Exposed Hsp70-binding site impacts yeast Sup35 prion disaggregation and propagation

Abstract

The dynamic balance between formation and disaggregation of amyloid fibrils is associated with many neurodegenerative diseases. Multiple chaperones interact with and disaggregate amyloid fibrils, which impacts amyloid propagation and cellular phenotypes. However, it remains poorly understood whether and how site-specific binding of chaperones to amyloids facilitates the concerted disaggregation process and modulates physiological consequences in vivo. Here, we identified binding sites of Ssa1, Sis1, and Hsp104 chaperones for Sup35, the protein determinant of yeast prion [PSI⁺] yeast. Our biophysical and genetic analyses with various Sup35 deletion mutants and amyloid conformations revealed that the Ssa1-binding to the region outside amyloid core plays a key role in facilitating disaggregation and propagation of yeast prions both in vitro and in vivo. Furthermore, we developed a reconstitution system, including the Ssa1-binding tag and the HAP/Caseinolytic protease P (ClpP) hybrid chaperones, and found that this reconstitution system successfully degraded distinct prion strain conformations. Together, these results show that the properly positioned, exposed Ssa1-binding region in amyloid fibrils influences the efficiency of amyloid disaggregation and propagation, and eventually prion strain phenotypes. More broadly, our findings provide molecular foundations for previous, puzzling observations of prion propagation in vivo, and offer insights into elimination of amyloid deposits in cells.

Keywords: amyloid; chaperone; disaggregation; yeast prion.

Fig. 1.

Identification of Ssa1, Sis1, or Hsp104 binding site in Sup35NM monomer by NMR. (A) Schematic diagram of NMR analysis for Sup35NM-7R–chaperone interactions. CSD and intensity ratio (I_−chap/I_+chap) derived from HSQC NMR spectra in the absence or presence of chaperone were plotted against 253 residues of Sup35NM. (B) ¹⁵N-labeled Sup35NM-7R monomer was mixed with nonlabeled Ssa1. Chaperone concentration-dependent changes of the NMR signal profiles were shown by different colors. 100% (mol/mol) of chaperone concentration corresponds to 50 μM of Sup35NM-7R. (C) Amino acid sequences of putative Ssa1-or Sis1-binding sites in Sup35NM identified by the HSQC NMR analysis. (D–G) ¹⁵N-labeled Sup35NM-7R monomer was mixed with nonlabeled, (D) Sis1, (E) Ssa1/Sis1, (F) Hsp104 WT, or (G) Hsp104 Y257A mutant, respectively. Chaperone concentration-dependent changes of the NMR signal profiles were shown as in (B). In the NMR profiles, the interaction regions of Sup35NM-7R with each chaperone are highlighted by light green. 2 mM ATP was included in (B) and (E), and 2 mM ATPγS was included in (F) and (G). Red horizontal dot lines in CSD and intensity profiles indicate the resolution limit of the chemical shift (0.005 ppm) in our NMR conditions and the relative 100% signal intensity, respectively. All NMR spectra were acquired at 30 °C, and representative data are shown (n ≥ 2).

Fig. 2.

Characterization of Sup35NM mutants lacking a chaperone binding site. (A) List of Sup35NM WT and various deletion mutants lacking a putative chaperone binding site. The Δ129 to 148 deletion mutant was additionally constructed as a proposed binding site of overexpressed Hsp104 (31). (B and C) In vitro disaggregation assay of Sup35NM amyloid by Ssa1/Sis1/Hsp104 (3Ch). Sup35NM WT or the deletion mutant fibrils (1 μM) were treated with buffer alone (black) or Ssa1/Sis1/Hsp104 [200%/200%/120% (mol/mol)], respectively, relative to 100% Sup35NM (1 μM). (B) Disaggregation of Sup35NM WT and various deletion mutant fibrils was monitored by ThT fluorescence. Values represent means ± SEM (n = 3). (C) Representative AFM images were shown after 3 h of the disaggregation assay (n = 3). (Scale bar, 2 μm.) (D and E) In vitro fibrilization assay of Sup35NM amyloid in the presence of Ssa1/Sis1/Hsp104. Sup35NM WT or Sup35NM deletion mutant monomers (5 μM) were treated with buffer alone or Ssa1/Sis1/Hsp104 [25%/25%/25% (mol/mol), respectively, relative to 100% Sup35NM (5 μM)] (D) De novo amyloid formation of Sup35NM WT and mutants was monitored by ThT fluorescence. Values represent means ± SEM (n = 2). Note that the black, gray, and pink lines overlap the x-axis. (E) Representative AFM images are shown after 20 h of the fibrilization assay (n = 3). (Scale bar, 2 μm.)

Fig. 3.

Deletion of a chaperone-binding site in Sup35 impaired [PSI⁺] propagation in vivo. (A) List of full-length Sup35 WT and various deletion mutants for in vivo analysis of [PSI⁺] propagation by plasmid-shuffling or cytoduction assay. (B) Schematic diagram of plasmid-shuffling experiments for investigating the effect of chaperone-binding site in Sup35 on [PSI⁺] propagation by color assay. The capability of Sup35 deletion mutants to maintain [PSI⁺(Sc4)] strain phenotypes was investigated. (C) Representative images of [PSI⁺(Sc4)] yeast colonies were shown after the plasmid-shuffling experiments (n = 3). Two parental [psi⁻] and [PSI⁺(Sc4)] strains were also shown in the far-Right panels as controls. (D) Fractions of distinct yeast color phenotypes of various Sup35 deletion mutants after plasmid-shuffling experiments (n = 3). Yeast color was classified into five groups (white, white/pink, pink, pink/sectoring, red). (E) Schematic diagram of cytoduction experiments for investigating the effect of chaperone-binding site on [PSI⁺] propagation by color assay. (F) Representative images of [PSI⁺(Sc4)] yeast colonies were shown after cytoduction experiments (n = 3). Two parental [psi⁻] recipient and [PSI⁺(Sc4)] donor strains were also shown in the far-Right panels as controls. (G) Fractions of distinct yeast color phenotypes of the various Sup35 deletion mutants after cytoduction experiments (n = 3). Yeast color was classified into five groups (white, white/pink, pink, pink/sectoring, red).

Fig. 4.

S17R amyloid shows resistance to disaggregation due to its buried Ssa1-binding site. (A) Sup35NM M1, M2, M3 mutants that contain a distinct Ssa1-binding site of residues 143 to 164 (blue). (B) Schematic diagram of the seeding reaction between Sup35NM WT, M1, M2, or M3 monomers and WT Sc4 or S17R mutant amyloid seeds (5%, mol/mol). The amyloid core regions of WT Sc4 and S17R fibrils, and the Ssa1/Sis1-binding site in the seeded fibrils are shown in red and blue, respectively. Note that M3[S17R] fibrils cannot adopt the S17R-type amyloid conformation due to the partial deletion of the S17R-type amyloid core region in M3 protein. (C) Disaggregation of Sup35NM M1, M2, or M3 amyloids that formed with Sc4 amyloid seeds. 5 μM Sup35NM WT (blue), M1 (green), M2 (orange), or M3 (purple) monomers were incubated with WT Sc4 seed (5%, mol/mol) to form Sc4 fibrils. Kinetics of amyloid disaggregation was monitored by ThT fluorescence in the presence of 2 μM Ssa1, 2 μM Sis1, and 1 μM Hsp104. ThT fluorescence data were normalized with the data of amyloid alone. Values represent means ± SEM (n = 3). (D) Disaggregation of Sup35NM WT and S17R amyloid fibrils by Hsp104/Sa1/Sis1. Disaggregation of 5 μM Sup35NM Sc4 WT (blue), or S17R (red) amyloid was monitored by ThT fluorescence changes in the presence of 2 μM Ssa1, 2 μM Sis1, and 1 μM Hsp104. Disaggregation of Sup35NM M1, M2, and M3 amyloids that were formed with S17R seeds. 5 μM Sup35NM M1 (green), M2 (orange), or M3 (purple) monomer was incubated with S17R seed (5%, mol/mol) to form S17R-type fibrils with an additional Ssa1 tag. Sc4 WT amyloids that were formed by WT monomer and Sc4 WT seeds (blue), and S17R mutant amyloids that were formed by WT monomer and S17R seeds (cyan) or by S17R monomer with S17R seeds (red) are shown as controls. Values represent means ± SEM (n = 3).

Fig. 5.

S17R amyloid with a Ssa1-binding tag showed an enhanced binding to Ssa1 and Hsp104. (A) S17R mutant fibrils show a lower binding affinity to Ssa1 than WT fibrils. Representative TIRF images were shown (n = 4). Time-dependent binding of Cy3-labeled Ssa1 and nonlabeled Sis1 to the mixture of Alexa488-labeled WT (blue) and SaraFluor650-labeled S17R mutant (red) fibrils was monitored by TIRF at 0, 120, and 240 s after the mixing. (Scale bar, 5 μm.) (B) Time-dependent fluorescence intensity changes of Cy3-labeled Ssa1 were monitored for WT (blue) or S17R mutant (red) fibrils. Values represent means ± SEM (n = 4). (C) S17R mutant fibrils show a lower binding to Hsp104 than WT fibrils. Representative TIRF images were shown (n = 4). Time-dependent binding of Hsp104-SNAP549 and nonlabeled Ssa1/Sis1 to the mixture of WT (blue) and S17R mutant (red) fibrils was monitored by TIRF at 0, 50, and 150 s after the mixing. (Scale bar, 5 μm.) (D) Time-dependent binding of Hsp104-SNAP549 was monitored for WT (blue) or S17R mutant (red) fibrils. Values represent means ± SEM (n = 4). (E) The number of amyloid fragmentation event by Ssa1/Sis1/Hsp104 during 200 s was counted for WT or S17R mutant fibrils. Values represent means ± SEM (n = 5). Statistical significance was analyzed by Student’s t test. ***P < 0.001. (F) M1[S17R] fibrils show a higher binding affinity to Ssa1 than S17R[S17R] fibrils. Representative TIRF images were shown (n = 4). Time-dependent binding of Cy3-labeled Ssa1 and nonlabeled Sis1 to the mixture of Alexa488-labeled M1[S17R] (blue) and SaraFluor650-labeled S17R[S17R] (red) fibrils was monitored by TIRF at 0, 120, and 240 s after the mixing. (Scale bar, 5 μm.) (G) Time-dependent fluorescence intensity changes of Cy3-labeled Ssa1 were monitored for M1[S17R] (blue) or S17R[S17R] (red) fibrils. Values represent means ± SEM (n = 4). (H) M1[S17R] fibrils show an enhanced binding to Hsp104 than S17R[S17R] fibrils. Representative TIRF images were shown (n = 4). Time-dependent binding of Hsp104-SNAP549 and nonlabeled Ssa1/Sis1 to the mixture of M1[S17R] (blue) and S17R[S17R] (red) fibrils was monitored by TIRF at 0, 50, and 150 s after the mixing. (Scale bar, 5 μm.) (I) Time-dependent fluorescence intensity changes of Hsp104-SNAP549 were monitored for M1[S17R] (blue) or S17R[S17R] (red) fibrils. Values represent means ± SEM (n = 4). (J) The number of amyloid fragmentation event by Ssa1/Sis1/Hsp104 during 200 s was counted for M1[S17R] or S17R[S17R] fibrils. Values represent means ± SEM (n = 5). Statistical significance was analyzed by Student’s t test. ***P < 0.001.

Fig. 6.

Development of chaperone system with HAP/ClpP chaperones and a Ssa1-binding tag degrades distinct prion strain amyloids. (A) Diagram showing that HAP transfers the Sup35NM monomer extracted from fibrils immediately to ClpP protease, resulting in degradation of the extracted monomer. By contrast, the monomers from Hsp104-mediated disaggregation have the potential to form fibrils once again by reacting with remaining seeds. (B) Disaggregation and degradation of Sup35NM Sc4 fibrils were monitored by ThT fluorescence in the absence (black) or presence of 2 μM Ssa1/2 μM Sis1/1 μM Hsp104/(blue), 2 μM Ssa1/2 μM Sis1/1 μM Hsp104/1.5 μM ClpP (green), 2 μM Ssa1/2 μM Sis1/1 μM HAP (red), or 2 μM Ssa1/2 μM Sis1/1 μM HAP/1.5 μM ClpP (orange), respectively. Note that under the given condition of the amyloid disaggregation with Ssa1/Sis1/Hsp104, Ssa1/Sis1/Hsp104/ClpP, or Ssa1/Sis1/HAP, the ThT fluorescence showed a recovery phase after 2 h, while that with Ssa1/Sis1/HAP/ClpP did not. Values represent means ± SEM (n = 3). (C) Representative AFM images are shown after 4 h of the treatment of Sc4 fibrils with buffer alone (Left), Ssa1/Sis1/Hsp104 (Center), or Ssa1/Sis1/HAP/ClpP (Right) (n = 3). (Scale bar, 1 μm.) (D) Seeding activity toward Sup35NM WT monomer was monitored by ThT fluorescence. Disaggregated products of WT Sc4 amyloid after the treatment with Ssa1/Sis1/Hsp104 or Ssa1/Sis1/Hsp104/ClpP for 24 h showed a seeding activity, while those with Ssa1/Sis1/HAP/ClpP did not. Values represent means ± SEM (n ≥ 2). Statistical significance was analyzed by one-way ANOVA with Dunnett’s multiple comparison test. ***P < 0.001. (E) Degradation of Sup35NM WT fibrils with a Ssa1-binding tag was monitored by ThT fluorescence. M1[WT] (green), M2[WT] (orange), M3[WT] (purple), and control WT[WT] (blue) Sc4 fibrils were treated with 2 μM Ssa1, 2 μM Sis1, 1 μM HAP, and 1.5 μM ClpP (4Ch), and ThT fluorescence was monitored. Values represent means ± SEM (n = 3). (F) Degradation of Sup35NM S17R fibrils with a Ssa1-binding tag was monitored by ThT fluorescence. M1[S17R] (green), M2[S17R] (orange), M3[S17R] (purple), and WT[S17R] (cyan), and control WT[WT] (blue) Sc4 and S17R[S17R] (red) fibrils were treated with 2 μM Ssa1, 2 μM Sis1, 1 μM HAP, and 1.5 μM ClpP, and ThT fluorescence was monitored. Values represent means ± SEM (n = 3).

Fig. 7.

Efficiency of amyloid disaggregation and prion strain phenotypes depend on how Ssa1/Sis1-binding site is exposed in the fibrils. In Sup35NM WT Sc4 amyloid that has the core region of residues 2 to 42, the Ssa1/Sis1-binding site (residues 143 to 164) is exposed. Therefore, Ssa1/Sis1 can efficiently bind to the Sc4 amyloid and initiate the whole cascade of amyloid disaggregation. This situation results in efficient amyloid fragmentation and thereby a large number of fragmented fibers (seeds), leading to strong [PSI⁺] strain phenotypes. By contrast, the S17R amyloid harbors the amyloid core of residues 81 to 148, which partially masks the Ssa1/Sis1-binding site, preventing Ssa1/Sis1 from interacting with the S17R amyloid. This situation results in less efficient amyloid fragmentation and thereby a small number of fragmented fibers (seeds), leading to weak and sectoring [PSI⁺] strain phenotypes.