TOPORS-mediated RAD51 SUMOylation facilitates homologous recombination repair

Abstract

Homologous recombination (HR) is critical for error-free repair of DNA double-strand breaks. Chromatin loading of RAD51, a key protein that mediates the recombination, is a crucial step in the execution of the HR repair. Here, we present evidence that SUMOylation of RAD51 is crucial for the RAD51 recruitment to chromatin and HR repair. We found that topoisomerase 1-binding arginine/serine-rich protein (TOPORS) induces the SUMOylation of RAD51 at lysine residues 57 and 70 in response to DNA damaging agents. The SUMOylation was facilitated by an ATM-induced phosphorylation of TOPORS at threonine 515 upon DNA damage. Knockdown of TOPORS or expression of SUMOylation-deficient RAD51 mutants caused reduction in supporting normal RAD51 functions during the HR repair, suggesting the physiological importance of the modification. We found that the SUMOylation-deficient RAD51 reduces the association with its crucial binding partner BRCA2, explaining its deficiency in supporting the HR repair. These findings altogether demonstrate a crucial role for TOPORS-mediated RAD51 SUMOylation in promoting HR repair and genomic maintenance.

Figure 1.

TOPORS interacts with RAD51. (A) HeLa cells, with or without 5 Gy of IR or 2 mM HU treatment, were lysed and subjected to immunoprecipitation followed by immunoblotting as indicated. (B) HEK293T cells transfected with HA-RAD51 along with or without GFP-TOPORS were treated with 5 Gy of IR and subjected to immunoprecipitation and immunoblots as indicated. Asterisk indicates nonspecific band. (C)In vitro pull-down assay carried out using immobilized control His or His-RAD51 fusion proteins on Ni⁺-NTA Agarose followed by recombinant GST-TOPORS protein. GST pull-downs were immunoblotted with antibodies as indicated. (D) HeLa cells, treated with or without 5 Gy of IR, were stained with indicated antibodies. Colocalization of TOPORS (red) and PML (green) is visible as a yellow merged signal, and colocalization of TOPORS (red) and γ-H2AX (cyan) is visible as a white merged signal. Nuclei were stained with DAPI. Scale bar: 5 μm. The percent colocalization between TOPORS and PML and between TOPORS and γ-H2AX is shown. Quantification of merged foci as a percentage of TOPORS–PML or TOPORS–γ-H2AX foci from total TOPORS foci. At least 50 cells from three independent slides were analyzed. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01. (E) Whole cell lysates of HeLa cells, with or without 5 Gy of IR treatment, immunoprecipitated using an anti-TOPORS antibody followed by immunoblotting using indicated antibodies. (F) Percent colocalization of RAD51 and TOPORS foci at the indicated times following IR treatment. Data are presented as mean ± SD (n = 3). (G) A PLA probe was used to detect colocalization of endogenous RAD51–TOPORS. DNA were counterstained with DAPI. The negative controls were obtained by omitting one of the primary antibodies. Representative images and a scatterplot of the PLA signal per nucleus are shown. Scale bar: 30 μm. Data are presented as mean ± SD (n =3). P-values between indicated samples were calculated using a Mann–Whitney test.

Figure 2.

TOPORS is phosphorylated at Thr515 in response to IR. (A) Schematic diagrams of the TOPORS protein domains with putative ATM phosphorylation sites indicated. (B) Determination of IR-induced phosphorylation sites in TOPORS by mass spectrometry. A peptide containing Thr515 phosphorylation is shown. HeLa cells transfected with control siRNA or ATM siRNA were treated with or without 5 Gy of IR (C) or 2 mM HU (D). Whole cell lysates were then subjected to immunoprecipitation using an anti-TOPORS antibody followed by immunoblotting using indicated antibodies. (E) HeLa cells transfected with control siRNA or ATR siRNA were treated with or without 2 mM HU. Whole cell lysates were then subjected to immunoprecipitation using an anti-TOPORS antibody followed by immunoblotting using indicated antibodies. (F) HEK293T cells transfected with control GFP vector, GFP-TOPORS-WT or GFP-TOPORS-T515A were treated with IR (5 Gy). Cell lysates were immunoprecipitated with anti-GFP antibody and subjected to immunoblot analysis with anti-phospho-threonine antibody. Asterisk indicates degradation products of GFP-TOPORS. (G) HA-RAD51-expressing HEK293T cells transfected with control GFP vector, GFP-TOPORS-WT or GFP-TOPORS-T515A were treated with 5 Gy of IR and subjected to immunoprecipitation and immunoblots as indicated. Asterisk indicates degradation products of GFP-TOPORS.

Figure 3.

TOPORS plays an important role in DSB repair. Control and TOPORS-depleted U2OS cells (A) or control and TOPORS^−/− MEF cells (B) were exposed to 5 Gy of IR and fixed at the indicated time points. Immunostaining experiments were performed using an anti-RAD51 antibody. DNA was counterstained with DAPI. Scale bar: 10 μm. Percentage of cell populations that shows >5 foci for RAD51 is shown. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01. (C) A schematic of the DR-GFP reporter system used to measure rates of HR. (D) HR efficiency as measured by FACS in control and TOPORS-depleted U2OS/DR-GFP cells. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01. (E) HR efficiency of U2OS/DR-GFP cells transfected with the indicated siRNA combinations. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01, ns = not significant. (F) IR-induced DNA damage as measured through a neutral comet assay in control and TOPORS-depleted U2OS cells. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01. (G) The number of chromosome aberrations as measured by metaphase chromosome spreads of control and TOPORS-depleted U2OS cells treated with 2 Gy of IR. Representative images and quantification of aberrations are shown. The results are shown as mean ± SD (n = 3). P-values between the indicated samples were calculated using a Mann–Whitney test. (H) Array CGH profiles of genomic DNA derived from TOPORS^+/+ and TOPORS^−/− MEFs. Genomic positions that fall above or below the dotted line indicate amplifications or deletions of regions of genome, respectively. (I) Colony forming ability of control and TOPORS-depleted U2OS cells treated with the indicated doses of IR. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01.

Figure 4.

RAD51 is SUMOylated both in vitro and in vivo. (A) Control and TOPORS-depleted HeLa cells were treated with or without 5 Gy of IR. Immunoprecipitations using the anti-RAD51 antibody were performed and the following immunoblot analyses were done using anti-SUMO1 and anti-SUMO2/3 antibodies. HC indicates heavy chain. (B) HeLa cells transfected with control, TOPORS, PIAS1 and PIAS4, PIAS1, PIAS4 or RAD51 siRNA were irradiated with 5 Gy of IR, immunoprecipitated with anti-RAD51 antibody and subjected to immunoblot analysis with anti-SUMO1 antibody. HC indicates heavy chain. (C) HEK293T cells were cotransfected with HA-RAD51, GFP-TOPORS, His-SUMO1 and Ubc9, treated with IR (5 Gy) or HU (2 mM), and subjected to immunoprecipitation and immunoblotting as indicated. Asterisks indicate nonspecific bands. (D) HEK293T cells were transfected with His-SUMO1, His-SUMO2 or His-SUMO3 along with the indicated plasmids, exposed to 5 Gy of IR and subjected to immunoprecipitation followed by immunoblotting as indicated. HC indicates heavy chain. (E) HEK293T cells transfected with control or Flag-Senp2 along with the indicated plasmids were exposed to 5 Gy of IR and subjected to immunoprecipitation and immunoblotting as indicated. Asterisk indicates nonspecific band. (F) His-tagged RAD51 proteins incubated with purified GST-tagged TOPORS, recombinant SUMO1/2/3 and/or SUMO E1/E2 as indicated and analyzed by immunoblotting using anti-His and anti-GST antibodies.

Figure 5.

RAD51 SUMOylation is required for HR-mediated DSB repair. (A) TOPORS knockdown HEK293T cells were transfected with GFP-TOPORS-WT or GFP-TOPORS (T515A) along with the indicated plasmids, exposed to 5 Gy of IR, and immunoprecipitated and immunoblotted as indicated. Asterisks indicate nonspecific bands. (B) TOPORS knockdown HeLa cells reconstituted with GFP-TOPORS-WT or GFP-TOPORS (T515A or T515E) were treated with 5 Gy of IR, fixed at 3 h and immunostained using an anti-RAD51 antibody. The percentage of cell populations that shows >10 foci for RAD51 in GFP-positive cells is shown. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01, ns = not significant. Asterisk indicates degradation products of GFP-TOPORS. (C) γ-H2AX foci of the same cells as described in (B). Cells were treated with 5 Gy of IR, fixed at 24 h and immunostained using an anti-γ-H2AX. The percentage of cell populations that shows >10 foci for γ-H2AX in GFP-positive cells is shown. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01, ns = not significant. Asterisk indicates degradation products of GFP-TOPORS. (D) HR efficiency of TOPORS knockdown U2OS/DR-GFP cells reconstituted with TOPORS-WT or TOPORS (T515A). The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01, ns = not significant. (E) A schematic of the domains of the human RAD51 including two putative SUMOylation sites and the SUMO-interacting motif (SIM). (F) HEK293T cells transfected with HA-RAD51-WT or mutants together with indicated plasmids were exposed to 5 Gy of IR. Whole cell lysates were analyzed by immunoprecipitation followed by immunoblotting as indicated. Asterisk indicates nonspecific band and HC indicates heavy chain. (G) An amino acid sequence alignment of the predicted consensus SUMO site with K57 and K70 highlighted in red. (H) HeLa cells expressing HA-RAD51-WT or HA-RAD51-K57R/K70R were treated with 5 Gy of IR, fixed at 3 h and immunostained using an anti-HA antibody. Nuclei were stained with DAPI. Scale bar: 10 μm. The percentage of cell populations that shows >5 foci for RAD51 is shown. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01. (I) Quantification of DNA damage through a neutral comet assay in control, RAD51 knockdown and RAD51 knockdown cells expressing indicated constructs after treatment with IR (5 Gy) at the indicated time points. Scale bar: 200 μm. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01, ns = not significant. (J) The efficiency of HR repair, measured using the assay depicted in Figure 3C, in RAD51-depleted DR-GFP-U2OS cells reconstituted with the indicated constructs. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01, ns = not significant. (K) RAD51 knockdown HeLa cells reconstituted with HA-RAD51-WT and HA-RAD51-K57R/K70R were treated with 2 Gy of IR. The number of chromosome aberrations was measured by metaphase chromosome spreads. The results are shown as mean ± SD (n = 3). P-values between the indicated samples were calculated using a Mann–Whitney test; ns = not significant. (L) Colony forming ability of the same cells as described in (J). Cells were treated with the indicated doses of IR. The results are shown as mean ± SD (n = 3), ^∗∗P < 0.01, ns = not significant.

Figure 6.

SUMOylation of RAD51 promotes its interaction with BRCA2. (A) Control and TOPORS-depleted HeLa cells were treated with or without 5 Gy of IR. Immunoprecipitations using an anti-RAD51 antibody were performed and the following immunoblot analyses were done using anti-BRCA2 or anti-RAD51 antibodies. (B) TOPORS knockdown HEK293T cells reconstituted with GFP-TOPORS-WT or GFP-TOPORS (T515A) were treated with or without IR (5 Gy), and subjected to immunoprecipitation and immunoblotting as indicated. RAD51 knockdown HeLa cells reconstituted with HA-RAD51-WT and HA-RAD51-K57R/K70R were treated with or without 5 Gy of IR (C) or 2 mM HU (D), and subjected to immunoprecipitation and immunoblotting as indicated. (E) Schematic model representing the role of TOPORS in RAD51 function in DSBs and replication stress. See text for details.