Abstract
Aim: The DNA-repair gene X-ray repair cross-complementing group 3 (XRCC3) is important in DNA double-strand break repair and plays a critical part in initiation of carcinogenesis. Triple-negative breast cancer (TNBC) is the most difficult breast cancer subtype with no existing gene-targeting drugs and little knowledge on its genetic etiology. This study aimed to investigate the contribution of the XRCC3 genotype to individual TNBC susceptibility. Materials and Methods: A total of 2,464 Taiwan citizens consisting of 1,232 breast cancer cases and 1,232 controls were enrolled in this case–control study, and genotyping of XRCC3 rs1799794, rs45603942, rs861530, rs3212057, rs1799796, rs861539 and rs28903081 were performed with polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). We also conducted risk-stratified sub-group analyses to determine the association between the genotype and age- and hormone-related characteristics of breast cancer sub-groups. Results: There was no significant difference between breast cancer and control groups in the distributions of the genotypic or allelic frequencies as for the XRCC3 rs1799794 (p=0.5195 and 0.9545), rs45603942 (p=0.3478 and 0.1449), rs861530 (p=0.4567 and 0.5081), rs3212057 (p=1.0000 and 1.0000), rs1799796 (p=0.8487 and 0.7315) and rs28903081 (p=1.0000 and 1.0000), respectively. However, the XRCC3 rs861539 TT genotype was more prevalent in patients with breast cancer [odds ratio (OR)=2.99, 95% confidence interval (CI)=1.62-5.55; p=0.0002], and especially among those who were younger than 55 years (OR=2.61, 95% CI=1.82-3.73; p=0.0001), with first menarche earlier than 12.2 years (OR=2.47, 95% CI=1.74-3.52; p=0.0001), with menopause at 49.0 years old or later (OR=2.53, 95% CI=1.76-3.62; p=0.0001), or with TNBC (OR=2.05, 95% CI=1.46-4.28; p=4.63*10−4). Conclusion: XRCC3 rs861539 TT is a potential predictive marker for TNBC in Taiwanese women and investigations in other populations are warranted for further universal application in cancer detection and prediction.
Breast cancer is one of the most common worldwide malignancies among women, and its morbidity and mortality have not decreased with the development of anticancer drugs (1-3). Among the sub-types of breast cancer, the most dangerous is triple-negative breast cancer (TNBC), which refers to breast tumors that do not express the proteins for estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor 2 (HER2) (4). TNBC is characterized by highly invasive tumors, with poor prognosis, likely to recur locally and metastasize distantly. Clinically, endocrinotherapy and anti-HER2 target treatment are ineffective. However, no standard therapy is available at present. In Taiwan, breast cancer ranks second among all cancers, noted for high incidence, high mortality, and early onset, and about 10% of breast cancer cases are TNBC (5, 6).
Platinum drugs may crosslink with DNA double strands, leading to DNA double-strand breaks (DSBs), inhibiting DNA replication and transcription and eventually causing cell-cycle arrest, DNA repair and cell death. Patients with TNBC are abnormal in a variety of genes and signal transduction pathways and have defects in DNA repair. Consequently, platinum drugs might be more efficacious in treatment of TNBC compared to alternative agents (7). Mounting evidence has shown that genetic polymorphisms in DNA repair genes may affect overall DNA repair ability, thereby impacting the efficacy of individualized chemotherapy (8-11).
The X-ray repair cross-complementing group 3 (XRCC3) gene, that is located on human chromosomes 14q32.3, encodes for a DNA-repair protein XRCC3. XRCC3 is a member of the DNA repair protein RAD51 homolog 1 (RAD51)-related protein family that plays a role in homologous recombination (HR) to repair DSBs and maintain the overall integrity of the human genome (12). In literature, several studies have been performed to evaluate the relationship between the rs861539 C/T polymorphism (also named Thr241Met, T241M, C18067T and C722T) of the XRCC3 gene and cancer risk, making it the most commonly studied polymorphism of the XRCC3 gene (13-16). In literature, there are some reports investigating the contribution of XRCC3 rs861539 polymorphism to breast cancer but the findings are controversial and no article discusses the subgroup of TNBC (17-25). Some studies have identified the T variant of XRCC3 rs861539 as being associated with increased risk for breast cancer (17-21), while others did not (22-25). Pooled and meta-analyses tended to show a small but significant increase in breast cancer risk for this polymorphism (26-29). However, none of the previous literature discussed the contribution of XRCC3 rs861539 to TNBC. The lack of analysis of TNBC may be the limited sample size of patients with TNBC and lack of detection and recording for triple-negative markers.
Following the central dogma of molecular biology, variants of the XRCC3 rs861539 polymorphism may affect the function of the encoded protein, leading to altered DNA-repair capacity, that increased the level of bulky DNA adducts in leukocytes of healthy individuals (13). Among patients with bladder cancer, it was also found that variants of XRCC3 rs861539 polymorphism affected the function of the encoded protein and consequently altered the DNA repair capacity of their cells (30). Thus, the rs861539 C/T polymorphism and other polymorphic sites may also play a role in the pathogenesis and development of breast cancer, especially of TNBC.
As far as we are aware of, XRCC3 genotype among Taiwanese has never been investigated in association with breast cancer, let alone TNBC. This study's genotyping work ascertained correlation between XRCC3 polymorphisms and breast cancer risk in a large population of Taiwanese women, with 1,232 cases and 1,232 controls. Additional analyses evaluated the contribution of XRCC3 genotypes to breast cancer with specific clinicopathological features, such as those of TNBC.
Materials and Methods
Investigated population and sample collection. A total of 1,232 patients diagnosed with breast cancer were recruited at the outpatient clinics of general surgery at the China Medical University Hospital in Taichung, Taiwan. Female patients were included and males were excluded. Clinical characteristics of patients (including histological details) were all defined by expert surgeons. The questionnaire included questions on history and frequency of alcohol consumption and smoking habits, and “ever” having such a habit was defined as more than twice a week for a period of more than half year. Self-reported alcohol consumption and smoking habits were evaluated and classified as categorical variables. Staining of protein expression on slides was performed, reviewed and scored by two independent pathologists. For ER, PR, and p53 immunoassaying (Santa Cruz Biotechnology, Santa Cruz, CA, USA), nuclear staining in 10% of neoplastic cells served as the positive cutoff; a Ki-67 labelling index of >30% was considered positive. HER2/neu immunochemistry (Santa Cruz Biotechnology) results were derived according to the package insert and guidelines of the American Society of Clinical Oncology and College of American Pathologists (31). All patients voluntarily participated, completing self-administered questionnaires and supplying peripheral blood samples. An equal number of age-matched healthy volunteers without breast cancer as controls were selected after initial random sampling from the hospital's Health Examination Cohort. Exclusion criteria of the control group included previous malignancy, metastasized cancer from other or unknown origin, and any familial or genetic disease.
Genotyping conditions. Genomic DNA was extracted from peripheral blood leucocytes using the QIAamp Blood Mini Kit (Blossom, Taipei, Taiwan, ROC). In this study, a total of seven polymorphic sites were analyzed for all the participants of both the control and case groups. Briefly, all seven polymorphic sites were genotyped by means of a polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP). PCR was performed on a BioRad Mycycler (BioRad, Hercules, CA, USA) following the manufacturer's instructions. Each PCR reaction consisted of 5 min initial cycle at 94°C for 5 min; 40 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s; and a final extension at 72°C for 10 min. Then the SNP-containing DNA amplicons were subjected to individual overnight digestion by restriction endonucleases following the manufacturer's instructions (see Table I for details). Following digestion, each sample was immediately analyzed by 3% agarose gel electrophoresis. Details of the primer sequences, and enzymatic digestion conditions for each SNP analyzed in this study are summarized in Table I.
Statistical analyses. To ensure that the controls used were representative of the general population and to exclude the possibility of genotyping error, the deviation of the genotypeic frequencies of XRCC3 SNPs in the controls from those expected under the Hardy–Weinberg equilibrium was assessed using the goodness-of-fit test. Pearson's chi-square test was used to compare the distribution of the XRCC3 genotypes between cancer cases and non-cancer controls. Cancer risk associated with the genotypes was estimated as odds ratios (ORs) and 95% confidence intervals (CIs) using unconditional logistic regression. Data differences were recognized as significant when the statistical p-value was less than 0.05.
Results
A total of 1,232 patients diagnosed with breast cancer and 1,232 matched controls were enrolled, as compared and summarized in Table I. Patients and controls were well-matched regarding their ages, ages at menarche, and ages when bearing their first child (p>0.05) (Table II). The patient group had significantly more smokers and alcohol drinkers than did the control group (p<0.05) (Table II).
The frequencies of the genotypes of the XRCC3 polymorphisms in the breast cancer and control groups are shown in Table III. The results of genotyping analysis revealed the distribution of XRCC3 rs861539 was significantly different between breast cancer and control groups (p=1.10×10−6), while the distributions of XRCC3 rs1799794, rs45603942, rs861530, rs3212057, rs1799796 and rs28903081 genotypes were not (p>0.05) (Table III). The ORs associated with CT and TT genotypes were 1.74 and 2.99 (95% CI=1.32-2.31 and 1.62-5.55; p=0.0001 and 0.0002) compared to the CC wild-type genotype.
The distributions of alleles for each of the XRCC3 genotypic polymorphisms in the breast cancer and control groups are shown in Table IV. The results of the analysis revealed that the distribution of XRCC3 rs861539 was significantly different between breast cancer and control groups (p=3.77×10−9), while the distribution of XRCC3 rs1799794, rs45603942 and rs861530 was not (p>0.05) (Table IV). Those for XRCC3 rs3212057 and rs28903081 were not analyzed since they lack genetic polymorphism in the Taiwanese population. The OR for carrying the minor T allele at XRCC3 rs861539 was 1.99 (95% CI=1.58-2.52; p=3.77×10−9) compared to carrying the major C allele (Table IV).
It is well-known that distinct sub-types of breast cancer may have different mechanisms of carcinogenesis, we therefore analyzed the association among XRCC3 rs861539 genotypes with age-related and clinicopathological characteristics of patients with breast cancer (Tables V and VI). For the analysis of age of onset of breast cancer, the results showed that CT and TT genotypes for XRCC3 rs861539 were related to increased risk of diagnosed with breast cancer onset when patients were younger than 55 years (OR=2.19 and 5.61, 95% CI=1.48-3.23 and 2.31-13.63, respectively; p=8.84×10−8), but not in those aged 55 years or older (p=0.3073). Carrying a T allele-bearing genotype (CT or TT) was also associated with 2.61-fold higher risk than carrying wild-type CC in those younger than 55 years (OR=2.61, 95% CI=1.82-3.73; p=0.0001) (Table V). For the analysis of age at menarche, the results showed that those with CT and TT genotypes at XRCC3 rs861539 were at higher risk of breast cancer among those who had first menarche at age younger than 12.2 years (OR=2.12 and 4.66, 95% CI=1.45-3.12 and 2.02-10.74, respectively; p=4.71×10−7), but not in those whose menarche started later (p=0.2250). Those with CT and TT genotypes were also at 2.47-fold higher risk than those with wild-type CC among those who were younger than 12.2 years (OR=2.47, 95% CI=1.74-3.52; p=0.0001) (Table V).The results showed that women with T allele-carrying genotypes were associated with higher risk of breast cancer regardless of age at their birth of their first child (Table V). Women with T allele-carrying genotypes at XRCC3 rs861539 in whom menopause started age 49 years or more were at higher risk of breast cancer (OR=2.13, 4.81 and 2.53, 95% CI=1.44-3.18, 2.09-11.05 and 1.76-3.62, respectively; p=4.18×10−7), but not those who were younger at menopause (p=0.2061) (Table V). After adjustment for family history, smoking and alcohol drinking habits, all the significant findings in Table V retained the same trends.
Most interestingly, the role of variant genotypes (CT plus TT) serving as a genetic early detector for breast cancer was specifically focused on patients with TNBC (OR=2.05, 95% CI=1.46-4.28; p=4.63*10−4), but not for non-TNBC patients (OR=2.05, 95% CI=1.46-4.28; p=0.3666) (Table VI). On the contrary, the difference in distribution of the variant genotypes among patients with breast cancer when stratified by Ki-67 status was not statistically significant (p=0.2503 and 0.1413 for Ki-67-negative and -positive sub-group comparisons) (Table VI). After the adjustment for individual characteristics including family history, smoking and alcohol drinking habits, the variant genotypes (CT plus TT) remained associated with TNBC, but not non-TNBC or Ki67 status.
Discussion
TNBC is frequently observed in young patients and is associated with larger, higher-grade tumors (4), in addition to higher recurrence rates of metastasis and death (32). Among breast cancer cases, 5-10% are believed to be hereditary and associated with certain gene mutations, such as mutations in BRCA1 and BRCA2 (33). Scientists believe that TNBC must have subtle and distinct genomic differences in the common pathways of breast carcinogenesis and are devoted to the search for useful markers, but positive findings are very limited. In 2014, we found that cyclin D1 (CCND1) A870G GG genotype was especially less prevalent in Taiwanese patients with TNBC (34). In 2015, Li and her colleagues found that the A allele of TNFA-308 was associated with higher metastasis risk among Caucasian and Asian patients with TNBC (35). In 2013, Xie and his colleagues found that the G allele of 8-oxoguanine glycosylase 1 (hOGG1) gene was associated with increased risk of TNBC in Chinese Han women (36). Investigating the responses of 60 patients with TNBC to platinum-based chemotherapy, Lu and his colleagues found that the genotype of one important nucleotide excision repair gene, excision repair cross-complementing group 2 (ERCC2), would determine the outcome of platinum-based chemotherapy. Patients with GG genotype had better response towards platinum-based chemotherapy than those with GA genotype at ERCC2 rs1799793 (37). These two genomic markers are both closely involved in excision repair systems.
In the current study, we found a novel biomarker for TNBC in the DSB repair system, XRCC3 rs861539. The results showed that T allele (or CT and TT genotypes) of XRCC3 rs861539 is a breast cancer marker for Taiwanese women (Tables III and IV). In addition, T allele carriers of XRCC3 rs861539 were more prevalent in patients with breast cancer who were younger at onset (<55 years), with earlier first menarche (<12.2 years), with later menopause (≥49 years) (Table V), or with TNBC (Table VI). There was no difference for women with different ages of first full pregnancy, or Ki-67 status (Tables V and VI).
We have collected the largest breast cancer population in Taiwan, 1,232 cases and age-matched controls, to investigate the contribution of XRCC3 genotypes to breast cancer risk (Table II). We found that the genotype of XRCC3 rs861539 was associated with breast cancer susceptibility while other polymorphisms were not (Tables III and IV). This significance is consistent with previous findings in Pakistani (20), Polish (21), UK (17), Portuguese (18) and Thai (19) populations. Since we have collected the largest breast cancer population in Taiwan, the strategy of enlarging the number of investigated subjects is not an urgent need.
In the literature, it is well-believed that estrogen exposure is closely related to breast carcinogenesis, and there is no denying that age is the strongest demographic risk factor for most malignancies, since statistically, 75% of malignancies occurred in patients older than 55 years (38). With a large enough sample size, we are confident for the evaluation of the contribution of this SNP to breast cancer with specific clinicopathological features by stratification analysis. The estrogen- and age-related indices included age at onset, age at menarche, age at birth of first child, and age at menopause (Table V). As mentioned above, we are also interested in evaluating the contribution of XRCC3 rs861539 to TNBC.
In the current study, 104 patients were identified as having TNBC, so named for its negative expression of ER, PR, and HER2/neu (39), and characterized by aggressiveness and higher rates of recurrence and metastasis. It is of great value to determine potential oncotargets for TNBC from analyzing patients' genotypes or phenotypes since the existing targeted therapies which are effective in other sub-types of breast cancer are not effective in dealing with TNBC. Typically, TNBC occurs in young patients whose disease is associated with genetic variations in BRCA1 and other hereditary genes, such as hOGG1 and EGFR2 (31, 36, 40). All our findings show that the longer the estrogen exposure, the higher the risk of TNBC is for women carrying CT and TT genotypes of XRCC3 rs861539. We also found that the full pregnancy for child was not a protective factor for TNBC among Taiwanese women (Table V). Although the up-regulation of Ki-67 was reported to be a potential indicator for TNBC (41), the expression of Ki-67 has no interaction with XRCC3 rs861539 genotype in determining the TNBC risk in this study (Table VI).
The XRCC3 protein plays an important in the homologous recombination (HR) repair system, that is the most common pathway for repairing the radiation-induced DNA DSBs together with non homologous end-joining (NHEJ) pathway (42). In addition to affecting DNA-repair capacity in patients with bladder cancer (30), the T variant of XRCC3 rs861539 was also reported to be associated with elevated levels of DNA adducts (13), chromosomal deletions (43), and sensitivity to ionizing radiation and cross-linking agents (44, 45). In the current study, although we did not measure DNA adducts or repair capacity of the cells from tumor sites of the patients and non-tumor sites from the patients and controls, our findings support the concept that individuals carrying CT and TT genotypes of XRCC3 rs861539 are at higher risk of breast cancer (17-21) and other cancer types (11, 46) than those carrying the wild-type CC genotype.
In conclusion, the current study indicated that the T allele of XRCC3 rs861539 may be regarded as a predictive marker for breast cancer and TNBC. The determination of XRCC3 genotypes among Taiwanese without cancer and those with breast cancer may contribute to a better understanding of the mechanisms of breast cancer by evaluating possible interactions between XRCC3 genotypes and well-established risk factors. Population-based association studies in other countries are warranted to validate the possibility of XRCC3 being a novel gene target for anticancer drug development for breast cancer, especially TNBC.
Acknowledgements
We appreciate Tzu-Chia Wang and Yun-Ru Syu for their excellent technical assistance. This study was supported by research grants from Taichung Armed Forces General Hospital (104-A11, 104-A13) and Taiwan Ministry of Health and Welfare Clinical Trial and Research Center of Excellence (MOHW104-TDU-B-212-113002). The Authors declare no conflict of interest involving any company or individual.
Footnotes
↵* These Authors contributed equally to this study.
- Received July 30, 2015.
- Revision received September 19, 2015.
- Accepted October 7, 2015.
- Copyright© 2015, International Institute of Anticancer Research (Dr. John G. Delinasios), All rights reserved