PRMT5-Dependent Methylation of the TIP60 Coactivator RUVBL1 Is a Key Regulator of Homologous Recombination

Abstract

Protein post-translation modification plays an important role in regulating DNA repair; however, the role of arginine methylation in this process is poorly understood. Here we identify the arginine methyltransferase PRMT5 as a key regulator of homologous recombination (HR)-mediated double-strand break (DSB) repair, which is mediated through its ability to methylate RUVBL1, a cofactor of the TIP60 complex. We show that PRMT5 targets RUVBL1 for methylation at position R205, which facilitates TIP60-dependent mobilization of 53BP1 from DNA breaks, promoting HR. Mechanistically, we demonstrate that PRMT5-directed methylation of RUVBL1 is critically required for the acetyltransferase activity of TIP60, promoting histone H4K16 acetylation, which facilities 53BP1 displacement from DSBs. Interestingly, RUVBL1 methylation did not affect the ability of TIP60 to facilitate ATM activation. Taken together, our findings reveal the importance of PRMT5-mediated arginine methylation during DSB repair pathway choice through its ability to regulate acetylation-dependent control of 53BP1 localization.

Keywords: 52BP1; DNA repair; PRMT5; RUVBL1; TIP60; arginine methylation; homologous recombination.

Graphical abstract

Figure 1

PRMT5 Is Required for Effective Double-Strand Break Repair (A) HeLa-shCTRL, shPRMT5 hairpin 1 (1), and shPRMT5 hairpin 2 (2) cells were stimulated with 3Gy IR and co-stained for γH2AX and 53BP1 foci at the time points indicated. (B and C) Quantification of (A) is shown (mean ± SEM; n = 3). (D) Reconstitution of PRMT5-depleted HeLa cells with wild-type or catalytically inactive (MD) PRMT5. Quantification of 53BP1 foci 24 hr after IR (3 Gy) is shown (mean ± SEM; n = 3). (E) Primary, non-immortalized prmt5^f/f and prmt5^f/f;CreERT2 mouse embryonic fibroblasts (MEFs) were cultured in the absence or presence of 4-hydroxy-tamoxifen (4-OHT; 500 nM for 24 hr), followed by 4 days in 4-OHT-deficient media, and then immunoblotted for PRMT5. (F–H) Cells were stimulated with 5Gy IR and co-stained for γH2AX and 53BP1 foci at the time points indicated (F). Quantification of γH2AX (G) and 53BP1 (H) foci is shown (mean ± SEM). (I) HeLa-shPRMT5 cells were exposed to increasing doses of IR and cell viability was determined by colony survival assay (mean ± SD; n = 3). (J) Analysis of chromatid gaps and breaks per metaphase is shown (mean ± SEM; ^∗∗p < 0.0098; n = 3). (K) Quantification of 53BP1 foci in mitosin-positive (late S/G2) cells 24 hr after 3Gy IR is shown (mean ± SEM; n = 3). Scale bar, 10 μm. See also Figure S1.

Figure 2

PRMT5 Promotes Homologous Recombination-Mediated DSB Repair Independently of ATM Activation (A) HeLa-shCTRL and shPRMT5 cells were stimulated with 3Gy IR and RAD51 foci analyzed 6 hr later. Quantification of number of cells with >10 RAD51 foci is shown below (mean ± SEM; ^∗∗p = 0.006; n = 3). (B) Quantification of RPA foci in CENPF-positive (late S/G2) cells 8 hr after 3Gy IR is shown (mean ± SEM; ^∗∗p = 0.006; n = 3). (C) Time course of phosphorylated and total protein levels is shown. (D) U2OS-DR3 cells were treated with siRNA as indicated before transfection of I-Sce1 and myc-PRMT5. The number of GFP-positive cells (HR proficient) was determined by flow cytometry (mean ± SD; n = 3). (E) PRMT5-depleted cells were treated with increasing doses of camptothecin (CPT; left panel) or Olaparib (right panel), and survival was determined by colony assay (mean ± SD; n = 3). (F) Cells were treated with CPT (1 μM) for the time points indicated and immunoblotted for phosphorylated and total proteins. Scale bar, 10 μm. See also Figure S2.

Figure 3

PRMT5 Binds to and Methylates RUVBL1 at R205 (A) SYPRO Ruby stain of PRMT5 protein complexes isolated from 293T cells stably expressing tetracycline-regulated TAP-tagged catalytically inactive PRMT5 (PRMT5 MD) is shown. (B) Co-immunoprecipitation of myc-PRMT5 MD with FLAG-tagged RUVBL1/RUVBL2 is shown. (C) Endogenous immunoprecipitation of RUVBL1 from HeLa cells co-immunoprecipitates endogenous PRMT5. (D) Transfected HeLa-shCTRL or shPRMT5 cell lines were treated with CHX/CAM and labeled with [³H]-methyl methionine. FLAG-RUVBL1 was immunoprecipitated, and incorporated methyl groups were detected by SDS-PAGE followed by autoradiograph. Quantification of [³H]-methyl signal intensity adjusted for immunoprecipitation (IP) and normalized to shCTRL is shown (mean ± SD; ^∗p = 0.03 and ^∗∗∗p = 0.0009). (E) Autoradiograph and immunoblots of 293T cells transfected with FLAG-RUVBL1 or FLAG-RUVBL1-R205K, treated with CHX/CAM and labeled with [³H]-methyl methionine. Quantification of [³H]-methyl signal intensity adjusted for IP and normalized to wild-type is shown (mean ± SD; n = 3). (F) Immunoblot analysis of endogenous RUVBL1-R205me2s in HeLa-shCTRL or shPRMT5 cell lines is shown. (G) Immunoblot analysis of endogenous RUVBL1-R205me2s in HeLa cells treated with Adox (100 μM; 24 hr) or GSK591 (5 μM; 24 hr) is shown. (H) Transfected 293T cells were damaged with 10Gy IR and myc-PRMT5 MD immunoprecipitated. Associated RUVBL1 was detected by FLAG immunoblotting. (I) Autoradiograph and immunoblots of 293T cells transfected with FLAG-RUVBL1, treated with CHX/CAM, and labeled with [³H]-methyl methionine for 4 hr before treating with 10Gy IR are shown. (J) Immunoblot analysis of whole-cell lysates transfected with FLAG-RUVBL1 and damaged with 10Gy IR. See also Figures S3–S5.

Figure 4

Mutation of RUVBL1 at R205 Impedes DSB Repair after Ionizing Radiation (A) Cells were exposed to increasing doses of IR and cell viability was determined by colony assay. Representative images are displayed under quantification, and the inset immunoblot depicts RUVBL1 knockdown (mean ± SD; n = 3). (B) Generation of siRNA-resistant RUVBL1 wild-type (WT) or RUVBL1 methyl-deficient (R205K) HeLa cell lines is shown. (C and D) Cells were transfected with siRUVBL1, exposed to 3Gy IR, and the number of cells with >10 γH2AX (C) or 53BP1 (D) foci were quantified (mean ± SEM; n = 3). (E) Quantification of chromatid gaps and breaks after 2Gy IR is shown (mean ± SEM; n = 3). (F and G) HeLa-RUVBL1 WT or HeLa-RUVBL1-R205K cells were transfected with the indicated siRNAs and exposed to 3Gy IR. Cell survival was assessed by colony assay (F) (mean ± SD; n = 3). Number of cells with >10 53BP1 foci was scored (G) (mean ± SEM; ^∗∗p = 0.0095; n = 3). Scale bar, 10 μm. See also Figure S5.

Figure 5

Methylation of RUVBL1 Is Required for HR, but Not ATM Signaling (A and B) HeLa-RUVBL1-WT and HeLa-RUVBL1-R205K cells were transfected with siRUVBL1, exposed to 3Gy IR, and harvested at the time points indicated. CENPF/mitosin-positive cells were scored for RPA (A) and RAD51 (B) foci formation (mean ± SEM; n = 3). (C) HeLa-RUVBL1-WT and HeLa-RUVBL1-R205K cells were transfected with siRUVBL1, exposed to 10Gy IR, and harvested at the time points indicated for immunoblot analysis. (D and E) HeLa-RUVBL1-WT and HeLa-RUVBL1-R205K cells were transfected with siRUVBL1, treated with increasing doses of CPT (D) or Olaparib (E), and survival was assessed by colony survival assay (mean ± SEM; n = 3). Scale bar, 10 μm. See also Figure S5.

Figure 6

PRMT5-Dependent Methylation of RUVBL1 at R205 Regulates 53BP1 Removal and Is Epistatic with TIP60 (A) HeLa-shCTRL or shPRMT5 cells were siRNA transfected, exposed to 3Gy IR, and cell viability was assessed by colony survival. Results are expressed normalized to the respective non-treatment control (mean ± SD; n = 3). (B and C) HeLa-shCTRL or shPRMT5 cells were siRNA transfected, exposed to 3Gy IR, and stained for 53BP1 (B) or CENPF/RPA (C) foci 24 and 8 hr after damage, respectively (mean ± SEM; n = 3). (D and E) HeLa-shCTRL or RUVBL1-R205K cells were siRNA transfected, exposed to 3Gy IR, and stained for 53BP1 (D) or CENPF/RPA (E) foci 24 and 8 hr after damage, respectively (mean ± SEM; n = 3). (F) Validation of siTIP60 knockdown is shown. (G) Cells were transfected with si53BP1 and the number of mitosin-positive cells with >10 RAD51 foci was quantified (mean ± SEM; n = 3). (H) Cells were transfected with si53BP1 and the number of CENPF-positive cells with >10 RPA foci was quantified (mean ± SEM; n = 3). (I) Validation of 53BP1 siRNA SMARTpool. See also Figure S6.

Figure 7

PRMT5-Dependent Methylation of RUVBL1 Regulates TIP60 Acetyltransferase Activity (A and B) HeLa shCTRL or RUVBL1-R205K cells were siRNA transfected and treated with TSA (0.5 μM) for 16 hr before exposure to 3Gy IR. The number of cells with >10 53BP1 foci 24 hr after damage (A) and the number of CENPF-positive cells with >10 RPA foci 8 hr after damage (B) were quantified (mean ± SEM; n = 3). (C) In vitro TIP60 histone acetyltransferase assay. 293T cells were transfected with the TIP60 piccolo complex (TIP60/EPC1/ING3) and wild-type or methyl-deficient RUVBL1. TIP60 activity was determined by immunoblotting for recombinant H4 acetylation. (D) HeLa shCTRL, RUVBL1-WT, or -R205K cells were siRNA transfected, and acetylated histone marks were determined by immunoblotting. (E–G) ChIP performed with H4K16Ac (E), 53BP1 (F), and RUVBL1 (G) antibodies in the presence of mCherry-Lac1-Fok1-DD-induced DSBs, in U20S-265-RUVBL1 wild-type or RUVBL1-R205K-expressing cells. Data represent the average percentage input across the four primer pairs expressed relative to RUVBL1 WT (mean ± SEM; n = 3). (H) Model of how PRMT5-dependent methylation of RUVBL1 impacts on HR repair. PRMT5 symmetrically dimethylates RUVBL1 at R205, which stimulates TIP60 HAT activity enabling H4K16 acetylation and the removal of 53BP1 from DNA breaks. After end resection, RPA loading, and subsequent displacement by RAD51, homology searching and strand invasion can proceed enabling error-free HR-mediated repair of DSBs. Removal of PRMT5 results in hypomethylated RUVBL1 and ineffective TIP60 activity, causing reduced H4K16 acetylation and 53BP1 retention. Consequently, end resectioning is greatly impaired during late S/G2 phase, resulting in defective RPA and RAD51 recruitment, inappropriate error-prone NHEJ, and genomic instability. See also Figures S6 and S7.