a competition model. All are derived using the GIPro plugin in Cytoscape. Green edges represent negative correlations, red edges are positive correlations and blue lines indicate protein-protein interactions. Negative correlations are 50  shown at p<0.25 and positive correlations are shown at p<0.05. Width of the line indicates strength of the correlation.  3.3.4 The Previously Uncharacterized Protein Vaf1 is Downregulated by the VID Complex One intriguing negative correlation found in our analysis was between the unknown ORF, YDL176W, which we call Vid Associated Factor (VAF1), and the other subunits of the Vacuolar Import and Degradation (VID) complex which has roles in both proteasomal and vacuolar degradation (Figure 3.4) (Brown et al., 2008; Santt et al., 2008; Snowdon and van der Merwe, 2012). Despite being implicated in many biological processes, a mechanistic understanding of the different roles of the VID complex is severely lacking. Vaf1 was previously found associated with the VID complex in a computational proteomics study; however, its role in the VID complex has not been further explored (Pitre et al. 2006). Thus, we used this example of negative correlations to characterize a previously unknown open reading frame (ORF), extend our knowledge of the VID complex and to assess if negative correlations have predictive value in determining novel functional interactions between proteins.   Figure 3.4: VAF1 is negatively correlated with the VID complex with respect to their genetic interaction profiles. A) Schematic indicating the physical interactions and genetic correlations linking components of the VID complex. Blue lines: physical interactions, Green lines: negative correlations, Red lines: positive correlations. Positive correlations are reported with a significance of p<0.05 and Negative Correlations a significance of p<0.25.  The Vacuolar Import and Degradation (VID) complex is a highly conserved multi-subunit E3 ligase whose function in mammalian cells is not well understood. In yeast, the 51  VID complex has been linked to many cellular processes depending on the nutrient status of the cell. To understand the relationship between VID and Vaf1, we first tested Vaf1 protein levels in a vid30Δ mutant, a core VID component (Pitre et al., 2006). We noticed that Vaf1-GFP was present at very low levels in wild-type cells, independent of the nutrient status. However, Vaf1-GFP levels were increased in vid30Δ mutants. This effect was rescued by a complementing VID30 plasmid (Figure 3.5A), suggesting that VID negatively regulates levels of Vaf1, which could explain the negative correlation. We then sought to understand why and how Vaf1 was downregulated by VID to determine if this relationship fit one of our models of negative correlation (eg. competition or direct inhibition). The VID complex is primarily known for its roles in proteasomal and vacuolar degradation but has also been reported to affect the transcription of some genes (van der Merwe et al., 2015). Thus, the effect of the VID complex on Vaf1 levels could be due to a reduction in transcription caused by mutation of Vid30. However, we found that the cellular level of Vaf1 expressed from the constitutive ADH promoter was still dependent on Vid30 (Figure 3.5B), suggesting that VID does not affect Vaf1 levels through transcriptional control.  It is reasonable that Vaf1 levels are higher in a vid30Δ mutant because VID targets Vaf1 for degradation. However, as VID targets proteins for both proteasomal and vacuolar degradation, we decided to test which mode of degradation is used by the VID complex to downregulate Vaf1. We found that there was a large increase in Vaf1 protein levels in a rpn4Δ mutant (needed for proteasome function). However, Vaf1 levels were unaffected in a pep4Δ mutant, which abolishes the activity of vacuolar proteases (Figure 3.5C). These results suggest that the VID complex negatively regulates cellular levels of Vaf1 by targeting it for proteasomal degradation.    52   Figure 3.5: Vaf1 binds strongly to and is downregulated by Vid30. A) Vaf1 was genomically tagged with GFP in vid30Δ mutants.  Either an empty plasmid (pRS416) or a pVID30 complementing plasmid was added. Cells grown to mid-log phase in synthetic media containing glucose (“F”, fed). Samples were subsequently starved for 18 hours in Synthetic Media media containing 2% EtOH as the sole carbon source (“S”, Starved). The cells were then transferred to fresh synthetic media with glucose harvested at  30 or 60 min time points.  The lysates were then subjected to SDS-PAGE and western blotting. The equivalent of O.4 OD600 and 0.05 OD600 was loaded for GFP and PGK lanes respectively. B) Vaf1 was HA-tagged and expressed from a heterologous ADH promoter in the indicated wild-type and mutant strains. PGK was used as a loading control. C) Vaf1-GFP was expressed in wild-type or the indicated mutant strains grown in rich media (YPD) and analyzed by western blotting. D) Lysates were prepared from cells expressing Fbp1-HA. Cells were grown in Synthetic media-Ura + 2% EtOH overnight and then shifted to SD-URA at time=0 min. PGK was used as a loading control. E) Vid24 was HA-tagged and expressed from a heterologous ADH promoter and in the indicated wild-type and mutant strains.  PGK was used as a loading control. F) Immunoprecipitation of samples with rabbit anti-GFP antibody. Loading of the IP samples was 20X relative to lysate for HA tagged proteins. G) Model of Vaf1 downregulation of VID. Vaf1may target the VID complex for a yet unknown target and then is downregulated, similar to the mode of action by VID24.  53  Thus far, our evidence that Vaf1 is downregulated by the VID complex supports a direct inhibition model where VID downregulates Vaf1 activity. However, another model is possible based on what is known about other VID components. For instance, Vid24, a component of the VID complex activates the VID complex to downregulate FBPase, and is subsequently targeted for degradation in a negative feedback loop (Brown et al., 2008). By extension, Vaf1 may activate VID and then subsequently be targeted for degradation. Loss of Vaf1 did not alter the rate of FBPase turnover after starvation and re-feeding (Figure 3.5D), suggesting it does not function with the VID complex to down-regulate gluconeogenesis. Thus, Vaf1 could activate VID to participate in other cellular roles.  To understand whether Vaf1 targets the VID complex for another role, we explored the relationship between Vid24 and Vaf1, because if both of these proteins target VID for different substrates, they may compete for binding with the core VID complex or regulate each other’s protein levels. We found that while levels of Vid24 and Vaf1 expressed under the ADH promoter were increased in vid30∆ mutants. Vaf1 is not affected by the deletion of Vid24 and deletion of Vaf1 does not affect Vid24 levels (Figure 3.5B and 3.5E). Therefore, Vaf1 and Vid24 do not appear dependent on each other. We also found that Vid24 can co-precipitate with Vaf1, suggesting they do not appear to compete for binding to same site on the VID complex (Figure 3.5F). Thus, if Vaf1 targets the VID complex for another role, it does not compete with Vid24. From the work thus far, the negative correlation can best be explained where the VID complex targets Vaf1 for degradation by the proteasome. Further work will be needed to determine if Vaf1 is simply a substrate of the VID complex, or if it has additional regulatory roles.  3.4 Discussion While it has been shown previously that highly correlated genetic interaction profiles are likely to predict functional similarities and anticorrelations are sometimes used to find proteins with opposing functions, using within complex anti-correlations to predict novel regulatory relationships has not been studied (Breslow et al., 2010; Costanzo et al., 2010; Michaut et al., 2011). Here we show how systematic negative correlation analysis can uncover new antagonistic relationships between proteins within a complex. Our analysis 54  identified many other negative correlations within complexes that control a wide variety of cellular processes including transcription, mitochondrial function, and lipid regulation, which will yield new and interesting avenues for future exploration. As more genetic interaction data is generated, we predict this technique will become an even more robust tool to identify functional relationships between proteins that act in a wide variety of cellular processes.  While the recent genetic interaction data we used in this analysis is a high-quality dataset yielding information for most of the genome and is the most comprehensive to date, coverage of the genome is far from comprehensive (Ryan et al., 2012). The incompleteness of the genetic interaction data may have biased our analysis, leading to an underestimation of total negative genetic correlations within a complex. As new data becomes available, many more negative correlations within complexes may be identified using this technique.  Additionally, we used one type of phenotypic data in our analysis (genetic interactions), however, many other phenotypic datasets exist such as chemicogenomic datasets (Giaever et al., 2002; Hillenmeyer et al., 2008; Hoepfner et al., 2014) Thus, it would be interesting to combine these datasets to create a more comprehensive profile of negative correlations in the cell.  Recently, various genetic interaction datasets have been combined and optimized to improve gene function prediction, suggesting that a similar approach may be used for integrating phenotypic datasets that can be used for correlation analysis (Michaut and Bader, 2012). Using different phenotypic datasets, we may be able to uncover a more comprehensive view of the regulatory relationships within a complex. We found that the uncharacterized ORF YDL176W, which we call Vaf1 (VID Associated Factor 1), has a negatively correlated genetic interaction profile compared to members of the VID complex. Our studies focused on the relationship of two subunits of the VID complex, Vid30 and Vid24, with Vaf1. However, there is evidence that only a subset of the VID proteins (Vid30, Vid24, and Vid28) are needed for both the vacuolar and proteasomal degradation of FBPase (Regelmann et al., 2003; Hung et al., 2004; Brown et al., 2008). In addition, Hxt7 turnover requires some but not all of the same proteins required for FBPase turnover, suggesting that there is a core complex that may associate with additional factors to perform different functions (Snowdon et al., 2008; Snowdon and van der Merwe, 55  2012) We only tested the effect of Vid30 and Vid24 on levels of Vaf1, thus we are yet unsure if other components also regulate Vaf1.  We found a negative association with Vaf1 and have explored many ways in which Vaf1 may act in a complex with VID as a functional component. We cannot, however, rule out that Vaf1 is merely a substrate of the VID complex. This is somewhat unlikely due to the strength of the physical interaction between Vid30, Vid24, and Vaf1 shown by co-immunoprecipitation. Often interactions between a ligase and its substrate are weak and difficult to detect (Rubel et al., 2013). According to Pitre et al, and in concordance with our own analysis, the binding between Vid30 and Vaf1 is strong (Pitre et al., 2006). Thus, we conclude that this interaction is more likely a regulatory interaction though more study is required to determine the role of the newly characterized protein, Vaf1.  3.5 Methods 3.5.1 Network Analysis 3.5.1.1 Data sources and Preprocessing The protein networks were generated using the GIPRO (http://wodaklab.org/gipro/), a Cytoscape plug-in developed for integrative analysis and visualization of genetic interactions, protein-protein interactions, and protein complexes. Genetic interactions were derived from large-scale genetic interaction screens (Ryan et al., 2012). The genetic interaction dataset was first filtered to remove dubious open reading frames (ORF) based on gene descriptions from Saccharomyces Genome Database. This filtered dataset was subsequently used to compute the uncentered Pearson correlation coefficients. Protein complexes were defined from a manually curated set of 443 literature supported (current to 2014) protein complex (CYC 2014 dataset). Protein-protein interactions were derived from the BIOGRID 2014 database (version 3.2.108). The dataset was thresholded according to the reported p-value.  3.5.1.2 Correlation Between Genetic Interaction Profiles The genetic interaction profile of a gene is a vector of scores quantifying its interactions with all other genes in the dataset. To measure the similarity between genetic 56  interaction profiles between each pair of genes, uncentered Pearson correlation (UPC) coefficient was computed using the filtered dataset. The UPC was computed as follows,    where ai and bi are interaction scores of gene a and gene b with gene i, respectively, and n is the total number of genes in the dataset. The choice of UPC over Pearson correlation is based on the ability of UPC to keep track of signs of genetic interactions (alleviating or aggravating) so as to favor identification of coherent interactions. For any pair of genes on the same chromosome, if their genomic coordinates are overlapping or the distance between them on the chromosome is less than 1500 base pairs (the maximum length of 3’-untranslated region is 1461bp in budding yeast (Nagalakshmi et al., 2008), their correlation cannot be trusted, because deletion of one gene may perturb the structure/function of the other gene. Thus, such pairs were excluded from analysis. Finally, the UPC scores of reciprocal genes (AB and BA) were averaged if the sign of the scores was consistent (both were positive or both were negative); otherwise, both pairs were discarded. For UPC scores from multiple alleles of the same gene correlating with another gene, the UPC scores were averaged to give rise to a single score for the particular gene pair.  3.5.2 Yeast Strains and Plasmids Yeast strains from this study were made by homologous recombination as described (Janke et al., 2004; Sheff and Thorn, 2004) (Table 3.2). Deletions were made by swapping the genetic locus with the marker of choice. Vaf1-GFP and Vid30-GFP were tagged with an extra bright GFP variant GFP+::His (Scholz et al., 2000) which was amplified from pLC1317, a gift from R. Rachubinsky. Promoter switches and N-terminal tagging were achieved using homologous recombination using a toolbox cassette system (Janke et al., 2004). The desired modification was amplified and transformed.  All plasmids were acquired from commercial sources.  57   Table 3.2: Strains used in negative correlations study (Chapter 3) Strain ID Genotype Source or Reference MDY989 MATa leu2Δ0 met15Δ0 ura3Δ0 vid30Δ::KAN VAF1-GFP+::HIS  This Study MDY1001 MATa leu2Δ0 met15Δ0 ura3Δ0 VAF1-GFP+::HIS This Study MDY1003 MATa leu2Δ0 met15Δ0 ura3Δ0 pep4Δ::KAN VAF1-GFP+::HIS This Study MDY1036 MATa leu2Δ0 met15Δ0 ura3Δ0 rpn4Δ::HPH VAF1-GFP+::HIS This Study MDY1021 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 NAT:: ADHpr-HA-VAF1 This Study MDY1022 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 vid30Δ::KAN NAT::ADHpr-HA-VAF1 This Study MDY1008 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 vid24Δ::KAN ADHpr-HA-VAF1 This Study MDY1028 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 NAT::ADHpr-HA-VID24 This Study MDY1007 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 vid30Δ::KAN NAT::ADHpr-HA-VID24 This Study MDY1027 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 vaf1Δ::KAN NAT::ADHpr-HA-VID24 This Study MDY1065 MATa leu2Δ0 met15Δ0 ura3Δ0 VID30-GFP+::HIS NAT::ADHpr-HA-VAF1 This Study MDY1045 MATa leu2Δ0 met15Δ0 ura3Δ0 VID30-GFP+::HIS NAT::ADHpr-HA-VID24 This Study MDY1068 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 FBP1-HA::NAT This Study LDY149 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 FBP1-HA::NAT vaf1Δ::KAN This Study LDY150 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 FBP1-HA::NAT vid30Δ::KAN This Study  Table 3.3: Plasmids used in the negative correlations study (Chapter 3) Plasmid ID Description Source or Reference pRS416 pURA3(CEN) ATCC#87521 pVAF1 (CEN) pVAF1::URA (CEN) MoBY ORF collection (Ho et al., 2009) pVID30 (CEN) pVID30::URA3 (CEN) MoBY ORF collection (Ho et al., 2009)  3.5.3 Co-immunoprecipitation and Western Blot: Cells were grown in glucose-containing media to mid-log phase before further steps were done unless otherwise noted. In Figure 3.5A, cells were grown to mid-log phase in media containing 2% ethanol instead of glucose. Western lysates were prepared from log phase cells by bead bashing, freezing and resuspending in Thorner buffer (8M Urea, 5% SDS, 50mM Tris pH 6.8, 0.4 mg/mL bromophenol blue and 1% β-mercaptoethanol). Lysates were heated to 95°C and the equivalent of 0.1 to 0.45 OD600 of cells was loaded onto 10% SDS-PAGE gels.  For starvation pulse-chase experiments, cells were grown to log phase in media containing glucose, then pelleted and washed with water to remove any trace of glucose. A 5 OD600 pellet was saved as the “fed” sample. The remaining culture was resuspended in media 58  containing 2% ethanol as the carbon source to ~0.3 OD600/mL and incubated at 30°C for 16-18 hours. A 5 OD600 pellet was removed as the “starved” sample or time 0. The remaining culture was spun down and then resuspended in glucose-containing medium. A 5 OD600 pellet was spun down at the desired time increments. Lysates were prepared by glass bead lysis and then subjected to SDS-PAGE and western blot analysis.  Co-IP conditions were adapted from Menssen et al., 2012. Fresh spheroplasts (20 OD600) were prepared by digesting cell walls with zymolyase (MJS BioLynx SK1204911) and frozen at -80°C before use. The frozen spheroplasts were mixed with 1mL lysis buffer (150mM NaCl; 50 mM Tris/HCl, pH 7.5; 50 mM NaF; 5 mM EDTA; 0.1% Triton X-100) + 1X protease inhibitor cocktail (Pierce, #78425) + 2mM PMSF and allowed to lyse for 10 min. The lysate was cleared by centrifugation at 13,000 RPM at 4°C for 5 min. An aliquot was saved for “lysate” lane. The lysate was incubated with 5µL Rabbit anti-GFP antibody (Life Technologies, #A6455) for followed by Protein A sepharose beads (GE Healthcare 17-5280-04). The beads were washed and proteins were eluted by heating at 95C for 5 minutes. Samples were loaded on 10% SDS-PAGE gels.  For both Westerns and co-immunoprecipitations, proteins were transferred overnight to nitrocellulose membranes and blotted with either mouse anti-GFP (Roche 11814460001) or mouse anti-HA (Covance MMS-101R) followed by goat anti-mouse conjugated to horseradish peroxidase (Jackson ImmunoResearch 115035146). The blot was developed with the enhanced chemiluminescent reagent West Pico (Pierce 34077) and exposed to Amersham Hyperfilm (GE Healthcare 28906839). 59  Chapter 4: Endosomal Regulators of Yeast Neo1 Identified Through Quantitative Enzymatic and Imaging Analysis  4.1 Synopsis Neo1 is the least well-studied member of the P4-ATPase family of putative phospholipid flippases, which regulate lipid membrane asymmetry.  A few roles for Neo1 have been identified including ER-Golgi and endosome to Golgi trafficking. To examine the role of Neo1 at endosomes, we used a two-stage process that first identified mutations that caused intracellular accumulation of a Snc1-based reporter followed by a high-content microscopy screen to identify specific regulators of Neo1.  We identified three groups of mutants that affected localization of known Neo1 binding partners Dop1 and Mon2, slowed the progression of Neo1 through endosomes, or caused Neo1 to go to the vacuole. We were able to show that Arl1 is needed for complex stability between Dop1 and Mon2, that Vps13 has a novel role in Neo1 sorting, and that the cargo-selective sorting nexin Snx3 is required for correct localization of Neo1.  A short motif in the N-terminus was necessary for Snx3 dependent sorting and full Neo1 function.  Moreover, the recognition of Neo1 by Snx3 was required for the correct sorting of other Snx3 cargos, suggesting that incorporation of Neo1 into recycling vesicles may influence their formation.   4.2 Introduction Vesicle trafficking is a highly conserved process that requires intricate coordination of cargo binding, lipid dynamics, and coat formation.  Phospholipid flippases are a family of P4-type ATPases that translocate phospholipids from the luminal to the cytosolic leaflet of the bilayer. This is believed to stimulate vesicle formation by promoting positive curvature, acting as a recruiting platform for transport machinery, and/or changing membrane electrostatics (Graham, 2004; van Meer et al., 2008; Muthusamy et al., 2009; Sebastian et al., 2012; Takeda et al., 2014). The yeast Saccharomyces cerevisiae has five putative flippases: Dnf1, Dnf2, Dnf3, Drs2, and Neo1 (Catty et al., 1997; Halleck et al., 1998; Hua et al., 2002). Neo1 is the only essential member of the family (Prezant et al., 1996; Takar et al., 60  2016). However, the combined deletion of Dnf1, Dnf2, Dnf3, and Drs2 is also lethal, emphasizing the importance of flippases in cellular function (Hua et al., 2002).  Despite their partially redundant functions, flippases may have preferences for different lipid species (Liu et al., 2007; Jacquot et al., 2012; Zhou et al., 2013; Roelants et al., 2015). While Drs2 activity increases the localized concentration of phosphatidylserine (Liu et al., 2008; Tsai et al., 2013; Xu et al., 2013; Hsu et al., 2014), Dnf1 preferentially flips phosphatidylethanolamine and its lysophospholipid derivative (Baldridge et al., 2013).  But this preference can be modulated by a point mutation to accept phosphatidylserine instead and complements cold sensitive defects caused by drs2Δ mutants (Baldridge and Graham, 2011). Temperature sensitive mutants of Neo1 have been shown to cause disruptions of PE and PS asymmetry, suggesting these might be primary lipids flipped by Neo1 (Takar et al., 2016). However, the activity and substrate preference of Neo1 and other members of the P4 ATPase family are requires further characterization.  Flippases also carry out distinct roles in different compartments, where they bind specific regulators that modulate their activity. While Dnf1 and Dnf2 are on the plasma membrane, Drs2 and Dnf3 reside in internal structures (Hua et al., 2002). The localization of Drs2 is regulated by AP-1 and Rcy1 which are needed for its trafficking to and from endosomes respectively (Liu et al., 2008; Hanamatsu et al., 2014). The C-terminus of Drs2 is thought to be autoinhibitory (Zhou et al., 2013).  At the Golgi, binding of the Drs2 C-terminus of to PI4P and the ArfGEF, Gea2, relieves this autoinhibition and promotes lipid translocation. In addition, Drs2 is thought to be required for AP-1 dependent trafficking, as deletion of Drs2 has been shown to have a decrease in clathrin coated vesicles and Drs2 mutants affect the trafficking of Snc1, an AP-1 cargo (Chen et al., 1999; Liu et al., 2008). Strikingly, Drs2 activity was also required for its own transport in the AP-1 pathway as disruption of Drs2 using a temperature-sensitive allele caused similar trafficking defects as mutation of AP-1 subunits (Liu et al., 2008). Thus, Drs2 is not only trafficked by AP-1, but its activity has been shown to be important for the formation of AP-1 vesicles themselves (Liu et al., 2008).  Less is known about Neo1 trafficking and its contributions to vesicle trafficking.  However, it has been shown that the Neo1 binding partners Dop1 and Mon2 are required for 61  correct trafficking of the v-SNARE, Snc1, similar to Neo1 (Gillingham et al., 2006; Yamagami et al., 2015), suggesting these proteins regulate Neo1 activity or trafficking. Neo1, Dop1 and Mon2 form a trimeric complex that colocalizes with Golgi and endosome markers at steady state (Wicky et al., 2004; Efe et al., 2005; Barbosa et al., 2010).  In addition, Mon2 binds Arl1 and Gga2, factors needed for Golgi to endosome trafficking suggesting that Neo1 participates in the formation of vesicles for this pathway (Jochum et al., 2002; Singer-Krüger et al., 2008). However, the role of Mon2 and Dop1 in Neo1 flippase activity and localization is not known. While the role of Drs2 in post-Golgi transport is relatively well-studied, the endosomal requirement for Neo1 is less well understood. Neo1 has two mammalian homologs ATP9A/B which have been shown to localize to the Golgi and endosomes respectively and require sequences in the N-terminal tail for this localization (Takatsu et al., 2011).  This suggests that Neo1 may have a role in endosomal trafficking, however this has not been studied.  To systematically identify Neo1 regulators at endosomes, we first used a genome-wide screen to find mutants that accumulate intracellular Snc1, and subsequently examined the localization of Neo1 in these mutants using a high-content microscopy screen. This two-stage process identified several factors that regulate Neo1 recycling from endosomes.  Specifically, we found that the sorting nexin Snx3 is needed for recycling of Neo1 from endosomes and that disrupting Snx3 sorting of Neo1 causes defects in trafficking of other Snx3-dependent proteins, providing further evidence that localization and activity of flippases are linked.  4.3 Results 4.3.1 A Quantitative Snc1-based Assay Identifies Potential Regulators of Neo1 Mutations in or depletion of Neo1 causes defects in COPI retrograde transport of proteins from the Golgi to the ER (Hua and Graham, 2003) and impairs the endosome to Golgi trafficking of Snc1, causing it to accumulate intracellularly (Mioka et al., 2014; Yamagami et al., 2015). Neo1’s role at early endosomes, as opposed to its ER to Golgi trafficking role, is not expected to be essential for viability (Takeda et al., 2014).  Thus, to 62  find genes that regulate the post-Golgi trafficking of Neo1, we first used a genome-wide screen to identify the set of viable mutants that cause internal accumulation of a Snc1-based reporter. Levels of the GFP-Snc1-Suc2 (GSS) chimera can be quantified by measuring the activity of invertase (Suc2) exposed at the cell surface using an agar plate-based assay (Burston et al., 2008; Dalton et al., 2014).  The set of mutants with higher cell surface reporter levels were previously used to uncover endocytosis regulators (Burston et al., 2009). Here, we focused on the subset of mutants with reduced cell surface levels of the Snc1chimeric reporter (Figure 4.1A and Table A.1).  We identified 275 top-scoring mutants using CLIK (Cutoff Linked to Interaction Knowledge) analysis (Dittmar et al., 2013), which sets a threshold based on the enrichment of genetic and physical interactions. Several of these mutants were known Snc1 regulators, including the sorting nexin, Snx4, epsins and components of the tethering complex GARP (Golgi Associated Retrograde Protein) (Conibear et al., 2003; Hettema et al., 2003; Zimmermann et al., 2010). The Neo1 binding partner, Mon2 was expected to be in the dataset but fell below the CLIK threshold.  Neither Neo1 nor its other binding partner, Dop1, was present in the Snc1 screen as both are essential. However, our screen identified other flippase and flippase regulators such as Dnf3, and the flippase non-catalytic subunits Cdc50 and Crf1 (Hua et al., 2002; Saito et al., 2004). A variety of other proteins including lipid regulators, kinases and phosphatases, and vesicle biogenesis proteins that were not previously associated with Snc1 transport were also identified (Figure 4.1B). Other hits, which corresponded to known regulators of DNA/RNA synthesis or mitochondrial function, may have indirect effects on Snc1 or Neo1 trafficking and were not pursued further, leaving 128 candidate genes (Table A.2). 63   64  Figure 4.1: Using a genome-wide invertase assay in conjunction with high content microscopy to identify potential regulators of Neo1. A) The chimera GFP-Snc1-Suc2 reaches the plasma membrane, is internalized, and recycled back to the Golgi. A reduction in surface levels of the chimera results in a higher (brighter) densitometry value as the reporter is stuck in internal structure such as the early endosome (EE). B) The Z score for the densitometry was computed and graphed. C) Genes in the CLIK group were hand annotated based on descriptions from the Saccharomyces Genome Database (www.yeastgenome.org). Annotations with their corresponding genes are listed in Table A.2. The proportional representation of the categories was graphed.  Sub-groups in the trafficking category are listed in the callout. D) Schematic showing the set-up for high content microscopy and follow-up analysis.  We next used high-throughput microscopy and quantitative image analysis to identify which of these 128 mutants altered Neo1 trafficking or interactions with Neo1’s binding partners Dop1 or Mon2 (Figure 4.1C). Neo1 was tagged with GFP at its N-terminus, which was previously determined not to alter its function (Wicky et al., 2004), and expressed from the ADH promoter at the chromosomal locus, which provided a bright signal without noticeably affecting localization. Synthetic genetic array (SGA; (Tong and Boone, 2006)) techniques were used to generate a set of haploid mutants co-expressing ADHpr-GFP-Neo1 and the Golgi marker Sec7-dsRed (McDonold and Fromme, 2014).  Parallel mutant arrays were constructed to compare the localization of ADHpr-RFP-Neo1 with Dop1-GFP or Mon2-GFP. High throughput imaging of haploid arrays containing each reporter combination was performed twice to reduce anomalies and ensure full coverage of the hit list. Mutants that disrupt Neo1 transport at endosomes are predicted to cause Neo1 depletion from Golgi compartments and accumulation in the vacuole or endosomal structures.  To detect these changes, we created in-house automated image quantitation pipelines that determined the degree of Neo1 colocalization with cellular markers in an unbiased manner. The overlap of Neo1/Dop1 or Neo1/Mon2 puncta was compared to find mutants that separated Neo1 from its known binding partners. To detect loss of Neo1 from Golgi structures, we measured changes in colocalization of Neo1 and Sec7 puncta. An increased brightness and size of Neo1 puncta was used to identify mutants with slowed Neo1 progression through endosomes. For all puncta measurements, values were normalized to the mean (t-statistic). This generated three distinct classes of mutants (Supplementary Table A.3 and Figure 4.1C) that were investigated in detail, as described below.  65  4.3.2 Arl1 Promotes Stability of Dop1 in Complex and on Membranes The first class of mutants altered the association of the Neo1 binding partners, Mon2 and/or Dop1 with Neo1 (Figure 4.2A). Mutation of the tethering complexes GARP (Golgi Associated Retrograde Proteins) and COG (Conserved Oligomeric Complex) subunits caused a strong reduction in the colocalization of Neo1 with both Dop1 and Mon2. This appeared to be a reduction in Neo1 puncta, while Dop1 and Mon2 remained punctate and is discussed and shown in later sections. Interestingly, a subset of mutants showed a specific decrease in the colocalization of Neo1 with Dop1 while Neo1 still colocalized with Mon2 (Figure 4.2A). This subset was enriched for components of the Arl pathway which aids in anterograde vesicle trafficking at the Golgi. The differential effect of Arl pathway mutants on the colocalization of Neo1 and Dop1 puncta was reimaged, analyzed and found to be significant, with p-values ranging from 0.0128 (mak10∆) to <0.0001 (arl1Δ) (Figure 4.2B).  To determine if the reduced overlap was due to changes in Neo1 or Dop1 localization to membranes, the average number of Dop1, Mon2, and Neo1 puncta/cell were analyzed in these mutants. Loss of the GARP tethering subunit, Vps51, reduced Neo1 puncta without significantly affecting the number of Dop1 or Mon2 puncta. This is consistent with an observed missorting of Neo1 in vps51Δ mutants, while Mon2 and Dop1 recruitment to membranes was unaffected (Figure 4.2C, D). This suggests that GARP regulates Neo1 recycling at endosomes, as described for other flippases (Takagi et al., 2012), and that Dop1 and Mon2 localization is not dependent on the localization of Neo1.  The number of Dop1 puncta/cell was significantly reduced in an arl1Δ, arl3Δ, and sys1Δ mutants (Figure 4.2C). These mutants also exhibited a reduced number of Neo1 puncta/cell (Figure 4.2C). While some Neo1 appeared to be missorted to the vacuolar membrane in these mutants, a substantial pool of Neo1 was still present in puncta that colocalized with Mon2 (Figure 4.2A and B). The observation that defects in the Arl pathway reduce Dop1 recruitment was unexpected. However, a significant alteration in Dop1 localization was only observed when arlΔ mutants expressed a tagged form of Neo1 (data not shown), suggesting that Dop1, but not Mon2 membrane recruitment is sensitive to combined defects in Neo1 and Arl1.  66    Figure 4.2: Arl1 promotes stability of Dop1 in complex and on membranes. A) Mon2/Neo1 and Dop1/Neo1 colocalization t-statistic values were plotted revealing mutants which have reduced Dop1/Neo1 but not Mon2/Neo1 colocalization. Colored points were highlighted for t-statistic values below 1.5 for Dop1/Neo1 colocalization. B/C) Live-cell fluorescence microscopy was repeated using SGA-derived strains containing Adh-RFP-Neo1 and either Dop1-GFP or Mon2-GFP for mutants in the Arl pathway, vps51Δ, and wild-type. B) Colocalization of Mon2-GFP with Adh-RFP-Neo1 and Dop1-GFP with Adh-RFP-Neo1 C) Analysis of Adh-GFP-Neo1, Dop1-GFP, and Mon2-GFP puncta/cell.  Values were graphed and subjected to One-Way ANOVA with a posthoc test using a Dunnett correction:  *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.  Error bars report standard error of the mean. n=3 D) arl1Δ mutant results in a decrease in co-immunoprecipitation of Mon2-HA and Dop1-GFP pulldown: 1.3X IP to lysate for both GFP and HA. The graph depicts the densitometry ratio of Mon2-HA IP compared to Dop1-GFP.  Densitometry values were normalized and WT and expressed as a percent, where WT equals 100%. n=3. Unpaired t-test was performed, ***p<0.001 E) Model diagram showing that when ARL1 is absent, Dop1 recruitment and binding to Mon2 is reduced, resulting in a decreased colocalization with Neo1.  Mon2 is needed to recruit Dop1 to the Golgi (Gillingham et al., 2006; Barbosa et al., 2010). Because loss of Arl1 selectively affected Dop1, we hypothesized that Arl1 contributes to the stable association of Dop1 and Mon2.  We found the amount of Dop1 that co-precipitated with Mon2 was reduced by ~30% in an arl1Δ mutant (Figure 4.2D), which 67  supports the idea that loss of Arl1 impairs Mon2-Dop1 binding (Figure 4.2E). Based on these observations, we hypothesize that Mon2 acts as a recruiting platform to stabilize the interaction of Arl1 and Dop1 at the membrane. By uncovering a role for Arl1 in Dop1-Mon2 complex formation, these results show that our systematic imaging analysis can detect subtle changes in localization that identified a rare disassociation between Mon2 and Dop1.   4.3.3 Vps13 and PI3K Regulate Neo1 Transit through Endosomes Defects in trafficking at endosomes can cause proteins to accumulate in endosomal compartments. A second set of mutants in our microscopy screen were those that caused Neo1 to accumulate in large punctate structures. In wild-type cells, Neo1 is largely present in small puncta (0.6-1.5µM diameter) that colocalize with the Golgi marker Sec7 and is less frequently found in large puncta (1.7-2.5µM diameter). To identify mutants that cause Neo1 to accumulate in endosomes, we defined an initial set of 10 mutants with a reduced number of total Neo1 puncta/cell (t-statistic < -0.33) and an increased number of large Neo1 puncta (t-statistic ≥ 0.25) (Table A.3 and Figure 4.3A). Re-imaging and analysis of these mutants identified 3 mutants with a statistically significant increase in the number of large Neo1 puncta/cell relative to a wild-type strain: vps13Δ, vps30Δ, and vps38Δ (Figure 4.3B).  Large puncta were similarly observed when a plasmid expressing GFP-Neo1 from its native promoter was introduced into vps13Δ and vps30Δ mutants, indicating the phenotype was not due to Neo1 overexpression.  The large punctate structures often co-labeled with FM4-64 (Figure 4.3C). To determine if these large puncta are aberrant endosomes, we compared the overlap of Neo1 puncta with Sec7 and the early and late endosome markers Snx4 and Snx3 (Strochlic et al., 2007). In vps30Δ and vps13Δ mutants, Neo1 showed a significantly reduced colocalization with the Golgi marker, Sec7 (p<0.0001) that was more pronounced for large puncta (p< 0.015), indicating these structures are not enlarged Golgi compartments (Figure 4.3D).  Instead, we observed a striking colocalization of the large Neo1 puncta with Snx3 and Snx4 (Figure 4.3E/F). Together, these data suggest Vps13 and Vps30/38 are important regulators of Neo1 trafficking through endosomes.  68   Figure 4.3: Vps30/38 (PI3K) and Vps13 regulate Neo1 transit through the endosomes. A) Total Neo1 puncta and Large Neo1 puncta per cell were identified from 2 sets of high content microscopy images of strains containing ADHpr-GFP-Neo1 + Sec7-dsRed, Mon2-GFP + ADHpr-RFP-Neo1 or Dop1-GFP + ADHpr-RFP-Neo1. Values are reported as t-statistics and averaged. Large Neo1 puncta were plotted against total Neo1 69  puncta/cell.  Labeled points were subject to further imaging and analysis.  Colored points were significant when after subsequent imaging and analysis. B) Microscopy of ADHpr-GFP-Neo1 with Sec7-dsRed was repeated and Large Neo1 puncta/cell was analyzed.  One-Way ANOVA with a posthoc test with Dunnett correction:  *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Error bars report standard error of the mean. n=3 C) Live-cell fluorescence microscopy of pGFP-Neo1 (pLD27) and the endosomal dye FM4-64 in BY4741 compared in wild-type, vps30Δ, and vps13Δ mutants.  D/E) Colocalization of Neo1 and large Neo1 puncta with D) Sec7-dsRed using integrated ADHpr-GFP-Neo1 E)Snx3-Ruby F) Snx4-Ruby with pADHpr-GFP-Neo1 (pMD124) for  wild-type compared to vps30Δ and vps13Δ E/F left) Live-cell fluorescence microscopy representative images. D/E right/F right ) Quantification. n=3 One-Way ANOVA with a posthoc test with Dunnett correction:  *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Error bars report standard error of the mean.  Scale bar represents 2µM.  4.3.4 Vps13 Is Required for Sorting at Early and Late Endosomes Vps30/38 are two subunits of the Type I Vps34 PI 3-kinase complex important for maintaining PI3P at endosomes (Kihara et al., 2001) suggesting that its role in Neo1 accumulation in endosomes is through its modulation of endosomal progression.  However, the function of Vps13 is still largely unknown and requires further investigation to understand how it contributes to Neo1 progression through endosomes. Vps13 is a large ~350kD protein with proposed roles in membrane contact site formation, and late endosome recycling (Robinson et al., 1988; Rothman et al., 1989; Brickner and Fuller, 1997; Park and Neiman, 2012; Park et al., 2013; Lang et al., 2015). Thus, it is surprising that we show that Vps13 has a defect in sorting of our Snc1 reporter GSS as this indicates an early endosome recycling role. We confirmed that a vps13Δ had significantly reduced levels of the GSS reporter (p<0.0001) at the cell surface, similar to Gcs1, a known regulator of Snc1 trafficking (Xu et al., 2013) (Figure 4.4A). This suggests that vps13Δ mutants have altered sorting at both early and late endosomes and we have expanded the known cargos sorted by Vps13. Recent work suggests Vps13 has a role in the function or maintenance of membrane contact sites, including the Nuclear-Vacuolar junction (NVJ), Vacuolar-Mitochondrial junction (VClamp) (Lang et al., 2015) and Endosome-Mitochondrial junctions (Park et al., 2016). The GFP-Snc1-Suc1 and Neo1 trafficking defects of vps13Δ mutants could be an indirect effect of loss of contact sites as opposed to a direct cargo sorting role. However, we did not find that loss of known contact site proteins caused reduced cell surface levels of the Snc1 reporter in our genome-wide screen (Table A.1, A.2).  We also determined that mutation of genes important for the formation of different membrane contact sites, including 70  nvj1∆ (NVJ), mdm12∆ (ERMES), and vps39∆ (v-CLAMP), did not cause the accumulation of Neo1 in large puncta (Figure 4.4B).   Figure 4.4: Vps13 is a novel regulator of Neo1, which aids trafficking through the endosomal pathway. A/C) Vps13 is a novel regulator of the Snc1-based reporter trafficking. Cell surface levels of the GFP-Snc1-Suc2 reporter were quantified using a plate based invertase overlay assay in yeast deletion strains or C) LGY1 plus complementing plasmids top) representative images displayed. Bottom: Quantitation of densitometry using Cell Profiler Analysis to identify the mean intensity of each spot.  Each group of four was average and treated as one sample. n=3 B/D) Live-cell fluorescence microscopy of B) pADHpr-GFP-Neo1 and D) ADHpr-GFP-Neo1 in B) membrane contact site mutants (n=6 for nvj1Δ vps39Δ, and mdm12Δ; n=8 for vps13Δ and wild-type) and D) vps13Δ with complementing plasmids (n=3) i) representative images scaled the same.  Scale bar =2µM   Statistics: One-Way ANOVA with a posthoc test using a Dunnettt correction:  *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Error bars report standard error of the mean. 71   Lang et al. (2015) isolated missense mutations in Vps13 that blocked Vps13 localization to the NVJ and rescued ERMES defects, possibly by enhancing v-CLAMP function (Lang et al., 2015). However, we found wild-type Vps13, Vps13L1627S, and Vps13D716H were equally effective in rescuing the defective trafficking of the Snc1 reporter or GFP-Neo1 in the vps13∆ strain (p<0.0001) (Figure 4.4C, D).  This suggests that the role of Vps13 in endosomal sorting is independent of its role at membrane contact sites. Together, this supports a new role for Vps13 in the endosomal recycling of Neo1 and the Snc1-based reporter GSS.   4.3.5 Identification of Factors that Regulate Neo1 Recycling from Endosomes  To find mutations that result in Neo1 recycling defects, we first selected a set of deletion strains with reduced localization of Neo1 at the Golgi, as defined by a decrease in the total number of Neo1 puncta/cell (t-stat < -0.33) and reduced colocalization with Sec7 (t-stat < -0.67) (Figure 4.5A). This list was further refined by eliminating mutants that accumulate Neo1 in endosomes (large puncta; t-statistic < 1). Interestingly from this analysis we also noticed that some mutants increased the colocalization with Sec7 suggesting that they could be anterograde regulators of Neo1 but as this was not the focus of this study, they were not pursued further (Figure A.1).  72   Figure 4.5: Known regulators as well as novel potential regulators cause Neo1 to missort to the vacuole. A) Colocalization of Adh-GFP-Neo1 with Sec7-dsRed as well as Adh-GFP-Neo1 puncta/cell were analyzed for two sets of high content images and student’s t-statistic values were graphed.  Mutants represented in the colored points were subjected to additional microscopy for small-scale validation.  B/C) Live-cell fluorescent microscopy of ADHpr-GFP-Neo1 with Sec7-dsRed was repeated and B) colocalization C) mean puncta intensity was analyzed.  One-Way ANOVA with a posthoc test with Dunnett correction:  *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Error bars report standard error of the mean n=3. D) Fluorescence microscopy of pGFP-Neo1 in wild-type compared to COG, GARP, Arl, and snx3Δ mutants with the endosomal dye FM4-64.  Scale bar represents 2µM. 73  This set of 22 mutants with reduced colocalization with Sec7 and reduced Neo1 puncta/cell was re-imaged and re-analyzed, identifying 16 with significantly reduced colocalization between Neo1 and Sec7 relative to a wild-type strain (Figure 4.5B). A subset of the 16 mutants also showed a significant reduction in the intensity of Neo1 puncta (Figure 4.5C), which suggests they have more severe missorting defects. These included proteins with known or suspected roles in endosome-to-Golgi trafficking, such as the COG and GARP tethering complexes (Conibear and Stevens, 2000; Suvorova et al., 2002; Fotso et al., 2005), the synaptojanin-like protein Inp53 (Ha et al., 2003), the v-SNARE Gos1 (McNew et al., 1998) and components of the GET complex, which promotes the membrane insertion of SNAREs and other tail-anchored proteins (Mateja et al., 2015). Mutants in the Arl pathway also showed a partial decrease in Sec7 colocalization with Neo1. We confirmed that Neo1 expressed from its native promoter is mislocalized to FM464-labeled vacuolar membranes in a subset of these mutants and in other mutants merely appears diffuse meaning it could be degraded or impeded in stranded vesicles (Figure 4.5C, D) In mutants lacking the sorting nexin Snx3, Neo1 is found in vacuolar rings as well as some remaining puncta suggesting a partial defect in the retrograde trafficking of Neo1 (Figure 4.5C). Taken together, these mutants describe a recycling pathway involving SNAREs, SNARE regulators, and tethering complexes that aid in vesicle docking and fusion with the Golgi. Snx3 was the only cargo-selective protein identified by this analysis, suggesting it might be responsible for recruiting Neo1 into retrograde carriers at endosomes.  4.3.6 Snx3-dependent Sorting of Neo1 Requires a Short Amino Acid Motif in the N-Terminal Tail Relatively few proteins (Ste13, Kex2, Pep12, and Ftr1) have been shown to require the cargo selective sorting nexin Snx3 for recycling from endosomes to the Golgi (Voos and Stevens, 1998; Hettema et al., 2003; Strochlic et al., 2007).  If Snx3 is responsible for sorting Neo1 into retrograde tubules, Neo1 may contain a cytosolic sorting signal that is recognized by Snx3. Members of the P4-ATPase family have a conserved structure, containing nucleotide-binding, actuator, phosphorylation, and membrane domains as well as regulatory N- and C-terminal cytosolic tails (Catty et al., 1997; Andersen et al., 2016).  For example, 74  the related flippase Drs2 interacts with AP-1 vesicle coat proteins through its C-terminal tail (Chantalat et al., 2004) and the N-terminus is needed for endocytosis of Dnf1 and Drs2 (Liu et al., 2007). Neo1 has a 195aa N-terminal cytosolic tail with several conserved motifs, and a much shorter C-terminal tail (21aa). Additionally, the N-terminal tails of the Neo1 homologs ATP9A and ATP9B are sufficient for their localization (Takatsu et al., 2011).  Thus the N-terminal tail of Neo1 is a candidate for a Snx3 sorting motif.  To determine if the N-terminal tail of Neo1 contains signals that confer Snx3-dependent sorting, we fused amino acids 1-195 of Neo1 to the transmembrane and luminal portions of alkaline phosphatase (ALP), creating a chimeric GFP-Neo1-ALP protein expressed under a constitutive ADH promoter (Figure 4.6A and Figure A.2). We found GFP-Neo1-ALP localization was similar to that of GFP-Neo1 in wild-type and snx3∆ cells, and like GFP-Neo1 exhibited a significantly reduced colocalization with Sec7 in snx3∆, arl1∆, vps51∆ (GARP) and cog7∆ mutants (Figure 4.6B). This suggests the Neo1 N-terminal tail contains sorting signals that specify its ability to be recycled from endosomes. In addition, GFP-Neo1-ALP largely localized to the vacuole rim in snx3Δ mutants (Figure 4.6B) suggesting that a Snx3 sorting signal is encoded in the first 195 amino acids of Neo1. 75   Figure 4.6: The N-terminal tail of Neo1 is sufficient for localization. A) Schematic of the Neo1-ALP chimera.  The first 195 amino acids of Neo1 are fused to the transmembrane and luminal portions of ALP.  B) Live-cell fluorescence microscopy of pAdh-GFP-Neo1 (pMD124) and pADHpr-GFP-Neo1-ALP (pMD192) in BY4741 compared to mutants left) representative images scaled the same.  Scale bar =2µM right) Quantitation of the colocalization between Neo1 and Neo1-ALP with Sec7-dsRed was analyzed. n=3 C) Live-cell fluorescence microscopy of BY4741 and snx3Δ with different Neo1-ALP chimeras. Left) Schematic representation of each chimera right) Quantitation of >200 cell measuring cells containing vacuolar rim staining. n=1 D) Live-cell fluorescence microscopy of BY4741 and snx3Δ with pADHpr-GFP-Neo1 +/- a mutation changing F65EM>AAA, P94LM>AAA, or an N-terminal truncation removing the first 100 amino acids (Δ100). Left) Representative images scaled the same.  Graph: Quantitation of >200 cells/trial and assessed 76  the  number of cells containing vacuolar rim staining. n=3 Statistics: One-Way ANOVA with a posthoc test using Dunnett correction:  *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.  Error bars report standard error of the mean.  To find the Snx3 signal, fragments of the Neo1 N-terminal tail were fused to the ALP transmembrane and luminal domains, and the number of cells exhibiting vacuolar rim localization was compared in wild-type and snx3∆ cells (Figure 4.6C and Figure A.3A).  Deletion of 54 or 80 membrane-proximal residues had no effect on Snx3-dependent sorting, nor did the removal of the N-terminal 50 residues (Figure 4.6C). However, further deletions caused localization to the vacuole rim in wild type cells, suggesting the Snx3 sorting signal lies between amino acid residues 50 and 104 (Figure 4.6C and Figure A.3B). This region contains two short stretches of conservation, surrounded by poorly conserved sequences (Figure A.2). We confirmed that removal of the first 100aa of Neo1 caused mislocalization to the vacuole (Figure 4.6D).  We then mutated conserved residues in each candidate motif to alanines in full length GFP-Neo1. We observed a significant relocalization of Neo1-F65EM>AAA to the vacuolar rim, whereas mutation of Neo1-P94LM>AAA had no effect (Figure 4.6D).  Thus, the conserved FEM residues may be part of a Snx3 sorting signal.  The Neo1-F65EM>AAA mutation caused significant mislocalization of the 1-140 chimera to the vacuolar rim. Surprisingly, when this F65EM motif was mutated in the 1-195 chimera, there was little increase in vacuolar rim localization (Figure A.3C). We also found that the GFP-Neo1-ALP chimera but not GFP-Neo1 lacking the first 101aa was competent for Snx3 sorting (Figure A.3C). These results suggest that a membrane-proximal portion of the N-terminal tail contains a second Snx3 sorting signal that is masked in the context of the native Neo1 protein. Taken together, these results show that the conserved FEM motif, which is similar to the acidic-hydrophobic sequence of amino acids found in other Snx3 signals (Nothwehr et al., 1993), is required for the Snx3-dependent sorting of Neo1.   4.3.7 Neo1 Sorting by Snx3 is Functionally Important  The loss of Snx3 causes a partial mislocalization of Neo1 to the vacuole rim. To determine if sorting by Snx3 affects viability, Neo1-F65EM>AAA and Neo1-P94LM>AAA plasmids were introduced into a strain expressing NEO1 from the GAL1 promoter.  When grown on dextrose to repress GAL1-NEO1 expression, the wild-type and mutant plasmids 77  could support cell growth and were not temperature sensitive, but strains with the empty plasmid were inviable (Figure A.4). Thus, these sequences are not required for the essential function of Neo1 and do not appear to destabilize Neo1. However, strains expressing Neo1-F65EM>AAA, but not wild-type Neo1 or the Neo1-P94LM>AAA mutant, exhibited sensitivity to neomycin and trifluoperazine, drugs that cause reduced growth in a heterozygous Neo1 mutant (Figure 4.7C) (Hoepfner et al., 2014). The Neo1-F65EM>AAA and snx3 mutants showed a similar pattern of drug sensitivity, suggesting that these snx3∆ mutant phenotypes result from the mislocalization and perturbation of Neo1 function.  Drs2 sorting into vesicles at the Golgi and endosomes is believed to contribute to the formation of these vesicles by promoting curvature formation (Wiederkehr et al., 2000; Furuta et al., 2007; Chen et al., 2011; Hanamatsu et al., 2014). To test if Neo1 inclusion in Snx3 vesicles is important for the trafficking of other Snx3 cargos, we examined the sorting of the well-characterized Snx3 cargo protein A-ALP (Nothwehr et al., 1993).  Approximately 30% of cells expressing the Neo1-F65EM>AAA mutant missorted pGFP-A-ALP to the vacuolar rim, which was a significant defect (p<0.0001) though not as severe as the ~90% missorting (p<0.0001) seen in a snx3∆ mutant (Figure 4.7B). No further increase in A-ALP missorting was observed in a snx3Δ Neo1-F65EM>AAA double mutant, suggesting that they work in the same pathway (Figure 4.7B). These results suggest that sorting of Neo1 by Snx3 into recycling tubules is important for the formation or function of these tubules (Figure 4.7C). 78   Figure 4.7: A small conserved region on the N-terminal tail of Neo1 is needed for sorting by Snx3 and for sorting of other Snx3 cargo proteins. A) Growth assay to assess drug sensitivity of Neo1 N-terminal tail truncations. Yeast were spotted in 10x dilution series on YPD + DMSO, YPD + 30µM Trifluoperazine and YPD + 4mg/mL Neomycin and grown at 30°C.  Images selected based on best representation and taken between 34 and 48 hours growth B) Live-cell fluorescence microscopy of pA-ALP (pMD130) in GALpr-Neo1 with covering pNeo1 or pNeo1 (F65EM>AAA) to test if Neo1 is needed to be in complex with Snx3 to aid in Snx3 dependent trafficking.  All images scaled the same. i) Representative images ii) Quantitation showing an increase in vacuolar rim staining in the pNeo1 (F65EM>AAA) mutant. Scale bar =2µM; Statistics: % puncta was compared using One-Way ANOVA with a posthoc test using Dunnett correction:  *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.  Error bars report standard error of the mean. C) Model diagram showing that 79  Neo1 is sorted into Snx3 containing vesicles which require the N-terminal F65EM motif on Neo1.  In addition, A-ALP is also partially missorted when the Neo1 N-terminal tail is mutated.  4.4 Discussion 4.4.1 Identification of Novel Endosome Trafficking Regulators   A genome-wide screen to uncover new regulators of Snc1 recycling at endosomes identified a large number of cargo-selective cargo adaptor proteins. We hypothesized that these adaptors may not bind Snc1, but instead may influence Snc1 recycling indirectly, by directing the sorting of one or more flippases. Snc1 sorting defects would result from the resultant loss of lipid asymmetry at endosomes. Drs2 and Neo1 are both needed for Snc1 sorting, but the regulation of Neo1 post-Golgi trafficking has not been studied. Our Snc1-based screen and high-content imaging analysis identified known and novel regulators of Neo1 with a high degree of sensitivity. This showed Snx3 (but not other vesicle adaptors) was important for sorting Neo1. In particular, this study followed up on a few of the mutants identified in the screen including Snx3, Arl1, and Vps13.  However, more Neo1 regulators remain. For example, several kinases and phosphatases were found to affect Snc1 trafficking but did not alter Neo1 localization (i.e. Casein Kinase A (CKA)). Activity of other yeast flippases, Dnf1 and Dnf2, is regulated by the kinases Fpk1 and Fpk2 (Nakano et al., 2008), whose activity in turn is regulated by upstream kinases (Roelants et al., 2010, 2015). Thus, these kinases/phosphatases identified in the screen remain as candidates that may affect activity but not localization of Neo1. Additionally, our screen may yield insights into other trafficking regulators of Neo1 outside of the endosomal system. For example, a few mutants caused Neo1 to accumulate in Golgi including two proteins that control exocytic vesicle trafficking, Ypt31 and TRAPP-II, and thus may regulate Neo1 anterograde trafficking (Tokarev et al., 2009). Thus, this dataset contains more potential flippase regulators that remain to be explored. Our screen tested for early endosome regulators however it is possible that Neo1 could be trafficked primarily via late endosomes, and affects early endosome lipid composition indirectly. ESCRT mutants have been shown to regulate the trafficking of Neo1 (Wicky et al., 2004). As well, the late endosome recycling complex, retromer is needed for Neo1 localization (Wu et al., 2016 and data not shown). Thus, it would be interesting to do a 80  future screen targeting the late endosome as a way to fully represent Neo1 endosomal regulators.   4.4.2 Extending Our Knowledge of the Known Neo1 Regulator Arl1 Known Neo1 regulators were identified in our screen, including the known Neo1 regulator, Arl1.  Previously, physical and genetic connections have linked Neo1 and Arl1 and we have been able to extend the understanding of Arl1 regulation showing that Arl1 is needed for Dop1-Mon2 complex stability as well as an independent role in Neo1 trafficking (Efe et al., 2005; Manlandro et al., 2012).  Mon2 binds to Arl1 and to GGA coat proteins which is thought to aid in anterograde vesicle trafficking from the Golgi. arl1Δ  and mon2Δ mutants reduce GGA recruitment, suggesting that Neo1 mislocalization in an arl1Δ mutant is due to lack of GGA recruitment (Jochum et al., 2002; Singer-Krüger et al., 2008). Singer Kruger (2008)  However, we did not see a large defect in Neo1 localization in a mon2Δ mutant which would be expected if GGA recruitment was required (Figure 4.2E and F) (Singer-Krüger et al., 2008).  Additionally, GGA1 and GGA2 single mutants did not show Neo1 mislocalization. Thus, while GGA could aid in Neo1 traffic, either a double mutant is required or another pathway may compensate such as recruitment of AP-1 or transit via the plasma membrane (Panic et al., 2003; Liu et al., 2008).  The independent role of Arl on Neo1 activity and localization requires further investigation. Neo1 mislocalization in mak3/10 mutants, was not as severe as sys1Δ, arl1Δ, arl3Δ mutants This suggests in mak3/10 mutants that leave some functional Arl1 on the membrane.  The defects, while small, were consistent and reproducible. Thus, our quantitative image analysis was capable of detecting small differences in localization reproducibly.   4.4.3 Endosomal Regulation of Neo1 by Vps13 We found that Neo1 accumulates in endosomal structures in cells lacking Vps13, or the PI3K subunits. However, regulation of Neo1 localization by Vps30/38 (PI3K) and Vps13 may be through differing mechanisms. PI3P is necessary for many endosomal proteins to be recruited to endosomes and deletion of Vps30 reduces PI3P by approximately 65%, which is 81  known to affect sorting nexin recruitment (Burda et al., 2002), suggesting that Neo1 accumulates in endosomes due to a defect in endosomal progression. Vps13, in contrast, is a large ~350kD protein with proposed roles in membrane contact site formation, and late endosome recycling (Robinson et al., 1988; Rothman et al., 1989; Park and Neiman, 2012; Park et al., 2013; Lang et al., 2015). We found that a vps13Δ mutant had reduced Dop1/Neo1 colocalization but not Mon2/Neo1 colocalization (Figure 4.2). In contrast, PI3K mutants show increased colocalization of both Dop1 and Mon2 with Neo1. Thus, not only are Vps13 and PI3K regulated differently, this suggests that Vps13 may be needed for Dop1 stability at endosomes (Figure 4.3). Thus, while Vps30/38 have known roles in endosomal maturation explaining the Neo1 defects, further work with Vps13 will be required to determine the role of Vps13 in the sorting of Neo1.  4.4.4 Recycling of Neo1 by Snx3 Only a few proteins (Ste13, Kex2, Pep12, and Ftr1) require both Snx3 and retromer for recycling from endosomes to the Golgi (Voos and Stevens, 1998; Hettema et al., 2003; Strochlic et al., 2007). However, other well-known retromer cargo proteins such as Vps10 do not require Snx3 (Voos and Stevens, 1998).  We found that a short conserved motif in the cytoplasmic tail of Neo1, residues F65EM, were required for Snx3 dependent sorting of Neo1.  The FEM motif fits into a broader conserved amino acid sequence in closely related yeast species (ExFEMRxV/I, Figure A.2) which is similar to signals found in Ste13 and Kex2 (Wilcox et al., 1992; Nothwehr et al., 1993).  These signals feature a consensus ΦEΦ, where Φ is a bulky hydrophobic. These sequences appear to be anchored by an upstream glutamic acid and a C-terminal hydrophobic (either isoleucine or valine).  Interestingly, Nothwehr (1993) showed that the mutation of the glutamic acid and final hydrophobic residue on its own did not cause mislocalization of A-ALP.  It appears then that Snx3 requires a somewhat conserved motif, though further work is needed to understand the context and specificity of this motif. Some of the knowledge gained in our study about Neo1 trafficking is corroborated by what is known about the mammalian homologs of Neo1, ATP9A and ATP9B. Thus, the regulators identified in our study may be directly applicable.   Specifically, we found that 82  Neo1 requires its N-terminus for localization, consistent with knowledge about ATP9A and ATP9B (Takatsu et al., 2011).  However, the FEM motif is not conserved in mammals suggesting that another signal may be needed for sorting of the mammalian proteins.  Finally, ATP9B (endosome localized) may be needed for sorting of mammalian Snx3-dependent proteins (P. Cullen, personal communication).  Thus, our study suggests that mechanistic contribution of flippases may be conserved.   Neo1 may play a role in endosomal recycling in a surprisingly similar mechanism to Drs2. Catalytic mutants of Drs2 cause defects in early endosome trafficking suggesting that flippase activity is needed to generate the curvature needed for efficient early endosome trafficking.  By extension, Neo1 activity may be needed to generate curvature required for late endosome recycling, suggesting a model where all flippases may work in a similar way to regulate vesicle trafficking.    4.5 Materials and Methods 4.5.1 Yeast Strains and Plasmids All yeast strains listed in Table 4.1 were made by homologous recombination (Longtine et al., 1998; Janke et al., 2004; Sheff and Thorn, 2004; Lee et al., 2013; Slubowski et al., 2015).  Strains not listed were made by SGA using parental strains listed in Table 1 (Tong and Boone, 2006).  Strains were propagated in rich medium (YPD: 1% yeast extract, 2% peptone, 2% dextrose) or SD minimal medium (0.17% yeast nitrogen base, 0.5% ammonium sulfate, 2% synthetic complete mix, 2% dextrose) supplemented with the appropriate amino acids.     83  Table 4.1: Strains used in the Neo1 study (Chapter 4) Strain ID Genotype Source Y7043 MATα his3Δ1 leu2Δ0 ura3Δ0 met15Δ0 LYS2+ can1Δ::STE2pr-LEU2 lyp1Δ cyh2Δ Dr. C. Boone (University of Toronto) BY4741 MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 GE Dharmacon, YSC1048 BY4742 Mat α his3∆1 leu2∆0 lys2∆0 ura3∆0 GE Dharmacon, YSC1049 MAT a deletion collection MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 geneX∆::KAN GE Dharmacon, YSC1053 MDY519 Y7043 SUC2::GFP-SNC1-SUC2::URA3 Burston et al., 2009 MDY1060 Y7043 NAT::ADHpr-GFP-NEO1 This Study BBY630 Y7043 NAT::ADHpr-GFP-NEO1, SEC7-dsRed::URA This Study MDY1145 Y7043 DOP1-GFPenvy::HIS3, NAT::ADHpr-RFP-NEO1 This Study MDY1146 Y7043 MON2-GFPenvy::HIS3, NAT::ADHpr-RFP-NEO1 This Study MDY1171 BY4741 DOP1-GFPenvy::HIS3 This Study LDY189 BY4741 MON2-GFPenvy::HIS3 This Study LDY165 BY4741 arl1∆::NAT This Study LDY188 LDY165 DOP1-GFPenvy::HIS3 This Study LDY190 LDY165 MON2-GFPenvy::HIS3 This Study vps51∆ BY4741 vps51∆::KAN MAT a deletion collection LDY187 BY4741 vps51∆::KAN, DOP1-GFPenvy::HIS3 This Study LDY190 BY4741 vps51∆::KAN, MON2-GFPenvy::HIS3 This Study MDY1118  BY4741 mon2∆::NAT, This Study LDY186 MDY1118 DOP1-GFPenvy::HIS3 This Study MDY1172 BY4741 MON2-3HA::HPH This Study MDY1181 BY4741 MON2-3HA::HPH, DOP1-GFPenvy::HIS3 This Study LDY171 MDY1181 arl1∆::NAT This Study LDY153 BY4741 vps13D::NAT This Study LDY155 BY4741 vps30D::NAT This Study BBY655 BY4741 SNX4-yomRuby2::KAN This Study LDY178 BBY655 vps13∆::NAT This Study LDY182 BBY655 vps30∆::NAT This Study MDY10 Y7043 snx4∆::NAT,  GFP-SNC1tm-SUC2::URA This Study MDY52 Y7043 vps13∆::NAT,  GFP-SNC1tm- SUC2::URA This Study MDY279 Y7043 gcs1∆::NAT,  GFP-SNC1tm- SUC2::URA This Study LGY1 BY4742 vps13∆::NAT,  GFP-SNC1tm- SUC2::LYS This Study mdm12∆ BY4741 mdm12∆::KAN MAT a deletion collection nvj1∆ BY4741 nvj1∆::KAN MAT a deletion collection vps39∆ BY4741 vps39∆::KAN MAT a deletion collection vps13Δ::NAT ADH-GFP-Neo1 MATa his3Δ1 leu2Δ0 ura3Δ0  LYS2+ can1Δ::STE2pr-LEU2 lyp1Δ cyh2Δ vps13Δ::KAN NAT::ADHpr-GFP-NEO1 (MDY1060 x BY4741 vps13Δ::KAN) This Study (SGA resultant strain) MDY1207 BY4741 SNX3-yomRuby2::KAN This Study    84  Strain ID Genotype Source LDY179 MDY1207 vps13∆::NAT This Study LDY180 MDY1207 vps30∆::NAT This Study snx3∆ BY4741 snx3∆::KAN MAT a deletion collection cog7∆ BY4741 cog7∆::KAN MAT a deletion collection MDY1178 BY4741 KAN::GAL-NEO1 This Study LDY183 MDY1178 snx3∆::NAT This Study vps9∆ BY4741 vps9∆::KAN MAT a deletion collection trs33∆ BY4741 trs33∆::KAN MAT a deletion collection ypt31∆ BY4741 ypt31∆::KAN MAT a deletion collection  Plasmids are listed in Table 4.2. All plasmids were made by homologous recombination in yeast, rescued in Escherichia coli and confirmed by sequencing. pMD124 was made by PCR amplifying NAT::ADHpr-GFP-Neo1 from genomic DNA and co-transforming with linearized pRS413. pMD195 was made by amplifying NAT::ADHpr-GFP with the first 195 amino acids of Neo1 as well as aa97-1701 from Pho8 with overlapping homology encoded on the primers.  These were co-transformed into cut pMD124.  To make pMD151, the ADH promoter was digested from pMD124 and replaced with 800bp 5’UTR of Neo1.  pMD192 was made by switching the HIS3 marker by digesting and replacing with the URA3 gene from pRS416.  pLD39 was made by removing sequences encoding GFP from pMD192 and co-transforming with a PCR product containing homology to the 5’UTR and N-terminal regions of Neo1.  For pLG1, Vps13 was amplified and co-transformed into the multiple cloning site.  Subsequent mutations were made using overlapping PCR products containing mutations encoded in the primers. These were co-transformed into cut wild-type plasmid.   Table 4.2: Plasmids used in the Neo1 study (Chapter 4) Plasmid ID Description Source or Reference pRS413 pHIS3 (CEN) ATCC#87518  pRS416 pURA3 (CEN) ATCC#77145 pMD124 pRS413-NAT:: ADHpr-GFP-NEO1 This study pLD27 pRS413-NAT::UTR-GFP-NEO1 This study pMD146 pRS413-NAT:: ADHpr-GFP-NEO1 (P94LM>AAA) This study pMD148 pRS413-NAT::UTR-GFP-NEO1 (P94LM>AAA) This study pMD151 pRS416-NAT::UTR-GFP-NEO1 This study pMD192 pRS413-NAT:: ADHpr-GFP-NEO1(1-195)-ALP This study pMD195 pRS413-NAT:: ADHpr-GFP-NEO1(50-195)-ALP This study pMD200 pRS413-NAT:: ADHpr-GFP-NEO1(1-195)-ALP (F65EM>AAA) This study pLD32 pRS413-NAT::ADHpr-yeGFP-NEO1(101-195)-ALP This study pLD33 pRS413-NAT::ADHpr-yeGFP-NEO1(132-195)-ALP This study 85  Plasmid ID Description Source or Reference pLD34 pRS413-NAT::ADHpr-yeGFP-NEO1(162-195)-ALP This study pLD36 pRS413-NAT:: ADHpr-GFP-NEO1(1-52)-ALP  This study pLD37 pRS413-NAT:: ADHpr-GFP-NEO1(1-104)-ALP  This study pLD38 pRS413-NAT:: ADHpr-GFP-NEO1(1-140)-ALP  This study pLD39 pRS416-NAT::NEO1 This study pLD41 pRS413-NAT:: ADHpr-GFP-NEO1 (50-140)-ALP This study pLD42 pRS413-NAT:: ADHpr-GFP-NEO1(101-140)-ALP This study pLD44 pRS413-NAT:: ADHpr-GFP-NEO1 (Δ100) This study pLD46 pRS413-NAT:: ADHpr-GFP-NEO1 (F65EM>AAA) This study pLD47 pRS416-NAT::UTR-GFP-NEO1 (F65EM>AAA) This study pLD48 pRS416-NAT::UTR-NEO1 (F65EM>AAA) This study pLD49 pRS416-NAT::UTR-GFP-NEO1 (Δ100) This study pLD50 pRS416-NAT::UTR-GFP-NEO1 (P94LM>AAA) This study pLD53 pRS416-NAT:: ADHpr-GFP-NEO1(1-140)-ALP (F65EM>AAA) This study pLG1 pRS416-VPS13 This study pMD204 pRS416-VPS13 (D716H) This study pMD184 pRS416-VPS13 (L1627S) This study pMD130 pRS413-A-ALP This study  4.5.2 Genome-wide Invertase Overlay Assay Raw images from the genome-wide invertase-based GSS reporter screen (Burston et al., 2009) were reanalyzed using GridGrinder (gridgrinder.sourceforge.net), and values were subjected to background subtraction, normalized to the plate median and filtered to eliminate absent or very slow-growing strains. Raw values were converted to a z-score and the final ranking was reported using the mean MATa and MATα values.  CLIK analysis (http://www.rothsteinlab.com/tools/clik) based on BIOGRID version 3.4.128 on October 1, 2015 was used to select a top hit list of 275 unique genes, which were subsequently filtered to remove genes with roles in mitochondrial, ribosome, DNA or RNA biology.    4.5.3 High-Content Imaging Screen Fluorescent reporters were introduced into 128 MAT a deletion mutants using Synthetic Genetic Array (SGA) technology (Tong and Boone, 2006).  Haploid strains were selected on synthetic complete –LEU –LYS –ARG –HIS/URA (strain dependent) +L-canavanine (c9758; Sigma Aldrich, St. Louis, MO, St. Louis, MO) +S-(2-aminoethyl)-L-cysteine (A-2636; Sigma Aldrich) + Nourseothricin-dihydrogen sulfate (5005000; Werner Bioagents, Jena, Germany).  All array manipulations were made using a pinning robot (BM5-BC-48; S&P Robotics Inc., Toronto, ON). 86  Colony arrays were pinned into 96-well plates containing 150µL synthetic complete medium + one 5mm glass bead (11-312C; Fisher Scientific) to prevent settling of yeast during incubation and grown overnight. Yeast cultures were diluted 1/10 and grown 3 hours shaking at 30°C.  The wells in a 384-well glass bottom Matriplate (MGB101-1-2-LG-2; Brooks, Chelmsford, MA) were treated with 0.1mg/mL type V Concanavalin A (C7275; Sigma Aldrich, St. Louis, MO) and allowed to dry.  The bottom of the plate was treated with SigmaCote (SL2; Sigma Aldrich) per manufacturer’s instructions.  60µL of the log-phase yeast culture was transferred to the 384-well plate and allowed to settle ~ 10 minutes. Each well was washed with 90 µL media before adding 60µL synthetic complete media.  Images were acquired at room temperature using a Leica TCS SP8 3X microscope (Leica Microsystems GmbH, Wetzlar, Germany) with an HC PL APO 63x/1.30 Glyc CORR CS objective (Leica) and an ORCA-Flash4.0 digital camera (Hamamatsu Photonics, Hamamatsu City, Japan). Imaging was automated using MetaMorph 7.7 software (MDS Analytical Technologies, Sunnyvale, CA) for x-y well position and the Adaptive Focus Control (Leica) for z position.  The screen was done twice for each mutant array and each well was imaged once, acquiring between 100 and 1000 cells/image. Small-scale imaging was carried out as described above using Concanavalin A-treated 96-well plates. Labeling with 4µM FM4-64 (Life Technologies, Carlsbad, CA) was carried out in synthetic media for 20 min at room temperature, and chased for >20 min. Images were adjusted using Metamorph 7.7 and Adobe Photoshop (Adobe, San Jose, California).  4.5.4 Quantitative Image Analysis Three parameters were quantified for each image: number of live cells, puncta per cell, and colocalization of puncta identified in both green and red channels. Image analysis was carried out using custom MetaMorph 7.7 journals. T-statistics were reported for all of the microscopy measurements.  This is a normalized score calculated by the following equation 𝑥 − 𝑋�𝑠 87  Where x is the raw value, 𝑋� is the mean of the sample set, and s is the standard deviation of the sample set. A value of zero indicates that the selected mutant behaves similarly to the average.  Live cells were identified and counted using the Count Nuclei application in one channel using two different thresholds: 1) lower threshold: to identify the number and location of all cells  2) high threshold: identify dead cells to be masked out. All puncta were identified using the Granularity application. To find puncta in only live cells, the dead cells were masked out by overlaying (using a Logical AND step) an inverted high threshold Count Nuclei resultant image. The number of puncta is then divided by the number of live cells to obtain the number of puncta per cell. Using the Colocalization application, the overlap of green and red puncta identified in previous steps was measured and divided by the total puncta area (either green or red channel). Because Neo1 was tagged with GFP or RFP, all figures report the t-statistic based on the percentage area of overlapping puncta divided by the total Neo1 puncta area fluorescence. Since the brightness of the fluorescently tagged proteins and dyes varied significantly from day to day, all comparisons were made between samples acquired on a given day, and an ANOVA was used to assess the statistical significance of differences between mutants and wild-type.  4.5.5 Small-scale Invertase Overlay Assay Strains containing a chromosomally integrated GFP-Snc1-Suc2 reporter were grown overnight in synthetic media with 2% fructose (SF). Overnight cultures were adjusted to 0.1 OD600/mL, spotted on SF plates and incubated at 30°C for 24 hours. Images were taken before and after incubation with overlay reagents (125mM sucrose, 100mM NaOAC pH 5.5, 0.4mM N-ethylmaleimide, 0.01mg/mL horseradish peroxidase, 8 Units/mL glucose oxidase, 0.6mg/mL O-dianisidine, 0.7% agarose) using a Canon Rebel T3I camera (Canon, Tokyo, Japan). Densitometry of each colony was assessed using the Cell Profiler MeasureObjectIntensity module and the intensity of each set of four was averaged for each replicate.   88  4.5.6 Sequence Alignments Closely related fungal homologs of Neo1 were identified using NCBI BlastP, and aligned using Tcoffee (Waterhouse et al., 2009; Di Tommaso et al., 2011). An image of the resulting alignment was created using Jalview version 2 (Waterhouse et al., 2009).  4.5.7 Co-immunoprecipitation Co-immunoprecipitation was conducted essentially as described in Whitfield et al. (2015) with the following changes: Frozen spheroplasts were lysed in lysis buffer (50mM Tris pH7.5, 100mM NaCl, 1% CHAPSO, 1mM PMSF, 1X Protease Inhibitors (ThermoFisher)). Additionally lysates was incubated 60 min with rabbit polyclonal anti-GFP (Dr. Luc Berthiaume, University of Alberta, Edmonton, AB) and Protein A Sepharose Beads (GE Healthcare, Mississauga, ON). Finally, beads were suspended in Thorner buffer (8M Urea, 5% SDS, 50mM Tris-Cl pH6.8, 0.1M EDTA, 0.4mg/mL Bromophenol blue, 1% β-Me) and subjected to SDS-PAGE.  4.5.8 Growth Assay Cells were grown overnight in synthetic media containing galactose (for growth complementation) or dextrose (drug spot assay). Cells were spun down and concentration was adjusted to 1 OD/mL. Strains were serially diluted 1/10; the first spot is 0.1OD/mL. Plates contain either synthetic medium minus uracil or YPD +/- the 4mg/mL Neomycin prepared in water (Sigma Aldrich, St. Louis, MO) or 30µM Trifluoperazine (Sigma Aldrich). Plates were incubated for the specified amount of time and imaged with a Canon Rebel T3I camera (Canon, Tokyo, Japan).  4.6 Acknowledgements We would like to acknowledge Dr. Luc Berthiaume, Department of Cell Biology, University of Alberta, Canada for his generous gift of rabbit anti-GFP serum. We thank Tony Cooke from Leica Microsystems and Dr. Dan Gottschling, Fred Hutchinson Cancer Research Institute, for helping us develop our automated imaging system. This work was supported by Canadian Institutes for Health Research grants 247169 and 365914. 89  Chapter 5: Discussion and Future Directions 5.1 Efficacy and Future of Using High-Throughput Studies  This study has taken advantage of high-throughput techniques to identify new regulators of endosomal trafficking pathways in yeast, using two different approaches to accomplish this goal.  The first approach used high-throughput data to find positive and negative correlations. Two different predictions from this correlation analysis were explored experimentally, resulting in the identification of an endosomal sorting role for the uncharacterized complex, BLOC, and a new substrate for the known complex, VID. The second approach used a high-content microscopy screen to generate novel insights into the role of flippases in endosomal trafficking. Through this work, many new questions have emerged, prompting exciting avenues for future experimentation.  We showed that correlation analysis can provide directed and relevant predictions to identify functional relationships between proteins. Additionally, we showed that negative correlations within complexes can be used to make robust predictions about inhibitory relationships between proteins.  These inhibitory relationships fell into two categories, either inhibition or competition for shared subunits.  However, there may be much more to learn from negative correlation predictions.  From our analysis, it appeared that negative correlations within complexes are quite rare and a standard p<0.05 cutoff may be too stringent. For example, a recent paper showed that apm1 and apm2, previously thought redundant µ-subunits for the AP-were shown to be required for sorting different cargos (Whitfield et al., 2015). Our analysis, identified a negative correlation between apm1 and apm2 however, the p-value associated with the prediction fell below the p<.25 cutoff. Thus, it is likely that our analysis missed real biological predictions and due to the rarity of negative correlations within complexes, a p-value may not be the best method to derive which predictions are biologically relevant. While there have been targeted screens for various condition-specific pathways in yeast, our work exemplifies the need for more genetic interaction and proteomics yeast studies completed under various nutrient conditions. Ubiquitin ligase research in endosomal trafficking has focused on ubiquitination of substrates by ligases, such as Rsp5, under various conditions such as nutrient limiting scenarios (Belgareh-Touzé et al., 2008). However, while 90  it is shown that the targets of ubiquitin ligases change in different nutrient status, downstream trafficking effects are less well studied in these cases (Clague et al., 2012). To further understand complexes like VID that have different composition and roles under different nutrient conditions, it is necessary that future genetic and proteomics studies focus on varying nutrient conditions to complement those already done and create a more comprehensive view of cellular function.   5.2 Mechanistic Relationship between Yeast BLOC and ESCRT Correlations analysis was used in this thesis to characterize the yeast BLOC complex. We explored the positive correlation identified between BLOC and ESCRTs as well as a negative correlation between BLOC and Smf1.  We found that yeast BLOC subunits localize to endosomes along with ESCRT but they do not require one other to localize, and that BLOC mutants caused a delay in Smf1 degradation similar to the ESCRT mutant, mvb12Δ. Thus, we showed that BLOC is an endosomal sorting complex, needed to target specific proteins to the vacuole. More broadly we also found that correlation analysis can predict the function of uncharacterized protein complexes by comparing profiles with those of known complexes.  However, while we were able to identify a similar function between the yeast BLOC complex and ESCRT, a major caveat of this study is that we have not identified if BLOC and ESCRT work directly with each other or in independent complementary pathways to promote sorting of proteins into the multivesicular body. This is an important question, as no other pathways beyond ESCRT are known to sort proteins into the multivesicular pathway in yeast.  There is evidence to support the model where BLOC and ESCRT work independently. For example, there is evidence that BLOC is needed for recruitment of the RabGAP, Msb3, which downregulates the Rab5 homolog, Vps21 (Peter et al., 2013). Overactive Vps21 has been shown to be the cause of membrane accumulation and disruption of endosomal maturation in ESCRT mutants (Russell et al., 2012), thus disruptions in BLOC may cause an endosomal sorting defect due to inappropriate downregulation of Vps21.  In addition, BLOC has been shown to bind to the vacuolar tethering complex HOPS, which is needed for appropriate fusion of the endosome with the lysosome (Rana et al., 2015). Finally, 91  a new complex has been identified in mammals called BORC which contains all three of the yeast BLOC homologs (Pu et al., 2015).  The function of this complex is not well understood but it is thought to contribute to maintenance and positioning of the lysosome (Pu et al., 2015).  Thus, if yeast BLOC acts more like BORC and less like the mammalian BLOC-1 complex, it may not have a role in direct cargo sorting and affect trafficking indirectly through changes in vacuole placement or endosomal progression through recruitment of Msb3 (Peter et al., 2013; Langemeyer and Ungermann, 2015). Thus, the positive correlation between BLOC and ESCRT may be explained by a shared general role in endosomal maturation.  However, this hypothesis does not yet explain why only some cargo proteins are affected by deletion of BLOC subunits.  We cannot rule out the possibility that BLOC also has a direct role in sorting proteins into the multivesicular body for degradation. There are several scenarios that allow BLOC to work directly with ESCRT for cargo incorporation into the multivesicular body. Large-scale proteomics has suggested that there is an interaction between yeast BLOC and ESCRT-0 subunits (Babu et al., 2012). One possibility is that BLOC binds to ESCRT-0 to aid in cargo recognition. In preliminary studies, we could not show strong binding by coimmunoprecipitation experiments from cell lysates (data not shown); suggesting that they may not bind directly or there is only a very transient interaction. Instead of binding to ESCRT-0, it is possible that BLOC binds and works with ESCRT-I to help recruit proteins for degradation.  In mammalian cells, there is evidence that BLOC-1 subunit BLOS1 binds to Tsg101, an ESCRT-I subunits as well as Snx2, a retromer component.  This binding is needed for Epidermal Growth Factor Receptor (EGFR) degradation (Zhang et al., 2014). BLOC could also work independently of ESCRT-0/I and still recruit ESCRT-III to generate multivesicular bodies. In yeast, Bro1 which has been shown to bind ESCTR-0 to help recruit cargo and initiate ESCRT-III assembly rather than through activation of ESCRT-I/II (Tang et al., 2016). To support this model, deletion of Bro1 has been shown to affect Smf1 trafficking (Eguez et al., 2004). Perhaps, BLOC binds to Bro1 and is able to activate downstream ESCRT complexes independent of ESCRT-0 and I.   It is likely that if BLOC works with a complex to aid in endosomal sorting, it will also bind tightly to this complex. Thus, testing the binding of BLOC to the different ESCRT complexes ESCRT-0, ESCRT-I, ESCRT-III, as well as Bro1 through coimmunoprecipitation 92  binding experiments or SILAC type analysis may yield clarity as to whether BLOC works directly with ESCRT. However, it is possible that binding may be indirect.  Therefore, it would be important to test the specificity of the interaction between BLOC and the binding complex through removing components of the potential bridging complexes. For example, if BLOC binds to ESCRT-I, by deleting ESCRT-0, binding to ESCRT-I should not be disrupted. In this way, comprehensive analysis of BLOC binding to ESCRTs and Bro1 can yield clues as to its function in endosomal sorting. In a complementary approach, turnover of the manganese transporter Smf1, whose trafficking was shown to be affected by BLOC could be assessed in ESCRT-0/BLOC, ESCRT-I/BLOC, Bro1/BLOC, and ESCRT-III/BLOC double mutants compared to the single mutants.  Smf1 turnover may be assessed in two different ways: (1) microscopy of GFP-Smf1, as was done in the study or (2) a manganese sensitivity assay, as a delay in Smf1 turnover will make the yeast sensitive to excess manganese. If the double mutant has the same Smf1 sorting defect, it is likely they work in the same pathway. Conversely, an additive effect where Smf1 trafficking is affected more in the double than either single mutant, would suggest that both contribute in independent ways. In this way, we can test if BLOC ESCRT-0, I, II or Bro1 work in the same pathway for sorting of Smf1.  Other potential cargo proteins Lyp1 and Can1 have also recently been identified and can be assessed using similar methods (Peter et al., 2013).  With the above experiments, we can more confidently answer whether or not BLOC works in a pathway directly to sort cargo for the vacuole.  Finally, it is possible, that BLOC does not work as a cargo adaptor with ESCRT but still acts in the same pathway by changing the cargo such as through modifications so that proteins are more easily recognized by ESCRT.  For instance, perhaps BLOC has roles in ubiquitination by recruiting ubiquitin adaptors or blocking access of deubiquitinating enzymes.  One way to test if BLOC has an effect on ubiqutin is through a co-immunoprecipitation of Smf1 and testing for the presence of ubiquitin in wild-type and BLOC mutants.  Additionally, fusing a deubiquitinating enzyme to BLOC may show if a change in the ubiquitin status is directly due to a ubiquitination step by BLOC. These experiments provide preliminary indications to study how BLOC contributes to endosomal sorting in yeast.  93   5.3 Identifying the Role of a Vaf1-containing VID Complex Through correlation analysis, we identified negative or anticorrelations that occur within a complex, and hypothesized that these negative correlations can predict inhibitory relationships. The negative correlations could be explained by two different mechanisms, direct inhibition or competition for shared subunits. We followed up on a prediction, showing that VID downregulates Vaf1 in a proteasome-dependent manner suggesting that it fits into the inhibition model. However, there is strong physical interaction between Vaf1 and VID, similar to Vid24, which participates with VID in FPBase downregulation but is also turned over by the VID complex (Regelmann et al., 2003). This leaves the question as to whether Vaf1 competes with Vid24 or other regulators to target VID to other substrates, which would be more consistent with a competition model.   Although we found that Vaf1 was targeted for degradation in a VID- and proteasome-dependent manner, a vid30Δ mutant caused a much smaller increase in Vaf1 levels compared to a rpn4∆ mutant that exhibits reduced proteasome levels (Figure 3.4).  This suggests that other ligases might be able to target Vaf1 for proteasomal degradation.  In order to determine other overlapping E3 ligases that target Vaf1 for degradation, we could screen known E2 and E3 ubiquitin ligases for a stabilizing effect on Vaf1 assessed by Western blot.  Thus, we might be able to tell if Vaf1 is specifically degraded by the VID complex, which will give information as to the role of Vaf1 and its relation to VID. Due to the many roles identified for the VID complex, including downregulation of gluconeogenesis, endocytosis, and an autophagic-type response after extended starvation, it is possible that a core complex binds to accessory factors that target the VID complex for its different roles (Brown et al., 2008; Santt et al., 2008; Snowdon et al., 2008; Snowdon and van der Merwe, 2012).  To ascertain if Vaf1 helps target VID to any of these roles or another role not yet found, first, the Vaf1-containing VID complex should be characterized.  Computational analysis to explore the internal configuration of the VID complex suggests that Vaf1 binds to Vid28, Vid30, Gid8 but none of the other components (Pitre et al., 2006).  Vid30, Vid28, and Gid8 have the most connections to other proteins in the complex and may represent the core complex (Pitre et al., 2006). However, a later paper using 94  coimmunoprecipitation presents an alternative topology (Menssen et al., 2012). By systematically deleting each of the known VID components and assessing binding between Vaf1 and Vid30, we may be able to identify the core complex that is needed for Vaf1 incorporation. To further assess the hypothesis that Vaf1 adapts the VID core complex for a cellular role, competitive binding for the VID core complex can be probed.  Due to the negative correlation between Gid7 and Vaf1, these are likely candidates to compete for binding, perhaps targeting the complex for different roles. Overexpression of any non-core members of the VID complex may disrupt binding of Vaf1 to the VID core complex.  Conversely, overexpression of Vaf1 may disrupt binding of other non-core members to the VID core complex.  In this way, the dynamics of the VID complex can be explored.  One possible role for the Vaf1-containing VID complex is regulation of endocytosis. VID30 contains a SPRY domain and along with Ssh4 and Ear1, these are the only yeast genes that do (Menssen et al., 2012). Ear1/Ssh4 bind to the ubiquitin ligase Rsp5 and is needed for the endocytosis of specific proteins from the surface. In addition, the VID complex has been shown to be required for the internalization of the hexose transporter Hxt7 and Hxt3 (Snowdon et al., 2008; Snowdon and van der Merwe, 2012). A vaf1 mutant has a small endocytosis defect in the screen of a chimeric Snc1 reporter (Burston et al., 2009). Thus, Vaf1 could adapt VID for the endocytosis of specific substrates. To test this model, we could determine if Vaf1 is needed in endocytosis of Snc1 or other proteins from the cell surface such as Hxt7. To find more potential VID/Vaf1 dependent endocytosis cargo proteins, a high-throughput microscopy screen of membrane proteins in wild-type, vaf1, vid30, and rsp5 mutants could be carried out to test if the progression of these proteins to the vacuole is altered. In this way, we may be able to characterize a new role of the VID complex and also explain the role of Vaf1 in the VID complex.   5.4 Mon2 and Dop1 Regulation of Neo1 In our high-content imaging screen, we compared the localization of Mon2 and Dop1 with Neo1. We predicted that a disruption in Neo1 localization would also disrupt Dop1 and Mon2 localization as they are part of a trimeric complex. However, we were surprised to find, that when Neo1 was sorted to the vacuole, Dop1 and Mon2 were still able to localize to 95  the membrane as puncta (Chapter 4).  Notably, members of the Arl pathway caused a specific disruption in Dop1 but not Mon2 colocalization with Neo1 and some of the reasons for this separation were tested in Chapter 4. However, how Neo1 is regulated by Mon2 and Dop1 and which aspects of Neo1 function require Dop1 and Mon2 still remains to be explored further. One hypothesis is that Dop1 and Mon2 help to regulate Neo1 by acting as localization determinants. However, it did not appear that deletion of Mon2 caused a disruption in the localization of Neo1. It is not known if disruption of Dop1 causes a change in Neo1 localization, though in a mon2Δ mutant, Dop1 recruitment to the membrane is severely affected (Gillingham et al., 2006). Thus, it is possible that either enough Dop1 is recruited to localize Neo1 or this is not the role Dop1 imparts on Neo1. It seems unlikely from this evidence that Dop1 and Mon2 are only needed for localization of Neo1.  A second and perhaps more favorable hypothesis then is that Neo1 activity is modulated at least in part by Dop1 and/or Mon2. Activity in other flippases is modulated by a β-subunit from the Cdc50 family (Saito et al., 2004; Chen et al., 2006; Lenoir et al., 2009; Bryde et al., 2010; Takahashi et al., 2011). Neo1 and its mammalian homologs do not bind any known members of the Cdc50 family (Barbosa et al., 2010; Takatsu et al., 2011).  Perhaps then binding of Dop1/Mon2 to Neo1 is needed for activity.  It is important to consider that even if Mon2 and Dop1 are required for Neo1 activity, they may not be needed for all Neo1 activity throughout the cell and may instead regulate Neo1 activity only at post-Golgi compartments. Both Dop1/Mon2 are needed for Snc1 trafficking (Gillingham et al., 2006). Thus, if Neo1 is needed for sorting of Snc1, and Dop1/Mon2 are also needed, this suggests that Dop1 and Mon2 may work with Neo1 to sort Snc1. One way to test if Dop1 and Mon2 are needed for Neo1 activity is to identify mutations in either Mon2 or Neo1 that disrupt their binding, but do not affect binding to other factors, and track recycling of Snc1 and Snx3-dependent cargos such as A-ALP. The same approach can be taken with Dop1 if disrupting binding of Dop1 and Neo1 is not lethal. These experiments will test if binding is required for correct sorting but will not assess activity directly.  96  However, it is possible that disruption of Dop1 and Mon2 causes a decrease in Neo1 trafficking from the Golgi, causing a shift in its localization that results in decreased activity. We saw that Neo1 remained punctate in a mon2Δ mutant, however, we did not determine if there was a shift in endosome or Golgi localization. Thus, detailed localization of Neo1 when Dop1 and Mon2 binding is disrupted will be needed to determine how Dop1 and Mon2 contribute to Neo1 activity at the late endosome.  It is interesting to think how Mon2 and Dop1 may modulate Neo1 activity. Because deletion of Dop1 is lethal, it is harder to test modulation of Neo1 activity in a direct way. Mon2, on the other hand, is not essential and may be easily tested. Mon2 is a very large protein, containing a Sec7 domain, which normally confers guanine nucleotide exchange (Efe et al., 2005; Mahajan et al., 2013).  However, there is evidence suggesting that the Mon2 Sec7 domain is not active but may act as a recruiting platform (Mahajan et al., 2013).  To test this idea, high-throughput protein interaction data can be mined to find Mon2 interacting proteins. Following up on these hits, testing localization of these proteins in a mon2Δ strain may indicate if Mon2 is needed for their recruitment. It is possible, however, that many proteins may be recruited by more than one factor, so multiple deletions may be required to see a defect. In all, these experiments represent a starting point to study the regulation of Neo1 by Dop1 and Mon2.  5.5 Role of Neo1 at the Late Endosome From our high-content microscopy screen, we identified that Neo1 sorting by Snx3 is required for sorting of other Snx3-dependent cargos.  This exciting discovery suggests that sorting of Neo1 like Drs2 is needed for trafficking of other proteins in that pathway. The mechanistic contribution of Neo1 to endosomal recycling is an exciting avenue for future experimentation.  A brief description of retromer-dependent vesicle trafficking will give context to a potential contribution of Neo1 to Snx3-dependent trafficking at the late endosome. In general, the cargo-selective portion of retromer with or without Snx3 recognizes cargo proteins and sequesters them.  The structural domain made up of two SNX-BAR proteins is thought to generate tubules that eventually become vesicles bound for the Golgi. Two 97  hypotheses exist for tubulation by BAR domains.  In one model, BAR domains are solely responsible for tubulation from otherwise flat membranes. Conversely, in another model, BAR domains maintain and enhance curvature that is generated by other mechanisms such as amphipathic helices or flippase activity (van Weering et al., 2012).  Our evidence that Snx3 sorting of Neo1 is needed for sorting of other Snx3 cargos supports a model where BAR domains maintain and enhance curvature, however, flippases are needed to generate the initial curvature leading to tubule formation at the late endosome.  However, we cannot rule out the possibility that flippases enhance the ability of SNX-BAR proteins to promote tubulation, perhaps by increasing tubule scission. Interestingly, the mammalian Drs2 homologs ATP8A1 and ATP8A2 are needed to recruit EHD1, a protein needed for endosomal tubule scission, suggesting a general requirement for flippases in retromer function (Lee et al., 2015). To explore this question, A-ALP recycling can be monitored in flippase catalytic mutants compared to wild-type Neo1 or the Neo1 FEM mutant, which is predicted to be absent from Snx3-dependent tubules, to determine if the activity is required for sorting at the late endosome. The chimeric Neo1-ALP is able to localize correctly but does not have flippase activity, suggesting activity is not needed for sorting in a strain that has wild type Neo1. However, direct testing of a catalytic mutant may be difficult, as our preliminary evidence suggests that a catalytic Neo1 mutant is not able to leave the endoplasmic reticulum. Thus, a temperature sensitive variant may have to be developed to use this approach.  Additionally, to get around this issue, we could test if overexpression of Neo1 causes A-ALP sorting to be more efficient; suggesting that increased activity is beneficial.  Also, it might be possible to switch the catalytic domain from another flippase; however, it is possible that there might be overlapping functionalities. Conversely, in vitro tubulation assays can be used to see if incorporating active or inactive flippase into liposomes will allow tubulation to occur in the presence of retromer SNX-BAR proteins. These experiments represent a starting point to identify if Neo1 activity is needed for late endosome recycling pathways. It is also interesting to explore whether other flippases may also contribute to late endosome recycling. We observed a partial defect in A-ALP sorting in yeast containing the Neo1 Snx3-binding mutant. So either Neo1 is not essential for retromer activity, or other 98  flippases also act at the late endosome. Dnf3 and Drs2 have the potential to also act in these pathways as they both have intracellular steady-state localization (Pomorski et al., 2003). Additionally, Dnf3 and Drs2 have a significant Snc1 internalization phenotype from our screen, suggesting they may also have a role in endosomal recycling. To test this model, first, the localization of Drs2 and Dnf3 can be monitored to determine if they are Snx3 or retromer-dependent. Then a mutation in Dnf3 or Drs2 combined with Neo1 FEM mutant can be tested for an enhancement of the A-ALP sorting defect.  In this way, we can determine the functional overlap of flippases at the late endosome. In all, evidence that flippases contribute to vesicle trafficking is emerging and by studying the role of Neo1 at late endosomes a more comprehensive view of flippase action can be achieved.  5.6 Conclusion Here we show that using data generated from high-throughput screens is an effective method to identify and characterize new aspects of biology.  We used a couple of different high-throughput techniques and made many new discoveries that enhance our understanding of endosomal sorting. However, these techniques are applicable to any area of biology and due to the conservation of yeast and mammals, findings based on studies in yeast can be applied to extend our knowledge of mammalian systems, creating a potential to inform therapies to treat disease. 99  References Amerik, A., Sindhi, N., and Hochstrasser, M. (2006). A conserved late endosome-targeting signal required for Doa4 deubiquitylating enzyme function. J. Cell Biol. 175, 825–835. Andersen, J. P., Vestergaard, A. L., Mikkelsen, S. A., Mogensen, L. S., Chalat, M., and Molday, R. S. (2016). P4-ATPases as Phospholipid Flippases Structure, Function, and Enigmas. Front. Physiol. 7, 275. Antonny, B., Beraud-Dufour, S., Chardin, P., and Chabre, M. (1997). N-terminal hydrophobic residues of the G-protein ADP-ribosylation factor-1 insert into membrane phospholipids upon GDP to GTP exchange. Biochemistry 36, 4675–4684. Babst, M., Katzmann, D. J., Estepa-Sabal, E. J., Meerloo, T., and Emr, S. D. (2002a). ESCRT-III: An Endosome-Associated Heterooligomeric Protein Complex Required for MVB Sorting. Dev. Cell 3, 271–282. Babst, M., Katzmann, D. J., Snyder, W. B., Wendland, B., and Emr, S. D. (2002b). Endosome-associated complex, ESCRT-II, recruits transport machinery for protein sorting at the multivesicular body. Dev. Cell 3, 283–289. Babst, M., Odorizzi, G., Estepa, E. J., and Emr, S. D. (2000). Mammalian Tumor Susceptibility Gene 101 (TSG101) and the Yeast Homologue, Vps23p, Both Function in Late Endosomal Trafficking. Traffic 1, 248–258. Babu, M. et al. (2012). Interaction landscape of membrane-protein complexes in Saccharomyces cerevisiae. Nature 489, 585–589. Baldridge, R. D., and Graham, T. R. (2011). Identification of residues defining phospholipid flippase substrate specificity of type IV P-type ATPases. Proc. Natl. Acad. Sci. 109, E290–E298. Baldridge, R. D., and Graham, T. R. (2012). Two-gate mechanism for phospholipid selection and transport by type IV P-type ATPases. Proc. Natl. Acad. Sci. U. S. A. 110, E358–E367. Baldridge, R. D., Xu, P., and Graham, T. R. (2013). Type IV p-type ATPases distinguish mono-versus diacyl phosphatidylserine using a cytofacial exit gate in the membrane domain. J. Biol. Chem. 288, 19516–19527. Barbosa, S., Pratte, D., Schwarz, H., Pipkorn, R., and Singer-Krüger, B. (2010). Oligomeric Dop1p is part of the endosomal Neo1p-Ysl2p-Arl1p membrane remodeling complex. Traffic 11, 1092–1106. Bean, B. D. M., Davey, M., Snider, J., Jessulat, M., Deineko, V., Tinney, M., Stagljar, I., Babu, M., and Conibear, E. (2015). Rab5-family guanine nucleotide exchange factors bind retromer and promote its recruitment to endosomes. Mol. Biol. Cell 26, 1119–1128. Belgareh-Touzé, N., Léon, S., Erpapazoglou, Z., Stawiecka-Mirota, M., Urban-Grimal, D., and Haguenauer-Tsapis, R. (2008). Versatile role of the yeast ubiquitin ligase Rsp5p in intracellular trafficking. Biochem. Soc. Trans. 36, 791–796. Beltrao, P., Cagney, G., and Krogan, N. J. (2010). Quantitative Genetic Interactions Reveal 100  Layers of Biological Modularity. Cell 141, 739–745. Bonangelino, C., Chavez, E., and Bonifacino, J. S. (2002). Genomic Screen for Vacuolar Protein Sorting Genes in Saccharomyces cerevisiae. Mol. Biol. Cell 13, 2001–2015. Bonifacino, J. S. (2004). The GGA proteins: adaptors on the move. Nat Rev Mol Cell Biol 5, 23–32. Bonifacino, J. S., and Lippincott-Schwartz, J. (2003). Coat proteins: shaping membrane transport. Nat. Rev. Mol. Cell Biol. 4, 409–414. Bonifacino, J. S., and Traub, L. M. (2003). Signals for Sorting of Transmembrane Proteins to Endosomes and Lysosomes. Annu. Rev. Biochem. 72, 395–447. Boone, C., Bussey, H., and Andrews, B. J. (2007). Exploring genetic interactions and networks with yeast. Nat. Rev. Genet. 8, 437–449. Bougen-Zhukov, N., Loh, S. Y., Lee, H. K., and Loo, L.-H. (2016). Large-scale image-based screening and profiling of cellular phenotypes. Cytom. Part A, 1–11. Breslow, D. K., Collins, S. R., Bodenmiller, B., Aebersold, R., Simons, K., Shevchenko, A., Ejsing, C. S., and Weissman, J. S. (2010). Orm family proteins mediate sphingolipid homeostasis. Nature 463, 1048–1053. Brickner, J. H., and Fuller, R. S. (1997). SOI1 encodes a novel, conserved protein that promotes TGN-endosomal cycling of Kex2p and other membrane proteins by modulating the function of two TGN localization signals. J. Cell Biol. 139, 23–36. Brigance, W. T., Barlowe, C., and Graham, T. R. (2000). Organization of the yeast Golgi complex into at least four functionally distinct compartments. Mol. Biol. Cell 11, 171–182. Broach, J. R. (2012). Nutritional control of growth and development in yeast. Genetics 192, 73–105. Brown, C. R., Wolfe, A. B., Cui, D., and Chiang, H.-L. (2008). The vacuolar import and degradation pathway merges with the endocytic pathway to deliver fructose-1,6-bisphosphatase to the vacuole for degradation. J. Biol. Chem. 283, 26116–26127. Bryde, S., Hennrich, H., Verhulst, P. M., Devaux, P. F., Lenoir, G., and Holthuis, J. C. M. (2010). CDC50 proteins are critical components of the human class-1 P 4-ATPase transport machinery. J. Biol. Chem. 285, 40562–40572. Burd, C., and Cullen, P. J. (2016). Retromer : A Master Conductor of Endosome Sorting. Burda, P., Padilla, S. M., Sarkar, S., and Emr, S. D. (2002). Retromer function in endosome-to-Golgi retrograde transport is regulated by the yeast Vps34 PtdIns 3-kinase. J. Cell Sci. 115, 3889–3900. Burri, L., and Lithgow, T. (2004). A Complete Set of SNAREs in Yeast. Traffic 5, 45–52. Burston, H. E., Davey, M., and Conibear, E. (2008). Genome-Wide Analysis of Membrane Transport Using Yeast Knockout Arrays. In: Membrane Trafficking, ed. A. Vancura, Totowa, NJ: Humana Press, 29–39. Burston, H. E., Maldonado-Báez, L., Davey, M., Montpetit, B., Schluter, C., Wendland, B., 101  and Conibear, E. (2009). Regulators of yeast endocytosis identified by systematic quantitative analysis. J. Cell Biol. 185, 1097–1110. Butland, G. et al. (2005). Interaction network containing conserved and essential protein complexes in Escherichia coli. Nature 433, 531–537. Carlton, J. G., and Cullen, P. J. (2005). Sorting nexins. Curr. Biol. 15, R819-20. Catty, P., De Kerchove D’Exaerde, A., and Goffeau, A. (1997). The complete inventory of the yeast Saccharomyces cerevisiae P-type transport ATPases. FEBS Lett. 409, 325–332. Chantalat, S., Park, S.-K., Hua, Z., Liu, K., Gobin, R., Peyroche, A., Rambourg, A., Graham, T. R., and Jackson, C. L. (2004). The Arf activator Gea2p and the P-type ATPase Drs2p interact at the Golgi in Saccharomyces cerevisiae. J. Cell Sci. 117, 711–722. Chen, C., Ingram, M. F., Rosal, P. H., and Graham, T. R. (1999). Role for Drs2p, a P-Type ATPase and Potential Aminophospholipid Translocase, in Yeast Late Golgi Function. Cell 147, 1223–1236. Chen, L., and Davis, N. G. (2002). Ubiquitin-independent entry into the yeast recycling pathway. Traffic 3, 110–123. Chen, S. H., Shah, A. H., and Segev, N. (2011). Ypt31/32 GTPases and their F-Box effector Rcy1 regulate ubiquitination of recycling proteins. Cell. Logist. 1, 21–31. Chen, S., Wang, J., Muthusamy, B. P., Liu, K., Zare, S., Andersen, R. J., and Graham, T. R. (2006). Roles for the Drs2p-Cdc50p complex in protein transport and phosphatidylserine asymmetry of the yeast plasma membrane. Traffic 7, 1503–1517. Chi, R. J., Liu, J., West, M., Wang, J., Odorizzi, G., and 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. Clague, M. J., Liu, H., and Urbe, S. (2012). Review Governance of Endocytic Trafficking and Signaling by Reversible Ubiquitylation. Dev. Cell 23, 457–467. Clark, M. R. (2011). Flippin’ lipids. Nat. Immunol. 12, 373–375. Clotet, J., Gari, E., Aldea, M., and Arino, J. (1999). The Yeast Ser / Thr Phosphatases Sit4 and Ppz1 Play Opposite Roles in Regulation of the Cell Cycle. Mol. Cell. Biol. 19, 2408–2415. Coleman, J. A., Kwok, M. C. M., and Molday, R. S. (2009). Localization, purification, and functional reconstitution of the P4-ATPase Atp8a2, a phosphatidylserine Flippase in photoreceptor disc membranes. J. Biol. Chem. 284, 32670–32679. Collins, M. O., Yu, L., and Choudhary, J. S. (2007). Analysis of protein phosphorylation on a proteome-scale. Proteomics 7, 2751–2768. Conibear, E., Cleck, J. N., and Stevens, T. H. (2003). Vps51p Mediates the Association of the GARP (Vps52/ 53/54) Complex with the Late Golgi t-SNARE Tlg1p. Mol. Biol. Cell 14, 1610–1623. Conibear, E., and Stevens, T. H. (2000). Vps52p , Vps53p , and Vps54p Form a Novel 102  Multisubunit Complex Required for Protein Sorting at the Yeast Late Golgi. Mol. Biol. Cell 11, 305–323. Costaguta, G., Stefan, C. J., Bensen, E. S., Emr, S. D., and Payne, G. S. (2001). Yeast Gga coat proteins function with clathrin in Golgi to endosome transport. Mol. Biol. … 12, 1885–1896. Costanzo, M. et al. (2010). The genetic landscape of a cell. Science 327, 425–431. Dalton, L., Davey, M., and Conibear, E. (2015). Large-Scale Analysis of Membrane Transport in Yeast Using Invertase Reporters. In: Membrane Trafficking:Second Edition, ed. B. L. Tang, New York: Springer Science+Business Media, 395–409. Darsow, T., Odorizzi, G., and Emr, S. D. (2000). Invertase Fusion Proteins for Analysis of Protein Trafficking in Yeast. In: Methods in Enzymology, 95–106. De, M., Abazeed, M. E., and Fuller, R. S. (2013). Direct binding of the Kex2p cytosolic tail to the VHS domain of yeast Gga2p facilitates TGN to prevacuolar compartment transport and is regulated by phosphorylation. Mol. Biol. Cell 24, 495–509. Decourty, L., Saveanu, C., Zemam, K., Hantraye, F., Frachon, E., Rousselle, J.-C., Fromont-Racine, M., and Jacquier, A. (2008). Linking functionally related genes by sensitive and quantitative characterization of genetic interaction profiles. Proc. Natl. Acad. Sci. U. S. A. 105, 5821–5826. Dell’Angelica, E. C. (2004). The building BLOC(k)s of lysosomes and related organelles. Curr. Opin. Cell Biol. 16, 458–464. Dell’Angelica, E. C. (2009). AP-3-dependent trafficking and disease: the first decade. Curr. Opin. Cell Biol. 21, 552–559. Dell’Angelica, E. C., Puertollano, R., Mullins, C., Aguilar, R. C., Vargas, J. D., Hartnell, L. M., and Bonifacino, J. S. (2000). GGAs: A family of ADP ribosylation factor-binding proteins related to adaptors and associated with the Golgi complex. J. Cell Biol. 149, 81–93. Denks, K., Vogt, A., Sachelaru, I., Petriman, N., Kudva, R., and Koch, H. (2014). The Sec translocon mediated protein transport in prokaryotes and eukaryotes. Mol. Membr. Biol. 31, 58–84. Dittmar, J. C., Pierce, S., Rothstein, R., and Reid, R. J. D. (2013). Physical and genetic-interaction density reveals functional organization and informs significance cutoffs in genome-wide screens. Proc. Natl. Acad. Sci. U. S. A. 110, 7389–7394. Dokudovskaya, S. et al. (2011). A conserved coatomer-related complex containing Sec13 and Seh1 dynamically associates with the vacuole in Saccharomyces cerevisiae. Mol. Cell. Proteomics 10, M110.006478. Dores, M. R., Paing, M. M., Lin, H., Montagne, W. A., Marchese, A., and Trejo, J. (2012). AP-3 regulates PAR1 ubiquitin-independent MVB / lysosomal sorting via an ALIX-mediated pathway. Mol. Biol. Cell 23, 3612–3623. Drin, G., and Antonny, B. (2010). Amphipathic helices and membrane curvature. FEBS Lett. 584, 1840–1847. 103  Dubuke, M. L., and Munson, M. (2016). The Secret Life of Tethers: The Role of Tethering Factors in SNARE Complex Regulation. Front. Cell Dev. Biol. 4, 1–8. Ear, P. H., Kowarzyk, J., Booth, M. J., Abd-Rabbo, D., Shulist, K., Hall, C., Vogel, J., and Michnick, S. W. (2016). Combining the Optimized Yeast Cytosine Deaminase Protein Fragment Complementation Assay and an In Vitro Cdk1 Targeting Assay to Study the Regulation of the γ-Tubulin Complex. In: Methods in Molecular Biology, ed. A. S. Coutts, and ed. L. Weston, New York: Springer Science+Business Media, 157–171. Efe, J. a, Plattner, F., Hulo, N., Kressler, D., Emr, S. D., and Deloche, O. (2005). Yeast Mon2p is a highly conserved protein that functions in the cytoplasm-to-vacuole transport pathway and is required for Golgi homeostasis. J. Cell Sci. 118, 4751–4764. Eguez, L., Chung, Y. S., Kuchibhatla, A., Paidhungat, M., and Garrett, S. (2004). Yeast Mn2+ transporter, Smf1p, is regulated by ubiquitin-dependent vacuolar protein sorting. Genetics 167, 107–117. Falcón-Pérez, J. M., Starcevic, M., Gautam, R., and Dell’Angelica, E. C. (2002). BLOC-1, a novel complex containing the pallidin and muted proteins involved in the biogenesis of melanosomes and platelet-dense granules. J. Biol. Chem. 277, 28191–28199. Fields, S., and Song, O. (1989). A novel genetic system to detect protein-protein interactions. Nature 340, 245–246. Fjorback, A. W. et al. (2012). Retromer binds the FANSHY sorting motif in SorLA to regulate amyloid precursor protein sorting and processing. J. Neurosci. 32, 1467–1480. Foote, C., and Nothwehr, S. F. (2006). The clathrin adaptor complex 1 directly binds to a sorting signal in Ste13p to reduce the rate of its trafficking to the late endosome of yeast. J. Cell Biol. 173, 615–626. Fotso, P., Koryakina, Y., Pavliv, O., Tsiomenko, A. B., and Lupashin, V. V (2005). Cog1p plays a central role in the organization of the yeast conserved oligomeric Golgi complex. J. Biol. Chem. 280, 27613–27623. Furuta, N., Fujimura-Kamada, K., Saito, K., Yamamoto, T., and Tanaka, K. (2007). Endocytic Recycling in Yeast Is Regulated by Putative Phospholipid Translocases and the Ypt31p / 32p – Rcy1p Pathway. Mol. Biol. Cell 18, 295–312. Gautam, R., Novak, E. K., Tan, J., Wakamatsu, K., Ito, S., and Swank, R. T. (2006). Interaction of Hermansky-Pudlak Syndrome genes in the regulation of lysosome-related organelles. Traffic 7, 779–792. Gavin, A.-C. et al. (2002). Functional organization of the yeast proteome by systematic analysis of protein complexes. Nature 415, 141–147. Gavin, A.-C. et al. (2006). Proteome survey reveals modularity of the yeast cell machinery. Nature 440, 631–636. Ghosh, P., and Kornfeld, S. (2004). The GGA proteins: key players in protein sorting at the trans-Golgi network. Eur. J. Cell Biol. 83, 257–262. Giaever, G. et al. (2002). Functional profiling of the Saccharomyces cerevisiae genome. 104  Nature 418, 387–391. Gillingham, A. K., Whyte, J. R. C., Panic, B., and Munro, S. (2006). Mon2, a relative of large Arf exchange factors, recruits Dop1 to the Golgi apparatus. J. Biol. Chem. 281, 2273–2280. Gonzalez, D. et al. (2014). Suppression of Mediator is regulated by Cdk8-dependent Grr1 turnover of the Med3 coactivator. Proc. Natl. Acad. Sci. U. S. A. 111, 2500–2505. Graham, T. R. (2004). Flippases and vesicle-mediated protein transport. Trends Cell Biol. 14, 670–677. Guarente, L. (1993). Synthetic enhancement in gene interaction: a genetic tool come of age. Trends Genet. 9, 362–366. Ha, S., Torabinejad, J., DeWald, D. B., Wenk, M. R., Lucast, L., De Camilli, P., Newitt, R. A., Aebersold, R., and Nothwehr, S. F. (2003). Synaptojanin-like Protein Inp53/Sjl3 Functions with Clathrin in a Yeast TGN-to-Endosome Pathway Distinct from the GGA Protein-dependent Pathway. Mol. Biol. Cell 14, 1319–1333. Halleck, M. S., Pradhan, D., Blackman, C., Berkes, C., Williamson, P., and Schlegel, R. A. (1998). Multiple members of a third subfamily of P-type ATPases identified by genomic sequences and ESTs. Genome Res. 8, 354–361. Hanamatsu, H., Fujimura-Kamada, K., Yamamoto, T., Furuta, N., and Tanaka, K. (2014). Interaction of the phospholipid flippase Drs2p with the F-box protein Rcy1p plays an important role in early endosome to trans-Golgi network vesicle transport in yeast. J. Biochem. 155, 51–62. Hayes, M. J., Bryon, K., Satkurunathan, J., and Levine, T. P. (2011). Yeast homologues of three BLOC-1 subunits highlight KxDL proteins as conserved interactors of BLOC-1. Traffic 12, 260–268. Helenius, A., and Aebi, M. (2004). Roles of N-linked glycans in the endoplasmic reticulum. Annu. Rev. Biochem. 73, 1019–1049. Hettema, E. H., Lewis, M. J., Black, M. W., and Pelham, H. R. B. (2003). Retromer and the sorting nexins Snx4/41/42 mediate distinct retrieval pathways from yeast endosomes. EMBO J. 22, 548–557. Hierro, A., Rojas, A. L., Rojas, R., Murthy, N., Effantin, G., Kajava, A. V, Steven, A. C., Bonifacino, J. S., and Hurley, J. H. (2007). Function architecture of the retromer cargo-recognition complex. Nature 449, 1063–1067. Hillenmeyer, M. E. et al. (2008). The Chemical Genomic Portrait of Yeast: Uncovering a Phenotype for All Genes. Science (80-. ). 320, 362–365. Ho, C. H. et al. (2009). A molecular barcoded yeast ORF library enables mode-of-action analysis of bioactive compounds. Nat. Biotechnol. 27, 369–377. Hoepfner, D. et al. (2014). High-resolution chemical dissection of a model eukaryote reveals targets, pathways and gene functions. Microbiol. Res. 169, 107–120. Hong, W. J., and Lev, S. (2014). Tethering the assembly of SNARE complexes. Trends Cell 105  Biol. 24, 35–43. Hsu, J.-W., Chen, Z.-J., Liu, Y.-W., and Lee, F.-J. S. (2014). Mechanism of action of flippase Drs2p in modulating GTP hydrolysis of Arl1p. J. Cell Sci., 2615–2620. Hsu, V., Lee, S., and Yang, J. (2009). The evolving understanding of COPI vesicle formation. Nat. Rev. Mol. Cell Biol. 1, 33–51. Hsu, V. W., and Yang, J.-S. (2009). Mechanisms of COPI vesicle formation. FEBS Lett. 583, 3758–3763. Hu, P. et al. (2009). Global functional atlas of Escherichia coli encompassing previously uncharacterized proteins. PLoS Biol. 7, 0929–0947. Hua, Z., Fatheddin, P., and Graham, T. R. (2002). An Essential Subfamily of Drs2p-related P-Type ATPases Is Required for Protein Trafficking between Golgi Complex and Endosomal / Vacuolar System. Mol. Biol. Cell 13, 3162–3177. Huh, W. K., Falvo, J. V, Gerke, L. C., Carroll, A. S., Howson, R. W., Weissman, J. S., and O’Shea, E. K. (2003). Global analysis of protein localization in budding yeast. Nature 425, 686–691. Hung, G. C., Brown, C. R., Wolfe, A. B., Liu, J., and Chiang, H. L. (2004). Degradation of the Gluconeogenic Enzymes Fructose-1,6- bisphosphatase and Malate Dehydrogenase Is Mediated by Distinct Proteolytic Pathways and Signaling Events. J. Biol. Chem. 279, 49138–49150. Jablonowski, D., Täubert, J. E., Bär, C., Stark, M. J. R., and Schaffrath, R. (2009). Distinct subsets of Sit4 holophosphatases are required for inhibition of Saccharomyces cerevisiae growth by rapamycin and zymocin. Eukaryot. Cell 8, 1637–1647. Jacquot, A., Montigny, C., Hennrich, H., Barry, R., le Maire, M., Jaxel, C., Holthuis, J., Champeil, P., and Lenoir, G. (2012). Phosphatidylserine stimulation of Drs2p·Cdc50p lipid translocase dephosphorylation is controlled by phosphatidylinositol-4-phosphate. J. Biol. Chem. 287, 13249–13261. Jahn, R., and Scheller, R. H. (2006). SNAREs — engines for membrane fusion. Nat. Rev. Mol. Cell Biol. 7, 631–643. Janke, C. et al. (2004). A versatile toolbox for PCR-based tagging of yeast genes: New fluorescent proteins, more markers and promoter substitution cassettes. Yeast 21, 947–962. Jensen, D., and Schekman, R. (2011). COPII-mediated vesicle formation at a glance. J. Cell Sci. 124, 1–4. Jensen, L. T., Carroll, M. C., Hall, M. D., Harvey, C. J., Beese, S. E., and Culotta, V. C. (2009). Down-Regulation of a Manganese Transporter in the Face of Metal Toxicity. Mol. Biol. Cell 20, 2810–2819. Jochum, A., Jackson, D., Schwarz, H., Pipkorn, R., and Singer-Krüger, B. (2002). Yeast Ysl2p, homologous to Sec7 domain guanine nucleotide exchange factors, functions in endocytosis and maintenance of vacuole integrity and interacts with the Arf-Like small GTPase Arl1p. Mol. Cell. Biol. 22, 4914–4928. 106  Kaake, R. M., Wang, X., and Huang, L. (2010). Profiling of protein interaction networks of protein complexes using affinity purification and quantitative mass spectrometry. Mol. Cell. Proteomics 9, 1650–1665. Kaksonen, M., Toret, C. P., and Drubin, D. G. (2005). A modular design for the clathrin- and actin-mediated endocytosis machinery. Cell 123, 305–320. Katzmann, D. J., Babst, M., and 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, 145–155. Kihara, A., Noda, T., Ishihara, N., and Ohsumi, Y. (2001). Two distinct Vps34 phosphatidylinositol 3-kinase complexes function in autophagy and carboxypeptidase y sorting in Saccharomyces cerevisiae. J. Cell Biol. 153, 519–530. Kirchhausen, T. (2000). Three ways to make a vesicle. Nat. Rev. Mol. Cell Biol. 1, 187–198. Krogan, N. J. et al. (2006). Global landscape of protein complexes in the yeast Saccharomyces cerevisiae. Nature 440, 637–643. Lachmann, J., Barr, F. a, and Ungermann, C. (2012). The Msb3/Gyp3 GAP controls the activity of the Rab GTPases Vps21 and Ypt7 at endosomes and vacuoles. Mol. Biol. Cell 23, 2516–2526. Lam, K. K. Y., Davey, M., Sun, B., Roth, A. F., Davis, N. G., and Conibear, E. (2006). Palmitoylation by the DHHC protein Pfa4 regulates the ER exit of Chs3. J. Cell Biol. 174, 19–25. Lang, A. B., Peter, A. T. J., Walter, P., and Kornmann, B. (2015). ER-mitochondrial junctions can be bypassed by dominant mutations in the endosomal protein Vps13. J. Cell Biol. 210, 883–890. Langemeyer, L., and Ungermann, C. (2015). BORC and BLOC-1: Shared Subunits in Trafficking Complexes. Dev. Cell 33, 121–122. Lee, S. et al. (2015). Transport through recycling endosomes requires EHD1 recruitment by a phosphatidylserine translocase. EMBO J. 34, 669–688. Lee, S., Lim, W. A., and Thorn, K. S. (2013). Improved Blue, Green, and Red Fluorescent Protein Tagging Vectors for S. cerevisiae. PLoS One 8, 4–11. Legrain, P., and Rain, J.-C. (2014). Twenty years of protein interaction studies for biological function deciphering. J. Proteomics 107, 93–97. Lemmon, M. (2008). Membrane recognition by phospholipid-binding domains. Nat. Rev. Mol. Cell Biol. 9, 99–111. Lenoir, G., Williamson, P., Puts, C. F., and Holthuis, J. C. M. (2009). Cdc50p plays a vital role in the ATPase reaction cycle of the putative aminophospholipid transporter Drs2p. J. Biol. Chem. 284, 17956–17967. Liu, K., Hua, Z., Nepute, J. A., and Graham, T. R. (2007). Yeast P4-Atpases Drs2p and Dnf1p Are Essential Cargos of the NPFXD/Sal1p Endocytic Pathway. Mol. Biol. Cell 18, 2991–3001. 107  Liu, K., Surendhran, K., Nothwehr, S. F., and Graham, T. R. (2008). P4-ATPase Requirement for AP-1 / Clathrin Function in Protein Transport from the trans -Golgi Network and Early Endosomes. Mol. Biol. Cell 19, 3526–3535. Liu, T.-T., Gomez, T. S., Sackey, B. K., Billadeau, D. D., and Burd, C. G. (2012). Rab GTPase regulation of retromer-mediated cargo export during endosome maturation. Mol. Biol. Cell 23, 2505–2515. Longtine, M. S., McKenzie, A., Demarini, D. J., Shah, N. G., Wach, A., Brachat, A., Philippsen, P., and Pringle, J. R. (1998). Additional modules for versatile and economical PCR-based gene deletion and modification in Saccharomyces cerevisiae. Yeast 14, 953–961. Luke, M. M., Della Seta, F., Di Como, C. J., Sugimoto, H., Kobayashi, R., and Arndt, K. T. (1996). The SAP, a new family of proteins, associate and function positively with the SIT4 phosphatase. Mol. Cell. Biol. 16, 2744–2755. Ma, X., Tarone, A. M., and Li, W. (2008). Mapping genetically compensatory pathways from synthetic lethal interactions in yeast. PLoS One 3. Macdonald, C., Stringer, D. K., and Piper, R. C. (2012). Sna3 is an Rsp5 Adaptor Protein that Relies on Ubiquitination for its MVB Sorting. Traffic 13, 586–598. Mahajan, D., Boh, B. K., Zhou, Y., Chen, L., Cornvik, T. C., Hong, W., and Lu, L. (2013). Mammalian Mon2/Ysl2 regulates endosome-to-Golgi trafficking but possesses no guanine nucleotide exchange activity toward Arl1 GTPase. Sci. Rep. 3, 3362. Mak, A. B. et al. (2010). A lentiviral-based functional proteomics approach identifies chromatin remodeling complexes important for the induction of pluripotency. Mol Cell Proteomics 9, 811–823. Manlandro, C. M. A., Palanivel, V. R., Schorr, E. B., Mihatov, N., Antony, A. A., and Rosenwald, A. G. (2012). Mon2 is a negative regulator of the monomeric G protein, Arl1. FEMS Yeast Res. 12, 637–650. Martina, J. A., Moriyama, K., and Bonifacino, J. S. (2003). BLOC-3, a protein complex containing the Hermansky-Pudlak syndrome gene products HPS1 and HPS4. J. Biol. Chem. 278, 29376–29384. Mateja, A., Paduch, M., Chang, H.-Y., Szydlowska, A., Kossiakoff, A. A., Hegde, R. S., and Keenan, R. J. (2015). Structure of the Get3 targeting factor in complex with its membrane protein cargo. Science 347, 1152–1155. McDonold, C. M., and Fromme, J. C. (2014). Four GTPases Differentially Regulate the Sec7 Arf-GEF to Direct Traffic at the trans-Golgi Network. Dev. Cell 30, 759–767. Mcgough, I., and Cullen, P. (2011). Recent Advances in Retromer Biology. Traffic 12, 963–971. Mcmahon, H. T., and Boucrot, E. (2015). Membrane curvature at a glance. J. Cell Sci. J. Cell Sci. J. Cell Sci. 128, 1065–1070. McNew, J. A., Coe, J. G. s, Sogaard, M., Zemelman, B. V., Wimmer, C., Hong, W., and Sollner, T. H. (1998). Gos1p, a Saccharomyces cerevisiae SNARE protein involved in Golgi 108  transport. FEBS Lett. 435, 89–95. van Meer, G., Voelker, D. R., and Feigenson, G. W. (2008). Membrane lipids: where they are and how they behave. Nat. Rev. Mol. Cell Biol. 9, 112–124. Menssen, R., Schweiggert, J., Schreiner, J., Kusevic, D., Reuther, J., Braun, B., and Wolf, D. H. (2012). Exploring the topology of the Gid complex, the E3 ubiquitin ligase involved in catabolite-induced degradation of gluconeogenic enzymes. J. Biol. Chem. 287, 25602–25614. van der Merwe, G., Cooper, T. G., and Vuuren, V. (2015). Ammonia Regulates VID30 Expression and Vid30p Function Shifts Nitrogen Metabolism toward Glutamate Formation Especially when Saccharomyces cerevisiae Is Grown in Low Concentrations of Ammonia. J. Biol. Chem. 276, 28659–28666. Michaut, M., and Bader, G. D. (2012). Multiple genetic interaction experiments provide complementary information useful for gene function prediction. PLoS Comput. Biol. 8. Michaut, M., Baryshnikova, A., Costanzo, M., Myers, C. L., Andrews, B. J., Boone, C., and Bader, G. D. (2011). Protein complexes are central in the yeast genetic landscape. PLoS Comput. Biol. 7. Michnick, S. W., Ear, P. H., Landry, C., Malleshaiah, M. K., and Messier, V. (2010). A toolkit of protein-fragment complementation assays for studying and dissecting large-scale and dynamic protein-protein interactions in living cells., Elsevier Inc. Mim, C., and Unger, V. M. (2012). Membrane curvature and its generation by BAR domains. Trends Biochem. Sci. 37, 526–533. Mioka, T., Fujimura-Kamada, K., and Tanaka, K. (2014). Asymmetric distribution of phosphatidylserine is generated in the absence of phospholipid flippases in Saccharomyces cerevisiae. Microbiologyopen 3, 803–821. Montigny, C., Lyons, J., Champeil, P., Nissen, P., and Lenoir, G. (2015). On the molecular mechanism of flippase- and scramblase-mediated phospholipid transport. Biochim. Biophys. Acta - Mol. Cell Biol. Lipids 1861, 767–783. Munro, S. (2001). What can yeast tell us about N-linked glycosylation in the Golgi apparatus ? FEBS Lett. 498, 223–227. Muthusamy, B. P., Natarajan, P., Zhou, X., and Graham, T. R. (2009). Linking phospholipid flippases to vesicle-mediated protein transport. Biochim. Biophys. Acta 1791, 612–619. De Nadal, E., Clotet, J., Posas, F., Serrano, R., Gomez, N., and Arino, J. (1998). The yeast halotolerance determinant Hal3p is an inhibitory subunit of the Ppz1p Ser  ͞Thr protein phosphatase. Proc. Natl. Acad. Sci. U. S. A. 95, 7357–7362. Nagalakshmi, U., Wang, Z., Waern, K., Shou, C., Raha, D., Gerstein, M., and Snyder, M. (2008). The Transcriptional Landscape of the Yeast Genome Define by RNA Sequencing. Science (80-. ). 320, 1344–1349. Nakano, K., Yamamoto, T., Kishimoto, T., Noji, T., and Tanaka, K. (2008). Protein Kinases Fpk1p and Fpk2p are Novel Regulators of Phospholipid Asymmetry. Mol. Biol. Cell 19, 109  1783–1797. Nakatsu, F., and Ohno, H. (2003). Adaptor Protein Complexes as the Key Regulators of Protein Sorting in the Post-Golgi Network Identification of AP complexes Recognition of sorting signals by AP complexes. Cell Struct. Funct. 28, 419–429. Nakayama, K., and Wakatsuki, S. (2003). The structure and function of GGAs, the traffic controllers at the TGN sorting crossroads. Cell Struct. Funct. 28, 431–442. Natarajan, P., Wang, J., Hua, Z., and Graham, T. R. (2004). Drs2p-coupled aminophospholipid translocase activity in yeast Golgi membranes and relationship to in vivo function. Proc. Natl. Acad. Sci. U. S. A. 101, 10614–10619. Nazarian, R., Falcón-Pérez, J. M., and Dell’Angelica, E. C. (2003). Biogenesis of lysosome-related organelles complex 3 (BLOC-3): a complex containing the Hermansky-Pudlak syndrome (HPS) proteins HPS1 and HPS4. Proc. Natl. Acad. Sci. U. S. A. 100, 8770–8775. Neklesa, T. K., and Davis, R. W. (2009). A genome-wide screen for regulators of TORC1 in response to amino acid starvation reveals a conserved Npr2/3 complex. PLoS Genet. 5, e1000515. Newell-litwa, K., Salazar, G., Smith, Y., and Faundez, V. (2009). Roles of BLOC-1 and Adaptor Protein-3 Complexes in Cargo Sorting to Synaptic Vesicles. 20, 1441–1453. Nikko, E., Sullivan, J. a, and Pelham, H. R. B. (2008). Arrestin-like proteins mediate ubiquitination and endocytosis of the yeast metal transporter Smf1. EMBO Rep. 9, 1216–1221. Nothwehr, S. F., Bruinsma, P., and Strawn, L. A. (1999). Distinct domains within Vps35p mediate the retrieval of two different cargo proteins from the yeast prevacuolar/endosomal compartment. Mol. Biol. Cell 10, 875–890. Nothwehr, S. F., Roberts, C. J., and Stevens, T. H. (1993). Membrane protein retention in the yeast Golgi apparatus: Dipeptidyl aminopeptidase A is retained by a cytoplasmic signal containing aromatic residues. J. Cell Biol. 121, 1197–1209. Obara, K., Sekito, T., and Ohsumi, Y. (2006). Assortment of Phosphatidylinositol 3-Kinase Complexes-Atg14p Directs Association of Complex I to the Pre-autophagosomal Structure in Saccharomyces cerevisiae. Mol. Biol. Cell 17, 1527–1539. Odorizzi, G., Babst, M., and Emr, S. D. (2000). Phosphoinositide signaling and the regulation of membrane trafficking in yeast. Trends Biochem. Sci. 25, 229–235. Odorizzi, G., Cowles, C. R., and Emr, S. D. (1998). The AP-3 complex: a coat of many colours. Trends Cell Biol. 8, 282–288. Orlean, P., and Menon, A. K. (2007). Thematic review series: lipid posttranslational modifications. GPI anchoring of protein in yeast and mammalian cells, or: how we learned to stop worrying and love glycophospholipids. J. Lipid Res. 48, 993–1011. Panchaud, N., Peli-Gulli, M.-P., and De Virgilio, C. (2013). SEACing the GAP that nEGOCiates TORC1 activation. Cell Cycle 12, 2948–2952. Panic, B., Whyte, J. R. C., and Munro, S. (2003). The ARF-like GTPases Arl1p and Arl3p 110  act in a pathway that interacts with vesicle-tethering factors at the Golgi apparatus. Curr. Biol. 13, 405–410. Di Paolo, G., and De Camilli, P. (2006). Phosphoinositides in cell regulation and membrane dynamics. Nature 443, 651–657. Park, J. S., and Neiman, A. M. (2012). VPS13 regulates membrane morphogenesis during sporulation in Saccharomyces cerevisiae. J. Cell Sci. 125, 3004–3011. Park, J. S., Okumura, Y., Tachikawa, H., and Neiman, A. M. (2013). SPO71 encodes a developmental stage-specific partner for Vps13 in Saccharomyces cerevisiae. Eukaryot. Cell 12, 1530–1537. Park, J. S., Thorsness, M. K., Policastro, R., McGoldrick, L. L., Hollingsworth, N. M., Thorsness, P. E., and Neiman, A. M. (2016). Yeast Vps13 promotes mitochondrial function and is localized at membrane contact sites. Mol. Biol. Cell 27, 2435–2449. Pashkova, N., Gakhar, L., Winistorfer, S. C., Sunshine, A. B., Rich, M., Dunham, M. J., Yu, L., and Piper, R. C. (2013). The Yeast Alix Homolog Bro1 Functions as a Ubiquitin Receptor for Protein Sorting into Multivesicular Endosomes. Dev. Cell 25, 520–533. van de Peppel, J., Kettelarij, N., van Bakel, H., Kockelkorn, T. T. J. P., van Leenen, D., and Holstege, F. C. P. (2005). Mediator expression profiling epistasis reveals a signal transduction pathway with antagonistic submodules and highly specific downstream targets. Mol. Cell 19, 511–522. Peter, A. T. J., Lachmann, J., Rana, M., Bunge, M., Cabrera, M., and Ungermann, C. (2013). The BLOC-1 complex promotes endosomal maturation by recruiting the Rab5 GTPase-activating protein Msb3. J. Cell Biol. 201, 97–111. Di Pietro, S. M., Falco, J. M., Setty, S. R. G., Marks, M. S., and Angelica, E. C. D. (2006). BLOC-1 Interacts with BLOC-2 and the AP-3 Complex to Facilitate Protein Trafficking on Endosomes □. Mol. Biol. Cell 17, 4027–4038. Piper, R. C., and Lehner, P. J. (2011). Endosomal transport via ubiquitination. Trends Cell Biol. 21, 647–655. Pitre, S. et al. (2006). PIPE: a protein-protein interaction prediction engine based on the re-occurring short polypeptide sequences between known interacting protein pairs. BMC Bioinformatics 7, 365. Pomorski, T., Lombardi, R., Riezman, H., Devaux, P. F., van Meer, G., and Holthuis, J. C. M. (2003). Drs2p-related P-type ATPases Dnf1p and Dnf2p are required for Phospholipid Translocation across the Yeast Plasma Membrane and Serve a Role in Endocytosis. Mol. Biol. Cell 14, 1240–1254. Poss, Z. C., Ebmeier, C. C., and Taatjes, D. J. (2013). The Mediator complex and transcription regulation. Crit. Rev. Biochem. Mol. Biol. 48, 575–608. Prezant, T. R., Chaltraw, W. E., and Fischel-Ghodsian, N. (1996). Identification of an overexpressed yeast gene which prevents aminoglycoside toxicity. Microbiology 142, 3407–3414. 111  Pu, J., Schindler, C., Jia, R., Jarnik, M., Backlund, P., and Bonifacino, J. S. (2015). BORC, a Multisubunit Complex that Regulates Lysosome Positioning. Dev. Cell 33, 176–188. Rana, M., Lachmann, J., and Ungermann, C. (2015). Identification of a Rab GAP cascade that controls recycling of the Rab5 GTPase Vps21 from the vacuole. Mol. Biol. Cell 26, 2535–2549. Raposo, G., and Marks, M. S. (2007). Melanosomes – dark organelles enlighten endosomal membrane transport. Nat. Rev. Mol. Cell Biol. 8, 786–797. Regelmann, J., Schule, T., Josupeit, F. S., Horak, J., Rose, M., Entian, K.-D., Thumm, M., and Wolf, D. H. (2003). Catabolite Degradation of Fructose-1,6-bisphosphatase in the Yeast Saccharomyces cerevisiae: A Genome-wide Screen Identifies Eight Novel GID Genes and Indicates the Existence of Two Degradation Pathways. Mol. Biol. Cell 14, 1652–1663. Reggiori, F., Black, M. W., and Pelham, H. R. (2000). Polar transmembrane domains target proteins to the interior of the yeast vacuole. Mol. Biol. Cell 11, 3737–3749. Richter, C., West, M., and Odorizzi, G. (2007). Dual mechanisms specify Doa4-mediated deubiquitination at multivesicular bodies. EMBO J. 26, 2454–2464. Riekhof, W. R., and Voelker, D. R. (2006). Uptake and utilization of lyso-phosphatidylethanolamine by Saccharomyces cerevisiae. J. Biol. Chem. 281, 36588–36596. Robinson, J. S., Klionsky, D. J., Banta, L. M., and Emr, S. D. (1988). Protein sorting in Saccharomyces cerevisiae: isolation of mutants defective in the delivery and processing of multiple vacuolar hydrolases. Mol. Cell. Biol. 8, 4936–4948. Rochette, S., Diss, G., Filteau, M., Leducq, J.-B., Dubé, A. K., and Landry, C. R. (2015). Genome-wide protein-protein interaction screening by protein-fragment complementation assay (PCA) in living cells. J. Vis. Exp., e52255. Roelants, F. M., Baltz, A. G., Trott, A. E., Fereres, S., and Thorner, J. (2010). A protein kinase network regulates the function of aminophospholipid flippases. Proc. Natl. Acad. Sci. U. S. A. 107, 34–39. Roelants, F. M., Su, B. M., von Wulffen, J., Ramachandran, S., Sartorel, E., Trott,  a. E., and Thorner, J. (2015). Protein kinase Gin4 negatively regulates flippase function and controls plasma membrane asymmetry. J. Cell Biol. 208, 299–311. Rothman, J. H., Howald, I., and Stevens, T. H. (1989). Characterization of genes required for protein sorting and vacuolar function in the yeast Saccharomyces cerevisiae. EMBO J. 8, 2057–2065. Rubel, C. E., Schisler, J. C., Hamlett, E. D., DeKroon, R. M., Gautel, M., Alzate, O., and Patterson, C. (2013). Diggin’ on U(biquitin): A Novel Method for the Identification of Physiological E3 Ubiquitin Ligase Substrates. Cell Biochem. Biophys. 67, 127–138. Russell, M. R. G., Shideler, T., Nickerson, D. P., West, M., and Odorizzi, G. (2012). Class E compartments form in response to ESCRT dysfunction in yeast due to hyperactivity of the Vps21 Rab GTPase. J. Cell Sci. 125, 5208–5220. Ryan, C. J. et al. (2012). Hierarchical Modularity and the Evolution of Genetic Interactomes 112  across Species. Mol. Cell 46, 691–704. Ryder, P. V, and Faundez, V. (2009). Schizophrenia: the “BLOC” may be in the endosomes. Sci. Signal. 2, pe66. Saito, K., Fujimura-Kamada, K., Nobumichi, F., Kato, U., Umeda, M., and Tanaka, K. (2004). Cdc50p, a Protein Required for Polarized Growth, Associates with the Drs2p P-type ATPase Implicated in Phospholipid Translocation in Saccharomyces cerevisiae. Mol. Biol. Cell 15, 3418–3432. Salazar, G., Craige, B., Styers, M. L., Doucette, M. M., Wainer, B. H., Angelica, E. C. D., Peden, A. A., Werner, E., and Faundez, V. (2006). BLOC-1 Complex Deficiency Alters the Targeting of Adaptor Protein Complex-3 Cargoes. 17, 4014–4026. Santt, O., Pfirrmann, T., Braun, B., Juretschke, J., Kimmig, P., Scheel, H., Hofmann, K., Thumm, M., and Wolf, D. H. (2008). The Yeast GID Complex , a Novel Ubiquitin Ligase ( E3 ) Involved in the Regulation of Carbohydrate Metabolism. Mol. Biol. Cell 19, 3323–3333. Schluter, C., Lam, K. K. Y., Brumm, J., Wu, B. W., Saunders, M., Stevens, T. H., Bryan, J., and Conibear, E. (2008). Global Analysis of Yeast Endosomal Transport Identifies the Vps55 / 68 Sorting Complex. Mol. Biol. Cell 19, 1282–1294. Schmidt, O., and Teis, D. (2012). The ESCRT machinery. Curr. Biol. 22, R116-20. Scholz, O., Thiel, A., Hillen, W., and Niederweis, M. (2000). Quantitative analysis of gene expression with an improved green fluorescent protein. p6. Eur. J. Biochem. 267, 1565–1570. Seaman, M. N. J. (2012). The retromer complex - endosomal protein recycling and beyond. J. Cell Sci. 125, 4693–4702. Seaman, M. N. J., Harbour, M. E., Tattersall, D., Read, E., and 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, 2371–2382. Seaman, M. N. J., McCaffery, J. M., and Emr, S. D. (1998). A membrane coat complex essential for endosome-to-Golgi retrograde transport in yeast. J. Cell Biol. 142, 665–681. Seaman, M. N. J., and Williams, H. P. (2002). Identification of Functional Domains of Yeast Sorting Nexins Vps5p and Vps17p. Mol. Biol. Cell 13, 2826–2840. Sebastian, T. T., Baldridge, R. D., Xu, P., and Graham, T. R. (2012). Phospholipid flippases: building asymmetric membranes and transport vesicles. Biochim. Biophys. Acta 1821, 1068–1077. Sen, A., Madhivanan, K., Mukherjee, D., and Aguilar, R. C. (2012). The epsin protein family: coordinators of endocytosis and signaling. Biomol. Concepts 3, 117–126. Setty, S. R. G. et al. (2007). BLOC-1 Is Required for Cargo-specific Sorting from Vacuolar Early Endosomes toward Lysosome-related Organelles. Mol. Biol. Cell 18, 768–780. Sheetz, M. P., and Singer, S. J. (1974). Biological Membranes as Bilayer Couples. A Molecular Mechanism of Drug-Erythrocyte Interactions. Proc. Natl. Acad. Sci. 71, 4457–4461. 113  Sheff, M. A., and Thorn, K. S. (2004). Optimized cassettes for fluorescent protein tagging in Saccharomyces cerevisiae. Yeast 21, 661–670. Shibasaki, S., Sakata, K., Ishii, J., Kondo, A., and Ueda, M. (2008). Development of a yeast protein fragment complementation assay (PCA) system using dihydrofolate reductase (DHFR) with specific additives. Appl. Microbiol. Biotechnol. 80, 735–743. Shields, S. B., and Piper, R. C. (2011). How ubiquitin functions with ESCRTs. Traffic 12, 1306–1317. Shin, H. W., Takatsu, H., and Nakayama, K. (2012). Mechanisms of membrane curvature generation in membrane traffic. Membranes (Basel). 2, 118–133. Singer-Krüger, B., Lasić, M., Bürger, A.-M., Hausser, A., Pipkorn, R., and Wang, Y. (2008). Yeast and human Ysl2p/hMon2 interact with Gga adaptors and mediate their subcellular distribution. EMBO J. 27, 1423–1435. Sitaram, A. et al. (2012). Differential recognition of a dileucine-based sorting signal by AP-1 and AP-3 reveals a requirement for both BLOC-1 and AP-3 in delivery of OCA2 to melanosomes. Mol. Biol. Cell 23, 3178–3192. Slessareva, J. E., Routt, S. M., Temple, B., Bankaitis, V. A., and Dohlman, H. G. (2006). Activation of the Phosphatidylinositol 3-Kinase Vps34 by a G Protein alpha Subunit at the Endosome. Cell 126, 191–203. Slubowski, C. J., Funk, A. D., Roesner, J. M., Paulissen, S. M., and Huang, L. S. (2015). HHS Public Access. Yeast 32, 379–387. Snowdon, C., Hlynialuk, C., and Van Der Merwe, G. (2008). Components of the Vid30c are needed for the rapamycin-induced degradation of the high-affinity hexose transporter Hxt7p in Saccharomyces cerevisiae. FEMS Yeast Res. 8, 204–216. Snowdon, C., and van der Merwe, G. (2012). Regulation of Hxt3 and Hxt7 turnover converges on the Vid30 complex and requires inactivation of the Ras/cAMP/PKA pathway in Saccharomyces cerevisiae. PLoS One 7, e50458. Stahelin, R. V, Scott, J. L., and Frick, C. T. (2014). Cellular and Molecular Interaction of Phosphoinositides and Peripheral Proteins. Chem Phys Lipids 182, 3–18. Stanley, P. (2011). Golgi glycosylation. Cold Spring Harb. Perspect. Biol. 3, 1–13. Stenmark, H. (2009). Rab GTPases as coordinators of vesicle traffic. Nat. Rev. Mol. Cell Biol. 10, 513–525. Strochlic, T. I., Setty, T. G., Sitaram, A., and Burd, C. G. (2007). Grd19/Snx3p functions as a cargo-specific adapter for retromer-dependent endocytic recycling. J. Cell Biol. 177, 115–125. Sullivan, J. A., Lewis, M. J., Nikko, E., and Pelham, H. R. B. (2007). Multiple Interactions Drive Adaptor-Mediated Recruitment of the Ubiquitin Ligase Rsp5 to Membrane Proteins In Vivo and In Vitro. Mol. Biol. Cell 18, 2429–2440. Supek, F., Supekova, L., Nelson, H., and Nelson, N. (1996). A yeast manganese transporter related to the macrophage protein involved in conferring resistance to mycobacteria. Proc. 114  Natl. Acad. Sci. U. S. A. 93, 5105–5110. Suvorova, E. S., Duden, R., and Lupashin, V. V. (2002). The Sec34/Sec35p complex, a Ypt1p effector required for retrograde intra-Golgi trafficking, interacts with Golgi SNAREs and COPI vesicle coat proteins. J. Cell Biol. 157, 631–643. Sztul, E., and Lupashin, V. (2009). Role of vesicle tethering factors in the ER-Golgi membrane traffic. FEBS Lett. 583, 3770–3783. Szul, T., and Sztul, E. (2011). COPII and COPI traffic at the ER-Golgi interface. Physiology (Bethesda). 26, 348–364. Takagi, K., Iwamoto, K., Kobayashi, S., Horiuchi, H., Fukuda, R., and Ohta, A. (2012). Involvement of Golgi-associated retrograde protein complex in the recycling of the putative Dnf aminophospholipid flippases in yeast. Biochem. Biophys. Res. Commun. 417, 490–494. Takahashi, Y., Fujimura-Kamada, K., Kondo, S., and Tanaka, K. (2011). Isolation and characterization of novel mutations in CDC50, the non-catalytic subunit of the Drs2p phospholipid flippase. J. Biochem. 149, 423–432. Takar, M., Wu, Y., and Graham, T. R. (2016). The essential Neo1 from budding yeast plays a role in establishing aminophospholipid asymmetry of the plasma membrane. J. Biol. Chem. 291, 15727–15739. Takatsu, H., Baba, K., Shima, T., Umino, H., Kato, U., Umeda, M., Nakayama, K., and Shin, H. W. (2011). ATP9B, a P4-ATPase (a putative aminophospholipid translocase), localizes to the trans-Golgi network in a CDC50 protein-independent manner. J. Biol. Chem. 286, 38159–38167. Takeda, M., Yamagami, K., and Tanaka, K. (2014). Role of phosphatidylserine in phospholipid flippase-mediated vesicle transport in Saccharomyces cerevisiae. Eukaryot. Cell 13, 363–375. Tang, S., Buchkovich, N. J., Henne, W. M., Banjade, S., Kim, Y. J., and Emr, S. D. (2016). ESCRT-III activation by parallel action of ESCRT-I/II and ESCRT-0/Bro1 during MVB biogenesis. Elife 5, e15507. Tarassov, K., Messier, V., Landry, C. R., Radinovic, S., Serna Molina, M. M., Shames, I., Malitskaya, Y., Vogel, J., Bussey, H., and Michnick, S. W. (2008). An in vivo map of the yeast protein interactome. Science 320, 1465–1470. Teis, D., Saksena, S., and Emr, S. D. (2008). Ordered Assembly of the ESCRT-III Complex on Endosomes Is Required to Sequester Cargo during MVB Formation. Dev. Cell 15, 578–589. Temkin, P., Lauffer, B., Jager, S., Cimermancic, P., Krogan, N. J., and von Zastrow, M. (2011). Snx27 mediates retromer tubule entry and endosome-to-plasma membrane trafficking of signaling receptors. Nat. Cell Biol. 13, 715–721. Tokarev, A. A., Taussig, D., Sundaram, G., Lipatova, Z., Liang, Y., Mulholland, J. W., and Segev, N. (2009). TRAPPII Complex Assembly Requires Trs33 or Trs65. 10, 1831–1844. Tong, A. H. Y., and Boone, C. (2006). Synthetic genetic array analysis in Saccharomyces 115  cerevisiae. In: Yeast Protocols, ed. W. Xiao, Clifton, N.J.: Humana Press, 171–192. Truschel, S. T. et al. (2009). ESCRT-1 function is required for Tyrp1 transport from early endosomes to the melanosome limiting membrane. Traffic 10, 1318–1336. Tsai, P.-C., Hsu, J.-W., Liu, Y.-W., Chen, K.-Y., and Lee, F.-J. S. (2013). Arl1p regulates spatial membrane organization at the trans-Golgi network through interaction with Arf-GEF Gea2p and flippase Drs2p. Proc. Natl. Acad. Sci. U. S. A. 110, E668-77. Usaj, M. M., Styles, E. B., Verster, A. J., Friesen, H., Boone, C., and Andrews, B. J. (2016). High-Content Screening for Quantitative Cell Biology. Trends Cell Biol. 26, 598–611. Vaibhava, V., Nagabhushana,  a., Chalasani, M. L. S., Sudhakar, C., Kumari,  a., and Swarup, G. (2012). Optineurin mediates negative regulation of Rab8 function by TBC1D17, a GTPase activating protein. J. Cell Sci. 8, 5026–5039. Valdivia, R., Baggott, D., Chuang, J., and Schekman, R. (2002). The Yeast Clathrin Adaptor Protein Complex 1 Is Required for the Efficient Retention of a Subset of Late Golgi Membrane Proteins. Dev. Cell 2, 283–294. Varshavsky A (1997). The ubiquitin system. Trends Biochem. Sci. 22, 383–387. Veraksa, A., Bauer, A., and Artavanis-Tsakonas, S. (2005). Analyzing protein complexes in Drosophila with tandem affinity purification-mass spectrometry. Dev. Dyn. 232, 827–834. Vilariño-Güell, C. et al. (2011). VPS35 Mutations in Parkinson Disease. Am. J. Hum. Genet. 89, 162–167. Voos, W., and Stevens, T. H. (1998). Retrieval of resident late-Golgi membrane proteins from the prevacuolar compartment of Saccharomyces cerevisiae is dependent on the function of Grd19p. J. Cell Biol. 140, 577–590. Wang, C. W., Hamamoto, S., Orci, L., and Schekman, R. (2006). Exomer: A coat complex for transport of select membrane proteins from the trans-Golgi network to the plasma membrane in yeast. J. Cell Biol. 174, 973–983. van Weering, J. R. T., Sessions, R. B., Traer, C. J., Kloer, D. P., Bhatia, V. K., Stamou, D., Carlsson, S. R., Hurley, J. H., and Cullen, P. J. (2012). Molecular basis for SNX-BAR-mediated assembly of distinct endosomal sorting tubules. EMBO J. 31, 4466–4480. Wernimont, A. K., and Weissenhorn, W. (2004). Crystal structure of subunit VPS25 of the endosomal trafficking complex ESCRT-II. BMC Struct. Biol. 4, 10. Whitfield, S. T., Burston, H. E., Bean, B. D. M., Raghuram, N., Maldonado-Baez, L., Davey, M., Wendland, B., and Conibear, E. (2015). The alternate AP-1 adaptor subunit Apm2 interacts with the Mil1 regulatory protein and confers differential cargo sorting. Mol. Biol. Cell 1, 1–30. Wicky, S., Schwarz, H., and Singer-Krüger, B. (2004). Molecular interactions of yeast Neo1p, an essential member of the Drs2 family of aminophospholipid translocases, and its role in membrane trafficking within the endomembrane system. Mol. Cell. Biol. 24, 7402–7418. Wiederkehr, A., Avaro, S., Prescianotto-baschong, C., Riezman, H., and Haguenauer-tsapis, 116  R. (2000). Traffic and The F-box Protein Rcylp Is Involved in Endocytic Membrane cerevisiae Recycling Out of an Early Endosome in Saccharomyces cerevisiae. J. Biol. Chem. 149, 397–410. Wilcox, C. A., Redding, K., Wright, R., and Fuller, R. S. (1992). Mutation of a tyrosine localization signal in the cytosolic tail of yeast Kex2 protease disrupts Golgi retention and results in default transport to the vacuole. Mol. Biol. Cell 3, 1353–1371. Williams, R. L., and Urbé, S. (2007). The emerging shape of the ESCRT machinery. Nat. Rev. Mol. Cell Biol. 8, 355–368. Wu, Y., Takar, M., Cuentas-Condori, A. A., and Graham, T. R. (2016). Neo1 and phosphatidylethanolamine contribute to vacuole membrane fusion in Saccharomyces cerevisiae. Cell. Logist. 6, e1228791. Xu, C., and Ng, D. T. W. (2015). O-mannosylation: The other glycan player of ER quality control. Semin. Cell Dev. Biol., 1–6. Xu, P., Baldridge, R. D., Chi, R. J., Burd, C. G., and Graham, T. R. (2013). Phosphatidylserine flipping enhances membrane curvature and negative charge required for vesicular transport. J. Cell Biol. 202, 875–886. Yamagami, K., Yamamoto, T., Sakai, S., Mioka, T., Sano, T., Igarashi, Y., and Tanaka, K. (2015). Inositol Depletion Restores Vesicle Transport in Yeast Phospholipid Flippase Mutants. PLoS One 10, e0120108. Yang, Q. et al. (2012). The BLOS1-interacting protein KXD1 is involved in the biogenesis of lysosome-related organelles. Traffic 13, 1160–1169. Yenush, L., Jose, M. M., Arino, J., and Serrano, R. (2002). The Ppz protein phosphatases are key regulators of K + and pH homeostasis : implications for salt tolerance , cell wall integrity and cell cycle progression. EMBO J. 21, 920–929. Zavodszky, E., Seaman, M. N. J., Moreau, K., Jimenez-Sanchez, M., Breusegem, S. Y., Harbour, M. E., and Rubinsztein, D. C. (2014). Mutation in VPS35 associated with Parkinson’s disease impairs WASH complex association and inhibits autophagy. Nat. Commun. 5, 3828. Zhang, A., He, X., Zhang, L., Yang, L., Woodman, P., and Li, W. (2014). Biogenesis of Lysosome-related Organelles Complex-1 Subunit 1 (BLOS1) Interacts with Sorting Nexin 2 and the Endosomal Sorting Complex Required for Transport-I (ESCRT-I) Component TSG101 to Mediate the Sorting of Epidermal Growth Factor Receptor into Endo. J. Biol. Chem. 289, 29180–29194. Zhou, X., and Graham, T. R. (2009). Reconstitution of phospholipid translocase activity with purified Drs2p, a type-IV P-type ATPase from budding yeast. Proc. Natl. Acad. Sci. U. S. A. 106, 16586–16591. Zhou, X., Sebastian, T. T., and Graham, T. R. (2013). Auto-inhibition of Drs2p, a yeast phospholipid flippase, by its carboxyl-terminal tail. J. Biol. Chem. 288, 31807–31815. Zimmermann, J., Chidambaram, S., and Fischer von Mollard, G. (2010). Dissecting Ent3p: 117  the ENTH domain binds different SNAREs via distinct amino acid residues while the C-terminus is sufficient for retrograde transport from endosomes. Biochem. J. 431, 123–134. 118  Appendices  Appendix A  : Supplementary Material for Chapter 4  Figure A.1: Some mutants cause a relocalization of Neo1 to the Golgi.  A) Live-cell fluorescent microscopy of ADHpr-GFP-Neo1 with Sec7-dsRed was repeated and colocalization was analyzed.  One-Way ANOVA with a posthoc test with Dunnett correction:  *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Error bars report standard error of the mean n=3. B) Fluorescence microscopy of pGFP-Neo1 (pLD27) in wild-type compared to mutants with the endosomal dye FM4-64.  All strains scaled the same. Scale bar represents 2µM.   119   Figure A.2: T-Coffee alignment. Compares the sequence of the first 195 amino acids of Neo1 in Saccharomyces Cerevisiae with other closely related yeasts.  Colors defined by ClustalX coloring.  120   Figure A.3: The Snx3 sorting signal is contained in the first 195 amino acids of Neo1. A) Live-cell fluorescent microscopy of BY4741 and snx3∆ with different Neo1-ALP chimeras.  Representative images are shown and scaled the same. Scale bar 2µM. B) Live-cell fluorescence microscopy of BY4741 and snx3∆ with different Neo1-ALP chimeras. left) Schematic representation of each chimera right) Quantitation of >200 cell measuring cells containing vacuolar rim staining. n=1 C) Live-cell fluorescence microscopy of BY4741 and 121  snx3Δ with 1-195 Neo1-ALP and 1-140 Neo1-ALP chimeras +/- a mutation changing 65FEM to AAA showing the mutation causes a disruption in the 1-140 Neo1-ALP chimera  i) ) Representative images scaled the same.  ii) Quantitation of cells containing vacuolar rim staining. n=3 Statistics: One-Way ANOVA with a posthoc test using Dunnett correction:  *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.  Error bars report standard error of the mean.    Figure A.4: Growth assay to assess viability of Neo1 N-terminal tail truncations. Yeast containing Galpr-Neo1 + either empty or Neo1 containing plasmids were spotted in 10x dilution series on synthetic medium without uracil and either dextrose which prevents expression of genomic Neo1 or galactose and grown at 30°C or 37°C.  Images were taken after 36 hours growth.   122  Table A.1: Genome-wide invertase Z Score. Low numbers represent more cell surface GSS localization. High numbers represent less cell surface localization. ORF  Gene Name Brief Description MAT α  Invertase Z Score MAT a  Invertase Z Score Average Mat a/α Invertase Z Score YAL002W VPS8 Membrane-binding component of the CORVET complex 1.11 0.88 0.99 YAL004W  Dubious open reading frame 0.03 -0.42 -0.20 YAL005C SSA1 ATPase involved in protein folding and NLS-directed nuclear transport -0.27 -0.66 -0.47 YAL007C ERP2 Member of the p24 family involved in ER to Golgi transport -0.08 -0.22 -0.15 YAL008W FUN14 Integral mitochondrial outer membrane (MOM) protein 0.01 0.48 0.25 YAL009W SPO7 Putative regulatory subunit of Nem1p-Spo7p phosphatase holoenzyme 1.67 0.09 0.88 YAL010C MDM10 Subunit of both the ERMES and the SAM complex  -0.82 -0.82 YAL011W SWC3 Protein of unknown function 1.20 1.80 1.50 YAL012W CYS3 Cystathionine gamma-lyase 0.24 -0.35 -0.06 YAL013W DEP1 Component of the Rpd3L histone deacetylase complex  -2.77 -2.77 YAL014C SYN8 Endosomal SNARE related to mammalian syntaxin 8 -0.22 0.30 0.04 YAL015C NTG1 DNA N-glycosylase and apurinic/apyrimidinic (AP) lyase 0.62 -0.05 0.29 YAL016W TPD3 Regulatory subunit A of the heterotrimeric PP2A complex  -0.83 -0.83 YAL017W PSK1 PAS domain-containing serine/threonine protein kinase 0.08 -0.27 -0.10 YAL018C LDS1 Protein Involved in spore wall assembly -0.53 0.41 -0.06 YAL019W FUN30 Snf2p family member with ATP-dependent chromatin remodeling activity -0.01 0.77 0.38 YAL020C ATS1 Protein required for modification of wobble nucleosides in tRNA -0.01 -0.73 -0.37 YAL021C CCR4 Component of the CCR4-NOT transcriptional complex  -0.12 -0.12 YAL022C FUN26 High affinity, broad selectivity, nucleoside/nucleobase transporter 0.27 0.04 0.15 YAL023C PMT2 Protein O-mannosyltransferase of the ER membrane -0.09 0.03 -0.03 YAL024C LTE1 Protein similar to GDP/GTP exchange factors -0.38 -0.05 -0.21 YAL026C DRS2 Trans-Golgi network aminophospholipid translocase (flippase) 2.71 -1.75 0.48 YAL027W SAW1 5'- and 3'-flap DNA binding protein 0.40 0.30 0.35 YAL028W FRT2 Tail-anchored ER membrane protein of unknown function -0.34 0.22 -0.06 YAL029C MYO4 Type V myosin motor involved in actin-based transport of cargos 0.43 0.44 0.44 YAL030W SNC1 Vesicle membrane receptor protein (v-SNARE) 2.19 1.66 1.93 YAL031C GIP4 Cytoplasmic protein that regulates protein phosphatase 1 Glc7p 0.16 0.18 0.17 YAL034C FUN19 Non-essential protein of unknown function 0.12 -0.08 0.02 YAL035W FUN12 Translation initiation factor eIF5B -0.27 -1.22 -0.75 YAL036C RBG1 Member of the DRG family of GTP-binding proteins 0.03 -0.37 -0.17 YAL037W  Putative protein of unknown function 0.89 0.14 0.52 YAL039C CYC3 Cytochrome c heme lyase (holocytochrome c synthase) 2.98 0.32 1.65 YAL040C CLN3 G1 cyclin involved in cell cycle progression 0.79 0.49 0.64 YAL042W ERV46 Protein localized to COPII-coated vesicles 0.46 0.34 0.40 YAL043C-a  #N/A 0.36 0.56 0.46 YAL044C GCV3 H subunit of the mitochondrial glycine decarboxylase complex -0.63 1.40 0.38 YAL045C  Dubious open reading frame 0.53 0.07 0.30 YAL046C BOL3 Protein involved in Fe-S cluster transfer to mitochondrial clients 0.28 -0.24 0.02 YAL047C SPC72 Gamma-tubulin small complex (gamma-TuSC) receptor  -2.10 -2.10 123  ORF  Gene Name Brief Description MAT α  Invertase Z Score MAT a  Invertase Z Score Average Mat a/α Invertase Z Score YAL048C GEM1 Outer mitochondrial membrane GTPase, subunit of the ERMES complex 0.58 0.33 0.46 YAL049C AIM2 Cytoplasmic protein involved in mitochondrial function or organization 0.43 0.13 0.28 YAL051W OAF1 Oleate-activated transcription factor -0.13 -0.19 -0.16 YAL053W FLC2 Putative calcium channel involved in calcium release under hypotonic -0.39 -1.33 -0.86 YAL054C ACS1 Acetyl-coA synthetase isoform 0.44 -0.01 0.22 YAL055W PEX22 Putative peroxisomal membrane protein -1.52 -0.86 -1.19 YAL056W GPB2 Multistep regulator of cAMP-PKA signaling -0.43 0.18 -0.12 YAL058C-A  #N/A 0.40 0.05 0.22 YAL058W CNE1 Calnexin -0.60 -0.23 -0.41 YAL059W ECM1 Pre-ribosomal factor involved in 60S ribosomal protein subunit export -0.37 -0.09 -0.23 YAL060W BDH1 NAD-dependent (R,R)-butanediol dehydrogenase -0.67 -0.08 -0.37 YAL061W BDH2 Putative medium-chain alcohol dehydrogenase with similarity to BDH1 -0.57 -0.43 -0.50 YAL062W GDH3 NADP(+)-dependent glutamate dehydrogenase -0.67 -0.40 -0.54 YAL064C-A TDA8 Putative protein of unknown function 0.33 0.02 0.17 YAL065C  Putative protein of unknown function -0.31 -0.42 -0.36 YAL066W  Dubious open reading frame 0.33 -0.54 -0.11 YAL067C SEO1 Putative permease -0.14 -0.57 -0.35 YAL068C PAU8 Protein of unknown function -1.46 -1.95 -1.71 YAR002C-A ERP1 Member of the p24 family involved in ER to Golgi transport 0.04 0.29 0.16 YAR002W NUP60 FG-nucleoporin component of central core of the nuclear pore complex 0.06 0.17 0.12 YAR003W SWD1 Subunit of the COMPASS (Set1C) complex 0.52 -0.71 -0.09 YAR014C BUD14 Protein involved in bud-site selection 0.99 0.13 0.56 YAR015W ADE1 N-succinyl-5-aminoimidazole-4-carboxamide ribotide synthetase -0.15 0.26 0.06 YAR018C KIN3 Nonessential serine/threonine protein kinase 0.35 0.14 0.25 YAR020C PAU7 Member of the seripauperin multigene family 0.07 0.62 0.35 YAR023C  Putative integral membrane protein 0.03 0.74 0.39 YAR027W UIP3 Putative integral membrane protein of unknown function 0.05 0.63 0.34 YAR028W  Putative integral membrane protein 0.28 0.10 0.19 YAR029W  Member of DUP240 gene family but contains no transmembrane domains 0.23 -0.57 -0.17 YAR030C  Dubious open reading frame -0.13 0.39 0.13 YAR031W PRM9 Pheromone-regulated protein -0.10 -0.39 -0.24 YAR035W YAT1 Outer mitochondrial carnitine acetyltransferase 0.86 0.44 0.65 YAR037W  #N/A 0.75 0.51 0.63 YAR040C  #N/A 0.68 0.24 0.46 YAR042W SWH1 Protein similar to mammalian oxysterol-binding protein 0.26 0.37 0.32 YAR043C-a  #N/A    YAR044W  Merged open reading frame 0.03 0.01 0.02 YAR047C  Dubious open reading frame 0.17 0.09 0.13 YAR050W FLO1 Lectin-like protein involved in flocculation 0.01 0.03 0.02 YBL001C ECM15 Non-essential protein of unknown function 0.14 0.20 0.17 YBL002W HTB2 Histone H2B 0.27 1.23 0.75 YBL003C HTA2 Histone H2A 0.02 -0.30 -0.14 124  ORF  Gene Name Brief Description MAT α  Invertase Z Score MAT a  Invertase Z Score Average Mat a/α Invertase Z Score YBL005W PDR3 Transcriptional activator of the pleiotropic drug resistance network 0.35 0.34 0.35 YBL006C LDB7 Component of the RSC chromatin remodeling complex  -1.35 -1.35 YBL007C SLA1 Cytoskeletal protein binding protein -2.65 -2.63 -2.64 YBL008W HIR1 Subunit of the HIR complex 1.02 1.35 1.19 YBL009W ALK2 Protein kinase -0.43 0.04 -0.20 YBL010C  Putative protein of unknown function -0.25 -0.25 -0.25 YBL011W SCT1 Glycerol 3-phosphate/dihydroxyacetone phosphate sn-1 acyltransferase 0.43 0.56 0.50 YBL012C  Dubious open reading frame 0.01 0.33 0.17 YBL013W FMT1 Methionyl-tRNA formyltransferase 0.11 0.70 0.41 YBL015W ACH1 Protein with CoA transferase activity 2.65 2.40 2.52 YBL016W FUS3 Mitogen-activated serine/threonine protein kinase involved in mating -0.62 -1.06 -0.84 YBL017C PEP1 Type I transmembrane sorting receptor for multiple vacuolar hydrolases 0.00 0.56 0.28 YBL019W APN2 Class II abasic (AP) endonuclease involved in repair of DNA damage -0.05 -0.12 -0.08 YBL021C HAP3 Subunit of the Hap2p/3p/4p/5p CCAAT-binding complex 3.26 1.64 2.45 YBL022C PIM1 ATP-dependent Lon protease    YBL024W NCL1 S-adenosyl-L-methionine-dependent tRNA: m5C-methyltransferase 0.04 0.65 0.35 YBL025W RRN10 Protein involved in promoting high level transcription of rDNA -1.15 -0.15 -0.65 YBL027W RPL19B Ribosomal 60S subunit protein L19B -1.88 -1.74 -1.81 YBL028C  Protein of unknown function that may interact with ribosomes 0.36 0.59 0.48 YBL029W  Non-essential protein of unknown function 0.42 0.38 0.40 YBL031W SHE1 Mitotic spindle protein  0.84 0.84 YBL032W HEK2 RNA binding protein involved in asymmetric localization of ASH1 mRNA  0.36 0.36 YBL033C RIB1 GTP cyclohydrolase II  0.38 0.38 YBL036C  Putative non-specific single-domain racemase 0.09 0.23 0.16 YBL037W APL3 Alpha-adaptin 0.84 0.74 0.79 YBL038W MRPL16 Mitochondrial ribosomal protein of the large subunit 0.52 -0.15 0.19 YBL039C URA7 Major CTP synthase isozyme (see also URA8) 0.43 0.30 0.36 YBL042C FUI1 High affinity uridine permease, localizes to the plasma membrane 0.27 -0.12 0.07 YBL043W ECM13 Non-essential protein of unknown function -0.03 -0.53 -0.28 YBL044W  Putative protein of unknown function -1.00 -0.12 -0.56 YBL045C COR1 Core subunit of the ubiquinol-cytochrome c reductase complex 2.03 1.13 1.58 YBL046W PSY4 Regulatory subunit of protein phosphatase PP4 -0.11 -0.01 -0.06 YBL047C EDE1 Scaffold protein involved in the formation of early endocytic sites -2.07 -2.38 -2.23 YBL048W RRT1 Protein of unknown function 0.34 1.67 1.01 YBL049W MOH1 Protein of unknown function, essential for stationary phase survival 0.56 0.39 0.47 YBL051C PIN4 Protein involved in G2/M phase progression and response to DNA damage 0.05 -0.10 -0.02 YBL052C SAS3 Histone acetyltransferase catalytic subunit of NuA3 complex 0.11 0.32 0.22 125  ORF  Gene Name Brief Description MAT α  Invertase Z Score MAT a  Invertase Z Score Average Mat a/α Invertase Z Score YBL053W  Dubious open reading frame 0.03 0.30 0.17 YBL054W TOD6 PAC motif binding protein involved in rRNA and ribosome biogenesis -0.18 0.19 0.00 YBL055C  3'-->5' exonuclease and endonuclease with a possible role in apoptosis 0.12 0.07 0.10 YBL056W PTC3 Type 2C protein phosphatase (PP2C) -0.87 0.07 -0.40 YBL057C PTH2 One of two mitochondrially-localized peptidyl-tRNA hydrolases 0.00 0.55 0.27 YBL058W SHP1 UBX (ubiquitin regulatory X) domain-containing protein -1.47  -1.47 YBL059W  Putative protein of unknown function 0.49 0.65 0.57 YBL060W YEL1 Guanine nucleotide exchange factor specific for Arf3p -0.45 0.45 0.00 YBL061C SKT5 Activator of Chs3p (chitin synthase III) during vegetative growth -0.34 0.20 -0.07 YBL062W  Dubious open reading frame -0.55 0.07 -0.24 YBL063W KIP1 Kinesin-related motor protein -0.49 -0.05 -0.27 YBL064C PRX1 Mitochondrial peroxiredoxin with thioredoxin peroxidase activity -0.24 0.16 -0.04 YBL065W  Dubious open reading frame -0.34 0.19 -0.08 YBL066C SEF1 Putative transcription factor  0.05 0.05 YBL067C UBP13 Ubiquitin-specific protease that cleaves Ub-protein fusions -0.39 1.87 0.74 YBL068W PRS4 5-phospho-ribosyl-1(alpha)-pyrophosphate synthetase, synthesizes PRPP -0.08 -0.18 -0.13 YBL069W AST1 Lipid raft associated protein 0.36 0.11 0.23 YBL070C  Dubious open reading frame 0.21 0.36 0.28 YBL071C  Putative protein of unknown function 0.33 0.76 0.55 YBL072C RPS8A Protein component of the small (40S) ribosomal subunit 1.19 1.70 1.44 YBL075C SSA3 ATPase involved in protein folding and the response to stress -0.07 0.10 0.01 YBL078C ATG8 Component of autophagosomes and Cvt vesicles -0.26 -0.27 -0.26 YBL079W NUP170 Subunit of inner ring of nuclear pore complex (NPC) 0.85 -0.61 0.12 YBL080C PET112 Subunit of the trimeric GatFAB AmidoTransferase(AdT) complex 2.67 0.96 1.82 YBL081W  Non-essential protein of unknown function 0.24 0.07 0.16 YBL082C ALG3 Dolichol-P-Man dependent alpha(1-3) mannosyltransferase -1.32 -1.41 -1.37 YBL083C  Dubious open reading frame  -1.49 -1.49 YBL085W BOI1 Protein implicated in polar growth 0.13 0.39 0.26 YBL086C  Protein of unknown function 0.12 0.35 0.23 YBL087C RPL23A Ribosomal 60S subunit protein L23A 0.87 0.25 0.56 YBL088C TEL1 Protein kinase primarily involved in telomere length regulation -0.07 0.44 0.19 YBL089W AVT5 Putative transporter 0.39 1.23 0.81 YBL090W MRP21 Mitochondrial ribosomal protein of the small subunit 2.25 0.25 1.25 YBL091C MAP2 Methionine aminopeptidase -0.29 -0.32 -0.31 YBL091C-A SCS22 Protein involved in regulation of phospholipid metabolism 0.45 -0.10 0.17 YBL093C ROX3 Subunit of the RNA polymerase II mediator complex  0.88 0.88 YBL094C  Dubious open reading frame -0.21 -0.06 -0.13 YBL095W MRX3 Protein that associates with mitochondrial ribosome -0.33 -0.37 -0.35 YBL096C  Non-essential protein of unknown function -0.41 0.15 -0.13 YBL098W BNA4 Kynurenine 3-monooxygenase 0.20 0.47 0.34 YBL099W ATP1 Alpha subunit of the F1 sector of mitochondrial F1F0 ATP synthase 1.19 0.78 0.98 126  ORF  Gene Name Brief Description MAT α  Invertase Z Score MAT a  Invertase Z Score Average Mat a/α Invertase Z Score YBL100C  Dubious open reading frame -0.61  -0.61 YBL101C ECM21 Protein involved in regulating endocytosis of plasma membrane proteins -1.08 0.05 -0.52 YBL102W SFT2 Tetra-spanning membrane protein found mostly in the late Golgi 0.62 1.13 0.87 YBL103C RTG3 bHLH/Zip transcription factor for retrograde (RTG) and TOR pathways 2.05 2.05 2.05 YBL104C SEA4 Subunit of SEACAT, a subcomplex of the SEA complex -0.48 -0.07 -0.28 YBL106C SRO77 Protein with roles in exocytosis and cation homeostasis 0.06 0.26 0.16 YBL107C MIX23 Mitochondrial intermembrane space protein of unknown function -0.24 -0.68 -0.46 YBR001C NTH2 Putative neutral trehalase, required for thermotolerance -0.79 -0.20 -0.49 YBR003W COQ1 Hexaprenyl pyrophosphate synthetase 3.17 -0.42 1.38 YBR005W RCR1 Protein of the ER membrane involved in cell wall chitin deposition -0.88 0.20 -0.34 YBR006W UGA2 Succinate semialdehyde dehydrogenase -0.12 0.33 0.10 YBR007C DSF2 Deletion suppressor of mpt5 mutation 0.49 0.73 0.61 YBR008C FLR1 Plasma membrane transporter of the major facilitator superfamily 0.56 0.27 0.42 YBR009C HHF1 Histone H4 0.55 0.31 0.43 YBR010W HHT1 Histone H3 -0.03 0.28 0.13 YBR012C  Dubious open reading frame -0.03 0.57 0.27 YBR013C  Putative protein of unknown function 0.01 0.32 0.16 YBR014C GRX7 Cis-golgi localized monothiol glutaredoxin -0.13 -1.03 -0.58 YBR015C MNN2 Alpha-1,2-mannosyltransferase 0.37 -0.83 -0.23 YBR016W  Tail-anchored plasma membrane protein with a conserved CYSTM module -0.02 0.05 0.01 YBR018C GAL7 Galactose-1-phosphate uridyl transferase 0.20 -0.34 -0.07 YBR019C GAL10 UDP-glucose-4-epimerase -0.52 -1.14 -0.83 YBR020W GAL1 Galactokinase 0.91 -0.71 0.10 YBR021W FUR4 Plasma membrane localized uracil permease 0.34 -0.31 0.02 YBR022W POA1 Phosphatase that is highly specific for ADP-ribose 1''-phosphate 0.07 -1.06 -0.50 YBR023C CHS3 Chitin synthase III -0.21 -1.06 -0.64 YBR024W SCO2 Protein anchored to mitochondrial inner membrane 0.37 -0.58 -0.10 YBR025C OLA1 P-loop ATPase with similarity to human OLA1 and bacterial YchF -0.49 -0.16 -0.33 YBR026C ETR1 2-enoyl thioester reductase 2.94 -1.10 0.92 YBR027C  Putative protein of unknown function -0.12 0.16 0.02 YBR028C YPK3 AGC kinase -0.02 0.20 0.09 YBR030W RKM3 Ribosomal lysine methyltransferase 0.16 -0.27 -0.06 YBR031W RPL4A Ribosomal 60S subunit protein L4A 0.36 0.40 0.38 YBR032W  Putative protein of unknown function -0.51 0.76 0.12 YBR033W EDS1 Putative zinc cluster protein, predicted to be a transcription factor 0.13 0.42 0.28 YBR034C HMT1 Nuclear SAM-dependent mono- and asymmetric methyltransferase 1.01 1.28 1.14 YBR035C PDX3 Pyridoxine (pyridoxamine) phosphate oxidase  -0.81 -0.81 YBR036C CSG2 Endoplasmic reticulum membrane protein -1.33 -0.33 -0.83 YBR037C SCO1 Copper-binding protein of mitochondrial inner membrane 4.63 2.98 3.80 YBR039W ATP3 Gamma subunit of the F1 sector of mitochondrial F1F0 ATP synthase  -1.03 -1.03 127  ORF  Gene Name Brief Description MAT α  Invertase Z Score MAT a  Invertase Z Score Average Mat a/α Invertase Z Score YBR039W ATP3 Gamma subunit of the F1 sector of mitochondrial F1F0 ATP synthase  -1.03 -1.03 YBR040W FIG1 Integral membrane protein required for efficient mating 0.39 0.13 0.26 YBR041W FAT1 Very long chain fatty acyl-CoA synthetase and fatty acid transporter -0.09 0.02 -0.03 YBR042C CST26 Putative transferase involved in phospholipid biosynthesis 0.04 0.05 0.05 YBR043C QDR3 Multidrug transporter of the major facilitator superfamily -0.01 -0.42 -0.22 YBR044C TCM62 Protein involved in assembly of the succinate dehydrogenase complex 1.26 2.02 1.64 YBR045C GIP1 Meiosis-specific regulatory subunit of the Glc7p protein phosphatase 0.20 0.28 0.24 YBR046C ZTA1 NADPH-dependent quinone reductase -0.54 0.03 -0.25 YBR047W FMP23 Putative protein of unknown function -0.23 -0.02 -0.12 YBR048W RPS11B Protein component of the small (40S) ribosomal subunit 1.58 0.91 1.24 YBR050C REG2 Regulatory subunit of the Glc7p type-1 protein phosphatase 0.26 0.12 0.19 YBR051W  Dubious open reading frame 0.13 0.14 0.14 YBR052C RFS1 Protein of unknown function -0.19 0.00 -0.09 YBR053C  Putative protein of unknown function 0.06 0.35 0.20 YBR054W YRO2 Protein with a putative role in response to acid stress -0.34 -0.20 -0.27 YBR056W  Putative glycoside hydrolase of the mitochondrial intermembrane space 0.02 -0.03 -0.01 YBR057C MUM2 Protein essential for meiotic DNA replication and sporulation 0.49 0.05 0.27 YBR058C UBP14 Ubiquitin-specific protease 0.42 -0.22 0.10 YBR059C AKL1 Ser-Thr protein kinase -0.64 -0.91 -0.77 YBR061C TRM7 2'-O-ribose methyltransferase -0.74 2.42 0.84 YBR062C  Protein of unknown function that interacts with Msb2p -0.24 0.57 0.17 YBR063C  Putative protein of unknown function 0.25 0.11 0.18 YBR064W  Dubious open reading frame 0.29 0.27 0.28 YBR065C ECM2 Pre-mRNA splicing factor -0.45 -0.44 -0.44 YBR066C NRG2 Transcriptional repressor 0.35 0.18 0.27 YBR067C TIP1 Major cell wall mannoprotein with possible lipase activity -0.15 0.23 0.04 YBR068C BAP2 High-affinity leucine permease -0.01 -0.12 -0.06 YBR069C TAT1 Amino acid transporter for valine, leucine, isoleucine, and tyrosine -0.06 0.27 0.11 YBR071W  Protein of unknown function found in the cytoplasm and bud neck -0.37 0.15 -0.11 YBR072W HSP26 Small heat shock protein (sHSP) with chaperone activity 0.20 0.14 0.17 YBR073W RDH54 DNA-dependent ATPase -0.15 0.04 -0.05 YBR074W PFF1 Multi-spanning vacuolar membrane protease -0.41 -0.05 -0.23 YBR075W  Merged open reading frame -0.45 0.14 -0.15 YBR076W ECM8 Non-essential protein of unknown function 0.03 -0.17 -0.07 YBR077C SLM4 Component of the EGO and GSE complexes -0.49 0.27 -0.11 YBR078W ECM33 GPI-anchored protein of unknown function -0.43 0.20 -0.12 YBR081C SPT7 Subunit of the SAGA transcriptional regulatory complex    YBR082C UBC4 Ubiquitin-conjugating enzyme (E2) 0.41 -0.23 0.09 YBR083W TEC1 Transcription factor targeting filamentation genes and Ty1 expression 0.95 0.88 0.91 YBR084C-A RPL19A Ribosomal 60S subunit protein L19A 0.22 0.20 0.21 YBR084W MIS1 Mitochondrial C1-tetrahydrofolate synthase 0.02 -0.61 -0.29 YBR085W AAC3 Mitochondrial inner membrane ADP/ATP translocator -0.06 -0.35 -0.20 128  ORF  Gene Name Brief Description MAT α  Invertase Z Score MAT a  Invertase Z Score Average Mat a/α Invertase Z Score YBR090C  Putative protein of unknown function 0.06 0.14 0.10 YBR090C-A  #N/A -0.12 0.03 -0.04 YBR092C PHO3 Constitutively expressed acid phosphatase similar to Pho5p 0.53 0.26 0.40 YBR093C PHO5 Repressible acid phosphatase -0.60 0.47 -0.07 YBR094W PBY1 Putative tubulin tyrosine ligase associated with P-bodies 0.20 0.22 0.21 YBR095C RXT2 Component of the histone deacetylase Rpd3L complex -1.20 -1.46 -1.33 YBR096W  Putative protein of unknown function 0.00  0.00 YBR097W VPS15 Serine/threonine protein kinase involved in vacuolar protein sorting    YBR097W VPS15 Serine/threonine protein kinase involved in vacuolar protein sorting    YBR098W MMS4 Subunit of structure-specific Mms4p-Mus81p endonuclease -0.01 0.24 0.11 YBR099C  Dubious open reading frame -0.09 0.03 -0.03 YBR100W  Merged open reading frame 0.07 0.29 0.18 YBR100W  Merged open reading frame 0.07 0.29 0.18 YBR101C FES1 Hsp70 (Ssa1p) nucleotide exchange factor -0.01  -0.01 YBR103W SIF2 WD40 repeat-containing subunit of Set3C histone deacetylase complex 1.19 1.08 1.14 YBR104W YMC2 Putative mitochondrial inner membrane transporter -0.14 0.47 0.17 YBR105C VID24 GID Complex regulatory subunit -0.38 -0.18 -0.28 YBR106W PHO88 Probable membrane protein -1.22 -0.69 -0.96 YBR107C IML3 Outer kinetochore protein and component of the Ctf19 complex -0.74 0.11 -0.31 YBR108W AIM3 Protein that inhibits barbed-end actin filament elongation 0.14 -0.39 -0.12 YBR111C YSA1 Nudix hydrolase family member with ADP-ribose pyrophosphatase activity 0.12  0.12 YBR112C CYC8 General transcriptional co-repressor 0.30  0.30 YBR113W  Dubious open reading frame -0.32 0.36 0.02 YBR114W RAD16 Nucleotide excision repair (NER) protein -1.69 0.07 -0.81 YBR115C LYS2 Alpha aminoadipate reductase 0.19 -0.47 -0.14 YBR116C  Dubious open reading frame  0.42 0.42 YBR117C TKL2 Transketolase 0.31  0.31 YBR119W MUD1 U1 snRNP A protein 1.69 -0.01 0.84 YBR120C CBP6 Mitochondrial protein required for translation of the COB mRNA 0.39 0.94 0.66 YBR125C PTC4 Cytoplasmic type 2C protein phosphatase (PP2C) -1.02 -0.02 -0.52 YBR126C TPS1 Synthase subun