Yeast prion protein derivative defective in aggregate shearing and production of new 'seeds'

Abstract

According to the nucleated polymerization model, in vivo prion proliferation occurs via dissociation (shearing) of the huge prion polymers into smaller oligomeric 'seeds', initiating new rounds of prion replication. Here, we identify the deletion derivative of yeast prion protein Sup35 (Sup35-Delta22/69) that is specifically defective in aggregate shearing and 'seed' production. This derivative, [PSI+], previously thought to be unable to turn into a prion state, in fact retains the ability to form a prion ([PSI+](Delta22/69)) that can be maintained in selective conditions and transmitted by cytoplasmic infection (cytoduction), but which is mitotically unstable in non-selective conditions. MorePSI+](Delta22/69) retains its mitotic stability defect. The [PSI+](Delta22/69) cells contain more Sup35 protein in the insoluble fraction and form larger Sup35 aggregates compared with the conventional [PSI+] cells. Moderate excess of Hsp104 disaggregase increases transmission of the [PSI+](Delta22/69) prion, while excess Hsp70-Ssa chaperone antagonizes it, opposite to their effects on conventional [PSI+]. Our results shed light on the mechanisms determining the differences between transmissible prions and non-transmissible protein aggregates.

Fig. 1. The Sup35-Δ22/69 protein can both induce the prion state and be turned into a prion. (A) Transient overproduction of Sup35N or Sup35N-Δ22/69 induces the appearance of Ade⁺ colonies in the [psi^– PIN⁺] strains bearing sup35-Δ22/69 or SUP35⁺. Yeast strains: GT247-1C ([sup35-Δ22/69], left column); GT255-2D ([SUP35⁺], central column) and GT159 (SUP35⁺, right column). Plasmids: left and central columns, pRS316GAL control (none), pRS316GAL-SUP35N (GAL-SUP35N) and pGAL-SUP35NΔBst (GAL-sup35N-Δ22/69); right column, pFL39GAL control (none), pFL39GAL-SUP35N (GAL-SUP35N) and pFL39GAL-SUP35NΔBst (GAL-sup35N-Δ22/69). The –Ade/glucose plates were photographed after 10 days at 30°C. Similar results were obtained at 25°C (not shown). The strains GT247-1D and GT249-1A (not shown) produced the same result as GT247-1C. The plasmid pGAL-SUP35, bearing a complete SUP35 gene under the GAL promoter, induced Ade⁺ formation at about the same efficiency as pRS316GAL-SUP35N (not shown). (B) Induction of prion derivatives of Sup35-Δ22/69 is increased at low temperature. Yeast strain: GT249-21A ([sup35-Δ22/69]). Plasmid designations are as above (A, left and central columns). After induction on –Ura/galactose medium for 3 days at 30°C (left) or for 3 days at 30°C followed by 1 month at 4°C (right), cultures were velveteen replica plated onto –Ade/glucose medium. Plates were photographed after 10 days at 30°C. Strains GT247-1C and GT247-1D (not shown) produced the same result as GT249-21A. (C) Transmission of the newly induced prion derivatives of Sup35-Δ22/69 by cytoduction. Ade⁺ derivatives, induced in GT247-1C ([sup35-Δ22/69], line 1) or in GT255-2D ([SUP35⁺], line 3) by pFL39GAL-SUP35N, were mated to the lawn of the recipient strain c10B-H49 on YPD, and following 1 day of incubation, transferred to medium selective for cytoductants (+Ade; see Materials and methods) and the same medium lacking adenine (–Ade) to identify cytoductants able to suppress the ade2-1 mutation. The original uninduced strains (lines 2 and 4) were used as cytoplasm donors in the control experiment. Plates were photographed after 3 (left column) or 7 (right column) days of incubation. The same results were observed with the recipient strain GT388 (not shown). (D) Instability of the newly induced prion derivatives of Sup35-Δ22/69. Yeast strains are as in (A), left and central columns. Plasmid: pFL39GAL-SUP35N. After induction on –Trp/galactose medium for 3 days at 30°C, cultures were replica plated either directly onto –Ade/glucose medium (plate on the top) or first onto complete synthetic glucose-containing medium and then (after 1 day of incubation) onto –Ade/glucose medium. Plates were photographed after 10 days at 30°C. Post-induction incubation on complete medium decreases the frequency of Ade⁺ colonies in sup35-Δ22/69 but not in SUP35⁺. The same result was obtained with the inducer plasmid pGAL-SUP35 and also in the case when YPD was used instead of the synthetic complete medium (not shown). (E) Centrifugation analysis of aggregation of the Sup35 and Sup35-Δ22/69 prions. Yeast strains GT247-1C (Sup35-Δ22/69) and GT255-2D (Sup35), bearing the plasmid pRS316GAL-SUP35N, were incubated on –Ura/galactose medium for 3 days and velveteen replica plated onto –Ade/glucose medium. The Ade⁺ colonies, induced in each strain, were collected after 2 weeks of growth at 30°C, pooled together and either used directly for protein isolation (from –Ade) or incubated in liquid YPD on the shaker at 30°C and 200 r.p.m. (–Ade → YPD) for 24 h (Sup35-Δ22/69) or up to 72 h (Sup35) prior to protein isolation. Uninduced cultures of the respective strains (before induction) were used as [psi^–] controls. Proteins were isolated and fractionated into the soluble (S) and insoluble (I) fractions by centrifugation at 12 000 g as described (Newnam et al., 1999), run on the SDS–PAGE gel, transferred onto the nitrocellulose membrane and reacted with the Sup35NM specific antibodies. The larger fraction of the Sup35-reacting material was insoluble in the Ade⁺ cultures bearing Sup35-Δ22/69, compared with those bearing full-size Sup35. However, insoluble Sup35-Δ22/69 disappeared after 24 h in YPD, while full-size Sup35 aggregates remained even after 72 h in YPD. (F) Fluorescence microscopy of the Sup35 and Sup35-Δ22/69 prion aggregates. Derivatives of the yeast strains GT255-2D ([SUP35⁺]) and GT247-1C ([sup35–22/69]), bearing the plasmids pRS316-NMGFP (Sup35NM–GFP) and pRS316-NMΔBstGFP (Sup35NM-Δ22/69–GFP), respectively, were transformed with the plasmid pFL39GAL-SUP35N, incubated on –Ura/galactose medium for 3 days and velveteen replica plated onto –Ade/glucose medium. Cells from Ade⁺ colonies, grown after 2 weeks of incubation at 30°C, were analyzed by fluorescence microscopy as described in Materials and methods. Uninduced cultures of the respective strains (before induction) were used as [psi^–] controls. Prion aggregates are seen as fluorescent clumps.

Fig. 2. Effects of the chaperone proteins on the Sup35-Δ22/69 prion. (A) The continuous presence of an extra copy of HSP104 increases the frequency of the detectable Ade⁺ derivatives in the sup35-Δ22/69 strain but decreases it in the SUP35⁺ strain. The yeast strains GT247-1C ([sup35-Δ22/69]) and GT255-2D ([SUP35⁺]) were co-transformed with the TRP1-based plasmid pFL39GAL-SUP35N and either URA3-based plasmid pYS104, containing the HSP104 gene (↑Hsp104), or matching control plasmid pRS316GAL. Co-transformants were incubated on –Trp, Ura/galactose medium for 3 days and velveteen replica plated onto –Ade/glucose medium. Plates were photographed after 10 days at 30°C. (B) Transient overproduction of Hsp104 does not influence Ade⁺ induction in the sup35-Δ22/69 background. The yeast strain GT247-1C ([sup35Δ22/69]) was co-transformed with the TRP1-based plasmid pFL39GAL-SUP35N and either URA3-based plasmid pGAL104-URA3, containing the HSP104 gene under the GAL promoter (↑Hsp104), or matching control plasmid pRS316GAL. Co-transformants were incubated on –Trp, Ura/galactose medium for 3 days and velveteen replica plated onto –Ade, Ura/glucose medium. Plates were photographed after 10 days at 30°C. (C) Moderate excess of Hsp104 promotes partial dissociation of the Sup35-Δ22/69 aggregates. Ade⁺ colonies induced in GT247-1C (Sup35-Δ22/69) by pFL39GAL-SUP35N in the presence of plasmid pYS104, containing the HSP104 gene (↑Hsp104), or matching control plasmid pRS316GAL (control), were collected after 2 weeks of growth on –Ade, Ura at 30°C, pooled together and used for protein isolation and centrifugation analysis of the Sup35-Δ22/69 aggregation, as described above (see the legend to Figure 1E). Centrifugation was performed either at 12 000 or 3000 g. The presence of excess Hsp104 increases the relative amount of the Sup35-Δ22/69 protein remaining in the soluble fraction at 3000 g, which appears to correspond to the smaller aggregates. (D) The presence of an extra copy of SSA1 decreases the frequency of Ade⁺ colonies in the sup35-Δ22/69 strain. The yeast strain GT247-1C ([sup35-Δ22/69]) was co-transformed with the TRP1-based plasmid pFL39GAL-SUP35N and either URA3-based plasmid pRS316K-SSA1-K10, containing the SSA1 gene (↑Ssa1), or matching control plasmid pRS316GAL. Co-transformants were incubated on –Trp, Ura/galactose medium for 3 days and velveteen replica plated onto –Ade, Ura/glucose medium. Plates were photographed after 14 days at 30°C.