Investigation of DNA repair-related SNPs underlying susceptibility to papillary thyroid carcinoma reveals MGMT as a novel candidate gene in Belarusian children exposed to radiation

Thyroid cancer (TC), with an age-standardised incident rate of 4.0 per 100,000 men and women in developed countries (http://​ci5.​iarc.​fr) is the most common type of endocrine malignancy and it is now the fifth most common cancer diagnosed in women [1]. Papillary thyroid carcinoma (PTC) along with follicular thyroid carcinoma (FTC) and Hürthle cell carcinoma (a subtype of FTC) is termed differentiated thyroid carcinoma (DTC). All together these TC types account for more than 90% of all TC and PTC is the most common subtype, representing approximately 80% of cases. During the last three decades incidence rates of TC, and in particular of PTC, have increased in nearly all countries [2]. This trend is partly attributable in high-resource countries to increased surveillance of the thyroid gland and improved detection methods [3], but it is also possible that changes in exposure to environmental factors and ionizing radiation, a well-established risk factor for this cancer especially when it occurs during childhood or as a young adult, could influence PTC incidence [4‐6]. Other risk factors for PTC include iodine deficiency and excess [7], previous history of benign thyroid disease, such as nodules and autoimmune thyroid disease, as well as a family history of DTC. Remarkably, DTC has a strong familial component and first-degree relatives of DTC patients have up to eight times higher risk of developing DTC than the general population, indicating that genetic factors play an important role in DTC risk [8, 9]. The observed familial risk could be partly explained by high-penetrance mutations in yet unidentified genes or by the additive effect of numerous low-penetrance variants [10, 11]. This latter hypothesis would explain the paucity of families with more than two affected members. Recent case-control studies, in particular Genome-Wide Association Study (GWAS) have highlighted a number of low-penetrance alleles contributing to sporadic DTC risk. The association of DTC with the 9q22 (rs965513) locus close to the thyroid specific factor FOXE1 has been detected in all association studies conducted so far in different populations [12]. Weaker associations have been reported, among others, with rs944289 at 14q13.3 (near NKX2–1), rs966423 at 2q35 (in DIRC3), rs334725 at 1p31.3 (in NFIA) and rs2439302 at 8p12 (in NRG1) [12‐14]. Since the associations described so far only explain about 10% of the familial risk of DTC [15], additional studies on specific populations are of particular relevance, especially studies of high-risk populations having been exposed to known environmental risk factors. Notably a sharp increase in incidence of thyroid malignancies, virtually all PTC, has been observed in children and adolescents that were exposed to radioactive fallout from the Chernobyl nuclear power plant accident in April 1986. Variation in clinical course, ranging from highly aggressive tumours developing after the shorter latency to more indolent carcinomas with longer latent period has been reported in this population [16], which suggests that some predisposing genetic factors could influence an individual’s sensitivity to radiation and therefore modify the risk of developing TC in exposed population [17, 18].

DNA repair is an important defence mechanism against DNA damage caused by normal metabolic activities and environmental factors [19]. DNA damage is recognized and processed by a series of distinct pathways called the “DNA damage response (DDR)” [20]. It includes direct repair (DR), base and nucleotide excision repair (BER and NER), mismatch repair (MMR), double strand break repair (DSBR) and interstrand cross-links repair system [21, 22]. Because ionizing radiation produces DNA lesions and DNA repair genes play a critical role in maintaining genome integrity, it has been proposed that inherited variations in such genes may reduce DNA repair capacity and influence cancer development in exposed subjects. Few case-control studies on radiation-related PTC have been conducted to date and published data suggest that alterations in expression levels and polymorphisms in some candidate DNA repair genes belonging to the different pathways are implicated in risk of developing thyroid cancer [23‐26]. In particular, others and we have reported association with the coding SNP rs1801516 (p.Asp1853Asn) in the ATM gene, a key regulator of signalling following DNA double-strand breaks (DSBs) [25‐27]. Other possibly involved DNA repair-related SNPs include rs25487 (p.Arg399Gln) in XRCC1 involved in BER [25, 27‐29] and rs13181 (p.Lys751Gln) in ERCC2 involved in NER [30]. To follow up with these findings, we hypothesized that alterations in other genes related to the different DNA repair mechanisms may be correlated with tumour susceptibility in subjects originated from the Gomel region of Belarus and having been exposed to radioactive fallout from the Chernobyl nuclear power plant accident during childhood. Here we evaluated the risk association between SNPs in DNA repair genes present on the Cancer SNP panel array (Illumina) and PTC in this very unique population. Findings of this study may increase understanding of PTC aetiology and help identify subjects who may be particularly susceptible to the carcinogenic effects of radiation on the thyroid.

Understanding the aetiology of PTC and increased susceptibility to exposure to ionising radiation is an important aim for radiation protection policy. Radiation exposure during childhood is a strong risk factor for PTC, and polymorphisms in DNA repair genes are likely to affect this risk, but few studies have been designed to determine the role of such genes as modulators of PTC risk [40]. For this reason we chose to evaluate the association between a panel of common candidate SNPs located within 43 DNA repair genes and PTC. Our results showed significant association for rs2296675 located in MGMT encoding the O⁶-methylguanine DNA methyltransferase, and suggestive association for variants in ERCC5 encoding a single-strand specific DNA endonuclease and variants in PCNA encoding the proliferating cell nuclear antigen. Hence our findings support the involvement of several mechanisms that could be mobilised by the follicular cells of the thyroid gland to repair the different types of DNA damages that could occur after exposure to radiation.

There are several limitations to this study that should be noted. First, the power to detect genes with small effect sizes may be low due to the relatively small number of subjects included in this study because of the uniqueness of the studied population. Indeed for SNPs with low MAF in controls (5%), the power of our study for evidencing an association between a candidate SNP and PTC could reach 80% only for an OR of 3.5 or higher for P < 0.05. For SNPs with MAF in controls of about 20%, our study had a power of 80% for evidencing an association if OR is about 1.90 P < 0.05.

Second, the genotyped markers in the Illumina SNP Cancer Panel array are scarce (on average 3.6 SNPs per cancer gene on the array, and 4.1 SNPs per DNA repair) and do not cover all of variations in the candidate genes. In addition, due to small sample size, we only included markers with a MAF ≥ 0.05, and common SNPs usually have small effects. Since the most significant SNPs that we identified are non-conding variants further sequencing and functional studies are required to confirm whether the disease associations of reported markers are causal.

Nevertheless, the association found between MGMT and PTC susceptibility is a novel interesting finding. MGMT is known to be one of the most important DNA repair proteins, and it catalyzes the transfer of the methyl group from O⁶-methylguanine adducts of double-stranded DNA induced by the alkylating agents to the cysteine residue in its own molecule and thus prevents the transition from G:C to A:T point mutations by removing alkyl adducts from the O⁶ position of guanine. Loss of MGMT expression has been associated with aggressive tumour behaviour and progression in several types of neoplasia, including esophageal, hepatocellular, lung, gastric and breast carcinomas [41‐44]. The gene that is located at chromosome 10q26 spans nearly 300 kb of genomic DNA, where heterozygous deletion can often be observed in glioblastoma multiform patients [45].

Interestingly, the minor allele G of rs2296675 in MGMT had been previously shown to increase overall cancer risk of cancer across multiple tissues [per minor allele OR = 1.30, 95% CI 1.19,1.43, P = 4.1 × 10⁻⁸] [46], but risk of thyroid cancer was not specifically examined in the published CLUE II cohort study. Nevertheless, a link between MGMT and thyroid cancer had already been established in few studies. First, it was shown that expression level of MGMT protein was significantly downregulated in malignant compared to benign thyroid lesions [47]. Secondly, the methylation status of MGMT has also been investigated in thyroid neoplasia [48, 49]. Correlation of MGMT expression and promoter methylation with genomic instability in PTC patients had been then reported [50]. Functional studies are needed to decipher the biological properties of the methyltransferase and to understand how sequence variants may alter its functions and lead to the development of PTC. To our knowledge, the suggestive association between PCNA and PTC risk has not been observed in other populations, while a protective effect for carriers of the minor allele of rs2227869 in ERCC5, a SNP that is not present on the SNP Cancer Panel array, was reported in the Portuguese population [51]. Hence replication of the association study with a focus on MGMT, ERCC5 and PCNA SNPs in other populations, including both irradiated and non-irradiated PTC patients and matched healthy subjects is also warranted. More widely, identification of genetic modifiers of radiation-associated carcinogenesis may thus be a step forward to allow future personalized cancer risk prediction and may serve in prevention endeavours or public health response to widespread radiation exposure.