Contribution of excision repair cross-complementing group 1 genotypes to triple negative breast cancer risk

Compared with other subgroups of breast cancer, triple negative breast cancer (TNBC) is considered to be the one with the greatest invasiveness and metastatic mobility, and the highest recurrence rate. Considering the lack of predictive markers for TNBC, we aimed to examine the contribution of excision repair cross complementing-group 1 (ERCC1) genotypes to TNBC. The rs11615 and rs3212986 of ERCC1 were investigated and evaluated for their associations with susceptibility to breast cancer, especially TNBC, in Taiwan. In this study, 1,232 breast cancer patients (104 were TNBC) and 1,232 healthy controls were recruited and their genotypes at ERCC1 rs11615 and rs3212986 were revealed by polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP) analysis. Our results indicated that genotypes of ERCC1 rs11615 (Ptrend = 2.2*10E-9), but not rs3212986 (Ptrend = 0.6181), were associated with breast cancer risk. In the allelic frequency distribution analysis, breast cancer patients carried the T allele of ERCC1 rs11615 a higher rate than the control subjects, further supporting the idea that ERCC1 rs11615 TT genotype is positively associated with breast cancer susceptibility. More importantly, the frequency of the ERCC1 rs11615 TT genotype was even higher among TNBC patients than among other subtypes of breast cancer patients (P = 0.0001, odds ratio = 1.73, 95% confidence interval = 1.15–2.63). The genotypes of ERCC1 rs11615 were not associated with Ki67 status. Our findings firstly show that the T allele of ERCC1 rs11615 can serve as a predictive biomarker for breast cancer and TNBC. We believe that ERCC1 could serve as a target for personalized treatment of breast cancer, especially for TNBC.


Introduction
Published statistics reveal that breast cancer is the most common cancer diagnosed among females worldwide [1]. Among the subgroups of breast cancer, triple negative breast cancer PLOS ONE | https://doi.org/10.1371/journal.pone.0202112 August 10, 2018 1 / 11 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 (TNBC) accounts for 10-20% of all newly diagnosed female breast cancers [2]. Since cells of this cancer lack the three common receptors, estrogen receptor (ER), progesterone receptor (PR), and hormone epidermal growth factor receptor 2 (HER-2), there are as yet no specific clinical drugs or targeting therapies for this kind of breast cancer. As a result, TNBC is characterized by high invasiveness, poor prognosis, and high chances of recurrence [3]. The abnormality of gene expression in TNBC patients is another concern of scientists. Given the shortage of targeted treatments for TNBC and its typical properties, the discovery of the biomarkers and medication for TNBC are considered to be in a high priority.
Environmental factors, such as ionizing radiation (IR) and ultraviolet (UV) radiation, are considered to be common causes of DNA damage in living organisms [4]. In cases of DNA damage, repair systems such as base excision repair (BER), nucleotide excision repair (NER), and mismatch repair (MMR) are believed to confer resistance to front line cancer treatments [5]. Defective DNA repair systems enable cancer cells to accumulate genomic alterations, which may also cause them to lose their normal growth regulation [6][7][8]. Therefore, any polymorphisms in genes involved in DNA repair systems may contribute to the etiology of carcinogenesis, including breast cancer initiation and progression.
NER functions as the main pathway to repair massive DNA damage such as that caused by UV light, environmental mutagens, and some cancer chemotherapeutic adducts of DNA [9]. In literature, cellular NER capacity was found to be deficient in the cells of breast cancer patients, especially those at sporadic stage I [10,11]. The deficiency could be explained by the lower expression of some NER proteins, such as XPA, XPF and CSB [11]. Among all the subtypes of breast cancer, the TNBC was of the lowest NER capacity [12]. The excision repair cross complementing-group 1 (ERCC1) gene is located on human chromosome 19q13.32, and encodes a DNA repair protein, ERCC1 [13]. ERCC1 plays an essential role in NER-pathway because of its damage recognition and excision ability [14]. A positive correlation was found between ERCC1 mRNA expression and DNA repair capacity in several studies [15,16]. The expression of ERCC1 was increased by cisplatin treatment in a time-and a dose-dependent manner in ovarian cancer cell lines [17]. However, the expression level of ERCC1 may have a dual function in breast carcinogenesis. It was reported that higher expression of ERCC1 is associated with favorable prognostic factors for early stage breast cancer patients [18], but with poor outcome for those metastatic TNBC patients treated with platinum-based chemotherapy [19]. Mutations in this gene contributed to the etiology of cerebro-oculo-facio-skeletal (COFS) syndrome [20], and polymorphisms in ERCC1 that alter its expression may influence overall genomic stability, and thus enhance personal susceptibility to cancer. Previously, ERCC1 variants have been found to be associated with carcinogenesis in various types of cancer, such as lung, colorectal, gastric and ovarian, as well as breast cancer [21][22][23][24][25]. Notably, the association of ERCC1 polymorphisms with breast cancer has been revealed in different countries, including Korea, United States, Iran, China and Thailand [20,[26][27][28][29][30]. As mentioned above, high expression of ERCC1 was found to be associated with poor patient outcomes for TNBC patients [18]. However, no previous literature has explored the contribution of ERCC1 polymorphisms to TNBC.
In the current investigation, we aimed at discovering the contributions of ERCC1 rs11615 and rs3212986 genotypes to breast cancer susceptibility in a large population of Taiwanese females, including 1,232 breast cancer cases and 1,232 healthy controls. These two SNPs were selected because they were reported to be associated with the susceptibility and outcomes of different cancers [21][22][23][24][25]. In addition, a query of the ClinVar and dbSNP databases turned out that these two SNPs were linked to "drug responses". Interestingly, Zhu and his colleagues showed that there were no statistical associations between ERCC1 rs11615 and the risk of breast cancer [30]. Moreover, we attempted to find in ERCC1 a useful biomarker for early prediction and detection of TNBC in the Taiwanese population.

Investigated sample collection
A total of 1,232 female patients diagnosed with breast cancer were enrolled in China Medical University Hospital (CMUH), Taichung, Taiwan. At the same time, controls were recruited from the Health Examination Cohort of CMUH [31]. These individuals had received a health checkup with history taking, complete physical examination, serial laboratory testing, and necessary image study. We excluded those with primary malignancy, metastatic cancer from other or unknown origin, and any hereditary or genetic disease. All the participants were volunteered, following self-administered questionnaires and furnishing of peripheral blood samples, for our following experiments. The content of the questionnaire includes former historical questions and the habits of alcohol consumption and cigarette smoking. These factors were recorded and further summarized in Table 1. All the enrolled individuals gave informed consent. Our study was evaluated and approved by the Institutional Review Board of China Medical University Hospital (DMR-99-IRB-108).

Genotyping conditions
Genomic DNA extraction from each investigated individual was prepared using the QIAamp Blood Mini Kit (Blossom, Taipei, Taiwan). The DNA obtained from peripheral blood leucocytes was stored at -80˚C after extraction. Amplified DNA products were subjected to digestion by BsrDI and MboII restriction endonucleases, respectively, for 2 h at 37˚C. Detailed information on primer sequences and enzymatic digestion conditions is summarized in Table 2. Polymerase chain reaction (PCR) was performed following the manufacturer's instructions on a BioRad Mycycler (BioRad, Hercules, CA, USA) after the digestion. For each PCR procedure, the conditions were set at 94˚C for 5 min initial cycle; 35 cycles of 94˚C for 30 s, 55˚C for 30 s and 72˚C for 30 s; and a final extension step at 72˚C for 10 min. After PCR amplification, the PCR products were separated by 3% agarose gel electrophoresis for approximately 40 min. Following digestion with BsrDI, PCR products of ERCC1 rs11615 originating from the C allele were uncut (393-bp), whereas the T allele was cut into fragments of 228-bp and 165-bp. Upon digestion with MboII, PCR products of ERCC1 rs3212986 originating from the G allele were uncut (367-bp), while the C allele was cut into fragments of 233-bp and 134-bp. All the genotypic processing was repeated blindly by two researchers, and all the genotyping results were 100% concordant.

Statistical analyses
All statistical analyses were performed using Stata 14.0. Student's t-test was applied to the comparison of ages between the breast cancer case and control groups. Pearson's Chi-square test was applied for comparing the distribution of the ERCC1 genotypes among the control and breast cancer groups. The association between ERCC1 genotypes and breast cancer risk was estimated by odds ratios (ORs) and 95% confidence intervals (CIs). Data differences were recognized as significant when the statistical p-value was less than 0.05. We calculated the statistical power of our analysis. With a sample size of 1,232 breast cancer cases and 1,232 controls, and a minor allele frequency of~30% for both SNPs in controls, we had 80% power to detect a minimum OR of 1.29 for overall breast cancer risk. For TNBC risk, with 104 cases and 1,232 controls, we had 80% power to detect a minimum OR of 1.78.

Comparison of demographics and lifestyles between the breast cancer case and control groups
The characteristics of the investigated population were summarized and shown in Table 1. Characteristics such as age, age at menarche, and the age at the first birth were all well-matched between patients and controls (p>0.05). Lifestyle factors like cigarette smoking and alcohol use were also considered in our study. The results revealed that both cigarette smoking and alcohol consumption were significantly different between groups. The number of smokers and alcohol drinkers in the patient group was much greater than that in the control group. The results demonstrated that these lifestyle-related factors may put the breast cancer patients at risk (p<0.0001). Amplified DNA products from the samples were digested by specific enzymes. Details are shown in Table 2.

Association of ERCC1 genotypes and breast cancer risk
In our current study, two ERCC1 polymorphisms (rs11615 C>T and rs3212986 T>G) were studied and compared between healthy controls and breast cancer patients. Table 3 analyzed and demonstrated the distribution of ERCC1 genotypes of each polymorphism between groups. The genotypes of ERCC1 rs3212986 were not significantly different among healthy controls and breast cancer patients (P trend = 0.6181, Table 3). In contrast, the distribution of ERCC1 rs11615 genotypes was found to be differently distributed among 1,232 controls and 1,232 breast cancer patients (P trend = 2.2 × 10 −9 ) (Table 3). However, detailed analysis of the results for ERCC1 rs11615 showed that it was the homozygous TT genotype, but not the heterozygous CT genotype, that related to the increasing risk of breast cancer (OR = 2.08 and 1.06, 95% CI = 1.64-2.64 and 0.89-1.26, P-value = 0.0001 and 0.5205, respectively) ( Table 3).
Recessive and dominant models of ERCC1 rs11615 were further compared in Table 3. Both models indicated a great positive association between the genotypes of ERCC1 rs11615 and breast cancer risk (OR = 2.03 and 1.29, 95% CI = 1.62-2.54 and 1.10-1.51, P = 0.0001 and 0.0016, respectively).

Association of ERCC1 allelic subtypes and breast cancer risk
To extend our study, we also analyzed the allelic frequencies of ERCC1 polymorphisms (rs11615 and rs3212986) among the investigated groups (1,232 controls and 1,232 breast cancer patients); the data are presented in Table 4. This was consistent with our findings that the distribution of ERCC1 rs11615 allelic frequencies was significantly associated with increased breast cancer risk (P = 8.8 × 10 −9 ), while ERCC1 rs3212986 allelic frequencies were not found to be related to breast cancer risk (P = 0.3028). In the patient group, the frequency of the variant T allele was much higher than the wild-type C allele (38.4% and 30.6%, respectively).

Association of ERCC1 rs11615 genotypes with breast cancer risk
A Chi-square test was performed to investigate the association of ERCC1 rs11615 genotypes with breast cancer risk. In the clinicopathologic characteristics analysis, there were 657 patients available for triple-negative status and 615 patients available for Ki67 status. Surprisingly, Table 5 revealed that ERCC1 rs11615 genotypes were differentially distributed among the breast cancer patients who showed positive triple-negative status (P = 0.0001). However, a more representative distribution of ERCC1 rs11615 genotypes was observed for the other factor, Ki67 status (OR = 1.05 and 1.05).
To summarize, these findings indicate that ERCC1 rs11615 (C>T) was associated with breast cancer risk. Therefore, ERCC1 rs11615 genotypes may serve as predictive markers for the early detection of breast cancer patients. More importantly, the variant forms of ERCC1 genotypes (CT and TT) contribute to an increased risk of developing TNBC.

Discussion
TNBC occurs more commonly in younger females, especially those with BRCA1 germline mutations [32,33]. Compared to other hormone receptor positive breast tumors, TNBC tumors are typically more invasive and aggressive, with greater risk of early relapse, which clinically enhances the difficulties of curing TNBC [34,35]. Given the suboptimal outcomes after chemotherapy, the search for quantifiable TNBC biomarkers for early prediction is urgently needed. To this end, we have previously validated several biomarkers for TNBC in a large Taiwanese population. In 2014, the Cyclin D1 (CCND1) A870G GG genotype was found to be infrequent in Taiwanese TNBC patients, which may contribute to distinguishing the TNBC patients from other breast cancer patients [36]. In 2015, we found X-Ray Repair Cross Complementing 3 (XRCC3) genotypes were associated with Taiwanese TNBC patients, suggesting that XRCC3 may be a potential predictive marker for TNBC [31]. Further, in 2016, the CC genotype of tissue inhibitor of metalloproteinase-1 (TIMP-1) rs4898 was also found to increase the risk for TNBC in Taiwan and may serve as a predictive marker for TNBC [37]. The details of intracellular signaling pathways, such as the cell cycle, extracellular matrix regulation and DNA repair, are worth of further investigations. The DNA repair system plays an essential role in preventing DNA damage accumulation, maintaining genomic stability and serving as anticancer gatekeepers of the cells. Several lines of evidence indicate that tumor cells were found with more DNA repair protein-related mutations, leading to partial or loss of their related functions, which may serve as one of the reasons for progression in cancer initiation and development [38][39][40]. Among the types of DNA damage, the double-strand breaks (DSBs) may represent the most severe and irreversible damage to the whole genome in the case they are not reversed by the DNA repair system immediately and properly when they are formed. Cells that survive DSBs and do not undergo apoptosis are prone to becoming cancer cells. Hence, we examined the genotypes of several DNA DSB repair genes, such as X-Ray Repair Cross Complementing 3 (XRCC3), XRCC6, XRCC7, which are involved in the DSB repair system. The associations of these DNA repair genes with multiple types of cancer and diseases, including nasopharyngeal carcinoma [41], lung cancer [42], leiomyoma [43], breast cancer [44], hepatocellular carcinoma [45], and especially TNBC [46], have been explored in the literature.
Besides the genes involved in the DSB repair system, we are also interested in examining the contribution of the central NER repair protein, ERCC1, to the etiology of TNBC. The genotypes of ERCC1 have been shown to be associated with other types of cancer, including colorectal cancer [22], bladder cancer [47], esophageal cancer [48], but not breast cancer, not to mention TNBC. The accumulated case-control results in other types of cancer showed that the genotypes of ERCC1 may also contribute to TNBC, but this has never been investigated. In 2012, Ozkan and his colleagues demonstrated that the expression of ERCC1 was significantly elevated in approximately two thirds of the TNBC patients. More valuably, it may serve as a predictor for the poor response to platinum-based chemotherapy [49]. In 2015, Dumont and his colleagues proposed that the genotypes of ERCC1 rs11615 and CYP1B1 rs1056836 can jointly predict the prognosis responses to neoadjuvant chemotherapy of breast cancer patients, especially ER positive ones [50]. However, the sample size was small, with only 118 women, and ERCC1 rs11615 could not serve as a TNBC marker. In addition, the need to combine with CYP1B1 rs1056836 may add information to early prediction, but suggests that ERCC1 rs11615 may serve as only a low-penetrant marker, but not a high-penetrant one. From the 5-fluorouracil-, doxorubicin-and cyclophosphamide-induced DNA repair viewpoint, the genotypes of either ERCC1 rs11615 or rs3212986 may cause the differential responses to these drugs in TNBC or other subtypes of breast cancer patients [51]. The clinical study contained 324 breast cancer patients, of which the number of TNBC patients was 33, so any genotype's association to TNBC is not conclusive or representative. In 2017, the prospective role of ERCC1 to be a promising marker for Caucasians was validated by El Kashef and his team [18]. In this study, we aim to validate the contribution of ERCC1 genotypes for TNBC patients who are Taiwanese, the oriental Han population with a different genetic background from Caucasians. To fulfill this aim, we collected a large sample of 1,232 breast cancer patients in Taiwan, which strongly increased the credibility and the importance of our findings. We found that the genotypes of ERCC1 rs11615 were associated with breast cancer susceptibility, while rs3212986 polymorphism was not. In detail, T allele (or CT and TT genotypes) of ERCC1 rs11615 is a novel biomarker for Taiwanese females (Table 3). In addition, the T allele of ERCC1 rs11615 was common in patients with breast cancer (Table 4). In a most recently updated meta-analysis in 2018 including 4,547 subjects, Li et al. reported that ERCC1 rs11615 genotypes were associated with the risk of breast cancer, especially in Asian populations [52]. Among the breast cancer patients, we investigated, 104 of them were confirmed to be TNBC patients since their tissues were negative for ER, PR, and HER2/neu expression. Our findings indicate that the representation of the ERCC1 rs11615 TT genotype was increased by about 10% (from 15% to 25%) among the patients with TNBC, compared with other breast cancer patients (Table 5). Inconsistent with this result, a previous study [46] reported that the overexpressed Ki-67 was a potential indicator for TNBC, it seems that the expression of Ki-67 has no linkage with the ERCC1 rs11615 genotype in determining the susceptibility of TNBC in this study (Table 5).
In conclusion, the present case-control study, with a very large sample, indicates that the T allele of ERCC1 rs11615 may potentially serve as a powerful marker for the prediction of breast cancer, especially for TNBC. Furthermore, it is the very first time that ERCC1 rs11615 polymorphism was found to be associated with the risk of TNBC. The identification of ERCC1 genotypes among Taiwanese individuals without cancer and those who suffer from breast cancer may lead to a better understanding of the mechanisms behind breast cancer. The feasibility of the ERCC1 gene being a therapeutic target in drug development and an alternative treatment for TNBC may be quite promising.