CtIP is required to initiate replication-dependent interstrand crosslink repair

Abstract

DNA interstrand crosslinks (ICLs) are toxic lesions that block the progression of replication and transcription. CtIP is a conserved DNA repair protein that facilitates DNA end resection in the double-strand break (DSB) repair pathway. Here we show that CtIP plays a critical role during initiation of ICL processing in replicating human cells that is distinct from its role in DSB repair. CtIP depletion sensitizes human cells to ICL inducing agents and significantly impairs the accumulation of DNA damage response proteins RPA, ATR, FANCD2, γH2AX, and phosphorylated ATM at sites of laser generated ICLs. In contrast, the appearance of γH2AX and phosphorylated ATM at sites of laser generated double strand breaks (DSBs) is CtIP-independent. We present a model in which CtIP functions early in ICL repair in a BRCA1- and FANCM-dependent manner prior to generation of DSB repair intermediates.

Figure 1. CtIP depletion sensitizes cells to ICL inducing agents.

(A) Left, Survival of HEK293 cells transfected with control or CtIP siRNAs and exposed to 8-MOP+UVA or angelicin+UVA. CtIP_1 and CtIP_2 are two independent siRNAs. Cells were treated with indicated drug 48 hours post transfection and survival was assessed 8 days post irradiation. Fraction surviving cells is calculated in respect to untreated cells. Right, Survival of control or CtIP depleted cells treated 8 days following exposure to indicated concentration of MMC. Bars indicate standard deviation between 3 independent experiments. (B) γH2AX staining of fixed HeLa cells 2 hours post microirradiation with 730 nm laser light. S-phase synchronized cells were treated with indicated drugs prior to laser microirradiation. Black line on phase images indicates region of nucleus targeted by the laser. Thymidine arrested cells were microirradiated in the presence of 8-MOP. (C) Quantification of cells scored as having bright γH2AX staining along laser tracks. For details on cell analysis see Experimental Procedures. Bars indicate standard deviation between 3 independent experiments.

Figure 2. Effects of depletion of CtIP on cell cycle and BrdU incorporation.

(A) Cell cycle profiles of control and siCtIP transfected HeLa cells 48 hours after transfection. Labels mark the cell populations in the G1 (left peak), S (saddle), and G2 (right peak) phases of the cell cycle based on their DNA content. (B) Summary of cell cycle distributions in control and siRNA transfected HeLa cells. Percentages shown are averages of the results of 3 experiments ± the standard deviations. There was no significant difference between the cell cycle distributions of control and CtIP depleted cells (P≥0.2). (C) Immunoblot analysis confirming knockdown of CtIP by two independent siRNAs in 3 independent experiments. Protein extracts were prepared at 48 hours post transfection. Blot was probed with anti-CtIP antibody as indicated on left. A Ku86 was used as a loading control. (D) BrdU incorporation is unaffected in CtIP depleted cells. siRNA transfected HeLa cells were synchronized with thymidine, were released into S-phase 48 hours post transfection, and fixed after 20 minutes BrdU incorporation. Cells were stained for BrdU. Left, Images of fixed control and CtIP depleted cells following staining for BrdU. Right, Quantification of relative BrdU signal in control and CtIP depleted cells.

Figure 3. CtIP depletion reduces γH2AX at ICLs but not at DSBs.

(A) γH2AX staining of S-phase control and CtIP depleted cells 2 hours post microirradiation with 730 nm laser light in the presence of 8-MOP to form ICLs (above) or 532 nm laser light to form DSBs (below). (B) Quantification of cells scored as having bright γH2AX staining along laser tracks in control (light grey), CtIP siRNA_1(grey) and CtIP siRNA_2 (dark grey) depleted cells. Bars indicate standard deviation between 5 independent experiments.

Figure 4. CtIP depletion reduces ATM-pS1981 at ICLs but not at DSBs.

(A) ATM-pS1981 staining of S-phase control and CtIP depleted cells 2 hours post microirradiation with 730 nm laser light in the presence of 8-MOP to form ICLs (top) or 532 nm laser light to form DSBs (below). (B) Quantification of cells scored as having bright ATM-pS1981 staining in control (light grey) CtIP siRNA_1(grey) and CtIP siRNA_2 (dark grey) depleted cells. Bars indicate standard deviation between 3 independent experiments.

Figure 5. CtIP depletion reduces FANCD2 at ICLs.

(A) FANCD2 staining of S-phase control and CtIP depleted cells 2 hours post microirradiation with 730 nm laser light in the presence of 8-MOP. (B) Quantification of cells scored as having bright FANCD2 staining along laser tracks in control or CtIP depleted cells.

Figure 6. CtIP depletion reduces RPA and ATR accumulation at ICLs.

(A) RPA staining of S-phase control and CtIP depleted cells 2 hours post microirradiation with 730 nm laser light in the presence of 8-MOP. (B) Quantification of cells scored as having bright RPA staining intensity along laser tracks in control or CtIP depleted cells. (C) GFP-CtIP or GFP-CtIPT847A in S-phase cells pre pre-and post-microirradiation with 730 nm light in the presence of 8-MOP. Cells were stained for RPA2. (D) Quantification of RPA2 along laser tracks in cells scored positive for GFP-CtIP or GFP-CtIPT847A accumulation. (E) ATR staining of S-phase control and CtIP depleted cells 2 hours post microirradiation with 730 nm laser light in the presence of 8-MOP. (F) Quantification of cells scored as having bright ATR staining along laser tracks in control or CtIP depleted cells. Bars indicate standard deviation between 3 independent experiments.

Figure 7. CtIP accumulation at ICLs is BRCA1–dependent.

GFP-CtIP in S-phase control and BRCA1 depleted cells pre-and post-microirradiation with 730 nm light in the presence of 8-MOP.

Figure 8. FANCM depletion reduces CtIP accumulation at ICLs but not at DSBs.

(A) Immunoblot of control cells and cells depleted for FANCM. (B) GFP-CtIP in S-phase control and FANCM depleted cells pre-and post-microirradiation with 730 nm light in the presence of 8-MOP. (C) Quantification of cells scored as having visible GFP-CtIP along laser tracks in control or depleted cells. Bars indicate standard deviation between 3 independent experiments. (D) GFP-CtIP in S-phase control and FANCM depleted cells pre- and post-microirradiation with green laser to generate DSBs. (E) Quantification of cells scored as having bright GFP-CtIP along DSB containing laser tracks. (F) Model of CtIP function in replication associated ICL repair. FANCM binds ICL stalled replication fork and remodels fork to enable CtIP and MRN access. BRCA1 ubiquitinates CtIP facilitating its localization to damaged chromatin. CtIP and MRN act to generate ssDNA at a stalled fork. RPA bound single stranded DNA activates ATR. ATR phosphorylates FANCD2 followed by FA core complex ubiquitination of the FANCI-FANCD2 complex. The monoubiquitinated FANCI-FANCD2 complex localizes to the damaged chromatin and facilitates ICL incision and downstream repair events. CtIP and MRN act again to resect ends, activate ATM, facilitate completion of repair by homologous recombination.