RAP80-directed tuning of BRCA1 homologous recombination function at ionizing radiation-induced nuclear foci

Abstract

In response to DNA double-strand breaks (DSBs), BRCA1 forms biochemically distinct complexes with certain other DNA damage response proteins. These structures, some of which are required for homologous recombination (HR)-type DSB repair, concentrate at distinct nuclear foci that demarcate sites of genome breakage. Polyubiquitin binding by one of these structures, the RAP80/BRCA1 complex, is required for efficient BRCA1 focal recruitment, but the relationship of this process to the execution of HR has been unclear. We found that this complex actively suppresses otherwise exaggerated, BRCA1-driven HR. By controlling the kinetics by which other BRCA1-interacting proteins that promote HR concentrate together with BRCA1 in nuclear foci, RAP80/BRCA1 complexes suppress excessive DSB end processing, HR-type DSB repair, and overt chromosomal instability. Since chromosomal instability emerges when BRCA1 HR function is either unbridled or absent, active tuning of BRCA1 activity, executed in nuclear foci, is important to genome integrity maintenance.

Figure 1.

BRCA1 IRIF formation in the presence and absence of RAP80. U2OS cells treated with control or RAP80 siRNA were irradiated with 5 Gy IR or mock-treated (designated as NT) and released for different periods of time. (A,B) Representative images showing BRCA1 and RAP80 IRIF at 2 h (A) and 8 h (B) post-IR. Nuclei were counterstained with DAPI. (C,D) Percentages of cells, noted above, that contain RAP80 (C) and BRCA1 (D) IRIF. At the indicated intervals, cells containing at least 10 distinct foci, as recognized by relevant antibodies in immunofluorescence experiments, were denoted and counted. At least 100 cells were counted for each category of foci at each time point. The experiment was repeated three times and error bars indicate standard deviation. Data in D were analyzed using two-way ANOVA and, when a significant difference between control and RAP80-depleted cells was present, the actual significance values were marked. (***) P < 0.001. (E) Distribution of control and RAP80-depleted cells in five distinct categories based on intensity of BRCA1 IRIF (class I–class V) at 4 h after 5 Gy IR. A representative cell is shown for each class. All images were obtained at the same magnification and exposure time. During the analysis, each nucleus in a given culture was characterized as belonging to one of the above-noted cell classes. At the completion of the experiment, the distribution of nuclei among these classes in control and RAP80-depleted cultures was assessed and then compared. Bars, 10 μm. Numbers of cells in each class and total numbers of cells counted, as well as their corresponding percentages in control or RAP80-depleted cells, are summarized in the table. All images were analyzed in parallel for each experiment, and the data represented in the table are that obtained, collectively, from five identical experiments.

Figure 2.

Dynamics of IRIF formation. U2OS cells were treated as described in Figure 1. At the indicated intervals, cells containing at least 10 distinct foci, as recognized by relevant antibodies in immunofluorescence assays, were identified and counted. Experiments were repeated three times, and 80–150 cells were counted for each category of foci at each time point for each experiment. (A) Percentages of cells containing CtIP foci. (B) Percentages of cells containing BRCA1 foci with colocalized CtIP. (C) Percentages of cells containing BACH1 foci. (D) Percentages of cells containing BRCA1 foci with colocalized BACH1. (E) Percentages of cells containing RAD51 foci. (F) Percentages of cells containing BRCA1 foci with colocalized RAD51. (G) Percentages of cells containing γH2AX foci. (H) Percentages of cells containing BRCA1 foci with colocalized γH2AX foci. Data were analyzed using two-way ANOVA and, when a significant difference between control and RAP80-depleted cells was present, the actual significance values were marked. (***) P < 0.001; (**) P < 0.01; (*) P < 0.5. Please see also Supplemental Figure S2.

Figure 3.

RAP80 depletion leads to an increased frequency of homology-mediated DSBR. (A) A schematic diagram summarizing the I-SceI-inducible DSBR reporter in U2OS-DR cells and the relevant HR measurement procedure. (B) A histogram showing the relative change in GFP-positive cells after I-SceI induction in cells transfected with control siRNA or siRNA targeting BRCA1, BACH1, CtIP, RAP80 (siRAP80-1), Abraxas, or BRCC36. The percentage of GFP-positive cells after I-SceI expression in control siRNA-treated samples was normalized to 1. Data were collected from three independent experiments. Error bars indicate standard deviation. (C,D) Rescue of increased HR in RAP80-depleted cells by expression of an siRNA-resistant RAP80 cDNA. (C) Histogram showing the relative change of I-SceI-induced HR in cells treated with control or RAP80 siRNA and cotransfected with an siRNA-resistant RAP80 cDNA (Res cDNA) or the vector DNA. The levels of GFP-positive cells in vector + control siRNA cotransfected cells were normalized to 1. Experiments were performed in triplicate, and error bars indicate standard deviation. (D) Western blots showing the efficiency of endogenous RAP80 depletion by siRNA and expression of a Myc-tagged RAP80 cDNA resistant to siRNA. Actin served as a loading control for Western blots, and the expression of a Myc-tagged GFP cDNA (cloned in the same vector) in parallel cell cultures served as controls for transfection efficiency. Please note the total amounts of protein loaded in lanes 3 and 4 (cf. actin bands) were less than those in lanes 1 and 2 in order to distinguish endogenous and tagged RAP80 using the same RAP80 antibody. (E,F) The excessive HR-mediated DSBR in RAP80-depleted cells is BRCA1- and RAD51-dependent. (E) HR-mediated repair of I-SceI-induced DSB. Experiments were performed as described in A. Specifically, U2OS-DR cells expressing an shRNA that targets luciferase (shLuc) or RAP80 (shRAP80) were transfected with control siRNA or an siRNA directed at BRCA1 or RAD51. The histogram shows the relative increase in GFP-positive cells of each category. The levels of GFP-positive cells in shLuc-expressing cells were normalized to 1. Experiments were performed in triplicate, and error bars indicate standard deviation. (F) Western blots reveal the efficiency of protein depletion by RAP80 shRNA and/or siRNA directed at BRCA1 or RAD51. Fifteen micrograms of protein was loaded in each lane, and actin (detected by blotting) served as a loading control.

Figure 4.

RAP80/BRCA1 complexes are required to suppress excessive DSB end processing. (A) Representative images showing IR-induced ssDNA foci in control, RAP80, Abraxas, or CtIP siRNA-treated U2OS cells. Cells were immunostained with an anti-BrdU antibody under nondenaturing conditions 2 h after 5 Gy IR treatment. Nuclei were counterstained with DAPI. Bars, 10 μm. (B) A histogram showing the percentage of cells containing >10 nuclear BrdU/ssDNA foci. Experiments were performed in triplicate, and at least 100 cells were counted in each experiment for each treatment. Error bars indicate standard deviation. (C) A histogram showing the average number of nuclear BrdU/ssDNA foci in cells containing 10 or more such foci. At least 120 nuclei were counted for each treatment. Error bars indicate standard deviation. (D) A schematic diagram showing the locations of four sets of primers used in the ChIP assays (marked as P1, P2, P3, or P4). The associated numbers correspond to the numbers of base pairs (bp) from the I-SceI recognition site to the position of a relevant 5′ primer. The genomic region that contains the sequences in which HR is expected is marked in green. (E) Results of RPA ChIP assays. (F) Results of RAD51 ChIP assays. The extent of RPA or RAD51 antibody binding at each location was calculated by comparing each antibody binding results with binding results obtained at the same DNA location with irrelevant IgG and normalized with the amounts of input DNA. Levels of RPA or RAD51 binding were all normalized to that obtained at position P1 in control cells (i.e., in the absence of RAP80 knockdown and I-SceI expression). (G) Representative images showing the recruitment/accumulation of RAD51 to UV laser-induced DSB stripes in cells treated with control or RAP80 siRNA. Bars, 10 μm. (H,I) Histograms showing the percentage of γ-H2AX-positive stripes in control siRNA or RAP80 siRNA-treated cells that were stained with RAD51 (H) or RAP80 (I) antibody, respectively. Data were collected from two independent experiments with at least 60 stripes scored in each experiment. Error bars indicate standard deviation.

Figure 5.

Rap80 depletion results in increased frequencies of chromosomal translocations. (A) A schematic diagram showing the Alu translocation reporter in HomAlu mouse ES cells and the relevant experimental procedure. The centromere of mouse chromosome 17 or 14 is indicated as a black dot. (Hygro) Hygromycin resistance gene; (Hprt) hypoxanthine phosphoribosyltransferase gene; (Alu) two identical Alu elements derived from the MLL gene; (Neo-SD and SA-Neo) two portions of the neomycin/G418 resistance gene with splice donor and splice acceptor; (3′ Puro and 5′ Puro) 3′ and 5′ portions of the puromycin resistance gene. Gray boxes adjacent to 5′ and 3′ Puro indicate homologous sequences from the puromycin cassette. Locations of PCR primers for analyzing translocation products and the predicted sizes of such PCR products are also indicated. (B) A histogram showing the results of Alu recombination/translocation assays. Relative frequencies of translocations after I-SceI expression were compared with that of control siRNA-treated samples (normalized to 1). Translocations in vector transfected (i.e., no I-SceI) samples were undetectable (i.e., frequency <5 × 10⁻⁷) and are not shown in the histogram. Data were collected from two independent experiments in which each assay was performed in triplicate. (C) Results of quantitative RT–PCR showing the efficiency of Rap80 siRNA treatment. The amount of Rap80 mRNA in control siRNA-treated cells was normalized to 100. (D) A partial gel image of PCR analysis on translocation products in genomic DNA derived from G418-resistant clones. The size of PCR products correspondent to NHEJ- or SSA-mediated translocation is indicated. At least 100 clones were analyzed each for cells treated with control, siRap80-1, or siRap80-2 siRNA. Only 12 PCR results are shown for control (lanes 1–6) and siRap80-1-treated (lanes 7–12) cells.

Figure 6.

Increased chromosomal instability after DNA damage in RAP80-depleted cells. (A) Representative metaphase spreads showing SCEs from U2OS cells transfected with control or RAP80 siRNA. Cells were treated with 200 nM etoposide (ETO) or DMSO for 24 h. Bars, 5 μm. (B) A dot graph depicting the number of SCEs in cells treated with control, RAP80, or BRCC36 siRNA. Each data point represents the total number of SCEs in an individual cell. Horizontal lines indicate the mean number of SCEs in each category of cells. (C) Representative metaphase spreads from cells treated with control or RAP80 siRNA. Arrows point to multiradial chromosomal structures. (D) A histogram summarizing percentages of spreads containing multiradial chromosomal structures 24 h after cells were exposed to 2 Gy IR or mock-treated.

Figure 7.

A model for HR-tuning function by RAP80/BRCA1-containing complexes present in IRIF. The main subunits of the RAP80 complex (i.e., RAP80, Abraxas, BRCC36, and BRCA1) are designated. Symbols for polyubiquitin and a generic P-Ser/Thr residue bound at the BRCA1 BRCT motifs are also included. The three HR-supporting BRCA1 complexes are also depicted, along with the hypothesized suppression by RAP80 complexes of their concentration in IRIF and their contribution to HR. The hatched circle symbolizes IRIF.