Nucleotide Exchange Factor for the Yeast Hsp70 Molecular Chaperone Ssa1p

Strains and media.

E. coli strains used were DH5α (endA1 hsdR17 supE44 thi-1 recA1 gyrA relA1 Δ[lacZYA-argF] U169 deoR [φ80 dlacΔ lacZ M15]) and BL21 (F⁻ ompT hsdS [r_B⁻ m_B⁻] gal). Bacteria were grown in Luria-Bertani (LB) medium supplemented with ampicillin (50 μg/ml) for plasmid selection (3). Yeast strains used in this study are described in Table 1 and were grown on yeast extract-peptone-dextrose (YPD) medium or YNB-N0 lacking amino acids (Difco) but supplemented with the appropriate nutrients (1).

DNA manipulation techniques.

Standard techniques were used (3), and restriction enzymes were used according to the manufacturer's instructions. Ready-To-Go PCR beads (Pharmacia Biotech) or Taq DNA polymerase (Gibco BRL) and Crocodile III Thermocycler (Appligene Oncor) were used for PCRs.

To overexpress Fes1p for protein purification (see below), the yBR101c ORF was amplified by PCR of genomic DNA. The oligonucleotides used were MK1 (5′-CGCGGATCCGTGAAAAGCTATTACAGTGGT-3′) and MK2 (5′-CGCGGATCCTAATACATACTTTACGGC-3′), which allowed cloning into pBluescript SK− vector (Stratagene) via the underlined BamHI sites. The cloned PCR product was verified by DNA sequence analysis as described previously (33). The plasmid for glutathione S-transferase (GST)-Fes1p fusion protein expression was obtained by in-frame cloning of the yBR101c ORF into the BamHI site of pGEX-5X-1 (Amersham Pharmacia Biotech).

The GST-Fes1p coding sequence was amplified by PCR with pGEX-5X-1-yBR101c as a template. The oligonucleotides used were MK3 (5′-GCGCGCACTAGTATGTCCCCTATACTAGGTTATTGG-3′) and MK4 (5′-CCCATCGATTCATAATACATACTTTACGGCTAAATAATC-3′), which allowed cloning into pGPD425 (47) via the underlined SpeI and ClaI sites, respectively. This construct was transformed into the strains from which yBR101c had been deleted (see below) and fully complemented the thermosensitive phenotype of these mutants (see Results).

A yBR101c disruption cassette was obtained by amplifying pRS400 (8) with primers MK5 (5′-AACGAAAGAGTAAAATAGAAAAAAAAACACATACATAACTCTGTGCGGTATTTCACACCG-3′) and MK6 (5′-AAATGGTGAATGTAATATCATTTTATTTCTACGGACGTAAAGATTGTACTGAGAGTGCAC-3′), which contain a 40-bp sequence homologous to the promoter and terminator portions of yBR101c, respectively, along with a 20-bp region of homology with the KanMX4 cassette in pRS400. Disrupted mutants were then obtained by transforming the resulting PCR fragment into yeast and selecting for G418-resistant colonies at a final concentration of 200 μg/ml. Disruption of the yBR101c ORF was confirmed in resistant colonies by using primers MK7 (5′-CTCATCGTCAGTCAGAAAGC-3′) and MK8 (5′-ATAAATGTTCGAGATTCCGG-3′), which flank the yBR101c ORF. In order to exclude possible mutations arising from the transformation procedure, the phenotype (see Results) of the yBR101c disruption in each strain was analyzed in three independent clones.

Localization of a GFP-tagged Fes1p.

A GFP-tagged version of Fes1p was ob tained by using the method described by Longtine et al. (40). A HIS3MX6-GAL1-GFP cassette was obtained by PCR amplification using oligonucleotides JMB59 (5′- CCCTTGAATTCGCAATAGACCACTGTAATAGCTTTTCTTTGTATAGTTCATCCATGC-3′) and JMB68 (5′-GGTAAGCTTTTCAGCTACTGGTTGGTCAAGTTGGGCCCTATTAAGGCGAGCTCGTTTAAACTGGATGG-3′). The resulting cassette bears regions of homology to yBR101c (underlined) that allow integration into the yeast genome by double recombination. The W3031b yeast strain was transformed with this construct, and integrants expressing GFP-Fes1p under the control of the GAL1 promoter were selected on minimal medium lacking histidine and checked by PCR and Southern blot analysis. This strain was grown on raffinose-containing medium and then transferred into galactose-containing medium for 8 h, and the cells were observed with an Olympus BX 51 fluorescence microscope.

Cellular fractionation was performed using 20 A₂₅₄ U of the lysate prepared for ribosome fractionation (see below). The lysate volume was brought up to 1 ml with polysome lysis buffer (PLB; see below) and centrifuged at 16,000 × g at 4°C for 15 min. The pellet was resuspended in 500 μl of PLB, and half of the supernatant was centrifuged at 150,000 × g at 4°C for 1 h. The pellet was resuspended in 500 μl of PLB. A 10-μl aliquot of each sample was resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and analyzed by Western blotting using anti-GFP (a gift from S. Subramani, University of California, San Diego), anti-Sec61p (60), and anti-Sse1p antibodies (a gift from J. Goeckeler, University of Pittsburgh). Western blots were developed using chemiluminescence and ECL reagents (Pierce) according to the manufacturer's instructions.

Ribosome fractionation.

The yeast strains were grown in 200 to 400 ml of YPD (RSY801 and RSY801; Δfes1) or YP-galactose (W303; GFP-FES1) at 30°C to log phase (A₆₀₀ = ∼1) and, where indicated, the cells were shifted for 30 min at 37°C. Cycloheximide was added at a final concentration of 50 μg/ml, and cells were pelleted and washed with ice-cold PLB (20 mM HEPES [pH 7.4], 100 mM NaCl, 20 mM MgCl₂, 100 μg of cycloheximide/ml) containing 0.2 μl of diethylpyrocarbonate (Sigma) per ml. The cells were resuspended in 2 ml of the same buffer, and lysates were prepared by beating the cells with glass beads eight times for 15 s, with a 30-s period of cooling on ice between each vortexing step. The lysate was cleared by two centrifugation steps, each at 5,000 rpm for 5 min in a microcentrifuge at 4°C. A fraction of the supernatant corresponding to 50 to 100 A₂₅₄ U was layered over a 30-ml linear sucrose gradient (7 to 47%, wt/vol) prepared in PLB. The sucrose gradient was centrifuged at 27,000 rpm for 4.5 h at 4°C in a Beckman SW28.1 rotor. Gradients were collected with continuous monitoring at 254 nm.

To determine the association of GFP-Fes1p and Ssa1p on ribosomes and polysomes, fractions representing the load, the cytosol (top of the gradient), fractions after the cytosol but before ribosome-containing fractions, the 80S ribosomes, and polysomes were collected and proteins were resolved by SDS-PAGE and analyzed by Western blotting using anti-GFP, anti-Ssa1p (a gift from C. Koehler, University of California, Los Angeles), and anti-L3 (a gift from J. Warner, Albert Einstein College of Medicine, New York, N.Y.) antisera.

Protein purification.

The GST-Fes1p protein was purified from E. coli strain BL21 transformed with the expression plasmid described above. A 50-ml culture in LB containing 50 μg of ampicillin per ml was grown overnight at 26°C and diluted into 2 liters of the same medium. After 2 h at 28°C, IPTG (isopropyl-β-d-thiogalactopyranoside) was added to a final concentration of 0.5 mM, and cells were grown for an additional 4 h. Cells were harvested and washed once in water and once in phosphate-buffered saline (pH 7.4)-2 mM EDTA before the cell pellet was snap-frozen in liquid nitrogen and stored at −70°C. The cell pellet was thawed and resuspended in ∼20 ml of buffer A (phosphate-buffered saline [pH 7.4], 5 mM EDTA, 1 mM β-mercaptoethanol, 1 mM phenylmethylsulfonyl fluoride, 1 μg of leupeptin per ml, 1 μg of pepstatin A per ml), and the cells were disrupted by sonication for 30 s six times at the highest setting with 1 min on ice between sonications. The broken cells were centrifuged at 13,000 rpm in a Sorvall SA600 rotor for 15 min, and the resulting supernatant was further centrifuged at 20,000 rpm in a Sorvall SA600 rotor for 20 min to obtain a cleared lysate. This lysate was loaded onto a 2-ml glutathione Sepharose 4B column (Pharmacia Biotech) equilibrated in buffer A containing 1% Triton X-100. The column was washed sequentially with 30 ml of the following: (i) buffer A containing 5% glycerol; (ii) buffer A containing 1 M NaCl and 5% glycerol; (iii) 50 mM Tris-Cl (pH 7.5), 10 mM magnesium acetate, 200 mM potassium acetate, 5 mM ATP, 5% glycerol; and (iv) buffer A containing 5% glycerol. GST-Fes1p was eluted with 10 ml of elution buffer (50 mM Tris-Cl [pH 8.0], 50 mM NaCl, 20 mM reduced glutathione, 5% glycerol), and 1-ml fractions were collected and analyzed by SDS-PAGE and Coomassie brilliant blue staining. Peak fractions were pooled, dialyzed against dialysis buffer (10 mM HEPES [pH 7.0], 50 mM NaCl, 10 mM dithiothreitol [DTT], 10% glycerol), and frozen in small aliquots in liquid nitrogen and stored at −70°C. GST-Fes1p was purified to near homogeneity as assessed by SDS-PAGE and Coomassie brilliant blue staining.

GST-Sls1p and Kar2p were purified as described before (32, 45), and purified Ssa1p and Ydj1p, prepared as published (63), were a kind gift from S. W. Fewell (University of Pittsburgh). Purified DnaK and GrpE proteins were obtained from StressGen.

Protein interaction assays.

GST pull-down assays were performed essentially as described in reference 32, with minor modifications. A total of 4 μg of GST-Fes1p or GST-Sls1p was incubated with 2 μg of either Ssa1p or Kar2p in the presence of 1 mM ATP or ADP in GST-binding buffer (20 mM HEPES [pH 6.8], 100 mM KCl, 5 mM MgCl₂, 0.1% NP-40, 2% glycerol, 1 mM DTT, 1 mM EDTA, and 1 mM phenylmethylsulfonyl fluoride) for 1 h at 4°C. A 20-μl aliquot of glutathione-Sepharose 4B (50% slurry) (Amersham Pharmacia Biotech) was added, and reaction mixtures were rotated for 30 min at 4°C. Reaction mixtures were centrifuged for 2 min at 13,000 rpm in a microcentrifuge, the supernatant was removed, and the pellet was washed three times with 100 μl of GST-binding buffer. The pellet fraction was then analyzed by SDS-PAGE using 8% polyacrylamide gels followed by Coomassie brilliant blue staining.

Nucleotide exchange and ATPase assays.

An Hsp70-[α-³²P]ATP complex (with Ssa1p or DnaK) was obtained as described previously (63). In brief, 25 μg of Hsp70 was incubated in complex buffer (25 mM HEPES-KOH [pH 7.5], 100 mM KCl, 11 mM magnesium acetate) containing 25 μM ATP and 100 μCi of [α-³²P]ATP (3,000 Ci/mmol; NEN) for 30 min on ice. Complexes were purified from free nucleotide by gel filtration on G-50 Nick Columns (Amersham Pharmacia Biotech). Peak complex fractions were pooled, the glycerol concentration was adjusted to 10%, and the samples were frozen in small aliquots in liquid nitrogen. For nucleotide exchange assays, 25 μl of complex was quickly thawed and mixed with 25 μl of complex buffer with or without the indicated cofactors (at a threefold molar excess over Ssa1p) at 30°C. At the specified time points, a 15-μl aliquot was removed and purified from free nucleotide on G-50 microspin columns (Amersham Pharmacia Biotech), and the eluate was mixed with 5 μl of stop buffer (36 mM ATP, 2 M LiCl, 4 M formic acid). The amount of remaining nucleotide was measured by scintillation counting and analyzed by thin-layer chromatography as described previously (58).

Single-turnover ATPase assays were performed essentially as described elsewhere (63). In brief, 25 μl of Ssa1p-[α-³²P]ATP complex (see above) was mixed with 25 μl of complex buffer containing or lacking the indicated cofactors (at a threefold molar excess over Ssa1p) at 30°C. At the specified time points, a 6-μl aliquot of the reaction mixture was removed and added to 2 μl of stop buffer on ice. Triplicate 2-μl aliquots of this mixture were spotted on a thin-layer chromatography plate and developed (63).

For steady-state ATPase assays, 4 μg of Ssa1p was mixed with a twofold molar excess of Ydj1p in ATPase buffer (50 mM HEPES [pH 7.4], 50 mM NaCl, 10 mM DTT, 2 mM MgCl₂) containing 50 μM ATP and 0.2 μCi of [α-³²P]ATP in a total volume of 20 μl. After 10 min at 30°C, a 4-μl aliquot was stopped with 4 μl of stop buffer, and 5-μl aliquots were added to 15 μl of buffer with or without the specified amount of cofactor (GST-Fes1p or GST-Sls1p). At the indicated time points, 4-μl aliquots were taken and added to 4 μl of stop buffer on ice. Triplicate 1-μl samples were analyzed by thin-layer chromatography as described above.

Identification of yBR101c as a homologue of SLS1 (yOL031c).

The identification of Sls1p as a nucleotide exchange factor for yeast BiP (6, 7, 32) raised the possibility for the existence of a similar factor for cytosolic Hsp70s. In order to address this question, we used the Psi-BLAST 2.0 software (http://www.ncbi.nih.nlm.gov) to identify Sls1p homologues in the yeast genome. The only gene uncovered in this search was yBR101c (Fig. 1), which encodes a predicted 290-amino-acid protein. The protein encoded by this gene, which we call Fes1p (mnemonic for factor exchange for Ssa1p), is significantly shorter than Sls1p and lacks a signal sequence, ER retention motif, nuclear localization sequence, and core glycosylation consensus sites, consistent with a cytosolic localization of Fes1p. Although the Sls1p homologues that we have found in the databases so far are poorly conserved, they share several clusters of amino acids that are present in Fes1p (Fig. 1).

The Δfes1 mutant is thermosensitive.

As a first step toward investigating the function of Fes1p, the encoding ORF was disrupted by using a kanamycin resistance cassette (8) in two unique wild-type genetic backgrounds (see Materials and Methods) (Table 1). In both cases, the disruption of FES1 did not affect viability but strongly impaired growth at high temperature (Fig. 2).

We also sought genetic interactions between Δfes1 and ydj1-151, a thermosensitive mutant of the Ydj1p J DnaJ homologue that is defective in its ability to activate Ssa1p (16). Genetic interactions were previously found between Δsls1 and a thermosensitive sec63 allele (32) which encodes a mutated form of the ER J protein Sec63p (55). Disruption of FES1 in the ydj1-151 strain partially suppressed the Δfes1 and ydj1-151 thermosensitive phenotypes (Fig. 2). Since several independent transformants were analyzed, we excluded the possibility that this result is due to a third mutation resulting from the transformation procedure. The observed genetic interactions suggest that Fes1p and Ydj1p may collaborate to facilitate a cellular pathway and/or coregulate another factor.

Fes1p is a soluble protein located in the cytosol.

In order to determine the subcellular localization of Fes1p, the protein was tagged at the N terminus with GFP. GFP-Fes1p is expressed under the control of the GAL1 promoter, and the corresponding strain grew as well as the isogenic wild type at 37°C (data not shown), suggesting that the GFP tag does not affect the function of Fes1p.

We used fluorescence microscopy to assess the localization of GFP-Fes1p after inducing its expression for 8 h in galactose-containing medium. As shown in Fig. 3A, the protein was found primarily in the cytosol, as expected from the primary sequence analysis (Fig. 1). The GFP-Fes1p protein appeared enriched at the nucleus (Fig. 3A), but whether this is physiologically relevant or whether this is an effect due to the overexpression of the fusion protein is not known. Indeed, the nuclear colocalization was more pronounced when the expression of the protein was induced for longer times (data not shown). In any event, GFP-Fes1p was excluded from the vacuole (Fig. 3A).

To confirm the cytosolic localization of the protein, subcellular fractionation was performed on crude lysates as described in Materials and Methods. As shown in Fig. 3B, GFP-Fes1p was found in soluble fractions, as was Sse1p, a cytosolic Hsp110 (14) (Fig. 3B; supernatant fractions), and relatively minor amounts of GFP-Fes1p were found in the crude membrane fraction (Fig. 3B; 16,000 × g pellet), where Sec61p, the ER translocation pore protein, is detected. It should be noted that Ssa1p is also detected in both supernatant and pellet fractions (14). We conclude that Fes1p is a soluble cytosolic protein.

Fes1p interacts with Ssa1p in an ATP-dependent manner.

To determine if Fes1p, like Sls1p (32), binds to Hsp70, the protein was overexpressed as a GST fusion protein in E. coli and purified to near homogeneity by affinity chromatography (see Materials and Methods). The GST-Fes1p fusion protein fully complemented the thermosensitive phenotype of the Δfes1 strain when placed under the control of a strong constitutive promoter (see Materials and Methods), suggesting that the GST tag does not affect the function of Fes1p.

Although S. cerevisiae possesses six cytosolic Hsp70s (Ssa1 to Ssa4, Ssb1, and Ssb2; [22]), we first examined the interaction between Fes1p and the cytosolic Hsp70, Ssa1p, because of its defined roles in protein translation, translocation, folding, and ERAD (21). We also purified GST-Sls1p and Kar2p (32, 45) to provide negative and specificity controls for the binding reactions. Binding of GST-Fes1p and GST-Sls1p to the Hsp70s was assessed by GST pull-down assays in either 1 mM ATP or 1 mM ADP (see Materials and Methods). After 1 h at 4°C, the mixtures were centrifuged and washed, and the pellet fractions were analyzed by SDS-PAGE and Coomassie blue staining. As shown in Fig. 4, Fes1p binds to Ssa1p, but not to Kar2p, and conversely Sls1p binds preferentially to Kar2p, as shown previously (32), although some nonspecific binding was also observed between Sls1p and Ssa1p (Fig. 4); however, the Sls1p-Ssa1p binding is ∼10-fold lower than the interaction between Sls1p and Kar2p. Also previously described for the Sls1p-Kar2p interaction (32), Fes1p binds more efficiently to the ADP-bound form of Ssa1p, suggesting that the interaction is specific. It should be noted that the binding experiments shown in Fig. 4 were performed at pH 7.4 (with similar results at pH 8.0; data not shown) and under these conditions there is some interaction between Sls1p and the ATP-bound form of Kar2p, consistent with the ability of Sls1p to release ATP from Kar2p (32). The possibility that the Hsp70s recognize a population of misfolded GST fusion proteins is excluded by the specificity of interaction between each chaperone and either Fes1p or Sls1p and by control experiments using purified GST to which neither factor was fused (data not shown). Overall, the preferred interactions between Kar2p and Sls1p and between Fes1p and Ssa1p are consistent with the compartmentalization of these chaperone pairs in the ER lumen and cytoplasm, respectively.

Fes1p is a nucleotide exchange factor for Ssa1p.

To determine whether Fes1p promotes nucleotide release upon binding to Ssa1p, we prepared an Ssa1p-[α-³²P]ATP complex (63) and performed nucleotide exchange assays (see Materials and Methods). The Ssa1p-[α-³²P]ATP complex was incubated at 30°C in the presence of buffer, GST-Fes1p, or GST-Sls1p, aliquots were removed over time, and Ssa1p-bound and free nucleotide were resolved by gel filtration chromatography. The amount of bound nucleotide was assessed by scintillation counting (Fig. 5A) and analyzed by thin-layer chromatography (Fig. 5B). In the presence of GST-Fes1p a rapid decrease (within 1 min) of bound nucleotide was observed compared to that of the buffer control. The thin-layer chromatography plate (Fig. 5B) shows that both ADP and, to a lesser extent, ATP were stripped by GST-Fes1p. Importantly, GST-Sls1p had very little effect on nucleotide release from Ssa1p (Fig. 5), a result which was consistent with the GST pull-down assays (Fig. 4). Also, the Fes1p-stimulated nucleotide release was most likely not due to its ability to act as a peptide substrate, which was previously shown to enhance nucleotide release from Ssa1p (75), because GST-Sls1p had less effect in this assay (Fig. 5A; compare striped bars to grey bars). These results show that Fes1p is a nucleotide exchange factor for Ssa1p.

In order to validate our results and to gain more insight into the specificity of these cofactors, we performed similar assays with the E. coli DnaK protein and its nucleotide exchange factor GrpE, which lacks homology to Sls1p and Fes1p. We prepared a DnaK-[α-³²P]ATP complex and performed nucleotide exchange assays as described for Ssa1p. As shown in Fig. 6A, GrpE induced the rapid release of nucleotide, consistent with previously published data (38, 42, 43), whereas Sls1p and Fes1p had no effect on the amount of nucleotide bound to DnaK (Fig. 6A; compare striped bars to grey and dotted bars). The thin-layer chromatography plate shows that GrpE quantitatively released ATP from DnaK (Fig. 6B). This result is consistent with previous findings that GrpE can release both ATP and ADP (9, 38, 42), although our preparation of the DnaK-[α-³²P]ATP complex had insufficient specific activity that [α-³²P]ADP was not detected by thin-layer chromatography. As described previously (37), we also found that GrpE was unable to trigger nucleotide release from Ssa1p (data not shown). These and previous results (32) suggest that the Sls1p and Fes1p class of exchange factors, like GrpE, bind preferentially to the ADP-bound form of Hsp70s, but can release both ADP and ATP from their respective chaperones. Further experiments will be required to determine if one nucleotide is more efficiently and/or rapidly released than the other by Fes1p and/or Sls1p.

Effects of Fes1p on the ATPase activity of Ssa1p.

We then examined the effects of the Fes1p and Sls1p nucleotide exchange factors on Ssa1p single-turnover ATPase activity (63). The Ssa1p-[α-³²P]ATP complex was incubated at 30°C in the presence or absence of GST-Fes1p, GST-Sls1p, and Ydj1p, which enhances ATP hydrolysis (16, 18, 19, 44, 45). The extent of ATP hydrolyzed over time was assessed by thin-layer chromatography and PhosphorImager analysis (63), and the results are shown in Fig. 7A. Ssa1p alone displayed a weak ATPase activity, which was efficiently stimulated by Ydj1p. Interestingly, Fes1p significantly inhibited the ATPase activity of Ssa1p both in the presence and in the absence of Ydj1p, whereas Sls1p had only a limited effect, probably due to its poor interaction with Ssa1p (Fig. 4). As discussed in another study (39), the observed inhibition of ATPase activity in single-turnover conditions is expected for a nucleotide exchange factor, since stripping the bound nucleotide will decrease the apparent rate of hydrolysis. Therefore, this result further supports the previous data showing that Fes1p is a nucleotide exchange factor for Ssa1p.

We next assessed the effect of Fes1p on the ATPase activity of Ssa1p under steady-state conditions, as it was previously shown that both GrpE and Sls1p stimulate the ATPase activity of their respective Hsp70s synergistically with J proteins (32, 38). Ssa1p was incubated for 10 min at 30°C in the presence of [α-³²P]ATP and Ydj1p to allow ATPase activation. Then, following addition of either buffer or the indicated amounts of GST-Fes1p, ATP hydrolysis was monitored over time by thin-layer chromatography and PhosphorImager analysis (Fig. 7B). Surprisingly, Fes1p inhibited the steady-state ATPase activity of Ssa1p in a concentration-specific manner. Control experiments showed no effect of purified GST, to which no factor was fused, on the ATPase activity of Ssa1p (data not shown). It is important to state that although our nucleotide exchange experiments demonstrate that Fes1p is able to induce nucleotide dissociation from Ssa1p, they do not exclude the possibility that Fes1p also prevents nucleotide binding per se. In fact, inhibition of nucleotide binding by Fes1p could give rise to the observed effects on Ssa1p ATPase activity. For example, mammalian HspBP1 inhibits Hsp70 ATPase activity by preventing ATP binding (54). Regardless, we conclude that Fes1p has a negative effect on both steady-state and single-turnover Ssa1p ATPase activity.

Fes1p is dispensable for posttranslational protein translocation, ERAD, and luciferase refolding.

Since Ssa1p is required for posttranslational protein translocation (5, 17, 20, 44), ERAD of the cystic fibrosis transmembrane conductance regulator in yeast (72), and protein folding (14, 36, 37), we examined whether Fes1p is involved in the same pathways using well-established in vivo and in vitro assays.

Mutations in SSA1 affect the posttranslational translocation of pre-pro-α-factor (ppαF) (5, 17, 20, 44); therefore, we examined the Δfes1 strain for accumulation of ppαF. We also performed in vitro translocation assays (12) using microsomes purified from wild-type and Δfes1 strains. However, in neither case was a translocation defect apparent (data not shown). Next, cystic fibrosis transmembrane conductance regulator degradation was monitored in the Δfes1 strain by cycloheximide-chase analysis as described previously (71, 72), but no difference in the degradation rates of the Δfes1 mutant and of the wild type was observed (data not shown). We then performed in vitro ERAD assays using purified microsomes and cytosol from both wild-type and mutant strains (46, 68). In these experiments, ∼70% of the translocated ΔGpαF was degraded in a cytosol- and ATP-dependent manner, regardless of the origin of the microsomes or cytosol (data not shown). Finally, we failed to detect any defects in Ssa1p-dependent refolding activity of luciferase in vitro when using wild-type or Δfes1 mutant cytosol, or upon addition of excess purified Fes1p (reference 14 and data not shown).

From these data, we conclude that Fes1p is not a cofactor for every aspect of Ssa1p function but might be involved in another specific pathway. By analogy, function-specific cofactors for Hsp70s have been observed with DnaJ homologues (50, 67, 70, 73).

Evidence for a function of Fes1p in protein translation.

Many reports indicate the importance of molecular chaperones in protein translation. The Ssb1p and Ssb2p Hsp70s (49, 52) and Hsp40 family members (Sis1p, zuotin, Ydj1p [13, 70, 73]) associate with ribosomes and have been implicated in efficient protein synthesis or translation initiation (22). Similarly, Ssa1p was shown recently to be associated with ribosomes through its interaction with Sis1p, a cytosolic J protein, and Pab1p, the poly(A) binding protein (31). The thermosensitive ssa1-45 mutant is defective for protein translation at high temperature by affecting the interaction of Pab1p with the initiation factor eIF4G (31).

To determine if deletion of FES1 affected protein translation, we first assessed the sensitivity of wild-type and mutant strains to cycloheximide, an inhibitor of translation elongation, in a halo assay (for an example, see reference 51). As shown in Fig. 8A, the Δfes1 strain was more sensitive to cycloheximide than the wild-type control, as revealed by the formation of a larger halo of inhibition at 30°C. Interestingly, while the ydj1-151 strain showed increased cycloheximide sensitivity at 35°C, as expected from former studies (13), the ydj1-151 Δfes1 double mutant was almost indistinguishable from the wild-type strain at both temperatures, which might explain the suppressive effect of the FES1 disruption on the ydj1-151 thermosensitive growth (Fig. 2). It should be noted that the severe growth inhibition of the Δfes1 strain at 35°C probably results from both thermosensitivity and cycloheximide sensitivity. We were unable to observe significant differences in sensitivity between the wild-type and Δfes1 strains for anisomycin (data not shown), an inhibitor of peptidyltransferase activity, suggesting the specificity of Fes1p action in translation.

It was previously shown that the depletion of active Ssa1p affected protein translation, as the incubation of the ssa1-45 mutant at various times at the nonpermissive temperature of 37°C progressively reduced the ability of this strain to incorporate [³⁵S]methionine into newly synthesized proteins (31). We recapitulated this experiment by measuring the incorporation of [³⁵S]methionine-cysteine into newly translated proteins following a 2-min pulse at 0, 10, 30, and 60 min after incubation of the wild-type and Δfes1 strains at 37°C. Whereas the wild-type strain displayed little change in [³⁵S]methionine-cysteine incorporation between 10 and 30 min, the translation rate decreased in the Δfes1 strain (Fig. 8B), similarly to what was observed for the ssa1-45 mutant (31). To verify that this effect was not a result of cell death of the Δfes1 mutant at 37°C, aliquots were taken at each time point, plated on complete media, and incubated at the permissive temperature of 26°C. Approximately equal numbers of colonies were obtained with wild-type and mutant strains at each time point; moreover, growth curves at 37°C showed that the Δfes1 mutant kept dividing after 2 h of incubation at 37°C, although at this time point a reduced rate compared to that of the wild-type strain was observed. However, until this time, similar growth rates were apparent (data not shown). Therefore, we conclude that the observed decrease in translation rate is due to a defect associated with the disruption of the FES1 gene.

Defects in protein translation may lead to changes in the ratio of free ribosomes to polysomes. The ssa1-45 strain showed an increase of 80S ribosomes concomitant with a decrease in polysomes when shifted to the restrictive temperature (31). To determine whether deletion of FES1 affected polysome profile, linear sucrose gradients (7 to 47%) were used to assess ribosome distribution in the wild-type and Δfes1 strains at 30°C and after a 30-min incubation at 37°C. As shown in Fig. 8C, the Δfes1 strain showed a slight increase in the 80S ribosomes relative to polysomes compared to the wild type at 30°C. At 37°C, the increase in 80S ribosomes was more pronounced (Fig. 8C; compare the 80S peak to the first polysome peak). The polysome profile for the wild-type strain after 30 min of incubation at 37°C was undistinguishable from that obtained at 30°C (data not shown).

To examine whether Fes1p is polysome associated, lysates prepared from the GFP-FES1 strain were fractionated through the sucrose gradient and proteins corresponding to the lysate (Fig. 8D, lane 1), the nonribosomal cytosolic fraction (top of the gradient, lane 2), the position between cytosol- and ribosome-containing fractions (lane 3), the 80S ribosomes (lane 4), and polysomes (lane 5) were separated by SDS-PAGE and analyzed by Western blotting. As expected, the L3 large subunit ribosomal protein was found in the lysate and ribosomal fractions and was excluded from soluble cytosolic and preribosomal fractions. Also, as published previously, Ssa1p was found in the 80S and polysome fractions (31). Finally, we identified Fes1p in these same fractions. These data are consistent with a role of Fes1p in protein translation and with its interaction with Ssa1p.