USP42 enhances homologous recombination repair by promoting R-loop resolution with a DNA-RNA helicase DHX9

Abstract

The nucleus of mammalian cells is compartmentalized by nuclear bodies such as nuclear speckles, however, involvement of nuclear bodies, especially nuclear speckles, in DNA repair has not been actively investigated. Here, our focused screen for nuclear speckle factors involved in homologous recombination (HR), which is a faithful DNA double-strand break (DSB) repair mechanism, identified transcription-related nuclear speckle factors as potential HR regulators. Among the top hits, we provide evidence showing that USP42, which is a hitherto unidentified nuclear speckles protein, promotes HR by facilitating BRCA1 recruitment to DSB sites and DNA-end resection. We further showed that USP42 localization to nuclear speckles is required for efficient HR. Furthermore, we established that USP42 interacts with DHX9, which possesses DNA-RNA helicase activity, and is required for efficient resolution of DSB-induced R-loop. In conclusion, our data propose a model in which USP42 facilitates BRCA1 loading to DSB sites, resolution of DSB-induced R-loop and preferential DSB repair by HR, indicating the importance of nuclear speckle-mediated regulation of DSB repair.

Fig. 1. A focused screen identified transcription-related nuclear speckle factors as HR regulators.

a U2OS cells treated with tubercidin or mock treated (dimethyl sulfoxide: DMSO) were subjected to immunofluorescence staining with the anti-SC35 antibody. Scale bar: 10 μm. b U2OS cells were incubated with tubercidin or DMSO followed by treatment with CPT. Cell extracts were subjected to immunoblotting analysis with the indicated antibodies. c Screen for nuclear speckle factors involved in HR regulation. The DR-GFP assay was performed with siRNA pools targeting nuclear speckle factors. The siRNA targeting CtIP or luciferase (control: Ctrl) was used as a positive or negative control, respectively, and indicated with blue markers. Homology directed repair efficiency is plotted as a Z-score. The inset magnifies the data for top hits with gene symbols. The genes previously reported to be involved in HR regulation are indicated with green. d A DR-GFP assay was carried out with two independent siRNAs targeting selected top hits from the initial screen. GFP-positive cell populations were normalized to mock treatment (Ctrl), which was set to 100% (mean ± SEM, n = 3). *p < 0.05. **p < 0.01. ***p < 0.005.

Fig. 2. USP42 promotes HR by facilitating DNA-end resection.

a U2OS cells transfected either with siRNA control (Ctrl) or siRNA targeting USP42 were treated with CPT or mock treated. Cell extracts were analysed by immunoblotting with the indicated antibodies. Asterisk indicates endogenous USP42. Immunoblotting with the anti-γH2AX antibody was used as a control for DNA damage induction. b A DNA-end resection assay was carried out with the cells transfected with the indicated siRNAs (mean ± SEM, n = 3). The siRNA targeting CtIP was a positive control. c RAD51 foci formation efficiency was examined with cells transfected with the indicated siRNAs (mean ± SEM, n = 3). The population of cells with >5 RAD51 foci was plotted. d Cells transfected with the indicated siRNAs were subjected to a clonogenic survival assay (mean ± SEM, n = 4 for siCtrl and siUSP42#2, n = 3 for siUSP42#1, siUSP42#3, and siXRCC4). The siRNA targeting XRCC4 was used as a positive control. e U2OS, USP42 KO, or USP42 KO cells complemented with either GFP or GFP-USP42 (FL) were treated with CPT and subjected to immunoblotting analysis with the indicated antibodies. End. USP42: endogenous USP42. f The indicated cell lines were subjected to a clonogenic survival assay [mean ± SEM, n = 6 for U2OS, USP42 KO + GFP and USP42 KO + GFP-USP42 (FL), n = 5 for USP42 KO]. ***p < 0.005.

Fig. 3. USP42 is required for efficient recruitment of BRCA1 to DSB sites.

a BRCA1 foci formation efficiency was examined with cells transfected with the indicated siRNAs. (Left) The representative images are shown. (Right) The population of cells with >10 BRCA1 foci was plotted (mean ± SEM, n = 3). The cells positive for CENPF staining were analysed. See Fig. S3 for the representative images of undamaged condition. b 53BP1 foci formation efficiency was examined with cells transfected with the indicated siRNAs. The number of foci per nucleus after IR were plotted as a Box and Whiskers plot (median, 5–95 percentile, n = 3). The cells positive for CENPF staining were analysed. c U2OS or USP42 KO cells were transfected with plasmids encoding either GFP or GFP-MRE11. Cell extracts were subjected to immunoprecipitation with an anti-GFP antibody followed by immunoblotting analysis with the indicated antibodies. ***p < 0.005.

Fig. 4. Nuclear speckle localization of USP42 is required for efficient HR.

a Schematic representation of USP42 protein and truncated mutants. The numbers represent amino acid residues. USP: ubiquitin-specific protease domain, P: proline-rich region, R: arginine-rich region, K: lysine-rich region. Localization to nuclear speckles of the indicated mutants are shown on the right. b Subcellular localization of various GFP-fused USP42 proteins. Cells transfected with the plasmids encoding the indicated GFP-USP42 were subjected to immunofluorescence staining with an anti-SC35 antibody. Scale bar: 10 μm. c U2OS or USP42 KO cells complemented with either GFP or GFP-USP42 (FL, ΔC or Δ946–1196) were treated with CPT and subjected to immunoblotting analysis with the indicated antibodies. Note that endogenous USP42 (End. USP42) and GFP-USP42 (FL) could only be detected in pellet fraction with an anti-USP42 antibody, while GFP and GFP-USP42 (ΔC) were detected in soluble fraction with an anti-GFP antibody. GFP-USP42 (Δ946–1196) was detected in both fractions. d The signal intensities of phosphorylated RPA2 that were normalized to the signal intensity of tubulin were plotted [mean ± SEM, n = 11 for U2OS and USP42 KO, n = 10 for USP42 KO + GFP, n = 8 for USP42 KO + GFP-USP42 (FL), n = 5 for USP42 KO + GFP-USP42 (ΔC), n = 3 for USP42 KO + GFP-USP42 (Δ946–1196)]. *p < 0.05. ***p < 0.005.

Fig. 5. USP42 is epistatic with DHX9 in the cellular survival after DSB induction and promotes resolution of DSB-induced R-loop.

a, b The interaction between USP42 and DHX9 was tested by coimmunoprecipitation with either GFP-USP42 a or GFP-DHX9 b followed by immunoblotting analysis with the indicated antibodies. Transfection with a plasmid encoding GFP was a negative control. c U2OS cells transfected with the indicated siRNAs were treated with CPT and then subjected to immunoblotting analysis with the indicated antibodies. d A DR-GFP assay was performed with the cells transfected with the indicated siRNAs (mean ± SEM, n = 3). e BRCA1 foci formation efficiency was examined with cells transfected with the indicated siRNAs. (Upper) The representative images are shown. (Lower) The population of cells with >10 BRCA1 foci was plotted (mean ± SEM, n = 3). The cells positive for CENPF staining were analysed. See Fig. S4B for the representative images of undamaged condition. f U2OS or USP42 KO cells transfected with the indicated siRNAs were subjected to a clonogenic survival assay (mean ± SEM, n = 5). g, h The indicated cells g or cells transfected with the indicated siRNAs h were treated with phleomycin (Phleo., +) and then further cultured for 2 h upon removal of phleomycin (2 h). Purified genomic DNA was subjected to slot blot analysis with an S9.6 antibody. Signal intensity was normalized to mock-treated samples (−), and then plotted as a Box and Whiskers plot (median, 5–95 percentile, n = 5 for U2OS, USP42 KO, and siBRCA1, n = 7 for siCtrl). *p < 0.05, ***p < 0.005.