Background

Hashimoto’s thyroiditis (HT), also known as chronic lymphocytic thyroiditis, is the most common form of autoimmune thyroid disease. HT is characterized by chronic inflammation of the thyroid gland and synthesis of thyroid peroxidase antibodies (TPO Ab) and/or thyroglobulin antibodies (TG Ab) [1]. The pathogenesis of this autoimmune disease is complex, and it is generally accepted that the infiltration of T lymphocytes and plasma cells leads to the destruction of thyroid tissue [2, 3].

HT is the most common cause of hypothyroidism in areas with excess iodine intake [4, 5]. On the worldwide scale, it is reported that the incidence of HT is higher in women than men (0.28 per 1000 population/year for women and 0.03 per 1000 population/year for men) [6]. In addition, according to Chabchoub et al. [7], the prevalence of HT is estimated to be 22.8% in Africa.

HT is multifactorial autoimmune diseases where interaction between genetic and environmental factors plays an important role [8,9,10]. Many recent studies also pointed the role of epigenetic modifications in HT pathogenesis [11]. However, family and twin studies suggested that genetic factors are the main drivers of HT. It is estimated that genetic predisposition contributes to about 80% to HT development [7, 12].

Several genetic studies suggested the contribution of two types of genes to HT susceptibility: immune regulatory genes (HLA-DR, CTLA-4, PTPN22) [13] and thyroid-specific ones (TG and TPO) [14]. Recently, several other genes (IL15, IL27, FOXP3, and SEPS1) have also been suggested to be associated with HT [15,16,17,18].

A first genome-wide association analysis (GWAS) on patients with autoimmune thyroid disease (AITD) used ImmunoChip and identified the association of four loci with HT: 2p25.1, 3q27.3/3q28, 6q15, and 11q21 [19]. Additionally, another GWAS by Inoue et al. [20] reported the association of ZFAT gene with HT severity but not with its development. Another GWAS by Oryoji et al. [21] directly compared HT and GD cases and indicated the association of rs7537605 in the VAV3 locus with HT. Finally, a GWAS by Eriksson et al. [22] has shown the association of rs6679677 near PTPN22, rs3184504 in SH2B3, and rs2517532 in the HLA class I region with a large spectrum of autoimmune diseases.

Importantly, there was only one GWAS that focused exclusively on HT to date. That study was performed in the Croatian population and included 405 cases and 433 controls, followed up by replication in 303 cases and 302 controls from two replication cohorts [23]. That study suggested an association of three novel HT variants: rs12944194 located 206 kb from SDK2, rs791903 inside IP6K3, and rs75201096 inside GNA14 [23]. These three genetic variants are located in genomic regions with good biological candidacy to HT/autoimmunity and with the regulation of thyroid function.

The main aim of our study was to analyze the association of these three newly suggested genetic variants with HT in a different ethnic group than originally reported, the non-Caucasian, Arab-Berber population from Algeria.

Methods

Subjects

We performed a case-control study that included one hundred patients with HT and one hundred and twenty-six healthy subjects from the Aures region of Algeria that belong to an Arab-Berber ethnic group. Cases and controls were recruited in the period of 2 years (from January 2016 to February 2018).

HT patients (96% of women), with a median age of 39.5 years and median body mass index (BMI) of 27.26 kg/m2, were recruited from three clinics: El Balsem el Chafy (Khenchela), Endocrinology & Metabolism Clinics of Dr. Mallem and Endocrinology & Metabolism Clinics of Dr. Heddar (Batna). HT was diagnosed by specialists in nuclear medicine based on several criteria that included clinical symptoms, ultrasound findings, biochemical measurements of thyroid-stimulating hormone (TSH), and thyroid antibody (TPO Ab), according to European Thyroid Association (ETA) guidelines (Management of Subclinical Hypothyroidism) [24]. All HT patients had high levels of TSH and TPO-Ab and were treated with levothyroxine sodium (Merck KGaA, Frankfurter Str. 250, Darmstadt, Germany). We excluded patients with a personal history of any other autoimmune disease.

A control group (64.3% of women), with a median age of 33 years and a median BMI of 24.73 kg/m2, were recruited from the University Hospital of Batna. All control participants were selected according to the following criteria: TPO Ab negative (5–34 UI/mL), TSH levels between 0.27 and 4.7 μIU/mL, and FT4 levels between 12 and 22 pmol/L. None of the control subjects had a personal history of HT or any other autoimmune disease.

Clinical and demographic characteristics of HT patients and the control group are presented in Table 1.

Table 1 Clinical and demographic characteristics of HT patients and control group
Full size table

The study was approved by the scientific-ethical committee of the University Hospital of Batna, under the number 524. Informed consent was obtained from all participants and the study was conducted according to the Declaration of Helsinki.

Blood samples and DNA extraction

Venous blood was collected in EDTA tubes (4 mL) from all study participants and stored at − 80° until extraction. Genomic DNA was extracted by the QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. DNA concentration was measured using Nanodrop ND-1000 Spectrophotometer (ND-1000, Thermo Fisher Scientific, Waltham, MA, USA).

Genotyping

We selected single nucleotide polymorphisms (SNPs) (rs12944194, rs791903, and rs75201096) from the recently published GWAS of HT in Caucasian population [23]. SNP genotyping was performed by real-time PCR using ABI PRISM 7500 Sequence Detection System (Applied Biosystems, Foster City, USA) and pre-developed TaqMan SNP genotyping assays.

Association analysis

All statistical tests were performed by Plink software version 1.07 [25]. Prior association analysis, we checked if genotypes of the three investigated SNPs were according to Hardy–Weinberg equilibrium (HWE) using an exact test [26]. Association of the three SNPs with HT was tested using a logistic regression model adjusted for age and BMI. Differences in the distribution of genotype frequencies of three SNPs were tested by Fisher’s exact test. In addition, we checked the association of these three SNPs with TPO Ab levels in the group of HT cases only, using a linear regression model adjusted for age and BMI. A logarithmic transformation of TPO Ab levels in HT cases was performed prior analysis to ensure normal distribution.

Odds ratio (OR) and confidence intervals (95% CI) were calculated using IBM SPSS Statistics for Windows, Version 20.0. (Armonk, NY: IBM Corp). We used Bonferroni corrected P value of 0.017 (0.05 corrected for 3 SNPs) for declaring significant associations.

Results

Genotype distribution of all SNPs in HT patients and the control group was according to HWE (Table 2).

Table 2 SNP characteristics in HT patients and control group
Full size table

We did not detect significant allelic association of the three investigated SNPs with HT at the Bonferroni-corrected significance threshold (Table 3). However, we observed nominal significance in the distribution of rs791903 genotype (P = 0.024) between HT and the control group. Recessive homozygotes (CC) were less frequent in HT cases than in controls (P = 0.025) (Table 4).

Table 3 Results of Association analysis for rs12944194, rs791903, and rs75201096 in HT case-control analysis
Full size table
Table 4 Genotype association of rs12944194, rs791903, and rs75201096 with HT
Full size table

No associations were detected with TPO Ab levels in the group of HT patients (Table 5).

Table 5 Results of association analysis of three SNPs with TPO-Ab levels
Full size table

Discussion

The goal of our study was to explore the association of three newly suggested HT-associated genetic variants with the risk of developing HT in 100 HT patients and 126 controls of Arab-Berber ethnicity. To our knowledge, this is the first genetic study that focused on the evaluation of the association of these genetic variants (rs12944194, rs791903, and rs75201096) with susceptibility to HT in non-Caucasian population from Algeria. Our main result is the observation of a nominally significant decrease in the frequency of rs791903 recessive genotype in HT cases versus controls.

The original GWAS study, performed by Brčić et al. [23], showed an association of rs12944194, rs791903, and rs75201096 genetic variants with HT in Croatian population. The present study did not confirm the allelic association of these three SNPs and HT in our population. However, we observed that the rs791903 recessive homozygotes (CC) were significantly less frequent in HT patients than in controls (P = 0.025, OR = 0.476) (Table 4).

Genetic variant rs791903 is located in the intronic region of inositol hexaphosphate kinase 3 (IP6K3) gene on chromosome 6 (6p21). This locus has already been strongly associated with several inflammatory and autoimmune diseases [27,28,29].

A recent study observed association of rs791903 with rheumatoid arthritis (RA) [30], which is an autoimmune diseases that is also present in 4.25% of patients with HT [31]. It is suggested that HT and RA share genetic background and our study adds more evidence to genetic contribution of rs791903 in, both, RA and HT. Furthermore, it has been reported that IP6K3 gene is expressed in the thyroid tissue [32], but until now, its exact role remains unclear.

Several HT-related autoimmune diseases, such as systemic lupus erythematosus (SLE), GD, and RA, were already associated with genetic variants in ITPR3 gene that belongs to thyroid hormone synthesis pathway [33]. ITPR3 gene is located very near to our SNP rs791903 (38 kb away). According to Oishi et al. [33], the rs3748079, located in the promoter region of ITPR3 gene and 114 kb from rs791903 (D′ = 0.671, r2 = 0.1557), was strongly associated with SLE in Japanese population. Also, two other markers of the same gene, rs9394159 and rs4713693, were associated with GD [34]. Genetic studies of AITD showed the existence of a common genetic background between HT and GD [35,36,37]. Almost 74% of SNPs related to GD have the same effect on HT [23]; therefore, the genetic overlap between HT and GD, as supported by our study, gives additional evidence of the role of this SNP in AITD.

The most important limitation of our study is the sample size. Future studies of the genetic background of HT in our ethnic group should include larger sample sizes to confirm the results from this study. It is important to note that there is a difference in gender distribution between our case and control group: our case group consists of 96% women (as is expected for this female prevalent disease) [38, 39] whereas the control group consists of 64.3% of women. The difference in the gender distribution between cases and controls should not affect our results as the latest scientific reports encourage using mixed-sex controls (those consisted of both males and females) in studies of sex-specific traits (such as HT, which is predominantly affecting women). This largest genetic study including 61,094 women and 53,769 men showed that there are no allele frequency differences at common SNPs between men and women [40]. The advantages of our study include the stringent selection of HT patients, which are characterized by the presence of significantly higher levels of TPO Ab than controls (< 0.001). Our patients also had a higher BMI than controls, which is in line with the observations from most studies [41]. Another advantage is the careful selection of controls in clinical settings. We excluded individuals with any type of thyroid or any other autoimmune disorder from controls. Controls had TPO Ab levels in the reference ranges, thus minimizing potential bias (Table 1).

Finally, it is now a common practice to analyze genetic variants in different ethnic groups than originally reported, as we now know that differences in the strength and even the direction of genetic signals across different populations exist. We believe that results of our study add to the current scientific knowledge on genetic susceptibility to HT. However, in the present time, when genetic influence on HT is just starting to be deciphered, there is no translational benefit for study subjects. Therefore, it is important to continue these efforts and perform replication studies and additional functional studies using other high-throughput analytic methods, to understand the mechanisms by which IP6K3 gene exerts its roles in HT. The final aim in the future is to benefit patients by translating new knowledge on genetic factors (thus biological mechanisms) into clinical practice.

Conclusion

In conclusion, this is the first genetic study that was focused on understanding the genetic basis of HT in Algerian population. Our results suggest the lack of allelic association of rs12944194, rs791903, and rs75201096 with HT. However, we found differences in the genotype distribution of rs791903 between HT and the control group, where recessive homozygotes (CC) were significantly less frequent in HT patients.

Due to the increased prevalence of HT in the Arab-Berber ethnical group from Algeria, future large-scale case-control studies are necessary for replication of our results.