Abnormal proteins can form aggresome in yeast: aggresome-targeting signals and components of the machinery

Abstract

In mammalian cells, abnormal proteins that escape proteasome-dependent degradation form small aggregates that can be transported into a centrosome-associated structure, called an aggresome. Here we demonstrate that in yeast a single aggregate formed by the huntingtin exon 1 with an expanded polyglutamine domain (103QP) represents a bona fide aggresome that colocalizes with the spindle pole body (the yeast centrosome) in a microtubule-dependent fashion. Since a polypeptide lacking the proline-rich region (P-region) of huntingtin (103Q) cannot form aggresomes, this domain serves as an aggresome-targeting signal. Coexpression of 103Q with 25QP, a soluble polypeptide that also carries the P-region, led to the recruitment of 103Q to the aggresome via formation of hetero-oligomers, indicating the aggresome targeting in trans. To identify additional factors involved in aggresome formation and targeting, we purified 103QP aggresomes and 103Q aggregates and identified the associated proteins using mass spectrometry. Among the aggresome-associated proteins we identified, Cdc48 (VCP/p97) and its cofactors, Ufd1 and Nlp4, were shown genetically to be essential for aggresome formation. The 14-3-3 protein, Bmh1, was also found to be critical for aggresome targeting. Its interaction with the huntingtin fragment and its role in aggresome formation required the huntingtin N-terminal N17 domain, adjacent to the polyQ domain. Accordingly, the huntingtin N17 domain, along with the P-region, plays a role in aggresome targeting. We also present direct genetic evidence for the protective role of aggresomes by demonstrating genetically that aggresome targeting of polyglutamine polypeptides relieves their toxicity.

Figure 1.

103QP forms aggresome in yeast. A) Scheme of major constructs used in this study. B) Benomyl treatment (25 μg/ml added simultaneously with induction of 103QP) disrupts formation of a single aggregate by 103QP. Top panel: fluorescent microscopy images of live yeast cells expressing 103Q and 103QP for 4 h. Bottom panel: percentage of cells with multiple and single aggregates in cells transformed with 103Q-GFP, 103QP-GFP, and 103QP-GFP on Benomyl treatment was quantified by counting multiple fields with total 200 cells. C) A component of the SPB, Spc72, colocalizes with the single 103QP aggregate. Fluorescent microscopy images of live yeast cells expressing 103QP-RFP for 6 h. The cells express endogenous levels of Spc72-GFP. D) Fluorescent microscopy images (top panel) of live cells show that 103Q expressed under the control of the MET promoter forms a single aggregate. Formation of this aggregate is disrupted by 25 μg/ml benomyl added simultaneously with induction. Bottom panel: graph quantifies percentage of cells with multiple and single aggregates in Met-103Q-GFP and Met-103Q-GFP plus Benomyl. E) 103Q was expressed under the control of MET promoter at the lower levels compared to 103Q and 103QP expressed under the control of GAL promoter. Top panel: immunoblot with anti-GFP antibody (that also recognizes CFP); slightly higher position of the band corresponding to 103Q expressed under MET control is due to the difference in the electrophoretic mobility of GFP and CFP. Bottom panel: loading was normalized by total proteins. Cdc48 is used as a control. Quantification of the Western blot was done by Quantity one software (Bio-Rad). Relative PolyQ levels are expressed as ratios of PolyQ to Cdc48 loading control.

Figure 2.

Proline-rich region targets polyQ aggregates to aggresome in trans. Fluorescent microscopy images of live cells. A) 103Q forms a complex with 25QP. Immunoblot with anti-FLAG antibody (detects both 103Q and 25QP). Cells expressing 103Q-RFP alone or together with 25QP were lysed and centrifuged for 10 min at 10,000 g. 25QP was immunoprecipitated from the resulting supernatants with anti-GFP antibody. Of note, anti-GFP antibody does not recognize RFP. G103Q, R25Q, and R25QP indicate GFP and RFP tags, respectively. B) Coexpression with 25QP, but not 25Q, promotes formation of aggresomes by 103Q. Quantification is shown in bottom panel by the percentage of cells with multiple and single aggregates in cells of 103Q coexpressed with 25Q, 103Q coexpressed with 25QP, and 103Q coexpressed with 25QP on treatment with Benomyl. C) Graph quantifies percentage of cells with multiple and single aggregates in 103Q coexpressed with 25Q, 103Q coexpressed with 25QP, and 103Q coexpressed with 25QP in the presence of Benomyl.

Figure 3.

Identification of aggresome-associated proteins. A) Isolated 103QP aggresomes and 103Q multiple aggregates separated on 2-D gel (Coomassie blue staining). Spots of Cdc48 and Bmh1 are circled. B) Representative MALDI-TOF mass spectrum shows the peptide map assignments for the identified protein Bmh1 (top panel). Sequence of the Bmh1 protein is shown on the right top corner; identified peptides are indicated in bold. Sequence assignment confirms that the peptide originates from Bmh1, not Bmh2 (bottom panel). Nanospray tandem mass spectrum of [M+2H]²⁺ m/z 746.4 corresponds to the peptide that appears in the MALDI mass spectrum with [M+H]⁺ m/z 1491.7. Complete amino acid sequence for the peptide corresponding to residues 29-42 in the Bmh1 protein is shown above the MS/MS spectrum; the defining b and y ions are marked. T, peptides that result from trypsin autolysis; Ac, acetylation.

Figure 4.

Cdc48 is critical for aggresome formation. A) Cdc48, but not Bmh1, associates specifically with 103QP aggregates. Immunoblot of purified 103QP and 103Q aggregates and control (labeled on the top) developed with anti-Cdc48, anti-Bmh1, or anti-GFP (for 103QP and 103Q) antibodies (labeled on the right side). B) Colocalization of Cdc48-GFP (expressed in the cells at the endogenous levels) with 103QP-RFP aggresomes. Fluorescent microscopy images of live cells. No colocalization is seen with 103Q aggregates, confirming the results obtained with the purified aggregates. C) Fluorescent microscopy images show the cdc48 mutation suppresses aggresome formation at the nonpermissive temperature. 103QP was induced in wild-type and cdc48-1 mutant for 6 h. Cells were shifted to 25°C 2 h after the beginning of induction. Similar inhibition of the aggresome formation was seen with the heat-sensitive cdc48-10 mutant (not shown). D) The cdc48 mutations make 103QP toxic. Wild-type and cdc48-1 cells with 103QP were spotted onto SG (minimal medium with galactose) or sd (minimal medium with glucose) plates after serial 3-fold dilutions and grown for 3 d at 30°C. E) The npl4 and ufd1 mutations suppress aggresome formation by 103QP. Fluorescent microscopy images of live cells expressing 103QP for 6 h after induction the cultures were transferred to nonpermissive temperatures. F) bmh1 deletion prevents colocalization of Cdc48-GFP with 103QP-RFP aggregates. bmh1 cells expressing 103QP for 6 h were fixed, and fluorescent microscopy images were gained.

Figure 5.

Bmh1 is essential for aggresome formation. Fluorescent microscopy images of live cells. A) bmh1 deletion leads to formation of multiple aggregates by 103QP (right panel). Fluorescent microscopy images of live cells. B) bmh1 deletion causes toxicity of 103QP. Wild-type and bmh1 cells expressing 103QP were spotted onto SG or sd plates after serial 3-fold dilutions and grown for 3 d at 30°C. C) Association of Bmh1 with soluble polyQ polypeptides in a P-region-independent manner. Lysates of cells expressing 25QP, 103Q, or 103QP were centrifuged at 10,000 g, and polyQ polypeptides were immunoprecipitated from the supernatant with anti-GFP antibody. The precipitates were then immunoblotted with anti-Bmh1 antibody (top panel) or anti-FLAG antibody for polyQ polypeptides (bottom panel). D) Deletion of the N17 region in 103QP abolished Bmh1 binding. Lysates of cells expressing 103QP or 103QP lacking N17 region were centrifuged at 10,000 g, and polyQ polypeptides were immunoprecipitated from the supernatant with anti-FLAG antibody. Lysates and the precipitates were then immunoblotted with antibodies against Bmh1 and GFP. E) Deletion of the N17 region in 103QP prevents aggresome formation. Wild-type cells expressing 103QP and ΔN17-103QP with deleted N17 region were induced for 6 h. Fluorescent microscopy images of live cells.