ATM and CDK2 control chromatin remodeler CSB to inhibit RIF1 in DSB repair pathway choice

Abstract

CSB, a member of the SWI2/SNF2 superfamily, is implicated in DNA double-strand break (DSB) repair. However, how it regulates this repair process is poorly understood. Here we uncover that CSB interacts via its newly identified winged helix domain with RIF1, an effector of 53BP1, and that this interaction mediates CSB recruitment to DSBs in S phase. At DSBs, CSB remodels chromatin by evicting histones, which limits RIF1 and its effector MAD2L2 but promotes BRCA1 accumulation. The chromatin remodeling activity of CSB requires not only damage-induced phosphorylation on S10 by ATM but also cell cycle-dependent phosphorylation on S158 by cyclin A-CDK2. Both modifications modulate the interaction of the CSB N-terminal region with its ATPase domain, the activity of which has been previously reported to be autorepressed by the N-terminal region. These results suggest that ATM and CDK2 control the chromatin remodeling activity of CSB in the regulation of DSB repair pathway choice.

Fig. 1

CSB interacts with RIF1 and is recruited by RIF1 to FokI-induced DSBs. a Immunofluorescence of U2OS-265 cells with or without induction of FokI expression. Fixed cells were stained with anti-CSB antibody. Nuclei were stained with DAPI in blue in this and following figures. Scale bars, 5 μm. b Immunofluorescence of U2OS-265 cells expressing Myc-tagged CSB with or without induction of FokI expression. Fixed cells were stained with anti-Myc antibody. Scale bars, 5 μm. c Quantification of the percentage of U2OS-265 cells with Myc-CSB accumulated at FokI-induced DSBs. U2OS-265 cells were treated with DMSO or ATM inhibitor KU55933 for 1 h prior to induction of FokI expression. A total of 250 Myc-expressing cells were scored for each independent experiment in a blind manner. Standard deviations, referred to as SDs in this and the subsequent figures, from three independent experiments are indicated. *P < 0.05 (Student t test). d Quantification of the percentage of parental and 53BP1 KO U2OS-265 cells with Myc-CSB accumulated at FokI-induced DSBs. Scoring was done as described in 1c. SDs from three independent experiments are indicated. *P < 0.05 (Student t test). e Immunofluorescence with an anti-Myc antibody. 48 h prior to FokI induction, U2OS-265 cells were transfected with siRNA against scramble DNA (siControl) or RIF1 (siRIF1). Scale bars, 5 μm. f Quantification of the percentage of siControl- and siRIF1-expressing U2OS-265 cells with Myc-CSB accumulated at FokI-induced DSBs. Scoring was done as described in 1c. SDs from three independent experiments are indicated. *P < 0.05 (Student t test). g Quantification of the percentage of siControl- and siCSA-expressing U2OS-265 cells with Myc-CSB accumulated at FokI-induced DSBs. Scoring was done as described in 1c. SDs from three independent experiments are indicated. h Coimmunoprecipitation with IgG and anti-CSB antibody in HCT116 cells treated with or without 20 Gy IR. Immunoblotting was performed with anti-CSB, anti-RIF1 and anti-53BP1 antibodies. Protein molecular weight markers in kDa are indicated in this and the subsequent figures. i Coimmunoprecipitation with anti-RIF1 antibody in parental (WT) and CSB knockout (KO) HCT116 cells treated with or without 20 Gy IR. Immunoblotting was performed with anti-RIF1, anti-53BP1 and anti-CSB antibodies

Fig. 2

RIF1 interacts with CSB and recruits CSB to DSBs via its CTD domain. a Schematic diagram of RIF1. NLS: nuclear localization signal; CTD: C-terminal domain. b Schematic diagram of CSB. c Quantification of the percentage of cells exhibiting Myc-CSB accumulated at the lac operator array. U2OS-265 cells were co-transfected with Myc-CSB and various mCherry-LacR-RIF1 alleles as indicated. A total of 250 cells positive for Myc-CSB expression were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. d Quantification of the percentage of cells exhibiting Myc staining accumulated at the lac operator array. U2OS-265 cells were co-transfected with mCherry-LacR-RIF1-C and various Myc-tagged CSB alleles as indicated. A total of 250 cells positive for expression of various Myc-tagged CSB alleles as indicated were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. e Quantification of the percentage of cells exhibiting Myc-CSB-C accumulated at the lac operator array. U2OS-265 cells were co-transfected with Myc-CSB-C and various mCherry-LacR-RIF1-C alleles as indicated. A total of 250 cells positive for Myc-CSB-C expression were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. f Coimmunoprecipitation with anti-Myc antibody in 293T cells expressing the vector alone or Myc-CSB-C in conjunction with various mCherry-LacR-CTD alleles as indicated. Immunoblotting was done with anti-Myc and anti-mCherry antibodies. g Quantification of the intensity of Myc-CSB signal at the site of FokI-induced DSBs. 24 h post transfection with siControl or siRIF1, U2OS-265 cells were transfected with the EGFP vector alone or various siRIF1-resistant EGFP-RIF1 alleles as indicated and induced for FokI expression 48 h post transfection. Analysis of Myc-CSB signal intensity was only done for cells positive for expression of Myc-CSB, EGFP and mCherry-LacR-FokI. The respective numbers of cells analyzed for siControl/EGFP, siRIF1/EGFP, siRIF1/EGFP-RIF1 and siRIF1/EGFP-Rif-DCTDIII were 131, 107, 120 and 124. *P < 0.05 (Mann–Whitney test). NLS nuclear localization signal; UBD ubiquitin-binding domain; WHD winged helix domain

Fig. 3

CSB interacts with RIF1 and inhibits RIF1 at DSBs. a Quantification of the percentage of cells exhibiting Myc staining accumulated at the lac operator array in U2OS-265 cells co-transfected with indicated alleles. A total of 250 cells positive for Myc staining were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. b Quantification of the percentage of cells exhibiting Myc staining at the site of FokI-induced DSBs. A total of 250 U2OS-265 cells expressing various Myc-tagged CSB alleles as indicated were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. c Quantification of the intensity of RIF1 signal at FokI-induced DSBs. The respective numbers of cells analyzed for parental and CSB KO were 275 and 277. *P < 0.05 (Mann–Whitney test). d Quantification of the percentage of cells exhibiting MAD2L2 at FokI-induced DSBs. A total of 500–550 cells were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. *P < 0.05 (Student t test). e Quantification of the percentage of cells exhibiting GFP-PTIP at FokI-induced DSBs. A total of 500 cells were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. f Quantification of the intensity of BRCA1 signal at FokI-induced DSBs. The respective numbers of cells analyzed for parental and CSB KO were 282 and 294. *P < 0.05 (Mann–Whitney test). g HR-mediated repair of I-SceI-induced DSBs in U2OS-DR-GFP WT and CSB-KO cells. SDs from three independent experiments are indicated. *P < 0.05 (Student t test). h NHEJ-mediated repair of I-SceI-induced DSBs. SDs from three independent experiments are indicated. *P < 0.05 (Student t test). i Quantification of the intensity of RIF1 signal at FokI-induced DSBs. The respective numbers of cells analyzed for the vector alone, Myc-CSB, Myc-CSB-GG and Myc-CSB-GGG were 298, 304, 204, and 209. *P < 0.05 (Mann–Whitney test). j Quantification of the intensity of BRCA1 signal at FokI-induced DSBs. The respective numbers of cells analyzed for the vector alone, Myc-CSB, Myc-CSB-GG and Myc-CSB-GGG were 279, 270, 291, 258. *P < 0.05 (Mann–Whitney test)

Fig. 4

CSB is recruited to FokI-induced DSBs in S phase to inhibit RIF1. a Quantification of the percentage of synchronized Myc-CSB-expressing U2OS-265 cells exhibiting indicated proteins at FokI-induced DSBs. For Myc-CSB, a total of 250 Myc-CSB-expressing cells were scored for each independent experiment in blind. For RIF1, a total of 500–550 cells were scored in blind for each independent experiment. SDs from three independent experiments are indicated. b Quantification of the percentage of synchronized Myc-CSB-expressing U2OS-265 cells exhibiting γH2AX at FokI-induced DSBs. A total of 500–550 cells were scored in blind for each independent experiment. SDs from three independent experiments are indicated. c Western analysis of U2OS-265 parental (WT) and 53BP1 KO cells. The γ-tubulin blot was used as a loading control here and the following figures. d Quantification of the percentage of U2OS-265 WT and 53BP1 KO cells with γH2AX at FokI-induced DSBs. Scoring was done as in 3b. SDs from three independent experiments are indicated. e Quantification of the percentage of U2OS-265 WT and 53BP1 KO cells with RIF1 at FokI-induced DSBs. Scoring was done as in Fig 3b. SDs from three independent experiments are indicated. f Quantification of the percentage of synchronized Myc-CSB-expressing WT and 53BP1 KO U2OS-265 cells with Myc-CSB at FokI-induced DSBs. A total of 250 Myc-CSB-expressing cells were scored for each independent experiment in blind. SDs from three independent experiments are indicated. *P < 0.05 (Student t test). g Quantification of the percentage of U2OS-265 WT and CSB-KO cells with γH2AX at FokI-induced DSBs. Scoring was done as in Fig. 3b. SDs from three independent experiments are indicated. h Quantification of the percentage of synchronized U2OS-265 WT and CSB KO cells with RIF1 at FokI-induced DSBs. Scoring was done as in Fig. 3b. SDs from three independent experiments are indicated. *P < 0.05 (Student t test). i Quantification of the percentage of synchronized U2OS-265 WT and CSB KO cells with BRCA1 at FokI-induced DSBs. Scoring was done as in Fig. 3b. SDs from three independent experiments are indicated. *P < 0.05 (Student t test). j Western analysis of U2OS-265 WT and CSB KO cells

Fig. 5

CSB evicts histones from the chromatin surrounding I-PpoI-induced DSBs in vivo. a Relative occupancy of histone H2A in ddI-PpoI-expressing hTERT-RPE parental cells. Cells were either untreated (untx) or treated with Shield-1 and 4-OHT and then collected at indicated times. The x-axis represents the distance in kb upstream and downstream from the I-PpoI-induced DSB on chromosome 1, which was set as 0. The y-axis represents the relative occupancy of H2A of treated cells relative to untreated cells. Standard error of the mean (SEM) from three independent experiments are indicated. b Relative occupancy of histone H2B in ddI-PpoI-expressing hTERT-RPE parental cells. Both x- and y-axes are as described in a. SEM from three independent experiments are indicated. c Relative occupancy of histone H2A in ddI-PpoI-expressing hTERT-RPE CSB KO cells. Both x- and y-axes are as described in a. SEM from three independent experiments are indicated. d Relative occupancy of histone H2B in ddI-PpoI-expressing hTERT-RPE CSB KO cells. Both x- and y-axes are as described in a. SEM from three independent experiments are indicated. e Relative occupancy of histone H2A in ddI-PpoI-expressing hTERT-RPE CSB KO cells complemented with the vector alone, Myc-tagged wild-type CSB or Myc-tagged mutant CSB-W851R. Both x- and y-axes are as described in 5a. SEM from three independent experiments are indicated. f Relative occupancy of histone H2B in ddI-PpoI-expressing hTERT-RPE CSB KO cells complemented with the vector alone, Myc-tagged wild-type CSB or Myc-tagged mutant CSB-W851R. Both x- and y-axes are as described in a. SEM from three independent experiments are indicated

Fig. 6

The N-terminus of CSB mediates its chromatin remodeling activity to repress RIF1 accumulation at sites of DSBs. a Relative occupancy of histone H2A in ddI-PpoI-expressing CSB KO hTERT-RPE cells complemented with various alleles as indicated. Both x- and y-axes are as described in Fig. 5a. SEM from three independent experiments are indicated. b Relative occupancy of histone H2B in ddI-PpoI-expressing CSB KO hTERT-RPE cells complemented with various alleles as indicated. Both x- and y-axes are as described in Fig. 5a. SEM from three independent experiments are indicated. c Western analysis of hTERT-RPE-IPpoI-CSB KO cells expressing various alleles as indicated. d Quantification of the percentage of Myc-CSB and Myc-CSB-ΔN30-expressing U2OS-265 cells exhibiting anti-Myc staining at FokI-induced DSBs. A total of 250 cells positive for anti-Myc staining were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. e Quantification of the percentage of cyclin A- and cyclin A + cells with 10 or more IR-induced RIF1 foci. hTERT-RPE CSB KO cells stably expressing various alleles as indicated were treated with 2 Gy IR and fixed 1 h post IR. A total of 500–550 cells were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. *P < 0.05. ns: P > 0.05 (Student t test). f Quantification of the percentage of cyclin A + cells with ≥ 10 IR-induced BRCA1 foci. Scoring was done as in e. SDs from three independent experiments are indicated. *P < 0.05. ns: P > 0.05 (Student t test). g Quantification of the percentage of cyclin A + cells with 10 or more IR-induced RAD51 foci. hTERT-RPE CSB KO cells stably expressing various alleles as indicated were treated with 2 Gy IR and fixed 4 h post IR. Scoring was done as in e. SDs from three independent experiments are indicated. *P < 0.05. ns: P > 0.05 (Student t test). h Clonogenic survival assays. SDs from three independent experiments are indicated. *P < 0.05 (Student t test) for comparison between CSB and ΔN30. i HR-mediated repair. SDs from three independent experiments are indicated. *P < 0.05. ns: P > 0.05 (Student t test). j Western analysis of U2OS-DR-GFP parental and CSB KO cells

Fig. 7

ATM controls the chromatin remodeling activity of CSB through S10 phosphorylation. a Western analysis of U2OS CSB KO cells stably expressing various alleles as indicated. Cells were either treated with or without 10 Gy IR. b Western analysis of anti-CSB immunoprecipitates from HCT116 cells treated with or without 10 Gy IR. c Western analysis. Myc-CSB-expressing U2OS CSB KO cells were treated with DMSO, ATM inhibitor KU55933, ATR inhibitor VE-821 or DNA-PKcs inhibitor NU-7026 for 1 h prior to 10 Gy IR. d Western analysis of ATM-deficient GM16666A and ATM-complemented GM16667A cells. Cells were transfected with the vector alone or Myc-CSB, followed by treatment with 10 Gy IR 48 h post transfection. e Relative occupancy of histone H2A in ddI-PpoI-expressing hTERT-RPE CSB KO cells complemented with various alleles as indicated. Both x- and y-axes are as described in Fig. 5a. SEM from three independent experiments are indicated. f Relative occupancy of histone H2B in ddI-PpoI-expressing hTERT-RPE CSB KO cells complemented with various alleles as indicated. Both x- and y-axes are as described in Fig. 5a. SEM from three independent experiments are indicated. g Quantification of the percentage of cyclin A- and cyclin A+ cells with 10 or more IR-induced RIF1 foci. hTERT-RPE CSB KO cells stably expressing various alleles as indicated were treated with 2 Gy IR and fixed 1 h post IR. A total of 500–550 cells were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. *P < 0.05. ns: P > 0.05 (Student t test). h Quantification of the percentage of cyclin A+ cells with ≥ 10 IR-induced BRCA1 foci. Scoring was done as in 7 g. SDs from three independent experiments are indicated. *P < 0.05. ns: P > 0.05 (Student t test). i Quantification of the percentage of cyclin A+ cells with 10 or more IR-induced RAD51 foci. Cells were treated with 2 Gy IR and fixed 4 h post IR. Scoring was done as in 7 g. SDs from three independent experiments are indicated. *P < 0.05. ns: P > 0.05 (Student t test)

Fig. 8

Cyclin A–CDK2 controls the chromatin remodeling activity of CSB through S158 phosphorylation. a Western analysis of U2OS-CSB-KO cells stably expressing the vector alone, Myc-CSB or Myc-CSB carrying a S158A mutation. Immunoblotting was performed with anti-CSB-pS158 and anti-Myc antibodies. b Western analysis of synchronized HCT116 cells. The arrow indicates the position of CSB-pS158. Asterisks indicate non-specific bands. c Western analysis. Asynchronous and synchronized HCT116 cells post release from a double thymidine block as indicated were treated with DMSO or the CDK inhibitor roscovitine. The arrow indicates the position of CSB-pS158. Asterisks indicate non-specific bands. d In vitro kinase assays with recombinant cyclin A-CDK2 and bacteria-expressed recombinant CSB fragments as indicated. e Relative occupancy of histone H2A in ddI-PpoI-expressing hTERT-RPE CSB KO cells complemented with various alleles as indicated. Both x- and y-axes are as described in Fig. 5a. SEM from three independent experiments are indicated. f Relative occupancy of histone H2B in ddI-PpoI-expressing hTERT-RPE CSB KO cells complemented with various alleles as indicated. Both x- and y-axes are as described in Fig. 5a. SEM from three independent experiments are indicated. g Quantification of the percentage of cyclin A− and cyclin A+ cells with 10 or more IR-induced RIF1 foci. hTERT-RPE CSB KO cells stably expressing various alleles as indicated were treated with 2 Gy IR and fixed 1 h post IR. A total of 500–550 cells were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. *P < 0.05. ns: P > 0.05 (Student t test). h Quantification of the percentage of cyclin A+ cells with ≥ 10 IR-induced BRCA1 foci. Scoring was done as in 8 g. SDs from three independent experiments are indicated. *P < 0.05. ns: P > 0.05 (Student t test). i Clonogenic survival assays of olaparib-treated hTERT-RPE CSB-KO cells complemented with various alleles as indicated. SDs from three independent experiments are indicated. *P < 0.05 (Student t test) for comparison between CSB and S158A

Fig. 9

CSB phosphorylations on S10 and S158 modulate the interaction of the N-terminus with the ATPase domain. a Immunofluorescence of U2OS-265 cells transfected with Myc-CSB-N in conjunction with either mCherry-LacR-CSB-ATPase or mCherry-LacR-CSB-C. Scale bars, 5 μm. b Quantification of the percentage of U2OS-265 cells from a with anti-Myc staining at FokI-induced DSBs. A total of 250 Myc-expressing cells were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. c Quantification of the percentage of U2OS-265 cells with anti-Myc staining at FokI-induced DSBs. U2OS-265 were co-transfected with various alleles as indicated. A total of 250 Myc-expressing cells were scored for each independent experiment in a blind manner. SDs from three independent experiments are indicated. *P < 0.05 (Student t test). d Model for control of CSB chromatin remodeling activity by ATM and CDK2 in DNA DSB repair pathway choice in S/G2. See the text for details