MRN complex-dependent recruitment of ubiquitylated BLM helicase to DSBs negatively regulates DNA repair pathways

Abstract

Mutations in BLM in Bloom Syndrome patients predispose them to multiple types of cancers. Here we report that BLM is recruited in a biphasic manner to annotated DSBs. BLM recruitment is dependent on the presence of NBS1, MRE11 and ATM. While ATM activity is essential for BLM recruitment in early phase, it is dispensable in late phase when MRE11 exonuclease activity and RNF8-mediated ubiquitylation of BLM are the key determinants. Interaction between polyubiquitylated BLM and NBS1 is essential for the helicase to be retained at the DSBs. The helicase activity of BLM is required for the recruitment of HR and c-NHEJ factors onto the chromatin in S- and G1-phase, respectively. During the repair phase, BLM inhibits HR in S-phase and c-NHEJ in G1-phase. Consequently, inhibition of helicase activity of BLM enhances the rate of DNA alterations. Thus BLM utilizes its pro- and anti-repair functions to maintain genome stability.

Fig. 1

BLM is recruited to AsiSI-induced DSBs. a BLM and γH2AX colocalize after DSB induction. U2OS–AsiSI–ER cells were grown asynchronously (−4-OHT) or in the presence of +4-OHT for 4 h. Cells were fixed and stained with anti-BLM and anti-γH2AX antibodies. Nucleus was stained by DAPI. Bar, 5 μM. b BLM is recruited to sequence-specific DSBs. U2OS–AsiSI–ER cells were grown asynchronously (−4-OHT) or in the presence of +4-OHT for 4 h. ChIP was carried out using anti-BLM antibody. Recruitment of BLM to the indicated AsiSI-induced DSBs or GAPDH loci were determined by ChIP-qPCR carried out using primers, which were either proximal or distal to the AsiSI sites. c Recruitment of BLM to the DSBs is helicase independent. Same as b, except BLM recruitment to the indicated AsiSI-induced DSBs was determined in U2OS–AsiSI–ER cells grown either in the absence or presence of ML216. 4-OHT treatment was for 4 h. d Biphasic BLM recruitment occurs at both proximal and distal positions with respect to the AsiSI site. Same as b, except BLM recruitment to the indicated AsiSI-induced DSB was determined after 4-OHT treatment for 0.5, 1, 2, and 4 h. ChIP was carried out using anti-BLM antibody. Recruitment of BLM to the indicated AsiSI-induced DSBs or GAPDH loci were determined by ChIP-qPCR. Distance from AsiSi site at which recruitment was measured was approximately 0.08, 0.8, 3, 6, and 9 kb. e BLM is co-recruited with NBS1 but not with DNA repair proteins to the DSBs in early phase of recruitment. Same as b except U2OS–AsiSI–ER cells were treated with 4-OHT for 0.5 h. Parallel ChIPs were carried out with anti-BLM, anti-NBS1, anti-XRCC4, and anti-RAD51 antibodies. Recruitment of BLM, NBS1, XRCC4, and RAD51 to the indicated DSBs or to the GAPDH loci was determined by carrying out ChIP-qPCR analysis. For all ChIP-qPCR analysis, the depicted values (mean ± standard deviation) were obtained from four independent experiments. Data were analyzed by unpaired two-tailed Student’s t-test

Fig. 2

Recruitment of BLM at the DSBs. a BLM co-immuoprecipitates with MRN complex and ATM after DSB generation. (Top) Extracts were made from U2OS–AsiSI–ER cells (±4-OHT, 4 h). Lysates were probed with antibodies against BLM (only A300-110A), ATM, pSer1981 ATM, MRE11, RAD50, NBS1, pSer343 NBS1, hsp90. (Bottom) Immunoprecipitations (±4-OHT, 4 h) were carried out with anti-BLM antibody or the corresponding IgG and the immunoprecipitates were probed with the above-mentioned antibodies. b, c Recruitment of BLM to the DSBs is dependent on the presence of NBS1, MRE11, and ATM in both 0.5 h and 4 h post induction of DSBs. ChIP was carried out using anti-BLM antibody on chromatin obtained from U2OS–AsiSI–ER cells grown in the presence of 4-OHT for either 0.5 or 4 h after transfection with either b Control siRNA or NBS1 siRNA or MRE11 siRNA, or c control siRNA or ATM siRNA. BLM recruitment to the indicated AsiSI-induced DSB or GAPDH loci were determined by ChIP-qPCR after either 0.5 or 4 h of 4-OHT treatment. d Recruitment of BLM to DSBs partially depends on MRE11 exonuclease activity. ChIP was carried out using anti-BLM antibody on chromatin obtained from U2OS–AsiSI–ER cells grown in presence of 4-OHT for either 0.5 or 4 h. In both the time points, the cells were grown either in +4-OHT or in +4-OHT + Mirin conditions. BLM recruitment to the indicated AsiSI-induced DSBs or GAPDH loci were determined by ChIP-qPCR. e Recruitment of BLM to DSBs depends on ATM kinase activity only in the early phase. ChIP was carried out using anti-BLM antibody on chromatin obtained from U2OS–AsiSI-ER cells grown in the presence of 4-OHT for either 0.5 or 4 h. In both the two time points, the cells were grown in +4-OHT or +4-OHT + KU 55933 conditions. BLM recruitment to the indicated AsiSI-induced DSBs or GAPDH loci were determined by ChIP-qPCR. For all ChIP-qPCR analyses, the depicted values (mean ± standard deviation) were obtained from four independent experiments. Data were analyzed by unpaired two-tailed Student’s t-test

Fig. 3

Ubiquitylated BLM at the DSBs interacts with NBS1. a, b RNF8-dependent polyubiquitylation of BLM is required for its interaction with NBS1. (Top) a, b In vitro ubiquitylated BLM WT or BLM 3K was probed with anti-BLM antibody. (Middle) a, b Exogenous NBS1, immunoprecipitated from HEK293T cells (±NCS), was probed with anti-Myc tag antibody. (Bottom) a, b Interaction carried out with either polyubiquitylated or non-ubiquitylated BLM and bound Myc-tagged NBS1 was detected with anti-BLM antibody. c Ubiquitylation is essential for BLM to interact with NBS1 in vivo. HCT116 BLM (−/−) cells (+NCS) were transfected with either EGFP-C1 BLM (WT) or (3K). (Top) Lysates were probed with antibodies against GFP, NBS1, hsp90. (Bottom) Lysates were immunoprecipitated with antibody against NBS1 and probed with anti-NBS1 and anti-GFP antibodies. d Lack of RNF8 prevents the interaction between BLM and NBS1. (Left) U2OS–AsiSi–ER cells were transfected with siControl or siRNF8. Extracts were probed with anti-RNF8, anti-BLM, anti-NBS1, anti-hsp90 antibodies. (Right) Immunoprecipitations, carried out with either anti-BLM antibody or IgG, were probed with anti-BLM, anti-NBS1 antibodies. e Polyubiquitylated BLM is complexed with NBS1 but not MRE11. (Top) RNF8-dependent in vitro ubiquitylation of BLM WT, in the presence of Ub WT or Ub with all lysines mutated (Ub KO), was detected using anti-BLM antibody. (Middle) Exogenous Myc-tagged NBS1 and MRE11 were immunoprecipitated and detected with anti-Myc tag antibody. (Bottom) Interactions were carried out with either ubiquitylated or non-ubiquitylated BLM with immunoprecipitated Myc-tagged NBS1 or MRE11 and detected with anti-BLM antibody. f Recruitment of BLM to DSBs is abrogated in the presence of Mirin and siRNF8. ChIP was carried out using anti-BLM antibody on chromatin obtained from U2OS AsiSI-ER cells (4 h, 4-OHT). BLM recruitment to the indicated AsiSI-induced DSBs or GAPDH loci were determined by ChIP-qPCR. g, h Formation of BLM foci are abrogated in presence of Mirin and siRNF8. Experiments were done in same conditions as in f. g Immunofluorescence was carried out with anti-BLM antibody. Nucleus was stained with DAPI. Bar, 5 μM. h Quantitation of BLM foci formed under different conditions. Data were analyzed by unpaired two-tailed Student’s t-test. Values mean ± standard deviation

Fig. 4

Cell cycle-dependent recruitment of HR and c-NHEJ factors is dependent on the helicase activity of BLM. a–c BLM, RAD51, and XRCC4 are co-recruited to the DSBs in S-phase. U2OS AsiSI were synchronized in S-phase (4 h post release from double thymidine block). Chromatin was prepared from these S-phase cells grown without (−4-OHT) or after 4 h of 4-OHT treatment. ChIP was carried out using a anti-BLM antibody; b anti-RAD51 antibody; c anti-XRCC4 antibody. BLM, RAD51, and XRCC4 recruitment to the indicated AsiSI-induced DSBs were determined by ChIP-qPCR. d–f BLM and XRCC4 are co-recruited to the DSBs in G1-phase. U2OS–AsiSi-ER cells were synchronized in G1-phase (18 h post release from double thymidine block). Chromatin was prepared from these G1-phase cells grown without (−4-OHT) or after 4 h of 4-OHT treatment. ChIP was carried out using d anti-BLM antibody; e anti-RAD51 antibody; f anti-XRCC4 antibody. Recruitment to the indicated AsiSI-induced DSBs or to the GAPDH loci were determined by ChIP-qPCR. g, h Helicase function of BLM is required for the cell cycle-dependent recruitment of RAD51 and XRCC4 to the DSBs. U2OS–AsiSI–ER cells were synchronized in either g S-phase or h G1 phase. Cells in either of the phases were subjected to 4 h of 4-OHT treatment, carried out either in the absence or presence of ML216. Parallel ChIPs were carried out with antibodies against RAD51 and XRCC4. The recruitment of RAD51 and XRCC4 to the indicated AsiSI-induced DSBs or to the GAPDH loci was determined by ChIP-qPCR. For all ChIP-qPCR analyses, the depicted values (mean ± standard deviation) were obtained from four independent experiments. Data were analyzed by unpaired two-tailed Student’s t-test. i, j Helicase function of BLM is required for the cell cycle-dependent recruitment of HR and c-NHEJ factors to the chromatin. U2OS–AsiSI–ER cells were grown as in g, h. Soluble and chromatin fractions were isolated from cells and western analysis carried out with antibodies against BLM, HR factors (RAD51, RAD54, MRE11), c-NHEJ factors (Ku70, Ku80, Ligase IV, XRCC4), fractionation controls (hsp90 for soluble fraction and SMC1 for chromatin fraction). In each case, a Coomassie gel is included to demonstrate that equal amount of protein was taken for the analysis

Fig. 5

BLM negatively regulates both HR and c-NHEJ during repair phase. a Lack of functional BLM enhances HR in S-phase. HR assays were carried out using extracts from S- or G1-phase-arrested U2OS–AsiSI-ER cells grown in −4-OHT, +4-OHT/PW, +4-OHT/PW + ML216, +4-OHT/PW + B02, +4-OHT/PW + SCR7 conditions. Fold change in recombination frequency is represented with respect to the control where extract was not added. The entire experiment was done three times. Statistical analysis was done by Student’s t-test. b Lack of functional BLM enhances total NHEJ using intrachromosomal substrates. NHEJ substrate (pJS296), I-SceI expression vector (pJS20) and mCherry2-C1 plasmids were co-transfected in asynchronously growing U2OS–AsiSI–ER cells grown in the absence or presence of ML216. The percentage of GFP-expressing cells was scored by FACS analysis. The experiment was done six times, data are presented as mean ± standard deviation and the statistical analysis was done by Student’s t-test. c, d Lack of functional BLM enhances c-NHEJ in both S- and G1-phase. c-NHEJ assays were carried out using extracts, which were prepared as in a. c Representative autoradiogram is shown. d Percentage efficiency of joint product formation is represented. The product formed after ML216 treatment in G1-phase is taken as 100% and all other products are calculated relative to the above condition. Data are presented as mean ± standard deviation and the statistical analysis was done by Student’s t-test. e Lack of functional BLM enhances nucleotide change at AsiSI sites. Genomic DNA was isolated from G1-phase-arrested U2OS AsiSI-ER cells grown in +4-OHT/PW, +4-OHT/PW + ML216 conditions. Thirty-five independent clones for each of the four AsiSI junctions were sequenced for both conditions. The percentage of nucleotide changes involving deletions, insertions and deletions with insertions are represented. Data were obtained from three independent experiments and represented as mean ± standard deviation, were analyzed by unpaired two-tailed Student’s t-test

Fig. 6

Mechanism of BLM recruitment to the DSBs and its effect on DNA repair. In the recruitment phase, MRN complex is recruited to the AsiSI-induced DSBs (a), which in turn recruits ATM (b). Chromatin-bound E3 ligase RNF8 polyubiquitylates BLM (red line, red dots representing ubiquitin residues). The kinase activity of ATM (brown lines) is essential for the BLM recruitment process, acting directly on BLM or via NBS1/MRE11. Single-stranded region obtained due to the exonuclease activity of MRE11 (green line) is also an essential requirement for BLM recruitment. Polyubiquitylated BLM interacts with NBS1 (blue line), which allows BLM to indirectly yet functionally interact with MRE11, leading to the optimal recruitment of BLM (c). The helicase activity of BLM is essential for the co-recruitment of c-NHEJ protein XRCC4 in G1-phase and RAD51 in S-phase (black lines) (d). During repair phase, BLM predominantly inhibits c-NHEJ in G1-phase and HR in S-phase. As a backup regulatory mechanism, BLM also inhibits c-NHEJ in S-phase (e). Hence BLM switches from a pro-repair protein (due to its role in the recruitment phase) to an anti-repair regulatory protein (due to its role in the repair phase), thereby maintaining genomic stability