Summary Selective protein degradation by the ubiquitin-proteasome system frequently involves substrate recognition via short linear motifs known as degradation signals or degrons. Whereas degrons located at protein N-termini (N-degrons) have been extensively studied, our understanding of C-terminal degrons (C-degrons) is comparatively limited. Previously, we showed that the yeast SCF ubiquitin ligase and one of its substrate receptor subunits, the F-box protein Das1, target a broad range of C-degrons and implicated SCF Das1 in orphan quality control. Here, we sought to determine how Das1 recognizes its substrates. By combining in vivo competition assays with structural modeling and mutational analysis, we demonstrate that distinct C-degrons compete for a common site on Das1, indicating a shared mode of recognition. We identify a positively charged pocket within the Das1 leucine-rich repeat domain as the C-degron binding site. Three basic residues at the base of this pocket are essential for Das1 function, likely mediating electrostatic interactions with the C-terminal carboxyl group of the degron, while additional residues contribute to substrate specificity. Comparative analysis reveals that this pocket and its function are conserved across the Saccharomycetaceae family, supporting a conserved role for Das1 in protein quality control.

The GID/CTLH ubiquitin ligase is a multisubunit E3 conserved across eukaryotes. GID/CTLH has been implicated in a variety of processes, including metabolic regulation, cell proliferation, embryonic development and cell differentiation. However, our understanding of substrate recognition by GID/CTLH remains incomplete. Here, we characterize the regulation, specificity and functions of the putative substrate receptor Gid11 in budding yeast. We find that upon its expression during the switch of carbon source from glucose to ethanol, Gid11 associates with the core GID complex, likely forming a large ring assembly. Through systematic mutagenesis, we identify a negatively charged pocket on Gid11 that binds substrates carrying N-terminal degrons starting with threonine and define the Thr/N-degron motif. Such Thr/N-degrons are exposed upon co-translational removal of the initiator methionine, while escaping N-terminal acetylation. Finally, we show that during the switch to ethanol, GIDGid11 targets for degradation factors linked to glycolysis, opposing the established role of GID in the degradation of gluconeogenic enzymes during the reverse switch from ethanol to glucose. Taken together, our results establish Gid11 as a bona fide GID substrate receptor and highlight GID as a regulator of metabolic transitions. ### Competing Interest Statement The authors have declared no competing interest.

Functional genomics with libraries of knockout alleles is limited to non-essential genes and convoluted by the potential accumulation of suppressor mutations in knockout backgrounds, which can lead to erroneous functional annotations. To address these limitations, we constructed genome-wide libraries of conditional alleles based on the auxin-inducible degron (AID) system for inducible degradation of AID-tagged proteins in the budding yeast Saccharomyces cerevisiae. First, we determined that N-terminal tagging is at least twice as likely to inadvertently impair protein function across the proteome. We thus constructed two libraries with over 5,600 essential and non-essential proteins fused at the C-terminus with an AID tag and an optional fluorescent protein. Approximately 90% of AID-tagged proteins were degraded in the presence of the auxin analog 5-Ph-IAA, with initial protein abundance and tag accessibility as limiting factors. Genome-wide screens for DNA damage response factors revealed a role for the glucose signaling factor GSF2 in resistance to hydroxyurea, highlighting how the AID libraries extend the yeast genetics toolbox.

The chitin synthase Chs3 is a multipass membrane protein whose trafficking is tightly controlled. Accordingly, its exit from the endoplasmic reticulum (ER) depends on several complementary mechanisms that ensure its correct folding. Despite its potential failure on its exit, Chs3 is very stable in this compartment, which suggests its poor recognition by ER quality control mechanisms such as endoplasmic reticulum-associated degradation (ERAD). Here we show that proper N-glycosylation of its luminal domain is essential to prevent the aggregation of the protein and its subsequent recognition by the Hrd1-dependent ERAD-L machinery. In addition, the interaction of Chs3 with its chaperone Chs7 seems to mask additional cytosolic degrons, thereby avoiding their recognition by the ERAD-C pathway. On top of that, Chs3 molecules that are not degraded by conventional ERAD can move along the ER membrane to reach the inner nuclear membrane, where they are degraded by the inner nuclear membrane-associated degradation (INMAD) system, which contributes to the intracellular homeostasis of Chs3. These results indicate that Chs3 is an excellent model to study quality control mechanisms in the cell and reinforce its role as a paradigm in intracellular trafficking research.

Single ribonucleoside monophosphates (rNMPs) are transiently present in eukaryotic genomes. The RNase H2-dependent ribonucleotide excision repair (RER) pathway ensures error-free rNMP removal. In some pathological conditions, rNMP removal is impaired. If these rNMPs hydrolyze during, or prior to, S phase, toxic single-ended double-strand breaks (seDSBs) can occur upon an encounter with replication forks. How such rNMP-derived seDSB lesions are repaired is unclear. We expressed a cell cycle phase restricted allele of RNase H2 to nick at rNMPs in S phase and study their repair. Although Top1 is dispensable, the RAD52 epistasis group and Rtt101Mms1-Mms22 dependent ubiquitylation of histone H3 become essential for rNMP-derived lesion tolerance. Consistently, loss of Rtt101Mms1-Mms22 combined with RNase H2 dysfunction leads to compromised cellular fitness. We refer to this repair pathway as nick lesion repair (NLR). The NLR genetic network may have important implications in the context of human pathologies.

Selective degradation of unnecessary or abnormal proteins by the ubiquitin-proteasome system is an essential part of proteostasis. Ubiquitin ligases recognize substrates of selective protein degradation and modify them with polyubiquitin chains, which mark them for proteasomal degradation. Substrate recognition by ubiquitin ligases often involves degradation signals or degrons, which are typically short linear motifs found in intrinsically disordered regions, e.g., at protein termini. However, specificity in selective protein degradation is generally not well understood, as for most ubiquitin ligases no degrons have been identified thus far. To address this limitation, high-throughput mutagenesis approaches, such as multiplexed protein stability (MPS) profiling, have been developed, enabling systematic surveys of degrons in vivo or allowing to define degron motifs recognized by different ubiquitin ligases. In MPS profiling, thousands of short peptides can be assessed in parallel for their ability to trigger degradation of a fluorescent timer reporter. Here, we describe common types of libraries used to identify and dissect degrons located at protein termini using MPS profiling in budding yeast, and provide protocols for their construction.

Faithful chromosome segregation in budding yeast requires correct positioning of the mitotic spindle along the mother to daughter cell polarity axis. When the anaphase spindle is not correctly positioned, a surveillance mechanism, named as the spindle position checkpoint (SPOC), prevents the progression out of mitosis until correct spindle positioning is achieved. How SPOC works on a molecular level is not well understood. Here we performed a genome-wide genetic screen to search for components required for SPOC. We identified the SWR1 chromatin-remodeling complex (SWR1-C) among several novel factors that are essential for SPOC integrity. Cells lacking SWR1-C were able to activate SPOC upon spindle misorientation but underwent mitotic slippage upon prolonged SPOC arrest. This mitotic slippage required the Cdc14-early anaphase release pathway and other factors including the SAGA (Spt-Ada-Gcn5 acetyltransferase) histone acetyltransferase complex, proteasome components and the mitotic cyclin-dependent kinase inhibitor Sic1. Together, our data establish a novel link between SWR1-C chromatin remodeling and robust checkpoint arrest in late anaphase.

N-terminal protein sequences and their proteolytic processing and modifications influence the stability and turnover of proteins by creating potential degrons for cellular proteolytic pathways. Understanding the impact of genetic perturbations of components affecting the processing of protein N-termini and thereby their stability, requires methods compatible with proteome-wide studies of many N-termini simultaneously. Tandem fluorescent timers (tFT) allow the in vivo measurement of protein turnover completely independent of protein abundance and can be deployed for proteome-wide studies. Here we present a protocol for Multiplexed Protein Stability (MPS) profiling of tFT-libraries encoding large numbers of different protein N-termini fused to tFT in the yeast Saccharomyces cerevisiae. This protocol includes fluorescence cell sorting based profiling of these libraries using a pooling approach. Analysis of the sorted pools is done by using multiplexed deep sequencing, in order to generate a stability index for each N-terminally peptide fused to the tFT reporter, and to evaluate half-life changes across all species represented in the library.

Selective protein degradation typically involves substrate recognition via short linear motifs known as degrons. Various degrons can be found at protein termini from bacteria to mammals. While N-degrons have been extensively studied, our understanding of C-degrons is still limited. Towards a comprehensive understanding of eukaryotic C-degron pathways, here we perform an unbiased survey of C-degrons in budding yeast. We identify over 5000 potential C-degrons by stability profiling of random peptide libraries and of the yeast C‑terminome. Combining machine learning, high-throughput mutagenesis and genetic screens reveals that the SCF ubiquitin ligase targets ~40% of degrons using a single F-box substrate receptor Das1. Although sequence-specific, Das1 is highly promiscuous, recognizing a variety of C-degron motifs. By screening for full-length substrates, we implicate SCF Das1 in degradation of orphan protein complex subunits. Altogether, this work highlights the variety of C-degron pathways in eukaryotes and uncovers how an SCF/C-degron pathway of broad specificity contributes to proteostasis.

Tandem fluorescent protein timers (tFTs) are versatile reporters of protein dynamics. A tFT consists of two fluorescent proteins with different maturation kinetics and provides a ratiometric readout of protein age, which can be exploited to follow intracellular trafficking, inheritance and turnover of tFT-tagged proteins. Here, we detail a protocol for high-throughput analysis of protein turnover with tFTs in yeast using fluorescence measurements of ordered colony arrays. We describe guidelines on optimization of experimental design with regard to the layout of colony arrays, growth conditions, and instrument choice. Combined with semi-automated genetic crossing using synthetic genetic array (SGA) methodology and high-throughput protein tagging with SWAp-Tag (SWAT) libraries, this approach can be used to compare protein turnover across the proteome and to identify regulators of protein turnover genome-wide.

A healthy and functional proteome is essential to cell physiology. However, this is constantly being challenged as most steps of protein metabolism are error-prone and changes in the physico-chemical environment can affect protein structure and function, thereby disrupting proteome homeostasis. Among a variety of potential mistakes, proteins can be targeted to incorrect compartments or subunits of protein complexes may fail to assemble properly with their partners, resulting in the formation of mislocalized and orphan proteins, respectively. Quality control systems are in place to handle these aberrant proteins, and to minimize their detrimental impact on cellular functions. Here, we discuss recent findings on quality control mechanisms handling mislocalized and orphan proteins. We highlight common principles involved in their recognition and summarize how accumulation of these aberrant molecules is associated with aging and disease.

Inactivation of N-terminal acetyltransferase A is found to alter Rpn4 as well as E3 ligase abundance, causing up-regulation of Ubiquitin–proteasome activity. In this context, Tom1 is also identified as a novel chain-elongating enzyme of the UFD-pathway. N-terminal acetylation is a prominent protein modification, and inactivation of N-terminal acetyltransferases (NATs) cause protein homeostasis stress. Using multiplexed protein stability profiling with linear ubiquitin fusions as reporters for the activity of the ubiquitin proteasome system, we observed increased ubiquitin proteasome system activity in NatA, but not NatB or NatC mutants. We find several mechanisms contributing to this behavior. First, NatA-mediated acetylation of the N-terminal ubiquitin–independent degron regulates the abundance of Rpn4, the master regulator of the expression of proteasomal genes. Second, the abundance of several E3 ligases involved in degradation of UFD substrates is increased in cells lacking NatA. Finally, we identify the E3 ligase Tom1 as a novel chain-elongating enzyme (E4) involved in the degradation of linear ubiquitin fusions via the formation of branched K11, K29, and K48 ubiquitin chains, independently of the known E4 ligases involved in UFD, leading to enhanced ubiquitination of the UFD substrates.

Selective protein degradation by the ubiquitin-proteasome system (UPS) is involved in all cellular processes. However, the substrates and specificity of most UPS components are not well understood. Here we systematically characterized the UPS in Saccharomyces cerevisiae. Using fluorescent timers, we determined how loss of individual UPS components affects yeast proteome turnover, detecting phenotypes for 76% of E2, E3, and deubiquitinating enzymes. We exploit this dataset to gain insights into N-degron pathways, which target proteins carrying N-terminal degradation signals. We implicate Ubr1, an E3 of the Arg/N-degron pathway, in targeting mitochondrial proteins processed by the mitochondrial inner membrane protease. Moreover, we identify Ylr149c/Gid11 as a substrate receptor of the glucose-induced degradation-deficient (GID) complex, an E3 of the Pro/N-degron pathway. Our results suggest that Gid11 recognizes proteins with N-terminal threonines, expanding the specificity of the GID complex. This resource of potential substrates and relationships between UPS components enables exploring functions of selective protein degradation.

Pooled genetic screening is a powerful method to systematically link genotype to phenotype and gain insights into biological processes, but applying it to visual phenotypes such as cell morphology or protein localization has remained a challenge. In their recent work, Fowler and colleagues (Hasle et al, 2020) describe an elegant approach for high‐throughput cell sorting according to visual phenotypes based on selective photoconversion. This allows combining the advantages of high‐content phenotyping by fluorescence microscopy with the efficiency of pooled screening to dissect complex cellular processes.

Polyubiquitin chains linked via lysine (K) 63 play an important role in endocytosis and membrane trafficking. Their primary source is the ubiquitin protein ligase (E3) Rsp5/NEDD4, which acts as a key regulator of membrane protein sorting. The heterodimeric ubiquitin-conjugating enzyme (E2), Ubc13-Mms2, catalyses K63-specific polyubiquitylation in genome maintenance and inflammatory signalling. In budding yeast, the only E3 proteins known to cooperate with Ubc13-Mms2 so far is a nuclear RING finger protein, Rad5, involved in the replication of damaged DNA. Here, we report a contribution of Ubc13-Mms2 to the sorting of membrane proteins to the yeast vacuole via the multivesicular body (MVB) pathway. In this context, Ubc13-Mms2 cooperates with Pib1, a FYVE-RING finger protein associated with internal membranes. Moreover, we identified a family of membrane-associated FYVE(type)-RING finger proteins as cognate E3 proteins for Ubc13-Mms2 in several species, and genetic analysis indicates that the contribution of Ubc13-Mms2 to membrane trafficking in budding yeast goes beyond its cooperation with Pib1. Thus, our results widely implicate Ubc13-Mms2 as an Rsp5-independent source of K63-linked polyubiquitin chains in the regulation of membrane protein sorting. This article has an associated First Person interview with the first author of the paper.

Whole-genome duplication has played a central role in the genome evolution of many organisms, including the human genome. Most duplicated genes are eliminated, and factors that influence the retention of persisting duplicates remain poorly understood. We describe a systematic complex genetic interaction analysis with yeast paralogs derived from the whole-genome duplication event. Mapping of digenic interactions for a deletion mutant of each paralog, and of trigenic interactions for the double mutant, provides insight into their roles and a quantitative measure of their functional redundancy. Trigenic interaction analysis distinguishes two classes of paralogs: a more functionally divergent subset and another that retained more functional overlap. Gene feature analysis and modeling suggest that evolutionary trajectories of duplicated genes are dictated by combined functional and structural entanglement factors.

Article Figures and data Abstract Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Tail-anchored (TA) proteins insert post-translationally into the endoplasmic reticulum (ER), the outer mitochondrial membrane (OMM) and peroxisomes. Whereas the GET pathway controls ER-targeting, no dedicated factors are known for OMM insertion, posing the question of how accuracy is achieved. The mitochondrial AAA-ATPase Msp1 removes mislocalized TA proteins from the OMM, but it is unclear, how Msp1 clients are targeted for degradation. Here we screened for factors involved in degradation of TA proteins mislocalized to mitochondria. We show that the ER-associated degradation (ERAD) E3 ubiquitin ligase Doa10 controls cytoplasmic level of Msp1 clients. Furthermore, we identified the uncharacterized OMM protein Fmp32 and the ectopically expressed subunit of the ER-mitochondria encounter structure (ERMES) complex Gem1 as native clients for Msp1 and Doa10. We propose that productive localization of TA proteins to the OMM is ensured by complex assembly, while orphan subunits are extracted by Msp1 and eventually degraded by Doa10. https://doi.org/10.7554/eLife.45506.001 Introduction Correct localization of proteins is essential to ensure their functionality and to establish the identity of individual cellular organelles. Most proteins are synthesized in the cytosol and targeted to subcellular compartments either co-translationally or after their synthesis is completed, post-translationally (Aviram and Schuldiner, 2017; Hegde and Keenan, 2011; Wasilewski et al., 2017). The signal recognition particle (SRP) recognizes N-terminal signal sequences of ER-targeted nascent polypeptides and recruits the ribosome-nascent chain complex to the Sec61 translocon (Voorhees and Hegde, 2016). Membrane proteins that are not recognized by SRP are captured by other cytosolic chaperons, which keep these proteins in an unfolded state guiding them to the ER, mitochondria or peroxisomes post-translationally (Aviram and Schuldiner, 2017; Hegde and Keenan, 2011; Wasilewski et al., 2017). Tail-anchored (TA) proteins are a specific class of membrane proteins that have a single transmembrane (TM) domain at their very C-terminus. They are involved in various cellular processes such as membrane fusion, protein translocation and regulation of apoptosis (Antonsson, 2001; Beilharz et al., 2003; Burri and Lithgow, 2004; Chen and Scheller, 2001; Kalbfleisch et al., 2007). The GET pathway (guided entry of TA proteins) ensures targeting of TA proteins to the ER (Hegde and Keenan, 2011). In brief, a cytosolic pre-targeting complex comprising Sgt2, Get4, and Get5 captures the TM segment of a TA protein after it emerges from the ribosome and loads it onto Get3 (Mariappan et al., 2010; Mateja et al., 2009; Schuldiner et al., 2008). This cytosolic ATPase targeting factor is subsequently recruited to the ER membrane via the Get1-Get2 receptor complex where, after ATP hydrolysis, the TA protein is inserted into the membrane (Mariappan et al., 2011; Schuldiner et al., 2008; Wang et al., 2011). Recently, the EMC complex (ER membrane protein complex) was shown to ensure ER targeting of a subset of TA proteins in mammalian cells (Guna et al., 2018). However, no dedicated pathway targeting TA proteins to the outer mitochondrial membrane (OMM) has been identified so far and it is unclear how they insert into the lipid bilayer (Vitali et al., 2018). From in vitro studies, it has been proposed that because of the low ergosterol content of the OMM TA proteins can insert unassisted (Kemper et al., 2008). Given these multiple, possibly overlapping, targeting mechanisms, insertion of TA proteins appears to be intrinsically prone to failure. Moreover, uncontrolled mitochondrial targeting requires protein quality control systems to remove mistargeted and surplus TA proteins. To ensure protein homeostasis (proteostasis), organelle-specific safeguards recognize orphan complex subunits, damaged and mistargeted protein species and direct them for degradation by the ubiquitin-proteasome system (Juszkiewicz and Hegde, 2018). In the ER, this is linked to polytopic E3 ubiquitin ligases that, in concert with a diverse set of additional factors, constitute the ER-associated degradation (ERAD) pathway (Avci and Lemberg, 2015; Mehrtash and Hochstrasser, 2018; Ruggiano et al., 2014). In yeast, three major branches of ERAD are centered around the polytopic E3 ubiquitin ligases Hrd1, Doa10 and Asi1/3 (Bays et al., 2001; Carvalho et al., 2006; Deak and Wolf, 2001; Foresti et al., 2014; Khmelinskii et al., 2014; Swanson et al., 2001). While Hrd1 is specific for proteins with misfolded ER luminal or aberrant TM domains, Doa10 primarily mediates degradation of ER proteins with cytosolic lesions (Carvalho et al., 2006). The Asi1/3 complex removes misfolded or mislocalized membrane proteins from the inner nuclear membrane, a sub-area of the ER (Foresti et al., 2014; Khmelinskii et al., 2014). In addition, Doa10 encompasses broader functionality by targeting soluble proteins for proteasomal degradation and was suggested to perform a quality control of proteins that fail membrane insertion (Ast et al., 2014; Kats et al., 2018; Maurer et al., 2016; Swanson et al., 2001). In this context, Doa10 triggers degradation of proteins with hydrophobic sequences at either end of the protein, such as hydrophobic C-terminal pro-domains of glycosylphosphatidylinositol (GPI)-anchored proteins (Ast et al., 2014) or hydrophobic N-termini (Kats et al., 2018). For retro-translocation and extraction of membrane-embedded ERAD substrates the required energy is provided by the AAA-ATPase Cdc48 (Ye et al., 2001). Cdc48 substrate recognition is mediated through its Ufd1-Npl4 cofactor, which binds ubiquitinated proteins (Meyer et al., 2000). While this process is best understood in the ER, analogous extraction pathways have been described for mitochondria, the Golgi complex and peroxisomes (Avci and Lemberg, 2015; Chen et al., 2014; Okreglak and Walter, 2014; Schmidt et al., 2019; Stewart et al., 2012). Likewise, Ubx2 which was thought to be an ERAD-specific Cdc48 adaptor, has been shown to also bind to the TOM (translocase of the outer membrane) complex, defining a mitochondrial protein translocation-associated degradation pathway (Mårtensson et al., 2019). In a related process referred to as mitochondria-associated degradation, Doa1 or Vms1 recruit the Cdc48 complex to OMM-anchored ubiquitinated proteins to target them for proteasomal degradation (Anton et al., 2011; Heo et al., 2010; Karbowski and Youle, 2011; Wu et al., 2016). Whereas this pathway is linked to the ubiquitin-proteasome system through the cytoplasmic E3 ubiquitin ligases Rsp5 and Mdm30, an alternative extraction mechanism relies on the AAA-ATPase Msp1 (known as ATAD1/Thorase in humans). Msp1 shows dual localization to the OMM and peroxisomes (Chen et al., 2014; Okreglak and Walter, 2014). In peroxisomes it has been suggested, that the TA protein Pex15 evades Msp1-dependent extraction by interacting with other peroxisomal proteins such as Pex3 (Weir et al., 2017). Deletion of Msp1 leads to stabilization of mistargeted peroxisomal Pex15 or the Golgi v-SNARE Gos1 in the OMM (Chen et al., 2014; Okreglak and Walter, 2014). In vitro studies suggest that Msp1 is sufficient to extract TA proteins from proteoliposomes (Wohlever et al., 2017). Although these reconstitution experiments showed that the Msp1 N-terminal TM domain is dispensable for the dislocation reaction, recently Li et al. suggested that a conserved negatively charged aspartate residue in the N-terminal portion of Msp1 facing the inter membrane space (IMS) is important for efficient recognition of Msp1 clients in vivo (Li et al., 2019). Moreover, a hydrophobic surface of the AAA-ATPase domain has been implicated in recognition of Msp1 clients (Li et al., 2019). However, no clear sequence consensus between all known Msp1 clients exists (Chen et al., 2014; Li et al., 2019; Okreglak and Walter, 2014) and the mechanism how Msp1 distinguishes between mislocalized and properly located TA proteins is ill defined. This, taken together with the finding that Cis1 can recruit Msp1 to TOM for degradation of stalled import intermediates (Weidberg and Amon, 2018), indicates that multiple routes can target proteins for Msp1-mediated dislocation. Furthermore, the fate of TA proteins extracted in the cytoplasm by Msp1 is unclear. To address these issues, we screened for additional factors important for clearing mistargeted TA proteins from the OMM. Our findings outline a previously unanticipated role of the ER-resident E3 ubiquitin ligase Doa10 as a major regulator in the control of OMM TA protein targeting. Results Screen for factors involved in turnover of mistargeted Pex15Δ30 To identify factors involved in degradation of mistargeted TA proteins in yeast, we studied a Pex15 variant lacking the 30 C-terminal residues (Pex15Δ30). Whereas wild type (wt) Pex15 localizes to peroxisomes, Pex15Δ30 constitutively mislocalizes to the OMM and exhibits Msp1-dependent turnover (Okreglak and Walter, 2014). Pex15Δ30 N-terminally tagged with the circular permuted cp8 superfolder GFP variant (referred to GFP hereafter), which shows less cytoplasmic background due to improved proteasomal degradation (Khmelinskii et al., 2016), localized predominantly to mitochondria and only to a minor extent to peroxisomes in wt and msp1Δ cells (Figure 1—figure supplement 1A and B). To be able to measure Pex15∆30 turnover in high-throughput in different mutants, we fused Pex15Δ30 to a tandem fluorescent protein timer (tFT) consisting of the fast-maturing superfolder GFP (sfGFP) and the slower-maturing mCherry (Figure 1A) (Khmelinskii et al., 2012). A population of newly synthesized fusion proteins exhibits first sfGFP fluorescence and acquires mCherry fluorescence over time. This property allows use of the mCherry/sfGFP ratio as a measure of protein turnover, whereby the ratio increases as a function of protein half-life (Khmelinskii et al., 2012). We introduced the tFT-Pex15Δ30 construct into an arrayed collection of non-essential yeast gene deletion strains (Winzeler et al., 1999) using high-throughput strain construction (Tong and Boone, 2006) and measured the mCherry and sfGFP fluorescence of colonies grown on agar (Figure 1A) (Khmelinskii et al., 2014; Khmelinskii et al., 2012). Because the tFT readout can be affected by perturbations unrelated to protein turnover, for example changes of mitochondrial physiology (Khmelinskii et al., 2012), we performed a control screen using tFT-tagged Tom5 (Figure 1A). Tom5 is a stable TA protein subunit of the preprotein translocase TOM. Gene deletions that significantly increased tFT-Pex15∆30 abundance and stability but did not affect tFT-Tom5 were considered as hits (Figure 1B, Supplementary file 1). According to this analysis, tFT-Pex15Δ30 was stabilized in msp1Δ as has been observed before using GFP-tagged constructs (Chen et al., 2014; Okreglak and Walter, 2014). Moreover, deletion of GET3, which was previously linked to trafficking of full-length Pex15 and shows a negative genetic interaction with msp1Δ (Costanzo et al., 2010; Okreglak and Walter, 2014), caused a significant increase in sfGFP level and mCherry/sfGFP ratio. tFT-Pex15Δ30 was also significantly stabilized in the absence of the ER-resident P-type ATPase Spf1, the pheromone regulated membrane protein Prm1 and, interestingly, the ER-resident E3 ubiquitin ligase Doa10, its associated E2 ubiquitin-conjugating enzyme Ubc7 and its tethering factor Cue1 (Figure 1B). A mutant lacking Ubc6, the second E2 ubiquitin-conjugating enzyme mediating protein degradation together with Doa10 (Weber et al., 2016), was not present in the yeast strain collection used for screening. To validate these results, we manually deleted each hit in a haploid strain expressing tFT-Pex15Δ30 and measured strain fluorescence with flow cytometry. Strains lacking MSP1, DOA10, UBC7, CUE1, GET3 or SPF1 consistently showed increased sfGFP fluorescence and mCherry/sfGFP ratios compared to wt (Figure 1C). Deletion of UBC6 also stabilized tFT-Pex15Δ30. In contrast, we did not observe stabilization of tFT-Pex15Δ30 in prm1Δ cells and excluded this mutant from further analysis. Next, we used fluorescence microscopy to examine the localization of tFT-Pex15Δ30 in each mutant. In msp1Δ, doa10Δ, ubc6Δ, ubc7Δ, and get3Δ mutant strains tFT-Pex15Δ30 accumulated in mitochondrial structures (Figure 1—figure supplement 1B and C). Figure 1 with 1 supplement see all Download asset Open asset Genome-wide screen to identify factors stabilizing Pex15Δ30 TA protein. (A) Overview of the tFT-Pex15Δ30 and tFT-Tom5 TA protein reporters. Tail-anchor is highlighted in blue. Strains, expressing either tFT-Pex15Δ30 or tFT-Tom5 from the TEF1 promoter were crossed into the yeast non-essential gene deletion collection (Winzeler et al., 1999) using automated mating and selection procedure. sfGFP and mCherry fluorescence was acquired from arrayed colonies grown on agar (n = 4). (B) sfGFP signal and mCherry/sfGFP ratio of the tFT-Pex15Δ30 and tFT-Tom5 reporters of each mutant shown as z-score (which resembles the standard deviations from the mean a data point is). Mutants with z-scores > 3 for sfGFP and mCherry/sfGFP ratio (at 5% false discovery rate) for Pex15Δ30 and not affecting Tom5 are highlighted in black. Dashed gray lines indicate the thresholds. GET mutants below the threshold are highlighted in dark gray. (C) Flow cytometry validation of generated yeast mutants as indicated. Mean sfGFP intensities and mCherry/sfGFP ratios normalized to wt (n = 4, ± SEM). https://doi.org/10.7554/eLife.45506.002 Disparate effects of spf1Δ and get3Δ on Pex15Δ30 turnover Whereas ablation of Get3 or the Doa10 complex lead to accumulation of Pex15Δ30 in mitochondria, deletion of SPF1 led to pronounced ER localization (Figure 2A and Figure 1—figure supplement 1C). This finding is consistent with a previous report demonstrating that mitochondrial TA proteins mislocalize to the ER in spf1Δ strains (Krumpe et al., 2012). This phenotype is thought to be caused by altered ER ergosterol levels in spf1Δ, mimicking conditions of the OMM. Interestingly, in addition to the ER localization, the overall levels of the tFT-Pex15Δ30 reporter increased in the spf1Δ mutant (Figure 1C). This indicates that spf1∆ cells cannot efficiently degrade Pex15Δ30. Western blot analysis revealed a characteristic 26 kDa remnant of sfGFP that persists vacuolar degradation in spf1∆ cells (Figure 2B) (Khmelinskii et al., 2016; Shintani and Klionsky, 2004). Monitoring tFT-Pex15Δ30 localization in spf1∆ strains by microscopy revealed that this accumulation is likely caused by transport of ER-inserted protein through the secretory pathway. Whereas in the ER only the fast maturing sfGFP and no mCherry signal can be detected, the latter is observed in the vacuole (Figure 1—figure supplement 1C). However, if Pex15Δ30 is, in addition, directly targeted into the vacuole or multivesicular body membrane is not resolved yet. Figure 2 with 1 supplement see all Download asset Open asset ER insertion and mitochondrial accumulation impedes efficient Pex15Δ30 degradation in spf1Δ and get3Δ. (A) Microscopy analysis of spf1Δ strains expressing GFP-Pex15Δ30 from the TEF1 promoter. Co-expression of chromosomally tagged cellular marker proteins: Cox4-mScarlet-i, mitochondria; Pex3-mScarlet-i, peroxisomes; Sec63-mScarlet-i, ER. Images are adjusted for optimal display range. Colocalization is colored in white in merge (green - GFP, magenta - mScarlet-i). Scale bar: 5 μm. (B) Western blot (WB) analysis of log phase grown wt, msp1Δ, and spf1Δ strains expressing tFT-Pex15Δ30. Probing with anti-GFP antibody detects full length protein (FL) and degradation resistant tFT intermediates (Khmelinskii et al., 2016): h, SDS-induced mCherry hydrolysis product; p, proteasomal-degradation resistant fragment; v, vacuolar degradation-resistant fragment. Pgk1 serves as loading control. (C) Microscopy analysis of get2Δ, get3Δ, get4Δ, and get5Δ strains expressing tFT-Pex15Δ30 compared to wt and msp1Δ. Scale bar: 5 μm. (D) Flow cytometry GFP measurement of strains from (C) normalized to wt (n = 3, ± SEM). (E) WB analysis of log phase grown strains from (C) with detection of degradation resistant tFT intermediates. https://doi.org/10.7554/eLife.45506.004 To understand if deletion of GET3 stabilizes Pex15Δ30 directly or indirectly, we investigated the effect of other components of the GET pathway on cellular levels of tFT-Pex15Δ30. Of note, in the screen, mutants lacking the GET components Sgt2, Get4 and Get5 showed effects just below the threshold (Figure 1B). Similar to get3Δ, deletion of the ER membrane receptor subunit Get2 or the cytosolic pre-targeting factors Get4 and Get5 resulted in increased levels of the tFT-Pex15Δ30 and accumulation in mitochondria (Figure 2C and D). Western blot analysis showed an increase of tFT-Pex15Δ30 level in a similar range as msp1Δ for all GET mutants (Figure 2E). Of note, the characteristic 26 kDa vacuolar tFT degradation intermediate observed for spf1∆ was not seen in get3Δ cells (Figure 2—figure supplement 1). Taken together, these results indicate that the GET mutants interfere with the tFT-Pex15Δ30 reporter by blocking canonical ER targeting function (involving the ER-resident membrane receptor Get2). Hence, Pex15Δ30 accumulation in mitochondria is most likely a secondary consequence of altered TA protein homeostasis in GET mutants (see below). The Doa10 E3 ligase is a major factor in Pex15Δ30 turnover Our screen revealed that Doa10 is important for degradation of Pex15Δ30 (Figure 1B). Rescreening of the tFT-Pex15Δ30 reporter against a library of mutants in the ubiquitination machinery confirmed that Ubc6, Ubc7 and Doa10 are the only significant hits (Figure 3—figure supplement 1A and B). Fluorescence measurements with flow cytometry and cycloheximide chase experiments independently revealed reduced turnover of the tFT-Pex15Δ30 reporter in the DOA10 mutant (Figure 1C and Figure 3—figure supplement 1C). Notably, tFT-Pex15Δ30 levels were higher in doa10Δ compared to msp1Δ, and further increased in the msp1Δdoa10Δ double mutant (Figure 3A and B). Different to an msp1∆get3∆ double deletion, which shows a fitness defect (Chen et al., 2014; Okreglak and Walter, 2014), fitness of the msp1Δdoa10Δ mutant was indistinguishable from wt (Figure 3—figure supplement 1D), excluding a general toxicity effect and pointing towards a direct role of Doa10 in Pex15Δ30 degradation. Consistent with ubiquitin-dependent degradation, addition of proteasome inhibitors delayed turnover of the Pex15Δ30 reporter (Figure 3—figure supplement 1E). Moreover, double deletion of MSP1 and DOA10 showed an additive stabilization effect of Pex15Δ30 (Figure 3A), suggesting that the two proteins do not act in a linear pathway. Figure 3 with 1 supplement see all Download asset Open asset Doa10 triggers degradation of cytoplasmic Pex15Δ30. (A) Flow cytometry measurements of wt, msp1Δ, doa10Δ and msp1Δdoa10Δ strains expressing tFT-Pex15Δ30. Mean GFP intensities normalized to wt (n = 4, ± SEM). (B) Western blot (WB) analysis from log phase grown yeast of (A) with full-length tFT-Pex15Δ30 (FL) and degradation resistant tFT intermediates: h, SDS-induced mCherry hydrolysis product; p, proteasomal-degradation resistant fragment; v, vacuolar degradation-resistant fragment. Unspecific bands are marked with star. Pgk1 serves as loading control. (C) Microscopy analysis of DOA10 deletion strains co-expressing GFP-Pex15Δ30 and mitochondrial marker protein Cox4-mScarlet-i, peroxisomal marker protein Pex3-mScarlet-i or ER marker protein Sec63-mScarlet-i. Colocalization is highlighted in white in merge image. Images are shown with optimal display range. Scale bar: 5 μm. https://doi.org/10.7554/eLife.45506.006 Doa10 is a major player of ERAD-C (Mehrtash and Hochstrasser, 2018; Ruggiano et al., 2014). However, we observed only mitochondrial accumulation of tFT-Pex15Δ30 in doa10Δ (Figure 3C), indicating that it acts on pre-inserted TA protein reporter. Moreover, Pex15Δ30 turnover was not affected in the absence of Hrd1 and Asi1/3, the other two ERAD E3 ligases (Figure 3—figure supplement 1A,B and C). An alternative degradation route for Pex15Δ30 might be vacuolar degradation. In line with this, western blot analysis of tFT-Pex15Δ30 in doa10Δ revealed accumulation of the vacuolar tFT degradation intermediate at 26 kDa (Figure 3C and Figure 2—figure supplement 1) (Khmelinskii et al., 2016; Shintani and Klionsky, 2004). Interestingly, additional deletion of MSP1 decreased abundance of this fragment suggesting that mitochondria act as a ‘sink’ for mistargeted tFT-Pex15Δ30. Along this hypothesis, the vacuolar sfGFP fragment was absent in the get3Δ mutant, supporting the idea that Get3 is required for membrane insertion and subsequent transport of tFT-Pex15Δ30 to the vacuole (Figure 2—figure supplement 1). Since in doa10Δ no prominent ER localization is observed (Figure 3—figure supplement 1B), we assume that tFT-Pex15Δ30 only very transiently exists in the ER. Next, we asked whether under conditions where ER targeting of TA proteins is distorted, also Doa10 activity is compromised. To this end, we studied turnover of the Doa10-dependent ERAD substrate Ste6* (Stolz et al., 2010). Interestingly, upon GET3 deletion Ste6* showed a minor stabilization (Figure 3—figure supplement 1F). This may be either caused by mistargeting of the E2 ubiquitin conjugating enzyme Ubc6 leading to a reduction of Doa10-dependent ubiquitination (Weber et al., 2016) or cytoplasmic accumulation of TA proteins saturating Doa10 activity. Over all, our results show that the E3 ubiquitin ligase Doa10 plays a prominent role in turnover of the mistargeted TA reporter Pex15Δ30 whereas the GET pathway plays an indirect role via maintaining the general TA protein homeostasis including but not limited to Ubc6. Doa10 targets cytoplasmic Pex15Δ30 for degradation to prevent mitochondrial mistargeting Doa10 and Msp1 localize to two different cellular compartments, the ER and the OMM, respectively. This poses the question in which order these factors interact with the tFT-Pex15Δ30 reporter. We hypothesized that Doa10 either acts directly on newly synthesized tFT-Pex15Δ30, thus preventing its insertion into the OMM, or after extraction from it, whereas Msp1 deals only with the fraction that is mistargeted to mitochondria. According to this model, cytoplasmic accumulation of Pex15Δ30 in doa10Δ strains enhances mitochondrial mistargeting and therefore saturates Msp1-dependent extraction. In order to test this, we overexpressed Msp1 in doa10Δ from a galactose inducible promoter (Figure 4—figure supplement 1A). To investigate the localization of overexpressed Msp1 we tagged it at the C-terminus with mNeonGreen. Under galactose induction for 4 hr Msp1 predominantly localizes to mitochondria (Figure 4—figure supplement 1B). Galactose induction of Msp1 in the doa10Δ background did not affect cellular tFT-Pex15Δ30 level, consistent with the requirement of Doa10 for degradation (Figure 4A, Figure 3—figure supplement 1C, and Figure 4—figure supplement 1C). Fluorescence microscopy, however, revealed that tFT-Pex15Δ30 was not detected any more in mitochondria. The GFP signal changed its localization to the cytosol and the ER, whereas mCherry was only detected in the vacuole (Figure 4B). This corroborates that tFT-Pex15Δ30 is only transiently localized in the ER and finally gets targeted to the vacuole (where the GFP signal is quenched). Similar observations were made using mutants in other components of the Doa10 ligase complex, CUE1, UBC7 or UBC6 deletion led to a stabilization of tFT-Pex15Δ30 (Figure 1C). Likewise, overexpression of Msp1 in cue1Δ or ubc6Δ did not restore tFT-Pex15Δ30 levels to wt (Figure 4A) and imaging showed relocalization of tFT-Pex15Δ30 from mitochondria to the cytosol, the ER and the vacuole (Figure 4—figure supplement 1D and E). These results show that in Doa10-deficient strains Msp1 levels are limiting when expressed from its endogenous promoter and argues that Doa10 defines the predominant route for proteasomal degradation of Pex15Δ30. Consistent with this, we observed also reduced poly-ubiquitination of Pex15Δ30 in the doa10∆ strain (Figure 4C and Figure 4—figure supplement 1F). Interestingly, in the absence of Doa10 a prominent single band resembling mono-ubiquitinated Pex15Δ30 was observed (Figure 4C). This indicates that alternative ubiquitination events not leading to proteasomal degradation can occur. The question whether vacuolar degradation compensating ablation of Doa10 depends on ubiquitin-dependent trafficking (Komander and Rape, 2012) remains to be addressed. Figure 4 with 1 supplement see all Download asset Open asset Msp1 overexpression clears mitochondrial accumulated Pex15Δ30 in doa10Δ. (A) Flow cytometry GFP measurements of tFT-Pex15Δ30 expressed in wt, DOA10, CUE1, UBC6 and MSP1 mutant yeast (n = 4, ± SEM). Msp1 protein expression is controlled from the galactose-inducible GAL1 promoter (GAL1pr-MSP1). (B) Localization of tFT-Pex15Δ30 in strains of (A) with and without Msp1 overexpression. Scale bar: 5 μm. (C) Ubiquitylation of TAP-Pex15Δ30 in wt and doa10Δ strains expressing 10xhistidine-tagged ubiquitin as assessed by Ni-NTA affinity purification and western blotting (WB). neg. refers to a strain expressing TAP-Pex15Δ30 but not 10xhistidine-tagged ubiquitin. Tub2 is used as loading control. https://doi.org/10.7554/eLife.45506.008 Screen for other Msp1 and Doa10 clients Our data indicates that the truncated Pex15Δ30 reporter is targeted by Msp1- and Doa10-dependent quality control. In order to identify more substrates for this pathway, we measured the abundance of 55 TA proteins (Burri and Lithgow, 2004, see Supplementary file 2) in wt, msp1Δ and doa10Δ strains. The corresponding sfGFP-tagged alleles expressed from the constitutive NOP1 promoter were obtained from an N-terminal library (Yofe et al., 2016) and fluorescence was analyzed from colonies. This screen identified six proteins with a more than 2-fold increase in their GFP level upon deletion of DOA10 (Figure 5A). One of them, Ubc6, had been shown to be a substrate of Doa10 in the context of ERAD (Figure 5A and B) (Walter et al., 2001). Other proteins that accumulated upon compromised function of Doa10 are Csm4, Pgc1, Sps2, Ydl241w. These proteins localized to the ER or other vesicular structures (Figure 5B) and did not accumulate in the msp1Δ mutant (Figure 5A). The strongest stabilization in doa10Δ was observed for Fmp32, which is an uncharacterized mitochondrial protein implicated in maintenance of the respiratory capacity of mitochondria (Paupe et al., 2015) (Figure 5A and B). Figure 5 with 1 supplement see all Download asset Open asset Doa10 controls targeting fidelity of mitochondrial TA proteins. (A) Mean GFP fluorescence of 55 N-terminally GFP-tagged TA proteins expressed from NOP1 promoter deleted for DOA10 or MSP1 compared to wt. Measurements were taken from colonies grown on agar (n = 4). Highlighted proteins are significantly enriched (Student’s t-test p<0.05, threshold gray dashed line). (B) Microscopy validation of significantly enriched proteins from (A) in wt compared to doa10Δ. Images for each protein investigated are shown with the same display range for wt and doa10Δ. Insets show optimized display range. Scale bar: 5 μm. (C) Microscopy analysis of sfGFP-tagged Fmp32 under its endogenous promoter. Mitochondria are stained with the mitochondrial dye Mitotracker Red. Scale bar, 5 μm. (D) Western blot (WB) analysis of log phase growing yeast from (C). neg. is a wt yeast referring to GFP specific bands. Unspecific bands are marked with star. Pgk1 is used as loading control. https://doi.org/10.7554/eLife.45506.010 A similar analysis in msp1Δ identified weak, but significant accumulation of 3 proteins: Pex15, Gos1 and Fmp32 (Figure 5A). This indicates that consistent with previous reports (Chen et al., 2014), a small fraction of full-length Pex15, as well as Gos1, mislocalized to mitochondria and are subject to Msp1-dependent extraction. In contrast to Pex15Δ30, full-length Pex15 and Gos1 do not accumulate in DOA10 deletion strains, indicating that in presence of active Mps1 they are efficiently targeted to their native destination. Again, the strongest effect of MSP1 deletion was observed for Fmp32 (1.8-fold increase) (Figure 5A). This indicates that like the Pex15Δ30 reporter, Fmp32 is extracted and degraded in a concerted action by Doa10 and Msp1. Interestingly, under its native promoter, sfGFP-tagged Fmp32 is almost undetectable in wt strains (Figure 5C and D). Deletion of MSP1 led to a robust increase in the levels of sfGFP-Fmp32 (Figure 5D). Of note, sfGFP-Fmp32 did not accumulate in the absence of Cis1, indicating that is extracted by the canonical Msp1 activity and not target for mitochondrial protein import stress response (Weidberg and Amon, 2018) (Figure 5—figure supplement 1). Deletion of DOA10 alone or in combination with MSP1 led to a dramatic increase of Fmp32 localizing both to the cytosol and mitochondria (Figure 5C and D). This result implies that the mitochondrial protein Fmp32 is constantly degraded by Doa10 with the assistance of Msp1 extrac

The heterodimeric ubiquitin-conjugating enzyme (E2), Ubc13-Mms2, catalyses K63-specific polyubiquitylation in genome maintenance and inflammatory signalling. In budding yeast, the only ubiquitin protein ligase (E3) known to cooperate with Ubc13-Mms2 so far is a nuclear RING finger protein, Rad5, involved in the replication of damaged DNA. We have now identified a family of membrane-associated FYVE-(type)-RING finger proteins as cognate E3s for Ubc13-Mms2 in several species. We show that budding yeast Pib1, a FYVE-RING finger E3 associated with internal membranes, exhibits exquisite selectivity for Ubc13-Mms2 and cooperates with the E2 in the multivesicular body pathway. Phenotypic analysis indicates that the contribution of Ubc13-Mms2 to membrane trafficking goes beyond its cooperation with Pib1, suggesting an involvement with additional E3s in the endocytic compartment. These results widely implicate Ubc13-Mms2 in the regulation of membrane protein sorting.

Tail-anchored (TA) proteins insert post-translationally into the endoplasmic reticulum (ER), the outer mitochondrial membrane (OMM) and peroxisomes. Whereas the GET pathway controls ER-targeting, no dedicated factors are known for OMM insertion, posing the question of how accuracy is achieved. The mitochondrial AAA-ATPase Msp1 removes mislocalized TA proteins from the OMM, but it is unclear, how Msp1 clients are targeted for degradation. Here we screened for factors involved in degradation of TA proteins mislocalized to mitochondria. We show that the ER-associated degradation (ERAD) E3 ubiquitin ligase Doa10 controls cytoplasmic level of Msp1 clients. Furthermore, we identified the uncharacterized OMM protein Fmp32 and the ectopically expressed subunit of the ER-mitochondria encounter structure (ERMES) complex Gem1 as native clients for Msp1 and Doa10. We propose that productive localization of TA proteins to the OMM is ensured by complex assembly, while orphan subunits are extracted by Msp1 and eventually degraded by Doa10.