The association between polymorphisms in microRNA genes and cervical cancer in a Chinese Han population

Several studies have confirmed the crucial roles of microRNAs (miRNAs) in cancer occurrence. In addition, single nucleotide polymorphisms (SNPs) in miRNA genes have been associated with various cancers. The aim of the present study was to investigate the association of SNPs in miRNA genes with cervical intraepithelial neoplasia (CIN) and cervical cancer in a Chinese Han population. We searched SNPs in nineteen miRNAs by sequencing healthy individuals (n=50). Then, a total of 400 patients with CIN, 609 patients with cervical cancer and 583 healthy individuals were recruited to genotype the SNPs using a Taqman assay. The results showed that only five of the nineteen miRNAs had SNPs (rs11134527 in pri-miR-218-2; rs74693964 in pri-miR-145; rs6062251 in pri-miR-133a2; rs531564 in pri-miR-124-1; and rs1834306 in pri-miR-100) in this Chinese Han population. The frequency of the rs11134527A allele was significantly higher in the control group than in CIN and cervical cancer groups (P=0.011 and 0.035, respectively). The frequency of the rs531564G allele was higher in the CIN and control groups than in the cervical cancer group (P=0.019 and 0.017, respectively). These results indicated that rs11134527 in pri-miR-218-2 and rs531564 in pri-miR-124-1 could be associated with CIN and cervical cancer in the Chinese Han population.


INTRODUCTION
Cervical cancer is one of the most common malignancies in women worldwide, with an estimated global incidence of 470,000 new cases and approximately 233,000 deaths per year [1]. Compared to developed countries, the incidence and mortality of cervical cancer is higher in low-resource countries [2]. High-risk human papillomavirus (HPV) infections are the main aetiological factors of cervical intraepithelial neoplasia (CIN) in women, and persisting high-risk HPV infections significantly facilitate the development of CIN to cervical cancer. However, evidence suggests that host genetic factors are also involved in the initiation and development of cervical cancer [3][4][5].
Recent studies have demonstrated the aberrant expression of miRNAs in cervical cancer and their contribution to tumour generation [11]. Functional studies have shown that miRNAs impact tumour cell fate through the regulation of various cell signalling pathways [12]. Thus, miRNAs act as oncogenic or anti-oncogenic molecules in the initiation and development of cervical cancer [13]. For example, miR-218 functions as a tumour suppressor through targeting focal adhesion pathways and the SLIT2-ROBO1 pathway, and inhibiting cancer cell migration and invasion. Single nucleotide polymorphisms (SNPs) in miRNA genes have been associated with various human diseases [14][15][16]. In 2016, Wang et al. reported that rs767649 in miR-155 is associated with cervical cancer [17].
In the present study, we searched for SNPs in nineteen miRNAs associated with cervical cancer [18]. Subsequently, we conducted an association study in patients with CIN, patients with cervical cancer and healthy individuals to investigate the relationship of the SNPs in miRNAs with CIN and cervical cancer in a Chinese Han population. Table 1 lists the characteristics of subjects in the present study. There was no significant difference in age between CIN, cervical cancer and control groups ( p>0.05). In the cervical cancer group, there were 91 individuals with adenocarcinoma (AC), 501 individuals with squamous cell carcinoma (SCC) and 17 individuals with other pathological types.

Model of inheritance analysis
Tables 4-6 present the results of the analyses of different models of inheritance for the five SNPs. The Akaike information criterion (AIC) and Bayesian

DISCUSSION
Evidence has shown that miRNAs are involved in the initiation and development of various types of cancers, and these molecules function as antioncogenes or oncogenes. SNPs in miRNAs, including primary, precursor and mature miRNA, may change the expression level and alter the binding affinity of miRNAs, and further regulate the expression of the target genes, thereby markedly impacting cell pathway regulation, which defines cellular fates [19,20]. In the present study, we observed that rs11134527 in pri-miR-218-2 and rs531564 in pri-miR-124-1 were associated with CIN and cervical cancer in a Chinese Han population.
In 2013, Yamamoto N et al. reported that the expression of miR-218 was significantly lower in cervical cancer tissues than in adjacent non-cancerous tissues, and the restoration of miR-218 significantly inhibited cancer cell proliferation, migration and invasion [21]. In addition, the anti-oncogene function of miR-218 has been reported in several types of cancers through targeting several oncogenes, such as Rictor (oral cancer), EGFR (non-small cell lung cancer), ROBO1 receptor (gastric cancer) and LAMB3 (cervical cancer) [22][23][24][25]. Several studies have investigated the association of rs11134527 with different diseases (Table 7) [26][27][28]; however, these studies provided contradictory results. In the present study, rs11134527 in pri-miR-218-2 was associated with CIN  [28] whose study showed that the G/G genotype was associated with a decreased risk of cervical cancer compared with AA and A/G genotypes. The discrepancy between these results might reflect different sample sizes and populations in the two studies, and unfortunately, the Shi et al. study did not focus on the association with CIN, thus, we could not compare the results between CIN and the control. In the present study, the allelic frequencies of rs11134527 were only significantly different in the comparison of CIN and cervical cancer groups with control group, but not in the comparison of CIN group with cervical cancer group.
Recent studies have reported that miR-124 suppresses a series of genes involved in the growth of human cancers. MiR-124 suppresses tumour growth through targeting STAT3 in colorectal cancer [29], CD151 in breast cancer [30], and ERK in cutaneous squamous cell carcinoma [31]. In cervical cancer, Wan et al. reported that miR-124 represses vasculogenic mimicry and cell motility by targeting amotL1 [32], and Zhang et al. showed that miR-124 inhibits the proliferation, invasion, migration and epithelial-mesenchymal transition of cervical cancer [33]. As an anti-oncogene, significantly lower miR-124 expression was detected in cervical cancer than in lowgrade squamous intraepithelial lesions [34]. Studies have shown that SNPs in miRNA genes might alter miRNA expression or function [35,36]. Thus, SNPs in miR-124 might be associated with miR-124 expression and are further associated with cervical cancer. Rs531564 has been significantly correlated with the risk of various cancers (Table 7), including cervical cancer and oesophageal squamous cell carcinoma (ESCC) [26,37,38]. In the present study, rs531564 in pri-miR-124-1 was associated with cervical cancer risk (P=0.017, OR=1.341; 95%CI: 1.054-1.706). Additionally, this SNP was correlated with the progression from CIN to cervical cancer (P=0.019; OR=1.385; 95%CI:1.054~1.819). These results were   MiR-145, a tumour suppressor, has been implicated in the death-promoting regulatory loop of P53 [39]. The downregulation of miR-145 has been associated with aggressive progression and poor prognosis in cervical cancer [40]. In 2015, Chacon-Cortes et al reported the association of rs353291 in miR-145 with breast cancer susceptibility [41]. However, in the present study, we investigated the association of rs74693964, another SNP in near Gene-3 of miR-145, with cervical cancer in a Chinese population; no significant association of rs74693964 with CIN and cervical cancer was detected.
MiR-133a plays a crucial role in myoblast proliferation and differentiation. Recent studies have shown the downregulation of miR-133a in human cancers. Accumulating evidence has indicated that the restoration of miR-133a expression could induce the suppression of cell proliferation, migration and invasion, and tumour cell apoptosis [42][43][44], indicating that miR-133a might act as a tumour suppressor in human cancers, including cervical cancer [45]. Zhou et al reported that rs8089787 and rs9948906 in miR-133a1 were associated with the risk of asthma in a Chinese Han population [46]. Paula et al analysed the association of rs13040413 and rs200375711 in miR-133a2 with long QT syndrome and observed no association between the two SNPs with long QT syndrome [47]. In the present study, we investigated the association of rs6062251 in the near Gene-3 of miR-133a2 with the initiation and development of cervical cancer. These results showed that the genotypic frequencies were significantly different between control, CIN and cervical cancer groups. The inheritance analysis showed that T/T-C/C genotype were correlated with increased risk of progression from CIN to cervical cancer in the overdominant model (P=0.007, OR=1.45; 95%CI: 1.11-1.90).
In recent years, miR-100 has been reported to target numerous biomolecules associated with carcinogenesis, suggesting that miR-100 functions as both a tumour promoter and a tumour suppressor. In cervical cancer, miR-100 functions as a potential tumour suppressive miRNA involved in tumour occurrence, development and drug resistance [48]. Li et al. [49] reported the reduced expression of miR-100 in CIN and cervical cancer. Zhu et al. reported that a SNP in miR-100 was associated with various types of diseases (Table 7) [50]. In the present study, we did not observe any association of rs1834306 in pri-miR-100 with the initiation and development of cervical cancer in a Chinese Han population. Similarly, Xiong et al. detected no relationship between rs1834306 in pri-miR-100 and cervical cancer risk in a Chinese Han population (Table 7) [51].
In current study, our results showed the allele or genotype of given miRNA played different roles in normal tissue, CIN and/or cervical cancer. The reason for the discrepancies could be the onset of cervical cancer is a multistep process that means progression of cervical cancer is from normal tissue to invasive cervical cancer through a series increasing grades of squamous intraepithelial lesions [52]. Among this multistep process, the miRNAs expression and their targets have shown discrepancies. For example, the expression of miR-145, miR-218 and miR-100 was progressive reduced and down-regulated from CIN 1 to 3 and cervical cancer versus normal tissue [53][54][55]. Furthermore, the miRNA expression alternation could change their targets function. In 2012, Qin et al. reported that miR-133b expression in CIN 2, 3 and cervical carcinoma had a gradual raise, which increased AKT and MAPKs (ERK1 and ERK2) phosphorylation augmenting tumorigenesis [56]. Their finding indicated miRNAs could play different roles in the progression of cervical cancer. Therefore, the different alleles or genotypes of SNPs in these miRNAs could function differently in the multistep stages of cervical cancer by altering their expression level and the binding affinity between them and their targets.
In summary, we investigated the association of the SNPs in five anti-oncogene miRNAs (pri-miR-218-2, pri-miR-145, pri-miR-133a2, pri-miR124-1 and pri-miR-100) with the CIN and cervical cancer in a Chinese Han population. The results showed that the A allele of rs11134527 in pri-miR-218-2 maybe associated with

Ethics statement
The present study was approved through the Institutional Review Board of the 3rd Affiliated Hospital of Kunming Medical University. The methods of this investigation were performed in accordance with the approved guidelines and principles expressed in the Helsinki Declaration of 1975, which was revised in 2008. All participants provided written informed consent.

Subjects
A total of 400 CIN and 609 cervical cancer patients were recruited at the 3rd Affiliated Hospital of Kunming Medical University from 2011 to 2015. The diagnosis of CIN and cervical cancer was confirmed by pathological examination according to "Diagnosis and Treatment Obstetrics and Gynaecology" and FIGO stage (International Federation of Gynaecology and Obstetrics, 2009). Subjects with oncotherapy history or other malignancy or incomplete clinical data were excluded from the present study. During the same period, 583 healthy women undergoing routine health check-ups at the same hospital were designated as the control group. The inclusion criteria of control subjects were HPV-negative women without any cervical lesions. All participants were self-reported as ethnically Han.

Genomic DNA extracted
Genomic DNA was extracted from peripheral lymphocytes using the QIAamp Blood Mini Kit (Qiagen, Hilden, German).

SNP genotyping using TaqMan assay method
SNPs, namely, rs11134527 in pri-miR-218-2, rs74693964 in pri-miR145, rs6062251 in pri-miR-133a2, rs531564 in pri-miR-124-1 and rs1834306 in pri-miR-100, were genotyped through PCR amplification using the TaqMan assay. The TaqMan assays including the primers and probes were designed and ordered by Applied Biosystems (https://bioinfo.appliedbiosystems.com/ genome-database/snp-genotyping.html). Briefly, these SNPs were genotyped in 384-well plates using TaqMan assays in the QuantStudio 6 Flex Fast Real-Time PCR System. Amplification was performed in a 5-μl reaction volume consisting of 2.5 μl 2 × TaqMan Master Mix, 0.125 μl 40 × Primer and TaqMan Probe (FAM VIC) dye mix, 1.375 μl ddH 2 O, and 1 μl template DNA (substituted by equivalent ddH 2 O in negative control). The following PCR cycle conditions were used: 95°C for 10 min; PCR stage 92°C for 10 s and 60°C for 1 min repeated in 40 cycles. Data acquisitions and analysis were performed on QuantStudio™ real-time PCR software. To identify the accuracy of SNP genotyping using the TaqMan assay, three positive controls and one negative control were simultaneously genotyped using the TaqMan assay.

Statistical analysis
The allelic and genotypic frequencies of the SNPs were calculated using a direct-counting method. Hardy-Weinberg equilibrium (HWE) was tested for the SNPs in the control group. A χ 2 test was used to determine differences in allelic and genotypic frequencies, and the odds ratios (OR) with associated 95% confidence intervals (CIs) of allele and genotype-specific risks were calculated. The ages among different groups were compared using one-way ANOVA, followed by the LSD test for multiple comparisons correction. The association of each SNP with CIN and cervical cancer was analysed for the model of inheritance using SNPStats software [57]. The Akaike information criterion (AIC) and Bayesian information criterion (BIC) were used to determine the best-fit model for each SNP. The model with the smallest AIC and BIC values corresponds to minimal expected entropy. The statistical analyses were performed using SPSS 13 (Chicago, IL). A P value less than 0.05 was considered statistically significant.

Author contributions
Yao Yufeng, Shi Li and Yang Hongying designed the current study; Li Chuanyin, Wang Xiaona and Zhang Yu performed the DNA extracting, sequencing and SNP genotyping; Yan Zhiling, Zhang Yu, Liu Shuyuan, Yang jie and Hong Chao collected the blood samples and the clinical data; Li Chuanyin drafted the manuscript; Yao Yufeng, Shi Li and Yang Hongying revised the manuscript critically. All authors agreed the final version of the manuscript.