Vulnerability of newly synthesized proteins to proteostasis stress

Abstract

The capacity of the cell to produce, fold and degrade proteins relies on components of the proteostasis network. Multiple types of insults can impose a burden on this network, causing protein misfolding. Using thermal stress, a classic example of acute proteostatic stress, we demonstrate that ∼5-10% of the soluble cytosolic and nuclear proteome in human HEK293 cells is vulnerable to misfolding when proteostatic function is overwhelmed. Inhibiting new protein synthesis for 30 min prior to heat-shock dramatically reduced the amount of heat-stress induced polyubiquitylation, and reduced the misfolding of proteins identified as vulnerable to thermal stress. Following prior studies in C. elegans in which mutant huntingtin (Q103) expression was shown to cause the secondary misfolding of cytosolic proteins, we also demonstrate that mutant huntingtin causes similar 'secondary' misfolding in human cells. Similar to thermal stress, inhibiting new protein synthesis reduced the impact of mutant huntingtin on proteostatic function. These findings suggest that newly made proteins are vulnerable to misfolding when proteostasis is disrupted by insults such as thermal stress and mutant protein aggregation.

Keywords: Heat-shock; Neurodegenerative disease; Protein aggregation; Proteomics; Proteostasis; Ubiquitin.

Fig. 1.

Heat-shock treatment causes protein ubiquitylation and aggregation. (A) Equivalent volumes of soluble and insoluble fractions from HEK293 cells treated as noted in the figure were analyzed by immunoblotting with anti-ubiquitin antibody (1:5000). Conditions (shown above blot): 1, 37°C control; 2, cells treated with 10 nM MG-132 for 12 h; 3, cells placed in a 42°C incubator for 1 h before harvesting. The image shown is representative of three independent repetitions of the experiment. (B) Coomassie Blue staining of a 10–20% Tris-Glycine gel from experiments performed in parallel with those shown in A in which 10 µl of each fraction generated by our protocol was loaded in each lane. The origin of the various fractions is identified at the bottom of the figure. BenchMark™ molecular mass markers were used for molecular mass estimations. The image shown is the representative of three distinct experiments.

Fig. 2.

Venn diagram illustrating the overlap in protein identification by LC-MS/MS in detergent-insoluble fractions from heat-shocked cells. The LC-MS/MS data from the present study of HEK293 cells (see Table S2; proteins meeting statistical significance in one or both LC-MS/MS experiments) was compared to previously published LC-MS/MS data from SH-SY5Y and CCF-STTG1 cells (see Tables 1 and 2 in Xu et al., 2012). Six proteins were found to lose solubility in all three cell lines: ubiquitin (UBQ), CDK1, FEN1, MATR3, GTF2I and SND1. Seven additional proteins were found in both SH-SY5Y cells and HEK293 cells: KPNA2, ANP32E, PDCD11, MYBBP1A, NSUN2, TYMS and RRP12. Five proteins were found in both CCF-STTG1 and HEK293 cells: BYSL, TARDBP, STAT1, TTLL12 and RBM12B.

Fig. 3.

Isolation of HEK293FT cells expressing Htt fragments fused to CFP. (A) Workflow diagram of fluorescence activated cell sorting (FACS). (B) Immunoblot using an antibody against Htt protein (1:1000) confirmed the expression of Htt fragments in transfected cells. (C) Representative images of the transfected cells in fluorescent and bright field, demonstrating the high frequency of inclusions in the cells transfected with Htt103Q–CFP. Scale bars: 200 µm.

Fig. 4.

Expression of mutant Htt fragments in HEK293FT cells causes collateral misfolding of cytosolic and nuclear proteins. At 48 h after transfection with expression vectors for Htt25Q–CFP or Htt103Q–CFP, cells were separated by FACS and then subjected to sequential detergent extraction and sedimentation (see Materials and Methods). Immunoblots of insoluble fractions showing that cells expressing Htt103Q–CFP specifically accumulated insoluble ubiquitin (A), MATR3 (B), FEN1 (C), TDP-43 and CDK1 (D). The images shown are representative of three independent transfection experiments.

Fig. 5.

Inhibiting protein synthesis for 30 min before heat-shock prevents the accumulation of insoluble polyubiquitin, CDK1, FEN1, TDP43 and MATR3. HEK293 cells were heat-shocked with and without CHX (treated cells were exposed to CHX for 30 min before heat-shock), and then proteins insoluble in DOC were separated by the small-scale protocol for detergent extraction and sedimentation (see Materials and Methods). (A) Immunoblot analysis with antibodies to ubiquitin demonstrated that CHX treatment nearly completely blocked the accumulation of detergent-insoluble polyubiquitin. SOD1 was detected by immunoblotting of the PBS soluble fractions to demonstrate that volumetric equalization of the samples provided a reliable means of maintaining similar protein content in the various soluble fractions. The images shown are representative of more than three independent experiments. (B) Using the small-scale method, DOC insoluble fractions were separated for immunoblot analysis with antibodies to ubiquitin (1:1000), CDK1 (1:500), FEN1 (1:500), TDP-43 (1:5000) and MATR3 (1:10,000). The treatment the cells received (37°C, 42°C heat shock, or CHX treatment), and the antibody used for immunoblotting are shown on the figure. (C) Immunoblots of DOC-insoluble fractions from three separate experiments were quantified. The data for each immunoblot were normalized to the value for the protein band visible in the lysates of heat-shocked cells, which was uniformly the most intense signal. The relative intensity of the protein band from cells at 37°C and cells treated with CHX is graphed as the mean relative percentage intensity (±s.d.). Statistical analyses of the data demonstrated that pre-treatment with CHX significantly reduced the level of insoluble CDK1, FEN1, TDP43 and MATR3 (P<0.05, Student's t-test, two-tailed, unequal variance).

Fig. 6.

CHX treatment of cells transfected with mutant Htt prevents the accumulation of insoluble polyubiquitin. HEK293FT cells were transiently transfected with expression vectors for HttN171-18Q and HttN171-82Q, or left untransfected, and then incubated for 24 h at 37°C. A set of the cells transfected with HttN171-82Q were treated with CHX for 6 h, and then harvested for small-scale detergent extraction and sedimentation. (A) Immunoblots for ubiquitin demonstrated that the levels of insoluble polyubiquitin were highest in the cells transfected with Htt-N171-82Q, and that CHX treatment can decrease the level of insoluble ubiquitin. Quantification (mean±s.d.) and statistical analyses of replicate experiments (n=3) demonstrated that the levels of insoluble ubiquitin were significantly (*P<0.05, Student's t-test, two-tailed, unequal variance) lower in untransfected cells, cells expressing Htt-N171-81Q, and cells expressing Htt-N171-82Q treated with CHX. For positive controls, untransfected cells were heat-shocked for 3 h with and without CHX pre-treatment (30 min at 37°C) (lanes 1–3). M, marker proteins, from the bottom up in A, 50 kDa, 60 kDa, 80 kDa and 100 kDa marker proteins. UTf, untransfected. (B) Immunoblots with anti-Htt antibody 2B4 detected high levels of insoluble Htt-N171-82Q proteins in these cells with trace levels of insoluble Htt-N171-18Q protein. Both Htt-N171-18Q and Htt-N171-82Q were detected in the PBS-soluble fractions of the same cell lysates. DOC-insoluble fractions were also immunoblotted for ubiquitin (lanes identified on the figure). (C). Immunostaining of similar transfected cell cultures demonstrated inclusion structures in a subset of cells expressing HttN171-82Q (arrows). The images shown are representative of three independent experiments.