Molecular chaperones of the Hsp110 family act as nucleotide exchange factors of Hsp70s
2006; Springer Nature; Volume: 25; Issue: 11 Linguagem: Inglês
10.1038/sj.emboj.7601138
ISSN1460-2075
AutoresZdravko Dragovic, Sarah A. Broadley, Yasuhito Shomura, Andreas Bracher, F. Ulrich Hartl,
Tópico(s)Plant biochemistry and biosynthesis
ResumoArticle11 May 2006free access Molecular chaperones of the Hsp110 family act as nucleotide exchange factors of Hsp70s Zdravko Dragovic Zdravko Dragovic Department of Cellular Biochemistry, Max-Planck-Institute of Biochemistry, Martinsried, Germany Search for more papers by this author Sarah A Broadley Sarah A Broadley Department of Cellular Biochemistry, Max-Planck-Institute of Biochemistry, Martinsried, Germany Search for more papers by this author Yasuhito Shomura Yasuhito Shomura Radioisotope Center School of Science, Hyogo University, Kamigori-cho, Ako-gun, Hyogo, Japan Search for more papers by this author Andreas Bracher Andreas Bracher Department of Cellular Biochemistry, Max-Planck-Institute of Biochemistry, Martinsried, Germany Search for more papers by this author F Ulrich Hartl Corresponding Author F Ulrich Hartl Department of Cellular Biochemistry, Max-Planck-Institute of Biochemistry, Martinsried, Germany Search for more papers by this author Zdravko Dragovic Zdravko Dragovic Department of Cellular Biochemistry, Max-Planck-Institute of Biochemistry, Martinsried, Germany Search for more papers by this author Sarah A Broadley Sarah A Broadley Department of Cellular Biochemistry, Max-Planck-Institute of Biochemistry, Martinsried, Germany Search for more papers by this author Yasuhito Shomura Yasuhito Shomura Radioisotope Center School of Science, Hyogo University, Kamigori-cho, Ako-gun, Hyogo, Japan Search for more papers by this author Andreas Bracher Andreas Bracher Department of Cellular Biochemistry, Max-Planck-Institute of Biochemistry, Martinsried, Germany Search for more papers by this author F Ulrich Hartl Corresponding Author F Ulrich Hartl Department of Cellular Biochemistry, Max-Planck-Institute of Biochemistry, Martinsried, Germany Search for more papers by this author Author Information Zdravko Dragovic1, Sarah A Broadley1, Yasuhito Shomura2, Andreas Bracher1 and F Ulrich Hartl 1 1Department of Cellular Biochemistry, Max-Planck-Institute of Biochemistry, Martinsried, Germany 2Radioisotope Center School of Science, Hyogo University, Kamigori-cho, Ako-gun, Hyogo, Japan *Corresponding author. Department of Cellular Biochemistry, Max-Planck-Institute of Biochemistry, Am Klopferspitz 18, 82152 Martinsried, Germany. Tel.: +49 89 8578 2244/2233; Fax: +49 89 8578 2211; E-mail: [email protected] The EMBO Journal (2006)25:2519-2528https://doi.org/10.1038/sj.emboj.7601138 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Hsp70 molecular chaperones function in protein folding in a manner dependent on regulation by co-chaperones. Hsp40s increase the low intrinsic ATPase activity of Hsp70, and nucleotide exchange factors (NEFs) remove ADP after ATP hydrolysis, enabling a new Hsp70 interaction cycle with non-native protein substrate. Here, we show that members of the Hsp70-related Hsp110 family cooperate with Hsp70 in protein folding in the eukaryotic cytosol. Mammalian Hsp110 and the yeast homologues Sse1p/2p catalyze efficient nucleotide exchange on Hsp70 and its orthologue in Saccharomyces cerevisiae, Ssa1p, respectively. Moreover, Sse1p has the same effect on Ssb1p, a ribosome-associated isoform of Hsp70 in yeast. Mutational analysis revealed that the N-terminal ATPase domain and the ultimate C-terminus of Sse1p are required for nucleotide exchange activity. The Hsp110 homologues significantly increase the rate and yield of Hsp70-mediated re-folding of thermally denatured firefly luciferase in vitro. Similarly, deletion of SSE1 causes a firefly luciferase folding defect in yeast cells under heat stress in vivo. Our data indicate that Hsp110 proteins are important components of the eukaryotic Hsp70 machinery of protein folding. Introduction Heat shock proteins of 70 kDa (Hsp70) comprise a major family of molecular chaperones involved in the cellular folding of newly synthesized proteins and the prevention of protein aggregation under stress conditions (Bukau and Horwich, 1998; Frydman, 2001; Hartl and Hayer-Hartl, 2002). Canonical Hsp70s, like the bacterial DnaK and the eukaryotic Hsc70, consist of a highly conserved ∼44 kDa N-terminal ATPase domain of regulatory function and a ∼27 kDa peptide-binding domain (PBD) with affinity to hydrophobic peptide segments exposed in non-native proteins (DeLuca-Flaherty et al, 1990; Zhu et al, 1996). In the ATP-bound state, peptide associates and dissociates rapidly from Hsp70, but remains tightly bound in the ADP-bound state. Repeated binding and release is thought to drive the folding of substrate proteins in an iterative process until the native state is achieved and hydrophobic segments are no longer exposed. For their physiological function, canonical Hsp70s require two auxiliary proteins, a J-domain protein and a nucleotide exchange factor (NEF). J-domain proteins trigger ATP hydrolysis by Hsp70 and induce tight peptide binding. NEFs (GrpE in bacteria, HspBP1/Fes1p and BAG-1 in eukaryotes) catalyze the release of ADP from Hsp70, allowing the binding of ATP, and thereby drive the folding cycle forward. A major difference between the bacterial and eukaryotic isoforms of Hsp70 lies in the kinetics of these reactions. Whereas the bacterial Hsp70 homologue DnaK is locked in the ADP-bound state and cannot function efficiently in the absence of the NEF GrpE, nucleotide exchange of eukaryotic Hsp70 becomes rate-limiting only under strong stimulation of ATP hydrolysis (Liberek et al, 1991; Brehmer et al, 2001). Eukaryotes contain a cytosolic protein class related to Hsp70 named after the mammalian Hsp110, which differs from Hsp70 by a more divergent ATPase domain and insertions of varying lengths in the C-terminal domain (Lee-Yoon et al, 1995; Easton et al, 2000). Various cytosolic Hsp110 isoforms have been described including Hsp110 (also known as Hsp105), Apg-1, and Apg-2 in mammalian cytosol; and Sse1p and Sse2p in the cytosol of yeast (Yasuda et al, 1995; Chen et al, 1996; Craven et al, 1996; Kaneko et al, 1997). The yeast genes encoding Sse1p and Sse2p are not individually essential, but deletion of both SSE1 and SSE2 results in synthetic lethality, while deletion of SSE1 alone confers a growth defect (Mukai et al, 1993; Shaner et al, 2004). Initially, Hsp110 family proteins were characterized as ‘holdases’ that keep denatured proteins in solution, but no endogenous client proteins have been described so far (Oh et al, 1997, 1999). More recent evidence suggests a cooperation between yeast Hsp110 Sse1p and the Hsp90 chaperone machinery (Lu and Cyr, 1998; Liu et al, 1999; Goeckeler et al, 2002; Zhao et al, 2005b). Another group of Hsp70-related proteins with C-terminal extensions resides in the ER, including mammalian Grp170 and yeast Lhs1p. The C-terminal segments of these Hsp170s differ from those of the Hsp110s (Chen et al, 1996; Craven et al, 1996). Association of Hsp110 with Hsp70 family members has also been reported, but only recently have studies established the interaction between mammalian Hsp105α and Hsp70; plant Hsp101 and Hsp70; Sse1p and Ssa1p; Sse1p and Ssb1p, a ribosome-associated yeast Hsp70 (Wang et al, 2000; Steel et al, 2004; Yamagishi et al, 2004; Shaner et al, 2005; Yam et al, 2005; Zhang and Guy, 2005). In the mammalian complex, only the ultimate C-terminus and a charged loop in the PBD of Hsp105α are dispensable for complex formation (Yamagishi et al, 2004). Hsp105α was additionally shown to synergistically increase the ATPase rate of Hsp70 in the presence of the J-domain protein Hsp40, but to inhibit folding of the model substrate firefly luciferase. Acceleration of ATP hydrolysis was also reported for the Sse1p–Ssa1p–Ydj1p system (Shaner et al, 2005). Perhaps the most intriguing clue as to how Hsp110 family proteins may regulate Hsp70 function came from a recent analysis of the ER-lumenal Lhs1p in yeast (Steel et al, 2004). Similar to the ER-resident NEF Sls1p/Sil1p, Lhs1p catalyzes nucleotide exchange on the ER-homologue of Hsp70, Kar2p, which results in an overall enhancement of Kar2p ATPase activity. On the other hand, Kar2p triggers the ATPase activity of Lhs1p. Here we report that human Hsp110 and the yeast homologues Sse1p and Sse2p function generally as NEFs for cytosolic Hsp70s. Mutational analysis revealed that a functional ATPase domain and the C-terminal region are important for NEF activity. Hsp110 releases peptide substrate from Hsp70 in an ATP-dependent manner. Equimolar addition of the yeast Hsp110, Sse1p, to the refolding reaction of thermally denatured firefly luciferase with yeast Hsp70 and Hsp40 considerably increased not only the yield but also the rate of refolding. Excess Sse1p, however, inhibited luciferase refolding. Consistently, deletion or overexpression of SSE1 resulted in a firefly luciferase folding defect in yeast cells subjected to heat shock. Taken together, the data indicate that Hsp110 proteins constitute an important additional element of the Hsp70 machinery of protein folding in eukaryotic cells. Results Hsp110s enhance nucleotide exchange on Hsp70s We set out to determine whether cytosolic Hsp110 class chaperones could act as NEFs for Hsp70 class chaperones. To this end, we used the fluorescent ADP analog, N8-(4-N′-methylanthraniloylaminobutyl)-8-aminoadenosine 5′-diphosphate (MABA-ADP). The fluorescence of MABA-ADP decreases upon release from Hsp70 and can be monitored using stopped flow fluorimetry (Theyssen et al, 1996; Packschies et al, 1997). MABA-ADP is an ideal spectroscopic probe for this purpose, as it has been shown to be similar to ADP with respect to Hsp70 in all kinetic and thermodynamic parameters tested to date. We first determined whether Sse1p, a yeast Hsp110 homologue, could stimulate nucleotide release from Ssa1p, a binding partner of Sse1p in the yeast cytosol (Shaner et al, 2005; Yam et al, 2005). Equimolar amounts of MABA-ADP and recombinant Ssa1p were incubated to form MABA-ADP/Ssa1p complexes, which were then mixed with an excess of ADP in the presence or absence of Sse1p (Figure 1A). Both resulting time-dependent MABA-ADP fluorescence signals exhibited a single-exponential decay, indicating a first-order reaction with a dissociation rate constant (koff) of 0.47±0.01 s−1 in the absence of Sse1p, and 5.73±0.06 s−1 in the presence of Sse1p (Supplementary Figure S1). The ∼12-fold acceleration of MABA-ADP dissociation in the presence of Sse1p indicates that this protein can indeed function as an NEF for Ssa1p and probably also for its close relatives, Ssa2p and the stress inducible isoforms, Ssa3p and Ssa4p. Next, we compared the nucleotide exchange efficiency of Sse1p with that of Fes1p, the only other known yeast cytosolic NEF (Figure 1B). In both cases, the time-dependent fluorescence signal exhibited a single exponential decay, but Fes1p increased the koff only up to 2.07±0.02 s−1, which was more than two times slower than the release obtained with Sse1p. These results suggest that Sse1p is a more efficient NEF than Fes1p. Figure 1.Hsp110 accelerates nucleotide exchange on Hsp70. (A) Accelerated dissociation of MABA-ADP from Ssa1p as monitored by stopped-flow fluorescence spectroscopy. Equimolar amounts of MABA-ADP and the indicated Hsp70 homologue were preincubated to form a complex, which was then mixed with an excess of ADP either in presence or absence of Sse1p. (B) Comparison of the efficiencies of nucleotide exchange in the presence of Sse1p and the canonical NEF, Fes1p. When applied at the same concentration (0.5 μM), Sse1p accelerates nucleotide release more efficiently than Fes1p. (C) Accelerated dissociation of MABA-ADP from Ssb1p. (D) The effect of human Hsp110 on the displacement of MABA-ADP from human Hsp70. (E) Release of MABA-ADP from Ssa1p by Sse2p. (F) Ssa1p does not trigger displacement of MABA-ADP from Sse1p. In all panels, effector-triggered and spontaneous dissociation is indicated by open and closed circles, respectively. Fes1p-triggered dissociation is indicated by closed squares. The molar ratio of effector to Hsp70 was 1:2 in all experiments. Download figure Download PowerPoint Yeast cytosol furthermore contains the highly similar, ribosome-associated proteins Ssb1p and Ssb2p, which belong to a more distantly related class of canonical Hsp70s. We therefore tested the effect of Sse1p on Ssb1p/MABA-ADP complexes and observed a ∼5-fold acceleration of MABA-ADP dissociation (Figure 1C), indicating that a conserved mechanism applies to the Sse1p interaction with both yeast Hsp70 homologues, Ssa1p and Ssb1p. Analogous experiments were also performed using the human homologues, Hsp70 and Hsp110 (HspH1) (Figure 1D). In this case, the dissociation rate constant of MABA-ADP was increased ∼5-fold in the presence of human Hsp110 (Supplementary Figure S1), suggesting that the nucleotide exchange function of Hsp110s on Hsp70s is conserved from yeast to humans. The cytosol of Saccharomyces cerevisiae contains another isoform of Sse1p, Sse2p. Sse1p and Sse2p have a sequence identity of 76%, but under normal growth conditions, the expression level of Sse2p stays far below that of its relative (Mukai et al, 1993; Ghaemmaghami et al, 2003). With this in mind, we tested whether Sse2p can compensate for its low concentration by increased nucleotide exchange activity on Ssa1p (Figure 1E), but found a similar ∼7-fold increase in the nucleotide exchange rate of MABA-ADP (Supplementary Figure S1), suggesting that the role of Sse2p may be to provide extra assistance to Sse1p during conditions of stress. Since Sse1p and Ssa1p might form a pseudo-symmetrical dimer that exerts similar effects on both of its partners, we also tested whether Ssa1p might conversely act as an NEF for Sse1p. For this purpose, we loaded Sse1p with MABA-ADP and monitored release triggered by mixing with Ssa1p and an excess of ADP (Figure 1F). Here, however, no accelerated dissociation of the nucleotide analog was observed, indicating the absence of mutuality in the Sse1p–Ssa1p complex with respect to nucleotide exchange. Similarly, Sse1p did not exert nucleotide exchange activity on Sse1p–MABA-ADP (data not shown). In summary, these data corroborate the clearly asymmetrical roles of Sse1p and Ssa1p in the heterodimer, despite their assumed overall structural similarity. Titration of NEFs To estimate the maximal nucleotide release rates caused by Sse1p versus Fes1p, we tested the effect of increasing concentrations of these NEFs on Ssa1p/MABA-ADP complexes. Already at 6 μM of Sse1p, a nucleotide release rate of 95±16 s−1 was obtained, reaching the upper limit of detection for our stopped-flow instrument (Figure 2A). After plotting the obtained koff values versus Sse1p concentration, it was clear that the titration curve did not reach saturation (Figure 2C). The best fit for the data points was a sigmoidal curve with a maximal release rate of 114±3 s−1. As saturation was not obtained, it remains unclear if the presented fit is the best description of the Sse1p–Ssa1p interaction. In contrast, saturation was reached in the titration experiment with Fes1p, and a maximal release rate of 29±3 s−1 with Kd of 5.1±0.9 μM was determined, based on a hyperbolic curve fit (Figure 2B and C). Taken together, these data confirm that Sse1p is a more efficient NEF than Fes1p. Figure 2.Dependence of nucleotide dissociation rate constant on NEF concentration. (A) Time course of displacement of MABA-ADP from Ssa1p–MABA-ADP complexes in the presence of increasing concentrations (0–6 μM) of Sse1p or (B) the canonical NEF Fes1p. For clarity, only the first second of traces has been shown. In both cases, higher concentrations of NEF lead to faster nucleotide exchange. (C) Observed dissociation rate constants were plotted versus NEF concentration. Each data point represents the mean of three measurements. The maximal nucleotide dissociation constant was estimated to be 114 s−1 for Sse1p and 29 s−1 for Fes1p. Download figure Download PowerPoint Functional analysis of Sse1p mutants Since Sse1p is a rather large multidomain protein compared to the conventional NEFs of Hsp70, HspBP1 and BAG-1, we wondered whether Sse1p could be functionally dissected into domains. In order to identify regions of Sse1p critical for nucleotide exchange on Hsp70, truncation mutants of Sse1p were generated and tested for nucleotide exchange activity on Ssa1p and Ssb1p using the stopped-flow assay (Figure 3). Consistent with previous reports, a fragment comprising the N-terminal ATP-binding domain (fragment 3, residues 1–383) could not be expressed as a soluble protein in Escherichia coli, precluding further analysis (Shaner et al, 2004). The remaining C-terminal fragment containing the PBD (fragment 4, residues 393–693) was soluble, but exhibited no NEF activity. Truncation mutants, in which a short unstructured region and two α-helical segments at the C-terminus of Sse1p were incrementally removed (secondary structure assignment based on prediction by the Jpred server (Cuff et al, 1998)), were successfully purified as soluble proteins, but also exhibited no nucleotide exchange activity (Figure 3A, fragments 5–7 and Figure 3B and C). The loss in NEF activity correlated with a reduced binding affinity of the mutant Sse1p proteins for Hsp70, as revealed in a binding assay with Ssb1 (Supplementary Figure S2). Sse1p mutants were incubated with Ssb1p, and the resulting complexes were immunoprecipitated with α-Ssb1p. Consistent with previous studies, Sse1p stably interacted with Ssb1p (Shaner et al, 2005; Yam et al, 2005). Introduction of the G233D mutation into Sse1p reduced its interaction with Ssb1p (this study and Shaner et al, 2005). Sse1p 1–674 (fragment 5) also interacted weakly with Ssb1p, while Sse1p 1–649 (fragment 6) and Sse1p 1–636 (fragment 7) did not detectably bind. Taken together, these findings suggest that both the ATP-binding domain and the less-conserved C-terminus of Sse1p are important for physical as well as functional interactions with Hsp70. Figure 3.Mutational analysis of Sse1p function. (A) Schematic representation of a set of Sse1p constructs assayed for their ability to accelerate nucleotide exchange on Ssa1p and Ssb1p. (B, C) Nucleotide exchange activity of the Sse1p constructs on Ssa1p and Ssb1p, respectively. Sse1p variants and Ssa1p or Ssb1p were mixed at a molar ratio of 5:1. In the bar graphs, the average nucleotide exchange activity of wt Sse1p was set to 1. The ATP binding mutant G233D retained only about 5% of wt activity, that is, koff is slightly above the spontaneous release rate. Download figure Download PowerPoint We also tested the influence of the Sse1p-G233D point mutant on the displacement of MABA-ADP from Ssa1p and Ssb1p (Figure 3B and C). In order to resolve rather subtle effects, a five-fold excess of Sse1p mutants over Ssa1p was used in these experiments. Surprisingly, the release rate of MABA-ADP was reduced more than 95% relative to wild-type (wt) Sse1p in the presence of Sse1p-G223D, suggesting that the nucleotide exchange activity of Sse1p requires a functional ATP binding domain. In a companion study, similar experiments were performed using another Sse1p mutant, K69M, which binds ATP (or ADP), but cannot hydrolyze ATP (Raviol et al, 2006). Surprisingly, the rate of MABA-ADP release from Ssa1p in the presence of the Sse1p K69M mutant was approximately the same as in the presence of wt Sse1p. However, the K69M mutant retains the capability to bind Sse1p (Shaner et al, 2005). In a limited proteolysis experiment probing for conformational changes of Sse1p in the absence and presence of nucleotide, no major changes in the fragmentation pattern were observed, suggesting only limited conformational rearrangements in Sse1p upon nucleotide binding (data not shown). Nucleotide-induced conformational shifts may however only become apparent in the complex with Ssa1p. On the other hand, we cannot exclude a specific effect of the G223D mutation in Sse1p. Effect of Hsp110 on the interaction of Hsp70 with peptide substrate In canonical Hsp70 proteins, ATP binding after nucleotide exchange triggers dissociation of the bound peptide substrate, allowing further cycles of protein folding. Accelerated substrate release in the presence of ATP is therefore another hallmark of NEF activity. In order to determine how Hsp110 influences substrate release from Hsp70, we utilized a fluorescent dye-labeled peptide probe, Dansyl-NRLLLTGC (D-NR). The peptide D-NR binds with high affinity and specificity to Hsp70, and its release from Hsp70 is characterized by a decrease in its fluorescence, which can be monitored using stopped-flow fluorimetry (Gragerov et al, 1994; Gässler et al, 2001). We first determined whether Hsp110 could influence the displacement of D-NR from nucleotide free Hsp70-D-NR complexes. Preformed nucleotide free Hsp70-D-NR complex was mixed with the indicated combinations of excess NR peptide, ATP, and Hsp110 (Figure 4A). Upon mixing of nucleotide free Hsp70-D-NR with NR, or with both NR and Hsp110, D-NR remained stably bound to Hsp70. Upon mixing the complex with NR and ATP, the half-time (t1/2) of release of D-NR from Hsp70 was estimated to be ∼16.9 s (Figure 4A and D). However, a substoichiometric amount of Hsp110 (Hsp110:Hsp70 ratio=1:2) together with NR and ATP accelerated D-NR release by four-fold, corresponding to a half-time of 3.9 s. These data indicate that Hsp110 enhances substrate release in an ATP-dependent manner. For comparison, similar experiments were performed in the presence of HspBP1 to analyze the influence of a conventional NEF on the displacement of peptide from nucleotide free Hsp70–D-NR complex (Figure 4B and D). Consistent with previous reports, HspBP1 also accelerated substrate release from Hsp70 in an ATP-dependent manner (Shomura et al, 2005). Another identical set of experiments using ADP–Hsp70–D-NR complex indicated that the presence of ADP did not alter the rate of D-NR release achieved by Hsp110 (Figure 4C and D). Taken together, these data suggest that Hsp110 releases substrates from Hsp70 like a conventional NEF. Figure 4.Hsp110 accelerates substrate release from Hsp70. (A) Hsp110 accelerates displacement of fluorescent dye-labeled peptide (D-NR) from nucleotide free Hsp70–D-NR complex in an ATP-dependent manner. Hsp70–D-NR was mixed with the indicated combination of unlabeled peptide (NR), Hsp110 (110), and ATP, and release of D-NR was monitored by stopped-flow fluorimetry. Components shown in parentheses indicate content of stopped-flow syringes, and ‘+’ indicates the mixing event. (B) Acceleration of D-NR release from nucleotide free Hsp70–D-NR complex triggered by the NEF HspBP1. (C) Hsp110-mediated acceleration of D-NR release from an ADP–Hsp70–D-NR complex in an ATP-dependent manner. Preformed ADP–Hsp70–D-NR complex was mixed with the indicated components and analyzed as in (A). (D) Bar graph representation of the peptide (D-NR) release half times from Hsp70–D-NR and ADP–Hsp70–D-NR complexes under the conditions tested. Download figure Download PowerPoint Hsp110 enhances the refolding activity of Hsc70 in an Hsp40-dependent manner Hsp110 class chaperones are thought to prevent aggregation by binding protein substrates during thermal denaturation and holding them in a folding-competent state (Oh et al, 1997, 1999). It is possible however, that Hsp110 has an additional role in the refolding process that may be mediated by its nucleotide exchange activity. We therefore asked whether Hsp110 influences the refolding of a heat-denatured substrate, in a manner that is independent of its holdase function during denaturation. Firefly luciferase was denatured at 42°C in the presence or absence of Hsc70, and refolding was initiated at 30°C by addition of ATP in the presence or absence of Hsp40 and/or Hsp110 (Figure 5A). When luciferase was denatured in the presence of Hsc70 to prevent aggregation, followed by refolding in the absence of other chaperones, it was reactivated to less than 10% of its prior activity level. The reactivation of luciferase was only slightly improved when either Hsp40 or Hsp110 was added to the refolding reaction, but was greatly improved when both Hsp40 and Hsp110 were added, with luciferase activity reaching more than 50% of its previous level. This is in line with an NEF function of Hsp110 because NEF activity is dispensable for eukaryotic Hsc70 without additional ATPase stimulation by Hsp40. The folding yield remained unchanged when the sequence of chaperones was reversed, that is, when Hsp110 was incubated with luciferase during unfolding (Figure 5B). Thus, Hsc70 was as effective as Hsp110 in prevention of protein aggregation, arguing against a predominant holdase function of Hsp110. Our data instead suggest that Hsp110, acting together with Hsc70 and Hsp40, promotes the refolding of denatured protein substrates. Figure 5.Hsp110 and Hsp70 mediate luciferase refolding in an Hsp40-dependent manner. (A) Luciferase refolding after preincubation with Hsc70. Firefly luciferase was heat-denatured at 42°C in the presence of Hsc70. Refolding was performed at 30°C in the presence of either Hsp110, Hsp40, or both, and ATP. In total, 100% activity corresponds to the enzyme activity of native luciferase in refolding buffer. (B) Luciferase refolding after preincubation with Hsp110 or Hsc70. Here, luciferase was denatured at 42°C in presence of Hsp110 or Hsc70 and folded at 30°C by adding ATP and different combinations of Hsc70, Hsp110, and Hsp40. Folding yields were similar irrespective of the order of Hsc70 and Hsp110 addition, indicating that the prevention of aggregation activity of Hsp110 is not a critical determinant for efficient protein refolding by the Hsp70–Hsp110 system. (C) Luciferase refolding by Ssa1p, Sse1p, and Ydj1p (yeast Hsp70/Hsp110/Hsp40). Luciferase was denatured in the presence of Ssa1p as described in (A). (D) Refolding of chemically denatured luciferase by Hsp70/Hsp110/Hsp40. (E) Hsp110 concentration dependence of luciferase refolding by the Hsp70/Hsp110/Hsp40 system. The concentrations of all other components were unchanged. Download figure Download PowerPoint We next tested whether Sse1p had the same effect as mammalian Hsp110 on Hsp70-assisted refolding. Firefly luciferase was heat-denatured in the presence of Ssa1p and refolding was initiated at 30°C in the presence or absence of Ydj1p and/or Sse1p and ATP (Figure 5C). Within 10 min of incubation, luciferase activity reached ∼40% of its prior activity when Ydj1p was added to the refolding reaction, but regained full activity when Sse1p was also present. Strikingly, addition of Sse1p accelerated the rate of refolding by more than 10-fold, an effect that was not observed with Fes1p as NEF (data not shown). The overall efficiency of refolding in the presence of Ssa1p, Sse1p and Ydj1p exceeded that observed with their mammalian counterparts. While this difference between the two systems might reflect different physiological folding demands, the ability to enhance refolding of substrates after denaturation is apparently conserved among Hsp110s. During heat denaturation, luciferase loses its activity rapidly but probably becomes only partially unfolded. To test whether the Hsp70/40/110 system can handle a completely unfolded protein resembling a nascent protein at the ribosome, we have performed similar experiments with luciferase denatured in guanidinium chloride. Hsc70 alone mediated refolding of such a substrate with a yield of up to 5% of its prior activity (Figure 5D). Addition of either Hsp40 or Hsp110 increased the yield of the refolded luciferase to 15 and 20%, respectively. However, when Hsp40 and Hsp110 were added together with Hsc70, 50% of luciferase was refolded. This agrees well with the data obtained for heat-denatured luciferase, suggesting that the Hsp70/40/110 folding system can handle both thermally and chemically denatured protein substrates. Titration of Hsp110 with the other components held at constant concentration revealed that Hsp110 is most effective for folding at an equimolar ratio to Hsc70 (Figure 5E). This could mean that Hsp110 operates as a stoichiometric component of a catalytic complex, but the details of the Hsp70/40/110 folding mechanism remain to be analyzed. We have also observed that Hsp110, when added in 10-fold excess, can have a negative effect on the refolding process, perhaps by accelerating substrate cycling to an extent where the reaction becomes unproductive for folding. Does Hsp110 also improve luciferase folding in vivo? To this end, we compared the activity of newly synthesized luciferase expressed in wt yeast, an SSE1 deletion strain (sse1Δ), and an SSE1 overexpressing strain (SSE1↑). Western blot analysis of extracts from these strains revealed that overexpression increased Sse1p levels ∼4-fold (data not shown). Equal amounts of yeast cells were transferred into media that induced luciferase synthesis and grown at 37°C. At various time points after induction, luciferase activity was measured in vivo and found to be significantly lower in the sse1Δ and SSE1↑ strains relative to wt (Figure 6A). Parallel analyses indicated that the kinetics of luciferase synthesis, as well as the solubility of luciferase, were similar among all three yeast strains (Figure 6B and data not shown), thus ruling out that Sse1p influences luciferase synthesis or solubility. Specific luciferase activity (SLA) determined 6 h postinduction revealed that deletion or overexpression of SSE1 reduced the SLA to 27 or 54% of wt levels, respectively (Figure 6A), suggesting the accumulation of substantial amounts of misfolded but solu
Referência(s)