Evaluation of Association of LOC105371267 Polymorphisms and Breast Cancer Susceptibility

Background: LOC105371267 (also known as PR-lncRNA1) was reported to be a p53-regulated lncRNA, which played essential roles in the pathogenesis of breast cancer (BC). We aimed to investigate the potential associations between LOC105371267 polymorphisms and BC risk. Method: Totally, 555 healthy individuals and 561 BC patients were recruited. Five candidate SNPs of LOC105371267 were genotyped with Agena MassARRAY system. Odds ratio (OR) and 95% condence intervals (CIs) were applied to evaluate the relationship of LOC105371267 with BC susceptibility. Additionally, stratication analyses based on clinical features and haplotype analysis were also conducted. Results: A decreased BC risk was observed rs3931698 GG genotype (OR = 0.30, P = 0.018) and recessive genetic model (OR = 0.30, P = 0.021). Stratied analysis with age also revealed that this SNP was associated with a lower risk at age < 52 years. Meanwhile, multiple clinical characteristics, including ER and PR status and stage were all correlated with SNPs rs6499221, rs3931698, rs3852740 and rs8044565. Conclusion: Four LOC105371267 SNPs (rs6499221, rs3931698, rs8044565 and rs3852740) were found to be correlated with development of BC. Additionally, ER, PR, and stage were also linked to LOC105371267 polymorphisms, providing novel diagnostic and therapeutic targets for of BC management.


Introduction
Breast cancer (BC) has been considered as one of prevailing cancers in the worldwide and the global annual incidence of BC has been continuously increased in the developing countries over the past decades [1,2]. Although tremendous achievements have been obtained in the diagnosis and treatment for BC [3][4][5], the underlying molecular mechanisms of BC has not been fully illuminated. Currently, it is well-acknowledged that unfavorable environmental risks, the pattern and lifestyle of individuals and genetic factors such as variants are all presumably associated with the initiation of BC [6].
Many researchers have empathized that genetic variants play essential roles in the cellular signaling of BC [7]. More encouragingly, increasing attention has been concentrated on the investigating the correlation between long noncoding RNAs (lncRNAs) polymorphisms and the BC pathogenesis. Ma et al evaluated the association between BC risk and LncRNA (LINC01585) using a GWAS method and they suggested that this LncRNA probably served as a novel therapeutic target for BC [8]. Moreover, Peng et al pointed out that lncRNA MALAT1 polymorphisms were correlated with the risk of BC based on the association analyses in a Chinese Han population [9].
Notably, overwhelming evidence has demonstrated that tumor suppressor p53 play essential roles in molecular mechanisms of cancer progression [10]. Moreover, p53-regulated lncRNAs were reported to contribute to the occurrence of different types of cancers [11]. For example, Liu et al highlighted that LncRNA loc285194, a p53-regulated lncRNA, served as a tumor suppressor in colon cancer via mediating the expression of miR-211 [12]. LOC105371267 (another p53-regulated lncRNA) might participate in the breast carcinogenesis. Unfortunately, few publications investigated the underlying relationship of p53-regulated lncRNAs and BC prevalence. Here, we carried out a hospital-based case-control study to assess the presumable correlation between LOC105371267 single-nucleotide polymorphisms (SNPs) and the susceptibility to BC in a Chinese population. In addition, we also investigated the association between LOC105371267 polymorphisms and clinical characteristics of BC.

Study population
In this case-control study, blood samples were collected from 561 patients with BC (cases) and 555 healthy individuals (controls), who were consecutively recruited from Shaanxi Provincial Cancer Hospital. All patients were newly diagnosed as breast carcinoma by the histopathological examination and none of them had undergone chemotherapy or radiotherapy before gathering samples. Moreover, those who had other cancer history or suffered from immunological, cardiovascular or hematologic disorders were excluded. The control subjects received from the physical examination center in the same hospital, who had not any medical illness, not family history of BC and were genetically unrelated to the included BC patients.
Additionally, the demographic data of participants and the clinical information of BC patients were acquired based on a standard questionnaire, including age, estrogen receptor (ER), progesterone receptor (PR), Ki67 status, tumor status, location and stage, lymph nodes metastasis and distance metastasis. All participants signed informed consent, and this work was approved by the Ethics Committee of Xizang Minzu University. All experiments were conducted in accordance with the World Medical Association Declaration of Helsinki.

SNPs genotyping assay
Total DNA isolation was undertaken from 5 mL of ethylenediamine tetraacetic acid (EDTA) -anticoagulated peripheral blood using GoldMag whole blood genomic DNA purifcation kit (GoldMag Co. Ltd., Xi an, China) according to the manufacturer′s protocol and subsequently was stored at − 80 °C for the following analysis. Five candidate SNPs of LOC105371267, including rs6499221, rs3931698, rs8044565, rs3852740 and rs111577197, were identi ed based on two databases the 1000 Genomes Project database (https://www.ncbi.nlm.nih.gov/variation/tools/1000genomes/) and dbSNP database (http://www.ncbi.nlm.nih.gov/proje cts/SNP/) using with the minor allele frequency (MAF) > 0.05 in Chinese Han population and call rate > 95% [13]. Moreover, functional prediction analysis of these SNPs were performed with web-based HaploReg v4.1 software (https://pubs.broadinstitute.org/mammals/haploreg/haploreg.php). Subsequently, these SNPs genotyping were carried out with the Agena MassARRAY system (Agena, San Diego, CA, U.S.A.) as described in a previous research [14] by two independent investigators. In addition, 10% of samples were randomly selected as blinded duplicates to evaluate the accuracy of SNP genotyping and exhibited 100% concordance.

Statistical analyses
The differences in demographic and clinical data between cases and controls were assessed by Pearson′s χ 2 test and Student t test. Hardy-Weinberg equilibrium (HWE) analyses for each SNP among controls were conducted by Fisher's exact test.
Pearson′s χ 2 test was also used to analyze the difference in allelic and genotype frequencies for each polymorphism between BC patients and healthy subjects. Accordingly, odds ratios (ORs) and their 95% con dence intervals (CIs) were calculated by logistic regression analysis after the adjustment for age to evaluate the correlation between LOC105371267 polymorphisms and BC risk using PLINK v1.07 software. Meanwhile, several genetic models (dominant, recessive and additive model) were utilized to estimate the relationship of LOC105371267 SNPs with the susceptibility to BC. Moreover, we performed multiple strati ed analyses in terms of age, ER, PR, lymph nodes metastasis, stage, Ki67 status, and tumor size. Additionally, the pairwise linkage disequilibrium (LD) were measured by the LD coe cient D' using the Haploview v4.2 software. Haplotype analysis was conducted by logistic regression analysis using the PLINK v1.07 software. All statistical analyses were carried out by SPSS v 18.0 software (Armonk, New York City, NY, USA) and two -sided P values were considered statistically different.

Characteristics of study population and SNP identi cation
A total of 1116 participants (561 BC patients and 555 controls) were recruited in the current study and baseline characteristics of these subjects were exhibited in Table 1. We noted that no signi cant difference was detected between cases and controls (P < 0.05) in terms of age. Five candidate SNPs of LOC105371267 (rs6499221, rs3931698, rs8044565, rs3852740 and rs111577197) were screened according to the criteria described above and successfully genotyped in included samples. The fundamental information of these SNPs was displayed in Table 2 and the genotypes frequency of all SNPs in control groups conformed to HWE (P > 0.05). Moreover, there were no signi cant differences in allele frequencies between patients and healthy controls (P > 0.05), implying that these SNPs were not correlated with the susceptibility to BC under the allele model.  Associations between LOC105371267 polymorphisms and BC risk The logistic regression model was employed to evaluate the associations between LOC105371267 SNPs and the risk of BC based on the adjustment with age (

Strati cation analysis
The strati ed analyses were subsequently performed to explore the correlation of BC susceptibility and multiple clinicopathologic indicators. As indicated in Table 4, the G allele of rs3931698 exhibited signi cant association with BC risk at age < 52 years in GG genotype (OR = 0.26, 95% CI: 0.71-0.97, P = 0.045) and additive model (OR = 0.68, 95% CI: 0.48-0.95, P = 0.025).We also found that AG genotype of rs6499221 was associated with BC risk under the age of 52 (OR = 1.48, 95% CI: 1.01-2.09, P = 0.046) For other SNPs, no signi cant associations were detected between age and the susceptibility to BC. In addition, rs3582740 was associated with a lower risk of ER-positive BC (OR = 0.73, 95% CI: 0.53-0.99, P = 0.043; Table S4) while rs6499221 raised the incidence of ER-positive patients under the additive genetic model (OR = 1.43, 95% CI: 1.02-2.02, P = 0.041; Table S1). The G allele of rs3931698 was correlated with a higher risk of PR-positive BC (GT vs TT: OR = 1.52, 95% CI: 1.01-2.29, P = 0.043; Table S2) whereas there was no correlation between other LOC105371267 polymorphisms and PR status (Table S1 and 3-5).
Afterwards, we further evaluated the impact of LOC105371267 SNPs on the severity of BC according to TNM staging (III-IV/I-II).
The results revealed that GT genotype of rs3921698 was overrepresented in patients with clinical III-IV stage compared to those with I-II stage (OR = 1.58, 95% CI: 1.04-2.40, P = 0.033; Table S2) and there was no correlation between other SNPs and TNM stage (Table S1 and [3][4][5]. Additionally, no statistical difference was estimated between selected ve SNPs in LOC105371267 and tumor size, Ki-67 status and lymph nodes metastasis based on the strati cation analyses (Table S1-5).

Haplotype analysis of LOC105371267 polymorphisms
The linkage disequilibrium (LD) and corresponding haplotypes of LOC105371267 SNPs were further investigated by Haploview software. Our ndings implied that SNPs rs3931698 and rs8044565 were in high LD block and formed three haplotypes (TC, GT and TT) (Fig. 1). Furthermore, none of haplotypes was related to the incidence of BC (P > 0.05, Table 5). Haplotypes were identi ed with the order of rs3931698 and rs8044565. P values, odd ratios and their 95% CI were estimated by logistic regression models with the adjustment for age.

Discussion
Recently, numerous researchers have concentrated on elucidating the correlations between lncRNAs and the susceptibility to BC. For example, Li et al carried out a GWAS-based association analysis between lncRNAs and BC prevalence, which suggested that lncRNAs polymorphism was linked to a higher BC risk probably via in uencing microRNA-mediated regulation in cell processes [15]. In this work, we performed the association analyses between LOC105371267 polymorphisms and the risk of BC based on a Chinese population. Five candidate SNPs (rs6499221, rs3931698, rs8044565, rs3852740 and rs111577197) were successfully genotyped. We found that carriers with rs3931698-G allele might have a lower incidence of BC. Strati ed by age, rs6499221 increased BC risk while rs3931698 reduced the risk at age < 52 years. Meanwhile, a higher risk was observed between rs3931698SNP and other two clinical indicators (PR status and stage). Rs6499221 and rs3852740 polymorphisms showed a decreased risk in ER-positive patients. Therefore, we speculated that LOC105371267 with SNPs (rs6499221, rs3931698, rs8044565 and rs3852740) might be responsible for the occurrence and development of BC. However, no signi cant relationship was found between LOC105371267 rs111577197 and BC prevalence.
LOC105371267 (also known as PR-lncRNA1) was reported to be a p53-regulated lncRNA. Sánchez et al highlighted that LOC105371267 could enhance cell apoptosis and cell cycle arrest by promoting the p53 signaling activation. Speci cally, they argued that PR-lncRNA1 regulated the p53 transcriptional network by the e cient binding of p53 to some of its target genes [16]. Furthermore, Li et al previously also pinpointed that PR-lncRNA1 interacted with a sequence-speci c RNA binding protein Sam68 and this complex could promote the p53-mediated transcription in human colon carcinoma cell lines [17]. These lines of evidence have led us to formulate the hypothesis that PR-lncRNA1 could be of pathogenic importance in BC. Our results rstly revealed that LOC105371267 polymorphisms were associated with the susceptibility to BC.
Age has been identi ed as a prominent risk factor in the BC initiation [18]. An early study suggested that BC patients with the oldest age were more vulnerable to rapid deterioration [19]. Moreover, Unlu et al highlighted that older women tend to have a higher BC risk compared with those younger women [20]. In the study, we found that PR-lncRNA1 SNP rs6499221 and rs3931698 were related to the risk of BC patients at age < 52 years. Additionally, several clinicopathological characteristics, including ER, PR, Ki-67, metastasis, stage and tumor size were also observed to participate in the BC pathogenesis [21][22][23]. Our study found that LOC105371267 polymorphisms might be associated with ER, PR, and stage of BC.
Although the association of four SNPs in PR-lncRNA1 with BC risk and several clinicopathological characteristics have been identi ed in the present work, there are still limitations. On the top of that, due to all participants were all enrolled in the same hospital and were Chinese Han population, the inherent selection bias cannot be excluded and our results cannot permit extrapolation of the results to other ethnic groups. In addition, the comprehensive clinical information and environmental factors should be included. Moreover, the precise molecular mechanisms of PR-lncRNA1 polymorphisms in BC progression remain to be deciphered. Despite the limitations mentioned above, the results of our study might provide evidence for the future studies about LOC105371267 with BC.