Microcephalin Regulates BRCA2 and Rad51-Associated DNA Double-Strand Break Repair

Microcephalin (MCPH1) is a member of the BRCA1 COOH terminal (BRCT) domain family of proteins that are involved in the cellular response to DNA damage (1, 2). MCPH1 contains a single NH₂ terminal and two COOH terminal BRCT domains. Biallelic mutations in the MCPH1 gene are associated with primary microcephaly (ref. 3; OMIM 251200) and premature chromosome condensation (PCC) syndrome (refs. 4, 5; OMIM 606858). Mutations in other DNA damage response genes, including Nijmegen breakage syndrome (NBS1; ref. 6), ataxia-telangiectasia (7), and ataxia-telangiectasia and Rad3-related protein (ATR; ref. 8) are also associated with microcephaly, suggesting that MCPH1 participates in the NBS1-associated, ATM-associated, and ATR-associated DNA damage response signaling pathways. Indeed, MCPH1 is rapidly recruited to nuclear foci after DNA damage (9). The localization of MCPH1, along with MDC1 (9), 53BP1, and γH2AX (10–12), to sites of DNA double-strand breaks is dependent on an interaction between the COOH terminal BRCT domains of MCPH1 and γH2AX (11, 12) but does not seem to involve MDC1 or 53BP1 (11). MCPH1-deficient cells display defects in intra-S and G₂-M phase cell cycle checkpoint activation in response to irradiation resulting in nuclear fragmentation, chromosomal instability and centrosome amplification (9, 10, 13, 14).

Other domains of MCPH1 also seem to have roles in mediating chromosome instability. Specifically, the single NH₂ terminal BRCT domain in MCPH1 is required for centrosome localization of the protein (12) and also has been implicated in rescue of the PCC phenotype observed in MCPH1-deficient cells from individuals with primary microcephaly (15). In addition, a central domain of MCPH1 that interacts with Condensin II through the CAPG2 subunit is required for homologous recombination activity in response to DNA double-strand breaks in mouse embryonic fibroblasts (MEF) reconstituted with MCPH1 (15).

Together, these findings suggest that MCPH1 is an important mediator of the response to DNA damage and maintenance of chromosomal stability. The recent observation that MCPH1 exhibits reduced expression in breast tumors and is associated with genomic instability and metastases (14) is in keeping with these results and suggests that MCPH1 may have a role in tumor suppression. Further studies aimed at fully understanding the mechanisms by which MCPH1 influences DNA damage signaling and repair may clarify the role of this protein in cancer.

Here, we investigate the role of MCPH1 in the DNA repair process at sites of DNA damage. We find that BRCA2 interacts with the COOH terminal BRCT domains of MCPH1 and that this interaction controls the enrichment of BRCA2 and Rad51 at sites of DNA double-strand breaks after irradiation. We show that MCPH1 does not influence formation of BRCA2-Rad51 complexes, suggesting that the interaction of MCPH1 with BRCA2 is required for recruitment and/or retention of these BRCA2-Rad51 complexes at sites of damage. We also verify that the interaction between MCPH1 and BRCA2-Rad51 complexes mediates BRCA2 and Rad51-dependent homologous recombination repair of DNA damage.

Plasmids and antibodies. FLAG-tagged BRCA2 fragments were cloned in pEV3S-FLAG3 vector. MCPH1 wild type and deletion mutants were cloned in pCS3-6xMyc vector. Glutathione S-transferase (GST)–tagged BRCA2 fragments (B2F1–B2F7) in pEGB and FLAG-tagged fragments (B2F1–B2F7) in pcDNA3.1 were provided by Dr. J. Chen (Yale University). The anti-MCPH1 rabbit polyclonal antibody was raised against residues 92 to 214 of MCPH1. The anti-BRCA2 rabbit polyclonal antibody was raised against the COOH terminal of BRCA2 (residues 2418–3201). The antisera were affinity purified with AminoLink Plus immobilization and purification kit (Pierce). Anti-γH2AX, 53BP1, MDC1, and Rad51 polyclonal antibodies were gifts from Dr. J. Chen (Yale University). Mouse anti-Flag (M2) antibody was purchased from Sigma. Mouse anti-HA antibody was purchased from Roche. Mouse anti-Myc antibody was purchased from Santa Cruz.

Short interfering RNA. Short interfering RNAs (siRNA) against MCPH1 were synthesized by Dharmacon, Inc. The siRNA duplexes were as follows: MCPH1 siRNA1 sense strand, 5′C TCT CTG TGT GAA GCA CCA; MCPH1 siRNA2 sense strand, 5′AA GCT CAG AAG AGA GGC GT. BRCA2 siRNAs (SMARTpool) were purchased from Dharmacon. Transfections were performed with 200 nmol/L siRNA using FuGene-6 (Roche) according to the manufacturer's instructions.

Coimmunoprecipitation. 293T cells were purchased from American Tissue Type Culture and maintained in DMEM supplemented with 10% bovine serum and 1% penicillin/streptomycin at 37°C in 5% CO₂ (v/v). VU423 FANCD1 (FA-D1) and A913 423/2-33 BRCA2-reconstituted FANCD1 cells were grown under similar conditions (16). Cells were lysed with NETN buffer [20 mmol/L Tris-HCl (pH 8.0), 100 mmol/L NaCl, 1 mmol/L EDTA, 0.5% Nonidet P-40] containing protease inhibitors (Complete, Boehringer Mannheim Biochemicals, Inc.) 2 h after exposure to 10 Gy of irradiation or in the absence of irradiation (data not shown). Whole-cell lysates obtained after centrifugation were incubated with 2 μg antibody and Protein A or Protein G sepharose beads (Amersham Biosciences) for 2 h at 4°C. For GST pull downs of GST-tagged proteins, cleared cell lysates were incubated with glutathione agarose beads (Sigma).

Immunofluorescence. 293T cells cultured on glass coverslips were irradiated (10 Gy), incubated for 3 h, and analyzed for formation of DNA damage repair foci. Cells were fixed with cold methanol for 20 min, permeabilized in 0.05% Triton-X 100 for 15 min, incubated in blocking buffer (16) and incubated with primary antibodies diluted in blocking buffer for 1 h at room temperature. After washing in PBS, cells were incubated with Alexa Fluor 488 and 568–conjugated goat anti-mouse and goat anti-rabbit IgG secondary antibodies (Molecular Probes). Cells were costained with 4′,6-diamidino-2-phenylindole. After a further wash in PBS, coverslips were mounted in Prolong (Molecular Probes). Cells were visualized and imaged using a Zeiss LSM510 confocal microscope. For each experiment, 100 nuclei transfected with expression constructs or siRNAs were evaluated. Effects on foci were consistently observed in >80% of transfected cells in each experiment.

Homologous recombination repair assay. The homologous recombination repair assay was carried out as described previously (17). Briefly, the efficiency of homology-directed repair was assessed using an I-SceI expression plasmid (pcBASce) and an I-SceI repair reporter plasmid (DR-GFP) composed of two differentially mutated green fluorescent protein (GFP) genes, one of which contains a unique I-SceI restriction site. Here, HeLa cells or V-C8 BRCA2 null hamster lung fibroblast cells that were stably transfected with a single copy of DR-GFP (16) were cotransfected with I-SceI and BRCA2 plasmids or MCPH1 siRNAs, and the number of GFP-expressing cells was assessed by flow cytometry after 48 h. Transfection efficiency was defined by immunofluorescence analysis of FLAG-tagged BRCA2 constructs and was used to normalize homologous recombination repair levels. Statistical significance relative to wild-type BRCA2 or vector control was determined using paired t tests and Fisher's protected least significant difference pair-wise comparison procedure.

MCPH1 interacts with BRCA2. Whereas it has been noted that MCPH1 is one of the first signaling proteins recruited to nuclear foci after DNA damage, there is also evidence that MCPH1 functions downstream of Chk1 and BRCA1 (13) and that MCPH1 is retained at sites of damage beyond the initial recognition stage (12). This suggests that MCPH1 contributes to DNA repair at multiple levels. To elucidate the involvement of MCPH1 in the response to DNA double-strand breaks, we explored potential interactions between MCPH1 and components of DNA damage signaling and repair pathways. Extracts from 293T cells ectopically expressing Myc-tagged MCPH1 were exposed to 10 Gy of irradiation and screened for interactions between MCPH1 and a series of proteins involved in the DNA repair process. We reproducibly found that myc-tagged MCPH1 coimmunoprecipitated with BRCA2 and Rad51 (Fig. 1A and B). In separate experiments, a significant amount of endogenous MCPH1 associated with endogenous BRCA2 and Rad51 (Fig. 1C and D). These interactions were also evident in the absence of irradiation. Together, these results suggest that MCPH1 interacts with BRCA2-Rad51 complexes in the presence and absence of DNA damage.

The MCPH1 BRCT domains are required for interaction with the NH₂ terminus of BRCA2. To identify regions of BRCA2 that are responsible for the MCPH1 interaction, we generated a series of constructs encoding fragments of BRCA2 (Fig. 2A; Supplementary Table S1). The fragment B2F1, encompassing BRCA2 residues 1 to 472, accounted for the interaction between BRCA2 and MCPH1 (Fig. 2A). Deletion mutants of B2F1 mapped the MCPH1 interaction region to residues 234 to 335 (B2F12; Fig. 2B). Further deletion analysis within B2F12 showed that residues 294 to 335 in B2F12Δ6 were sufficient for the interaction (Fig. 2C). However, deletion of residues 293 to 323 in B2F12.3 disrupted the interaction (Supplementary Fig. S1), as did deletion of residues 325 to 335 in B2F12ΔR6, B2F12ΔR7, and B2F12ΔR8 (Supplementary Table S1; Fig. 2C). The combined results suggest a complex MCPH1 interaction site located between residues 294-335 of BRCA2. We also generated deletion mutants of MCPH1 that lacked each of the three BRCT domains (Fig. 2D; Supplementary Table S1) and assessed their ability to interact with BRCA2. GST-tagged B2F1 failed to interact with MCPH1 in the absence of the MCPH1 BRCT2 domain (residues 642–720; Fig. 2D). In contrast, deletion of the BRCT1 (residues 1–96) and BRCT3 domains (residues 753–823) had no effect on the interaction (Fig. 2D). These findings indicate that a previously undefined BRCA2 NH₂ terminal domain and the MCPH1 BRCT2 domain are necessary for the interaction of BRCA2 with MCPH1.

Localization of BRCA2 at sites of DNA damage is dependent on MCPH1. Whereas MCPH1 is one of the first proteins recruited to sites of radiation-induced DNA damage, the interaction of MCPH1 with BRCA2 and Rad51 suggests that MCPH1 may function at multiple levels within the DNA damage signaling pathway. Here, we focused on determining the influence of MCPH1 on distal DNA damage signaling pathways. Specifically, we evaluated whether MCPH1 influenced the presence of BRCA2 and Rad51 at sites of DNA damage. As expected, MCPH1 was rapidly recruited to DNA damage foci after irradiation, where it colocalized with 53BP1, γH2AX (Supplementary Fig. S2a and c), and Rad51 (Fig. 3A) in all cells examined. However, when MCPH1 was depleted by siRNAs (Supplementary Fig. S2b and c), a substantial reduction in the number and intensity of Rad51 foci (Fig. 3A) and BRCA2 foci (Supplementary Fig. S2d) was consistently observed. In contrast, the MCPH1 siRNAs had little influence on recruitment of 53BP1 (Fig. 3A; Supplementary Fig. S2c) and MDC1 (Supplementary Fig. S2d) to foci, suggesting that the proximal DNA damage response pathway remained intact. This is consistent with recent findings that MCPH1 functions in an γH2AX-dependent but MDC1-independent DNA damage response pathway (11). Together, these results indicate that MCPH1 is an important mediator of the BRCA2-associated and Rad51-associated double-strand DNA break repair process.

To determine whether the influence of MCPH1 on DNA damage repair foci is dependent on the interaction of MCPH1 with BRCA2 and Rad51, we assessed the effect of MCPH1 mutants with deletions of the NH₂ terminal BRCT1 and COOH terminal BRCT2 and BRCT3 domains on radiation foci. Absence of the NH₂ terminal BRCT domain had no effect on localization of MCPH1, BRCA2, or Rad51 to DNA damage foci (Fig. 3B). This is consistent with the proposed involvement of this MCPH1 domain in centrosome regulation and PCC but not DNA repair (12, 15). In contrast, the BRCT2 and BRCT3 deletion mutants did not localize to foci but instead displayed a dispersed nuclear and cytoplasmic expression pattern (Fig. 3B; Supplementary Fig. S3). The absence of the BRCT2 and BRCT3 deletion mutants from the foci had no effect on recruitment of 53BP1 and γH2AX (Supplementary Fig. S3) but substantially reduced the number and intensity of BRCA2 and Rad51 foci (Fig. 3B). These results indicate that MCPH1 can influence the DNA damage response signaling pathway distal to γH2AX and are consistent with a direct effect of MCPH1 on the recruitment and/or retention of BRCA2 and Rad51 at the DNA repair foci.

MCPH1 does not influence formation of BRCA2-Rad51 complexes. To determine the influence of MCPH1 on the established interaction between BRCA2 and Rad51, we generated a deletion mutant (BRCA2-Δ328-351) of full-length BRCA2 (Supplementary Table S1) that disrupts part of the MCPH1 interaction site defined by the deletion mapping studies (Fig. 2C and D). Similarly to BRCA2 siRNA, this mutant had no influence on the localization of MCPH1, MDC1, and 53BP1 to sites of DNA damage (Fig. 3C; Supplementary Fig. S4a and b) but substantially reduced the number and intensity of BRCA2 and Rad51 foci (Fig. 3C; Supplementary Fig. S4b). Likewise, the B2F12 fragment that contains the MCPH1 interaction site and binds to MCPH1 inhibited the localization of MCPH1, BRCA2, and Rad51 to DNA damage foci (Supplementary Fig. S5a) in a dominant negative manner. In contrast, the B2F12ΔR6 construct, in which the MCPH1 interaction domain is disrupted, had no influence on MCPH1, BRCA2, or Rad51 focus formation (Supplementary Fig. S5B). These results suggest that the presence of Rad51 and BRCA2 at DNA damage foci is dependent on the interaction between BRCA2 and MCPH1.

Importantly, we also found that the BRCA2-Δ328-351 deletion mutant retained the ability to coimmunoprecipitate with Rad51 after DNA damage (Fig. 3D). Similarly, deletion of the MCPH1, BRCT2, and BRCT3 domains that mediate the interaction with BRCA2 and the presence of BRCA2 and Rad51 at foci had no effect on the BRCA2-Rad51 complex (Fig. 3D). We further explored this effect using BRCA2-deficient FANCD1 cells and FANCD1 cells reconstituted with wild-type BRCA2. After DNA damage, MCPH1 coimmunoprecipitated with Rad51 from the BRCA2 reconstituted cells, but not from the BRCA2-deficient cells (Fig. 3D), suggesting that the Rad51 interaction with MCPH is dependent on BRCA2. Together, these results indicate that MCPH1 interacts with and controls the presence of the BRCA2-Rad51 complex at sites of DNA damage.

MCPH1 influences BRCA2-dependent homologous recombination repair. Given the direct involvement of the BRCA2-Rad51 complex in homologous recombination repair of DNA double-strand breaks, these results suggest that MCPH1 may influence homologous recombination repair. We evaluated the influence of MCPH1 on homologous recombination using an in vivo assay (18–22) that depends on repair of a double-strand break at a unique I-Sce1 restriction endonuclease site (Fig. 4A). Depletion of MCPH1 from HeLa cells by siRNA significantly reduced homologous recombination repair activity relative to controls (P = 0.002), suggesting that MCPH1 levels influence this form of DNA damage repair (Fig. 4B). To determine whether the interaction of MCPH1 with BRCA2 accounts for the influence of MCPH1 on homologous recombination, we evaluated a series of BRCA2 deletion constructs (Fig. 2; Supplementary and Table S1) using the in vivo assay. Expression of B2F12 and B2F12Δ5, which contain the MCPH1 interaction domain, dominant-negatively inhibited homologous recombination activity (P = 0.002; Fig. 4C). In contrast, three BRCA2 fragments (B2F12ΔR6, B2F12ΔR7, and B2F12ΔR8) that do not interact with MCPH1 did not significantly disrupt homologous recombination activity (Fig. 4C). To confirm these effects in the context of full-length BRCA2, the influence of the BRCA2-Δ328-351 deletion mutant on homologous recombination activity relative to wild-type BRCA2 and a known deleterious missense mutant (D2723H) was assessed in BRCA2-null V-C8 cells containing the DR-GFP construct. Consistent with the results described above, the BRCA2-Δ328-351 deletion mutant caused a partial but statistically significant reduction in activity relative to wild-type BRCA2 (P = 0.009) similarly to D2723H (P = 0.001; Fig. 4D). These data indicate that the interaction between MCPH1 and BRCA2 mediates the homologous recombination repair activity of the BRCA2-Rad51 complex.

Although several studies suggest that MCPH1 is one of the first proteins recruited to sites of irradiation-induced foci where it is thought to mediate the response to DNA damage (10–14), little is known about how MCPH1 influences DNA repair. Recent studies have suggested that MCPH1 controls the cell cycle response to DNA damage by binding and mediating Chk1 and Cdc25a activity (13) and by regulating both the S and G₂-M checkpoints. In addition, the interaction between MCPH1 and Condensin II is apparently not associated with PCC but is required for homologous recombination repair of DNA double-strand breaks (15), suggesting that MCPH1 may have a role in the process of chromosome condensation and perhaps chromatin remodeling in response to DNA damage.

Here, we report that the presence of BRCA2-Rad51 complexes at sites of DNA double-strand breaks is dependent on an interaction between the NH₂ terminus of BRCA2 and the COOH terminal tandem-BRCT domains of MCPH1. Importantly, BRCA2 and Rad51 form complexes in the absence of MCPH1, albeit not at sites of DNA damage, suggesting that MCPH1 is required for the recruitment and/or retention of BRCA2-Rad51 complexes to repair foci. We also show that MCPH1 can regulate homologous recombination repair of DNA double-strand breaks in a BRCA2-dependent manner. These findings are consistent with the observation that MCPH1^−/− MEFs exhibit defects in homologous recombination repair of DNA damage (15). Thus, MCPH1 seems to contribute to homologous recombination repair through interactions with BRCA2 and the Condensin II complex (15).

Our findings extend our knowledge of MCPH1 and suggest that it forms an important link between the initial sensors and effectors of the DNA damage response, as defined by its rapid recruitment to DNA repair foci through an interaction with γH2AX, and the process of DNA repair, as established by its ability to regulate the localization of BRCA2-Rad51 complexes to sites of damage. Further efforts are needed to better understand how MCPH1 integrates these components of the DNA damage response signaling pathway.

Interestingly, it has recently been shown that Akt1 can repress homologous recombination through cytoplasmic retention of BRCA1 and Rad51 (23). It remains to be seen whether Akt1 signaling influences MCPH1-dependent recruitment and/or retention of BRCA2-Rad51 complexes at sites of damage. Furthermore, MEFs deficient in the SIRT1 histone deacetylase have been shown to exhibit significantly reduced recruitment of BRCA1, Rad51, and 53BP1 to DNA double-strand breaks and reduced homologous recombination activity (24). Given the role of SIRT1 in chromatin remodeling and the recent finding that SIRT1 localizes to DNA breaks to promote repair (25), further studies aimed at understanding the relationship between SIRT1 and components of the DNA signaling pathway, such as MCPH1 and BRCA2 are needed.

The ability of MCPH1 to regulate BRCA2 activity suggests that down-regulation or mutation of MCPH1 could lead to defects in DNA repair similar to those associated with BRCA2 inactivation. Importantly, down-regulation of MCPH1 has been observed in a number of tumors. Thus, MCPH1 may function as a tumor suppressor. Whereas mutation of both MCPH1 alleles is associated with PCC and primary microcephaly (5), mutation of a single allele may predispose or contribute to cancer development. Future studies aimed at detecting mutations in this gene and studies using knockout mice will test whether MCPH1 functions as a tumor suppressor in vivo. In addition, because MCPH1 seems to mediate BRCA2 function, it may be possible to use poly(ADP-ribose) polymerase inhibitors or DNA cross-linking agents that seem to be particularly effective against BRCA2-deficient cells to treat tumors with reduced levels of MCPH1.

No potential conflicts of interest were disclosed.

Grant support: NIH grants CA102701 and CA116167 (F.J. Couch).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.