Hsp70–Hsp110 Chaperones Deliver Ubiquitin-Dependent

Home , Chaperone (protein), Hsp70

RESEARCH ARTICLE Hsp70–Hsp110 chaperones deliver ubiquitin-dependent and -independent substrates to the 26S proteasome for proteolysis in yeast Ganapathi Kandasamy and Claes Andréasson*

ABSTRACT proteasome by shuttling factors that associate with both the During protein quality control, proteotoxic misfolded proteins are ubiquitin chain and the 19S regulatory particle of the proteasome recognized by molecular chaperones, ubiquitylated by dedicated (Elsasser et al., 2004; Husnjak et al., 2008; Su and Lau, 2009). quality control ligases and delivered to the 26S proteasome for Delivered proteins are unfolded, deubiquitylated and translocated degradation. Proteins belonging to the Hsp70 chaperone and Hsp110 into the 20S proteolytic chamber of the proteasome for degradation. (the Hsp70 nucleotide exchange factor) families function in the Ubiquitin tagging is dispensable for the proteasomal degradation degradation of misfolded proteins by the ubiquitin-proteasome of a subset of cellular proteins. In such ubiquitin-independent system via poorly understood mechanisms. Here, we report that the degradation, unstructured tails that interact directly with the Saccharomyces cerevisiae Hsp110 proteins (Sse1 and Sse2) function proteasome function as degrons (Ben-Nissan and Sharon, 2014; in the degradation of Hsp70-associated ubiquitin conjugates at the Takeuchi et al., 2007; Yu et al., 2016a,b). Classical examples of post-ubiquitylation step and are also required for ubiquitin-independent proteins that undergo such ubiquitin-independent degradation in Saccharomyces cerevisiae proteasomal degradation. Hsp110 associates with the 19S regulatory include ornithine decarboxylase (ODC), particle of the 26S proteasome and interacts with Hsp70 to facilitate the Rpn4 and Pih1 (Gödderz et al., 2011; Paci et al., 2016; Xie and delivery of Hsp70 substrates for proteasomal degradation. By using a Varshavsky, 2001). Thus the physical targeting of substrate proteins highly defined ubiquitin-independent proteasome substrate, we show directly to the proteasome and the susceptibility of flexible peptide that the mere introduction of a single Hsp70-binding site renders its stretches to mediate a translocation into the core proteolytic chamber degradation dependent on Hsp110. The findings define a dedicated govern rates of cellular protein turnover. and chaperone-dependent pathway for the efficient shuttling of The abundant Hsp70 family of chaperones (hereafter Hsp70) cellular proteins to the proteasome with profound implications for functions at the heart of PQC systems, and is critical for both the understanding protein quality control and cellular stress management. folding and proteasomal degradation of misfolded proteins. Hsp70 associates with misfolded proteins in a manner controlled by KEY WORDS: Protein degradation, Proteasome, Ubiquitin, its ATPase cycle and co-chaperones. The Hsp110 sub-family of Chaperone, Hsp70, Quality control proteins (hereafter Hsp110) is an abundant co-chaperone relative of Hsp70 that plays a central role in Hsp70 function. Hsp110 INTRODUCTION [Saccharomyces cerevisiae (hereafter yeast) proteins Sse1 and Proteolytic removal of misfolded proteins is an important process to Sse2], transiently associates with Hsp70 and accelerates nucleotide maintain proteostasis and to limit the damage caused by proteotoxic exchange, which results in large conformational changes of Hsp70 stress. Cellular protein quality control (PQC) systems selectively and release of substrates from the chaperone (Andréasson et al., recognize misfolded proteins, keep them associated with molecular 2008b; Dragovic et al., 2006; Raviol et al., 2006). Such controlled chaperones and target them for proteolytic degradation (McClellan association and release play critical roles in regulating the et al., 2005; Park et al., 2007). Failure to degrade misfolded proteins proteasomal degradation of misfolded proteins during PQC by PQC results in the accumulation of proteotoxic misfolded (Abrams et al., 2014; Gowda et al., 2013). Studies in yeast have proteins and has been linked to age-associated neurodegenerative shown that reducing the Hsp70 levels (yeast proteins Ssa1, Ssa2, Ssa3 diseases, including Parkinson’s and Huntington’s disease (Forloni and Ssa4) results in the accumulation of aggregated and ubiquitin- et al., 2002). modified misfolded proteins (Fang et al., 2011; Lee et al., 2016; The ubiquitin-proteasome system (UPS) is a critical component Shiber et al., 2013). Similarly decreasing the levels of Hsp110 in PQC (Ciechanover, 1994; Glickman and Ciechanover, 2002). (sse1Δ) results in the accumulation of misfolded model proteins Misfolded proteins are recognized by specialized ubiquitin ligases (Escusa-Toret et al., 2013; Guerriero et al., 2013; Heck et al., 2010; and covalently tagged with ubiquitin chains that serve as Mandal et al., 2010; McClellan et al., 2005) and of ubiquitin degradation signals (Eisele and Wolf, 2008; Theodoraki et al., conjugates following a heat shock (Gowda et al., 2013). These 2012). The polyubiquitylated proteins are then delivered to 26S findings support a role for Hsp70 and Hsp110 in misfolded protein degradation, perhaps to maintain misfolded proteins in a soluble form Department of Molecular Biosciences, The Wenner-Gren Institute Stockholm during transit to the proteasome, but experiments are plagued by University, SE-10691, Stockholm, Sweden. indirect phenotypes stemming from the role of Hsp70 and Hsp110 in protein biogenesis (Gowda et al., 2013). Thus, the mechanistic *Author for correspondence ([email protected]) function of Hsp70 and Hsp110 in proteolysis is not well understood. G.K., 0000-0001-8110-2567; C.A., 0000-0001-8948-0685 The transfer of misfolded proteins from chaperone systems to the UPS may rely on simple kinetic competition for substrate binding as

Received 13 September 2017; Accepted 14 February 2018 well as on more intricate mechanisms that physically link chaperones Journal of Cell Science

1 RESEARCH ARTICLE Journal of Cell Science (2018) 131, jcs210948. doi:10.1242/jcs.210948

to the UPS machinery. For example in the yeast cytoplasm, the analysis was expanded with the distinct UFD substrate UbV76–EGFP degradation of misfolded proteins that are associated with the (Dantuma et al., 2000). This UFD substrate was also dependent on chaperone Hsp70 depends on the nucleotide exchange factor Fes1 Hsp110 for its degradation, and again the unmodified and being released from the chaperone and then recognized by PQC monoubiquitylated forms were found to be distributed in the ubiquitin ligases (Gowda et al., 2016, 2013, 2018). In metazoan cells, soluble and insoluble fractions (Fig. 1D). To assess the functional mechanisms involve the direct docking of the ubiquitin ligase CHIP requirements of Sse1 in the protein degradation pathway, we tested to Hsp70, and the nucleotide exchange factor BAG-1 coordinates the previously described Sse1 mutants deficient in ATP binding release of the misfolded protein from Hsp70 with proteasome (G205D), ATP hydrolysis (K69M) and substrate binding (denoted interactions (Demand et al., 2001; Hohfeld and Jentsch, 1997; Lüders sbd; L433A, N434P, F439L and M441A) (Andréasson et al., 2008b; et al., 2000). A third example involves the ubiquitylation of proteins Garcia et al., 2017; Shaner et al., 2004). Sse1 depends on ATP damaged by acute heat shock, which has been proposed to depend on binding but not on hydrolysis to interact with Hsp70 (Andréasson the coordinated interactions of the PQC ubiquitin ligase Rsp5, the et al., 2008b). We found that the proteasomal degradation of UbV76– Hsp40 chaperone Ydj1 and the misfolded protein (Fang et al., 2014). Ura3–HA was strongly impeded by the lack of ATP binding, but not Nevertheless, our understanding of the basic PQC mechanisms that by ATP hydrolysis or substrate binding (Fig. 1F). Thus, Hsp110 link chaperone systems and the UPS is at the best rudimentary. functions together with Hsp70 in the proteasomal degradation Here, we investigate the function of Hsp110 in proteasomal pathway. degradation by employing a novel yeast strain with a temperature- Next, we focused on the potential post-ubiquitylation role of sensitive Hsp110 function (sse1-200 sse2Δ). We find that Hsp110 Hsp110 in proteasomal degradation following the turnover of functions together with Hsp70 in the degradation of aggregation- misfolded ubiquitylated proteins after heat shock (Fang et al., 2011; prone proteins involving both ubiquitin-dependent and -independent Gowda et al., 2013). Consistent with the previous reports, western proteasomal pathways. Hsp110 is required to keep Hsp70-associated blot analysis of ubiquitin-conjugate levels in wild-type (WT) cells proteasome substrates soluble and interacts with 19S regulatory transferred from 25°C to 37°C showed an accumulation of particle of the proteasome, suggesting coordinated recruitment of ubiquitylated proteins during the first 20 min after the temperature Hsp70–substrate complexes to the 26S proteasome for degradation. upshift, which were subsequently cleared by proteasomal The data defines a novel PQC pathway that enables the direct and degradation to reach the initial levels after 30 min (Fig. S1A). In efficient delivery of Hsp70-associated aggregation-prone proteins to sse1-200 sse2Δ cells, the accumulation of ubiquitin-conjugate the proteasome. levels was exacerbated and the accumulated ubiquitylated proteins remained and even increased 30 min after the temperature upshift, RESULTS reflecting the continued production of newly synthesized misfolded Hsp110 is required for the efficient degradation of ubiquitin- proteins and impaired degradation. To specifically analyse the modified and ubiquitin-independent proteasome substrates degradation of ubiquitin conjugates, we arrested translation with Heat stress induces the rapid accumulation of ubiquitin conjugates cycloheximide after 30 min of heat stress. Both the soluble and consisting of misfolded and low-solubility proteins (Fang et al., insoluble fractions turned over in cells expressing functional Sse1, 2014, 2011). Reducing the function of the essential Hsp110 protein while, by contrast, cells that lacked functional Hsp110 displayed an (sse1Δ) results in increased levels of cellular ubiquitin conjugates overall delay in the clearance of ubiquitin conjugates (Fig. S1B). and stabilization of the ubiquitin-fusion degradation (UFD) pathway Moreover, the resilient ubiquitin conjugates in sse1-200 sse2Δ cells substrate UbV76–Ura3–HA (UbV76 is a ubiquitin with a G76V were enriched in the insoluble fraction. We assessed whether substitution) (Gowda et al., 2013). However, the previously used the soluble ubiquitin-conjugate accumulation was exacerbated in experimental setup based on hypomorphic sse1Δ strains has been Hsp110-depleted cells with defective proteasomal degradation. In plagued by interpretation problems stemming from indirect strains with an epitope-tagged 19S regulatory particle protein Rpt1 biosynthetic protein folding defects. To circumvent such (rpt1-FH), the association between the 19S regulatory particle and the problems, we employed the newly developed temperature- 20S core particle of the proteasome subcomplexes is severely sensitive sse1-200 sse2Δ strain that enables the rapid and transient impaired, resulting in the accumulation of ubiquitylated proteins due inactivation of all Hsp110 activity (Kaimal et al., 2017). First, we to impeded proteolysis (Verma et al., 2000). Indeed, following heat- followed the turnover of the UFD pathway substrate (Johnson et al., shock at 37°C, the rpt1-FH strain accumulated soluble ubiquitin 1995). After a temperature upshift to 37°C, UbV76–Ura3–HA was conjugates and the levels were increased 1.5-fold under Hsp110- degraded in Sse1-complemented sse1-200 sse2Δ cells but was depleted conditions suggesting that Hsp110 functions in targeting stabilized in cells lacking Hsp110 activity (sse1-200 sse2Δ)orsse1- soluble misfolded proteins for proteasomal degradation (Fig. S1C). 200 sse2Δ cells expressing the Hsp70-binding mutant Sse1-2,3 We have recently shown that Hsp110 is required for Hsp104- (Fig. 1A). Importantly, a monoubiquitinylated species of UbV76– dependent disaggregation of insoluble proteins (Kaimal et al., 2017). Ura3–HA accumulated upon the inactivation of Sse1 pointing to a However, hsp104Δ cells did not display an increase in ubiquitin post-ubiquitylation function of Hsp110 in the degradation of this conjugates following heat stress, demonstrating that the acute process protein. To rigorously rule out that folding defects in the Sse1 is little affected by the Hsp104-dependent disaggregation (Fig. S1D). mutant caused the stabilization phenotype, translation was arrested Hsp110 may play a more general role in targeting proteins with cycloheximide immediately before the shift to 37°C and this associated with Hsp70 to the proteasome, independently of their did not impact on the dependency of Hsp110 for proteasomal ubiquitylation status. We directly tested the role of Hsp110 in the degradation (Fig. 1B). Similarly, the monoubiquitinylated species proteasomal degradation of ubiquitin-independent substrates by remained stable upon Hsp110 inactivation. Analysis of the solubility employing the model protein comprising the degradation signal of UbV76–Ura3-HA by means of centrifugation of lysates showed that from ODC (ODS) conjugated to Ura3–HA (ODS–Ura3–HA) a portion of the protein failed to undergo proteasomal degradation in (Gödderz et al., 2011). Indeed, this substrate was also stabilized in the absence of Hsp110 function and accumulated in the insoluble the sse1-200 sse2Δ strain transiently shifted to 37°C, and the fraction, including the monoubiquitylated species (Fig. 1C). The degradation was restored by SSE1 but not by sse1-2,3 Journal of Cell Science

2 RESEARCH ARTICLE Journal of Cell Science (2018) 131, jcs210948. doi:10.1242/jcs.210948

Fig. 1. Sse1 is indispensable for the degradation of ubiquitin-dependent and ubiquitin-independent proteasome substrates. (A) Degradation of the UFD substrate UbV76–Ura3–HA in sse1-200 sse2Δ strains transformed with vector control (VC) or corresponding plasmids carrying SSE1 or sse1-2,3.UbV76–Ura3– HA expression from the CUP1 promoter was induced by addition of 100 μM copper at OD600=0.2 to 0.8 and cells were shifted from 25°C to 37°C for 30 min before cycloheximide (CHX) addition. Samples were taken at the indicated time after CHX addition. Quantification of the western blot signals is presented in the right panel. Error bars denote s.d.; n≥5. (B) Same as in A except that CHX was added immediately before cells were shifted to 37°C. (C) Levels of UbV76–Ura3–HA in total (T) cell-free lysates or in soluble (S) and pellet (P) fractions after centrifugation at 21,100 g. Strains as in A were treated with CHX, shifted to 37°C and samples were taken at the indicated times. The asterisk (*) denotes a cross-reactive band. (D) Proteasomal degradation of Ubv76–EGFP in soluble and pellet fractions (21,100 g centrifugation) prepared from sse1-200 sse2Δ cells transformed with plasmid carrying SSE1 or VC after CHX addition. The position of migration of monoubiquitylated Ubv76–EGFP is indicated (Ub1). Ubv76–EGFP expression was induced by the addition of 100 μM copper. (E) Proteasomal degradation of the ubiquitin-independent proteasome substrate ODS–Ura3–HA in sse1-200 sse2Δ strains transformed with vector control (VC) or corresponding plasmids carrying SSE1 or sse1-2,3. Western blot signals were quantified from n≥3. Error bars denote s.d. (F) An experiment as in B but studying the UbV76– Ura3–HA substrate turnover rate in sse1-200 sse2Δ strains transformed with plasmids carrying either SSE1 or mutants defective in ATP binding (G205D), ATP hydrolysis (K69M) and substrate binding (sbd; L433A, N434P, F439L, and M441A). Quantification of the UbV76–Ura3-–HA signal is presented on the right. Error bar denotes s.d.; n≥3. The asterisk (*) denotes a cross-reactive band. (G) Same as in F, but degradation of ubiquitin-independent proteasome substrate ODS–Ura3–HA was analysed at 37°C. Quantification of the ODS–Ura3–HA signal is presented on the right. Error bars denote s.d.; n≥3. Journal of Cell Science

3 RESEARCH ARTICLE Journal of Cell Science (2018) 131, jcs210948. doi:10.1242/jcs.210948 complementation (Fig. 1E). Similar to what is seen for the but not structurally unrelated Hsp70 nucleotide-exchange factors, are ubiquitylated substrates, the turnover of ODS–Ura3–HA depended part of a specific proteasomal degradation pathway. on Sse1 ATP binding, but not hydrolysis or substrate binding (Fig. 1G). Thus, the data supports the notion that Hsp110 functions in Transient Hsp110 inactivation does not generally impair UPS dynamic complexes with Hsp70 in the targeting of a multitude of degradation ubiquitin-dependent and -independent substrates to the proteasome. The impaired degradation of proteasome substrates upon Hsp110 inactivation raises the concern that the experimental setup may Other Hsp70 nucleotide exchange factors do not substitute cause a general proteolysis defect, for example by out-titrating for Hsp110 in the protein degradation pathway available proteasomes with misfolded protein substrates. To assess Reasoning that Hsp110 is an important nucleotide-exchange factor the functionality of the UPS in sse1-200 sse2Δ cells transferred for cytosolic Hsp70, we tested whether overexpression of other to the non-permissive temperature, we first tested the well- Hsp70 nucleotide-exchange factors restored the degradation. The characterized temperature-sensitive protein GFP–Ubc9ts that armadillo-type nucleotide-exchange factor Fes1 supported growth of unfolds at the restrictive temperature (Betting and Seufert, 1996). the sse1-200 sse2Δ strain at 37°C when overexpressed (Fig. 2A) Paradoxically, GFP–Ubc9ts was previously reported to display a (Sadlish et al., 2008) but failed to restore the degradation of the degradation defect in sse1Δ cells (Escusa-Toret et al., 2013; UFD pathway substrate UbV76–Ura3–HA (Fig. 2B). Similarly, Kaganovich et al., 2008), but using the sse1-200 sse2Δ strain we overexpression of the structurally unrelated nucleotide-exchange found no evidence for significant stabilization (Fig. 3A). Similarly, factor Snl1ΔN (Snl1 lacking its N-terminal transmembrane domain) following the protein levels and bioluminescence of the derivative ts failed to restore the degradation of UbV76–Ura3–HA or the ubiquitin- yNluc–EGFP–Ubc9 (Masser et al., 2016) revealed that independent substrate ODS–Ura3–HA (Fig. 2C,D) (Sondermann degradation of this proteasome substrate was not dependent on et al., 2002). In striking contrast, expressing the Hsp110 Sse2 from Hsp110 and the UPS degradation was functional in sse1-200 sse2Δ the SSE1 promoter completely rescued the growth defect and the cells following a shift to 37°C (Fig. 3B,C). It is likely that, during substrate degradation phenotype (Fig. 2E,F). Thus, our data biogenesis in sse1Δ cells, GFP–Ubc9ts fails to fold and accumulate demonstrate that either of the two Hsp110 proteins Sse1 or Sse2, as stable aggregates that do not turn over (Gowda et al., 2013).

Fig. 2. Overexpressed Sse2 but not the Hsp70 nucleotide exchange factor Fes1 can mediate the function of Hsp110 in proteasomal degradation. (A) Growth phenotype of the sse1-200 sse2Δ strain transformed with plasmids carrying SSE1 or sse1-G205D (an ATP-binding mutant) or overexpressing

(OE) FES1. Cell suspensions were spotted on to solid medium and incubated at 37°C. (B) Degradation of the UFD substrate UbV76–Ura3–HA was analysed as in Fig. 1A. A quantification of the specific UbV76–Ura3–HA signal is presented in the right panel. Error bars denote s.d.; n≥3. (C) Degradation of UbV76–Ura3–HA was examined in sse1-200 sse2Δ cells transformed with empty vector control (VC) or plasmids carrying SSE1 or overexpressing Snl1ΔN. The asterisk (*) denotes a cross-reactive band. (D) Degradation of the ubiquitin-independent proteasome substrate ODS–Ura3–HA in sse1-200 sse2Δ cells transformed with plasmids carrying SSE1 or overexpressing Snl1ΔN. Western blot signals were quantified from n≥3 experiments. Error bars denote s.d. (E) Growth phenotype of the sse1-200 sse2Δ strain transformed with plasmids carrying SSE1, sse1-G205D or SSE2 under control of the SSE1 promoter. Cells were spotted on to

YPD agar plates and incubated at 37°C for 4 days. (F) Same as in Fig. 1B. The yeast strains in E were used to study the UbV76–Ura3–HA turnover rate at 37°C. The asterisk (*) denotes a cross-reactive band. Western blot signals were quantified from n≥3. Error bars denote s.d. Journal of Cell Science

4 RESEARCH ARTICLE Journal of Cell Science (2018) 131, jcs210948. doi:10.1242/jcs.210948

Fig. 3. The UPS is functional upon inactivation of Sse1 by heat shock. (A) Degradation of the temperature-sensitive and misfolding protein GFP– Ubc9ts in the sse1-200 sse2Δ strain transformed with VC or a corresponding plasmid carrying SSE1. GFP– Ubc9ts expression was induced with 2% galactose at 25°C and cells were shifted to 37°C in a medium containing 2% glucose. Quantification of the western blot signals is presented in the lower panel. Error bars denote s.d.; n≥6. (B) Degradation of yNluc–EGFP– Ubc9ts was analysed in strains as in A but on glucose medium at the indicated time after treatment with cycloheximide (CHX) to stop translation. The asterisks (*) represents degradation products. Error bars denote s.d.; n≥3. (C) Decay of yNIuc–EGFP– Ubc9ts in B was followed by undertaking bioluminescence measurements. Error bars denotes s.d.; n≥5. (D) Degradation of Stp1–HA in strains as in A. Cells were grown in amino acid-free medium, shifted to 37°C and CHX was added. Samples were analysed after the indicated time. Error bars denote s.d.; n≥3.

We also analysed the rapid proteasomal degradation of the latent sse2Δ and ssa1-45 cells (Fig. 4B). Normalized to the 19S regulatory cytoplasmic transcription factor Stp1 (Pfirrmann et al., 2010) in the particle subunit protein Rpt5, the ubiquitin signal was reduced by sse1-200 sse2Δ strain at the non-permissive temperature. Again, no 60% for the sse1-200 sse2Δ cells and 50% for the ssa1-45 cells significant stabilization was observed upon Hsp110 inactivation compared to the WT condition. Thus, the delivery of ubiquitin (Fig. 3D) supporting the notion that the UPS overall remains conjugates to the proteasome is impeded upon Hsp110 inactivation functional in sse1-200 sse2Δ cells transiently shifted to 37°C and and the conjugates remain associated with Hsp70. that Hsp110 together with Hsp70 plays a specialized role in the degradation of a subset of proteasome substrates. Hsp110 interacts with the 19S regulatory particle and recruits Hsp70 to the proteasome Hsp70 associates with ubiquitin-modified proteins that are We focused on the physical interaction of Hsp70 and Hsp110 with targeted for Hsp110-dependent proteasomal degradation proteasomes. Consistent with proteomic data (Guerrero et al., 2006), We tested whether Hsp70 associates with the ubiquitin-modified Sse1 was found to co-purify with 26S proteasomes (Fig. 5A). proteins that accumulate following a heat shock by Importantly, to maintain the integrity of the 26S proteasome, the immunoprecipitating the constitutively expressed Hsp70 protein entire purification was performed in the presence of an ATP Ssa2. Soluble ubiquitin conjugates and the ubiquitin-binding shuttle regeneration system. Since Sse1 is released from Hsp70 in the factor Dsk2 co-purified with Ssa2 (Fig. 4A; Fig. S2A). Importantly, presence of ATP, the interaction was not likely to be mediated via in sse1-200 sse2Δ cells, the amounts of soluble ubiquitin-conjugates Hsp70 (Andréasson et al., 2008b; Raviol et al., 2006). To find the that associated with Ssa2 doubled, suggesting that Hsp110 has an Sse1 interaction site on the proteasome, we separately purified the important role in shuttling these substrates to downstream 20S core (Pre1–FLAG) and 19S regulatory (Rpt1–FLAG) particle degradation by the proteasome, perhaps by mediating their release from heat-shocked cells. The use of ATP-depleted conditions from Hsp70 by accelerating nucleotide exchange. Indeed, facilitated the separation of the two particles and also enabled Hsp70 immunoprecipitation of Sse1–mCherry from a system with an to remain associated with its substrates. Consistent with the ATP regeneration system resulted in dramatically reduced levels of localization of the ubiquitin receptors on the proteasome (Grice associated ubiquitin conjugates and Ssa1 (Fig. S2B). and Nathan, 2016), we found that Hsp70 (Ssa1 and Ssa2) as well as We directly tested whether Hsp110 promotes the delivery of Sse1 associated with the 19S regulatory particle (Fig. 5B). Hsp70-associated substrates to the proteasome by purifying the 26S Purification of mCherry-tagged Sse1 confirmed that the 19S proteasome and quantitatively assessing the association of regulatory particle subunit (Rpt5) remained associated both in the ubiquitylated proteins. Specifically, a FLAG-6×His tag was presence or absence of ATP, although with reduced levels under genetically fused to PRE1 and intact 26S proteasomes were ATP conditions (Fig. S3A). Upon analysis of Sse1 co-purification purified in the presence of ATP by performing anti-FLAG with 19S regulatory (Rpt5–FLAG) particles, we found an chromatography (Verma et al., 2000). Strikingly, compared to association in the absence of ATP or with the non-hydrolysable what was seen in WT cells, substantially less ubiquitylated proteins analogue γ-ATP, and again the interaction was strongly reduced in were associated with the 26S proteasomes purified from sse1-200 the presence of ATP (Fig. S3B). Calculation of the fraction of total Journal of Cell Science

5 RESEARCH ARTICLE Journal of Cell Science (2018) 131, jcs210948. doi:10.1242/jcs.210948

Fig. 4. Sse1 facilitates the shuttling of Hsp70- associated ubiquitin conjugates to the 26S proteasome. (A) WT and sse1-200 sse2Δ strains pre grown at 25°C were subjected to heat shock at 37°C for 30 min and Ssa2–HA was immunoprecipitated. Untagged strains were used as control. The asterisk (*) denotes a cross- reactive band. The quantification of the specific signal from Hsp70-bound ubiquitin conjugates from two independent experiments is presented in the right panel. Data shows the average of two independent experiments with individual data points. (B) 26S Pre1-FLAG proteasomes were purified from heat-shocked (37°C, 30 min) WT, sse1-200 sse2Δ and ssa1-45 cells in the presence of ATP. The relative co-purifying ubiquitin-conjugate signals from two independent experiments were quantified. Data shows average of two independent experiments with individual data points.

Sse1 that co-purified with the 19S proteasome showed that <1% of Arg) to avoid ubiquitylation. Western blot analysis revealed that the population eluted under our nucleotide-free experimental ODS–HA–APPY was expressed at levels below the detection limit, conditions. To further characterize the complex between Sse1 and but that inhibition of proteasomal degradation with MG132 for the 19S regulatory particle, we tested the Hsp70-binding mutant 40 min resulted in readily detectable protein accumulation Sse1-2,3. Indeed, we found that Sse1-2,3 forms a complex with the (Fig. 6B). Inactivation of the Hsp70-binding site in the APPY 19S regulatory particle (Fig. 5C). Finally, we asked whether Hsp70 peptide by replacing the three central leucine residues with alanine in turn depended on Hsp110 for its interaction with the proteasome. residues (ODS–HA–APPY*) still allowed for efficient turnover of Intriguingly, inactivation of Hsp110 (sse1-200 sse2Δ at 37°C) the protein by the proteasome. Hence ODS–HA–APPY, resulted in reduced Hsp70 association with the 19S regulatory independently of its interaction with Hsp70, efficiently targets particle (Fig. 5D; Fig. S3C). Thus, the data show that Sse1 interacts and is degraded by the proteasome. with the 19S regulatory particle of the proteasome and this Next, we tested the requirement for Hsp110 function in the interaction is sensitive to ATP. Moreover Sse1 is required for degradation and found that upon Hsp110 inactivation (sse1-200 Hsp70 to associate with the proteasome. sse2Δ, 30 min at 37°C) a robust signal accumulated (Fig. 6C). Complementation by SSE1 restored degradation while the Hsp70- Hsp110 is required for the efficient degradation of a high- binding and nucleotide-exchange deficient sse1-2,3 mutant did not. affinity proteasome substrate with a single defined Hsp70- Importantly, mutation of the Hsp70-binding site in the APPY binding site peptide made the degradation largely independent of Hsp110. Next, Finally, we considered the possibility that direct binding by we attempted to increase the Hsp70 association with ODS–HA– Hsp70 makes the proteasomal degradation of substrates dependent APPY by introducing a second Hsp70-binding peptide. The on Hsp110. If this were the case, inactivation of Hsp110 would resulting ODS–HA–2×APPY generally accumulated to higher result in loss of nucleotide exchange accompanied by slower steady state levels than the construct with a single APPY Hsp70- release kinetics from Hsp70 and impaired targeting to the binding site and still depended on Hsp110 for its degradation proteasome. To test this hypothesis, we made use of an artificial (Fig. 6D). These observations are consistent with the notion that and unstructured protein built from the ODS (the N-terminal Hsp70 binding impedes proteasomal degradation and that Hsp110 unstructured tail of the ubiquitin-independent substrate ODC that alleviates this inhibition. Thus, a single Hsp70-binding site renders binds the proteasome), two HA tags (HA) and the well- the protein dependent on Hsp110 for its efficient degradation characterized Hsp70-substrate peptide APPY (Gödderz et al., indicating that the mere interaction with Hsp70 overrides the 2011; Pfund et al., 2001; Yu et al., 2016b) (Fig. 6A). Importantly, targeting to the proteasome via ODS binding. Taken together with this ODS–HA–APPY protein was designed entirely without lysine the entire data set, we conclude that Hsp110 facilitates the targeting residues (lysine 28, 29 and 33 of the ODS sequence were changed to of Hsp70–substrate complexes to the proteasome. Journal of Cell Science

6 RESEARCH ARTICLE Journal of Cell Science (2018) 131, jcs210948. doi:10.1242/jcs.210948

(Fig. 7). In this model, the chaperone Hsp70 interacts with the substrates and keeps them in a soluble state and competent for ubiquitylation but inhibits their efficient targeting to the proteasome. Hsp110 then associates with Hsp70-ADP in complex with substrate, triggering nucleotide exchange and thereby promoting targeting to the proteasome. The binding of Hsp110 to the 19S regulatory particle potentially facilitates proteasomal docking of Hsp70–substrate complexes. By functioning in parallel to canonical targeting of ubiquitylated proteins to the proteasome, this pathway fast-tracks the delivery of ubiquitin-modified proteins and even enable non-ubiquitylated proteins to be targeted for degradation. The notion that Hsp70 and Hsp110 function directly in the targeting proteins to the proteasome is supported by several key observations. First, the rapid inactivation of Hsp110 results in accumulation of soluble and Hsp70-associated ubiquitin conjugates as well as specific ubiquitin-dependent and -independent proteasome substrates. The stable ubiquitin conjugates that fail to reach the 26S proteasome remain associated with Hsp70 or undergo aggregation. In the same experimental conditions, other proteasome substrates are degraded normally, ruling out a general proteolysis defect. Second, Hsp70 and ubiquitin-conjugate association with the proteasome depend on Hsp110. Finally, the highly defined ubiquitin-independent proteasome substrate ODS–HA–APPY depends on Hsp110 for efficient proteasomal degradation but only when it carries an Hsp70-binding site. These observations, together with the extensive literature on the transient high-affinity interactions between Hsp70 and its nucleotide-exchange factor Hsp110 (Andréasson et al., 2008b; Bracher and Verghese, 2015; Raviol et al., 2006), support the notion that Hsp110 is a dedicated nucleotide exchange factor that assists in the targeting of Hsp70- ADP and its associated substrates to the 26S proteasome complex. The interaction between Hsp110 and the 19S regulatory particle Fig. 5. Proteasome-associated Sse1 recruits Hsp70 at the 19S regulatory suggests that Hsp110 functions as an Hsp70-docking site at the particle of 26S proteasome. (A) FLAG–6His-tagged 26S proteasomes proteasome that coordinates substrate recruitment with localized FH (Pre1 ) were immunoprecipitated in the presence of an ATP-regenerating release from the chaperone. Detailed characterization of this system supplemented with 5 mM ATP (ARS) from soluble extracts prepared interaction will enable direct testing of its requirement for the from sse1-200 sse2Δ cells transformed with a centromeric plasmid vector control (VC) or an equivalent plasmid carrying the SSE1 locus (SSE1). Pre- function of Hsp110 in proteasomal degradation. Nevertheless, the grown cells at 25°C were heat shocked at 37°C for 30 min before harvest. data presented here demonstrate that Hsp110 functions together Purified proteasomes were eluted with FLAG peptide and subjected to western with Hsp70 in the targeting of ubiquitin-modified and unmodified blot analysis. (B) Western blot analysis of the 19S and 20S particle bound Ssa1 substrates to the proteasome. or Ssa2 (Ssa1/2), and Sse1 chaperones purified from chromosomally tagged We find that Hsp70 binding to ODS–HA–APPY inhibited FH FH Rpt1 or Pre1 , respectively, in WT strains. Silver staining of the purified 19S proteasomal degradation specifically upon Hsp110 inactivation and and 20S particle of the yeast proteasome. Data represents two independent experiments. (C) sse1-200 sse2Δ cells carrying RPT1FH was transformed with that introduction of a second Hsp70-binding site impaired the a centromeric plasmid vector (VC) or an equivalent plasmid carrying SSE1 or degradation even further. We consider it a possibility that the sse1-2,3. Cells were heat shocked at 37°C for 30 min before harvest. The 19S behaviour of this well-defined and ubiquitin-independent model regulatory particle was purified in the absence of the ARS, and the interaction substrate reveals a fundamental function of Hsp70. According to this of Sse1 chaperones with 19S RP was analysed by western blotting. The levels understanding, Hsp70 association keeps proteins soluble and available of co-purifying Sse1 were quantified from two independent experiments. The for interaction with various folding and PQC factors but inhibits specific co-purifying Sse1 signals were subtracted from non-specific targeting to the proteasome. Hsp110 overcomes this inhibition of background signal. Data shows average of two independent experiments with individual data points. (D) Same as in C, but the interaction of Ssa1/2 with a proteasomal targeting via nucleotide exchange interactions with purified 19S regulatory particle was analysed by western blotting in the Hsp70–substrate complexes. Currently, we consider two possible presence or absence of Sse1 function. The levels of 19S regulatory particle- explanations for the failed targeting when Hsp110 is inactivated: either associated Ssa1/2 were quantified from two independent experiments. The Hsp70 keeps its associated substrates spatially separated from the non-specific background signal was subtracted from specific co-purifying proteasome (sequestration), or Hsp110 functions at the proteasome to Ssa1/2 signals. Data shows average of two independent experiments with release the substrates from Hsp70, which sterically hinders initiation of individual data points. translocation into the proteolytic chamber. Hsp110 (Sse1) has previously been proposed to be involved DISCUSSION specifically at the ubiquitylation step of proteasomal degradation of We find that Hsp110 and Hsp70 are important factors for the misfolded proteins (Heck et al., 2010; Mandal et al., 2010). These proteasomal degradation of ubiquitin-modified and unmodified studies relied on the use of the pleiotropic sse1Δ mutation, which proteins and propose a model that summarizes our findings unfortunately causes major cytosolic protein folding problems and Journal of Cell Science

7 RESEARCH ARTICLE Journal of Cell Science (2018) 131, jcs210948. doi:10.1242/jcs.210948

Fig. 6. The degradation of an Hsp70-bound ubiquitin- independent model substrate targeted to the proteasome depends on Hsp110. (A) Representation of the model substrate ODS–HA–APPY (85 amino acids, 9.4 kDa) including the proteasome-binding tail ODS from yeast ODC, an HA tag and the Hsp70-binding peptide APPY. Leu to Ala mutations (*) abolish high-affinity Hsp70 binding to APPY. (B) Western blot analysis of pdr5Δ cells (the pdr5Δ deletion is introduced to reduce the efflux of MG132) expressing ODS–HA–APPY or its derivative with defective Hsp70-binding (ODS–HA– APPY*). Cells growing at 25°C were treated with 10 µM of the proteasome inhibitor MG132 (+) and samples were prepared after 40 min. (C) sse1-200 sse2Δ cells expressing ODS–HA–APPY or ODS–HA–APPY* together with a centromeric plasmid vector (VC), or an equivalent plasmid carrying SSE1 or sse1-2,3, growing at 25°C were transferred to 37°C and samples were taken after 30 min. Graph shows the comparative quantification of the ODS–HA–APPY or ODS–HA–APPY* signal from six independent biological replicates. (D) Same as in C, except that the steady-state levels of ODS–HA–2×APPY were monitored in sse1-200 sse2Δ cells carrying centromeric plasmid vector (VC) or an equivalent plasmid expressing SSE1. The graph shows comparative quantification of the ODS–HA–APPY and ODS–HA– 2×APPY levels. Data points and mean are shown for n≥3 in B and D.

transcriptional stress responses (Koplin et al., 2010; Willmund et al., underlie at least some of the degradation defects that have been 2013; Yam et al., 2005). Aggregation of misfolded model proteins reported for sse1Δ cells (Gowda et al., 2013). For this reason, we do expressed in this mutant or pleiotropic impairment of the UPS not find sse1Δ strains very useful for the mechanistic interpretation of the involvement of Hsp110 in the UPS system. By using transient inactivation of Hsp110 (sse1-200 sse2Δ), we document that the UPS remains largely functional with ongoing turnover of several proteasome substrates including the destabilized model protein GFP–Ubc9ts and the latent transcriptional factor Stp1. GFP–Ubc9ts has previously been shown not to turn over in sse1Δ cells (Escusa- Toret et al., 2013) demonstrating that transient Hsp110 inactivation results in a more specific phenotype. This phenotype includes the impaired degradation of heat-induced ubiquitin conjugates, two distinct UFD pathway substrates and the ubiquitin-independent proteasome substrates ODS–Ura3–HA and ODS–HA–APPY. Importantly, proteins accumulate not only in the aggregate fraction but also as soluble species, indicating that Hsp110 has a role in their degradation beyond simply keeping proteins soluble. Thus, the use of rapid and transient Hsp110 inactivation allowed us to unmask the function of this protein in UPS. Our findings raise the question of whether proteasome substrates have special biophysical properties that necessitate escorting chaperones for their proteasomal targeting. Indeed, some proteasome substrates are known to be prone to aggregation, Fig. 7. Model for the role of Hsp70 and Hsp110 in the degradation of including proteins with intrinsically disordered or unstructured proteasome substrates. Hsp70 forms dynamic complexes with substrate regions (Erales and Coffino, 2014; Jiménez et al., 2012; Ng et al., proteins. Hsp110 binds to the 19S regulatory particle and recruits Hsp70-ADP in 2013). Moreover, polyubiquitin chains have recently been complex with substrate, and then triggers nucleotide exchange and thereby promoting the targeting of Hsp70 substrates for proteasomal degradation. In the shown to form aggregate fibrils under conditions of heat and absence of Hsp110 function, Hsp70 and its associated substrates fail to reach shearing (Morimoto et al., 2015), and ubiquitylation has the proteasome, which enhances their aggregation. See Discussion for details. potential to destabilize and unfold proteins upon conjugation Journal of Cell Science

8 RESEARCH ARTICLE Journal of Cell Science (2018) 131, jcs210948. doi:10.1242/jcs.210948

(Hagai and Levy, 2010; Morimoto et al., 2016). Thus, the biophysical plasmid pDG23 is an YCplac111 derivative that contains the CUP1 – characteristics of proteasome substrates, including ubiquitin- -promoter-driven Ubv76 GFP coding sequence derived from modified and ubiquitin-independent species, predict a role for pYES2-Ubv76-GFP (Heessen et al., 2003). The plasmid pGA4 contains the yNluc–GFP–Ubc9ts coding sequence expressed from the TDH3 general chaperones, such as Hsp70, in escorting aggregation-prone – – proteins to the 26S proteasome. According to our model, Hsp70 promoter (Masser et al., 2016). ODS HA APPY and derivatives were constructed through gene synthesis and introduced into a centromeric URA3 association with any protein that presents hydrophobic Hsp70- vector under the TDH3 promoter. binding sites would result in the rapid and Hsp110-dependent targeting to the proteasome. As an outcome, the chaperone- Preparation of total cell extracts dependent proteasome-targeting pathway selects misfolded and Total protein extracts were prepared as described previously (Silve et al., aggregation-prone proteins for fast-tracked degradation and even 1991). Briefly, 1 ml cell culture was mixed with 250 μl ice-cold NaOH. acts on non-ubiquitylated species. After 10 min of incubation on ice, 250 μl 50% trichloroacetic acid was In mammalian cells, the nucleotide exchange factor BAG-1 is added. Following centrifugation at 21,100 g for 10 min, the supernatant was part of proteasomal degradation pathways. BAG-1 interacts with the aspirated off, the pellet material were rinsed with 1 M Tris base and proteasome via an ubiquitin-like (UBL) domain, and ubiquitylation resuspended in 50 μl SDS-sample buffer (4% SDS, boiling for 5 min). of BAG-1 by the E3 ligase CHIP enhances the association (Alberti Typically total proteins extracted from cells at a 0.2 optical density at 600 nm et al., 2002; Lüders et al., 2000). Thus, a dedicated UBL domain (OD600) per ml were separated by SDS-PAGE and quantitatively analysed by western blotting. with low affinity for the proteasome acts in concert with a conditional ubiquitin chain that provides additional binding sites Fractionation of cell extracts for the proteasome. The here-identified Hsp110-dependent 100 ml of logarithmically growing cells (OD600=1.0) was harvested by proteasomal pathway may function in an analogous manner, centrifugation and transferred into 1.5 ml screw-capped tubes, frozen in perhaps representing an evolutionary more fundamental liquid nitrogen and stored at −80°C. Two pre-chilled stainless steel balls proteasomal pathway. Accordingly, Hsp110 interacts with the 19S (8.0 mm) were added to the screw-capped tubes and the frozen pellets were regulatory particle and engages Hsp70-ADP via transient ground into a fine powder in a Mini-Beadbeater (BioSpec Products, nucleotide-exchange interactions. Hsp70 substrates that carry Bartlesville, OK) at 2500 rpm (three times for 30 s each with repeated ubiquitin chains or engage the proteasome directly via preferred freezing in liquid nitrogen). The frozen ground pellet material was μ unstructured tails (Fishbain et al., 2015) provide additional binding suspended in 500 l ice-cold LWB150 buffer [40 mM Hepes-KOH pH 7.4, 150 mM KCl, 5 mM MgCl , 5% (v/v) glycerol), 0.2% (v/v) Triton interfaces that accelerate the degradation. Hence, such a kinetic 2 X-100], supplemented with Complete protease inhibitor cocktail (Roche mechanism ensures that the Hsp110-dependent proteasome Diagnostics), 1 mM PMSF, 40 mM NEM and an ATP-regenerating system pathway selectively targets proteins that stay associated with with 5 mM ATP, as described previously (Verma et al., 2000). Alternatively, Hsp70 for prolonged periods of time as a result of failed folding. 10 ml cells at OD600=1 were harvested, suspended in 200 μl LWB150 buffer The concept is supported by the study of the Hsp70 client Tau supplemented with protease inhibitors, 2 mM PMSF, 40 mM NEM and revealing that tight complexes between Hsp70 and Tau results in subjected to glass bead lysis for 30 s in 0.5 ml screw capped tubes. Unlysed enhanced turnover while transient interactions favour retention cells were removed by centrifugation at 21,100 g for 30 s. Protein (Young et al., 2016). In future studies, it will be interesting to concentrations were determined by means of the Bradford assay and equal characterize in detail the binding of Hsp110 to the 26S proteasome amounts of lysate were used for centrifugation at 21,100 g for 10 min at 4°C. to better understand the regulation of Hsp70 in chaperone-assisted After centrifugation the total, soluble and pellet fractions were boiled with SDS sample buffer and equal volumes of sample were used for western blot protein degradation and the quality control pathway. analysis.

MATERIALS AND METHODS Western blot analysis and protein stability Yeast strains and plasmids For protein stability assays, logarithmically growing cells grown at 25°C All yeast strains used in this study are derivatives of the BY4741 genetic were treated with 100 mg/l cycloheximide. Cells were harvested and the background. Details of the strains, plasmids and primers are listed in protein levels were determined by western blot analysis. Quantification of Tables S1, S2 and S3, respectively. The sse1-200 sse2Δ yeast strain signal intensities was performed as described previously (Gowda et al., (CAY1337), the SSE1-ymCherry strain (CAY1357) and the WT strain 2013). Immunoblotting was carried out with antibodies specific for GFP (1: (CAY1015) have been described previously (Kaimal et al., 2017). To tag the 5000 dilution; cat. no 11814460001, Roche Life Science), hemagglutinin proteasome (PRE1-FLAG-6×His) pJD416 (Verma et al., 2000) was (HA) (1:2500 dilution; cat. no 12013819001, Roche Life Science and linearized with the BsmI restriction enzyme and used to transform 1:1000 dilution; cat. no 901515, Millipore), Pgk1 (1:10,000 dilution; cat. no CAY1015, CAY1337 and the ssa1-45 strain CAY1250 to URA+.To 459250, Thermo Fisher Scientific), ubiquitin (1:1000 dilution; cat. no obtain genomic tagging of RPT1-FLAG-6×His, the assembly PCR product Z0458, DAKO), Ssa1/2 (1:50,000 dilution; rabbit serum), Sse1 (1:50,000 containing the C-terminal segment of RPT1 fused in frame to the FLAG- dilution; rabbit serum), anti-rabbit-IgG conjugated to horseradish 6×His-TCUP1-Ura3 open reading frame (ORF) was generated by using the peroxidase (HRP) (1:10,000 dilution; cat. no 31460, Thermo Fisher GA229, GA230, GA231 and GA232 primers. The assembly PCR product Scientific), anti-mouse-IgG conjugated to HRP (1:10,000 dilution; cat. no was used to transform CAY1015 and CAY1337 to URA+. The positive 32230, Thermo Fisher Scientific), anti-rabbit-IgG conjugated to IRDye 680 clones of RPT1FH were verified by analytical PCR by using the GA233 and (1:10,000 dilution; cat. no 926-68071, Li-Cor Odyssey), anti-rabbit-IgG GA234 primers. CAY1038 (hsp104Δ) is a derivative of BY4741 conjugated to IRDye 800 (1:10,000 dilution; cat. no 92632211, Li-Cor (EUROSCARF, Germany). A fragment encompassing SSA2-HA was Odyssey), anti-mouse-IgG conjugated to IRDye 680 (1:10,000 dilution; cat. amplified from CAY1191 by using primers GA235 and GA236 and was no 926-68070, Li-Cor Odyssey) and anti-mouse 800 (1:10,000 dilution; cat. introduced into CAY1337. Plasmid pCA503 is an SSE1-harbouring no 92632210, Li-Cor Odyssey). Proteins were visualized by either using derivative of the HIS3 vector pCA502 (Andréasson et al., 2008a) and its chemiluminescence HRP substrate and enhancer (SuperSignal West Dura derivative pCA899 carries the sse1-2,3 allele [mutations: A280T, Extended-Duration Substrate) or SuperSignal™ West Femto maximum N281T, E572Y, E575A (Polier et al., 2008)]. The derivative pCA1018 sensitivity Substrate (Thermo Fisher Scientific), or near-infrared was obtained by introducing the G205D mutation by site-directed fluorescence detection (Odyssey Fc, LI-COR Biosciences, Lincoln, NE). mutagenesis. In pCA1019 the entire SSE1 ORF of pCA503 was replaced Protein quantification was performed with the Image Studio Lite software with the SSE2 ORF by seamless homologous recombination in yeast. The (LI-COR Biosciences) or ImageJ software. Journal of Cell Science

9 RESEARCH ARTICLE Journal of Cell Science (2018) 131, jcs210948. doi:10.1242/jcs.210948

Bioluminescence activity assay Andréasson, C., Fiaux, J., Rampelt, H., Mayer, M. P. and Bukau, B. (2008b). Bioluminescence of yNluc-EGFP-Ubc9ts was monitored as described Hsp110 is a nucleotide-activated exchange factor for Hsp70. J. Biol. Chem. 283, previously (Masser et al., 2016). Yeast cultures grown at 25°C were treated 8877-8884. ts Ben-Nissan, G. and Sharon, M. (2014). Regulating the 20S proteasome ubiquitin- with cycloheximide and the decay of yNluc-EGFP-Ubc9 at 37°C was independent degradation pathway. Biomolecules 4, 862-884. followed with an the Orion II Microplate Luminometer (Berthold Detection Betting, J. and Seufert, W. (1996). A yeast Ubc9 mutant protein with temperature- Systems, Pforzheim, Germany). sensitive in vivo function is subject to conditional proteolysis by a ubiquitin- and proteasome-dependent pathway. J. Biol. Chem. 271, 25790-25796. Immunoprecipitation Bracher, A. and Verghese, J. (2015). The nucleotide exchange factors of Hsp70 Logarithmic phase yeast cultures grown at 25°C were shifted to 37°C for molecular chaperones. Front. Mol. Biosci. 2, 10. Ciechanover, A. (1994). The ubiquitin-proteasome proteolytic pathway. Cell 79, 30 min before harvesting (2500 g for 5 min). Cell extracts were prepared by 13-21. grinding in liquid nitrogen (see above) and the soluble fraction containing Dantuma, N. P., Lindsten, K., Glas, R., Jellne, M. and Masucci, M. G. (2000). Ssa2–HA or Sse1–mCherry was incubated with Anti-HA Affinity Matrix Short-lived green fluorescent proteins for quantifying ubiquitin/proteasome- from rat IgG1 (Roche Diagnostics) or RFP-Trap® (Chromotek) either dependent proteolysis in living cells. Nat. Biotechnol. 18, 538-543. without or with an ATP-regenerating system supplemented with 5 mM ATP Demand, J., Alberti, S., Patterson, C. and Höhfeld, J. (2001). Cooperation of a (Sigma-Aldrich, St Louis, MO). After extensive washing, the bound ubiquitin domain protein and an E3 ubiquitin ligase during chaperone/proteasome material was eluted with SDS sample buffer. coupling. Curr. Biol. 11, 1569-1577. Dragovic, Z., Broadley, S. A., Shomura, Y., Bracher, A. and Hartl, F. U. (2006). Molecular chaperones of the Hsp110 family act as nucleotide exchange factors of Purification of proteasomes Hsp70s. EMBO J. 25, 2519-2528. Proteasomes were purified as described previously (Verma et al., 2000). Eisele, F. and Wolf, D. H. (2008). Degradation of misfolded protein in the cytoplasm Briefly, logarithmically growing yeast cells at 25°C were heat stressed at 37°C is mediated by the ubiquitin ligase Ubr1. FEBS Lett. 582, 4143-4146. ̈ for 30 min before harvesting. Soluble cell extracts were prepared as Elsasser, S., Chandler-Militello, D., Muller, B., Hanna, J. and Finley, D. (2004). Rad23 and Rpn10 serve as alternative ubiquitin receptors for the proteasome. described above and proteasomes were purified either in the presence or J. Biol. Chem. 279, 26817-26822. absence of the ATP-regenerating system supplemented with 2 mM ATP or Erales, J. and Coffino, P. (2014). Ubiquitin-independent proteasomal degradation. ® γATP by using Anti-FLAG M2 Affinity Gel or EZview™ Red ANTI- Biochim. Biophys. Acta 1843, 216-221. FLAG® M2 Affinity Gel (Sigma-Aldrich, St Louis, MO). After extensive Escusa-Toret, S., Vonk, W. I. M. and Frydman, J. (2013). Spatial sequestration of washing, bound proteasomes were eluted by incubation of the beads with misfolded proteins by a dynamic chaperone pathway enhances cellular fitness 50 µl (150 µg/ml) FLAG peptide (Sigma-Aldrich) at 4°C and were during stress. Nat. Cell Biol. 15, 1231-1243. visualized by silver staining and western blotting. Fang, N. N., Ng, A. H. M., Measday, V. and Mayor, T. (2011). Hul5 HECT ubiquitin ligase plays a major role in the ubiquitylation and turnover of cytosolic misfolded proteins. Nat. Cell Biol. 13, 1344-1352. Acknowledgements Fang, N. N., Chan, G. T., Zhu, M., Comyn, S. A., Persaud, A., Deshaies, R. J., ̈ We thank R. Jurgen Dohmen (University of Cologne, Germany) and Per Rotin, D., Gsponer, J. and Mayor, T. (2014). Rsp5/Nedd4 is the main ubiquitin O. Ljungdahl (Stockholm University, Sweden) for valuable discussions. We also ligase that targets cytosolic misfolded proteins following heat stress. Nat. Cell Biol. thank Palnimurugan Rangasamy (Centre for Cell and Molecular Biology, India) for 16, 1227-1237. support with reagent sharing and for discussion. The plasmid pCA047 is a kind gift Fishbain, S., Inobe, T., Israeli, E., Chavali, S., Yu, H., Kago, G., Babu, M. M. and from P. O. Ljungdahl. The plasmids pJD416 and pDG258 were kind gifts from Matouschek, A. (2015). Sequence composition of disordered regions fine-tunes R. J. Dohmen and pDG23 were obtained from Nico Dantuma and Daniela protein half-life. Nat. Struct. Mol. Biol. 22, 214-221. Gödderz (Karolinska Institute, Sweden). The plasmid p413TEF-FLAG-SSE1sbd Forloni, G., Terreni, L., Bertani, I., Fogliarino, S., Invernizzi, R., Assini, A., CEN/ARS HIS3 is a kind gift from Tommer Ravid (The Hebrew University of Ribizzi, G., Negro, A., Calabrese, E., Volonté, M. A. et al. (2002). Protein Jerusalem). misfolding in Alzheimer’s and Parkinson’s disease: genetics and molecular mechanisms. Neurobiol. Aging 23, 957-976. Competing interests Garcia, V. M., Nillegoda, N. B., Bukau, B. and Morano, K. A. (2017). Substrate The authors declare no competing or financial interests. binding by the yeast Hsp110 nucleotide exchange factor and molecular chaperone, Sse1, is not obligate for its biological activities. Mol. Biol. Cell 28, Author contributions 2066-2075. Glickman, M. H. and Ciechanover, A. (2002). The ubiquitin-proteasome proteolytic Conceptualization: G.K., C.A.; Methodology: G.K., C.A.; Validation: G.K.; Formal pathway: destruction for the sake of construction. Physiol. Rev. 82, 373-428. analysis: G.K.; Investigation: G.K., C.A.; Resources: G.K., C.A.; Data curation: G.K.; Gödderz, D., Schäfer, E., Palanimurugan, R. and Dohmen, R. J. (2011). The Writing - original draft: G.K., C.A.; Writing - review & editing: G.K., C.A.; Visualization: N-terminal unstructured domain of yeast ODC functions as a transplantable and G.K., C.A.; Supervision: C.A.; Project administration: C.A.; Funding acquisition: C.A. replaceable ubiquitin-independent degron. J. Mol. Biol. 407, 354-367. Gowda, N. K. C., Kandasamy, G., Froehlich, M. S., Dohmen, R. J. and Funding Andreasson, C. (2013). Hsp70 nucleotide exchange factor Fes1 is essential for This work was supported by the Vetenskapsrådet (Swedish Research Council) [2015- ubiquitin-dependent degradation of misfolded cytosolic proteins. Proc. Natl. Acad. 05094-3 to C.A.], the Cancerfonden (Swedish Cancer Society) (CAN2016/361 to Sci. USA 110, 5975-5980. C.A.) and Carl Tryggers stiftelse för vetenskaplig forskning (CTS15-30 to C.A.). Gowda, N. K. C., Kaimal, J. M., Masser, A. E., Kang, W., Friedlander, M. R. and Andreasson, C. (2016). Cytosolic splice isoform of Hsp70 nucleotide exchange Supplementary information factor Fes1 is required for the degradation of misfolded proteins in yeast. Mol. Biol. Supplementary information available online at Cell 27, 1210-1219. ̈ http://jcs.biologists.org/lookup/doi/10.1242/jcs.210948.supplemental Gowda, N. K. C., Kaimal, J. M., Kityk, R., Daniel, C., Liebau, J., Ohman, M., Mayer, M. P. and Andréasson, C. (2018). Nucleotide exchange factors Fes1 and HspBP1 mimic substrate to release misfolded proteins from Hsp70. Nat. Struct. Mol. Biol. 25, 83-89. References Grice, G. L. and Nathan, J. A. (2016). The recognition of ubiquitinated proteins by Abrams, J. L., Verghese, J., Gibney, P. A. and Morano, K. A. (2014). Hierarchical the proteasome. Cell. Mol. Life Sci. 73, 3497-3506. functional specificity of cytosolic heat shock protein 70 (Hsp70) nucleotide Guerrero, C., Tagwerker, C., Kaiser, P. and Huang, L. (2006). An integrated mass exchange factors in yeast. J. Biol. Chem. 289, 13155-13167. spectrometry-based proteomic approach: quantitative analysis of tandem affinity- Alberti, S., Demand, J., Esser, C., Emmerich, N., Schild, H. and Höhfeld, J. purified in vivo cross-linked protein complexes (QTAX) to decipher the 26 S (2002). Ubiquitylation of BAG-1 suggests a novel regulatory mechanism during proteasome-interacting network. Mol. Cell. Proteomics 5, 366-378. the sorting of chaperone substrates to the proteasome. J. Biol. Chem. 277, Guerriero, C. J., Weiberth, K. F. and Brodsky, J. L. (2013). Hsp70 targets a 45920-45927. cytoplasmic quality control substrate to the San1p ubiquitin ligase. J. Biol. Chem. Andréasson, C., Fiaux, J., Rampelt, H., Druffel-Augustin, S. and Bukau, B. 288, 18506-18520. (2008a). Insights into the structural dynamics of the Hsp110-Hsp70 interaction Hagai, T. and Levy, Y. (2010). Ubiquitin not only serves as a tag but also assists reveal the mechanism for nucleotide exchange activity. Proc. Natl. Acad. Sci. USA degradation by inducing protein unfolding. Proc. Natl. Acad. Sci. USA 107,

105, 16519-16524. 2001-2006. Journal of Cell Science

10 RESEARCH ARTICLE Journal of Cell Science (2018) 131, jcs210948. doi:10.1242/jcs.210948

Heck, J. W., Cheung, S. K. and Hampton, R. Y. (2010). Cytoplasmic protein quality Pfund, C., Huang, P., Lopez-Hoyo, N. and Craig, E. A. (2001). Divergent control degradation mediated by parallel actions of the E3 ubiquitin ligases Ubr1 functional properties of the ribosome-associated molecular chaperone Ssb and San1. Proc. Natl. Acad. Sci. USA 107, 1106-1111. compared with other Hsp70s. Mol. Biol. Cell 12, 3773-3782. Heessen, S., Dantuma, N. P., Tessarz, P., Jellne, M. and Masucci, M. G. (2003). Polier, S., Dragovic, Z., Hartl, F. U. and Bracher, A. (2008). Structural basis for the Inhibition of ubiquitin/proteasome-dependent proteolysis in Saccharomyces cooperation of Hsp70 and Hsp110 chaperones in protein folding. Cell 133, cerevisiae by a Gly-Ala repeat. FEBS Lett. 555, 397-404. 1068-1079. Hohfeld, J. and Jentsch, S. (1997). GrpE-like regulation of the hsc70 chaperone by Raviol, H., Sadlish, H., Rodriguez, F., Mayer, M. P. and Bukau, B. (2006). the anti-apoptotic protein BAG-1. EMBO J. 16, 6209-6216. Chaperone network in the yeast cytosol: Hsp110 is revealed as an Hsp70 Husnjak, K., Elsasser, S., Zhang, N., Chen, X., Randles, L., Shi, Y., Hofmann, K., nucleotide exchange factor. EMBO J. 25, 2510-2518. Walters, K. J., Finley, D. and Dikic, I. (2008). Proteasome subunit Rpn13 is a Sadlish, H., Rampelt, H., Shorter, J., Wegrzyn, R. D., Andreasson, C., Lindquist, novel ubiquitin receptor. Nature 453, 481-488. S. and Bukau, B. (2008). Hsp110 chaperones regulate prion formation and Jiménez, B., Ugwu, F., Zhao, R., Ortı,́ L., Makhnevych, T., Pineda-Lucena, A. and Houry, W. A. (2012). Structure of minimal tetratricopeptide repeat domain propagation in S. cerevisiae by two discrete activities. PLoS ONE 3, e1763. protein Tah1 reveals mechanism of its interaction with Pih1 and Hsp90. J. Biol. Shaner, L., Trott, A., Goeckeler, J. L., Brodsky, J. L. and Morano, K. A. (2004). Chem. 287, 5698-5709. The function of the yeast molecular chaperone Sse1 is mechanistically distinct Johnson, E. S., Ma, P. C. M., Ota, I. M. and Varshavsky, A. (1995). A proteolytic from the closely related hsp70 family. J. Biol. Chem. 279, 21992-22001. pathway that recognizes ubiquitin as a degradation signal. J. Biol. Chem. 270, Shiber, A., Breuer, W., Brandeis, M. and Ravid, T. (2013). Ubiquitin conjugation 17442-17456. triggers misfolded protein sequestration into quality control foci when Hsp70 Kaganovich, D., Kopito, R. and Frydman, J. (2008). Misfolded proteins partition chaperone levels are limiting. Mol. Biol. Cell 24, 2076-2087. between two distinct quality control compartments. Nature 454, 1088-1095. Silve, S., Volland, C., Garnier, C., Jund, R., Chevallier, M. R. and Haguenauer- Kaimal, J. M., Kandasamy, G., Gasser, F. and Andreasson, C. (2017). Tsapis, R. (1991). Membrane insertion of uracil permease, a polytopic yeast Coordinated Hsp110 and Hsp104 activities power protein disaggregation in plasma membrane protein. Mol. Cell. Biol. 11, 1114-1124. Saccharomyces cerevisiae. Mol. Cell. Biol. 37, e00027-17d. Sondermann, H., Ho, A. K., Listenberger, L. L., Siegers, K., Moarefi, I., Wente, Koplin, A., Preissler, S., Ilina, Y., Koch, M., Scior, A., Erhardt, M. and Deuerling, S. R., Hartl, F. U. and Young, J. C. (2002). Prediction of novel Bag-1 homologs E. (2010). A dual function for chaperones SSB-RAC and the NAC nascent based on structure/function analysis identifies Snl1p as an Hsp70 co-chaperone polypeptide-associated complex on ribosomes. J. Cell Biol. 189, 57-68. in Saccharomyces cerevisiae. J. Biol. Chem. 277, 33220-33227. Lee, D. H., Sherman, M. Y. and Goldberg, A. L. (2016). The requirements of yeast Su, V. and Lau, A. F. (2009). Ubiquitin-like and ubiquitin-associated domain Hsp70 of SSA family for the ubiquitin-dependent degradation of short-lived and proteins: significance in proteasomal degradation. Cell. Mol. Life Sci. 66, abnormal proteins. Biochem. Biophys. Res. Commun. 475, 100-106. 2819-2833. Lüders, J., Demand, J. and Höhfeld, J. (2000). The ubiquitin-related BAG-1 Takeuchi, J., Chen, H. and Coffino, P. (2007). Proteasome substrate degradation provides a link between the molecular chaperones Hsc70/Hsp70 and the requires association plus extended peptide. EMBO J. 26, 123-131. proteasome. J. Biol. Chem. 275, 4613-4617. Theodoraki, M. A., Nillegoda, N. B., Saini, J. and Caplan, A. J. (2012). A network Mandal, A. K., Gibney, P. A., Nillegoda, N. B., Theodoraki, M. A., Caplan, A. J. and Morano, K. A. (2010). Hsp110 chaperones control client fate determination in of ubiquitin ligases is important for the dynamics of misfolded protein aggregates the hsp70-Hsp90 chaperone system. Mol. Biol. Cell 21, 1439-1448. in yeast. J. Biol. Chem. 287, 23911-23922. Masser, A. E., Kandasamy, G., Kaimal, J. M. and Andréasson, C. (2016). Verma, R., Chen, S., Feldman, R., Schieltz, D., Yates, J., Dohmen, J. and Luciferase NanoLuc as a reporter for gene expression and protein levels in Deshaies, R. J. (2000). Proteasomal proteomics: identification of nucleotide- Saccharomyces cerevisiae. Yeast 33, 191-200. sensitive proteasome-interacting proteins by mass spectrometric analysis of McClellan, A. J., Scott, M. D. and Frydman, J. (2005). Folding and quality control affinity-purified proteasomes. Mol. Biol. Cell 11, 3425-3439. of the VHL tumor suppressor proceed through distinct chaperone pathways. Cell Willmund, F., del Alamo, M., Pechmann, S., Chen, T., Albanese,̀ V., Dammer, 121, 739-748. E. B., Peng, J. and Frydman, J. (2013). The cotranslational function of ribosome- Morimoto, D., Walinda, E., Fukada, H., Sou, Y.-S., Kageyama, S., Hoshino, M., associated Hsp70 in eukaryotic protein homeostasis. Cell 152, 196-209. Fujii, T., Tsuchiya, H., Saeki, Y., Arita, K. et al. (2015). The unexpected role of Xie, Y. and Varshavsky, A. (2001). RPN4 is a ligand, substrate, and transcriptional polyubiquitin chains in the formation of fibrillar aggregates. Nat. Commun. 6, 6116. regulator of the 26S proteasome: a negative feedback circuit. Proc. Natl. Acad. Morimoto, D., Walinda, E., Fukada, H., Sugase, K. and Shirakawa, M. (2016). Sci. USA 98, 3056-3061. Ubiquitylation directly induces fold destabilization of proteins. Sci. Rep. 6, 39453. Yam, A. Y.-W., Albanese,̀ V., Lin, H.-T. J. and Frydman, J. (2005). Hsp110 Ng, A. H. M., Fang, N. N., Comyn, S. A., Gsponer, J. and Mayor, T. (2013). cooperates with different cytosolic HSP70 systems in a pathway for de novo System-wide analysis reveals intrinsically disordered proteins are prone to folding. J. Biol. Chem. 280, 41252-41261. ubiquitylation after misfolding stress. Mol. Cell. Proteomics 12, 2456-2467. Young, Z. T., Rauch, J. N., Assimon, V. A., Jinwal, U. K., Ahn, M., Li, X., Dunyak, Paci, A., Liu, P. X. H., Zhang, L. and Zhao, R. (2016). The proteasome subunit B. M., Ahmad, A., Carlson, G. A., Srinivasan, S. R. et al. (2016). Stabilizing the Rpn8 interacts with the small nucleolar RNA Protein (snoRNP) assembly protein Hsp70-Tau complex promotes turnover in models of Tauopathy. Cell Chem. Biol. Pih1 and mediates its ubiquitin-independent degradation in Saccharomyces cerevisiae. J. Biol. Chem. 291, 11761-11775. 23, 992-1001. Park, S.-H., Bolender, N., Eisele, F., Kostova, Z., Takeuchi, J., Coffino, P. and Yu, H., Kago, G., Yellman, C. M. and Matouschek, A. (2016a). Ubiquitin-like Wolf, D. H. (2007). The cytoplasmic Hsp70 chaperone machinery subjects domains can target to the proteasome but proteolysis requires a disordered misfolded and endoplasmic reticulum import-incompetent proteins to degradation region. EMBO J. 35, 1522-1536. via the ubiquitin-proteasome system. Mol. Biol. Cell 18, 153-165. Yu, H., Singh Gautam, A. K., Wilmington, S. R., Wylie, D., Martinez-Fonts, K., Pfirrmann, T., Heessen, S., Omnus, D. J., Andreasson, C. and Ljungdahl, P. O. Kago, G., Warburton, M., Chavali, S., Inobe, T., Finkelstein, I. J. et al. (2016b). (2010). The prodomain of Ssy5 protease controls receptor-activated proteolysis of Conserved sequence preferences contribute to substrate recognition by the transcription factor Stp1. Mol. Cell. Biol. 30, 3299-3309. proteasome. J. Biol. Chem. 291, 14526-14539. Journal of Cell Science