Genetic variants within the cancer susceptibility region 8q24 and ovarian cancer risk in Han Chinese women

Accumulating evidence suggests that genetic variants at chromosome 8q24 confer susceptibility to various types of cancer. This case-control study was designed to explore the relationship between genetic variants at 8q24 and ovarian cancer risk in Han Chinese women. Two variants (rs13281615 A > G and rs6983267 T > G) were genotyped in 377 ovarian cancer cases and 1034 cancer-free controls using TaqMan allelic discrimination assay. Logistic regression analysis revealed that the G allele of rs6983267 was significantly associated with increased risk of ovarian cancer (additive model: adjusted OR = 1.21, 95% CI = 1.01–1.43, P = 0.048; recessive model: adjusted OR = 1.51, 95% CI = 1.06–2.15, P = 0.023). However, no significant association was observed between rs13281615 and ovarian cancer. In stratified analysis, the risk effect of rs6983267 variant remained significant in premenopausal women (additive model: adjusted OR = 1.62, 95% CI = 1.18–2.23, P = 0.003). Summarily, this study suggested that 8q24 rs6983267 may contribute to the susceptibility of ovarian cancer in premenopausal Han Chinese women, supporting the pleiotropy of 8q24 in carcinogenesis.


INTRODUCTION
Ovarian cancer is the leading cause of cancer mortality among global women, with an estimation of 240,000 new cases and 150,000 deaths in 2012 [1]. Up to now, the specific etiology of ovarian cancer remains to be determined. It has been proposed that environmental and behavioral factors may play significant roles in ovarian cancer, such as occupation [2], parity [3], oral contraceptive use [4], and family history of cancer [5]. Furthermore, growing evidence has suggested that genetic variants may also contribute to the risk of ovarian cancer [6].
A segment of 1Mb within chromosome 8q24 has emerged as a susceptibility region for multiple cancers, including prostate, colon, breast, and ovarian cancers, in genome-wide association studies (GWAS) [7][8][9][10]. This region contains no known genes but is bounded at its telomeric end by oncogenes c-MYC and PVT1. Overexpression of c-MYC occurs in various types of cancer including ovarian cancer, and reduction of c-MYC expression by RNA interference can inhibit tumor growth [11]. An association between c-MYC expression and risk allele of rs6983267 at 8q24 was revealed in colorectal cancer. In addition, a noncoding RNA PVT1 plays an important role in carcinogenesis, and the increase of PVT1 expression was shown to relate with the GG genotype of rs13281615 at 8q24 [12].
Despite the critical role of 8q24 variants in cancer susceptibility, there has been no study examining the relationship between 8q24 and the risk of ovarian cancer in Chinese women. Therefore, we conducted this casecontrol study with 377 cases and 1034 healthy controls Research Paper www.impactjournals.com/oncotarget to explore the association of two potentially functional variants, rs6983267 and rs13281615 at 8q24 with ovarian cancer risk in Han Chinese women.

RESULTS
Demographic and clinical characteristics of 377 ovarian cancer cases and 1034 cancer-free controls are presented in Table 1. The primary pathological type of cases was epithelial ovarian cancer (86.7%). A total of 103 (27.3%) patients were diagnosed at stage I or II, and 274 (72.7%) were at stage III or IV. The age, occupation, and family history of cancer were comparable between cases and controls (all P > 0.05). However, ovarian cancer cases were more likely to have early menarche age, abortion history, and postmenopausal status, but less likely to use oral contraceptives than cancer-free controls (all P < 0.05).
The genotype distributions of rs13281615 and rs6983267 are shown in Table 2. Calling rates of the two variants were above 99% and their genotype frequencies in the control group were consistent with Hardy-Weinberg equilibrium (P > 0.05). No obvious linkage disequilibrium between the two variants was detected by LD analysis (D'=0.155, r 2 =0.015). A significant association was observed between rs6983267 and ovarian cancer risk (additive model: adjusted OR = 1.21, 95% CI = 1.01-1.43, P = 0.048; recessive model: adjusted OR = 1.51, 95% CI = 1.06-2.15 P = 0.023). However, there was no significant association between rs13281615 and ovarian cancer risk.
Furthermore, stratified analysis of the associations was performed according to demographic and clinical variables (Table 3 and Supplementary Table 2). The risk effect of rs6983267 on ovarian cancer was shown to be significant in premenopausal women (additive model: adjusted OR = 1.62, 95% CI = 1.18-2.23, P = 0.003), with significant heterogeneity in strata of menopausal status (P = 0.037).

DISCUSSION
In the current study, we investigated the relationship between two variants (rs6983267 and rs13281615) at 8q24 identified by the GWAS of multiple cancers and ovarian cancer risk. We revealed that the G allele of rs6983267 was significantly associated with increased risk of ovarian cancer in premenopausal Han Chinese women.
Chromosomal 8q24 has been considered the most important susceptibility region for various types of malignancy, including prostate [7,13], colorectal [14][15][16], and breast cancer [9,17]. However, the mechanism by which 8q24 variants affect cancer susceptibility is not fully understood. None of identified variants reside within known genes, of which there are few across 8q24. Functional evidence has indicated that this risk region may act as a regulatory hub by physical interactions with several neighboring genes important for carcinogenesis such as c-Myc and PVT1 [18]. Oncogene c-Myc encodes phosphoproteins that participate in the regulation of cell proliferation, apoptosis, and differentiation [19]. Previous studies demonstrated that gene amplification and increased expression of c-Myc could promote the development of ovarian cancer [20,21]. PVT1 is located adjacent to c-MYC and functions as a noncoding RNA with many alternatively spliced isoforms. Both PVT1 copy number gains and overexpression have been implicated in the development of many tumors, including ovarian cancer [22]. Small interfering RNA-mediated reduction in either PVT1 or MYC expression can inhibit cellular proliferation of ovarian cancer [22].
The variant rs6983267 has been associated with both prostate and colorectal cancers in different ethnic populations [16,[23][24][25]. It is located far away from any coding sequences, and the nearest gene is c-MYC. Functional studies suggest that rs6983267 is located in a transcriptional enhancer and exhibits long-range physical interaction with c-Myc [26]. The variant has enhancerrelated histone marks and can form a 335-kb chromatin loop to interact with the c-MYC promoter, resulting in elevated c-Myc expression [27]. In addition, transpiration factors such as TCF4 may bind preferentially to the risk allele G of rs6983267, enhancing the responsiveness to Wnt signaling [28]. These findings provide a potential mechanism for the source of association between rs6983267 and ovarian cancer. However, additional studies are required to prove the interaction of rs6983267 with c-MYC and other transcription factors in the pathogenesis of ovarian cancer.
Previous studies identified rs13281615 at 8q24 as an independent susceptibility locus for breast cancer [17,29]. The GG genotype of rs13281615 was significantly associated with estrogen receptor positivity, higher tumor grade and higher proliferation index in breast cancer. In addition, PVT1 expression was elevated in breast cancer tissues and the increase was related with the GG genotype of rs13281615 [12]. Overexpression of PVT1 has been demonstrated in in a variety of cancer types including ovarian cancer [30]. Increased expression of PVT1 in ovarian cancer cells may promote cisplatin resistance by regulating apoptotic pathways [31]. In the current study, we did not find significant association between rs13281615 and ovarian cancer risk in Han Chinese women. However, insufficient statistical power due to relatively small sample size in our study may also account for the lack of association. Independent studies with larger sample size are needed to validate our results.
Several limitations need to be mentioned. First, for lacking of samples, we conducted only one stage case-control study. Second, the biological function of rs6983267 in ovarian cancer was not further investigated in this study. Despite these limitations, this study was the first to provide evidence that rs6983267 at 8q24 may contribute to the risk of ovarian cancer in Han Chinese women. Replication studies in diverse populations and functional analyses are warranted to confirm the role of 8q24 variants in ovarian cancer.

Study population
This study was approved by the institutional review board of Nanjing Medical University, and each participant signed an informed consent before the enrollment. Briefly, a total of 377 patients with ovarian cancer were recruited from the areas of relatively high incidence, including cities of Nantong, Wuxi and Nanjing in east China's Jiangsu province, as previously described [32]. All ovarian cancer cases had definite histopathological diagnosis and met the following inclusion criteria: (1) Han Chinese; (2) local residences (at least 5 years); (3) without tumor history in any other organs.
Cancer-free controls were randomly selected from a community-based cohort of over 30,000 participants for non-infectious disease screening program in Jiangsu Province, during the same period as the cases were recruited [33]. The controls were genetically unrelated Han Chinese women, having a local residence (at least 5 years) and no history of cancer. A one-on-one interview was carried out by a trained interviewer to collect

Genotyping
Genomic DNA was extracted from leukocyte pellets by traditional proteinase K digestion, followed by phenolchloroform extraction and ethanol precipitation. Variants were genotyped using the TaqMan allelic discrimination assay on ABI PRISM 7900 HT platform (Life Technologies, CarIsbad, USA). Information of the primers and probes was shown in Supplementary Table 1. The genotyping was performed without knowing the subjects' case or control status. In each 384-well reaction plate, two negative controls were added for quality control. To make sure the reproducibility of genotyping, 50 cases and 50 controls were randomly selected to be retested, yielding a 100% concordance.

Statistical analysis
Differences in demographic characteristics, exposure variables, and genotypes frequencies between the cases and controls were evaluated by the χ² test for categorical variables and Student's t test for continuous variables. Hardy-Weinberg equilibrium (HWE) was tested using goodness-of-fit χ² test. Haploview was employed to analyze linkage disequilibrium (LD) parameters (i.e., D′ and r 2 ). The association between variants and ovarian cancer risk was estimated by computing odds ratios (ORs) and their 95% confidence intervals (CIs) from logistic regression analysis in three genetic models (additive, dominant, and recessive). Each model makes different assumptions about the genetic effect, as previously described [34]. Age and statistically significant variables in univariate analysis were included in multivariate analysis. In stratified analysis, the χ 2 -based Q test was used to assess the heterogeneity of associations between subgroups. All statistical analyses were performed with Statistical Analysis System software (v9.1.3, SAS Institute, Cary, NC). A two-sided P value of less than 0.05 was used as the criterion for statistical significance.