Longitudinal Associations Between TPO Gene Variants and TPOAb Seroconversion In A Population Based Study: Tehran Thyroid Study (TTS)

22 Background: Autoimmune thyroid diseases (AITD) are among the most common autoimmune diseases in the world. They are usually accompanied by the presence of anti- thyroid antibodies as the early predictive marker. Genetic determinants of the 25 susceptibility to develop thyroid antibodies are still poorly understood. This study aimed 26 to investigate the relation between thyroid peroxidase (TPO) gene variants (53 SNPs) and 27 positive TPOAb and also to evaluate the effect of some environmental factors on changes 28 from negative to positive TPOAb (Seroconversion). 29 Methods: Participants from the Tehran Thyroid Study (TTS) in phases 1 and 2 (N=5317, ≥ 30 20 years) were evaluated for the positive TPOAb and its relationship with 53 SNPs from 31 TPO gene (a cross-sectional approach). At the second stage of the study (a longitudinal 32 approach), negative TPOAb participants (control group, N= 4815) were followed up for 33 about 5.5 (5.54±1.62) years until they have had positive results for TPOAb (“TPOAb 34 seroconversion”). The association between TPO gene polymorphisms and TPOAb 35 seroconversion was evaluated using logistic regression analysis and SKAT (sequence 36 kernel association test) package. 37 Results: In cross-sectional analyses, 17 SNPs were associated with TPOAb positivity (521 38 positive TPOAb participants) after the adjustment for age, sex, body mass index (BMI), 39 smoking, the number of parity and oral contraceptive consumption (P <0.05). In 40 longitudinal analyses, there was an association between TPOAb seroconversion and four 41 SNPs before, and three SNPs after adjustment (P <0.05). 42 Conclusions: TPOAb seroconversion could be affected by some thyroid peroxidase gene 43 variants.


48
Autoimmune thyroid diseases (ATIDs) are among the most prevalent type of autoimmune 49 disorders (1). Although the exact pathogenesis of these disorders is not yet understood, 50 there is increasing evidence in favor of a role of genetic factors in collaboration with 51 environmental triggers (2). The basis for development of these disorders is production of 52 antibodies against cellular and molecular structures of thyroid gland. Although thyroid 53 peroxidase antibody (TPOAb) has not been identified as a direct cause of thyroid cell 54 destruction, there is a strong association between TPOAb and autoimmune thyroid 55 disorders and they are present in the serum of 90% to 95% of Hashimoto thyroiditis 56 patients (3). This association make them a reliable serological marker for diagnosis of 57 AITDs. The prevalence of anti-thyroid antibodies is between 5% to 24% among different 58 communities. This prevalence has been reported above 10% in a study that has been 59 performed in the framework of National Health and Nutrition Examination Survey 60 (NHANES); with a prevalence of 13% for TPOAb and 11.5% for Thyroglobulin 61 Antibody (TgAb) (4). The prevalence of TPOAb was reported 12.8% in Tehran Thyroid 62 Study (TSS) (5).

63
Genetic background plays the most important role in predisposition to an autoimmune 64 disorder (6-8). Preliminary studies for determination of genetic contribution to 65 autoimmune thyroid disorders performed by "candidate gene identification" approach 66 and mainly focused upon the genes having a role in the regulation of the immune system. 67 With the introduction of the Genome-wide association studies (GWAS), it has become 68 possible to perform genotyping on numerous individuals and at a high rate (9).  The purpose of this study was to investigate the relation between the variants in TPO locus 83 (53 SNPs) and TPOAb positivity/seroconversion and also to evaluate the effect of some 84 environmental factors on the conversion.

87
This study was conducted in the framework of the Tehran Thyroid Study (TTS); a cohort 88 study, being performed in the context of Tehran Lipid and Glucose Study (TLGS), to 89 collect comprehensive information on the thyroid diseases and their long-term 90 consequences in the population of Tehran, the capital of Iran. TLGS and TTS have been 91 described extensively elsewhere (13). Briefly, TTS has been started at 1997. It designed 92 in two stages, first stage was cross-sectional (phase 1) and second was a longitudinal 93 study (phase 2, 3, and 4). The length of each study phase was about three years and the 94 intervals between phases were four years. A total population of TTS was 5783. In the 95 first phase 4174 and in second phase 1609 new subjects participated. The subjects of TTS 96 were adults (aged ≥ 20 years) with thyroid function test results (5).

97
In the present study, the first stage was a cross-sectional (the first phase of TTS), and the 98 second one was a longitudinal study (phases 2-4 of TTS). The study population was TTS 99 participants who had genotype data for selected polymorphisms of the TPO gene and also 100 had information for TPOAb test results at baseline (first and second phase of TTS).

101
Pregnant women (n=40) were excluded. In the cross-sectional stage, we examined the 102 correlation of different polymorphisms genotypes of the TPO gene with positive TPOAb.

103
In the second stage of the study, TPOAb positive subjects (n=521) were excluded from 104 the analysis, and negative TPOAb subjects (n=4237) were examined in subsequent 105 phases until TPOAb seroconversion (until phase 4). Flowchart of participants through the 106 study is shown in Fig.1.

107
The effect of polymorphisms on seroconversion was evaluated in the presence of some 108 probable effective factors, such as age, sex, BMI, smoking, the number of parity and the 109 use of Oral Contraceptive pills (OCPs).

111
TPOAb measurements were performed on frozen serum samples. Measurement for all 112 samples were done in the same day using IEMA (Immunoenzymometric assay) method 113 by monoband Inc. Lake Forest, CA 92630, USA kit. Inter and intra-assay CVs were 3.9% 114 and 4.7%, respectively. The normal (negative) range defined for this kit was less than 35 115 IU/ml (5). Body mass index (BMI) was calculated as weight in kilograms divided by the 116 height in meters squared. Weight and height were taken by trained health care provider 117 and were measured according to the standard protocol. Data on Parity, smoking and OCP 118 consumption (biphasic or triphasic contraceptive tablets) were obtained by questionnaire.

119
Smoking status was categorized as ever (daily or sometimes consumption) and never 120 smokers by question at questionnaire (14).   Forty-nine SNPs with MAF greater than 0.05 (4 SNPs had MAF<0.05) were considered 132 for association analysis by logistic regression (Supplementary Table 1 (Table 3). Among these SNPs, rs6605278 showed statistically significant association 182 before and after adjustment.

183
In longitudinal approach, our results showed that age and BMI had significant effect on the 184 association between aforementioned SNPs and seroconversion. Age had a protective 185 effect on the TPOAb seroconversion risk while BMI increased it (P<0.001). For almost 186 all polymorphisms, the effect of the confounders was the same; Exceptions: In the two 187 SNPs (rs9678469, rs4927616) the effect of age was not significant. In two SNPs 188 (rs1514684, rs13431646) the effect of smoking was not significant. In one SNP 189 (rs9678469) the effect of BMI was not significant. About number of parities, significant 190 association (protective effect) was found only in two SNPs (rs938330 and rs13431646).

191
And about OCP consumption, just in one SNP (rs11682968) significant association with 192 positive TPOAb (seroconversion) was found.

194
Genetic analysis using the PLINK software showed that after adjustment for the variables of  (19). This is conceivable that slight 242 changes in the TPO structure occurs following single residual substitution, may trigger 243 autoimmune reaction.

244
The next associated SNP from the cross stage is rs7048722 that has previously reported as 245 associated variant (17). In our study, this intronic variant is present in the associated 246 block no. 11. We recognized rs11675434, which is located near the TPO gene, in block no.4. This SNP has 248 been reported in a GWAS meta-analyses in 18,297 individuals for TPOAb-positivity and 249 in 12,353 individuals for TPOAb serum levels (11) and showed significant association 250 with both phenotypes (p=1.5x10 -6 & 1.4x10 -13 respectively). Considering the strong 251 association of this variant, it seems probable that this block has been recognized as