Defective ribosomal products challenge nuclear function by impairing nuclear condensate dynamics and immobilizing ubiquitin

Abstract Nuclear protein aggregation has been linked to genome instability and disease. The main source of aggregation‐prone proteins in cells is defective ribosomal products (DRiPs), which are generated by translating ribosomes in the cytoplasm. Here, we report that DRiPs rapidly diffuse into the nucleus and accumulate in nucleoli and PML bodies, two membraneless organelles formed by liquid–liquid phase separation. We show that nucleoli and PML bodies act as dynamic overflow compartments that recruit protein quality control factors and store DRiPs for later clearance. Whereas nucleoli serve as constitutive overflow compartments, PML bodies are stress‐inducible overflow compartments for DRiPs. If DRiPs are not properly cleared by chaperones and proteasomes due to proteostasis impairment, nucleoli undergo amyloidogenesis and PML bodies solidify. Solid PML bodies immobilize 20S proteasomes and limit the recycling of free ubiquitin. Ubiquitin depletion, in turn, compromises the formation of DNA repair compartments at fragile chromosomal sites, ultimately threatening cell survival.

. DRiPs accumulate in subnucleolar compartments, without affecting the dynamics of NCL and NPM1 (related to Fig 1).

A
DRiP labeling in HeLa cells that were treated with OP-puro (25 lM) alone or combined with the translation inhibitors CHX (50 lg/ml) and anisomycin (ANS; 5 lg/ ml). Scale bars: 10 lm. B Hela cells overexpressing GFP-TDP43 were left untreated or treated with the nuclear import inhibitor ivermectin (25 lM) for 4 h, prior to fixation and imaging. Scale bars: 20 lm. C DRiP labeling in GFP-NCL HeLa Kyoto cells that were left untreated (Control) or treated with OP-puro (25 lM) alone or combined with Ivermectin (25 lM). Scale bars: 10 lm. D RT-qPCR analysis of the expression levels of precursor and mature ribosomal RNAs in HeLa cells treated as described (MG132 10 lM; puromycin 10 lg/ll; actinomycin D/Act.D 4 lM). Statistical significance via one-way ANOVA; *P = 0.01; **P = 0.001; n = 3 independent experiments, AE s.e.m. Approximate location of the qPCR primers is shown: 45S, 18S 5 0 J (5 0 Junction); 18S; 5.8S 5 0 J (5 0 Junction); and 5.8S; 28S (from left to right). E, F Quantitation of the fluorescence intensity recovery after photobleaching (FRAP) of GFP-NPM1 (E) and GFP-NCL (F). Cells were left untreated, or treated with OPpuro alone (25 lM), or combined with MG132 (20 lM). The average recovery curve (N = 17-22) is shown in black, while the fitting curve is shown in red. Gray shows the SD.  Figure EV2. Upon stress, inhibition of translation prevents the recruitment of HSPA1A and HSPA8 inside nucleoli (related to Fig 2).
A Staining of newly synthesized RNA with 5-ethynyl uridine (EU, 200 lM) alone or with actinomycin D (Act. D; 4 lM) for 4 h, followed by click chemistry. Nucleic acid is stained with DAPI. B DRiP labeling in HeLa cells treated with OP-puro (25 lM) alone or combined with Act. D (4 lM) for 1 h. C HeLa cells were either left untreated (Control) or treated as follows: HS at 42°C for 3 h (HS), alone or combined with CHX (50 lg/ml) or Act. D (3 lM); single CHX or Act. D treatment for 3 h; transcriptional stress for 6 h (MG132 10 lM with Act. D 4 lM) (Audas et al, 2016), alone or combined with CHX (50 lg/ml). After cellular treatment, total RNA was extracted and mRNA levels of HSPA1A were measured by RT-qPCR (RPL0 was used for normalization). Statistical significance via one-way ANOVA; n = 3, AE s.e.m. D GFP-NCL HeLa Kyoto cells were left untreated or exposed to HS at 42°C for 2 h alone or with CHX; as control, cells were also treated with CHX alone for 2 h (50 lg/ ml). Cells were then fixed and stained for HSPA1A and DAPI. E Staining of HSPA1A and HSPA8 in GFP-NCL HeLa Kyoto cells treated with Act. D (3 lM) alone or combined with HS at 42°C for 2 h. ▸ Figure EV3. DRiPs that accumulate upon proteasome inhibition or temperature upshift are ubiquitinated and do not colocalize with nuclear speckles (related to Fig 3).
A Subcellular distribution of DRiPs and SC35, used as a marker for nuclear speckles, in HeLa cells that were left untreated or treated as indicated. Scale bars: 10 lm. B Immunoprecipitation (IP) of puromycylated proteins from nucleolar or nucleoplasmic extracts. HeLa cells were treated as indicated for 2 h (25 lM puromycin, 20 lM MG132) before fractionation. Western blots against puromycin and ubiquitin are shown. C HeLa cells were treated with puromycin (5 lg/ml) for 1 h, followed by recovery for 6 h in drug-free medium (À) or in the presence of MG132 (10 lM) or NH 4 Cl (20 mM). Cytoplasmic and nuclear proteins were fractionated. Clearance of puromycylated and ubiquitinated proteins were analyzed by immunoblotting in both fractions. TUBA4A and LMNB1 were used as cytoplasmic and nuclear loading controls. Quantitation of the number of PML-NBs enriched for polyUb (> 2); n = 2,052-8,452; statistical significance via one-way ANOVA; P < 10 À10 . E, F PML are stress-responsive membraneless organelles whose number changes in response to proteasome inhibition, heat shock, and accumulation of DRiPs. HeLa cells were left untreated or treated as described in the graphics (for drug concentration, see Fig 5). EV8 Figure EV5. Solid PML-NBs sequester 20S proteasomes and compromise H2A ubiquitination and 53BP1 foci formation (related to Fig 7).
A GFP-PSMA7 HeLa Kyoto cells were left untreated or treated as described. Cells were fixed and subjected to click chemistry to visualize DRiPs. Nucleic acid was stained with DAPI. Scale bars: 10 lm. B, C Hela cells were left untreated or treated for 4 h with MG132 (10 lM), alone or combined with CHX (50 lg/ml). (B) Cells were fixed and processed for immunostaining of H2A-Ub and 53BP1. Scale bars: 10 lm. (C) Expression levels of ubiquitin, H2A-Ub, and H2A in total protein extracts. TUBA4A was used as loading control. D HeLa cells were left untreated or treated for 4 h with CHX, MG132, OP-puro, combined as indicated and using the concentrations previously reported. Representative pictures of 53BP1 distribution and quantitation of the % of cells with 53BP1 foci/nucleus are shown. Number of cells counted/condition: 681-1,214 in three independent experiments; statistical significance via one-way ANOVA; P = not significant (n.s.). Scale bars: 20 lm. E Cells treated as described, using the concentrations previously reported, were allowed to recover in drug-free medium for 20 h. Cells were then fixed, and 53BP1 and DRiPs were stained. Quantitation of the % of cells with 53BP1 foci/nucleus is shown. Number of cells counted/condition: 668-1,678 in three independent experiments; statistical significance via one-way ANOVA; P = not significant (n.s.). Scale bars: 10 lm.