Fig 2: USP14 is phosphorylated at S143 as identified by quantitative mass spectrometry.(A) Workflow for phospho-proteomic experiments. SH-SY5Y cells overexpressing Flag-WT-USP14 were lysed and processed as described in Methods. Immunoprecipitated Flag-WT-USP14 was eluted using glycine, neutralized, enzymatically digested, and processed for LC-MS/MS analysis of phospho-peptides. Below panel, LC-MS/MS annotated mass spectrum of ASGEMASAQYITAALR peptide spanning pS143 residue identified from a representative experiment. The experiment was repeated four times with similar results.(B) Graphical illustration of human USP14 showing the N-terminal ubiquitin-like (UBL) and C-terminal ubiquitin-specific protease (USP) domains. Lower half depicts the sequence alignment and the conservation of the S143 residue across eukaryotes.(C) Protein secondary structure elements (SSE) distribution by residue index throughout the protein structure monitored throughout the MD simulation are shown for WT-USP14 and USP14 phosphorylated at S143 (pS143-USP14). α-helices and β-strands are shown in orange and blue, respectively. The black rectangular boxes indicate the areas with the largest conformational changes. MD simulations were done as described in Methods.(D) SH-SY5Y cells were transfected with Flag-tagged WT-USP14, S143A-USP14 or S143D-USP14 for 24hr, and lysates immunoprecipitated with Flag antibody. The eluates were analyzed by immunoblotting using anti-PSMD2, anti-PSMC2, and anti-Flag antibodies. Input lanes, total cell lysate. A representative experiment is shown and was repeated three times with similar results. Note: Reduced interaction of Flag S143D-USP14 with the 19S RP components PSMD2 and PSMC2.(E) USP14-deleted were reconstituted with Flag-tagged WT-USP14 or S143A-USP14 or S143D-USP14 for 24hr and lysates were analyzed by native-gel electrophoresis followed by immunoblotting for Flag antibody to detect Flag-USP14 bound 26S proteasomes. Right panel shows quantification of the densitometry ratio of Flag-S143A-USP14 or Flag-S143D-USP14 containing 26S complexes normalized to Flag-WT-USP14 containing 26S complexes. Values are means ± S.E.M. ∗∗∗∗p ≤ 0.0001, n = 3. p-value was calculated by one-way ANOVA. Note, the reduction in Flag-tagged S143D-USP14 bound to 26S proteasomes.

Fig 3: 19S RP subunit depletion does not inhibit the catalytic activity of 20S CPs.a, Aggregation was induced in dualPIM cells plus or minus 1.25 µM or 2.5 µM NMS-873 before imaging for 19 h. DualPIM aggregate average size at 19 h and b, degradation rate are shown (p = 0.049 NMS2.5 µM vs CM). Of note, higher NMS-873 concentrations enhanced lysosomal clearance but also had toxic effects from 10 h. c, PSMB2, PSMB5, PSMC1 or PSMC5 knockdown efficiency in dualPIM cells by western blot. d, DualPIM cell transfected with siCtr, siPSMB2, siPSMB5, siPSMC1, or siPSMC5 were analysed by immunoblot with anti-Ub and tubulin antibodies. Tubulin served as loading control. e, Proteasomes from dualPIM cells depleted of PSMC1, PSMC5, PSMB2 or PSMB5, or treated with 100 nM BTZ were separated by native gel electrophoresis, and decorated with anti-PSMA6 (20 CP) or PSMC2 (19 RP) antibodies. Proteasomal activity was measured using the ABP4 probe. f, U2OS cells were transfected with the indicated siRNAs and LysoTrackerRed-positive puncta number per cell was counted. g, Aggregation was induced in mCherryPIM (magenta) cells for 5 h and the percentage of only 19S RP (white arrows) and double 19S RP/20S CP-positive (purple arrows) puncta was determined. h, DualPIM cells were transfected with both siCtr or siPSMC5 and a plasmid expressing V5-DNAJB6H31Q for 24 h, before aggregation induction for 5 h and IF staining using anti-V5 antibodies. Representative images are shown and the signal intensity of V5-DNAJB6H31Q in the dualPIM aggregate ROIs was quantified. Images in g and h were acquired using the SP8X DLS confocal/light sheet microscope and in f with the IncuCyte S3 system. Data are mean values +/- SD. Error bars represent the SD of 3 (4 for g) independent experiments. p-values were calculated by a two-tailed t-test with Bonferroni correction. Scale bars, 20 µm.