DNA polymerase η promotes nonhomologous end joining upon etoposide exposure dependent on the scaffolding protein Kap1

Abstract

DNA polymerase eta (Pol η) is a eukaryotic member of the Y-family of DNA polymerase involved in translesion DNA synthesis and genome mutagenesis. Recently, several translesion DNA synthesis polymerases have been found to function in repair of DNA double-strand breaks (DSBs). However, the role of Pol η in promoting DSB repair remains to be well defined. Here, we demonstrated that Pol η could be targeted to etoposide (ETO)-induced DSBs and that depletion of Pol η in cells causes increased sensitivity to ETO. Intriguingly, depletion of Pol η also led to a nonhomologous end joining repair defect in a catalytic activity-independent manner. We further identified the scaffold protein Kap1 as a novel interacting partner of Pol η, the depletion of which resulted in impaired formation of Pol η and Rad18 foci after ETO treatment. Additionally, overexpression of Kap1 failed to restore Pol η focus formation in Rad18-deficient cells after ETO treatment. Interestingly, we also found that Kap1 bound to Rad18 in a Pol η-dependent manner, and moreover, depletion of Kap1 led to a significant reduction in Rad18-Pol η association, indicating that Kap1 forms a ternary complex with Rad18 and Pol η to stabilize Rad18-Pol η association. Our findings demonstrate that Kap1 could regulate the role of Pol η in ETO-induced DSB repair via facilitating Rad18 recruitment and stabilizing Rad18-Pol η association.

Keywords: DNA polymerase η; Kap1; Rad18; etoposide; nonhomologous end joining.

Figure 1

Pol η accumulates at ETO-induced damage sites and promotes NHEJ repair.A, localization of GFP-Pol η after ETO treatment. GFP-Pol η–transfected U2OS cells were treated with ETO (2.5 μM) and further incubated for 24 h. The cells were then fixed and DAPI-stained. Scale bars: 3 μm. B, the proportions of GFP-Pol η–expressing cells with more than 30 foci were determined. Data represent means ± SEM from three independent experiments. C, representative images of cells stained with DAPI and 53BP1 after ETO treatment. ShNC-treated and shPol η–treated MRC5 cells were treated with 50 μM ETO for 30 min and further cultured. At the indicated time points, cells were treated with 0.5% Triton-X100 for 10 min followed by immunofluorescence. Scale bars: 10 μm. D, quantification of the number of 53BP1 foci in (C). For each cell line at each time point, at least 100 cells were measured. E, quantification of the percentage of cells with more than 10 53BP1 foci. The upper panels show immunoblots indicating the Pol η levels. Tubulin: loading control. For each cell line at each time point, at least 200 cells were counted. Data represent means ± SEM from three independent experiments. F, representative images of cells costained with 53BP1 and cyclin A after ETO treatment. ShNC- and shPol η-MRC5 cells were treated with 50 μM ETO for 30 min and further cultured for 30 h, followed by immunofluorescence. G1 represents cyclin A–negative cells. Scale bars: 5 μm. DAPI, 4, 6-diamidino-2-phenylindole; ETO, etoposide; NHEJ, nonhomologous end joining.

Figure 2

Pol η deficiency leads to NHEJ repair defect and cellular hypersensitivity to ETO.A, Pol η shRNA–treated cells complemented with Flag-Pol η or not were transfected with NHEJ reporter (I-SceI endonuclease encoding plasmid and pDsRed2-N1). NHEJ repair efficiency (GFP⁺/DsRed⁺) were analyzed by FACS after 96 h. The lower panels show the Pol η levels through immunoblotting. Tubulin: loading control. B, the recruitment of Pol η to ETO-induced DSBs is independent of its catalytic activity. The proportions of GFP-Pol η WT- or CI (catalytic inactive) mutant-expressing cells with more than 30 foci were determined. Data represent means ± SEM from three independent experiments. C, the role of Pol η in NHEJ is independent of its catalytic activity. EJ5 cells overexpressing Flag-vector, or Flag-Pol η WT, or CI were transfected with I-SceI endonuclease. NHEJ repair efficiency (GFP⁺) was analyzed by FACS after 48 h. The lower panels show immunoblots indicating the Pol η levels in different conditions. D, MRC5 (NC, siPol η, and siKu80), XPV, and GFP-Pol η–complemented XPV (XPV-GFP-Pol η) cells were treated with indicated ETO and further incubated for 7 to 10 days. The number of clones was determined. Surviving fraction was expressed as a percentage of mock-treated cells. Experiment was repeated three times, giving similar results. The representative curve is shown. Error bar: s.d., n = 3. E, immunoblots indicating the Pol η and Ku80 levels in (D). Tubulin: loading control. DSB, double-strand break; ETO, etoposide; FACS, fluorescence activated cell sorting; NHEJ, nonhomologous end joining.

Figure 3

Kap1 is a novel Pol η-interacting protein. A, list of proteins identified by mass spectrometry analysis. Triton-insoluble fractions of HEK293T cells expressing Flag-Pol η were lysed and subjected to tandem-affinity purification. Flag-Pol η immunoprecipitated proteins were resolved by SDS-PAGE followed by mass spectrometry analysis. B, Kap1 and DNA-PKcs associate with Pol η. HEK293T cells transfected with Flag-Pol η were lysed and immunoprecipitated with anti-Flag M2 beads, followed by immunoblotting with anti-DNA-PKcs, anti-Kap1, anti-PARP1, anti-Ku70, and anti-Flag antibodies. C, purified GST or GST-Kap1 proteins were incubated with His-Pol η protein. The bound proteins were resolved by SDS-PAGE and analyzed by immunoblotting with anti-His antibody and staining with Ponceau S. D, HEK293T cells expressing full-length (FL) or deletion mutants of Flag-Pol η (K1-K7) were lysed and immunoprecipitated with Flag M2 agarose beads followed by immunoblotting with anti-Kap1 and anti-Flag antibodies.

Figure 4

Kap1 depletion impairs the recruitment of Pol η to laser- and ETO-induced damage sites.A, U2OS cells were transfected with siKap1 or siNC oligos. Forty-eight hours later, the cells were transfected with GFP-Pol η followed by microirradiation. The proportion of cells with Pol η accumulation was quantified. Data represent means ± SEM from three independent experiments. B, the protein levels of Kap1 of cells in (A) were detected by Western blotting. β-Actin: loading control. C, Kap1 stable knockdown (shKap1-UTR) or negative control (shNC) cells expressing GFP-Pol η were complemented with Myc-Kap1 or not. Twenty-four hours later, the cells were exposed to ETO (2.5 μM) for 24 h. The proportions of GFP-Pol η–expressing cells with more than 30 foci were determined by counting at least 200 cells in each experiment. Data represent means ± SEM from three independent experiments. D, the protein levels of Kap1 and overexpressed Myc-Kap1 in (C) were detected by Western blotting. Tubulin: loading control. E, U2OS cells were transfected with FL (1–713) or truncated (301–713, 556–713 and 1–512) GFP-Pol η; 24 h later, the cells were exposed to ETO (2.5 μM) for 24 h. The proportions of GFP-Pol η–expressing cells with more than 30 foci were determined by counting at least 200 cells in each experiment. Data represent means ± SEM from three independent experiments. F, HEK293T cells transfected with FL or truncated (301–713, 556–713 and 1–512) GFP-Pol η were lysed and immunoprecipitated with anti-GFP agarose beads, followed by immunoblotting with anti-Kap1 and anti-GFP antibodies. ETO, etoposide.

Figure 5

Kap1 regulates Pol η focus formation in a Rad18-dependent manner after ETO treatment.A, HEK293T cells expressing GFP-Rad18 and Flag-Pol η were treated with ETO (50 or 100 μM) for 2 h and further incubated for 2 h. The cells were lysed and immunoprecipitated using anti-Flag M2 agarose beads followed by Western blot using anti-GFP and anti-Flag antibodies. B, Rad18 wildtype (WT) or knockout (KO) U2OS cells transfected with GFP-Pol η and Myc-Rad18 or Myc-Kap1 were treated with ETO (2.5 μM) for 24 h. The proportion of GFP-Pol η–expressing cells with more than 30 foci was determined. Data represent means ± SEM from three independent experiments. C, the Rad18, Myc-Rad18, and Myc-Kap1 levels in (B) were detected by Western blot. Tubulin: loading control. D, representative images of cells stained with Rad18 after ETO treatment. U2OS cells were transfected with siKap1 or siNC oligos. Seventy hours later, the cells were treated with ETO (100 μM) for 3 h followed by immunostaining with anti-Rad18 antibody. Scale bar: 10 μm. E, HEK293T cells expressing Myc-Rad18 were lysed and immunoprecipitated using anti-Myc agarose beads. The immunoprecipitates were examined via Western blot using anti-Kap1 and anti-Myc antibodies. F, HEK293T cells were transfected with Myc-Kap1 and wildtype GFP-Rad18 or mutant GFP-Rad18-△PID (lacks Pol η–interacting domain). The cells were lysed and immunoprecipitated using anti-GFP agarose beads followed by Western blot using anti-Myc and anti-GFP antibodies. ETO, etoposide; SE, short exposure; LE, long exposure.

Figure 6

Kap1 forms a ternary complex with Rad18 and Pol η.A, HEK293T cells were transfected with Myc-Rad18 and WT GFP-Pol η or GFP-Pol η-301-713 (lacks Kap1-interacting domain). The cells were lysed and immunoprecipitated using anti-GFP agarose beads followed by Western blot using anti-Myc, anti-Kap1, and anti-GFP antibodies. B, the protein levels of Kap1 and Myc-Kap1 in (C) were detected by Western blotting. Tubulin: loading control. C, HEK293T cells were transfected with siKap1 or siNC oligos. Forty-eight hours later, the cells were transfected with Flag-Pol η, GFP-Rad18, and Myc-Kap1 or not. The cells were lysed and immunoprecipitated using anti-GFP agarose beads followed by Western blot using anti-Flag, anti-Kap1, and anti-GFP antibodies. D, the protein levels of Rad18 in (E) were detected by Western blotting. Tubulin: loading control. E, Rad18 WT or KO HEK293T cells were transfected with Flag-Pol η. The whole-cell lysates were immunoprecipitated with anti-Flag M2 agarose beads. The immunoprecipitates were immunoblotted with anti-Kap1 and anti-Flag antibodies. F, model of Pol η in ETO-induced DSB repair. Post ETO exposure, Kap1 promotes Rad18-dependent focus formation of Pol η by facilitating Rad18 recruitment and associating with Pol η, which forms a ternary complex to stabilize Rad18–Pol η association. DSB, double-strand break; ETO, etoposide; LE, long exposure; SE, short exposure.

Figure 7

Rad18 and Kap1 promote NHEJ repair, while Rad18 phosphorylation on S409/S434 is not required for Pol η recruitment post ETO.A, representative images of cells stained with DAPI and 53BP1 after ETO treatment and quantification of the percentage of cells with more than 10 53BP1 foci. Rad18 WT and KO cells were treated with 50 μM ETO for 30 min and further cultured. At the indicated time points, cells were treated with 0.5% Triton-X100 for 10 min followed by immunofluorescence. For each cell line at each time point, at least 200 cells were counted. Data represent means ± SEM from three independent experiments. Scale bars: 10 μm. B, siRad18-, siKap1-, or siKu70-treated EJ5 cells were transfected with I-SceI endonuclease. NHEJ repair efficiency (GFP⁺) was analyzed by FACS after 48 h. The upper panels show the immunoblots indicating the levels of Rad18, Kap1, and Ku70 in different conditions. Tubulin: loading control. C, Rad18 WT or KO U2OS cells transfected with GFP-Pol η and Myc-Rad18 (WT or S409A/S434A mutant) were treated with ETO (2.5 μM) for 24 h. The proportion of GFP-Pol η–expressing cells with more than 30 foci was determined. Data represent means ± SEM from three independent experiments. The lower panels show the immunoblots indicating the related Rad18 levels. Tubulin: loading control. ETO, etoposide; FACS, fluorescence activated cell sorting; NHEJ, nonhomologous end joining.