Hsp70 chaperones as modulators of prion life cycle: novel effects of Ssa and Ssb on the Saccharomyces cerevisiae prion [PSI+]

Abstract

[PSI(+)] is a prion isoform of the yeast release factor Sup35. In some assays, the cytosolic chaperones Ssa1 and Ssb1/2 of the Hsp70 family were previously shown to exhibit "pro-[PSI(+)]" and "anti-[PSI(+)]" effects, respectively. Here, it is demonstrated for the first time that excess Ssa1 increases de novo formation of [PSI(+)] and that pro-[PSI(+)] effects of Ssa1 are shared by all other Ssa proteins. Experiments with chimeric constructs show that the peptide-binding domain is a major determinant of differences in the effects of Ssa and Ssb proteins on [PSI(+)]. Surprisingly, overproduction of either chaperone increases loss of [PSI(+)] when Sup35 is simultaneously overproduced. Excess Ssa increases both the average size of prion polymers and the proportion of monomeric Sup35 protein. Both in vivo and in vitro experiments uncover direct physical interactions between Sup35 and Hsp70 proteins. The proposed model postulates that Ssa stimulates prion formation and polymer growth by stabilizing misfolded proteins, which serve as substrates for prion conversion. In the case of very large prion aggregates, further increase in size may lead to the loss of prion activity. In contrast, Ssb either stimulates refolding into nonprion conformation or targets misfolded proteins for degradation, in this way counteracting prion formation and propagation.

Figure 1.—

Effects of excess Ssa1 on [PSI⁺] in the presence of excess Sup35 or Sup35N. (A and B) Excess Ssa1 increases de novo [PSI⁺] induction by overproduced Sup35N in the [psi⁻ PIN⁺] background. Transformants of the strain GT159 were incubated in the liquid media selective for the plasmids and containing 2% galactose and 2% raffinose (Gal + Raf) instead of glucose to induce the P_GAL promoter. Plasmids: “Control,” pRS316GAL + pLA1; “↑Sup35N,” pRS316GAL + pLA1-SUP35N; “↑Ssa1 + ↑Sup35N,” pGAL-SSA1 + pLA1-SUP35N. (A) Aliquots were taken after various periods of incubation and plated onto the glucose medium selective for both plasmids (−Ura-Leu); grown colonies were velveteen replica plated onto complete organic medium (YPD) and −Ade medium with glucose (−Ade/Glu) and scored for [PSI⁺] by color and growth, respectively. Aggregation of Sup35 in the [PSI⁺] cells results in translational readthrough of the ade1-14 (UGA) reporter, which restores growth on −Ade medium and leads to white or pink (as opposed to red) color on complete YPD medium. Two independent transformants were each tested in two independent experiments for each strain/plasmid combination. No fewer than 500 colonies were scored for each time point per culture. Averages of four repeats are shown. (B) Aliquots of 3-day cultures from A were directly spotted onto −Ade/Glu (P_GAL-repressing) medium, selective for [PSI⁺]. [PSI⁺] induction is detected as papillation after 10 days. (C–E) Excess Ssa1 or Ssb1 increases loss of [PSI⁺] in the presence of overproduced Sup35 in the [PSI⁺] background. (C) Transformants of the strong [PSI⁺] strain GT81-1C were grown in the selective liquid Gal + Raf medium to induce overproduction of Sup35. Aliquots were taken at 24-hr intervals and plated onto plasmid-selective Glu medium. Resulting colonies were velveteen replica plated onto YPD and −Ade/Glu to score for the presence of [PSI⁺]. Numbers correspond to the percentage of [psi⁻] colonies. For each strain/plasmid combination, two independent transformants were tested, each in two independent experiments. No fewer than 500 colonies were scored for each time point per culture. Averages of four repeats are shown. No [PSI⁺] loss was observed in the control cultures without overproduced Sup35 (not shown on graph). No systematic effects of excess Ssa on growth rates in the presence of excess Sup35 was detected; excess Ssb slightly increased growth under these conditions (not shown). Plasmids: “↑Sup35,” average of CEN-GAL-SUP35 + YEp13 and CEN-GAL-SUP35 + pRS424 combinations (no systematic differences were observed between these cultures); “↑Sup35 + ↑Ssa1,” CEN-GAL-SUP35 + pLH101; “↑Sup35 + ↑Ssb1,” CEN-GAL-SUP35 + pRS424-SSB1. (D) Aliquots of the same cultures as in C, taken after 3 days of incubation in Gal + Raf medium, were spotted onto YPD medium to assess [PSI⁺] loss qualitatively by appearance of red color. Plasmids: “Control,” pRS316GAL + pRS424; “↑Sup35,” CEN-GAL-SUP35 + pRS424; “↑Sup35 + ↑Ssa1,” CEN-GAL-SUP35 + pLH101; “↑Sup35 + ↑Ssb1,” CEN-GAL-SUP35 + pRS424-SSB1. (E) Transformants of the weak [PSI⁺] strain OT55 were grown on the solid synthetic galactose (Gal) medium, selective for the plasmids, and streaked out on YPD medium to detect [PSI⁺] loss by the appearance of red or sectored colonies. Plasmids: “↑Sup35,” CEN-GAL-SUP35 + pRS313; “↑Sup35 + ↑Ssa1,” CEN-GAL-SUP35 + pC211. At least six independent transformants were tested in each case with the same result. Similar results were obtained with the plasmid pRS316GAL-SUP35N, coding for Sup35N (not shown).

Figure 2.—

Effects of various members of the Ssa family on [PSI⁺]. (A) Excess Ssa1 or Ssa2, produced from P_SSA2 promoter, increases suppression of ade1-14 (UGA) in the weak [PSI⁺] strain OT55. Plasmids: “Control,” pRS313; “↑Ssa1,” pC211; “↑Ssa2,” pN2. All plasmids contain HIS3 marker. Images of −His-Ade plates were taken after 4 days. (B and C) Excess Ssa1, Ssa3, or Ssa4 produced from P_TEF1 promoter, increases suppression in OT55, detected as growth on −Ura-Ade medium, which selects for the plasmid and suppression. Plasmids: “Control,” pRS316; “↑Ssa1,” pTEF-SSA1; “↑Ssa3,” pTEF-SSA3; “↑Ssa4,” pTEF-SSA4. Images of plates were taken after 4 days. (D) Excess Ssa1 or Ssa2 protect [PSI⁺] from inhibition by excess Hsp104. Weak [PSI⁺] strain OT55 was transformed with Hsp104-coding plasmid or empty vector control and the corresponding Ssa-coding plasmid or control in all possible combinations. Growth on −Ura−His−Ade medium, which selects for both plasmids and [PSI⁺], was monitored after 7 days. Differences in suppression between the control and ↑Ssa1 (or ↑Ssa2) cultures, shown in A, are not evident in B due to a longer incubation period. Plasmids: “Control,” pRS316 + pRS313; “↑Ssa1,” pRS316 + pC211; “↑Ssa2,” pRS316 + pN2; “↑Hsp104,” pYS104 + pRS313; “↑Hsp104 + ↑Ssa1,” pYS104 + pC211; “↑Hsp104 + ↑Ssa2,” pYS104 + pN2. Similar results were obtained with the strong [PSI⁺] strains OT56 and GT81-1C (not shown). (E and F) Excess Ssa3 (E) or Ssa4 (F) protects [PSI⁺] from inhibition by excess Hsp104. Strong [PSI⁺] strain GT81-1C was transformed with P_GAL-HSP104, P_GAL-SSA3 (or P_GAL-SSA4), and the empty control plasmids in various combinations. Plates were grown on Gal medium for 3–4 days and then velveteen replica plated to the Glu medium selective for plasmids and lacking adenine. Plates were photographed after 10 days of incubation. (E) Plasmids: “Control,” pRS316GAL + pLA1; “↑Hsp104,” pRS316GAL + pH 28; “↑Ssa3,” pRS316GAL-SSA3 + pLA1; “↑Hsp104 + ↑Ssa3,” pH 28 + pRS316GAL-SSA3. (F) Plasmids: “Control,” pYCL1 + pLA1; “↑Ssa4,” pYCL1-GAL-SSA4 + pLA1; “↑Hsp104,” pYCL1 + pH 28; “↑Hsp104 + ↑Ssa4,” pH 28 + pYCL1-GAL-SSA4. Similar effect as for pRS316GAL-SSA3 was observed with the plasmid pGAL-SSA1 (not shown). Similar results were detected with the strain OT56 (not shown). (G) Excess Ssa2 increases [PSI⁺] induction by overproduced Sup35. Transformants of the [psi⁻ PIN⁺] strain OT60 were incubated on Gal medium and spotted onto −Ade/Glu medium. [PSI⁺] formation was detected as growth on −Ade after 7 days. Plasmids: “Control,” pRS316Gal + pRS313; “↑Sup35,” CEN-GAL-SUP35 + pRS313; “↑Sup35 + ↑Ssa2,” CEN-GAL-SUP35 + pN2. Similar results were observed for the Sup35N-coding plasmid pRS316GAL-SUP35N and for the strain GT159 (not shown). (H and I) Excess Ssa3 (H) and Ssa4 (I) increase [PSI⁺] induction by overproduced Sup35. Transformants of the [psi⁻ PIN⁺] strain GT159 were incubated on Gal medium and then spotted onto −Ade. [PSI⁺] formation was detected as growth on −Ade after 7 days. (H) Plasmids: “Control,” pRS316GAL + pLA1; “↑Sup35,” pRS316GAL + pLA1-SUP35; “↑Sup35 + ↑Ssa3,” pRS316GAL-SSA3 + pLA1-Sup35. (I) Plasmids: “Control,” pYCL1 + pLA1; “↑Sup35,” pYCL1 + pLA1-SUP35; “↑Sup35 + Ssa4,” pYCL1-GAL-SSA4 + pLA1-SUP35. In each experiment shown, at least six transformants were checked for each strain/plasmid combination with the same result. In all experiments in A–F, differences in growth were also confirmed by color assay on YPD medium (not shown). In experiments H and I, the same results were observed for the strain OT60 (not shown).

Figure 3.—

Overproduction of Ssa or Ssb does not affect levels of Hsp104. Ssa and Ssb overexpressor plasmids were transformed into the yeast strain GT81-1C. Plasmids: (A) pRS316 (Control), pTEF-SSA1 (P_TEF1-SSA1), pTEF-SSA3 (P_TEF1-SSA3), and pTEF-SSA4 (P_TEF1-SSA4). (B) pRS316GAL (Control) and pGAL-SSA1 (P_GAL-SSA1). (C) pYCL1 (Control), pYCL1-GAL-SSA3 (P_GAL-SSA3), and pYCL1-GAL-SSA4 (P_GAL-SSA4). For each plasmid, at least two independent transformants were used for protein isolation (data for one representative transformant are shown in each case). Yeast cultures were grown in −Ura medium to OD 0.5 (P_TEF1 constructs) or in Gal-Leu medium to OD 0.25 (P_GAL constructs). Proteins were isolated as described previously (Newnam et al. 1999) and analyzed by SDS-PAGE and Western blotting. Protein levels were determined by densitometry as described in materials and methods. Several measurements with different exposures were performed in each case to make sure that values remain within the linear range of assay sensitivity. Ade2 protein, identified by reaction to Ade2 antibody (not shown), was used as a loading control. Numbers show intensity of the Hsp104 bands in overexpressor strains relative to control, normalized by Ade2. Standardized errors did not exceed 5% of the total value.

Figure 4.—

Effects of the Ssa/Ssb chimeric constructs on [PSI⁺]-mediated suppression and compensation of the ssb1/2Δ-associated phenotypes. (A) Effects of Ssa-Ssb chimeras on [PSI⁺] protection from inhibition by excess Hsp104 in the strong [PSI⁺] strain GT81-1C. Transformants containing the P_GAL-HSP104 plasmid, pH 28, and pTEF-SSA1 (AAA), pTEF-SSB1 (BBB), or one of the chimeric constructs of the pTEF series (see Table 2 and materials and methods) were velveteen replica plated onto the −Ade/Glu medium selective for plasmids following galactose induction of P_GAL-HSP104. AAA, AAB, and BAA constructs counteract [PSI⁺] curing by excess Hsp104, as could be seen from increased growth on −Ade in comparison with the control strain. In contrast, BBB, ABA, ABB, and BBA constructs further increase curing, resulting in almost complete lack of growth. (B) Effects of Ssa-Ssb chimeras on [PSI⁺]-mediated nonsense suppression in the weak [PSI⁺] strain OT55. Transformants were replica plated onto −Ura-Ade media to monitor nonsense suppression by growth. Images were taken after 7 days. Similar differences in suppression were detected by color on YPD medium (not shown). All constructs containing the peptide-binding domain of Ssb (ABA, ABB, and BBA) exhibit an antisuppression effect, while only AAA causes increase of suppression. (C) Effects of Ssa-Ssb chimeras on nonsense suppression in the [PSI⁺]* ssb1/2Δ strain GT202. Suppressor activity of [PSI⁺]* is inhibited in the presence of Ssb (Chernoff et al. 1999). Only ABB and BBA reproduce this inhibitory effect of wild-type Ssb (BBB). (D) Compensation of ssb1/2Δ-associated drug sensitivity and GuHCl sensitivity phenotypes in the [psi⁻ PIN⁺] background. The ssb1/2Δ strain GT157 was transformed with plasmids coding for Ssa, Ssb, and chimeric constructs and tested for growth on YPD media containing 35 μg/ml hygromycin (Hyg), 5 mm GuHCl (GuHCl), and 0.2 mg/ml paromomycin (Par). Images were taken after 5 days. Hygromycin sensitivity is compensated only by complete Ssb (BBB), while sensitivities to paromomycin and GuHCl are also ameliorated by BBA and ABB. Similar results were observed for the isogenic [PSI⁺] strain GT146, except that compensation for sensitivities to paromomycin and GuHCl by ABB was less efficient (not shown). None of the plasmids used influenced growth on the −Ura medium, selective for plasmids but containing adenine (data not shown). In all experiments, SSA1 (AAA) and SSB1 (BBB) genes and chimeric constructs were expressed from the P_TEF1 promoter. In most cases, similar results were obtained when P_SSA1 (for constructs beginning with A) or P_SSB1 (for constructs beginning with B) promoters were used (not shown). Control plasmid, pRS316. See materials and methods for more details regarding chimeric constructs and their designations.

Figure 5.—

Effects of excess Ssa on size and distribution of prion polymers. (A) “Gel entry” assay: total lysates of the strong [PSI⁺] strain OT56, with either control plasmid YEp13 or the SSA1 plasmid pLH101 (↑Ssa1), were incubated in sample buffer containing 2% SDS for 5 min at 37° (nondenaturing conditions) or 100° (denaturing conditions), run on a 10% polyacrylamide separation gel with a 1.8% agarose stacking gel and transferred to a nitrocellulose membrane. High-molecular-weight polymers are not capable of entering the polyacrylamide gel under nondenaturing conditions. The arrows at the top of the polyacrylamide gel indicate where the Sup35 polymers are trapped. Presence of excess Ssa1 increases the amount of the soluble Sup35 protein that is capable of entering the gel under nondenaturing conditions. Equal amounts of total protein were loaded in each case. Note that some polymeric protein remains at the top of the gel even under denaturing conditions and that nondenatured polymeric Sup35 is apparently not as immunologically reactive as denatured Sup35. Similar results were obtained for the weak [PSI⁺] strain OT55 (not shown). (B) SDD-AGE assay: total cell lysates of the [PSI⁺] yeast strains OT55 and OT56, with or without the SSA1 plasmid, were separated by electrophoresis in a 1.8% agarose gel, transferred to PVDF membrane, and reacted to Sup35C antibody as described in materials and methods. As previously reported (Kryndushkin et al. 2003), the average size of the Sup35 polymers in the weak [PSI⁺] strain OT55 is larger than that in the strong [PSI⁺] strain OT56. Excess Ssa1 increases the average size of the Sup35 polymers in both strains and increases the proportion of the monomeric Sup35 protein vs. polymeric Sup35 protein, most evidently in OT56. Similar results were reproduced in three independent experiments. No polymeric Sup35 was observed in the extracts of the isogenic [psi⁻] strains GT17 and OT60 (not shown). Strains: [PSI⁺]^W, OT55; [PSI⁺]^S, OT56. Ssa1 overexpressor plasmid (↑Ssa1), pLH101 (YEp13 was used as a control). Yeast cultures were grown in −Leu medium, selective for the plasmids.

Figure 6.—

Direct interactions between Hsp70 and Sup35 proteins. (A) Ssa and Ssb, but not Hsp104, interact with Sup35NM in vitro. Extracts of the E. coli strain, overproducing Sup35NM-(His)₆ (“After Sup35NM-His”), and of the identical strain containing the pET-20b (+) plasmid that did not produce Sup35NM-(His)₆ (“Control”), were run through a Ni²⁺ resin under conditions where Sup35NM-(His)₆ is immobilized on this resin. Then these samples of resin were treated with extracts of the [PSI⁺] yeast strain OT56. Immobilized proteins were eluted from the resin, run on SDS-PAGE gel, and reacted to the appropriate antibodies. Total yeast lysates not exposed to the columns were run on the SDS-PAGE gel for comparison. The same observations were made with the extracts of strain OT55 (not shown). Similar data were obtained in at least three independent experiments for each protein. Results of one typical experiment are shown. (B) In vivo interaction of Sup35 with Ssa1 and Ssb1. Proteins were isolated from the strain GT159, transformed with the plasmid pmCUP1-SUP35-HA that expresses HA-tagged Sup35 from P_CUP1 promoter. Production of Sup35-HA was induced by 100 μm CuSO₄. Hsp, Sup35, and Ade2 proteins were identified by SDS-PAGE and Western blot either directly in total lysates or after immunoprecipitation with antibody against HA or Ade2 (IP). Bands reacting to Ssa and Ssb antibodies (but not to Hsp104 antibody) are visible in HA immunoprecipitates but not in Ade2 immunoprecipitates. In the cases of Ssa and Hsp104, a similar result was obtained when endogenous Sup35 was immunoprecipitated with the Sup35C antibody from the strain GT159 that did not contain P_CUP1-SUP35-HA plasmid (data not shown).

Figure 7.—

Model for the effects of Ssa and Ssb proteins on [PSI⁺] “life cycle.” See comments in the discussion.