Introduction

Telomeres are structures at the ends of eukaryotic chromosomes that protect chromosome termini from fusion and prevent the loss of essential genes due to the incomplete replication of the lagging DNA strand, solving the “end replication problem” [1, 2]. Mean telomere length in humans shows considerable interindividual variation and strong genetic determination [3]. Telomeres are extended by telomerase, which consists of a catalytic telomerase reverse transcriptase (TERT) protein, the telomerase RNA component (TERC), and the associated proteins. Telomerase is naturally expressed only in embryonic cells, germline cells, basal/stem cells of highly regenerative tissues, and activated lymphocytes. In other cells, telomerase activity is repressed during embryonic differentiation [4], resulting in continual telomere shortening within each mitotic cycle until telomeres reach a short critical length that induces the activation of cellular checkpoints and leads to a cell cycle arrest referred to as replicative senescence. Cells lacking critical checkpoint responses continue to divide and eventually enter a second growth arrest state termed “cellular crisis,” when severe telomere dysfunction and associated genomic instability cause apoptotic cell death [1].

Telomerase activity and telomere length have been hypothesized to be linked to the process of carcinogenesis, though the connection seems to be complicated. On the one hand, telomere erosion provides a tumor-suppressive function establishing a limit on the proliferative capacity of a cell and preventing the accumulation of somatic mutations. The inability to repress telomerase in somatic tissues can extend proliferative lifespan and promote carcinogenesis [5]. On the other hand, this mechanism to restrict cell growth can lead to the accumulation of senescent cells in elderly persons or in individuals with constitutively short telomeres. Senescent cells no longer divide but may remain metabolically active and secret metalloproteases, inflammatory cytokines, and growth factors and therefore create a tumorigenic microenvironment that promotes further malignant transformation of preneoplastic cells [2, 5]. Beyond that, the absence of a functional DNA-damage response may convert senescence and crisis induced by telomere dysfunction from a suppressor to a promoter of tumorigenesis. In this case, the inheritance of short telomeres or the impairment of telomere length maintenance in stem cells may drive tumor progression. Cells that bypassed replicative senescence need to inactivate p53 and activate telomere maintenance mechanisms to escape from crisis [1]. This can be achieved either by telomerase reactivation, or by the upregulation of the alternative lengthening of telomeres pathway. Telomerase can be reactivated via genetic or epigenetic mechanisms. Genetic mechanisms involve somatic mutations in the TERT gene promoter, which are frequently observed in many human cancers, particularly those that arise from tissues with low rates of self-renewal [6]. Epigenetic mechanisms of telomerase upregulation involve hypermethylation of specific regions in the TERT promoter, as it was shown for pediatric brain tumors and pancreatic cancer [7, 8].

Given the role that telomerase and its regulation seem to play in carcinogenesis, it has been proposed that sequence variants in telomerase genes such as TERT gene can affect individual cancer predisposition. This hypothesis was confirmed by a large body of evidence from candidate gene studies, meta-analysis [9, 10], genome-wide association studies (GWAS) [11–17], and fine-scale mapping studies [18–20] that revealed the association of different TERT polymorphisms with various types of cancer. However, to date, no study has tested these associations in ethnical Russians. In our study, we examined the association of single nucleotide polymorphisms (SNPs) in the TERT gene with the risk of the breast and prostate cancer in Russian residents of Western Siberia. We selected three polymorphisms for the analysis—rs2736100 T > G, rs7726159 C > A, and rs2853669 T > C, reasoning from a strong evidence for their functionality provided by previous studies. Intronic SNPs rs2736100 and rs7726159 were found to be the top polymorphisms associated with telomere length by recent GWAS [12, 21, 22]. Polymorphism rs285366 is located in Ets2 binding site in the TERT promoter, which has been suggested to be a positive regulator of the TERT gene. Its variant allele was shown to reduce promoter activity and associate with shorter mean leukocyte telomere length in several studies [23–25]; therefore, it plausibly affects Ets2 binding.

Materials and methods

Patients

Women with breast cancer, men with prostate cancer, and control groups were enrolled during an epidemiological study conducted by the Altai Branch of the N. N. Blokhin Cancer Research Centre of the Russian Academy of Medical Science in 2006–2009. All cases and controls were Caucasian Russians and lived in the Altai region of Russia. All the individuals gave signed informed consent, and the study was approved by the local ethics committee.

The group of women with breast cancer consisted of 660 patients with a histologically confirmed diagnosis of breast cancer (mean age 56.3 ± 7.8 years, range 46–83 years). All women had sporadic breast cancer. The control group included 523 women without oncological diseases (mean age 61.0 ± 15.7 years, range 19–89 years). The group of prostate cancer patients consisted of 372 men with histologically verified prostate cancer (mean age 69.2 ± 8.5 years, range 37–104 years). The most patients had sporadic form except six patients (1.6 %) with familial history of prostate cancer. The control group included 363 men with no personal or family history of cancer (mean age 61.3 ± 16.6 years, range 31–91 years).

Genomic DNA was isolated from leukocytes in venous blood by proteinase K digestion, followed by phenol/chloroform extraction, and ethanol precipitation. DNA samples were stored at −20 °C in a freezer compartment.

Genotyping

Genotyping was carried out by real-time PCR allelic discrimination with TaqMan probes. PCR was performed in 20-μL reaction volumes containing 20–100 ng of genomic DNA, 65 mM Tris–HCl (рН 8.9), 24 mM ammonium sulfate, 3.5 mM MgCl2, 0.05 % Tween 20, 0.2 mM dNTP, 0.3 mM of each primer, 0.1 mM of each probe (Supplemental Table 1), and 1.0 U of Taq polymerase. PCR thermal cycling conditions were as follows: denaturation for 3 min at 96 °С followed by 48 cycles of 8 s at 96 °С and 40 s at 60 °С. Amplification procedure was conducted using CFX96 Thermal Cycler (Bio-Rad, USA).

Statistical analysis

To evaluate the effects of the polymorphisms on cancer susceptibility, odds ratio (OR) and 95 % CI were calculated by logistic regression analysis adopting co-dominant, additive, dominant, and recessive models of inheritance. All data were adjusted for age. To choose the inheritance model that best fits the data, Akaike’s information criterion was used. The expected frequency of genotypes in the control group was tested for Hardy-Weinberg equilibrium using exact test. Differences were considered statistically significant at P < 0.05. Statistical analyses were performed using the GenABEL statistical package for the R language (version 3.0.2, http://www.r-project.org; glm function). Haplotype frequencies and the corresponding OR and CI 95 % values were calculated using the haplo.stats statistical package for the R language (version 3.0.2; haplo.score and haplo.glm functions). Linkage disequilibrium was analyzed based on D’ and r 2 values calculated using the CubeX program (http://genes.org.uk/software/cubex/). To estimate the statistical power of study, genetic power calculator (http://pngu.mgh.harvard.edu/~purcell/gpc/cc2.html) was used.

Results

The genotypes of polymorphisms rs2853669, rs2736100, and rs7726159 in the TERT gene were determined in the groups of women with breast cancer, men with prostate cancer, and corresponding control groups. The distribution of genotypes in the control groups was in accordance with Hardy-Weinberg equilibrium for all studied polymorphisms (Table 1). Variant allele C of the polymorphism rs2853669 was associated with the risk of prostate cancer assuming сo-dominant (TC vs. TT OR = 1.65, P = 0.002), additive (OR = 1.42, P = 0.005), and dominant (OR = 1.64, P = 0.001) models of inheritance. The association remained significant after the implementation of Bonferroni correction for multiple testing. According to Akaike’s information criterion, dominant model was preferable (Table 2). Variant allele A of the polymorphism rs7726159 was inversely associated with prostate cancer risk (сo-dominant model: AA vs. CC OR = 0.42, P = 0.002; additive model: OR = 0.69, P = 0.002; dominant model: OR = 0.67, P = 0.01; recessive model: OR = 0.48, P = 0.005; additive model the most suitable, Table 2), and the revealed association remained significant after applying Bonferroni correction. Alleles of polymorphism rs2736100 showed no association with prostate cancer risk, and no association was observed between all studied polymorphic loci and the risk of breast cancer.

Table 1 Allele and genotype frequencies in the case and the control groups
Full size table
Table 2 Association of polymorphisms in TERT gene with the risk of breast and prostate cancer adopting different models of inheritance
Full size table

We also determined the haplotype frequencies and studied the association of TERT haplotypes with the risk of breast and prostate cancer (Table 3). The rs2853669 T - rs2736100 G - rs7726159 A haplotype, containing both low-risk alleles, showed the association with reduced risk of prostate cancer (OR = 0.58, P = 0.001). The remaining haplotypes were not associated with the risk of breast cancer and prostate cancer. We evaluated the linkage disequilibrium of studied SNPs (Fig. 1). rs2736100 and rs7726159 were tightly linked with D′ = 0.97, r 2 = 0.53; rs2853669 and rs2736100 were linked with D′ = 0.73, r 2 = 0.21; and rs2853669 and rs7726159 were not in linkage disequilibrium (D′ = 0.19, r 2 = 0.03).

Table 3 Association of TERT haplotypes with the risk of breast and prostate cancer
Full size table
Fig. 1
figure 1

The linkage disequilibrium between studied SNPs in ethnical Russians

Full size image

Discussion

In this case–control study, we studied the association of three functional single nucleotide polymorphisms (SNPs) in the TERT gene with the risk of breast and prostate cancer in ethnical Russians. For all tested SNPs, genotype distribution in the control groups was in accordance with Hardy-Weinberg equilibrium (Table 1). The observed frequencies of variant alleles were similar or close to those reported in different populations of European ancestry [12, 15, 16, 19, 26–29].

We revealed the associations of alleles rs2853669 T and rs7726159 A, as well as a haplotype containing both alleles, with the decreased risk of prostate cancer (Tables 2 and 3). Our findings confirm the result of recent GWAS [12], that reported the association of allele rs7726159 A with a decrease in prostate cancer risk, and are consistent with the fine-mapping study [18], that demonstrated the association between allele rs2853669 C and increased prostate cancer risk. Alleles rs2853669 T and rs7726159 A have previously been shown to be associated with longer telomeres [12, 21, 22, 24, 29, 30] or high telomerase and TERT promoter activity [23–25]. Along with these studies, our study provides evidence that shorter telomeres could be a risk factor for prostate cancer. To date, the relationship between telomere length and cancer still remains elusive [31, 32], and molecular mechanisms have been proposed for both direct and inverse correlation. It seems probable that this relationship is complex and cancer-specific. With regard to prostate cancer, two studies have been published examining this relationship. The first study revealed no correlation [33], while the second study suggested the association between shorter telomere length and increased risk of prostate cancer when measured in pre-diagnostic samples [34]. This relationship should be studied more extensively in larger prospective studies.

We observed no association between the tested SNPs and the risk of breast cancer. Our result is consistent with the results of previous studies regarding the polymorphism rs2736100 [19, 20, 26], but is in contradictory with the findings reported by some other studies regarding rs7726159 and rs2853669 polymorphisms. Allele rs7726159 A slightly increased the risk of breast cancer in the fine-mapping study [19] and GWAS [12], and allele rs2853669 C was inversely associated with this oncological disease in the abovementioned fine-mapping study [19], as well as in the meta-analysis of case–control studies [10]. Notably, the associations of these two SNP with breast cancer and prostate cancer are oppositely directed, providing evidence that these types of cancer differ in their etiology, and telomerase activity and telomere length could have an opposite effect on the risk of their development. Our inability to confirm the associations of rs7726159 and rs2853669 with breast cancer risk could be explained by the action of potential ethnicity-specific factors. But it could also be the result of a low statistical power of our study. In previous studies, OR values for rs7726159 A and rs2853669 C were correspondingly 1.05 and 0.94 per allele. In order to have sufficient statistical power to detect such weak associations, a study sample of more than 10,000 subjects is needed, while our study had an 80 % statistical power to reveal associations with OR value about 1.3 or greater (0.8 or less).

Besides limited sample size, some other limitations of our study should be acknowledged. First of all, we analyzed only three polymorphisms in the TERT gene, which are most likely to influence TERT function according to published studies. However, there are other SNPs located in the TERT gene, which were suggested to be functional and which are not tightly linked with SNPs tested in our study, e.g., rs2735940 [24, 30] and rs2736098 [17]. It would be interesting to analyze them in future studies. Secondly, we did not measure gene-environment interactions. Despite these limitations, our study is the first study investigating the association of TERT polymorphisms with prostate and breast cancer in Russian population, and we had enough statistical power to reveal the influence of SNPs in this gene on prostate cancer risk.

Conclusion

We revealed an inverse associations of alleles rs2853669 T and rs7726159 A of the TERT gene with the risk of prostate cancer in ethnical Russians. Along with previous studies, our results provide evidence that TERT polymorphisms are involved in the development of this malignancy. We did not reveal any association with the risk of breast cancer, probably due to the limited sample size. Thus, larger studies are warranted to further investigate this association in the Russian population.