53BP1 mediates productive and mutagenic DNA repair through distinct phosphoprotein interactions

Abstract

The DNA damage response (DDR) protein 53BP1 protects DNA ends from excessive resection in G1, and thereby favors repair by nonhomologous end-joining (NHEJ) as opposed to homologous recombination (HR). During S phase, BRCA1 antagonizes 53BP1 to promote HR. The pro-NHEJ and antirecombinase functions of 53BP1 are mediated in part by RIF1, the only known factor that requires 53BP1 phosphorylation for its recruitment to double-strand breaks (DSBs). Here, we show that a 53BP1 phosphomutant, 53BP18A, comprising alanine substitutions of the eight most N-terminal S/TQ phosphorylation sites, mimics 53BP1 deficiency by restoring genome stability in BRCA1-deficient cells yet behaves like wild-type 53BP1 with respect to immunoglobulin class switch recombination (CSR). 53BP18A recruits RIF1 but fails to recruit the DDR protein PTIP to DSBs, and disruption of PTIP phenocopies 53BP18A. We conclude that 53BP1 promotes productive CSR and suppresses mutagenic DNA repair through distinct phosphodependent interactions with RIF1 and PTIP.

Figure 1. Characterization of a separation of function mutant 53BP1

(A) Top: Representative flow cytometry plots measuring CSR after stimulation of WT and BRCA1^−/−53BP1^−/− B cells infected with retroviruses expressing 53BP1^DB (amino acids 1-1710), the N-terminal mutant 53BP1^8A or empty vector (EV). Numbers represent the percentages of IgG1 switched cells. B220 is a B cell marker. Bottom: Dot plot indicating IgG1 CSR as a percentage of WT value in the same experiment. Three independent experiments are shown. **: p<0.001 (two-tailed unpaired t-test); BRCA1/53BP1+DB vs. BRCA1/53BP1+8A, p>0.1 which is not significant (ns). (B) BRCA1^−/−53BP1^−/− B cells were reconstituted with empty vector, 53BP1^DB and 53BP1^8A retroviruses and treated with PARPi. The arrows indicate representative images of aberrant chromosomes. Dot plot indicates the total aberrations per cell in three independent experiments. At least 50 metaphases were analyzed for each genotype in each experiment. **: p<0.01 (two-tailed unpaired t-test);ns: not significant. (C) Western blot analysis of 53BP1 expression in WT B cells and BRCA1^−/−53BP1^−/− B cells stimulated and infected with empty vector, 53BP1^DB or 53BP1^8A. (D) BRCA1^−/−53BP1^−/− B cells infected with EV, 53BP1^DB or 53BP1^8A retroviruses were assayed for IRIF (10 Gy, 2 hour recovery) for RIF1 (red, top panel) and 53BP1 (red, lower panel). Cells were counterstained with DAPI (blue). Scale bar represents 10 μm. See also Figure S1.

Figure 2. Response of PTIP to different DNA damaging agents

(A) 53BP1^−/− B cells were reconstituted with empty vector, 53BP1^DB, 53BP1^8A, or 53BP1^28A retroviruses that were FLAG-tagged. Cells were irradiated (10 Gy, 45 minute recovery) and immunoprecipitation was performed with anti-FLAG antibodies. Western blot analysis of PTIP and FLAG are shown for input (left panel) and immunoprecipitated protein (right panel). (B) Isogenic immortalized WT and PTIP^−/− MEFs were either untreated or treated with irradiation (IR, 2 Gy), cisplatin (CisPt, 0.5 μM), camptothecin (CPT, 10nM) or PARP inhibitor (PARPi, 1μM) and chromosomal aberrations (chromatid breaks, chromosome breaks and radials) were quantified in at least 50 metaphase spreads for each genotype and each treatment. Data from an independent experiment is shown in Figure S3. (C) WT (green lines) and PTIP^−/− (blue line) MEFs were treated with different doses of the above drugs and colony formation was quantified relative to colonies formed in untreated cells from the same genotype. An experiment performed in parallel demonstrated that 1 μM PARPi treatment is toxic for BRCA1-mutant MEFs (red line). See also Figures S2-S5.

Figure 3. PTIP is dispensable for CSR to IgE

PTIP^f/fCD19^CRE (PTIP^−/−), (Rif1^f/fCD19^CRE) RIF1^−/− and littermate WT B cells were stimulated with αCD40 plus IL-4 and analyzed for IgG1 and IgE CSR on day 5. (A) Representative flow cytometry plots. The percentages of IgG1 switched cells (upper left quadrant) and IgE switched cells (lower right quadrant) is indicated. (B) Dot plot indicates IgG1 and IgE CSR in PTIP^−/− and RIF1^−/− as a percentage of the WT value in the same experiment. **:p<0.01 (two-tailed unpaired t-test);ns: not significant.

Figure 4. PTIP is required for NHEJ of dysfunctional telomeres

(A) WT and PTIP^−/− MEFs were infected with a retrovirus expressing either an empty vector or shRNA against TRF2 (shTRF2), and phosphorylated KAP1 (pKAP1) levels were measured by flow cytometry. (B) γ-H2AX (green) in telomere-dysfunction induced foci (TIF) generated in shTRF2-infected WT cells. PNA probe is shown in red, and images are merged on top of DAPI (blue). Scale bar represents 10 μm. (C) Representative images of a metaphase spread from WT and PTIP^−/− MEFs infected with shTRF2. Telomere fusions are visualized by a telomeric PNA probe (red) and DAPI (blue). Arrows point to representative telomeric fusions. (D) Quantitation of telomeric fusion frequencies. At least 1800 chromosomes from each genotype were analyzed. Mean value derived from 3 independent experiments. **: p<0.01 (two-tailed unpaired t-test). (E) Distribution of telomeric fusions per metaphase in WT and PTIP^−/− MEFs. At least 30 cells were examined in each of 3 independent experiments. p(chi-squared)<1×10⁻⁵.

Figure 5. Ablation of PTIP rescues homologous recombination in BRCA1 deficient cells

(A) WT, BRCA1^−/−, PTIP^−/− and BRCA1^−/−PTIP^−/− B cells were pulsed with CFSE and stimulated with (red) or without (green) PARPi. CFSE signal diminishes with increasing division. BRCA1^−/− cells are sensitive to PARPi (arrow indicates sluggish cells) but loss of PTIP in BRCA1-deficient cells rescues the proliferation defect. (B) Analysis of genomic instability (radial chromosomes, chromatid breaks and chromosome breaks) in metaphases from B cells treated with 1 μM PARPi. At least 50 metaphases were analyzed for each genotype. (C) B cells were stimulated for 2 days, irradiated with 10 Gy and the percentage of cells with immunofluorescent RAD51 foci were quantified (at least 400 cells counted for each genotype). Data in (B) and (C) represent mean of three experiments +/− standard deviations. **: p<0.05 (two-tailed unpaired t-test), ns: not significant. (D) High-throughput microscopy quantification of RPA foci per cell in WT and PTIP^−/− MEFs that were either untreated or treated with 30 Gy IR. Top: representative image of chromatin bound RPA in irradiated WT and PTIP^−/− cells. Bottom: quantitation of RPA foci. Bar indicates the mean number of RPA foci per cell, and the blue box designates cells with more than 15 foci, whose percentage is indicated above each box. **: p<0.001 (E) BRCA1^−/−PTIP^−/− B cells were reconstituted with PTIP^WT or PTIP^W663R retroviruses (expressing a GFP marker driven by an internal ribosome entry site) and treated with PARPi. Cells were sorted (GFP^positive =infected and GFP^negative =uninfected) and metaphases were analyzed for radial chromosomes (n=50 metaphases analyzed in each case). See also Figures S6-S9.

Figure 6. Recruitment of PTIP to DSBs is ATM and phospho-53BP1 dependent but RIF1-independent

(A) WT, 53BP1^−/− and ATM^−/− MEFs were infected with a FLAG-tagged WT PTIP retrovirus. Cells were irradiated with 10 Gy, and FLAG (red) IRIF together with γ-H2AX (green) were assessed four hours post-IR. DAPI is indicated in blue. (B) WT, 53BP1^−/−, ATM^−/− and 53BP1^−/− MEFs reconstituted with 53BP1^28A were treated with Hoecsht 33342 and then irradiated with a 364 nm laser line. Cells were allowed to recover for 15 minutes before processing for immunfluorescence analysis of PTIP and γ-H2AX. Hoechst counterstain is indicated in blue. (C) Cells expressing GFP-PTIP were irradiated with 10 Gy, and PTIP^GFP (green) and RIF1 (red) IRIF were assessed four hours later. Representative image is shown. 82% of PTIP IRIF colocalized with RIF1 foci and 78% of RIF1 colocalized with PTIP foci (n≥800 foci examined. Cells had on average 28 foci). (D) RIF1 IRIF (red) in irradiated WT and PTIP^−/− MEFs. (E) RIF1 (red) and PTIP (red) recruitment to laser scissors damage in WT and RIF1^−/− MEFs. Damaged cells are indicated by γ-H2AX tracks (green). Scale bars represents 10 μm. See also Figure S10.

Figure 7. PTIP and RIF1 association with DSBs is dependent on distinct phosphorylation sites on 53BP1

(A) 53BP1^−/− MEFs (reconstituted with 53BP1^DB, 53BP1^8A, 53BP1^7A, or 53BP1^15A were co-stained with RIF1 (red) and PTIP (green). Scale bar represents 10 μm. (B) Quantitation of percent 53BP1^DB, 53BP1^8A, 53BP1^7A or 53BP1^15A cells with greater than ten 53BP1, PTIP or RIF1 foci. At least 100 cells were analyzed for each genotype. (C) Integrated intensity of individual IRIF in 53BP1^−/− MEFs reconstituted with DB or 7A. Average RIF1 foci intensity (red line) is 1.6 fold greater in DB vs. 7A (** p<0.001, one-tailed unpaired t-test), and a greater percentage of very intense foci (z-score>3) are generated in 53BP1^DB compared to 53BP1^7A (blue box). (D) Model for regulation of 53BP1 pro-NHEJ and anti-HR activities by distinct phospho-interactions with RIF1 and PTIP respectively. PTIP binds to the 8S/TQ sites. RIF1 recruitment is largely dependent on C-terminal 7S/TQ sites, but RIF1 may also be stabilized by interactions with 8S/TQ. An unknown factor (X) may bind directly to phosphorylated 53BP1 and mediate RIF1 recruitment, whereas PTIP interaction with 53BP1 is direct (Munoz et al., 2007). See also Figures S11-S13.