Correction
7 Aug 2018: Campo C, Köhler A, Figlioli G, Elisei R, Romei C, et al. (2018) Correction: Inherited variants in genes somatically mutated in thyroid cancer. PLOS ONE 13(8): e0202208. https://doi.org/10.1371/journal.pone.0202208 View correction
Figures
Abstract
Background
Tumour suppressor genes when mutated in the germline cause various cancers, but they can also be somatically mutated in sporadic tumours. We hypothesized that there may also be cancer-related germline variants in the genes commonly mutated in sporadic well-differentiated thyroid cancer (WDTC).
Methods
We performed a two-stage case-control association study with a total of 2214 cases and 2108 healthy controls from an Italian population. By genotyping 34 single nucleotide polymorphisms (SNPs), we covered a total of 59 missense SNPs and SNPs located in the 5' and 3' untranslated regions (UTRs) of 10 different genes.
Results
The Italian1 series showed a suggestive association for 8 SNPs, from which three were replicated in the Italian2 series. The meta-analysis revealed a study-wide significant association for rs459552 (OR: 0.84, 95%CI: 0.75–0.94) and rs1800900 (OR: 1.15, 95%CI: 1.05–1.27), located in the APC and GNAS genes, respectively. The APC rs459552 is a missense SNP, located in a conserved amino acid position, but without any functional consequences. The GNAS rs1800900 is located at a conserved 5'UTR and according to the experimental ENCODE data it may affect promoter and histone marks in different cell types.
Conclusions
The results of this study yield new insights on WDTC, showing that inherited variants in the APC and GNAS genes can play a role in the etiology of thyroid cancer. Further studies are necessary to better understand the role of the identified SNPs in the development of WDTC and to functionally validate our in silico predictions.
Citation: Campo C, Köhler A, Figlioli G, Elisei R, Romei C, Cipollini M, et al. (2017) Inherited variants in genes somatically mutated in thyroid cancer. PLoS ONE 12(4): e0174995. https://doi.org/10.1371/journal.pone.0174995
Editor: Amanda Ewart Toland, Ohio State University Wexner Medical Center, UNITED STATES
Received: November 15, 2016; Accepted: March 17, 2017; Published: April 14, 2017
Copyright: © 2017 Campo et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: We are grateful to the study participants and the study investigators and coordinators for work in recruitment of subjects. The study has received financial support from the Istituto Toscano Tumori.
Competing interests: No potential conflicts of interest were disclosed.
Introduction
Thyroid cancer (TC) is a common endrocrine malignancy. The majority of all thyroid cancers are well-differentiated (WDTC). Among them the most frequent subtype is papillary TC (PTC), followed by follicular TC (FTC); they represent 80–84% and 6–10% of all thyroid carcinomas, respectively [1]. The incidence of TC in the world has been increasing in recent years and it accounts for about 1% of all oncological diseases based on the 2012 GLOBOCAN data [2]. Explanations for the increasing thyroid cancer incidence are controversial [3]. Some experts suggest that the rise in the number of new cancers is due to the increased diagnostic intensity; other experts believe that it is associated to lifestyle changes. Many individual and environmental factors have been considered as risk factors for TC. Indeed, thyroid gland seems to be an organ particularly vulnerable to ionizing radiation and the risk for TC in individuals exposed to radiation at young age persists throughout life [4,5]. Benign nodules/adenomas and goiter seem to have an important role in the predisposition to TC [6], as well as an inherited genetic susceptibility [7,8]. Common genetic variants in low-penetrance genes may interact with each other and with the environment, regulating TC susceptibility [9,10].
Many association studies, both designed candidate gene and genome-wide approaches, have been conducted to identify genetic variants affecting individual susceptibility to WDTC. In particular pathways involving DNA repair (OGG1, XRCC1, XRCC2, XRCC3, ATM), cell-cycle control (TP53), detoxification (CYP1A1, GSTs) and thyroid function (TG) have been intensively studied. Some genetic variants of these genes have been shown to be TC susceptibility factors in humans [11–13]. Genome wide association studies (GWASs) of WDTC have reported associations with single nucleotide polymorphisms (SNPs) at 2q35 (DIRC3), 8p12 (NRG1), 9q22.33 (FOXE1) and 14q13.3 (NKX2-1). These genetic risk factors were first identified in Icelandic and European studies, then replicated and confirmed across different populations [14–23]. Moreover, new risk loci at 20q11.22-q12 (DHX35) and 14q24.3 (BATF) were recently discovered in a GWAS conducted on an Italian population and followed by replication on three different cohorts (Italian, Polish, Spanish) [23]. Other seven loci showed evidence of association with WDTC only among Italians [22,23].
This study aimed to identify novel SNPs involved in WDTC susceptibility, selected in genes somatically mutated in the disease. These genetic variants were analyzed in a two-stage case-control association study. After identification of the candidate genes/polymorphisms associated with the disease, further verification was done by analyzing the functional consequences using in silico prediction tools.
Materials and methods
Study population
The first Italian case group consisted of 1523 WDTC patients (85% PTC, 11% FTC, 27.4% males, median age 53 years), while the control group comprised 1610 healthy individuals (66.9% males, median age 57 years). The SNPs which showed significant associations (p <0.05) were replicated in a second Italian sample set, totaling 691 affected individuals (100% PTC, 22.4% males, median age 47 years) and 498 controls (23.1% males, median age 53 years), Table A in S1 File. The patients were contacted during their routine follow-up by the Department of Endocrinology and Metabolism (University Hospital of Cisanello, Pisa). All patients had a histologically or cytologically confirmed WDTC. The healthy controls were blood donors (University Hospital A. Meyer, Florence) and workers (mainly physicians, nurses and paramedical staff), of the same Hospital of Pisa, without known thyroid disease. The study was authorized by the Pisa University Medical Ethics Committee (CEAVNO, North-West Tuscany Ethical Committee). According to the Helsinki declaration, all individuals signed an informed consent to participate to the study and to allow genetic analyses on their biological samples.
Gene and SNP selection
The gene selection was done using the browser COSMIC (http://cancer.sanger.ac.uk/cosmic). Starting from the 20 top genes mutated in PTC, we considered those with a mutation frequency higher than 5%, obtaining a total of 11 genes (APC, BRAF, CDH1, CDKN2A, CTNNB1, EGFR, GNAS, IDH1, SMAD4, TP53, TSHR). We considered all gene transcripts, then selected all missense variants and SNPs located in the 5´UTR and 3´UTR, with a global minor allele frequency (MAF) >10% and MAF >5% in the Italian population. We applied a tagging SNP approach and checked the linkage disequilibrium (LD) of those variants in the Hap Map Project (http://hapmap.ncbi.nlm.nih.gov/) and in the 1000 Genomes Project (http://www.1000genomes.org/). By selecting SNPs with a pair wise LD r2<0.8, we reduced the number of SNPs to be genotyped to 34 (5´UTR: 6, 3´UTR: 22 and missense: 6), covering 59 SNPs located in 10 different genes, Table B in S1 File. The BRAF gene was not further included in our analysis since it did not present any variants fulfilling our selection criteria.
Genotyping
Genotyping was performed using allele specific PCR based TaqMan Assays (Life Technologies) and KASPar v4.0 Assays (Kbioscience). HOT FIREPol Probe qPCR Mix Plus (ROX) from Solis BioDyne (Tartu, Estonia) was used for the TaqMan Assays, whereas the KASPar genotyping reaction from LGC was prepared according to the KBioscience’s conditions and products (KASP Assay mix, containing the target specific primers, and KASP Master mix). PCR reactions were performed in a 384-well plate format using 5ng DNA in a total reaction volume of 5μl per well. Endpoint genotyping detection was performed using ViiA7 Real-Time PCR System (Applied Biosystems, Weiterstadt, Germany).
Statistical analysis
The genotype distribution in controls was tested for Hardy-Weinberg equilibrium (HWE) using the chi square (χ2) test. SNPs with a significant deviation from HWE (p<0.005) were excluded from the analysis. Odd ratios (ORs) and 95% confidence intervals (95% CIs) were calculated by logistic regression and adjusted for the covariates age and sex (PROC LOGISTIC, SAS Version 9.2; SAS Institute, Cary, NC). For all SNPs, ORs were calculated according to the allelic model using the major allele as reference. For SNPs with allelic pValue <0.05, we also evaluated the associations according to co-dominant and dominant model using the major homozygote genotype as reference. The pValue <0.05 was considered statistically significant. The power of the study was evaluated using the Quanto Version 1.2 (USC Biostats), assuming a case-control ratio of 1:1 in phase1 and 1:0.7 in phase 2, a baseline risk of 0.001, MAF of 0.1–0.5, OR of 1.0–1.5 according to a log-additive inheritance model and a significance level of 0.05 (2-sided in phase 1; 1-sided in phase 2), Table C in S1 File. Furthermore, the meta-analysis of the two Italian sets was conducted. ORs and 95% CIs were calculated using both fixed-effect (Mantel-Haenszel) and random-effect (DerSimonian Laird) models. The significance of overall OR was evaluated by the Z-test. Heterogeneity between studies was evaluated with a Q-test. Heterogeneity was considered significant when p<0.05. In case of no significant heterogeneity, point estimates and 95% CI were assessed using the fixed-effect model; otherwise, random effects model was employed. Bonferroni´s correction was used, taking into account the number of SNPs involved in the meta-analysis, and the novel statistical significant threshold was 6.25×10−3.
In silico analysis
For each SNP significantly associated with WDTC risk, we carried-out in silico analysis to determine its predicted functional effect and to understand the molecular events controlling the gene expression. We performed SIFT and PolyPhen predictions to determine the possible effect of amino acid substitutions on protein function [24]. We used the HaploRegv4.1 software (http://www.broadinstitute.org/mammals/haploreg/haploreg.php) to investigate the influence of the SNPs on gene expression [25]. It uses LD and SNP information from the 1000 Genomes Project to map known genetic variants onto ENCODE, providing a potential mechanism for a SNP influence. ENCODE is a database which visualizes the influence of SNPs based on their effect on chromatin structure, histone modifications, sequence conservation across mammals, regulatory protein binding sites and their effect on regulatory motifs and transcription factor binding sites (TFBSs). This tool includes also an expanded library of expression quantitative trait loci (eQTL) [26]. As a complementary analysis we used the tool RegulomeDB (http://regulomedb.org/) to identify DNA features and regulatory elements that contain the coordinate of the SNP.
Results
A case-control association study was conducted with patients with sporadic WDTC, mainly PTC, and healthy controls from an Italian population. The two subtypes, papillary and follicular TC, were not analyzed separately due to the small number of FTC (11% of cases). Genotype distribution of three SNPs (rs3864004 in CTNNB1; rs2227983 and rs10228436 in EGFR) in controls deviated from HWE (p <0.005) and they were excluded from further analyses. The associations according to the allelic model between the SNPs and WDTC are shown in Table 1. The analysis adjusted for the covariates age and sex, revealed a protective role (pValue <0.05) for 5 SNPs: rs459552 within the APC gene; rs884904 within EGFR; rs8022600 and rs2268477 within TSHR; rs7229678 within the SMDA4 gene. An increased risk of WTDC (pValue <0.05) was found for the rare allele carriers of rs7144481 within TSHR gene, and of rs1800900 and rs7121 within the GNAS gene.
(*) SNPs with a p-ValueHWE <0.005 in controls were excluded from further analyses.
Regarding the associated SNPs (pValue <0.05), we also evaluated the genotype risks in the co-dominant and dominant models. For rs1800900 within the GNAS gene, the association of minor homozygote genotype in comparison to the major homozygote genotype was stronger than the association according to the allelic model (pValue = 8.82 x 10−3 vs. pValue = 0.01). A dominant model described the association of rs7144481 in TSHR with WDTC better than the allelic model (pValue = 0.01 vs. pValue = 0.04).
To validate these findings, we conducted a replication study of the 8 SNPs, statistical significant at the 0.05 level, by genotyping an additional Italian set of 691 cases and 498 controls. Only three SNPs showed ORs in the same direction and at the same level as in the first set (rs459552, rs7144481 and rs1800900). The combined analysis of the two data sets identified two promising variants associated with WDTC risk: rs459552 within APC, (OR = 0.84, 95CI% = 0.75–0.94, pValue = 4.00x10-3, allelic model) and rs1800900 within GNAS, (OR = 1.15, 95CI% = 1.05–1.27, pValue = 4.00x10-3, allelic model). All the results are shown in Tables 2 and 3. In summary, the joint analysis detected a decreased risk of WDTC for one SNP (rs459552) in APC and a positive association for one SNP (rs1800900) within GNAS.
To explore if the SNPs identified as WDTC risk markers have potential regulatory functions, in silico methods were used to predict the effect of amino acid substitutions on protein function and to understand the molecular events controlling gene expression. The results shown in Table 4 summarize all the information obtained by utilizing the HaploRegv4.1 software and RegulomeDB. The GNAS gene SNP rs1800900 was identified by SiPhy to be located in a conserved region. As reported by the ENCODE project data, rs1800900 may affect DNAse hypersensitivity sites and enhancer and promoter histone marks in different cell types. Several other linked SNPs show similar functional characteristics. Moreover, Regulome analysis revealed that according to the ChIP-seq data this SNP is involved in GATA1 protein binding sites (data not shown).
The APC gene SNP rs459552 is causing a missense amino acid change (Val1822Asp) in three Ensembl transcripts. However, in all three transcripts it seems to be tolerated by SIFT and considered to be benign by PolyPhen prediction. According to HaploReg v4.1 predictions, this variant alters the binding sites of Bbx and LRH1 transcription factors and it is located in a conserved amino acid position.
Discussion
To identify new risk variants predisposing to WDTC we investigated genes somatically mutated in the disease for inherited germline polymorphisms, and performed an association study in two case-control sets from Italy. By genotyping of 34 SNPs, we got information of 53 missense SNPs and SNPs located in the 5’ and 3’UTRs of 10 different genes. The Italian1 series showed a suggestive association for 8 SNPs and three of them provided evidence of association in the replication set. Finally, the meta-analysis revealed a study-wide significant association for rs459552 and rs1800900, in the APC and GNAS genes, respectively.
Although the origin of the samples was the same, there were some differences in the two data sets. First, the Italian1 series was dominated by female cases, while men were the larger group among controls. The Italian2 series was more homogeneous regarding the gender distribution, with a female dominance both among cases and controls. Also, both the cases and controls were older in the first set than in the second one. However, there was no significant difference in allele frequencies between women and men among cases or controls. To avoid sampling biases the analysis was adjusted for age and sex, but the replication series did not have enough statistical power to confirm all the statistically significant results of the first data set. In the first phase, we had an over 80% power to detect an OR of 1.2 for variants with MAF > 20%, however, in the second phase the power was reduced to 56% for variants with MAF of 20%, increasing to 71% for variants with MAF of 50%. As most of the associations were on opposite directions in the two sample sets, they could be considered as false positives in the first sample set. However, the strength of the association of rs459552 (pValue = 4.00x10-3) and rs1800900 (pValue = 4.00x10-3), together with the in silico data suggests that these two associations may be real.
Our gene selection was based on somatic TC mutation data available at the COSMIC browser. Comparison with the TCGA data showed very similar mutation spectrum, with BRAF being the most commonly mutated gene in PTC [27]. The mutation frequencies in general were, however, usually lower than in COSMIC, also for the genes included in our study. In the current data APC is reported to be mutated in 10% of PTC by COSMIC, but only in 0.7% of PTC by TCGA. For GNAS, COSMIC reports a 3% mutation frequency, while in TCGA GNAS is not included among the top 25 genes.
The Adenomatous Polyposis Coli (APC) gene encodes a tumor suppressor protein that acts as an antagonist of the Wnt signaling pathway. It is also involved in other processes including cell migration and adhesion, transcriptional activation, and apoptosis. Defects in this gene cause familial adenomatous polyposis (FAP), an autosomal dominant pre-malignant disease. FAP is an inherited condition caused by germline mutations in the APC gene [28]. In addition, somatic mutations of APC are observed in sporadic colorectal neoplasm, suggesting that disruption of this tumor suppressor gene may play a role in both familial as well as acquired colorectal tumorigenesis [29]. APC has been found to be expressed in normal as well as in neoplastic human thyroid tissue, in which multiple forms of specific RNA transcripts have been detected; it may be hypothesized that alterations in APC are likely candidates for a pathogenic role in thyroid tumorigenesis [30]. A study of mutations in Beta-catenin and APC genes in PTC, FTC and benign thyroid tumors suggested that mutations in these genes are rare and that the activation of the Wnt signaling pathway may not contribute to pathogenesis in human PTC and FTC [31]. However, a better insight into the correlation of APC gene mutations on the clinical manifestation of PTC requires further investigation as suggested by a study, which investigated gene variants associated with malignant thyroid disease in FAP. In this patient group carrying an APC mutation the prevalence of thyroid malignant diseases was estimated to be 3.7% [32].
In our study, the variant allele carriers of the APC SNP rs459552 had a 16% decreased risk of WDTC. The SIFT and PolyPhen predictions did not reveal any damaging effect for this SNP. The effects of APC rs459552 substitutions on colorectal cancer (CRC) risk has been investigated but no association between the genetic variant and CRC risk in the Scottish population was observed [33]. Instead, previous studies investigating the possible influence of rs459552 on the association between dietary and lifestyle factors and CRC showed significant results. Indeed, the polymorphism of APC gene may have functional significance, especially in the presence of diet, life-style, or other environmental exposures [34]. In a group of Portuguese subjects, a significant interaction was found between the rs459552 polymorphisms and dietary intake of high fat, cholesterol, calcium, and fiber for CRC risk [35].
Our statistical analysis showed a strong association between rs1800900, located in the GNAS gene, with WDTC. GNAS encodes the α-subunit of the Gs protein (Gsα), which binds GTP and stimulates adenylyl cyclase [36]. Somatic mutations in Gsα, and in TSHR gene, are the main reason for autonomously functioning, TSH-independent, thyroid nodules in iodine-deficient regions of the world [37,38]. In Gsα, activating mutations result in constitutive activation of the cAMP cascade, which determines clonal expansion of the affected thyroid cells and consequent thyroid autonomy [39]. Studies from Japan reported that somatic mutations of the Gsα and TSHR genes were caused by point mutations with single amino acid substitutions, which were confined to the hyperfunctioning nodule and which were not presented in neighboring normal thyroid parenchyma [40–42]. No information about rs1800900 has been reported in literature. This SNP seems to be involved in GATA1 protein binding. It has been recognized that GATA1, binding erythroid enhancers of specific genes, acts as a transcriptional activator and it is involved in the differentiation of hematopoietic stem cells to erythrocytes [43]. GATA1 collaborates with BRG1 to disrupt chromatin, regulating the reorganization of nucleosomes at enhancer elements and the binding of transcription factors to those specific DNA sequences [44].
In conclusion, our study provides evidence for the hypothesis that polymorphic variants in the APC and GNAS genes, which are commonly mutated in sporadic TC, may play a role in the etiology of TC. Further analysis and additional investigations are needed to confirm our results and to validate the accuracy and the specificity of our in silico predictions.
Supporting information
S1 File.
Table A. Characteristics of the study population. Table B. SNP selection. In bold all the SNPs genotyped. SNPs with MAF<0.10 were excluded from the analysis as were the SNPs with a bad sequence (*). From SNPs in high LD (r2>0.80; **) only one SNP was genotyped. a) Assay not available. Instead a SNP in high LD (r2 = 0.99) was genotyped, rs2518720. Table C. Statistical power of the TC association study according to Quanto 1.2 power calculation tool. Based on our study data we used a case-control ratio of 1:1 in phase 1 and 1:0.7 in phase 2, a baseline risk of 0.001, risk allele frequency between 0.1 and 0.5, odds ratio between 1.0 and 1.5, according to a log-additive inheritance model at a significance level of 0.05 (2-sided in phase 1; 1-sided in phase 2).
https://doi.org/10.1371/journal.pone.0174995.s001
(DOCX)
Acknowledgments
We are grateful to the study participants and the study investigators and coordinators for work in recruitment of subjects. The study has received financial support from the Istituto Toscano Tumori. All authors had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Author Contributions
- Conceptualization: KH FG SL AF.
- Data curation: KH FG SL AF.
- Formal analysis: CC.
- Funding acquisition: KH FG SL AF.
- Investigation: CC AK GF.
- Methodology: CC AF.
- Project administration: CC AK GF KH AF MC.
- Resources: CC RE CR FB KH AF.
- Software: CC AK GF.
- Supervision: AF.
- Validation: CC AF.
- Visualization: CC.
- Writing – original draft: CC.
- Writing – review & editing: CC AF.
References
- 1. Paschke R, Lincke T, Muller SP, Kreissl MC, Dralle H, et al. (2015) The Treatment of Well-Differentiated Thyroid Carcinoma. Dtsch Arztebl Int 112: 452–458. pmid:26205749
- 2. Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, et al. (2015) Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer 136: E359–386. pmid:25220842
- 3. Morris LG, Sikora AG, Tosteson TD, Davies L (2013) The increasing incidence of thyroid cancer: the influence of access to care. Thyroid 23: 885–891. pmid:23517343
- 4. Dilas LT, Bajkin I, Icin T, Paro JN, Zavisic BK (2012) [Iodine and thyroid gland with or without nuclear catastrophe]. Med Pregl 65: 489–495. pmid:23297615
- 5. Niazi AK, Niazi SK (2011) Endocrine effects of Fukushima: Radiation-induced endocrinopathy. Indian J Endocrinol Metab 15: 91–95. pmid:21731864
- 6. Meinhold CL, Ron E, Schonfeld SJ, Alexander BH, Freedman DM, et al. (2010) Nonradiation risk factors for thyroid cancer in the US Radiologic Technologists Study. Am J Epidemiol 171: 242–252. pmid:19951937
- 7. Hemminki K, Dong C (2001) Population-based study of familial medullary thyroid cancer. Fam Cancer 1: 45–49. pmid:14574015
- 8. Pellegriti G, Frasca F, Regalbuto C, Squatrito S, Vigneri R (2013) Worldwide increasing incidence of thyroid cancer: update on epidemiology and risk factors. J Cancer Epidemiol 2013: 965212. pmid:23737785
- 9. Canzian F, Amati P, Harach HR, Kraimps JL, Lesueur F, et al. (1998) A gene predisposing to familial thyroid tumors with cell oxyphilia maps to chromosome 19p13.2. Am J Hum Genet 63: 1743–1748. pmid:9837827
- 10. Sturgis EM, Li G (2009) Molecular epidemiology of papillary thyroid cancer: in search of common genetic associations. Thyroid 19: 1031–1034. pmid:19803789
- 11. Garcia-Quispes WA, Perez-Machado G, Akdi A, Pastor S, Galofre P, et al. (2011) Association studies of OGG1, XRCC1, XRCC2 and XRCC3 polymorphisms with differentiated thyroid cancer. Mutat Res 709–710: 67–72. pmid:21414327
- 12. Landa I, Robledo M (2011) Association studies in thyroid cancer susceptibility: are we on the right track? J Mol Endocrinol 47: R43–58.
- 13. Damiola F, Byrnes G, Moissonnier M, Pertesi M, Deltour I, et al. (2014) Contribution of ATM and FOXE1 (TTF2) to risk of papillary thyroid carcinoma in Belarusian children exposed to radiation. Int J Cancer 134: 1659–1668. pmid:24105688
- 14. Gudmundsson J, Sulem P, Gudbjartsson DF, Jonasson JG, Sigurdsson A, et al. (2009) Common variants on 9q22.33 and 14q13.3 predispose to thyroid cancer in European populations. Nat Genet 41: 460–464. pmid:19198613
- 15. Matsuse M, Takahashi M, Mitsutake N, Nishihara E, Hirokawa M, et al. (2011) The FOXE1 and NKX2-1 loci are associated with susceptibility to papillary thyroid carcinoma in the Japanese population. J Med Genet 48: 645–648. pmid:21730105
- 16. Wang YL, Feng SH, Guo SC, Wei WJ, Li DS, et al. (2013) Confirmation of papillary thyroid cancer susceptibility loci identified by genome-wide association studies of chromosomes 14q13, 9q22, 2q35 and 8p12 in a Chinese population. J Med Genet 50: 689–695. pmid:23847140
- 17. Liyanarachchi S, Wojcicka A, Li W, Czetwertynska M, Stachlewska E, et al. (2013) Cumulative risk impact of five genetic variants associated with papillary thyroid carcinoma. Thyroid 23: 1532–1540. pmid:23659773
- 18. Gudmundsson J, Sulem P, Gudbjartsson DF, Jonasson JG, Masson G, et al. (2012) Discovery of common variants associated with low TSH levels and thyroid cancer risk. Nat Genet 44: 319–322. pmid:22267200
- 19. Takahashi M, Saenko VA, Rogounovitch TI, Kawaguchi T, Drozd VM, et al. (2010) The FOXE1 locus is a major genetic determinant for radiation-related thyroid carcinoma in Chernobyl. Hum Mol Genet 19: 2516–2523. pmid:20350937
- 20. Jones AM, Howarth KM, Martin L, Gorman M, Mihai R, et al. (2012) Thyroid cancer susceptibility polymorphisms: confirmation of loci on chromosomes 9q22 and 14q13, validation of a recessive 8q24 locus and failure to replicate a locus on 5q24. J Med Genet 49: 158–163. pmid:22282540
- 21. Pereda CM, Lesueur F, Pertesi M, Robinot N, Lence-Anta JJ, et al. (2015) Common variants at the 9q22.33, 14q13.3 and ATM loci, and risk of differentiated thyroid cancer in the Cuban population. BMC Genet 16: 22. pmid:25879635
- 22. Kohler A, Chen B, Gemignani F, Elisei R, Romei C, et al. (2013) Genome-wide association study on differentiated thyroid cancer. J Clin Endocrinol Metab 98: E1674–1681. pmid:23894154
- 23. Figlioli G, Kohler A, Chen B, Elisei R, Romei C, et al. (2014) Novel genome-wide association study-based candidate loci for differentiated thyroid cancer risk. J Clin Endocrinol Metab 99: E2084–2092. pmid:25029422
- 24. Flanagan SE, Patch AM, Ellard S (2010) Using SIFT and PolyPhen to predict loss-of-function and gain-of-function mutations. Genet Test Mol Biomarkers 14: 533–537. pmid:20642364
- 25. Ward LD, Kellis M (2012) HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Res 40: D930–934. pmid:22064851
- 26. Li H, Deng H (2010) Systems genetics, bioinformatics and eQTL mapping. Genetica 138: 915–924. pmid:20811929
- 27. Cancer Genome Atlas Research N (2014) Integrated genomic characterization of papillary thyroid carcinoma. Cell 159: 676–690. pmid:25417114
- 28. Cameselle-Teijeiro J, Ruiz-Ponte C, Loidi L, Suarez-Penaranda J, Baltar J, et al. (2001) Somatic but not germline mutation of the APC gene in a case of cribriform-morular variant of papillary thyroid carcinoma. Am J Clin Pathol 115: 486–493. pmid:11293895
- 29. Zeki K, Spambalg D, Sharifi N, Gonsky R, Fagin JA (1994) Mutations of the adenomatous polyposis coli gene in sporadic thyroid neoplasms. J Clin Endocrinol Metab 79: 1317–1321. pmid:7962323
- 30. Kumamoto K, Ishida H, Ohsawa T, Ishibashi K, Ushiama M, et al. (2015) Germline and somatic mutations of the gene in papillary thyroid carcinoma associated with familial adenomatous polyposis: Analysis of three cases and a review of the literature. Oncol Lett 10: 2239–2243. pmid:26622826
- 31. Miyake N, Maeta H, Horie S, Kitamura Y, Nanba E, et al. (2001) Absence of mutations in the beta-catenin and adenomatous polyposis coli genes in papillary and follicular thyroid carcinomas. Pathol Int 51: 680–685. pmid:11696170
- 32. Martayan A, Sanchez-Mete L, Baldelli R, Falvo E, Barnabei A, et al. (2010) Gene variants associated to malignant thyroid disease in familial adenomatous polyposis: a novel APC germline mutation. J Endocrinol Invest 33: 603–606. pmid:20935450
- 33. Theodoratou E, Campbell H, Tenesa A, McNeill G, Cetnarskyj R, et al. (2008) Modification of the associations between lifestyle, dietary factors and colorectal cancer risk by APC variants. Carcinogenesis 29: 1774–1780. pmid:18375958
- 34. Slattery ML, Samowitz W, Ballard L, Schaffer D, Leppert M, et al. (2001) A molecular variant of the APC gene at codon 1822: its association with diet, lifestyle, and risk of colon cancer. Cancer Res 61: 1000–1004. pmid:11221825
- 35. Guerreiro CS, Cravo ML, Brito M, Vidal PM, Fidalgo PO, et al. (2007) The D1822V APC polymorphism interacts with fat, calcium, and fiber intakes in modulating the risk of colorectal cancer in Portuguese persons. Am J Clin Nutr 85: 1592–1597. pmid:17556698
- 36. Weinstein LS, Yu S, Warner DR, Liu J (2001) Endocrine manifestations of stimulatory G protein alpha-subunit mutations and the role of genomic imprinting. Endocr Rev 22: 675–705. pmid:11588148
- 37. Parma J, Duprez L, Van Sande J, Cochaux P, Gervy C, et al. (1993) Somatic mutations in the thyrotropin receptor gene cause hyperfunctioning thyroid adenomas. Nature 365: 649–651. pmid:8413627
- 38. Lyons J, Landis CA, Harsh G, Vallar L, Grunewald K, et al. (1990) Two G protein oncogenes in human endocrine tumors. Science 249: 655–659. pmid:2116665
- 39. Gozu HI, Bircan R, Krohn K, Muller S, Vural S, et al. (2006) Similar prevalence of somatic TSH receptor and Gsalpha mutations in toxic thyroid nodules in geographical regions with different iodine supply in Turkey. Eur J Endocrinol 155: 535–545. pmid:16990652
- 40. Nishihara E, Amino N, Maekawa K, Yoshida H, Ito M, et al. (2009) Prevalence of TSH receptor and Gsalpha mutations in 45 autonomously functioning thyroid nodules in Japan. Endocr J 56: 791–798. pmid:19550078
- 41. Kosugi S, Hai N, Okamoto H, Sugawa H, Mori T (2000) A novel activating mutation in the thyrotropin receptor gene in an autonomously functioning thyroid nodule developed by a Japanese patient. Eur J Endocrinol 143: 471–477. pmid:11022192
- 42. Vanvooren V, Uchino S, Duprez L, Costa MJ, Vandekerckhove J, et al. (2002) Oncogenic mutations in the thyrotropin receptor of autonomously functioning thyroid nodules in the Japanese population. Eur J Endocrinol 147: 287–291. pmid:12213664
- 43. Dore LC, Crispino JD (2011) Transcription factor networks in erythroid cell and megakaryocyte development. Blood 118: 231–239. pmid:21622645
- 44. Hu G, Schones DE, Cui K, Ybarra R, Northrup D, et al. (2011) Regulation of nucleosome landscape and transcription factor targeting at tissue-specific enhancers by BRG1. Genome Res 21: 1650–1658. pmid:21795385