Introduction

Colorectal cancer (CRC) is the second most common cancer in females and the third in males with 1.2 million new annual cases and 608,700 deaths in 2008 worldwide1. The vast majority of cases (90%) occur in people over 502. Apart from hereditary CRC syndromes, the development of this cancer type is still poorly understood3,4. Both germline and somatic genetic variations have been proposed as contributing factors in CRC development5,6,7. Recent large scale studies reported significant risk of different germline variations for CRC development6,8. Novel reports suggest a potential influence of single nucleotide polymorphisms (SNPs) of microRNA-related genes (miRNAs) for the risk of cancer development9.

The first description of miRNA appeared in 199310,11. The discovery of these small non-coding RNAs has opened a new investigation platform in molecular biology12. MiRNA are ~22 nucleotide sequences of RNA found in both prokaryotes and eukaryotes that are intimately involved in cell differentiation, cell cycle progression and apoptosis10,11. Recent studies have shown aberrant miRNA expression patterns in a range of human diseases including many cancers13,14. Deregulation of miRNAs can influence carcinogenesis through mRNA targets encoding tumor suppressor genes or oncogenes5. Either phenomena – over expression and silencing or switching off of specific miRNAs, have been described in the carcinogenesis of CRC15. The discovery of miRNAs has opened new opportunities for SNPs in cancer research16,17. MiRNAs have potential to regulate multiple genes; therefore, variations in genes encoding or related to miRNAs may produce pronounced regulatory effects that may modify the risk of different human diseases including cancer18. SNPs of the genes encoding miRNAs can alter miRNA expression and may influence cancer risks19,20. Growing number of case-control studies have shown associations between the polymorphisms of the genes encoding miRNAs and the risk of different malignancies. In this study, we selected five different SNPs miR-27a T>C (rs895819), miR-146a G>C (rs2910164), miR-196a-2 C>T (rs11614913), miR-492 G>C (rs2289030) and miR-608 C>G (rs4919510), which have been reported to influence cancer risks9,21,22,23 and performed a case-control study among CRC patients. Above mentioned SNPs are located in mature or pre-mature regions of miRNAs with potential implications in gene expression regulation9,16. It is also worth pointing out, that all of these miRNAs have been shown to be deregulated in colorectal cancer or target carcinogenesis related genes13,15.

Selected miRNAs and their polymorphisms have been studied in CRC as well as other cancer types9,24,25. The list of previously published case-control studies on rs895819, rs2910164, rs11614913, rs2289030 and rs4919510 SNPs in CRC is presented in Table 1. Several studies have linked miR-196a-2 gene polymorphism (rs11614913) with increased risk of CRC26,27,28, meanwhile other studies could not confirm this association29,30,31,32. A meta-analysis by Srivastava et al. found that miR-196a-2 gene polymorphism was associated with colorectal cancers9. MiR-146a has been shown to play an important role in tumor genesis by promoting cell proliferation and colony formation in NIH 3T3 cells33,34. SNP of miR-146a (rs2910164) was shown to be associated with decreased risk of CRC in Asian populations26,35,36, while no significant role of this polymorphism was observed in European population31,32 as well as in one Japanese study37. Pathi et al. showed that novel anti-inflammatory drug ethyl 2-((2,3-bis(nitrooxy)propyl)disulfanyl)benzoate (GT-094) decreases miR-27a expression in colon cancers cells38. Hezova et al. study did not show significant associations between miR-27a rs895819 polymorphism and CRC susceptibility in European population31, but studies on this SNP in other population are missing. MiR-492 is deregulated in colorectal cancer tissues when compared to normal colon mucosa39, but rs2289030 of this miRNA has not been previously studied with respect to CRC risk. Several reports showed that SNP of miR-608 (rs4919510) was linked with prognosis and survival in patients with CRC40,41,42, but not overall cancer risks42.

Table 1 Previous case-control studies on miR-27a (rs895819), miR-146a (rs2910164), miR-196a-2 (rs11614913), miR-492 (rs2289030) and miR-608 (rs4919510) SNPs and colorectal cancer risk
Full size table

Most of the currently available genotyping studies related to miR-27a T>C (rs895819), miR-146a G>C (rs2910164), miR-196a-2 C>T (rs11614913), miR-492 G>C (rs2289030) and miR-608 C>G (rs4919510) in CRC patients come from Asian populations and report partially conflicting results (Table 1). Furthermore, to our best knowledge there are no previous reports on miR-492 rs2289030 and CRC risk. Here in this study we evaluated the role of five above mentioned miRNA SNPs in a case-control study including 428 controls and 193 CRC cases of European descent.

Results

Characteristics of the subjects

Characteristics of the subject groups are presented in Table 2. Control group consisted of 428 individuals: 112 (26.2%) males and 316 (73.8%) females; mean age was 63.2 ± 10.6 years. There were 193 subjects within CRC group: 109 (56.5%) males and 84 (43.5%) females; mean age was 67.2 ± 10.3 years. In accordance with real-life age, subjects differed significantly according to age and gender distribution between the groups (P < 0.001). There were more males in CRC group and this groups was older than the controls; however, then performing logistic regression analysis odds ratios (OR) were adjusted for these covariates as explained in the methods section below. Individuals in control and CRC groups were recruited in Lithuania and Latvia (Table 2).

Table 2 Characteristics of subjects within control and colorectal cancer groups
Full size table

Associations of miRNA SNPs and risk of CRC

Genotype distributions for all five polymorphisms in the study control group were similar to those expected for Hardy-Weinberg equilibrium: rs895819 (P = 0.780), rs2910164 (P = 0.583), rs11614913 (P = 0.128), rs2289030 (P = 0.546), rs4919510 (P = 0.075). Genotype and allele distributions for miR-27a C>G (rs895819), miR-146a C>G (rs2910164), miR-196a-2 C>T (rs11614913), miR-492 C>G (rs2289030) and miR-608 C>G (rs4919510) gene polymorphisms in control and CRC patient groups are presented in Table 3. No significant associations were observed for diseases under study following correction for multiple testing. We observed a tendency for miR-146a C allele to be associated with lower risk of CRC when compared to G allele (OR = 0.68, 95% CI 0.49–0.95, P = 0.025), but difference did not reach the adjusted significance threshold. MiR-492 C allele was marginally associated with higher risk of CRC when compared to G allele (OR = 1.56, 95% CI 1.00–2.42, P = 0.047). All the other comparisons between control and CRC groups did not reveal significant associations or trends for five SNPs of miRNAs under different genetic models.

Table 3 Genotype frequencies of miR-27a, miR-146a, miR-196a-2, miR-492, miR-608 SNPs in controls and colorectal cancer patients
Full size table

Discussion

In this paper we present a case-control study of five gene polymorphisms – miR-27a (rs895819), miR-146a (rs2910164), miR-196a-2 (rs11614913), miR-492 (rs2289030) and miR-608 (rs4919510) including 428 controls and 193 CRC patients of European descent. MiRNA deregulation is a well-established event in colorectal carcinogenesis5,15; therefore, we expected that SNPs related to miRNAs could be associated with CRC. The polymorphisms mentioned above have been linked with the risk of different cancers; however, the data of these polymorphisms in CRC patients are scarce and reports show contradictory results. Our data suggest that miR-27a, miR-146a, miR-196a-2, miR-492 and miR-608 gene polymorphisms are not associated with the development of CRC in European population. To our best knowledge, there are only two reports on association between rs2910164, rs11614913 and CRC in Caucasian population, while rs2289030 has not been examined in any of the previous studies.

MiR-196a-2 SNP (rs11614913) was investigated in several case-control studies with respect to CRC risk. A report by Lv et al. showed that CT, TT genotypes and T allele were associated with an increased risk of CRC compared with the CC genotype and C allele (CT vs. CC: OR = 7.34, 95% CI 3.76–14.34, P < 0.001; TT vs. CC: OR = 1.99, 95% CI 1.63–2.42, P < 0.001, respectively)26. Opposite results were shown by Zhan et al. who found that CC genotype and C allele were associated with a significantly increased risk of CRC compared with the TT genotype and T allele (CC vs. TT: OR = 1.74, 95% CI 1.11–2.73, P = 0.015, C vs. T: OR = 1.32, 95% CI 1.06–1.65, P = 0.014), however, no significant association between this polymorphism and CRC progression was observed29. CC genotyped was also named as a risk genotype for CRC by Min et al.27 and Zhu L et al.28 studies. The other studies by Chen et al. and Hezova et al. did not observe the link between miR-196a-2 SNP and CRC risks30,31. Although a recent meta-analysis of seven studies suggested that rs11614913 might contribute to the reduced risk of CRC43, our results do not support these findings.

MiR-146a SNP (rs2910164) has been extensively studied in different cancers, but reports on CRC patients, especially in Caucasian populations, still remain scarce. Lv et al. found that carriers of rs2910164 CC genotype had a significantly decreased risk of CRC (OR = 0.58, 95% CI 0.37–0.93, P = 0.02) and C allele was associated with reduced risk of CRC (OR = 0.80, 95% CI 0.66–0.97, P = 0.02)26. Similar findings for miR-146a SNP have been detected in a study by Ma et al. including 1203 controls and 1147 CRC patients, which showed that carriers of GC/CC genotypes had a reduced the risk of colon cancer35. Opposite results were revealed in European studies by Hezova et al.31 and Vinci et al.32 where miR-146a SNP was not linked with CRC; furthermore, negative results have also been previously reported in Asian populations27,37. The results of our study support the negative findings of the latter studies as no significant association between miR-146a polymorphism and the risk of CRC was determined. A similar conclusion on rs2910164 and CRC risks is drawn by a meta-analysis by Wan et al.43.

Gene polymorphism of miR-27a has been poorly investigated in CRC patients in comparison to other types of cancer. Meta-analysis by Hu et al. concluded that miR-27a rs895819 polymorphism may contribute to the susceptibility of overall cancer risk24. To date, only one study carried out by Hezova et al. examined the role of this SNP in CRC development in Caucasians31. The results of the latter study did not show any associations between miR-27a rs895819 polymorphism and CRC susceptibility in 212 controls and 197 CRC patients. These results are in concordance with our study, but more extensive comparison of present results due to the lack of studies cannot be performed.

Polymorphism of miR-492 (rs2289030) has been yet little explored in cancer related case-control studies. In a study with non-small cell lung cancer no risk was determined for carriers of different rs2289030 genotypes44. The results our own group showed no role of this SNP in the development of gastric cancer45. A study on colorectal cancer by Lee et al. has demonstrated that progression-free survival of the patients with the combined miR-492 CG and GG genotype was significantly worse than that of the patients with the miR-492 CC genotype22, but the risk of CRC with respect to rs2289030 genotypes or alleles has not been analyzed in this study. Our present data do not show a significant role of miR-492 SNP for CRC development, but further studies including other populations from different ethnical backgrounds are warranted.

A recent meta-analysis did not detect significant associations for miR-608 rs4919510 in terms of overall cancer risks or specific types of cancer24. Previous studies that have analyzed the role of miR-608 (rs4919510) SNP in relation to CRC risk are scarce. A case-control study including 245 colorectal cancer patients and 446 controls did not find an association with CRC risk, but their results showed that GG genotype was associated with an increased risk of death in white population and reduced risk of death in African Americans42. Our results are in line with this study as rs4919510 genotypes or alleles of miR-608 were not linked with the presence of CRC. Two other studies on rs4919510 were not aimed to determine overall risk of CRC development, but rather looked at looked prognostic role of this SNP. Their results showed that miR-608 SNP was significantly linked with recurrence and survival in CRC patients40,41.

Overall, our study did not show significant association between miR-27a (rs895819), miR-146a (rs2910164), miR-196a-2 (rs11614913), miR-492 (rs2289030) and miR-608 (rs4919510) and the risk of CRC. These findings are in line with our previous research which did not reveal association between these SNPs and gastric cancer or premalignant gastric conditions45. Although some of the studies have observed significant differences for various cancer-types, the ultimate role of microRNA related SNPs in cancer development in still not clear. The differences among the conclusions drawn by separate studies may result from study design, subtypes of CRC or different ethical backgrounds9,24. The current meta-analysis suggest potential role of these SNPs in CRC development, but further research in this field is mandated9,24,25,43,46. To date there are 19 genome wide association studies (GWAS) in CRC patients according to National Human Genome Research Institute; but none of them have linked rs895819, rs2910164, rs11614913, rs2289030 and rs4919510 with the development of cancer47. Nevertheless, due to different limitations of GWAS studies, smaller association studies on potential candidate SNPs remain relevant.

Our study has certain limitations that have to be acknowledged. The number of individuals within CRC and control groups is not large for conclusive genotyping studies; however, we believe that our data will be valuable for future meta-analysis on miRNA related SNPs. Due to a small number of individuals we were not able to perform appropriate haplotype analysis in order to evaluate potential effects of combined genotypes of these miRNA SNPs. Unfortunately, we do not have a validation group; therefore, tendencies that were observed for miR-146a and miR-492 SNPs should be explored in further larger scale studies. Our control group might be biased by the fact that control subjects were hospital based – outpatient dyspeptic patients without previous cancer history. Survival data as well as smoking and alcohol consumption data were available only for a small proportion of subjects; therefore, stratification analyses were not performed. Due to a small number of individuals we were not able to perform genotyping sub-analysis for CRC regarding disease stage, tumor location or molecular subtypes, which may have different genetic susceptibility.

Methods

Ethics statement

The study was approved by the Ethics Committees of the Lithuanian University of Health Sciences and Central Medical Ethics Committee of Latvia. All patients have signed an informed consent form to participate in the study. All the methods applied in the study were carried out in accordance with the approved guidelines.

Study population

Patients and controls were recruited during the years 2007–2013 at two gastroenterology centers in Lithuania (Department of Gastroenterology, Lithuanian University of Health Sciences, Kaunas) and Latvia (Riga East University Hospital and Digestive Diseases Centre GASTRO, Riga). Control group consisted of patients with dyspeptic symptoms, who had no alarm symptoms and no history of previous malignancy. CRC patients had histological verification of colorectal adenocarcinoma and were recruited from out-patient and stationary departments. In total 621 (428 controls and 193 CRC patients) individuals were included in the study. There were 265 subjects from Latvian group (201 controls and 64 CRC) and 356 subjects from Lithuanian group (227 controls and 129 CRC). All patients were of European ethnicity.

DNA extraction and genotyping

Genomic DNA from samples was extracted using salting out method from peripheral blood leukocytes. DNA samples were stored at −20°C until analysis. SNPs of miR-27a C>G (rs895819), miR-146a C>G (rs2910164), miR-196a-2 C>T (rs11614913), miR-492 C>G (rs2289030) and miR-608 C>G (rs4919510) were genotyped by using predesigned TaqMan® assays with a 7500™ real-time cycler, in accordance with the manufacturer's instructions (Life Technologies, CA, USA). Genotype assignments were manually confirmed by visual inspection with the SDS 2.0.5 software compatible with the TaqMan® system. After genotyping 5% of samples in each genotype group were selected for repetitive analysis with 100% concordance rate. Samples that failed to genotype were recorded as undetermined.

Statistical analysis

Age is shown as means and standard deviations and was compared using unpaired Student's t-test. Statistical analysis of genotyping data was performed using PLINK software version 1.0748. Hardy-Weinberg equilibrium was assessed for each of SNPs. Association between control and CRC with gene polymorphisms was calculated using logistic regression analysis with adjustment for age, gender and country of birth with 95% confidence intervals (CI). The analysis was carried out using homozygous, heterozygous, allelic, recessive and dominant models. An adjusted significance threshold α = 0.01 (0.05/5) was used to determine significant differences.