The small heat shock protein Hsp31 cooperates with Hsp104 to modulate Sup35 prion aggregation

Abstract

The yeast homolog of DJ-1, Hsp31, is a multifunctional protein that is involved in several cellular pathways including detoxification of the toxic metabolite methylglyoxal and as a protein deglycase. Prior studies ascribed Hsp31 as a molecular chaperone that can inhibit α-Syn aggregation in vitro and alleviate its toxicity in vivo. It was also shown that Hsp31 inhibits Sup35 aggregate formation in yeast, however, it is unknown if Hsp31 can modulate [PSI⁺] phenotype and Sup35 prionogenesis. Other small heat shock proteins, Hsp26 and Hsp42 are known to be a part of a synergistic proteostasis network that inhibits Sup35 prion formation and promotes its disaggregation. Here, we establish that Hsp31 inhibits Sup35 [PSI⁺] prion formation in collaboration with a well-known disaggregase, Hsp104. Hsp31 transiently prevents prion induction but does not suppress induction upon prolonged expression of Sup35 indicating that Hsp31 can be overcome by larger aggregates. In addition, elevated levels of Hsp31 do not cure [PSI⁺] strains indicating that Hsp31 cannot intervene in a pre-existing prion oligomerization cycle. However, Hsp31 can modulate prion status in cooperation with Hsp104 because it inhibits Sup35 aggregate formation and potentiates [PSI⁺] prion curing upon overexpression of Hsp104. The absence of Hsp31 reduces [PSI⁺] prion curing by Hsp104 without influencing its ability to rescue cellular thermotolerance. Hsp31 did not synergize with Hsp42 to modulate the [PSI⁺] phenotype suggesting that both proteins act on similar stages of the prion cycle. We also showed that Hsp31 physically interacts with Hsp104 and together they prevent Sup35 prion toxicity to greater extent than if they were expressed individually. These results elucidate a mechanism for Hsp31 on prion modulation that suggest it acts at a distinct step early in the Sup35 aggregation process that is different from Hsp104. This is the first demonstration of the modulation of [PSI⁺] status by the chaperone action of Hsp31. The delineation of Hsp31's role in the chaperone cycle has implications for understanding the role of the DJ-1 superfamily in controlling misfolded proteins in neurodegenerative disease and cancer.

Keywords: Amyloids; Chaperone; DJ-1; Hsp104; Hsp31; Neurodegenerative diseases; Small Heat Shock Protein; Sup35; Yeast prion; [PSI+].

FIGURE 1.

Hsp31 overexpression decreases the level of PrD-Sup35 aggregates. (A) GAL-driven PrD-Sup35-EYFP was overexpressed for 24 h at 30°C in [psi⁻ PIN⁺] cells with or without overexpression of DsRed tagged Hsp31. PrD-Sup35-EYFP aggregates appeared as ribbon-like vacuolar peripheral rings. Hsp31 remains cytoplasmically diffuse in these cells. Elevated levels of Hsp31 decreased the presence of Sup35 aggregates in individual cells. (B) Quantitation of the number of cells with one or more Sup35-EYFP foci. The average of at least 3 independent experiments was plotted; error bars represent mean ±SEM. (*** unpaired Student's t-test; p ≤ 0.001). (C) Quantitation of the level of Sup35-EYFP fluorescence aggregates using flow cytometry. Aggregates are associated with higher fluorescence. Elevated levels of Hsp31 lowered the median fluorescence intensity (FI – arbitrary units) of Sup35-EYFP compared to vector control. Values represent mean ±SEM of 3 independent biological replicates (* unpaired Student's t-test; p ≤ 0.01). (D) Cellular lysate of cells describe in A and B was analyzed by semi-denaturing agarose electrophoresis and SDS-PAGE. Overexpression of Hsp31 suppresses the level of SDS-resistant Sup35 aggregates as detected with anti-GFP antibody. Lower panel shows the expression of Hsp31 detected by anti-DsRed antibody.

FIGURE 2.

Hsp31 transiently inhibits Sup35 prion induction. (A) To induce prion formation, a GAL-driven vector expressing PrD-Sup35-EYFP was expressed in the [psi⁻ PIN⁺] strain containing the ade1-14 nonsense mutation. Constitutive expression of Hsp31 was driven by the GPD promoter and [PSI⁺] formation was scored by quantifying white color colonies on ¼ YPD plates. Top panel represents the vector controls for the expression plasmids in the bottom panel where Hsp31-DsRed, Hsp42 and Hsp104 were overexpressed. (B) PrD-Sup35-EYFP was expressed for 6 h to transiently induce prion formation. Hsp31 overexpression decreased the prion induction level. Only one vector control is shown because other vectors had similar levels of prion induction. Error bars represent ±SEM (** unpaired Student's t-test; p ≤ 0.001, n = 3). (C) Time course of prion formation by pAG424-GAL-PrD-Sup35-EYFP with varied expression times in the presence of GPD-Hsp31. Error bars represent mean ±SEM (** unpaired Student's t-test; p ≤ 0.001, n = 3) ns = not significant. All experiments in this figure were biological replicates.

FIGURE 3.

Hsp31 is required for optimal Hsp104-induced curing of the [PSI⁺] phenotype. (A) To determine the effect of Hsp31 on [PSI⁺] prion curing, the WT and hsp31Δ [PSI⁺] strains harboring the GPD-Hsp31 expression vector (pAG415-GPD-Hsp31-DsRed) or the vector (pAG415-GPD-ccdB-DsRed) were grown for 12 h at 30°C before plating on ¼ YPD plates. (B) The WT and hsp31Δ [PSI⁺] strains with no vector were also grown and treated in the same way. Plates were grown for 2–3 d at 30°C and transferred at 4°C for increased color development. No difference in colony color was observed in these strains. (C) Low-level overexpression of Hsp104 was used to induce prion curing in [PSI⁺] hsp31Δ and WT strains. Cells were grown in liquid media for 12 h at 30°C before plating on ¼ YPD plates. Significantly less prion curing was observed in the [PSI⁺] hsp31Δ strain (**unpaired Student's t-test; p ≤ 0.001, n = 3). (D) High-level overexpression of Hsp104 was used to induce prion curing in [PSI⁺] hsp31Δ and WT strains. A 100% curing level was observed in WT strain. In the [PSI⁺] hsp31Δ strain, 100% curing was never achieved. White color colonies were plotted for the WT and [PSI⁺] hsp31Δ strain (****unpaired Student's t-test; p ≤ 0.0001, n = 3 biological replicates). (E) Hsp104 expression under the GAL promoter for 2 to 72 h in WT and [PSI⁺] hsp31Δ strain. At each indicated time point, cells were plated on ¼ YPD plates. Percentage of prion curing was calculated at each point for both WT and [PSI⁺] hsp31Δ strain. The plotted graph is one representation of 3 independent biological repeats. (unpaired Student's t-test; p ≤ 0.001 at 24, 48 and 72 h; n = 3). (F) Western blot demonstrating the relative expression levels of Hsp104. Time points for the pGAL plasmid are represent time after switching the strain to inducing galactose media. Equal amount of cells lysates were loaded in each lane and anti-histone H3 antibody was used as a loading control.

FIGURE 4.

Expression of Hsp31 in combination with Hsp104 increases prion curing. (A) Hsp31 and p426-GPD-Hsp104 were co-transformed in the [PSI⁺] strain. Vector without inserts served as controls. (B) Quantification of the experiments in panel A. Prion curing was increased from about 2.5% to 6% when Hsp31 was co-expressed with Hsp104 compared to the control strain (*unpaired Student's t-test; p ≤ 0.001, n = 3). (C) [PSI⁺] hsp31Δ strain harboring plasmids for Hsp104 and Hsp31. (D) Quantification of experiments describe in panel C. The combination of Hsp104 and Hsp31 increased prion curing in the [PSI⁺] hsp31Δ strain consistent with WT strain in A-B. (** One-way ANOVA; p ≤ 0.001, n = 3). (E) Image of p426-GPD-Hsp42 transformed cells demonstrating curing compared to vector only. (F) Hsp31 and p426-GPD-Hsp42 were co-transformed in the [PSI⁺] strain and quantified. The combination of Hsp42 and Hsp31 did not increase curing (ns = not significant). Data and images shown are representative of at least 3 independent biological experiments for all panels.

FIGURE 5.

Hsp31 interacts with Hsp104 and deletion of HSP31 does not alter Hsp104's thermotolerance response. (A) HSP31 deletion does not impair Hsp104's function in thermotolerance. Exponentially growing cells of the [PSI⁺] hsp31Δ and [PSI⁺] strains were drawn from the culture and decimal serial dilutions were plated onto YPD plates and incubated for 2 d at 30°C in each case. Both strains showed a comparable basal tolerance (top right image) and induced tolerance after pretreatment at 37°C (bottom right image) for 30 min to a 50°C heat shock treatment. Left images are non-treated cultures that serve as control. (B) Endogenous level of Hsp104 was determined in exponentially growing cultures of [PSI⁺] hsp31Δ and [PSI⁺] strains in YPD media using Hsp104 specific antibody. (C) Immunoprecipitation of Hsp31 from HSP31-9myc strain with overexpression of Hsp104 either under GPD or GAL promoter, using anti-myc antibody followed by immunoblotting with anti-Hsp104 antibody. Middle panel shows the successful pull down of Hsp31-9myc in all strains using anti-myc antibody. The lower panel Hsp104 was immunoprecipitated using anti-Hsp104 antibody followed by immunoblotting with anti-myc antibody. Equal amount of cells lysates were loaded in each lane.

FIGURE 6.

Hsp31 and Hsp104 reduce Sup35 prion toxicity. Hsp31, Hsp104 or the indicated combination of both along with GAL-PrD-Sup35-EYFP or full length Sup35 were overexpressed in [PSI⁺] and [PSI⁺] hsp31Δ strains. Decimal serial dilutions were plated onto selection plates with 2% glucose that serve as control or 2% galactose to induce the expression. Plates were incubated at 30°C for 3 d before producing the images. Hsp31 or Hsp104 rescued toxicity of GAL-PrD-Sup35-EYFP or full length Sup35 in these strains. Combination of Hsp31 and Hsp104 greatly reduce the toxicity compared to when they are individually expressed.

FIGURE 7.

Hsp31 acts early in the prionogenesis process. (A) Hsp31 overexpression together with Hsp104 decreases the aggregation of Sup35 formed by overexpression of GAL-PrD-Sup35-EYFP. Crude lysates of cells expressing Hsp31 (pAG415-GPD-HSP31-DsRed plasmid), Hsp104 (p426-GPD-Hsp104 plasmid) or both together were subjected to sedimentation analysis. Lysates were ultracentrifuged into P (pellet) and S (soluble) fractions and analyzed by immunoblotting using GFP-specific antibody. (B) Model depicting the intervention of Hsp31 during the prionogenesis process but lack of involvement in an established chaperone cycle. Hsp31 and Hsp104 physically interact but it remains to be determined if this interaction is involved in a handoff of substrates.