Breast Cancer Risk Is Associated with the Genes Encoding the DNA Double-Strand Break Repair Mre11/Rad50/Nbs1 Complex

The evolutionarily conserved Mre11-Rad50-Nbs1 (MRN) complex, consisting of proteins encoded by the genes Mre11, Rad50, and Nbs1, was recently shown to play a crucial role in DNA double-strand break (DSB) repair by recruiting the nuclear protein kinase ataxia telangiectasia mutated to DSB sites, leading to activation of this DNA repair network. Given the fact that carriers of defective mutation and polymorphic variants of ataxia telangiectasia mutated are at higher risk of developing breast cancer, we hypothesized a role of the MRN genes in determining breast cancer susceptibility. This hypothesis was examined both in a case control study of 559 breast cancer patients and 1,125 healthy women of single-nucleotide polymorphisms in Mre11, Rad50, and Nbs1 and by the in vivo detection of binding between Mre11 and BRCA1, encoded by the breast cancer susceptibility gene BRCA1. We were also interested in defining whether any association between MRN genes and breast cancer was modified by reproductive risk factors reflecting the level of estrogen exposure or susceptibility to estrogen exposure, as estrogen is known to initiate breast cancer development due to its metabolites causing DSB formation. Support for the hypothesis came from the observations that (a) one single-nucleotide polymorphism in Nbs1 was significantly associated with breast cancer risk, and a trend toward an increased risk of developing breast cancer was found in women harboring a greater number of putative high-risk genotypes of MRN genes (an adjusted odds ratio of 1.25 for each additional putative high-risk genotype; 95% confidence interval, 1.10-1.44); (b) this association between risk and the number of putative high-risk genotypes was stronger and more significant in women thought to be more susceptible to estrogen, i.e., those with no history of full-term pregnancy, those older (≥26 years of age) at first full-term pregnancy, or those having had fewer (<2) full-term pregnancies; the risk effect conferred by harboring a higher number of high-risk genotypes of MRN genes was more significant in women without a history of breast feeding; and (c) Mre11 and BRCA1 were shown to form a complex in vivo, and this interaction was increased by irradiation. This study supports the role of the MRN pathway in breast cancer development, further strengthening the suggestion that mechanisms regulating DSB repair may play a mutator role driving breast cancer pathogenesis. (Cancer Epidemiol Biomarkers Prev 2007;6(10):2024–32)

Double-strand breaks (DSB) are extremely cytotoxic DNA lesions, and cells have therefore developed an extensive array of responses that lead to damage repair, thus preventing cell death (1-4). The nuclear protein kinase ataxia telangiectasia mutated (ATM) is regarded as the primary activator of this network, and the recruitment of ATM to DNA DSBs is thought to be the critical step in its activation and function (5). However, although ATM has an affinity for DNA, this recruitment has been suggested to require and to be facilitated by specific partner proteins. Recently, the evolutionarily conserved Mre11/Rad50/Nbs1 (MRN) complex was implicated in ATM recruitment to DSBs (6-8). This complex is involved in the initial processing of DSBs due to its nuclease activity and DNA binding capability, which reside in the Mre11 protein and partially depend on the interaction of Mre11 with Rad50, which provides the energy source for the MRN complex (9, 10). However, in terms of the protein required for ATM recruitment, it has only recently been recognized that Nbs1 recruits activated ATM to sites of DNA damage, then promotes its phosphorylation and the triggering of subsequent steps in the DNA damage response (11). These results show that the MRN complex functions as a DNA DSB sensor upstream of ATM signaling. However, this critical role of the complex in initiating DSBs repair does not exclude its involvement in mechanisms downstream of ATM signaling in the intra-S or G₂-M cell cycle checkpoint in response to DNA damage, as suggested in earlier studies (9, 10). These functional links between ATM and the MRN complex help explain similarities in the clinical and cellular phenotypes associated with deficiency of each of the genes encoding these proteins. Hypomorphic mutations in Mre11 and Nbs1 result, respectively, in the human genetic instability disease, ataxia-telangiectasia–like disorder, and Nijmegen breakage syndrome (10, 12, 13). These single-gene disorders overlap with one another and with ataxia-telangiectasia, caused by ATM deficiency, all being characterized by neurologic abnormalities, radiosensitivity, impairment of the cellular response to DSBs, and genomic instability. The close relationship between ATM and the MRN complex prompted us to hypothesize a role of the MRN genes in determining breast cancer susceptibility. The rationale underlying this hypothesis is that carriers of either defective mutations or polymorphic variants of ATM are at higher risk of developing breast cancer (14, 15). Furthermore, we have adopted a model (i.e., the hide-then-hit hypothesis; ref. 16) to explain the lack of cancer predisposition in ataxia-telangiectasia–like disorder patients or no breast cancer phenotype observed in Nijmegen breakage syndrome patients and suggest that a disparate spectrum of disease phenotypes can be differently caused by mutated forms or hypomorphic/polymorphic variants of the same genes. This is because, in contrast to common genetic diseases, cancer formation requires an extended period of time for the essential genomic changes to accumulate. Genetic variants or low-penetrance alleles of DNA repair genes (such as Mre11/RAD50/NBS1 examined in the present study) may therefore have a chance to escape the lethality phenotype by not triggering obvious cell cycle surveillance, and the cells accumulate the necessary genomic instability leading to cancer development. In addition, the possibility of manifesting the tumorigenic phenotype depends not only on the joint effect of individual genes but also on the interaction between genes and risk factors. To test this hypothesis, we did this investigation based (a) on a case control study to estimate the breast cancer risk associated with harboring putative high-risk genotypes of MRN genes, (b) on an examination of the joint effect of MRN genotypes and well-established risk factors of breast cancer in determining cancer risk, and (c) on an in vivo study of the interaction between MRN and BRCA1, encoded by the breast cancer susceptibility gene BRCA1. The combination of these three lines of evidence provides an essential insight into the tumorigenic contribution of the MRN complex during breast cancer formation.

Study Population. This case control study is part of an ongoing cooperative study aimed at understanding the causes of breast cancer in Taiwan, which is characterized by low incidence (17), early tumor onset (18), hormone dependency (19, 20), and novel genomic alterations (21, 22). Because of the low incidence, which suggests an overall lower effect of common risk factors (23), and because of its homogenous genetic background (24), the Taiwanese population has certain advantages for studying the effects of subtle genetic variations, such as single nucleotide polymorphisms (SNP). Furthermore, the use of a genetically homogenous population (i.e., the Taiwanese population) reduces the chance of false positives due to population stratification (25, 26). The present study included 559 female breast cancer patients and 1,125 healthy female controls. All subjects gave their informed consent. The recruitment of both cases and controls and considerations regarding methodologic issues (such as study design, sampling scheme, and potential bias) have been described and addressed in detail previously (16, 19, 20, 25, 26).

Questionnaire. Two experienced research nurses were assigned to administer a structured questionnaire to both cases and controls. The information collected and the validity of this questionnaire have been addressed and confirmed in our previous studies (16, 19, 20, 25, 26).

Genotyping. Genomic DNA was extracted from the buffy coat isolated from whole-blood samples using a QIAamp DNA extraction kit (Qiagen, Inc.) following the manufacturer's protocol.

Using the current data from HapMap⁷

on haplotype blocks of MRN genes in the Chinese population, we selected SNPs in each block to detect genetic variation in these three candidate genes and five SNPs were selected for each MRN gene. These SNPs were chosen because they are evenly distributed throughout the entire genes, a total of 15 SNPs being genotyped (Mre11: rs535801, rs569143, rs601391, rs684507, and rs1061945; Nbs1: rs1805794, rs1805790, rs709816, rs1061302, and rs1063045; Rad50: rs2252775, rs2301713, rs3798134, rs2240032, and rs2244012). Because there have not been any reports of an association between genotypic and phenotypic changes in the SNPs of the MRN genes, these selected SNPs were used as markers to reflect possible linkage disequilibrium (LD) between themselves and different alleles of a gene of unidentified phenotypic variation. We used more than one SNP per gene to have an unbiased definition of the allelic and haplotypic statuses of each gene.

All SNPs were genotyped using a MassARRAY (SEQUENOM, Inc.). The PCR primers and extension primers for all SNPs were designed using Spectro-Designer software (SEQUENOM, Inc.). To ensure that the observed polymorphisms were correct and not the results of experimental variation, the results were confirmed by repeating 25% of the assays and by directly sequencing 10% of the specimens, and no inconsistent genotype was found.

Data Analysis. Univariate and multivariate analyses were used to determine risk factors and to establish background risk profiles for breast cancer, and important reproductive risk factors were used as indices to estimate the level of estrogen exposure or susceptibility to estrogen exposure in the subsequent analysis.

The genotypic frequency of each SNP of the individual genes was compared between cases and controls. Differences in genotypic frequency of individual SNPs between cases and controls were tested using multiple logistic regression models (27) with simultaneous consideration of known risk factors of breast cancer, and the adjusted odds ratio for the association was estimated.

Our interest in identifying joint effects of MRN genotypes and reproductive risk factors in determining breast cancer risk stems from the fact that exposure to estrogen, reflected by reproductive risk factors, is known to initiate breast cancer development due to estrogen metabolites causing DSB formation (20, 28). If MRN genes playing a role in DSB repair were important in breast tumorigenesis, the relationship between breast cancer risk and reproductive risk factors would not be the same in women harboring different MRN genotypes; this was evaluated using both joint and stratified methods (29). We calculated the risk of breast cancer associated with the combination of the putative high-risk genotypes of the MRN genes and a reproductive risk factor. Using β estimates from the logistic regression model (30), in which we used a set of dummy variables representing different combinations of gene (i.e., whether or not harboring the putative high-risk genotypes) and risk factor, we assessed the relative excess risk due to harboring putative high-risk genotypes within reproductive risk factor strata (joint method). Furthermore, the risk of breast cancer associated with polymorphism of the MRN genes was compared between women with or without a reproductive risk factor (stratified method).

In examining the interaction between components of the MRN complex and BRCA1, we were particularly interested in identifying an interaction between Mre11 and BRCA1, as BRCA1 has been shown to inhibit the nuclease activity of Mre11 (31), and it has been suggested that this inhibition is required for precise DSB repair (32). To this end, a pull-down assay was done. BRCA1 and Mre11 were cloned, respectively, into the expression vectors pXJ-Myc (Myc epitope–tagged in the pXJ vector) and pcDNA3.1-His (Invitrogen), and the Myc-tagged BRCA1 or/and His-tagged Mre11 were expressed in 293T cells and the cells harvested after 50 h; one set of cells expressing both proteins was exposed to a single dose of 10 Gy radiation 48 h after transfection and harvested 2 h later. Protein extracts were prepared as described previously (32), and His-tagged Mre11 was pulled-down using Ni-NTA agarose (Sigma). The affinity purified complex and whole-cell extract were separated by SDS-PAGE and transferred to a nitrocellulose membrane, which was probed with monoclonal antibodies against polyhistidine (HIS-1, Sigma-Aldrich) or myc (9E11, Santa Cruz) and polyclonal antiactin antibody (A2066, Sigma-Aldrich), followed by horseradish peroxidase–conjugated secondary antibody and enhanced chemiluminescence reagent (33).

The present study included 559 female patients with pathologically confirmed infiltrating ductal carcinoma of the breast and 1,125 healthy female controls. The risk profile of this series of study subjects (Table 1) was similar to that reported in our previous breast cancer studies (16, 19, 20, 25, 26), and using multiple logistic regression analysis, a significantly increased risk was found to be conferred by a family history of breast cancer in female first-degree relatives [yes versus no; adjusted odds ratio (aOR), 1.50; 95% confidence interval (95% CI), 1.12-2.00]. Of the various reproductive risk factors, pregnancy-related risk factors were consistently found to be highly associated with an increased risk. Compared with controls, cases had a lower frequency of a history of full-term pregnancy (FTP; no history versus having at least one FTP; aOR, 1.51; 95% CI, 1.05-2.18) and were older at first FTP (≥26 years versus <26 years; aOR, 1.34; 95% CI, 1.04-1.72). Furthermore, significant protection was conferred by a greater number of FTPs (history of ≥3 FTPs versus no history of FTP; aOR, 0.63; 95% CI, 0.43-0.92) or a history of breast feeding (yes versus no; aOR, 0.70; 95% CI, 0.56-0.89). The significant protection conferred by pregnancy against the development of breast cancer has been suggested to be due to its causing permanent differentiation of the vulnerable breast stem cells, thus reducing susceptibility to estrogen exposure (34). These risk factors were therefore used in the subsequent analysis to examine the presence of a joint effect on breast cancer risk of agents causing DSBs (i.e., estrogen exposure or susceptibility to estrogen exposure reflected by significant pregnancy-related risk factors) and the MRN genes. No association was found between cancer risk and smoking status, radiation exposure, hormone replacement therapy, or dietary intake of specific kinds of foods or vegetables, but obese women (body mass index, >24 kg/m²) showed a significantly higher risk (aOR, 1.30; 95% CI, 1.04-1.63).

Fifteen SNPs of the three MRN genes were genotyped in an initial screening of 192 cases and 192 controls. Of these, one (rs1061945) in Mre11 was infrequent (frequency of the less frequent allele, <0.01) and, so, was not genotyped in the remainder of the samples. The remaining 14 SNPs (Table 2) were genotyped in all cases and controls. The genotyping results for individual SNPs in the controls were compared with the expected distribution of genotypes estimated on the basis of the observed allelic frequency of each SNP, and no statistically difference was found (i.e., none of the loci gave a significant P value in the χ² test). This suggests that the SNPs examined were in Hardy-Weinberg equilibrium and that possible bias due to genotyping error was less likely (35). All SNPs in the same gene were found to be in strong LD (P < 0.01) in both cases and controls (Table 2). This finding is consistent with results for SNPs in other genes in Taiwanese (16, 25, 26), because the Taiwanese population is genetically homogenous and LD between SNPs is much stronger than in other populations (24).

To explore a possible association between breast cancer and individual polymorphisms in MRN genes, the genotypic distribution of polymorphisms of individual genes was compared between cases and controls (Table 3). For the RAD50 variant examined, it is notable that the more frequently found allele is the one associated with an increased risk. Furthermore, the test for differences in the distribution of haplotype frequencies between cases and controls was done with a global test using the program FASTEHPLUS with 10,000 permutations (36, 37). However, possibly because the MRN genes were in strong LD, the analysis based on haplotypes was similar to that based on individual SNPs (Supplementary Table S1), and as a result, we chose the SNP which showed the most significant P value in the multivariate logistic regression analyses to represent the allelic status of individual MRN genes (Table 3). When the effects of breast cancer risk factors were simultaneously adjusted, a trend to an increased breast cancer risk was consistently found to be associated with harboring one additional high-risk allele of individual MRN genes (Table 3), although only Nbs1 polymorphism (i.e., the homozygous variant genotype of Nbs1) showed a borderline significant association with breast cancer development. On the basis of the risk profiles of the individual genotypes of each MRN gene (Table 3), in the case of Mre11 and Nbs1, the homozygous wild-type and heterozygous genotypes were grouped together and compared with the homozygous variant genotype, whereas in the case of Rad50, the homozygous variant and heterozygous genotypes were grouped together and compared with the homozygous wild-type genotype. After these groupings, the Nbs1 genotype remained a significant determinant for breast cancer development and was associated with a 1.29-fold increase in risk (95% CI, 1.00-1.69; Table 3). To comprehensively assess the relative contribution of individual MRN genes to the association with breast cancer development, we did a logistic regression analysis considering the effects of individual genes and known risk factors of breast cancer. Consistent with the findings in Table 3, the breast cancer risk associated with susceptibility genotypes varied for the three genes, being more significant for Nbs1 (aOR, 1.29; 95% CI, 1.00-1.69) than for Mre11 (aOR, 1.25; 95% CI, 0.97-1.62) or Rad50 (aOR, 1.21; 95% CI, 0.95-1.53).

Because individual MRN proteins act collectively by forming a trimeric complex, which plays multiple functions upstream and downstream of ATM in the DNA DSB repair pathway (4, 9, 10), the prediction was that any defect due to missense variants in individual genes of this complex would act by dominantly interfering with the function of the normal allele of the other partner. Thus, to determine whether a joint effect of these MRN genes was associated with breast cancer development, we examined the breast cancer risk associated with the number of these putative high-risk genotypes, using women with all three putative low-risk genotypes as the reference group. As shown in Table 4, the risk of breast cancer increased significantly with the number of putative high-risk genotypes (P_trend = 0.003), one additional putative high-risk MRN genotype being associated with a 1.25-fold increase in risk (95% CI, 1.10-1.44) and the highest risk (aOR, 2.35; 95% CI, 1.28-4.34) being seen in women harboring the high-risk genotype of all three MRN genes. Furthermore, on the basis of the defined functions of the MRN complex (4, 9, 10), each combination of two proteins in this complex displays specific and independent functions required for DSB repair or for cellular responses to DSB lesions. Two examples are that (a) the binding of Nbs1 stimulates only the endonuclease activity of Mre11, whereas the binding of Rad50 to Mre11 stimulates both its exonuclease and endonuclease activities and (b) Nbs1 bound to Mre11 is involved in checkpoint responses (4, 9, 10). We then dissected the complex into the individual functional components and looked for possible joint effects contributed by any combination of two genes of this complex. To exclude a false combination effect due to an unequal contribution of individual genes, we separately estimated the risk associated with harboring different putative high-risk genotypes using a dummy variable coding scheme (30) and women with all putative low-risk genotypes as the reference group. The results were consistent with the presence of a joint effect, a higher risk being seen in women harboring high-risk genotypes of both MRN genes (Table 4).

We investigated whether estrogen exposure in conjunction with the MRN susceptibility genotypes resulted in an increased risk of cancer using the joint method (Fig. 1) and the stratified method (Table 5). To carry out this analysis, we first classified our women into two groups, those with two or three and those with zero or one putative high-risk genotypes in the three MRN genes, as such a definition would give sufficient statistical power to address the relevant questions. The reference group consisted of women harboring 0 or 1 putative high-risk genotypes and less susceptible to estrogen exposure (having a history of FTP, younger at first FTP, or a greater number of FTPs). Our hypothesis was supported by the finding that, in the absence of the estrogen-related risk factors, the harboring of two to three putative high-risk genotypes was associated with a small nonsignificant increase in risk, whereas, in the presence of any one of these risk factors, the harboring of the same number of high-risk genotypes was associated with a much greater and significant combined risk of breast cancer (Fig. 1). The increase in breast cancer risk in this subgroup of women was more than additive. These results indicate the presence of a combined effect between the MRN complex and susceptibility to estrogen exposure during breast tumorigenesis. To further verify this finding, the stratified method was used, in which we examined whether the breast cancer risk associated with having a higher number of high-risk genotypes (i.e., harboring two or three putative high-risk genotypes) was modified by a history of breast feeding. Modification of the risk effect was supported by our findings, because a significant aOR associated with a higher number of putative high-risk MRN genotypes or with harboring one additional putative high-risk genotype of the MRN complex was only seen in those women with no history of breast feeding (Table 5). The possibility of a difference in statistical power in the detection of cancer risk due to different sample sizes in the subsets of women can be excluded, because the subsets with or without a history of breast feeding included a similar number of women.

To look for experimental evidence supporting a breast tumorigenic contribution of the MRN genes, we determined whether there was a physical interaction between Mre11 and BRCA1. His-tagged Mre11 and Myc-tagged BRCA1 were overexpressed in 293T cells, and a Ni-NTA pull-down assay was done. As shown in Fig. 2, in cells expressing both proteins (lanes 4 and 5), co–pull down of the BRCA1 prey with the Mre11 bait was observed. The negative controls consisted of material pulled down by Ni-NTA from nontransfected cells (lane 1) or cells expressing only one of the proteins (lanes 2 and 3). These data are consistent with the suggestion that BRCA1 and the MRN complex act together in vivo as a surveillance complex to guard genomic integrity (32). It is interesting that this interaction was DNA damage–dependent, because it was increased by irradiation in the absence of any increase in the amounts of two proteins in the extracts (lane 6). Because the pull down was done with MRE11, showing little change (lanes 4 to 6 in the raw of IB anti-His in Fig. 2), such an increase would be in agreement with an increase in the quantity of BRCA1 being pulled down by MRE11.

On the basis of a multigenic model and using both epidemiologic and experimental approaches, the present study examined the contribution to breast cancer tumorigenesis of the genes encoding the proteins in the MRN complex, the key complex involved in DSB repair (1-4). This study addressed not only the breast tumorigenic risk associated with the MRN genes, but also whether there is an interaction between these genes and reproductive risk factors in relation to breast cancer risk and whether MRN proteins physically interact with the breast cancer suppressor BRCA1. Our study permits a more precise evaluation of the breast cancer risk associated with MRN genes and a better insight into breast tumorigenesis initiated by the DSB repair genes and how this is modified by estrogen exposure.

Consistent with our hypothesis, the findings of the present study, showing that genotypic polymorphisms of the MRN genes were jointly associated with an increased breast cancer risk, provide evidence supporting a contribution of MRN genes to breast tumorigenesis. Interestingly, an association between a specific low-penetrance mutation of RAD50 and susceptibility to hereditary breast cancer has been reported recently, and mutation carriers show increased spontaneous chromosomal aberrations in their peripheral blood T lymphocytes (38), consistent with the role of MRN in maintaining genomic stability. In a recent study which looked for mutation in the consensus coding sequences of 13,023 genes in breast cancer tumor tissue, Mre11 was identified as one of the hotspots of mutation (39). Together, these findings provide essential support for the biological plausibility of our results. However, in considering whether our findings represent a true association between the SNPs of the MRN genes and breast cancer, the most important issue is the interpretation of the identified association between SNPs and the trait. The present study used a candidate gene approach based on SNPs locating in the genes of the DSB repair pathway. Because the SNPs analyzed are mostly in introns, do not affect amino acid coding, and therefore probably do not directly affect protein function, the observed associations between breast cancer risk and SNPs should be interpreted as the presence of LD between these SNPs and other SNPs in exons (resulting in functional polymorphism) or in regulatory regions (affecting the expression of these genes). We attempted to use more than one SNP in these genes to assign the haplotypes and to examine haplotype effects on cancer risk, but the information generated by haplotype analysis was limited due to strong LD between SNPs in the same gene. However, because of this same reason, i.e., the presence of significant pair-wide LD in all of the SNPs, it is reasonable to expect that any increased breast cancer risk caused by unknown sequence variants spanning each MRN gene would most likely be reflected indirectly by at least one of the genotyped SNPs. In addition, it is less likely that the observed association reflects the effects of other adjacent genes, because (a) based on the newly published haplotype map of the human genome, no well-defined cancer-associated genes are located in the same haplotype blocks of the MRN genes and (b) the closest putative cancer gene MMP16 (which is suggested to be abnormally expressed in renal cell carcinoma) is 1.6 Mb 5′ of Nbs1. Furthermore, genetic heterogeneity is less of a concern in Taiwan than in the United States (24), and as a result, potential bias due to population stratification is less likely to be significant in our study and the chance that the functional variants targeted by the same SNPs of individual MRN genes are different in cases and controls due to differences in the genetic background of the two groups is small. However, we recognize that the sequencing of the entire gene and promoter region is the definitive approach to identifying all of the important sequence variants and that a large-scale evaluation of these variants and functional assessments are needed to address this question.

The multigenic approach used in this study provided a unique opportunity to evaluate the relative importance of the individual MRN genes in breast cancer development and led to Nbs1 being identified as the most significant of the three. The extreme C-terminal region of Nbs1 contains a conserved ATM interaction motif, deletion of which results in loss of the ability to recruit ATM to DSB sites (11), suggesting a critical role of Nbs1 in DSB repair. Interestingly, this C-terminal region is distinct from the Mre11 interaction region of Nbs1, which links Nbs1 to Mre11 (11). Phenotypically, a missense variant of Mre11 affecting the nuclease activity of Mre11 does not significantly affect the ability of ATM to phosphorylate downstream effector proteins for DSB repair (8). In addition, although an in vitro experiment has suggested that the exonuclease activity of Mre11 is required to trim the ends of DSBs, thus facilitating DSB repair, other studies have questioned whether such activity is required in vivo for efficient DSB repair (9). On the other hand, chromosome instability, a phenotypic characteristic of tumorigenesis, is not observed in mouse cells harboring a homozygous Rad50 mutation (10). Although the strength of association, measured as the aOR and 95% CI in epidemiologic studies, between SNPs of susceptibility genes and cancer risk does not necessarily correspond to the relative importance of the molecular function played by each individual gene, these results from molecular and cellular studies and animal models, suggesting that Nbs1 is clearly the most important gene of those coding for the MRN complex during DSB repair, are still of significance. It is notable that a recent epidemiologic study also shows that Nbs1 polymorphisms and haplotypes contribute to the etiology of sporadic breast cancer in young non-Hispanic White women (40).

Evidence about the mechanism by which estrogen causes DNA damage, thus initiating breast cancer development, is provided not only by molecular and cellular studies (28), but also by epidemiologic observations (19, 20). The suggested mechanism is that estrogen metabolites (i.e., CE-Qs) bind to DNA, leading to the formation of depurinating adducts and resulting in DSB formation. On the basis of this evidence, we examined the breast tumorigenic role of the MRN genes by exploring whether breast cancer risk was linked to a joint effect of MRN genotypes and pregnancy-related risk factors. If these susceptibility genes were associated with breast cancer development via the hypothesized mechanism involving the regulation of DSB repair, the relationship between cancer risk and susceptibility genotypes would be expected to be more significant in that subset of women who were expected to be more susceptible to estrogen, indicated by no history of pregnancy or breast feeding, an older age at first FTP, or fewer FTPs, as these risk factors are an indicator of increased susceptibility to estrogen exposure (41). The findings are consistent with these expectations, yielding additional support for involvement of the MRN genes in breast cancer development. More importantly, these observations may help explain the issue of tissue specificity and why the DSB repair mechanisms are of particular importance in the development of breast cancer, as the risk factors of increased estrogen exposure or increased susceptibility to estrogen exposure presumably reflect the extent of DSB formation or the degree of susceptibility to DSB formation. Consequently, breast cells that have lost DSB-related checkpoint/repair due to the harboring of at-risk genotypes have a growth advantage over DSB checkpoint/repair-proficient cells and are selected for by the microenvironment imposed by estrogen-related risk factors, resulting in an increased risk of developing breast cancer. Interestingly, our inference about the interaction between DSB repair genes and estrogen exposure is supported by genetic evidence that reproductive history might influence cancer risk in women with BRCA1 or BRCA2 mutation (42).

Our attempt to identify an interaction between the MRN complex and BRCA1 was based on the extension of how we view the etiology of tumorigenesis from single-gene mechanisms to multigenetic or to etiologic pathway-wide networks (43). Consequently, the consideration of whether there is a causal link between a putative cancer-associated gene and tumor development might be extended to whole tumorigenic networks. Given this revised concept, it has become essential to explore the presence of functional interactions between the MRN complex and breast cancer susceptibility genes, including ATM, Chk2, and BRCA1. In addition to the upstream role of the MRN complex in recruiting ATM (6-8, 11), thus initiating a series of responses to repair DSBs as mentioned above, there is recent evidence that this recruitment requires not only Nbs1 but also BRCA1 and that the absence of either protein results in a failure to recruit activated ATM to DSB sites (44). Furthermore, we have shown that ATM, Chk2, BRCA1, and the MRN complex interact to determine the fidelity of DNA DSBs end-joining (32, 43). These recently identified pathways are critical evidence for a breast tumorigenic role of the MRN genes. In addition, due to its nuclease activity and DNA binding capability, the MRN complex is involved in the initial processing of DSBs (45). These activities reside in Mre11, which is required to trim the ends of DSBs, thus facilitating DSB repair, especially during the error-free mechanism of homologous recombination by producing the overhanging DSB ends required for efficient homologous recombination (3, 4). On the other hand, because BRCA1 inhibits the nuclease activity of Mre11 (31) and Chk2 phosphorylates BRCA1, leading to inhibition of MRN complex foci formation (46), it has been suggested that this inhibition is required for precise end-joining because exonuclease digestion would result in the loss of nucleotides, leading to imprecise end-joining if these digested ends were rejoined directly (32). Thus, in the present study, our finding of a damage-dependent interaction between BRCA1 and Mre11 is in line with this hypothesized model. To further confirm this interaction is of functional importance, in our ongoing experiment, we have identified interaction domains of BRCA1 and Mre11 and have done functional assays to examine phenotypic changes in mutant Mre11 (unpublished observations). All these findings of physical and functional interactions between individual MRN components and proteins encoded by well-documented breast cancer susceptibility genes provide essential support for a tumorigenic contribution of MRN genes during breast cancer formation.

The great interest in defining the breast tumorigenic role of the genes involved in DSB repair, such as the MRN genes, arose from an attempt to confirm the “mutator phenotype” hypothesis (47). Given that tumorigenesis is a multistep process, resulting from a series of genomic alterations, which lead to the progressive disordering of the normal mechanisms controlling the growth, differentiation, and death of the cell, this hypothesis suggests that, to account for the high frequency of genomic alterations required for tumor progression, the genomes of cancer cells are unstable and that defective mutators cause these genomic instabilities. The fact that the familial breast cancer susceptibility genes BRCA1 and BRCA2 are involved in the homologous recombination pathway for DNA DSB repair (48) and our previous finding that the extent of DSB-initiated chromosomal instability in tumors increases significantly, as tumors progress to poorer grades or later stages (21) support the idea that breast cancer pathogenesis is driven by DSB-initiated chromosomal instability, and we hypothesize that the mechanisms regulating DSB repair may play a mutator role. In our recent studies (16, 25, 26, 49), we have identified specific molecular DSB repair mechanisms, functional aberrations of which are related to an increased risk of breast cancer and advanced pathologic/clinical manifestations. The results of the present study provide additional evidence supporting the mutator role of DSB repair mechanisms during breast cancer formation.