Breast Cancer Association Studies in a Han Chinese Population using 10 European-ancestry-associated Breast Cancer Susceptibility SNPs

Background: Genome-wide association studies (GWAS) have identified various genetic susceptibility loci for breast cancer based mainly on European-ancestry populations. Differing linkage disequilibrium patterns exist between European and Asian populations. Methods: Ten SNPs (rs2075555 in COL1A1, rs12652447 in FBXL17, rs10941679 in 5p12/MRPS30, rs11878583 in ZNF577, rs7166081 in SMAD3, rs16917302 in ZNF365, rs311499 in 20q13.3, rs1045485 in CASP8, rs12964873 in CDH1 and rs8170 in 19p13.1) were here genotyped in 1009 Chinese females (487 patients with breast cancer and 522 control subjects) using the Sequenom MassARRAY iPLEX platform. Association analysis based on unconditional logistic regression was carried out to determine the odds ratio (OR) and 95% confidence interval (95% CI) for each SNP. Stratification analyses were carried out based on the estrogen receptor (ER) and progesterone receptor (PR) status. Results: Among the 10 SNPs, rs10941679 showed significant association with breast cancer when differences between the case and control groups in this Han Chinese population were compared (30.09% GG, 45.4% GA and 23.7% AA; P = 0.012). Four SNPs (rs311499, rs1045485, rs12964873 and rs8170) showed no polymorphisms in our study. The remaining five SNPs showed no association with breast cancer in the present population. Immunohistochemical tests showed that rs2075555 was associated with ER status; the AA genotype showed greater association with ER negative than ER positive (OR = 0.54, 95% CI, 0.29–0.99; P = 0.046). AA of rs7166081 was also associated with ER status, but showed a greater association with ER positive than negative (OR = 1.59, 95% CI = 1.04–2.44; P = 0.031). However, no significant associations were found among the SNPs and PR status. Conclusion: In this study using a Han Chinese population, rs10941679 was the only SNP associated with breast cancer risk, indicating a difference between European and Chinese populations in susceptibility loci. Therefore, confirmation studies are necessary before utilization of these loci in Chinese.


Introduction
Breast cancer is one of the most common malignancies in females, and causes serious health damage to females worldwide. In European and American countries, the incidence of breast cancer is ~14% (Jemal et al., 2006). In China, the incidence of breast cancer is increasing annually (Yu et al., 2007). Moreover, compared to European and American patients, the onset of breast cancer begins 10 to 15 years earlier in Chinese patients.
The cause of breast cancer has not yet been fully elucidated. Embryonic mutations account for about 20-40% of the hereditary breast cancer cases (Walsh et al., 2007). Sporadic breast cancer is often caused by genetic susceptibility and environmental factors that interact with each other. Because various mutations explain a proportion of breast cancers, widespread genetic polymorphisms in a population results in individuals with differing breast cancer susceptibility risks. Gene polymorphisms result in various responses when humans encounter cancer-causing toxins, which can inhibit or accelerate the occurrence of breast cancer. Genetic polymorphisms are a key component of the differences in breast cancer susceptibilities among individuals. Therefore, such loci polymorphisms could be used as biomarkers for developing individualized prevention strategies. While the role of one susceptibility locus might be limited, multiple loci might have a cascading effect; for example, the risk of breast cancer is increased by approximately six times in females carrying 14 common risk alleles compared to those carrying one (Pharoah et al., 2008). Therefore, susceptibility gene polymorphism loci play an important role in breast cancer. Genome-wide association studies (GWAS) have identified various genetic susceptibility  (Raskin et al., 2008;Stacey et al., 2008;Walker et al., 2010;Fergus et al., 2012). Differing linkage disequilibrium patterns and environmental factors exist between European and Asian populations, and thus GWAS-identified single nucleotide polymorphisms (SNPs) in one population may not show significance in another (McCracken et al., 2007;Long et al., 2010). Thus, a suitable breast cancer assessment risk model must be developed in Chinese female patients. This study evaluated associations between variant loci and breast cancer using rs2075555 in COL1A1, rs12652447 in FBXL17, rs10941679 in 5p12/MRPS30, rs11878583 in ZNF577, rs7166081 in SMAD3, rs16917302 in ZNF365, rs311499 in 20q13.3, rs1045485 in CASP8, rs12964873 in CDH1 and rs8170 in 19p13.1, which have been previously reported in GWAS of European-ancestry populations. GWAS must be conducted in non-European populations to determine the genetic basis of breast cancer susceptibility.

Study subjects
A total of 487 breast cancer cases and 522 cancerfree controls were recruited from Nanfang Hospital of Southern Medical University, Guangdong, China (248 female patients) and Affiliated Hospital of Medical College of Qingdao University, Shandong, China (239 female patients) from 2009 to 2012. All breast cancer cases were confirmed histopathologically without restriction of age or histological type, and no patients had a history of other cancers. The histological information included estrogen receptor (ER) and progesterone receptor (PR) status, as determined immunohistochemically. Cancerfree controls were recruited randomly from Nanfang Hospital (306 females) and Affiliated Hospital of Qingdao University (216 females). All subjects were unrelated Chinese females of Han ancestry.

Genotyping
After obtaining informed consent, 5-ml peripheral blood samples were collected and delivered to the laboratory in the frozen state. Genomic DNA was extracted from a 200-μl peripheral blood sample using the Tiangen™ Genomic DNA Kit (Tiangen, Beijing, China) according to the manufacturer's instructions and stored at -80°C until use. Ten SNPs were selected based on data from the 1000 Genomes Project (http://www.1000genomes.org/). All of the SNPs were genotyped using the SEQUENOM MassARRAY matrix-assisted laser desorption ionizationtime of flight mass spectrometry platform (Sequenom, San Diego, California, USA). Primers were designed using a semiautomated method (Assay Design 3.1, Sequenom). The call rate for each assay was set at >90%.

Statistical analysis
The Hardy-Weinberg equilibrium (HWE) was calculated using Haploview 4.2 (Daly Lab, Cambridge, MA). The genotype and allele distributions between case and control subjects were compared. Association analysis based on unconditional logistic regression after  The P value is counted by the web-based tool SNPstats, the corresponding OR is counted after age adjustment age-adjustment in case-control study was carried out by calculating the odds ratio (OR) and 95% confidence interval (CI) for each SNP in codominant and dominant genetic models. Further stratifying analysis was conducted based on data ER or PR status. The statistical tests were implemented using the web-based tool SNPstats (http:// bioinfo.iconcologia.net/SNPstats), and the significance level was set at 0.05.

Results
Quality control for all SNPs with Hardy-Weinberg Equilibrium (HWE) P>0.05 in healthy controls, genotyping rate >80% and minor allele frequency (MAF) >0.01 was conducted. Statistical power of our study was shown in (Table 1). The MAF and OR was assumed at 0.03-0.30 and 1.1-2.0, respectively, and the power ranged from 0.0661 to 0.9999. In our study, 5 of 10 polymorphisms had a MAF greater than 0.15, only one SNP lower than 0.06. Therefore, the statistical power of this study was considered to be adequate.
Case patients and control subjects were comparable regarding age and geographical regions using a frequencymatched design ( Table 2). The mean ages of patients and control subjects were 48.20 ± 9.850 and 46.99 ± 9.647 years, respectively. Additionally, the hormonal receptors of the case were distributed as follows: ER-positive in 288 cases, ER-negative in 199 cases, PR-positive in 273 cases and PR-negative in 214 cases (Table 2).
Of the 10 SNPs analyzed in the present case-control study, rs10941679 showed significant differences between the case and control groups (  The P value is counted by the web-based tool SNPstats, the corresponding OR is counted after age adjustment DOI:http://dx.doi.org/10.7314/APJCP.2014.15.1.85 Breast Cancer Associations in a Han Chinese Population of 10 European-ancestry-associated SNPs two significant SNPs (Table 4). Compared to the CC-CA genotypes, the AA of rs2075555 in COL1A1 showed greater association with negative ER than with positive ER status (OR = 0.54; 95% CI, 0.29-0.99; P = 0.046). rs7166081 in SMAD3 also showed significant associations with ER status, but was associated with positive ER status in the recessive (OR = 1.59; 95% CI = 1.04-2.44; P = 0.031), codominant (OR = 2.01; 95% CI =1.22-3.33; P = 0.022) and dominant models (OR=1.64; 95% CI = 1.09-2.44; P = 0.017; Table 4). Although no significant association was found among these six SNPs and the PR status, marginal associations with PR status were observed among rs2075555 (OR = 0.57; 95% CI = 0.31-1.05; P = 0.069) and rs12652447 OR = 1.56; 95% CI =0.99-2.45; P = 0.052) in the recessive model (Table 5). We also evaluated the association between the SNPs and breast cancer risk in each subgroup which were stratified by the ER or PR status. Three breast cancer-associated SNPs, rs2075555, rs10941679 and rs7166081, in general, showed a stronger association with ER/PR-positive tumor than with ER/PR-negative (P<0.05) ( Table 4, 5).

Discussion
Breast cancer is one of the most common cancers, and its incidence is ranked first among cancers in females (Jemal et al., 2011). Compared with western countries, China has a lower incidence of breast cancer (http:// globocan.iarc.fr/factsheets/cancers/breast.asp). However, due to industry and lifestyle changes, the incidence of breast cancer in China is increasing, and breast cancer is being diagnosed in younger patients. In recent years, studies involving low-penetrance allelic variants have been conducted mainly using GWAS (Zheng et al., 2009; The P value is counted by the web-based tool SNPstats, the corresponding OR is counted after age adjustment Long et al., 2010;Haixin et al., 2013). These studies employed a large number of common genetic SNPs to identify associations with disease that rely on linkage disequilibrium (LD) patterns in the human genome. GWAS provides a powerful tool to identify novel markers for susceptibility and prognosis of disease. It has been used to identify new breast cancer susceptibility alleles by evaluating the association of genetic variants at various loci on different chromosomes (LD) in a large number of cases and controls by analyzing a panel of 1000 SNPs simultaneously. However, mixed populations may cause a potential false-positive type I error expansion due to multiple comparisons and other issues. High-quality validation studies are necessary to ensure that real associations with the disease are identified. Furthermore, as the majority of GWAS have been performed in European populations, the replication of GWAS-identified loci in specific Asian populations such as the Chinese or Japanese is also necessary (Easton et al., 2007;Hunter et al., 2007;Sueta et al., 2012). Therefore, we performed a verification study in a Chinese Han population using 10 susceptibility SNPs that had been reported previously in European populations. Stacey et al. reported that the genetic susceptibility locus rs10941679 in chromosome 5p12 was associated with breast cancer risk (Stacey et al., 2008). Another study showed that this locus was associated with breast cancer risk only among females with more than 30 years of menstruation (odds ratio = 1.15, 95%CI: 1.05-1.26) (Zheng et al., 2009). In the present study, rs10941679 in 5p12/MRPS30 was associated with breast cancer risk under codominant (OR = 0.60; 95% CI = 0.43-0.84), dominant (OR= 0.64; 95%CI = 0.47-0.89), and overdominant models (OR = 0.69; 95%CI = 0.52-0.91). However, a fine-mapping analysis of 5p12 from the Shanghai Breast Cancer Study (Long et al., 2010) reported that rs10941679 showed a null association breast cancer risk (P=0.07), which was inconsistent with our finding. One reason for this difference may be that the samples in the Shanghai Breast Cancer Study just from shanghai. Meanwhile, the Chinese population samples of study carried by Chan et al. may from Southern Chinese descent (P=0.74) (Chan et al., 2012). Samples of present study are from Southern and Northern of China, respectively. The inconsistent result of these studies maybe mainly caused by the difference of geographic population.
Clinically, ER and PR status is an important biomarker to predict the response to endocrine treatment, which is the first and most efficacious target treatment for breast cancer with ER-positive and PR-positive histological markers (Mario et al., 2011). Breast cancer patients with ER-and/or PR-negative tumors experience higher mortality after their diagnosis compared with females with ER-positive and/or PR-positive disease (Anderson et al., 2001;Dunnwald et al., 2007). It is thought that variants in patients negative for these receptors could be related to the poor prognosis. Therefore, it is important to identify SNPs associated with breast cancer ER and PR positive/ negative status. Our study revealed that rs2075555 in COL1A1 was associated with increased ER-negative breast cancer risk, and rs7166081 in SMAD3 appeared to be associated with ER-positive breast cancer. Using the generalized estimating equations (GEE) model of breast cancer, a previous report (Joanne et al., 2007) showed that rs2075555 in COL1A1 was associated with breast cancer (p = 8.0 × 10 -8 ). However, another study (Daniele et al., 2011) reported that rs2075555 in COL1A1 did not show an association with breast cancer risk. Walker et al. provided evidence that SNPs rs7166081 in SMAD3 was associated with increased breast cancer risk for BRCA2 mutation carriers (OR = 1.25, 95% CI= 1.07-1.45, P = 0.004) (Walker et al., 2010). In our study, although rs2075555 in COL1A1 and rs7166081 in SMAD3 variants showed no associations with breast cancer, they were associated with the ER-negative/positive status in breast cancer, which may be useful for determination of the prognosis and molecular subtyping for endocrine treatment.
Fergus et al. found that rs16917302 in the ZNF365 locus was associated with both ER-positive and ERnegative status in BRCA2 mutation carriers (Fergus et al., 2012). Daniele et al. also show that SNP 5p12-rs10941679 was significantly associated with a greater risk of PRpositive than PR-negative breast cancer (P = 0.0028), which was inconsistent with our finding. One reason for this difference may be that the consortium included large well-established cohorts in the United States and Europe. Additionally, Asian females might have different lifestyles or environmental exposures that affect their risk of breast cancer. Genetic interactions with other SNPs that differ in frequency between populations could also manifest as a heterogeneity effect.
In conclusion, this was a verification study in a Chinese Han population using 10 susceptibility SNPs reported previously in European populations. However, only one SNP (rs10941679 in 5p12) was confirmed, indicating differences between European and Chinese populations in terms of breast cancer susceptibility loci. Therefore, confirmation studies are necessary before these loci in can be applied in Chinese populations.