Protein quality control at the endosome: a novel model

Transporters must be sustained in proper working function at the plasma membrane to prevent serious harm to a cell. A dysfunctional transporter may cause leaking of ions, loss of membrane potential, disturbed cell signaling, and even cell death. Therefore, cautious monitoring of transporters occurs at the plasma membrane. A transporter which becomes unfolded or unstable is quickly endocytosed and taken to the early endosome compartment of the cell. However, these unfolded transporters are not automatically degraded, as a cell attempts to preserve them by allowing time for refolding. It is this process we refer to as quality control: where the decision is made whether to degrade the dysfunctional protein to maintain cell integrity or attempt to refold it to conserve energy. The endosome provides a safe place where this decision can be made. Properly folded and functional transporters will return to the plasma membrane to resume pumping, while those deemed dysfunctional will continue on to be degraded. In the case of the yeast uracil transporter Fur4, unfolded/destabilized Fur4 is ubiquitinated (signal for degradation added) by Rsp5, the only ubiquitin ligase known to work at the plasma membrane in yeast. After being endocytosed and brought to the early endosome, the decision to degrade or recycle Fur4 is made there. How exactly that quality control decision is made is not well understood. It is known that ubiquitination status plays a major role in that decision, and that both ubiquitination and deubiquitination can occur at the early endosome. However, studies have shown that Fur4 can recycle in spite of being in a ubiquitin tagged state in some mutant strains, suggesting the quality control decision is not so black and white. We attempt to elucidate this quality control process. Our results, and those published, have led us to a model whereby ubiquitination status is not the sole deciding factor involved in sorting at the early endosome, but where retention of ubiquitinated cargo is actually the key sorting step. Herein we describe a complex of proteins working together at the early endosome to carry out quality control.

PROTEIN QUALITY CONTROL AT THE ENDOSOME: A NOVEL MODEL by Katelyn Rose Froehlich A thesis submitted to the faculty of the The University of Utah in partial fulfillment of the requirements for the degree of Master of Science Department of Biology The University of Utah December 2015 Copyright © Katelyn Rose Froehlich All Rights Reserved The University of Utah Graduate School STATEMENT OF THESIS APPROVAL The thesis of Katelyn Rose Froehlich has been approved by the following supervisory committee members: Markus Babst , Chair 8/11/2015 Date Approved Kent Golic , Member 8/11/2015 Date Approved Julie Hollien , Member 8/11/2015 Date Approved and by Denise Dearing , Chair/Dean of the Department/College/School of Biology and by David B. Kieda, Dean of The Graduate School. ABSTRACT Transporters must be sustained in proper working function at the plasma membrane to prevent serious harm to a cell. A dysfunctional transporter may cause leaking of ions, loss of membrane potential, disturbed cell signaling, and even cell death. Therefore, cautious monitoring of transporters occurs at the plasma membrane. A transporter which becomes unfolded or unstable is quickly endocytosed and taken to the early endosome compartment of the cell. However, these unfolded transporters are not automatically degraded, as a cell attempts to preserve them by allowing time for refolding. It is this process we refer to as quality control: where the decision is made whether to degrade the dysfunctional protein to maintain cell integrity or attempt to refold it to conserve energy. The endosome provides a safe place where this decision can be made. Properly folded and functional transporters will return to the plasma membrane to resume pumping, while those deemed dysfunctional will continue on to be degraded. In the case of the yeast uracil transporter Fur4, unfolded/destabilized Fur4 is ubiquitinated (signal for degradation added) by Rsp5, the only ubiquitin ligase known to work at the plasma membrane in yeast. After being endocytosed and brought to the early endosome, the decision to degrade or recycle Fur4 is made there. How exactly that quality control decision is made is not well understood. iv It is known that ubiquitination status plays a major role in that decision, and that both ubiquitination and deubiquitination can occur at the early endosome. However, studies have shown that Fur4 can recycle in spite of being in a ubiquitin tagged state in some mutant strains, suggesting the quality control decision is not so black and white. We attempt to elucidate this quality control process. Our results, and those published, have led us to a model whereby ubiquitination status is not the sole deciding factor involved in sorting at the early endosome, but where retention of ubiquitinated cargo is actually the key sorting step. Herein we describe a complex of proteins working together at the early endosome to carry out quality control.TABLE OF CONTENTS ABSTRACT .......................................................................................................... iii LIST OF FIGURES .............................................................................................. vi ABBREVIATIONS ............................................................................................... vii ACKNOWLEDGMENTS ....................................................................................viii Chapters 1. INTRODUCTION ............................................................................................ 1 1.1 Protein Homeostasis ........................................................................... 1 1.2 Trafficking of Plasma Membrane Proteins ........................................... 3 1.3 Fur4 as a Model Protein in Yeast ........................................................ 6 2. A NOVEL MODEL: ENDOSOMAL QUALITY CONTROL AND RETENTION SYSTEM ............................................................................................................. 12 2.1 Introduction ........................................................................................ 12 2.2 Materials and Methods ....................................................................... 17 2.3 Results ............................................................................................... 20 2.4 Discussion .......................................................................................... 28 3. CONCLUSION ............................................................................................... 40 3.1 Concluding Remarks and Future Directions ....................................... 40 REFERENCES ................................................................................................... 42LIST OF FIGURES Figures Page 1.1: General Trafficking Route of a Plasma Membrane Protein .......................... 10 1.2: Fur4 Conformational Changes ..................................................................... 11 2.1: Protein Interaction Map ................................................................................ 31 2.2: EQRS (Endosomal Quality control & Retention System) Model ................... 32 2.3: Localization of EQRS Proteins ..................................................................... 33 2.4: Microscopy on rsp5-1 Cells .......................................................................... 34 2.5: Microscopy on ubp2Δ Cells .......................................................................... 35 2.6: 5-FU Growth Assay with ubp2Δ Cells .......................................................... 36 2.7: Western Blot Data ........................................................................................ 37 2.8: Microscopy on rup1Δ Cells .......................................................................... 39 ABBREVIATIONS 5-FU ................................................................................................. 5-FluoroUracil CFTR ............................. Cystic Fibrosis Transmembrane conductance Regulator DUB ................................................................................ DeUBiquitinating enzyme ER ..................................................................................... Endoplasmic Reticulum ESCRT ................................. Endosomal Sorting Complex Required for Transport EQRS ........................................... Endosomal Quality control & Retention System GFP ............................................................................. Green Fluorescent Protein ILV ........................................................................................ IntraLumenal Vesicle LID .................................................................................. Loop Interaction Domain MCS ...................................................................................... Multiple Cloning Site MVB ........................................................................................ MultiVesicular Body PM ........................................................................................... Plasma Membrane QC .................................................................................................. Quality Control RFP .................................................................................. Red Fluorescent Protein ACKNOWLEDGMENTS This project was conceived by Markus Babst. I would not have had the pleasure of working on it were it not for the many opportunities I am grateful to have been offered. I thank the Biology Department and the University of Utah for inviting me to join the Molecular, Cellular and Evolutionary Biology graduate program. I especially thank my mentor, Markus, for the invitation to join his lab. Markus was a fantastic PI and taught me more in a few years than I had learned in all my other years of school. I greatly appreciate having been able to learn how to do research under his guidance. I thank Matt Curtiss, our former lab technician, for showing me the ropes during my rotation in the Babst Lab. I thank my lab mates for their time, valuable input, help, and friendship. Justin Keener and Shrawan Mageswaran were senior graduate students at the time I joined, and they were of immense help and guidance during my first year in the lab. Akshay Moharir, Daniel Anand, and Will Barbeau joined the lab around the time I did, and I was very glad to have a cohort where we could all learn together and teach each other. I also thank the members of my committee, Kent Golic and Julie Hollien, for contributing their time and energy to helping me with this thesis project. I also owe gratitude to the NIH for providing funding.CHAPTER 1 INTRODUCTION 1.1 Protein Homeostasis Protein homeostasis, or proteostasis, is the maintenance of well functioning proteins in a cell required for its success and survival. Unfolded proteins are a detriment to the cell because of loss of function, formation of protein aggregates, and triggering of cellular stress responses. Unfolded proteins at the plasma membrane are especially detrimental to a cell, as they can cause leaking of ions and loss of membrane potential across the plasma membrane (PM). Therefore, a cell must remove unfolded proteins from the plasma membrane immediately through what is called endocytosis, and eliminate the dysfunctional proteins through degradation. The downside to this overly cautious system is that it can be energetically costly. Degraded proteins have to be replaced through new protein synthesis (which requires a lot of energy) to maintain functioning proteins at the plasma membrane. To conserve energy, proteins that are endocytosed are not always degraded. These proteins are given time to refold, safely inside the cell, and then can be recycled back to the plasma membrane. This process allows for the balance of cell integrity with energy conservation. A cell can remove potentially harmful proteins from the2 plasma membrane, reevaluate/refold them, and then recycle and reuse the functional proteins while degrading the dysfunctional (1). This system, whereby unfolded proteins are taken care of by being either refolded or degraded, is referred to as protein quality control (QC). There are different, but similar, protein quality control systems in place at other compartments in the cell as well as the plasma membrane. The system in place at the Endoplasmic Reticulum (ER) acts as new proteins are synthesized and folded, before they are transported to their destination in the cell. Generally, unfolded/misfolded proteins at the ER are held back and reevaluated by chaperone proteins, which assist in refolding, or when unsuccessful, aid in the protein being sent for degradation (2, 3). The idea is similar to that at the plasma membrane, but the proteins involved in quality control, and the processes by which they work, are very different at each location. For one thing, chaperones may not have as much access to membrane proteins at the endosome, so other QC mechanisms may be in place. While ER quality control still lends itself to many unanswered questions, also obscure is the quality control system that works after a protein has reached the plasma membrane. The importance of having a protein quality control system in place is evident in diseases that result from loss of proteostasis. Alzheimer's disease, Parkinson's disease, cystic fibrosis, prion diseases, and many more, all involve protein un-/mis-folding or protein aggregation (4-6). Whether the observed imbalance in proteostasis is a causative factor or symptomatic attribute of the disease is often unknown. Cystic fibrosis, however, is a clear example of the direct effects of protein misfolding. A ΔF508 mutation, which renders the 3 mammalian cystic fibrosis transmembrane conductance regulator (CFTR) protein unable to fold properly, is present in the vast majority of cystic fibrosis patients (7). Even native CFTR, a large and cumbersome protein, simply takes too long to fold efficiently, and so it is often subjected to ER quality control and never makes it out of the ER. Accordingly, about 75% of CFTRΔF508 protein is unable to fold and is held back in the endoplasmic reticulum (8). Hence, not enough of the protein is able to make it to the plasma membrane where it serves as a chloride channel, and the repercussion is an imbalance of chloride ions, leading to many problems. The native CFTR protein that escapes the ER and makes it to the plasma membrane is still very unstable and frequently endocytosed (9). CFTR has large cytosolic domains, and less region within the membrane, so it is able to be accessed by chaperones post-Golgi. About 75% of the CFTR endocytosed is, in turn, recycled back to the plasma membrane (10). So, the route of recycling instead of being degraded is very common for some proteins, and may be the norm. It is also a vital pathway, as the degradation of all endocytosed native CFTR would also likely result in a cystic fibrosis phenotype, one would predict. Clearly, there is a very important quality control process going on directly after endocytosis, about which we know very little, and it will be discussed herein. 1.2 Trafficking of Plasma Membrane Proteins Proteins have predefined destinations in a cell, and specific routes to travel in getting there. Proteins are tagged with specific amino acid sequences that specify where they should be sent (called sorting signals) (11), much like an 4 address on a postal package. Generally, ER targeted membrane proteins that lack specific sorting signals are sent to the plasma membrane as a default pathway(12). Other membrane proteins contain signals to send them to specific compartments in a cell. This sorting decision occurs at the "post office" of the cell, the Golgi apparatus.(12) After being synthesized in the ER, membrane proteins are sent through the Golgi for packaging and, if lacking a sorting signal, then travel in vesicles out to fuse with the plasma membrane (12). If the protein becomes unfolded or unstable at the PM, it is ubiquitinated (signal for degradation added), endocytosed, and taken to the early endosome (Fig. 1.1a) (13-16). The early endosome likely serves as a safety compartment in the cell, as it is here that a protein can be reevaluated, and a decision made whether to recycle or degrade the protein (14, 17). If the protein appears folded, it can be deubiquitinated (signal for degradation removed), and recycled from the early endosome back to the PM, to continue functioning there (Fig. 1.1b) (17). If the protein is unable to refold, it continues down the pathway to degradation (Fig. 1.1c). The early endosome matures into a late endosome, while taking its cargo (contents) with it (15). At the late endosome stage, the Endosomal Sorting Complex Required for Transport (ESCRT) proteins are recruited (18). The ESCRTs concentrate cargo proteins into a pool on the endosomal membrane (15, 19). The ESCRTs also facilitate the deformation of the membrane, to create an invagination into which the deubiquitinated cargo is deposited (Fig. 1.1d). For cargo to be deposited into intralumenal vesicles (ILVs), ubiquitin, the modification that has been added onto the cargo protein as a degradation signal, is removed and recycled (20, 21). At 5 this stage, in which formation of intralumenal vesicles (ILVs) occurs, the late endosome has matured into what is called a multivesicular body (MVB) (19). The MVB then fuses with the lysosome (in higher eukaryotes, vacuole in plants/fungi), and releases its ILVs there (Fig. 1.1e) (1, 19). In the acidic lumen of the lysosome/vacuole, ILVs are exposed to lipases and peptidases, which degrade both the lipids and the proteins, respectively (1, 19). From there, the broken up amino acids can be recycled and go to the cytoplasm for new protein synthesis. Ubiquitin, as previously mentioned, is considered the universal tag for degradation (22). The 76 amino acid ubiquitin moiety is added to both soluble and membrane proteins as a signal to the cell that they need to be degraded (17, 23). A single ubiquitin is sufficient as the signal for degradation of membrane proteins (24), but polyubiquitin chains are common (15, 25). Three types of enzymes act to add the ubiquitin modification to target proteins: E1, E2, and E3 enzymes. E1 enzymes are ubiquitin-activating enzymes, which prepare ubiquitin for addition to a target protein. E2 enzymes are ubiquitin-conjugating enzyme, which are responsible for chain assembly and elongation (26). E3 enzymes are the ubiquitin ligases - they covalently attach the ubiquitin moiety onto a target protein, specifically at a lysine residue (27, 28). In the yeast Saccharomyces cerevisiae, there is only one known E3 ubiquitin ligase that works on proteins at the plasma membrane, called Rsp5 (29-32). RSP5 knockout strains are not viable (31), but an rsp5-1 strain contains a hypomorphic allele of RSP5 that results in a far less active ubiquitin ligase (33). In this strain, proteins are stabilized on the plasma membrane because they are unable to be efficiently ubiquitinated (16, 34, 35). Mutation of the lysine residues 6 within a protein that are targeted for ubiquitination also results in the stabilization of the mutant protein on the plasma membrane (35, 36). Either monoubiquitin or polyubiquitin chains may be formed on a cargo protein. Polyubiquitin chains can be made through differing linkages as well, meaning another ubiquitin can be added onto any one of the 7 lysines within ubiquitin (37). All of these variations may signal different fates for target proteins (38). The standard convention is that K63 polyubiquitin chains are typically formed for the degradation of transmembrane proteins through the MVB pathway, while K48 linkages are used to direct soluble proteins to the proteasome for degradation (15, 39, 40). 1.3 Fur4 as a Model Protein in Yeast To study quality control at the plasma membrane, we work with the yeast Saccharomyces cerevisiae as our model organism. S. cerevisiae is a good choice as it is hardy, inexpensive, and easy to genetically modify as we keep it in the haploid state. Our chosen model protein is Fur4. Fur4 is the yeast uracil transporter. It is a member of the nucleobase cation symport-1 (NCS-1) family; it pumps uracil in along with a proton (symport), using the proton gradient to move uracil against its gradient (41, 42). Fur4, at 633 amino acids, is a large transporter, containing 12 transmembrane domains (43). Because the transport of uracil requires large conformational changes in Fur4, this protein is rather unstable. The physical arrangement of these domains and loop regions, and their movements during pumping, can be estimated based on knowledge of the crystal structure of a NCS-1 family bacterial transporter homolog, Mhp1 (42, 44). 7 Based on that structure, Keener & Babst (2013) predicted Fur4 to have an N-terminal LID region, or Loop Interaction Domain, which as it sounds, interacts with the loop regions of the transmembrane domains on the cytosolic side (35). Also present on the N-terminus is a degron, a series of phosphorylation and ubiquitination sites (lysines); presence of which is sufficient for recognition and degradation of a target protein (35, 45, 46) (Fig. 1.2). Many studies have found that the yeast uracil permease Fur4 is downregulated upon addition of substrate (i.e., uracil) to the culture (16, 35, 47, 48). Downregulation efficiency was similar to that seen with heat shock or hydrogen peroxide treatment, known triggers of protein unfolding and endocytosis (35). Keener & Babst (2013) proposed that the flexible Fur4 must go through large conformational changes to allow the transport of uracil, and during this time out of the ground state, the protein may look unfolded (35). Their idea is that the LID region actually acts like a lid on Fur4, keeping closed in the ground state (35). In the ground state, within the N-terminal 110 amino acids are residues that make contact with the loops of the transmembrane domains; this interaction is what holds the LID "shut" while also contributing to the protein's stability (35). While the LID is "shut" the degron is tucked in, preventing its accessibility (35) (Fig. 1.2a). But during pumping of substrate, the LID must open up to allow binding and release of uracil to the cytosol. During this time spent open, the LID-loop interactions are lost and the degron is exposed, and thus accessible to ubiquitination (35) (Fig. 1.2b). Under this model, simple diffusion and chance determine ubiquitination probability. During degron exposure, if the ubiquitin ligase Rsp5 comes along, it will ubiquitinate Fur4. Therefore, longer 8 time spent in the non-ground state equals longer degron exposure, and greater chance of a run in with Rsp5, which leads to ubiquitination and endocytosis. This is called an intrinsic sensing mechanism in that Fur4 itself senses when it is unfolded by loss of LID-loop interactions and degron exposure, and thus presents iteslf for ubiquitination (35). This QC mechanism works independently of chaperones, as opposed to CFTR quality control (1, 35). Support for the LID-degron idea comes from experimental manipulation of Fur4 protein. Deletion of the N-terminal 60 amino acids, what is deemed the degron, resulted in stabilization of Fur4 on the plasma membrane, even under conditions that should result in its downregulation (e.g., heat shock, peroxide treatment, and high substrate addition) (35). Alternatively, lysine-to-arginine mutation of residues within the degron also results in Fur4 stabilization under these conditions (35). This demonstrates that the degron, and specifically lysine ubiquitination sites, are necessary for Fur4 downregulation. Fur4 in the rsp5-1 strain is also stabilized at the plasma membrane, in spite of cell stress and substrate addition (35). This shows ubiquitination is necessary for Fur4 downregulation. Since we have a supported model of Fur4 behavior at the plasma membrane, it is a good transporter to use in our studies. Additionally, its flexible nature and instability make Fur4 a model protein for our quality control experiments: unfolding can easily be induced, causing ubiquitination and endocytosis of Fur4. Endocytosis of Fur4 can be induced by three modes: 1) cell starvation, which is known to cause downregulation of nutrient transporters; 2) stressing the 9 cell via heat shock, peroxide treatment, etc.; or 3) addition of high uracil concentration. The first mode is understood to occur as part of starvation response pathways like TORC1 (Target Of Rapamycin Complex 1), which are involved in nutrient sensing and turnover of proteins in response to low nutrients (49). The second mode is more straightforward - the use of known cell stressors to cause protein unfolding by disrupting chemical interactions that hold the protein in its folded conformation. The third mode, substrate induced downregulation, had been observed previously, but was not well understood until the new mechanism was proposed and published by Kenner & Babst in 2013 (35). The second and third modes both use the same mechanism resulting in endocytosis - the LID-degron system described above - and thus, these are the modes we employ to induce endocytosis in QC experiments (35). 10 Fig. 1.1: General Trafficking Route of a Plasma Membrane Protein a. A plasma membrane protein undergoes endocytosis, and is sent to the endosome. b. Endosomal membrane recycles, and can carry cargo protein with it, to fuse back with the plasma membrane. c. Alternatively, a protein continues down the path to degradation, traveling from endosome to MVB. d. ESCRTs are recruited to facilitate deposition of cargo proteins into ILVs. e. MVB fuses with vacuole/lysosome, releasing its contents for degradation there. (Figure provided by Markus Babst.) a b c d e 11 Fig. 1.2: Fur4 Conformational Changes Without substrate bound, Fur4 is in the ground state, with LID closed and degron tucked away from Rsp5 access. During substrate pumping, Fur4 undergoes large conformational changes that cause the LID to open up, exposing the degron. Fur4 deviates from the ground state any time uracil binds, but can revert to the ground state when no longer bound. - CHAPTER 2 A NOVEL MODEL: ENDOSOMAL QUALITY CONTROL AND RETENTION SYSTEM 2.1 Introduction 2.1.1 Early Endosome Dynamics After endocytosis of a plasma membrane protein, it travels to the early endosome, where a decision is made whether to continue down the path to degradation or to recycle back to the plasma membrane. But how is this QC decision made and carried out? In the case of Fur4, we understand the quality control mechanism in place at the plasma membrane, but we do not know much about the second QC system in place at the early endosome. There is plenty of evidence of dynamic ubiquitination and deubiquitination of cargo proteins, which points to a method of executing this decision (1, 25, 50, 51). At the early endosome, ubiquitin ligase can add ubiquitin moieties to a protein, while ubiquitin proteases can cleave off ubiquitin moieties. Again, the only known ubiquitin ligase working at the early endosome is Rsp5 (29-32). There are many deubiquitinating enzymes (or DUBs) in a cell - potentially around 100 in human and 17 in yeast - possibly of gene families that evolved to function at different compartments of a cell or on specific substrate (52, 53). We focused on 13 deubiquitinating enzymes Ubp2 and Ubp7 as likely to be working at the early endosome, based on evidence of their action on Ede1, an endocytic protein, and based on their interactions with proteins known to be functioning around this stage (25, 54). A simplistic idea follows that upon reaching the early endosome, a cargo protein either is deubiquitinated and recycled back to the plasma membrane, or it is reubiquitinated/polyubiquitinated and goes along with the maturing endosome to be degraded in the vacuole. In this scenario, ubiquitination status is the key sorting signal to determine protein fate. Degradation is the signal mediated path, requiring presence of ubiquitin (the signal), while recycling to the plasma membrane is the default path taken in absence of ubiquitin. If this is the case, the system could be upset by perturbation of the enzymes Rsp5, Ubp2, or Ubp7. Removal of Rsp5 from the system would, we expect, result in cargo proteins lacking ubiquitin sorting signals, and recycling by default back to the plasma membrane. (This idea of course is convoluted by the fact that Rsp5 works at the plasma membrane too, so initial endocytosis would be largely impaired.) On the other hand, removal of the DUBs from the system would result in cargo proteins remaining ubiquitinated or growing long polyubiquitin chains, and going forth with the maturing endosome to be eventually degraded in the vacuole. This is what is expected if quality control at the early endosome is a straightforward system whereby ubiquitin ligase adds ubiquitin to a protein that appears unfolded, and DUBs remove ubiquitin from an apparently folded protein. In the case of Fur4, and based on the work done by Keener & Babst (2013), "appearing unfolded" means LID opening and degron exposure, while appearing folded means the 14 degron is neatly tucked away, inaccessible to Rsp5. So, the decision made at the early endosome would be dependent on whether a protein appears unfolded or folded and on corresponding ubiquitination status. Rechecking of proteins at the early endosome is very important as this step allows the balance of cell integrity with energy conservation, and it seems to be the stage when the final decision is made on a protein's fate. It is also a very common process to recycle PM proteins, as in the example of CFTR that was given (10, 55). Proteins may become temporarily unfolded, but they can be safely brought inside the cell for reevaluation by the second QC system, and then ubiquitin ligases/proteases can modify the protein accordingly to facilitate the decision. However, the quality control scenario presented is a very simplistic idea and current research suggests the process is not so black and white. 2.1.2 Key Proteins Involved at the Early Endosome To delve into the complexity of the quality control system at the early endosome, a deeper look at the proteins at work is warranted. There is plenty of evidence that Rsp5, Ubp2, and Ubp7 do not work independently of each other or of other proteins (25, 50, 51, 56). On the contrary, many of these proteins physically interact with each other (25). Pull down assays (with addition of mutation of predicted binding sites) plus immunoblotting were used to develop a protein interaction map, as published (25). That data, in addition to similar experiments, informed the interaction map shown in Fig. 2.1. According to the map, Rsp5 and Hua1, a protein of unknown function, physically interact with each other and with Hse1, an ESCRT-0 protein (25). Both proteins also interact with 15 Sla1, part of the upstream endocytic machinery (57). Ubp7 interacts with Hse1 as well (25). The protein Rup1 stands for Rsp5-Ubp2 interaction Protein; essentially, Rup1 tethers the ubiquitin ligase to the DUB (50). The importance of these interactions has not been fully elucidated yet, but it is known that their interaction is important for controlling sorting efficiency. For example, it seems Rsp5 requires Rup1 for tethering to Ubp2, in order for Rsp5 to efficiently ubiquitinate cargo (50, 51). It has been proposed that this may be due to an autoubiquitination process carried out by Rsp5 (56). The idea is that through autoubiquitination, Rsp5 could actually downregulate itself, adding the signal for degradation to itself, and then Ubp2 acts on Rsp5 by removing the ubiquitin it adds to itself. But others argue this is not the reason for Rsp5's requirement of Ubp2 for efficient ubiquitination of cargo, due to the fact that deletion or overexpression of Ubp2 does not alter steady state levels of Rsp5 (50). The four main proteins we have focused on are Rsp5, Rup1, Ubp2, and Hua1. All are soluble proteins and all are recruited to the endosomal membrane (according to a mass of data from the Saccharomyces Genome Database). We think these four proteins function in quality control at the endosome. We believe they work after the upstream endocytic machinery, and before the downstream ESCRT machinery takes over. But how exactly they function together at the endosome is unknown. 2.1.3 New Model of QC at the Endosome How do the ubiquitin ligase and DUBs work together at the early endosome to determine protein fate? This is the big question that the present 16 study attempts to address. The simple model of dynamic ubiquitination and deubiquitination does not suffice. Evidence at odds with this simple model is abundant. The biggest problem is that UBP2 deletion experiments have shown increased recycling of cargo, and less degradation (51). Published growth assays have shown hypersensitivity to drugs in ubp2Δ cells, indicating a greater presence of the transporter at the plasma membrane, indicative of increased recycling (51). Pulse chase experiments confirmed this, with Fur4-GFP remaining on the plasma membrane even after addition of uracil substrate (51). This is in stark contrast to what is expected. Deletion of the deubiquitinating enzyme Ubp2 should lead to more ubiquitination and thus more degradation of plasma membrane proteins, but only the first prediction holds true. Deletion of UBP2 does result in increased polyubiquitination, as demonstrated through immunoblotting against ubiquitin in ubp2Δ cells (58). But in looking at Fur4, it did not lead to concomitant increased degradation as expected. There is clearly more going on at the early endosome than just ubiquitination versus deubiquitination of cargo proteins. The goal of this study was to attempt to reconcile these unexpected results and explore new models of quality control at the early endosome. Based on current knowledge, we came up with a new model whereby the aforementioned proteins - particularly Rsp5, Ubp2, Hua1, and Rup1 - work in a complex to retain cargo proteins at the early endosome. We have dubbed this model EQRS = Endosomal Quality control & Retention System. The idea is that while a cargo protein goes through a cycle of ubiquitination and deubiquitination, it is transiently bound to Rsp5 or Ubp2, and likely bound to Rup1 or Hua1 through 17 attached ubiquitin moieties. All of these interactions keep the cargo protein at the early endosome. After a period of cycling, whatever remains will continue with the maturing endosome, through the MVB pathway, to eventually be degraded. While ubiquitin is still the signal for degradation, and lack thereof should default to recycling, under this model, retention is the key sorting step. So disruption of EQRS components could disrupt the whole complex and thus retention, causing a loss of regulated sorting at the endosome. (Fig. 2.2) The model draws comparisons from the system of glycosylation at the ER, whereby keeping a soluble protein in a cycle of glycosylation and de-glycosylation reactions retains the protein in the ER (2). 2.2 Materials and Methods 2.2.1 Strains & Growth Conditions Saccharomyces cerevisiae strains used for this study are listed in Table 2.1. Strains were created as follows: PCR was done on integration plasmid templates containing a selectable marker (e.g., G418 resistance, HIS auxotrophy), and fluorescent tag where needed, with primers designed to add 50 bp flanking regions of homology from target genes; transformation into wildtype yeast strain SEY 6210 allowed homologous recombination of the marker/tag into the genome; selection markers allowed for picking of colonies with the edited genome. New strains were confirmed by PCR analysis of genomic DNA. Yeast strains were grown at 30°C in common YPD (Yeast extract-Peptone-Dextrose), or in YNB (Yeast Nitrogen Base) supplemented with all amino acids except auxotrophic marker(s) to maintain plasmid. For inducing expression of copper 18 promoted Fur4-GFP (pJK19), 50 μM copper sulfate was added to the growth media. Induction of substrate-induced downregulation was done by adding 2g/100ml uracil to the media. Heat shock treatment involved swirling culture tubes in a 37°C water bath to shock, and then transfer to a 37°C shaking incubator for between 15-45 minutes. 2.2.2 DNA Manipulations Plasmids used for this work are listed in Table 2.1. Genes were amplified by PCR from genomic SEY6210. Restriction enzyme cloning was then used to insert the gene into a pRS4 series vector, with promoters as specified in Table 2.1. Plasmids were amplified in XL1-Blue strain of Escherichia coli, grown in standard LB media at 37°C, and selected for with addition of 100g/ml ampicillin to the media. For construction of the ubp2C745A mutant: a plasmid was created through PCR mutagenesis and homologous recombination. PCR amplification of two fragments of UBP2 was done on a UBP2 containing plasmid (pKF17), using M13R universal primer as forward with a UBP2 reverse primer with base substitution, and T7 universal primer as reverse with a UBP2 forward primer with base substitution. Use of M13R and T7 created regions of homology to the MCS of a pRS vector. The two PCR fragments could then be transformed with a gapped vector into yeast, and homologous recombination stitched together the three fragments, creating the plasmid pKF19. pKF19 size was confirmed by plasmid prep and test digest, and sanger sequencing confirmed presence of the C745A mutation. 19 2.2.3 Fluorescence Microscopy A deconvolution microscope (DeltaVision, Applied Precision) was used to perform all fluorescent microscopy in the present study. Cells were typically grown to 0.6-0.8 OD600/ml (log phase) before harvesting for microscopy. The different strains were analyzed for Fur4-GFP cellular distribution, observing presence/absence and relative amount of protein in compartments along the endocytic trafficking route: plasma membrane, endosomes, MVBs, vacuole. 2.2.4 5-Fluorouracil (5-FU) Growth Assays YNB plates were made without uracil (and without specific plasmid selection markers when required), and 5-Fluorouracil (5-FU) was added to the plates prior to pouring. 2μM 5-FU was determined to be the optimal concentration for these growth assays. Cultures were allowed to grow to ~0.7 OD600/ml before harvesting for the assay. Cells were diluted in 1M sorbitol to a final concentration of 0.5 OD600/1ml. A 1:5 dilution series was done in sorbitol in a 96-well plate. 3μl of the cells in sorbitol were plated, and images were captured after 1-4 days of growth at 30°C. 2.2.5 Western Blotting Cells were grown to log phase and a total of 3 OD600 of cells were harvested for protein preps. After centrifiguation, cells were resuspended in 6M urea sample buffer plus 100μl/ml β-mercaptoethanol. Cells were vortexed with glass beads and heated at 65°C for 5 minutes (2x). 8% SDS-PAGE gels were used for visualizing Fur4-GFP. Primary antibodies used were mouse-αGFP and 20 rabbit-αSnf7 (loading control), while secondary antibodies were goat-αmouse (800 channel) and goat-αrabbit (680 channel). Blots were imaged on Licor Odyssey scanner and analyzed with Image Studios 4.0. 2.3 Results 2.3.1 Localization of EQRS Components Our first priority was to confirm that the four key proteins in the EQRS complex are recruited to the endosome. Rsp5 has already been localized to endosomes, as well as the plasma membrane (32). To localize the other components, we made genomic integrations of GFP tags on RUP1, HUA1, and UBP2. Hua1-GFP had such low and diffuse expression, it was difficult to tell whether it specifically localized to the endosome. Rup1-GFP and Ubp2-GFP were also diffuse in the cell, but possibly appeared concentrated in small dots near the plasma membrane (likely endosomes). For further proof, we looked at the GFP-tagged proteins in a vps4Δ strain (which accumulates late endosomes), and co-localized with mCherry-DID2 (an ESCRT factor tagged with an RFP). There was no overlap of the two fluorescent markers, indicating that Rup1 and Ubp2 are not in late endosomes, but work upstream of the ESCRTs (Fig. 2.3). 2.3.2 Fur4-GFP Trafficking in Hypomorphic rsp5-1 Use of an hypomorphic RSP5 allele results in stabilization of Fur4 on the plasma membrane, even under heat shock (Fig. 2.4a). However, when we look at Fur4-GFP in rsp5-1 strain MYY808 by microscopy, there were still some endosomal structures visible after 60 minutes +uracil (Fig. 2.4b). In this strain, 21 Rsp5 activity is severely impaired, but not abolished. So, Rsp5 can still ubiquitinate at the PM, though much less efficiently, and limited endocytosis does occur. However, once at the early endosome, Rsp5 always loses the tug-of-war with the DUBs, and thus proteins are deubiquitinated and recycle. The observed phenotype in rsp5-1 strain represents both stabilization of Fur4 on the plasma membrane and increased recycling of Fur4 to the PM. 2.3.3 Deletion of EQRS Component Ubp2 Deletion of UBP2 results in increased polyubiquitination. Since ubiquitin is supposed to be the signal for degradation, it is surprising that deletion of UBP2 does not lead to increased degradation as expected. Our retention model is based on this anomaly and attempts to explain it. Therefore, while it has been published, we wanted to confirm it with our own experiments. We looked by fluorescence microscopy at trafficking of Fur4-GFP in a UBP2 deletion strain (ubp2Δ, named MCY66) and observed similar effects (Fig. 2.5). Fur4 was stabilized on the plasma membrane, in spite of treatments like heat shock or substrate addition, which usually induce downregulation. In fact, the ubp2Δ cells looked more like rsp5-1 cells (Fig. 2.4) than like wildtype cells (Fig. 2.5). Therefore, two strains that are expected to have polar opposite phenotypes, if Rsp5 functions solely as a ubiquitin ligase and Ubp2 functions solely as a deubiquitinase, actually yield very similar phenotypes. This indicates additional functions for these proteins at the early endosome. Additionally, we used growth assays to appraise relative amount of transporters on the plasma membrane. Transporters like Fur4 do not have an 22 on/off switch; if they are at the plasma membrane, they are pumping so long as there are substrate and protons to pump. Therefore, more transporters at the PM equals more substrate uptake. A toxic uracil analog called 5-Fluorouracil (5-FU) can be used to indicate uptake of substrate and thus relative amount of transporters on the PM. 5-Fluorouracil gets incorporated in RNA and inhibits RNA processing, thus inhibiting protein production and cell growth (59). Cells with more Fur4 transporters on the plasma membrane will take up more 5-FU and will grow more slowly. Hence, sensitivity to 5-FU is an indicator of recycling of Fur4 to the plasma membrane - more uptake of the drug is indicative of more recycling. The ubp2Δ strain MCY66 is more sensitive to 5-FU than wildtype, consistent with increased recycling of Fur4 (Fig. 2.6). As a third approach, we used western blotting to look at relative amounts of Fur4 degradation, plus or minus treatment. Looking at the amount of free GFP indicates the amount of Fur4-GFP protein that has entered and/or been degraded in the vacuole (since GFP is slow to degrade and can persist in the vacuole for hours). Under steady state conditions, more free GFP was present in the wildtype strain versus the ubp2Δ strain MCY66 (normalized to Snf7 loading control). This indicates more degradation of Fur4 in wildtype cells than in ubp2Δ cells. After heat shock, there is still more free GFP in wildtype than in ubp2Δ cells (Fig. 2.7). This result is consistent with microscopy and growth assay data, and supports the idea that ubp2Δ causes increased recycling of Fur4. These results from microscopy, growth assay, and western blotting strongly suggest that Ubp2 has other functions in addition to DUB activity, or that there is further regulation at the early endosome in addition to signal mediated 23 governance by ubiquitin. If our EQRS model is accurate, and Ubp2 functions in retention of cargo with the aid of the other proteins, the complete deletion of Ubp2 likely disrupts the EQRS complex. This explains why deletion of Ubp2 increases recycling - because retention is impaired and even ubiquitinated proteins escape the early endosome. Thus, we wanted to look at how an active site mutant version of Ubp2 would affect Fur4 trafficking, versus a complete deletion of Ubp2 protein. This would allow us to look at the singular effect of loss of DUB activity, without disrupting retention activity. 2.3.4 Complementation with Ubp2 Active Site Mutant We engineered a mutation (C745A) in the active site of Ubp2. This mutation abolishes the deubiquitinase activity of Ubp2, while still allowing it to bind ubiquitin and to interact with its normal protein partners (56, 60). We replaced endogenous Ubp2 with ubp2C745A, followed Fur4-GFP trafficking in the cell, and used growth assays and western blots to look at how the mutant affected degradation of Fur4. If Ubp2's role in QC at the early endosome is simply deubiquitination, a deletion and a point mutant should influence the system in the same way; they should lead to the same outcomes. However, if Ubp2 has a broader role in QC - retention of cargo proteins - then the two manipulations will yield very different results. ubp2Δ causes increased recycling, we presume, because of loss of retention of Fur4. We predicted ubp2C745A would, on the contrary, cause increased degradation of Fur4, because retention would still be possible while deubiquitination would not. Interestingly, when the mutant ubp2C745A (pKF19) was added as 24 complement to ubp2Δ strain (MCY66), Fur4 trafficking as seen by microscopy was clearly different from the ubp2Δ without complement added, and wildtype (Fig. 2.5). Fur4 was then degraded more efficiently. In fact, even under steady state conditions with no treatment, Fur4 was already visible in higher amounts inside the cell (in endosomes/MVBs/vacuole) than in ubp2Δ alone or wildtype (Fig. 2.5). This phenotype suggests decreased recycling of Fur4. 5-Fluorouracil growth assays were consistent with microscopy results. A full deletion of UBP2 grew the worst of strains tested (Fig. 2.6), as would be expected with more recycling of Fur4 (more uptake of 5-FU). On the contrary, complementing the ubp2Δ cells with mutant ubp2C745A (pKF19) rescued their growth, suggesting increased Fur4 degradation. Surprisingly, wildtype strain SEY6210 and the wildtype UBP2 complement (pKF17) in ubp2Δ strain grew the best. We had predicted the strain with ubp2C745A complement would have grown best because of high amounts of Fur4 degradation. Western blot showed overall lower levels of total Fur4 protein in ubp2C745A complement versus just ubp2Δ (Fig. 2.7). Also, with the mutant complement there wasn't much change in relative Fur4 degradation before or after heat shock, while in wildtype and ubp2Δ there is a clear increase in degradation following heat shock. Interestingly, the relative amount of Fur4 degradation occurring in ubp2C745A complement before heat shock, is about the same amount that occurs in ubp2Δ or wildtype after heat shock. The same trend is evident when looking at relative amount of Fur4 protein normalized to Snf7: the amount of Fur4 present in ubp2C745A complemented cells before heat shock is about the same amount of Fur4 present in ubp2Δ or wildtype after heat shock. So it appears even at steady 25 state there is probably more degradation occurring with the ubp2C745A mutant than in ubp2Δ alone or wildtype cells. This differs slightly from the 5-FU growth assay, but is more like what was expected. As predicted, while ubp2Δ causes increased recycling of Fur4, ubp2C745A complement reverses the phenotype causing increased degradation of Fur4. Thus far, results mostly support our EQRS model, or at least do not reject it. 2.3.5 Double Knockout of UBP2 and UBP7 As stated, there are two DUBs likely working at the early endosome. Deletion of UBP7 did not give such a strong phenotype, as compared to deletion of UBP2, which is why we focused on Ubp2. But if both DUBs were simultaneously deleted we predicted we might see more degradation, since this would theoretically nearly abolish deubiquitination at the early endosome. And microscopy and western blot data weakly suggest this may be correct. Western blot quantification shows slightly increased degradation of Fur4 in the double knockout after heat shock, versus wildtype and single ubp2Δ (Fig. 2.7). Interestingly, when complemented with ubp2C745A, the double knockout strain follows a similar trend as the single ubp2Δ complemented: degradation of Fur4 is increased even before heat shock, and there is little change in Fur4 degradation after heat shock. By microscopy, there is not a clear phenotype of ubp2Δ-ubp7Δ at steady state. Upon treatment though, our ubp2Δ-ubp7Δ strains, KRY5-1 and KRY5-2, had more efficient downregulation and degradation than wildtype or ubp7Δ alone. After 30 minutes at high uracil concentration, the double knockout strains had 26 mostly endocytosed Fur4 off the plasma membrane, while a lot was remaining at the PM in wildtype cells at 30 minutes. The double knockout strains showed most of the GFP signal in vacuoles or endosomes/MVBs, while wildtype had less internal signal. 2.3.6 Deletion of EQRS Component Rup1 Rup1 is known to physically interact with and tether Ubp2 to Rsp5 (50). The purpose this serves is not well understood, but it has been shown that Rup1 modulates activity of Rsp5 and Ubp2 (50, 56). Could Rup1 be involved in the recruitment of both enzymes to the early endosome? If so, we predict deletion of RUP1 would completely dismantle the system. We looked by microscopy at Fur4-GFP trafficking in a rup1Δ strain (MCY64) (Fig. 2.8). A phenotype was not very conspicuous, but upon induced downregulation by uracil addition, we noticed quicker internalization (or just less recycling), and possibly, a delay in Fur4 reaching the vacuole. 60 minutes after uracil addition, many of these cells still showed a large amount of GFP signal coming from endosomes/MVBs, not as much from the vacuole (where most signal is in wildtype after 60 minutes uracil). It seems Fur4 is still efficiently endocytosed in rup1Δ cells, and possibly, undergoes less recycling than wildtype. Also, Fur4 becomes stuck in endosomes and doesn't make it to the vacuole as efficiently as usual in rup1Δ cells. By growth assay, rup1Δ appears less sensitive to 5-FU than wildtype. This data supports the idea of a slight recycling defect when Rup1 is absent. If this is the case, deletion of RUP1 does not fully dismantle the system as 27 expected. However, it may perturb both recycling and degradation, which could explain why we see Fur4 stuck in endosomes/MVBs. 2.3.7 Deletion of EQRS Component Hua1 By microscopy, deletion of HUA1 alone had no real discernible phenotype. However, in a 5-FU growth assay hua1Δ did appear to have a growth advantage over wildtype. It is possible that deletion of HUA1 slightly decreases recycling or increases degradation, but the effect is so mild it is not detectable by microscopy. 2.3.8 Double Knockout of HUA1 and RUP1 The lack of obvious phenotypes from deletion of HUA1 or RUP1 may be because the EQRS system has some redundancy. We thought combining HUA1 and RUP1 deletions in one strain might reveal a clearer phenotype. A phenotype did become more evident in RUP1-HUA1 double knockout cells after overnight growth. Typically after overnight growth, when cells have entered stationary phase and gone into starvation mode, almost all Fur4 has been endocytosed and is in the vacuole being degraded. This is what was observed in wildtype cells and in ubp2Δ-ubp7Δ cells: almost all GFP signal was seen in the vacuole. rup1Δ-hua1Δ cells, however, were loaded with endosomes containing Fur4-GFP and even had some signal remaining at the plasma membrane after overnight growth to stationary phase. This supports the idea that deletion of RUP1 causes a delay in cargo reaching the vacuole, while simultaneous deletion of HUA1 seems to amplify the effect. The strain has not yet been analyzed by western blot or growth assay. 28 2.4 Discussion We confirmed that deletion of UBP2 yields increased recycling, but that complementation with a Ubp2 active site mutant increases degradation. Thus, a simple concept of tug-of-war between Rsp5 ubiquitinating and Ubp2 deubiquitinating cargo proteins at the early endosome is insufficient to account for the decision to degrade or recycle, respectively. Additionally, there are other endosome localized proteins that interact with Rsp5 and Ubp2 that cannot be ignored, for their interaction must serve a purpose. All four proteins - Rup1, Hua1, Rsp5, and Ubp2 - are ubiquitin binding proteins, which suggests a mode for engagement with cargo. We also know Rup1 has a modulatory effect on Rsp5 and Ubp2, based on previously published data (50, 56). It has been suggested that deletion of Ubp2 results in increased recycling of cargo because of autoubiquitination of Rsp5 without competing deubiquitination, thus resulting in decreased ubiquitin ligase activity (56). But our Ubp2 active site mutant data rejects that idea, as the mutant should also cause increased Fur4 recycling, if loss of DUB activity and thus lower amounts of Rsp5 were the explanation. Our EQRS model is bolstered by the UBP2 data. Complete deletion of UBP2 would disrupt the EQRS complex and even ubiquitinated cargo could escape retention. This accounts for the increased recycling observed in cells with ubp2Δ. However, with an active site mutant version of Ubp2, where DUB activity is diminished, the EQRS complex can still form and retain cargo. This accounts for increased degradation seen in ubp2Δ cells complemented with ubp2C745A. Furthermore, Hua1 may have a scaffolding role or be involved in handoff 29 to ESCRT-0, as it is known to physically interact with Hse1 (25). Rup1 could also have a hand in maturation of endosomes. These are just ideas that require further testing, but it could explain why rup1Δ and rup1Δ-hua1Δ cells have cargo stuck in endosomes with delayed progression to vacuoles. When the cells should thoroughly downregulate (i.e., during starvation), Fur4 may be endocytosed normally, but become stuck at endosomes during the ubiquitination-deubiquitination cycle, yet with some escaping back to the plasma membrane (but almost immediately being endocytosed again). It appears Ubp7 does not play a crucial role at the endosomes as does Ubp2. UBP7 deletion alone does not have a strong phenotype. However, combination with UBP2 deletion does appear to increase degradation, so it may be that deletion of the endosomal DUBs drastically reduces deubiquitination and ability to recycle. The microscopy and growth assay data seems reliable since these experiments have been repeated and given the same results. However, it must be noted that the western blot data comes mostly from one blot. A repeat blot showed some of the same trends (but weak) or else no significant trends. So, more blots really need to be performed to get significant data. Also, in our sample preparation procedure, unfortunately, there tends to be degradation occurring during the prep. We have since found a new protocol that should quickly stop all activity in the cell to prevent degradation occurring during the preps. Western blots need to be repeated with this new sample prep procedure, and for the purpose of providing replicate data. 30 Table 2.1: Strains and Plasmids Used in This Study Strains Description Genotype Source SEY 6210 WT MATα leu2-3,112 ura3-52 his3-Δ200 trp1-Δ901 lys2-801 suc2-Δ9 (61) MYY808 rsp5-1 MYY808 MATα, MDM1, smm1, his3, leu2, ura3 (62) MCY64 rup1Δ SEY 6210, rup1::KanMX This study MCY65 ubp7Δ SEY 6210, ubp7::KanMX This study MCY66 ubp2Δ SEY 6210, ubp2::KanMX This study KRY1 Rup1-GFP SEY 6210, Rup1-GFP::KanMX This study KRY2 Ubp2-GFP SEY 6210, Ubp2-GFP::KanMX This study KRY4 hua1Δ SEY 6210, hua1::KanMX This study KRY5-1 ubp2Δubp7Δ SEY 6210, ubp2::TRP1, ubp7::KanMX This study KRY8 hua1Δrup1Δ SEY 6210, hua1::TRP1, rup1::KanMX This study Plasmids pJK19 P(CUP1)-FUR4-GFP URA3 (pRS416) P(CUP1)-FUR4-GFP (35) pKF17 P(UBP2)-UBP2 LEU2 (pRS415) P(UBP2)-UBP2 This study pKF19 P(UBP2)-ubp2C745A LEU2 (pRS415) P(UBP2)-ubp2C745A This study pMB103 P(VPS4)-vps4E233Q URA3(pRS416) P(VPS4)-vps4E233Q (63) Created by M. Babst p415CherryDID2 P(VPS21)-mCherry-DID2 LEU2(pRS415) P(VPS21)-mCherry-DID2 (64) 31 Fig. 2.1: Protein Interaction Map Connecting lines show physical interactions between proteins. 32 Fig. 2.2: EQRS (Endosomal Quality control & Retention System) Model In this model, the green EQRS box represents a complex of the four key proteins - Ubp2, Rsp5, Rup1, and Hua1. A protein (e.g., Fur4) cycles through dynamic deubiquitination and (poly)ubiquitination. Whatever remains ubiquitinated can be caught by the retention complex and held back in the endosome. Without ubiquitin moieties attached, EQRS is not able to retain Fur4 and it escapes, recycling to the PM. Anything held back by EQRS goes along with the maturing endosome to be degraded in the vacuole/lysosome. The EQRS machinery likely is reused by falling off around the late endosome stage as the cargo is handed over to the ESCRTs. 33 Fig. 2.3: Localization of EQRS Proteins The plasma membrane of the cell has been outlined in white dashed line. These cells express a plasmid containing a vps4 dominant negative mutant which causes accumulation of late endosomes. Rup1 or Ubp2 (as indicated) is present where GFP is seen. The ESCRT factor Did2 is present where red mCherry is visible, and is used as a marker of late endosomes. The images show no overlap of GFP and mCherry, revealing that EQRS components accumulate on compartments that are distinct from late endosomes/MVBs. This suggests EQRS functions upstream of the ESCRTs. Additionally, without the vps4 mutation, Rup1 and Ubp2 are more diffuse in the cell, and harder to specifically localize. The fact that they accumulated on late endosomes here suggests that ESCRT factor vps4 may be necessary for recycling of the EQRS components. 34 a. b. Fig. 2.4: Microscopy on rsp5-1 Cells a. Wildtype shows very efficient downregulation of Fur4 after 60 minutes heat shock at 37°C, while Fur4 is stabilized on the PM in rsp5-1 strain before and after heat shock. b. Pictured is Fur4-GFP in rsp5-1 after 60 minutes +uracil. The signal in dots close to the plasma membrane indicates endosomal compartments. Efficiency of endocytosis has been greatly reduced in rsp5-1, but clearly not obliterated. 35 Fig. 2.5: Microscopy on ubp2Δ Cells White dashed lines mark the PM. WT shows normal downregulation after uracil addition. Most GFP signal is in the vacuole, with a little in MVBs. ubp2Δ shows much less GFP in vacuoles than WT, and much more still remaining at the PM, indicative of increased recycling of Fur4. But when complemented with Ubp2 active site mutant, there is a switch to increased degradation. It looks very similar to WT after uracil, with no signal remaining at the PM. Before uracil, the complement shows much more internal signal, so there is increased degradation even at steady state. 36 WT ubp2Δ UBP2 comp.-ubp2Δ ubp2C745A comp.-ubp2Δ Fig. 2.6: 5-FU Growth Assay with ubp2Δ Cells The UBP2 complement and WT grew best, which indicates less Fur4 on the PM (more degradation). upb2Δ was most sensitive to 5-FU, indicating more Fur4 on the PM (more recycling). 37 a. Fig. 2.7: Western Blot Data a. Blot Image: Lanes match in order with the quantification graphs on the following page. The overall trend visible here is a decrease in Fur4 protein after heat shock (every 2nd lane) and lower Fur4 protein amounts overall in the double ubp2Δ-ubp7Δ (lanes 7-10). b. Quantification: -/+ equals before (-) or after (+) heat shock treatment for 20 minutes. 38 b. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Relative Fur4 Degradation Strains - + 0 2 4 6 8 10 12 14 16 18 20 Amount Fur4 (normalized to Snf7) Strains - +39 Fig. 2.8: Microscopy on rup1Δ Cells After 30 minutes +uracil, all GFP signal appears in endosomes/MVB, instead of the majority being in the vacuole as with WT (refer to Fig. 2.4). CHAPTER 3 CONCLUSION 3.1 Concluding Remarks and Future Directions Maintaining functional proteins at the plasma membrane is vital for nutrient uptake, ion exchange for maintenance of membrane potential, and for cell-to-cell communication. Proteins can become unfolded or unstable on the plasma membrane, so a quality control system must be in place to take care of these proteins. Transporters like Fur4 may become unstable just as a byproduct of their natural pumping mechanism. Hence, Fur4 requires monitoring and removal from the plasma membrane to preserve cell integrity. At the same time, cells need to balance integrity with energy conservation, so there is a second QC system in place at the endosome to determine whether a PM protein really needs to be degraded or if the protein is still able to function and should be conserved. If a protein appears unfolded, it is quickly endocytosed to prevent it from being detrimental to the cell, possibly causing loss of membrane potential among other problems. After endocytosis, the protein can either continue down the MVB pathway to be degraded in the lysosome/vacuole, or it can be recycled back to the plasma membrane. The choice between these two fates is made at the early endosome. How this decision is made is incompletely understood. A widely 41 accepted idea in the field is that ubiquitination status determines a protein's fate, and that the processes of ubiquitination and deubiquitination occurring at the early endosome control that fate. However, numerous experiments on ubp2Δ strains suggest a more complex system of quality control at the early endosome. It's clear that deletion of UBP2 causes increased recycling of proteins to the plasma membrane, the polar opposite phenotype expected. So how can this be explained? That is how the EQRS model was born, to explain these unexpected results. There are many open questions and theories that should be explored through further research of this topic. A future goal is to understand how the retention complex is organized and its timing. What are the functional roles of all the key players in our EQRS model? We still do not know what Hua1 does. Which components are essential? Ubp7 likely is not essential since its deletion did not have an obvious phenotype, and there is probably some redundancy with Ubp2. How are the EQRS factors recruited to the endosome? At what point does EQRS hand off cargo to the ESCRTs? We believe EQRS functions upstream of the ESCRTs, so there is likely a step where EQRS leaves prior to or in conjunction with ESCRT recruitment. Answering these questions will help us understand the vital second quality control system at work at the early endosome. In spite of leaving us with many questions, the present work has added support to our EQRS model for the most part, leaving us with many future directions in which to go. REFERENCES 1. M. Babst, Quality control: quality control at the plasma membrane: one mechanism does not fit all. J. Cell Biol. 205, 11-20 (2014). 2. J. L. Brodsky, Cleaning up: ER-associated degradation to the rescue. Cell. 151, 1163-7 (2012). 3. R. Hampton, ER-associated degradation in protein quality control and cellular regulation. Curr. Opin. Cell Biol., 476-482 (2002). 4. M. S. Hipp, S. H. Park, F. U. Hartl, Proteostasis impairment in protein-misfolding and -aggregation diseases. Trends Cell Biol. 24, 506-514 (2014). 5. K. Cuanalo-Contreras, A. Mukherjee, C. Soto, Role of protein misfolding and proteostasis deficiency in protein misfolding diseases and aging. Int. J. Cell Biol. 2013, 1-10 (2013). 6. W. Scheper, D. a T. Nijholt, J. J. M. Hoozemans, The unfolded protein response and proteostasis in Alzheimer disease: Preferential activation of autophagy by endoplasmic reticulum stress. Autophagy. 7, 910-911 (2011). 7. G. L. Lukacs, a. S. Verkman, CFTR: Folding, misfolding and correcting the DF508 conformational defect. Trends Mol. Med. 18, 81-91 (2012). 8. C. L. Ward, R. R. Kopito, Intracellular turnover of cystic fibrosis transmembrane conductance regulator. Inefficient processing and rapid degradation of wild-type and mutant proteins. J. Biol. Chem. 269, 25710-25718 (1994). 9. G. L. Lukacs et al., The DF508 mutation decreases the stability of cystic fibrosis transmembrane conductance regulator in the plasma membrane. J Biol Chem. 268, 21592-21598 (1993). 10. A. Swiatecka-Urban et al., The short apical membrane half-life of rescued ΔF508-cystic fibrosis transmembrane conductance regulator (CFTR) results from accelerated endocytosis of ΔF508H-CFTR in polarized human airway epithelial cells. J. Biol. Chem. 280, 36762-36772 (2005). 43 11. B. Alberts et al., in Molecular Biology of the Cell, M. Anderson, S. Granum, Eds. (Garland Science, New York, NY, ed. 5th, 2007), pp. 695-748. 12. B. Alberts et al., in Molecular Biology of the Cell, M. Anderson, S. Granum, Eds. (Garland Science, New York, NY, ed. 5th, 2007), pp. 749-812. 13. L. Hicke, H. Riezman, Ubiquitination of a yeast plasma membrane receptor signals its ligand-stimulated endocytosis. Cell. 84, 277-287 (1996). 14. M. Jovic, M. Sharma, J. Rahajeng, S. Caplan, The early endosome: A busy sorting station for proteins at the crossroads. Histol. Histopathol. 25, 99-112 (2010). 15. E. Lauwers, Z. Erpapazoglou, R. Haguenauer-Tsapis, B. André, The ubiquitin code of yeast permease trafficking. Trends Cell Biol. 20, 196-204 (2010). 16. J. M. Galan, V. Moreau, B. Andre, C. Volland, R. Haguenauer-Tsapis, Ubiquitination mediated by the Npi1p/Rsp5p ubiquitin-protein ligase is required for endocytosis of the yeast uracil permease. J. Biol. Chem. 271, 10946-10952 (1996). 17. M. J. Clague, H. Liu, S. Urbé, Governance of endocytic trafficking and signaling by reversible ubiquitylation. Dev. Cell. 23, 457-67 (2012). 18. J. H. Hurley, The ESCRT complexes. Crit. Rev. Biochem. Mol. Biol. 45, 463-487 (2010). 19. R. C. Piper, D. J. Katzmann, Biogenesis and function of multivesicular bodies. Annu. Rev. Cell Dev. Biol. 23, 519-547 (2007). 20. a Y. Amerik, J. Nowak, S. Swaminathan, M. Hochstrasser, The Doa4 deubiquitinating enzyme is functionally linked to the vacuolar protein-sorting and endocytic pathways. Mol. Biol. Cell. 11, 3365-3380 (2000). 21. N. Luhtala, G. Odorizzi, Bro1 coordinates deubiquitination in the multivesicular body pathway by recruiting Doa4 to endosomes. J. Cell Biol. 166, 717-729 (2004). 22. A. Ciechanover, Proteolysis: from the lysosome to ubiquitin and the proteasome. Nat. Rev. Mol. Cell Biol. 6, 79-87 (2005). 23. L. Hicke, R. Dunn, Regulation of membrane protein transport by ubiquitin and ubiquitin-binding proteins. Annu. Rev. Cell Dev. Biol. 19, 141-172 (2003). 24. J. Terrell, S. Shih, R. Dunn, L. Hicke, A function for monoubiquitination in 44 the internalization of a G protein-coupled receptor. Mol. Cell. 1, 193-202 (1998). 25. J. Ren, Y. Kee, J. M. Huibregtse, R. C. Piper, Hse1, a Component of the Yeast Hrs-STAM Ubiquitin-sorting Complex, Associates with Ubiquitin Peptidases and a Ligase to Control Sorting Efficiency into Multivesicular Bodies. Mol. Biol. Cell. 18, 324-335 (2007). 26. Y. Ye, M. Rape, Building ubiquitin chains: E2 enzymes at work. Nat. Rev. Mol. Cell Biol. 10, 755-764 (2009). 27. S. Urbé, Ubiquitin and endocytic protein sorting. Essays Biochem. 41, 81-98 (2005). 28. S. Jentsch, The ubiquitin-conjugation system. Annu. Rev. Genet., 179-207 (1992). 29. N. Belgareh-Touzé et al., Versatile role of the yeast ubiquitin ligase Rsp5p in intracellular trafficking. Biochem. Soc. Trans. 36, 791-796 (2008). 30. J. Horák, The role of ubiquitin in down-regulation and intracellular sorting of membrane proteins: Insights from yeast. Biochim. Biophys. Acta - Biomembr. 1614, 139-155 (2003). 31. C. Hein, J.-Y. Springael, C. Volland, R. Haguenauer-Tsapis, B. André, NPI1, an essential yeast gene involved in induced degradation of Gap1 and Fur4 permeases, encodes the Rsp5 ubiquitin-protein ligase. Mol. Microbiol. 18, 77-87 (1995). 32. G. Wang et al., Localization of the Rsp5p Ubiquitin-Protein Ligase at Multiple Sites within the Endocytic Pathway. Mol. Cell. Biol. 21, 3564-3575 (2001). 33. G. Wang, J. Yang, J. M. Huibregtse, Functional domains of the Rsp5 ubiquitin-protein ligase. Mol. Cell. Biol. 19, 342-352 (1999). 34. J. Liu, A. Sitaram, C. G. Burd, Regulation of copper-dependent endocytosis and vacuolar degradation of the yeast copper transporter, Ctr1p, by the Rsp5 ubiquitin ligase. Traffic. 8, 1375-1384 (2007). 35. J. M. Keener, M. Babst, Quality Control and Substrate-Dependent Downregulation of the Nutrient Transporter Fur4. Traffic. 14, 412-427 (2013). 36. C. Marchal, R. Haguenauer-Tsapis, D. Urban-Grimal, Casein kinase 1-dependent phosphorylation within a PEST sequence and ubiquitination nearby lysines signal endocytosis of yeast uracil permease. J. Biol. Chem. 45 275, 23608-23614 (2000). 37. W. Dubiel, C. Gordon, Ubiquitin pathway: Another link in the polyubiquitin chain? Curr. Biol. 9, 554-557 (1999). 38. C. M. Pickart, D. Fushman, Polyubiquitin chains: Polymeric protein signals. Curr. Opin. Chem. Biol. 8, 610-616 (2004). 39. Z. Erpapazoglou et al., A dual role for K63-linked ubiquitin chains in multivesicular body biogenesis and cargo sorting. Mol. Biol. Cell. 23, 2170-2183 (2012). 40. P. Xu et al., Quantitative Proteomics Reveals the Function of Unconventional Ubiquitin Chains in Proteasomal Degradation. Cell. 137, 133-145 (2009). 41. P. Hopkins, N. R. Chevallier, R. Jund, A. A. Eddy, Use of plasmid vectors to show that the uracil and cytosine permeases of the yeast Saccharomyces cerevisiae are electrogenic proton symports. FEMS Microbiol. Lett. 49, 173-177 (1988). 42. S. Weyand et al., Structure and molecular mechanism of a nucleobase-cation-symport-1 family transporter. Science. 322, 709-713 (2008). 43. R. Jund, E. Weber, M.-R. C. Chevallier, Primary structure of the uracil transport protein of Saccharomyces cerevisiae. Eur. J. Biochem. 171, 417-24 (1988). 44. T. Shimamura et al., Molecular basis of alternating access membrance transport by the sodium hydantoin transporter Mhp1. Science. 328, 470-473 (2010). 45. T. Ravid, M. Hochstrasser, Diversity of degradation signals in the ubiquitin-proteasome system. Nat. Rev. Mol. Cell Biol. 9, 679-690 (2008). 46. C. Marchal, R. Haguenauer-Tsapis, D. Urban-Grimal, A PEST-like sequence mediates phosphorylation and efficient ubiquitination of yeast uracil permease. Mol. Cell. Biol. 18, 314-321 (1998). 47. M. O. Blondel et al., Direct Sorting of the Yeast Uracil Permease to the Endosomal System Is Controlled by Uracil Binding and Rsp5p-dependent Ubiquitylation. Mol. Biol. Cell. 15, 883-895 (2004). 48. K. Séron, M. O. Blondel, R. Haguenauer-Tsapis, C. Volland, Uracil-induced down-regulation of the yeast uracil permease. J. Bacteriol. 181, 1793-1800 (1999). 46 49. C. B. Jones et al., Regulation of Membrane Protein Degradation by Starvation-Response Pathways. Traffic. 13, 468-482 (2012). 50. Y. Kee, N. Lyon, J. M. Huibregtse, The Rsp5 ubiquitin ligase is coupled to and antagonized by the Ubp2 deubiquitinating enzyme. EMBO J. 24, 2414-2424 (2005). 51. M. H. Y. Lam et al., Interaction of the deubiquitinating enzyme Ubp2 and the E3 ligase Rsp5 is required for transporter/receptor sorting in the multivesicular body pathway. PLoS One. 4 (2009), doi:10.1371/journal.pone.0004259. 52. S. Urbe et al., Systematic survey of deubiquitinase localization identifies USP21 as a regulator of centrosome- and microtubule-associated functions. Mol. Biol. Cell. 23, 1095-1103 (2012). 53. A. Y. Amerik, S.-J. Li, M. Hochstrasser, Analysis of the Deubiquitinating Enzymes of the Yeast Saccharomyces cerevisiae. Biol. Chem. 381, 981-992 (2005). 54. J. S. Weinberg, D. G. Drubin, Regulation of Clathrin-Mediated Endocytosis by Dynamic Ubiquitination and Deubiquitination. Curr. Biol. 24, 951-959 (2014). 55. F. R. Maxfield, T. E. McGraw, Endocytic recycling. Nat. Rev. Mol. Cell Biol. 5, 121-132 (2004). 56. M. H. Y. Lam, A. Emili, Ubp2 Regulates Rsp5 Ubiquitination Activity In Vivo and In Vitro. PLoS One. 8 (2013), doi:10.1371/journal.pone.0075372. 57. S. D. Stamenova, R. Dunn, A. S. Adler, L. Hicke, The Rsp5 Ubiquitin Ligase Binds to and Ubiquitinates Members of the Yeast CIN85-Endophilin Complex, Sla1-Rvs167. J. Biol. Chem. 279, 16017-16025 (2004). 58. Y. Kee, W. Muñoz, N. Lyon, J. M. Huibregtse, The deubiquitinating enzyme Ubp2 modulates Rsp5-dependent Lys 63-linked polyubiquitin conjugates in Saccharomyces cerevisiae. J. Biol. Chem. 281, 36724-36731 (2006). 59. S. S. Cohen, J. G. Flaks, H. D. Barner, M. R. Loeb, J. Lichtenstein, The Mode of Action of 5-Fluorouracil and Its Derivatives. Proc. Natl. Acad. Sci. U. S. A. 44, 1004-1012 (1958). 60. R. T. Baker, J. W. Tobias, A. Varshavsky, Ubiquitin-specific proteases of Saccharomyces cerevisiae: Cloning of UBP2 and UBP3, and functional analysis of the UBP gene family. J. Biol. Chem. 267, 23364-23375 (1992). 61. J. S. Robinson, D. J. Klionsky, L. M. Banta, S. D. Emr, Protein sorting in 47 Saccharomyces cerevisiae: isolation of mutants defective in the delivery and processing of multiple vacuolar hydrolases. Mol. Cell. Biol. 8, 4936-4948 (1988). 62. H. a. Fisk, M. P. Yaffe, A role for ubiquitination in mitochondrial inheritance in Saccharomyces cerevisiae. J. Cell Biol. 145, 1199-1208 (1999). 63. S. Kametaka et al., Canonical interaction of cyclin G associated kinase with adaptor protein 1 regulates lysosomal enzyme sorting. Mol. Biol. Cell. 18, 2991-3001 (2007). 64. B. A. Davies et al., Coordination of substrate binding and ATP hydrolysis in Vps4-mediated ESCRT-III disassembly. Mol. Biol. Cell. 21, 3396-3408 (2010).