A network of protein interactions determines polyglutamine toxicity

Abstract

Several neurodegenerative diseases are associated with the toxicity of misfolded proteins. This toxicity must arise from a combination of the amino acid sequences of the misfolded proteins and their interactions with other factors in their environment. A particularly compelling example of how profoundly these intramolecular and intermolecular factors can modulate the toxicity of a misfolded protein is provided by the polyglutamine (polyQ) diseases. All of these disorders are caused by glutamine expansions in proteins that are broadly expressed, yet the nature of the proteins that harbor the glutamine expansions and the particular pathologies they produce are very different. We find, using a yeast model, that amino acid sequences that modulate polyQ toxicity in cis can also do so in trans. Furthermore, the prion conformation of the yeast protein Rnq1 and the level of expression of a suite of other glutamine-rich proteins profoundly affect polyQ toxicity. They can convert polyQ expansion proteins from toxic to benign and vice versa. Our work presents a paradigm for how a complex, dynamic interplay between intramolecular features of polyQ proteins and intermolecular factors in the cellular environment might determine the unique pathobiologies of polyQ expansion proteins.

Fig. 1.

Recruitment of 25Q htt exon I proteins into inclusions formed by 103Q htt exon I is independent of the amino acid sequences flanking the polyQ stretch. Shown are three examples of fluorescent microscopic images of yeast cells coexpressing the indicated htt exon I proteins with 25Q or 103Q. Areas of colocalization are shown in yellow in merged images (overlay). (Scale bar: 5 μm.)

Fig. 2.

The proline-rich region and the FLAG tag can alter polyQ toxicity in trans. (A) Cells coexpressing the indicated htt exon I proteins or a vector control were spotted on plates that either induce (+Induction) or repress (−Induction) protein expression in three 5-fold dilutions. Toxicity is reflected by reduced growth of cells under inducing conditions. (B) A dot blot was performed with protein lysates from cells expressing the indicated combinations of different htt exon I proteins. The blot was probed with an anti-FLAG antibody. (C) Cells containing FLAGhtt103QΔProCFP and either a vector control or the indicated 25Q htt exon I proteins were spotted on plates that either induce (+Induction) or repress (−Induction) protein expression in three 5-fold dilutions. (D) Cells containing FLAGhtt25QΔProCFP and either a vector control or the indicated 103Q htt exon I proteins were spotted on plates that either induce (+Induction) or repress (−Induction) protein expression in three 5-fold dilutions.

Fig. 3.

The RNQ1 prion status determines the aggregation of 103Q htt exon I proteins independent of the amino acids flanking the polyQ region. (A) Yeast cells that harbor Rnq1 in their prion state, in their nonprion state, or with a deletion of RNQ1 (Δrnq) were induced for the expression of the indicated htt exon I proteins and analyzed by fluorescence microscopy. (Scale bar: 15 μm.) (B) Yeast cells expressing Rnq-YFP (green) and the indicated htt exon I protein fused to CFP (red) were analyzed by fluorescence microscopy. Areas of colocalization are shown in yellow in overlaid images. (Scale bar: 10 μm.)

Fig. 4.

Q-rich proteins that modulate polyQ toxicity colocalize with 103Q htt proteins. Cells expressing the GFP fusion proteins Def1 (A), Yir003w (B), and Ylr274c (C) (all green) together with the indicated 103Q htt exon I protein fused to DsRed (RFP; red) were analyzed by fluorescence microscopy. Areas of colocalization are shown in yellow in overlaid images. (Scale bar: 5 μm.)

Fig. 5.

Overexpression of endogenous Q-rich proteins can induce polyQ toxicity. (A) Wild-type (W303) cells containing plasmids for the overexpression of PAN1, ENT2, CYC8, PUF2, or SNF5 fused to YFP under the control of a Cu₂SO₄-inducible promoter (CUP1) were spotted on plates that either moderately induce (100 μM additional Cu₂SO₄, +Induction) or weakly induce (no additional Cu₂SO₄, −Induction) expression. Shown are four 5-fold dilutions of cells per sample. (B) Fluorescence microscopy of cells expressing Pan1, Ent2, Cyc8, Puf2, or Snf5 fused to YFP under moderately inducing conditions (100 μM Cu₂SO₄). (C) Cells expressing the indicated htt exon I proteins together with a plasmid control or each of Pan1, Ent2, Cyc8, Puf2, or Snf5 fused to YFP were spotted on plates that either moderately induce (100 μM additional Cu₂SO₄, +Induction) or only weakly induce (no additional Cu₂SO₄, −Induction) expression. Four 5-fold dilutions for each sample were spotted. (D) Fluorescence microcopy of cells coexpressing Snf5 fused to YFP or Cyc8 fused to YFP (both red) together with the indicated htt exon I protein. Areas of colocalization are shown in yellow in overlaid images. (Scale bar: 5 μm.)