Association of rs3746444 in MiRNA-499 and rs3787016 in Long Non-coding RNAs POLR2E with Prostate Cancer: an updated meta-analysis and systematic review


 Background

It has been proved that a significant proportion of prostate cancer diagnoses may be associated with a strong hereditary component. Micro-RNAs and Long Non-coding RNAs are a group of a noncoding small molecule, which play critical roles in signalling pathways, metabolism, apoptosis and cancer development. Exercising a meta-analysis to examine the relationship between rs3746444, rs3787016 and prostate cancer (PCa) risk was the main objective of our study.
Results

10 case-control studies were included, while 6 on rs3787016 and 4 on rs3746444. On the whole, the genotypes carrying the T allele, in which were verified an increased risk between rs3787016 and PCa risk.( T vs C: odds ratio(OR) = 1.802, 95% CI 1.015–3.198, P value of heterogeneity(PH) = 0.00%; CT vs CC: OR = 1.179,95% CI 1.091–1.275, PH = 0.41; TT vs CC: .OR = 1.418, 95% CI 1.229–1.636, PH= 0.961; TT + CT vs CC: OR = 1.216, 95% CI 1.129–1.310, PH= 0.408; TT vs CT + CC: OR = 1.328, 95%CI 1.159–1.521, PH = 0.987). A vital relevance was performed by this meta-analysis in three models between rs3746444 and PCa risk, in which a tendency of increased PCa risk could be found (C vs T: OR = 1.197, 95% CI 1.035–1.384, PH= 0.837; CC vs TT: OR = 1. 137, 95% CI 0.813–1.590, PH = 0.392; CC + CT vs TT: OR = 1.426, 95% CI 1.166–1.743, PH= 0.406.).
Conclusion

Our findings proposed that rs3746444 in MiRNA-499 and rs3787016 in Long Non-coding RNAs POLR2E polymorphisms increased the risk of developing PCa. Further functional studies are warranted to reveal the role of the polymorphism in carcinogenesis.

Background For being the second most frequent cancer diagnosis made in men, prostate cancer hits the list leading cause of death around the world at number ve [1,2]. Giving no warning signals at earlier period, prostate cancer often performed as tardy progress that may only stand in need of active monitoring. Nevertheless, the clinical conformity case was often identi ed at or toward an end or late period or stage of development with poor prognosis. With the constant upgrading of potential diagnostic biomarkers of prostate cancer and the technological innovations of surgical techniques, diagnosed in the early stage, prostate cancer can be cured by radical resection, local radiotherapy or ablation. When it turned to advanced diseases, the survival rate still has declined rapidly when bone metastasis and lymphatic metastasis occurred, although endocrine therapy and chemotherapy have made considerable progress [3]. As the subject of quite a few studies, compared to other common tumor, the development or morbid condition of prostate cancer and further study is needed. Including old aging, ethnicity, individual heredity factors and family history [2,4] both were the generally-accepted risk factors of prostate cancer. Thus, diagnosing the individuals in danger of hereditary prostate cancer and potential other factors which contribute to the early detection and handle the development of PCa was urgent for us [5]. So far, when it turned to the populations of European and Non-European lineage, in which more than 177 singlenucleotide polymorphisms (SNPs) have been corroborated a liated with PCa by a great number of genome-wide association studies. Numerous SNPs have been apparently investigated and veri ed to increase risk facilitating to the advancement of PCa [2]. Likewise, advancement of PCa maybe was caused by both various expressions of multitudinous oncogenes and cancer suppressors in which MicroRNAs (MiRNAs) or Long Non-coding RNAs (LncRNAs) play important parts [6,7]. To veri ed genetic decisive factors of prostate cancer risk, more than 36 genome-wide association studies (GWAS) on PCa have reported proximately 582 exclusive prostate cancer risk SNPs, which was based on the NHGRI-EBI catalogue of published GWASs. First and foremost, many of these reported SNPs have subsequently been certi ed in forcefulness case-control studies [8,9].
Located at 16p13 encodes a subunit of RNA polymerase II, the POLR2E LncRNA gene was in charge of producing messenger RNA (mRNA) [10,11]. Rare studies have identi ed the impact of rs3787016 polymorphism on the risk of PCa [12][13][14][15], which was the same as esophageal squamous cell carcinoma (ESCC) [16], while the ndings were inconsistent and con icting. Notwithstanding, hardly any studies evaluated the association between these SNPs with the risk of PCa as well as quite a few investigations have been reported the functional signi cant effects of nucleotide changes in genes encoding of LncRNA POLR2E and MiRNA-499 [6,7]. While in the present investigation, we aimed to estimate the association between LncRNA POLR2E and MiRNA-499 gene polymorphisms with the risk of PCa patients [12][13][14][15][18][19][20][21], owing to the inconsistencies regarding the impact of LncRNA or MiRNA gene polymorphisms on prostate cancer affectability.

Study Characteristics
Association between rs3746444 and risk of PCa  (Fig. 4).

Meta-regression
For PCa risk, we detected a substantial heterogeneity in heterozygote model of rs3746444, recessive model of rs3746444 and allele model of rs3787016. The meta-regression analysis we conducted was designed to verify the heterogeneity of sources and the results were shown in the supplemental material ( Figure S1, Figure S2 and Figure S3). Nevertheless, the results cannot detect the source of the heterogeneity in heterozygote model of rs3746444,recessive model of rs3746444. But we found that ethnicity was a major source of heterogeneity in allele model of rs3787016 (p = 0.001).

Publication bias
We used Begg's test and Egger's test to assess publication bias. No evidence of publication bias was shown in all genotype models of rs3787016 and rs3746444 (P-value not is shown) (Fig. 5).

Discussion
Involved in process of carcinogenesis, the studies of polymorphism genes have attracted more and more attention, making genetic susceptibility to malignant tumors a hot subject in the international academy.
Through imperfect pairing with target mRNAs of protein-coding genes, evidence that MiRNAs played an important role in various crucial biological processes has been showed by earlier researches [22,23]. Noncoding RNA includes highly plentiful and functionally important RNAs, such as MicroRNAs and long noncoding RNAs. Changes of MicroRNAs or non-coding RNAs had a profound impact on the tumorigenesis of prostate cancer [24,25]. SNPs in MicroRNAs and non-coding RNAs regions may involve in these processes by way of changing their maturation and expression as oncogenes. Acting as the precursors of small non-coding RNAs, LncRNAs can produce MicroRNAs as well as endogenous small interfering RNA or served as a microRNA sponge to inhibit microRNA activity [26,27]. Meanwhile, several LncRNA also play vital roles, activating or suppressing oncogenes [28,29].
Rs3787016, localizes to the fourth intron of RNA polymerase II polypeptide E (POLR2E) 16 gene, has been investigated by several researchers on its association with cancer risk [16,[30][31][32][33][34][35][36][37].The most important function of LncRNAs is involvement in the transcriptional or posttranscriptional regulation of gene expression [38]. Abnormal expressions of these RNAs facilitate tumor-cell proliferation, invasion and survival in cancer development and progression [39][40][41]. There is a tremendous amount of developmental biological mechanisms of LncRNAs function in cancer that is unclaimed by most of existing theories. To sum up, there was an increased risk in genotypes carrying the T allele. But we found an obvious heterogeneity (I 2 = 98.5%, P H =0.000) in terms of allele model of rs3787016. In order to ascertain the heterogeneity of sources, we conducted subgroup analyses and the results detect that including ethnicity (p G = 0.000), genotype methods (p G = 0.000) and source of controls (p G = 0.000) both were the sources of the heterogeneity. Meanwhile, we also found that the study of Cao DL et al and the study of JHH et al were major sources of heterogeneity in allele model of rs3787016 (from 36.8-98.5%) by performing a sensitivity analysis, we conducted a meta-regression analysis and the results were shown that ethnicity was a major source of heterogeneity in allele model of rs3787016. Lacking in amplitude or quantity, as regards sources of controls, more studies in different populations should be involved in identifying convincible effects. Particularly we nd that there is still a lack of research on the east asian race. Before these reported ndings will contribute to clinical decision-making, additional studies with a larger sample size and in different ethnic populations are needed to con rm or further reinforce our present ndings.
MicroRNA-499 rs3746444 has been proved to be signi cantly associated with increased risk of cancer of the respiratory system, digestive system,urinary system and gynecological system [42][43][44][45][46][47][48][49][50][51][52][53].Existing studies have been veri ed that the MiRNA-499(rs3746444 T > C) gene polymorphism can control the expression level of SOX gene. Due to the up-regulation of SOX6 gene, the anti-apoptosis function of MiRNA-499(rs3746444 T > C) can be reversed [54]. Deregulation of SOX gene leads to activation Wnt/βcatenin signaling pathway, and the pathway correlated with cancerogenesis. With the changes of SOX gene expression, the rs3746444 variants might play a critical role in the tumor formation. There is an increased risk was identi ed in genotypes most carrying the C allele. We also found obvious heterogeneity in terms of heterozygote model of rs3746444 (I 2 = 54.4%, P H =0.087) and recessive model of rs3746444 (I 2 = 67.9%, P H =0.025).To ascertain the heterogeneity of sources, we conducted subgroup analyses and the results were shown that ethnicity, country and genotype methods were major sources of heterogeneity.
When we conducted the sensitivity analysis of the rs3787016, we found that signi cant decreases of pooled heterogeneity after excluding the study of Cao  was not considered so that our unadjusted estimates should be con rmed by further studies. Still, our results were supported by existing literature, while further well-designed large studies are warranted to clarify the possible role of these polymorphisms in more details.
Nevertheless, supported by the existing literature, our investigation is necessary to take in more welldesigned large-scale studies to clarify the possible role of these polymorphisms at full length.

Conclusion
In conclusion, our ndings proposed that rs3746444 in MiRNA-499 and rs3787016 in Long Non-coding RNAs POLR2E polymorphisms increased the risk of developing PCa. Further functional studies are warranted to reveal the role of the polymorphism in carcinogenesis.

Search Strategy
Aiming to systematically retrieve studies describing the relationship between rs3787016, rs3746444 and prostate cancer risk, we searched PubMed and EMBASE on March 1, 2020, and conducted the literature search as well as references list of included studies and reviews once more on Jun 4, 2020, which guaranteed the data of our research was current and available. The search attached importance to following subject terms and keywords: rs3787016, rs3746444, prostate cancer, of which detailed strategies are provided in the supplemental material.

Study Selection
As for literature eligibility, two investigators accomplished the assessments separately, while contradictory conclusions were resolved in terms of debate and unanimity. Inclusion criteria were as follows: (a) studies that was digging at the association of LncRNA POLR2E/miR-499 and PCa in full-text; (b) utilizing an irrelevant case-control design; (c) study published in English; (d) data provided make conducting an OR with 95% Cl possible; (e) su cient counting of genotypes or allele frequencies.
Exclusion criteria: (1) comment, review, meta-analysis, and editorial; (2) Only the largest sample was included when it turned to multiple publications from the same population/area ( Figure 6).

Data Extraction
The following information was draw out and inducted to specially designed forms from the included studies by two investigators separately : rst author's last name, year of publications, country of origin, ethnicity, study-design (sources of controls), sample size in cases and controls, the number of cases and controls with variant allele and wild type, genotyping methods and hardy weinberg equilibrium (HWE) of the studies respectively. For each publication, all data complying with the selection criteria were reviewed and extracted independently by two of the investigators. If dissent existed, the authors rechecked the original data and held a discussion to reach a consensus. When it still existed, discrepancies would be adjudicated by a third reviewer until the consensus was achieved on every item ( would be used if P H was no more than 0.05 (P H <0.05), and when P H was > 0.05, xed effects model was used 21 . Sensitivity analysis was also tested by removing one study at a time, to evaluate the effect of removal and effect of size of each study on the homogeneity of the whole. Furthermore, subgroup analyses were strati ed by ethnicity, source of controls, country, genotype methods, HWE (only in rs3746444). Publication bias was conducted by Begg's funnel plots and further assessed by the Egger's regression test [57,58]. When an asymmetric plot was shown or Begg 's test was P<0.05, we considered it as a signi cant publication bias. Besides, Hardy-Weinberg equilibrium (HWE) was implemented to identify the effective records in our study [59]. All the statistical analyses and graphics above were available to run under Stata 12.0 software (Stata Corp LP 4905, Lakeway Drive College Station, TX 77845 USA). A two tailed P<0.05 was regarded as signi cant, except for speci ed conditions, which a certain Pvalue was declared, The Newcastle-Ottawa Quality Assessment Scale (NOS) [60] was used to evaluate the quality of include studies (Figure 7). Abbreviations CI, con dence interval; HB, hospital-based; HWE, Hardy-Weinberg equilibrium; LncRNA, Long Non-coding RNA; MiRNA, Micro-RNA; OR, odds ratio; PB, population-based; PCa, Prostate Cancer; P H , P value of heterogeneity; P G , heterogeneity between groups; Unknown, not provided in the original reference.

Declarations Acknowledgments
We acknowledge that Third A liated Hospital of Guangzhou Medical University for supporting the work of our study.

Authors' Contributions
Xuebin Liu and Ping Liu designed the study. Yan Liu, Zhaozhang Wang, and Gaorui Zhou completed the literature eligibility assessment, extraction, and analysis of data. Zihao Zou, Jianfeng Zhong and Nan Deng reviewed the results. Xuebin Liu and Yuwu Li wrote the report. All authors participated in the discussion and modi cation of the text. All authors approved the nal version of the paper.

Funding
Not applicable.

Availability of data and materials
The datasets used and analysed during the current study are available from the corresponding author on reasonable request.
Ethics approval and consent to participate Not applicable.    Begg's funnel plot of publication bias test for the allelic model of rs3746444 (a) and rs3787016 (b) Figure 6 Search strategies in PUBMED and EMBASE Figure 7 Quality assessment of the included studies according to NOS