Cmr1/WDR76 defines a nuclear genotoxic stress body linking genome integrity and protein quality control

Abstract

DNA replication stress is a source of genomic instability. Here we identify changed mutation rate 1 (Cmr1) as a factor involved in the response to DNA replication stress in Saccharomyces cerevisiae and show that Cmr1--together with Mrc1/Claspin, Pph3, the chaperonin containing TCP1 (CCT) and 25 other proteins--define a novel intranuclear quality control compartment (INQ) that sequesters misfolded, ubiquitylated and sumoylated proteins in response to genotoxic stress. The diversity of proteins that localize to INQ indicates that other biological processes such as cell cycle progression, chromatin and mitotic spindle organization may also be regulated through INQ. Similar to Cmr1, its human orthologue WDR76 responds to proteasome inhibition and DNA damage by relocalizing to nuclear foci and physically associating with CCT, suggesting an evolutionarily conserved biological function. We propose that Cmr1/WDR76 plays a role in the recovery from genotoxic stress through regulation of the turnover of sumoylated and phosphorylated proteins.

Figure 1. Cmr1 associates with RPA-bound chromatin.

(a) Representation of the workflow used for the SILAC-based identification of protein complexes associated with DNA repair factors. Yeast strains expressing YFP-tagged (IG54-11D and IG46-1B) or untagged (IG45-8A) proteins were cultured in SILAC media and harvested in log phase after treatment with DNA-damaging agents. Protein complexes from SILAC lysates were affinity purified separately with GFP-Trap. Proteins were trypsin proteolysed and peptides were identified by liquid chromatography (LC)–MS/MS. (b) Identification of Rfa1-YFP and Rad52-YFP interacting proteins. The plots show log(10) SILAC ratios from GFP-tagged bait versus control from forward and reverse (SILAC label swap) experiments. Dots indicate identified proteins. Cmr1 is highlighted in red and some of the known interactors for Rfa1 and Rad52 are indicated in black in the respective plots. (c) Identification of Cmr1 interacting proteins. Cmr1-YFP (IG71-2B) was used as bait for the pull down; Cmr1 is highlighted in red; CCT-chaperonin complex subunits and Rfa1 are indicated in black. (d) Cmr1 relocalization into foci. Representative images of untreated and MMS-treated cells are shown. Arrowheads indicate selected foci. Scale bar, 2 μm. (e) Quantification of Cmr1 foci. Cmr1-YFP localization was examined by fluorescence microscopy in IG66. Cells were grown to exponential phase and imaged after treatment with zeocin (200 μg ml⁻¹), MMS (0.05%), CPT (5 μg ml⁻¹), 4-NQO (0.2 μg ml⁻¹), HU (200 mM), EMS (0.5%) for 2 h, or 1 h after ultraviolet irradiation (25 J m⁻²). Error bars represent 95% confidence intervals. Two to 3 replicates of 100–200 cells were analysed for each condition. (f) hMLH1 expression causes accumulation of Cmr1 foci. Cells expressing Cmr1-YFP (IG66) were transformed with pEH333 for ectopic expression of hMLH1 or with an empty vector (pEH334). Error bars represent 95% confidence intervals. Two to 3 replicates of 100–200 cells were analysed for each strain.

Figure 2. Cmr1 localizes into perinuclear foci.

(a) Cmr1 foci localize at the nuclear periphery. Cells expressing Cmr1-YFP (IG66) were transformed with pNEB21 for expression of Nup49-CFP and pML84 for expression of NLS-yEmRFP, and treated with 0.05% MMS. Cmr1 localization relative to the nucleoplasm and the nuclear membrane is shown in both unbudded (G1) and budded (S/G2) cells. Nuclear zooms are at 150%. (b) Cmr1 foci in α-factor-arrested cells. Cells (IG66) were cultured with α-factor for 1 h, followed by treatment with 0.05% MMS for 2 h. (c) Quantification of Cmr1 foci. The percentage of cells (IG66) with Cmr1-YFP foci was quantified in asynchronous and G1-arrested populations, before and after treatment with 0.05% MMS for 1 h, and 30, 60 and 90 min after the drug was removed. Two to 3 replicates of 100–200 cells were analysed for each condition. Error bars represent 95% confidence intervals. (d) Schematic representation of Cmr1 and truncation constructs. (e) The WD40 domain of Cmr1 is necessary and sufficient for relocalization into foci. Cmr1-YFP constructs containing the full-length protein (pIG13), the N-terminal (pIG14) or WD40-containing portion of the protein (pIG15) were expressed from a 2-μm plasmid in cells (ML657) harbouring NLS-yEmRFP, and relocalization of Cmr1 into nuclear foci was monitored after treatment with 0.05% MMS. Scale bars, 2 μm.

Figure 3. Characterization of INQ.

(a) Genome-wide analysis of Cmr1 co-localizing proteins. Haploid cells expressing GFP-tagged query proteins and Rad52-RFP as a nuclear marker (IG72-5C) were imaged by high-content fluorescence microscopy, untreated or treated for 2 h with 75 μg ml⁻¹ MG132. Proteins re-localizing into perinuclear foci were further tested for co-localization with Cmr1. Confirmed co-localizing proteins are listed. Proteins that also co-localize with Cmr1 after MMS treatment are highlighted in bold. (b) Cmr1 foci are induced by genetic impairment of proteasome function. Gene deletion strains expressing Cmr1-YFP and NLS-yEmRFP were imaged by high-content fluorescence microscopy. Strains exhibiting more than threefold increase in the percentage of spontaneous Cmr1 foci compared with wild type were manually retested. Only the mutants giving a result significantly different from the wild type are reported in the figure. Six mutants (arp6Δ, slx1Δ, fpr1Δ, whi2Δ, sgs1Δ and csm2Δ) exhibited elevated Cmr1-YFP foci levels in the screen, but not on manual retesting. Red and blue dashed lines represent the threefold thresholds for the automated and manual analyses, respectively. IG66 served as the wild-type reference strain for the manual re-testing. (c) Cmr1 foci are dependent on Hsp42 and Btn2. Cells deleted for BTN2 (IG239-2B) or HSP42 (IG238-9D) and expressing Cmr1-YFP were treated with MMS or MG132 for 2 h. Two to 3 replicates of 100–200 cells were analysed for each condition. Error bars represent 95% confidence intervals. (d) Cmr1 defines INQ. Cells expressing Cmr1-YFP (IG66), Cherry-VHL (pESC-mCherry-VHL) and Nup49-CFP (pNEB21) were grown at 25 °C to log phase in synthetic complete medium lacking tryptophan and uracil, and with 2% raffinose as a carbon source. Cherry-VHL expression was induced by addition of 3% galactose for 3 h, followed by a shift to 37 °C and treatment with 75 μg ml⁻¹ MG132 in 2% glucose for 1 h before imaging. Arrowheads mark VHL and Cmr1 foci. Images were deconvolved using the Volocity software (PerkinElmer). Scale bar, 2 μm. (e) Quantification of the foci described in d. Cherry-VHL foci (n=214) located inside and outside the nuclear periphery were assessed for co-localization with Cmr1-YFP. (f) Cmr1 and the CCT–chaperonin complex co-localize at perinuclear foci. Cells express Cmr1-yEmRFP (IG111), Nup49-CFP (pNEB21) and Cct6-YFP (pIG20). Orange arrowhead, Cmr1 and Cct6 co-localizing at a perinuclear focus. Yellow arrowhead, Cct6 focus. Scale bar, 2 μm. (g) Cmr1 is not degraded during the DNA-damage response. G1-arrested cells (IG174) were released into YPD containing 200 μg ml⁻¹ cycloheximide (CHX) or 0.05% MMS. After 60 min of MMS treatment, CHX and 75 μg ml⁻¹ MG132 or CHX and MG132 were added. Cmr1-TAP and tubulin were analysed by immunoblotting, using cmr1Δ (DP1) as a negative control. Cmr1 protein levels relative to the sample taken before addition of CHX are indicated below the blot.

Figure 4. Cmr1 interacts with chromatin and replication factors.

(a) Schematic representation of the principle of the bimolecular fluorescence complementation (BiFC) assay. N-terminal (VN) and C-terminal (VC) non-fluorescent fragments of Venus fluorescent protein are fused to putative interacting proteins, to assess their physical association by the appearance of a fluorescence signal. (b) Gene Onthology (GO) enrichment analysis of Cmr1 interaction partners in BiFC. Significantly overrepresented GO biological process terms are shown. Bars indicate the percentage of Cmr1 interactors belonging to the indicated GO term as determined using BinGO. P-values were calculated by Fisher’s t-test and corrected using the Benjamini and Hochberg false discovery rate correction. (c) Cmr1 interaction with Mcm3 is enhanced by MMS. The strain from the BiFC screen expressing Cmr1-VC, Mcm3-VN and NLS-yEmRFP was subjected to fluorescence microscopy before and after treatment with 0.05% MMS for 2 h. Scale bar, 2 μm. (d) Quantification of the intensity of the Cmr1-Mcm3 interaction signal in cells from experiment in c. Two to 3 replicates of 100–200 cells were analysed for each condition. The box plot displays nuclear fluorescence intensities in arbitrary units (AU), where the line across the box identifies the median sample value, the ends of the box are the 25th and 75th percentiles, and whiskers represent minimum and maximum values.

Figure 5. Cmr1 is involved in the response to replication stress.

(a) Mrc1 and Cmr1 foci co-localize during replication stress. Co-localization between Cmr1-YFP and Mrc1-CFP was assessed in untreated cells (IG160-4A) and after treatment with MMS for 2 h. Representative images are shown. Scale bar, 2 μm. Arrowhead indicates INQ focus. (b) Quantification of co-localization of Mrc1 and Cmr1 in response to HU, MMS and MG132 (n>200). (c) Checkpoint-defective Mrc1AQ protein accumulates in the nucleus. Mrc1-YFP (IG147) and Mrc1AQ-YFP (IG315) were imaged. Scale bars, 2 μm. (d) Quantification of Mrc1AQ protein levels. Images from c were quantified (n>150). The box plot displays fluorescence intensities in arbitrary units (AU), where the line across the box identifies the median sample value, the ends of the box are the 25th and 75th percentiles, and whiskers represent minimum and maximum values. (e) Checkpoint-defective Mrc1AQ exhibits reduced recruitment to INQ. Percentage of cells with Mrc1-YFP or Mrc1AQ-YFP foci was quantified after treatment with 0.05% MMS for 2 h. Error bars represent 95% confidence intervals (n>150). (f) Mrc1 focus formation requires BTN2 and HSP42. Mrc1-YFP localization was assessed in wild-type (IG147), btn2Δ (CC1–3B) and hsp42Δ (CC2–6B) cells. Representative images of Mrc1 localization are shown. Scale bars, 2 μm. (g) Quantification of Mrc1 foci in hsp42Δ and btn2Δ mutants shown in f. Error bars represent 95% confidence intervals. Two replicates, n>250. (h) Schematic representation of Mrc1 fusion to a fluorescent timer. The ratio between the sfGFP (fast maturing) and mCherry (slow maturing) fluorescence intensities was calculated as a measure of Mrc1 protein turnover. (i) Quantification of Mrc1 protein turnover. The fold difference between the GFP and mCherry nuclear fluorescence was measured in wild-type (CC98) and btn2Δ (CC102-9C) strains. Box plot were displayed as in d. Two replicates, n>100. (j) CMR1 deletion suppresses the DNA damage sensitivity of checkpoint mutants. Tenfold serial dilutions were plated on YPD or YPD containing the indicated drug. Strains were ML8–9A (wt), DP1 (cmr1Δ), IG156-7D (mrc1Δ), IG156-6C (mrc1Δ cmr1Δ), IG257-9C (pph3Δ), IG257-2C (pph3Δ cmr1Δ), IG177-9C (ctf18Δ) and IG177-8C (ctf18Δ cmr1Δ). (k) cmr1Δ is synthetic sick with homologous recombination mutants. Tenfold serial dilutions were plated on YPD or YPD containing 0.001% MMS. Strains were ML8-9A (wt), DP1 (cmr1Δ), IG164-1D (mre11Δ), IG164-2B (mre11Δ cmr1Δ), IG162-1D (rad52Δ) and IG162-2D (rad52Δ cmr1Δ).

Figure 6. Cmr1 promotes adaptation to the DNA damage checkpoint.

(a) Cmr1 genetic interaction with checkpoint deactivation pathways is functionally associated with INQ. Wild-type (ML8–9A), cmr1Δ (DP1), dia2Δ (CC4–19D), dia2Δ cmr1Δ (CC4-3B), pph3Δ (IG257-9C), cmr1Δ pph3Δ (IG257-2C), dia2Δ pph3Δ (IG296-2C), cmr1Δ dia2Δ pph3Δ (IG296-49B), hsp42Δ pph3Δ dia2Δ (IG322-12A) and btn2Δ pph3Δ dia2Δ (IG323-19D) strains were plated. (b) Mutants of INQ are checkpoint adaptation defective. Strains with a conditional cdc13-1 mutation and otherwise wild type (ML815-8A), exo1Δ (DLY1296), cmr1Δ (ML808-12D), rpd3Δ (ML807-2D), btn2Δ (ML821-4C) or hsp42Δ (ML822-4C) were examined. (c) Mutants of INQ are defective in Rad53 dephosphorylation during adaptation. cdc13-1 mutants were grown at 25 °C before being shifted to the restrictive temperature of 37 °C. Samples were harvested at 0, 6 and 24 h after temperature shift. Rad53 phosphorylation was detected by immunoblotting and the relative shift (Rad53^P/(Rad53+Rad53^P)) for each of the samples was quantified in the lower panel. Error bars reflect the s.d. of two independent experiments. (d) Negative genetic interaction between cmr1Δ and slx8Δ. Wild-type (ML8–9A), cmr1Δ (DP1), slx8Δ (NEB290-1B) and cmr1Δ slx8Δ (IG256-5D) strains were plated. (e) Cmr1 and Slx8 promote turnover of SUMO foci. Wild-type, cmr1Δ, slx8Δ and cmr1Δ slx8Δ cells (same strains as in d) ectopically expressing yEmRFP-Smt3 (pML133) were imaged. Percentage of cells with Smt3 foci was quantified. Error bars represent 95% confidence intervals (n>100). (f) Sumoylated and polysumoylated proteins accumulate in the cmr1Δ mutant. Cells from the experiment in d were transformed with a plasmid expressing 3myc-Smt3 (pRS313-3myc-Smt3), harvested in log phase and immunoprecipitation using anti-myc-coupled dynabeads was performed on whole-cell extracts. The temperature-sensitive ubc9-1 mutant (IG246-2C) grown at 37 °C was used as a negative control. Bands corresponding to high-molecular weight (high-MW) SUMO conjugates were quantified from two independent experiments using ImageJ. Error bars indicate s.d. (g) Co-localization of Slx8 and SUMO at INQ. Cells expressing Cmr1-CFP, Slx8-YFP and yEmRFP-Smt3 (IG302-1D transformed with pML133) were imaged. Representative image of co-localization between Smt3 and Slx8 at INQ (Cmr1) is shown. Scale bar, 2 μm. (h) INQ contains sumoylated proteins. Cmr1 and Smt3 localization was monitored in wild type (IG66) and ubc9^ts mutant (IG246-2C) after incubation at 37 °C for 2 h in the presence of 0.05% MMS. Error bars represent 95% confidence intervals (n>100).

Figure 7. hWDR76 interaction network and subnuclear localization suggest conservation of Cmr1 function.

(a) Domain organization of Cmr1 and human WDR76. Human WDR76 shares 29% sequence similarity with Cmr1 from S. cerevisiae. Filled boxes indicate WD40 repeats. NLS, putative nuclear localization signal. (b) Human WDR76 interacts with HELLS, SUGT1, XRCC5, XRCC6 and the CCT–TRiC complex. SILAC-labelled HeLa cells were transfected with GFP-WDR76 or empty vector. GFP-WDR76 and its interacting proteins were enriched using GFP-Trap resin. Proteins were resolved by SDS–PAGE and digested in-gel with trypsin. Peptides were analysed on a quadrupole Orbitrap mass spectrometer. The plot shows log(10) SILAC ratios of proteins associated with GFP-WDR76 compared with background. WDR76 is highlighted in red and several other interactors are also indicated. (c) Human WDR76 localizes into nuclear foci. Twenty-four hours after GFP-WDR76 transfection, U2OS cells were treated with 10 μM MG132 or 1.5 mM MMS for 2 h. Immunofluorescence analyses of 53BP1 were performed with anti-53BP1 antibody. DAPI was used to stain nuclei. Scale bar, 20 μm. (d) Human WDR76 does not co-localize with PCNA. Stably expressing RFP-PCNA U2OS cells were transfected with GFP-WDR76. 24 h after transfection, cells were fixed and stained with DAPI. Scale bar, 20 μm. (e) Model for the role of Cmr1 in promoting replication recovery. See Discussion for details.