Mammalian SWI/SNF complexes facilitate DNA double-strand break repair by promoting gamma-H2AX induction

Abstract

Although mammalian SWI/SNF chromatin remodeling complexes have been well established to play important role in transcription, their role in DNA repair has remained largely unexplored. Here we show that inactivation of the SWI/SNF complexes and downregulation of the catalytic core subunits of the complexes both result in inefficient DNA double-strand break (DSB) repair and increased DNA damage sensitivity as well as a large defect in H2AX phosphorylation (gamma-H2AX) and nuclear focus formation after DNA damage. The expression of most DSB repair genes remains unaffected and DNA damage checkpoints are grossly intact in the cells inactivated for the SWI/SNF complexes. Although the SWI/SNF complexes do not affect the expression of ATM, DNA-PK and ATR, or their activation and/or recruitment to DSBs, they rapidly bind to DSB-surrounding chromatin via interaction with gamma-H2AX in the manner that is dependent on the amount of DNA damage. Given the crucial role for gamma-H2AX in efficient DSB repair, these results suggest that the SWI/SNF complexes facilitate DSB repair, at least in part, by promoting H2AX phosphorylation by directly acting on chromatin.

Figure 1

Inactivation of the SWI/SNF complexes results in increased DNA damage sensitivity and decreased DSB repair. (A) (top) tet-VP16 and B05-1 cells were incubated with or without tetracycline for 4 days, and cell lysates were analyzed for the expression of flag-BRG-1 by immunoblottings. The expression of actin was also analyzed as internal control. (Bottom) Inactivation of the SWI/SNF complexes renders cells hypersensitive to IR. Cells grown as per in (A) were irradiated by various doses before the viability was determined by colony formation assays. Data are presented as mean±standard deviation (s.d.) from triplicates. (B) Neutral comet assays show that SWI/SNF inactivation leads to inefficient DSB repair. B05-1 (the first three graphs) and tet-VP16 cells (the last graph) grown as in (A) were irradiated by indicated doses, and the cells were collected immediately (0 h) or at the indicated time points after irradiation for comet assays. Each graph is depicted as mean±s.d. from three independent experiments. Immunoblot analysis for flag-BRG-1 expression in three independent experiments is shown next to the corresponding graphs. (C) The pictures show representative comet images of B05-1(+tet) and B05-1(−tet) cells untreated (0 Gy), or immediately (0 h) and 10 h after irradiation by 25 Gy.

Figure 2

The expression of the known DSB repair genes is not significantly affected by SWI/SNF inactivation. (A) The fold changes of the expression of thus far known 24 DBS repair genes by SWI/SNF inactivation are summarized (these are all below the cutoff value). The data sources for the fold change of each gene are indicated in the last column. Note that the majority of the genes (17 genes) were changed by less than 1.2-fold by SWI/SNF inactivation, and that Rad51 and Rad51D were decreased by only 1.3-fold and the remaining five genes were rather increased by SWI/SNF inactivation. (B) (Top) The effects of SWI/SNF inactivation on the expression of the DSB repair genes that were not included in the mouse 11K gene chip. RT–PCR was performed using total RNA isolated from B05-1(+tet) and B05-1(−tet) cells. The mRNA expression of GAPDH was analyzed as internal control. The predicted sizes of the PCR products are as follow: RAD54B, 614 bp; EME1, 681 bp; LIG4, 641 bp; Artemis, 655 bp; GAPDH, 361 bp. The first lane (M) is 100-bp standard size marker (the most bottom band is 400 bp). (Bottom) The expression of flag-BRG-1 and α-tubulin (internal control) from the cells used in the top panel was analyzed by immunoblottings.

Figure 3

Cells inactivated for the SWI/SNF complexes are compromised in the induction of γ-H2AX after DNA damage. (A) (Top) Immunoblot analysis for the expression of flag-BRG-1 and actin (internal control) from the cells used below. (Middle and bottom) Indicated cells were untreated (0 Gy) or irradiated by 25 Gy, and histones were acid-extracted at various time points for analysis of the levels of γ-H2AX and H2A (loading control) by immunoblottings. (B) (Top) Immunoblot analysis for the expression of flag-BRG-1 and actin (internal control) from the cells used below. (Bottom) The effects of SWI/SNF inactivation on the induction of γ-H2AX after irradiation by various doses. The experiments were carried out as in (A) except that the expression of H2AX was analyzed as loading control. (C) (Top) Immunoblot analysis for the expression of flag-BRG-1 and actin (internal control) from the cells used below. (Bottom) The effects of SWI/SNF inactivation on the formation of γ-H2AX foci were analyzed by immunofluorescence microscopy. Indicated cells were untreated (0 Gy) or irradiated by 0.5, 1, 2 and 6 Gy, and the cells were fixed after 1 h for immunostaining with γ-H2AX antibodies. The graph shows the average number of γ-H2AX foci determined by counting about 50 nuclei per sample. The error bar indicates mean±s.d. from three independent experiments. (D) Representative confocal images of γ-H2AX foci taken 15, 30 and 60 min after irradiation by 2 Gy. The nuclei were visualized by DAPI staining.

Figure 4

Downregulation of both BRG-1 and hBrm results in γ-H2AX defect, inefficient DSB repair and increased DNA damage sensitivity. (A) The effects of BRG-1/hBrm downregulation on the γ-H2AX induction after irradiation. HeLa cells were mock-transfected (−) or cotransfected with BRG-1 and hBrm siRNAs (+) for 48 h before irradiation by indicated doses. After 1 h, cells were collected and divided into two for preparation of whole-cell lysates and acid extracts of histones. The whole-cell lysates were analyzed for the expression of BRG-1, hBrm and actin (internal control) by immunoblottings. The acid extracts were divided into two for analyzing the levels of γ-H2AX and the expression of H2A and H2AX by immunoblottings in separate gels. The γ-H2AX blots show both short and long exposures for better comparison of γ-H2AX levels. A representative of five independent experiments is shown. (B) The effects of BRG-1/hBrm downregulation on the kinetics of γ-H2AX induction. HeLa cells transfected as in (A) were irradiated by 25 Gy and collected at various time points for analyzing the expression of indicated proteins. A representative of five independent experiments is shown. (C) The effects of BRG-1/hBrm downregulation on the formation of γ-H2AX foci. (Left) HeLa cells cotransfected with BRG-1 and hBrm siRNAs were irradiated by 2 Gy, and after 1 h cells were fixed and double stained with antibodies for BRG-1 and γ-H2AX. Confocal images were taken so as to capture both nontransfected and transfected cells in the same picture. The nuclei were visualized by DAPI staining. (Right) γ-H2AX foci were counted using the confocal image in the left panel and depicted as a graph. (D) The effects of BRG-1/hBrm downregulation on the viability after DNA damage. HeLa cells transfected as in (A) were irradiated by indicated doses before viability was evaluated by colony formation assays. Data are presented as mean±s.d. from triplicates. Immunoblot analysis of siRNA knockdown of BRG-1 is shown next to the graph. (E) The effects of BRG-1/hBrm downregulation on DSB repair. HeLa cells were transfected as in (A) and irradiated by 50 Gy and the cells were collected at the indicated time points before subjecting to neutral comet assays. Each graph is depicted as mean±s.d. from two independent experiments. Immunoblot analysis of siRNA knockdown of BRG-1 for the two independent experiments is shown next to the graph.

Figure 5

The effects of the SWI/SNF complexes on γ-H2AX are independent of ATM, DNA-PK and ATR. (A) siRNA knockdown of BRG-1 and hBrm has no effect on the expression of ATM, DNA-PKcs and ATR. HeLa cells mock transfected or cotransfected with BRG-1 and hBrm siRNA for 48 h were untreated (0 Gy) or irradiated by 10 Gy. After 1 h, cell lysates were analyzed for the expression of the indicated proteins by immunoblottings. (B) ATM activation after DNA damage normally occurs in SWI/SNF-inactivated cells. B05-1(+tet) and B05-1(−tet) cells were untreated (0 Gy) or irradiated by 10 Gy, and the cell lysates were prepared at the indicated time points for analysis of the diagnostic autophosphorylation of ATM at Ser-1981 as well as the expression of ATM by immunoblottings. The expression of flag-tagged proteins was analyzed to ensure the induction of the dominant negative BRG-1, and the expression of α-tubulin was analyzed as internal control. (C) siRNA knockdown of BRG-1 and hBrm has no effect on the activation of ATM after DNA damage. HeLa cell lysates prepared as in (A) were analyzed for the expression of the indicated proteins. (D) Nbs1 is phosphorylated in SWI/SNF-inactivated cells after DNA damage. The expression of phospho-Nbs1(Ser-343) and Nbs1 was analyzed by immunoblottings at 1 h postirradiation (30 Gy). The expression of flag-BRG-1 and α-tubulin (internal control) was also analyzed. (E) ATM foci can be formed in SWI/SNF-inactivated cells after DNA damage. B05-1(+tet) and B05-1(−tet) cells were fixed at 1 h after irradiation by 2 Gy, and double stained by antibodies for γ-H2AX and p-ATM(Ser-1981) before confocal images were captured. The nuclei were visualized by DAPI staining. (F) siRNA knockdown of BRG-1/hBrm has no effect on the formation of ATM foci after DNA damage. HeLa cells cotransfected with BRG-1 and hBrm siRNAs for 48 h were untreated (0 Gy) or irradiated by 2 Gy. Irradiated cells were collected after 15, 30, 60 and 240 min for double staining with antibodies for BRG-1 and p-ATM(Ser-1981). Confocal images were taken so as to capture both nontransfected and transfected cells in the same picture for each sample. The nuclei were visualized by DAPI staining. (G, H) DNA-PKcs and ATR can form foci in SWI/SNF-defected cells after DNA damage. B05-1(+tet) and B05-1(−tet) cells were irradiated by 2 Gy (G) or 10 Gy (H) followed by incubation for 2 h. Cells were collected for double staining with antibodies for γ-H2AX and p-DNA-PKcs(Thr-2609) (G), or with antibodies for γ-H2AX and ATR (H). The nuclei were visualized by DAPI staining.

Figure 6

The SWI/NF complexes bind to DSB-surrounding chromatin via interaction with γ-H2AX. (A) BRG-1 and mouse Brm (mBrm) bind to chromatin after DNA damage. NIH-3T3 cells were untreated or irradiated by 10 Gy, and after 1 h cells were collected and subjected to detergent extraction. Sequentially fractionated extracts (Fractions I–III) and pellet (Fraction IV) were subjected to immunoblottings with the antibodies against BRG-1, mBrm and Nbs1. A representative of four independent experiments is shown. (B) Time-course analysis for the chromatin binding of BRG-1 and mBrm (data not shown) after DNA damage. NIH-3T3 cells were untreated (0 Gy) or irradiated by 10 Gy, and at various time points after irradiation, cells were collected and subjected to detergent extraction chromatin retention assays as in (A). Equal protein loading was ensured by Ponceau staining (data not shown). Immunoblots for fraction III are shown. A representative of four independent experiments is shown. (C) Binding of BRG-1 and mBrm (data not shown) to chromatin increases in proportion to irradiated IR doses. NIH-3T3 cells were untreated (0 Gy) or irradiated by 2, 5, 10, 25 and 50 Gy. After 1 h, cells were collected and subjected to detergent extraction chromatin retention assays as in (A). Equal protein loading was ensured by Ponceau staining (data not shown). Immunoblots for fraction III are shown. A representative of three independent experiments is shown. (D) BRG-1 is specifically retained in the chromatin overlapping with γ-H2AX after detergent extraction. NIH-3T3 cells untreated (0 Gy) or irradiated by 5 Gy, and after 1 h, the cells were detergent fractionated in situ. Cells were then fixed and double stained with the antibodies for γ-H2AX and BRG-1 with the nuclei labeled with DAPI. The third row shows a representative staining pattern of the irradiated cells shown in the second row. (E) The phosphorylation of Flag-H2AX at Ser-139 (Flag-γ-H2AX) was analyzed using the histones extracted from 293T cells that stably express N-terminal Flag-tagged H2AX after irradiation as indicated. The size differences permitted a detection of Flag-tagged proteins and their endogenous counterparts on the same immunoblot gel using the antibodies for γ-H2AX or H2AX. The first lane is nonirradiated 293T cells harboring empty vector. (F) Flag-H2AX expressing 293T cells were examined for nuclear focus formation by double staining with the Flag and γ-H2AX antibodies 1 h after irradiation by 2 Gy. (G) 293T cells expressing an empty vector or Flag-H2AX were subjected to ChIP at 1 h after irradiation by 10 Gy. After formaldehyde crosslinking and sonication, chromatin was immunoprecipitated by anti-Flag antibodies and co-precipitation of BRG-1, hBrm and Nbs1 was analyzed by immunoblottings with the indicated antibodies.

Figure 7

The effects of SWI/SNF inactivation on DNA damage checkpoint. (A) (Left) Analysis of G2/M checkpoint in B05-1 cells. At 1 h after irradiation by indicated doses, cells were fixed for double staining with PI and the anti-phospho-H3 antibodies followed by FACS analysis. A representative FACS result is shown. (Right) The frequency of phospho-H3-positive cells (mitotic cells) at 1 h after irradiation was represented by a percentage (irradiated/nonirradiated). An average from three independent experiments is plotted as graph with mean±s.d. (B) The effects of siRNA downregulation of BRG-1 and hBrm on DNA damage G2/M checkpoints. HeLa cells were mock transfected or transfected with BRG-1 and hBrm siRNA for 48 h, and irradiated by the indicated doses. The mitotic cells were analyzed as in (B). An average from three independent experiments is plotted as graph with mean±s.d. Note that the overall levels of G2 arrest of HeLa cells in response to 0.5 and 1 Gy are much lower than those of NIH-3T3 cells. (C) SWI/SNF inactivation has no effect on the G2/M checkpoint in response to adriamycin. Cells were treated with 0.5 μM of adriamycin for 1 h, and harvested for analysis of mitotic cells as in (A). Percentages of phospho-H3-positive mitotic cells are indicated. (D) S-phase checkpoint analysis. (Top) The levels of new DNA synthesis were determined by BrdU incorporation method after cells were untreated (Un), treated with 0.5 μM of adriamycin (Adr), irradiated by 10-Gy IR, or 32 μM of cisplatin (Cis). The number of BrdU-positive nonirradiated cells was set to 100%, and an average percentage from three independent experiments was plotted as graph with mean±s.d. (Bottom) Immunoblot analysis shows that the expression of flag-BRG-1 was properly induced by tetracycline depletion. BrdU indicates the control reactions without BrdU and other abbreviations are same as the top panel.