Polymorphisms in Autophagy-Related Gene IRGM Are Associated with Susceptibility to Autoimmune Thyroid Diseases

Background To date, studies have shown that polymorphisms in an autophagy-related gene, IRGM, are linked with different diseases, especially autoimmune diseases. The present study aimed to examine the roles of IRGM polymorphisms in autoimmune thyroid diseases (AITD). Methods Three polymorphisms in IRGM gene (rs10065172, rs4958847, and rs13361189) were genotyped in 1569 participants (488 with Graves' disease, 292 with Hashimoto's thyroiditis, and 789 healthy controls) using PCR-based ligase detection reaction method. Gene-disease associations were evaluated for the three SNPs. Results T allele of rs10065172, A allele of rs4958847, and C allele of rs13361189 were all higher in Graves' disease patients than controls, and the ORs were OR = 1.207 (P = 0.022), OR = 1.207 (P = 0.027), and OR = 1.200 (P = 0.027), respectively. After adjusting for sex and age, rs10065172 and rs13361189 were still associated with GD under both the allele model and dominant model, and the adjusted ORs for rs10065172 were 1.20 (P = 0.033) and 1.33 (P = 0.024), while the adjusted ORs for rs13361189 were 1.19 (P = 0.042) and 1.33 (P = 0.026), respectively. No significant difference was found between Hashimoto's thyroiditis patients and controls. Haplotype analysis found that CTA frequency was distinguishingly higher in Graves' disease patients (OR = 1.195, P = 0.030). The frequency of TCG haplotype was distinguishingly lower in AITD and Graves' disease patients (OR = 0.861, P = 0.044; OR = 0.816, P = 0.017). Conclusions Our study reveals IRGM as a susceptibility gene of AITD and Graves' disease for the first time.


Introduction
Autoimmune thyroid disease (AITD) is the most prevalent autoimmune disease and mainly has two diverse types, Graves' disease (GD) and Hashimoto's thyroiditis (HT). AITD has a huge preponderance of females [1]. AITDs are characterized by lymphocytic infiltration responsive to thyroid antigens including thyroid-stimulating hormone receptor (TSHR), thyroid antigens including thyroglobulin (Tg), and thyroid peroxidase (TPO) [2]. In GD, anti-TSHR autoantibodies can bind to TSHR and stimulate the production and secretion of thyroid hormones, resulting in hyperthyroidism [3]. HT is characterized by the presence of thyroid autoantibodies against TPO and Tg [4].
Although the etiology of AITD remains elusive, the current paradigm is that the interplay between gene and environmental factors causes AITD [5,6]. It is estimated that genetic factor accounts for about 70-80% of liability to the development of AITD [7]. The HLA-DR locus is the most important susceptibility gene involved in AITD [5]. Other identified susceptibility genes of AITD mainly have the following two groups: (1) immunityrelated genes: protein tyrosine phosphatase-22 (PTPN22), CD40, CTLA4, and so on; (2) thyroid related genes: 2 BioMed Research International thyroglobulin (TG) and thyrotropin receptor (TSHR) genes [5].
Autophagy is an "autodigestive" process, which plays an important role in the delivery of intracellular components for terminal degradation and recycling [8]. This process has been associated with diverse aspects of innate and adaptive immunity [9], and abnormalities in the autophagic pathway have been involved in numerous autoimmune diseases, such as rheumatoid arthritis, systemic lupus erythematosus (SLE), and multiple sclerosis [10]. Immunity-related guanosine triphosphatase is critically important in defending against pathogens through regulating the progress of autophagy [11]. In the human genome, only immunity-related guanosine triphosphatase family M (IRGM) encodes a functional IRG among the three IRG genes (IRGC, IRGQ, and IRGM) [12]. IRGM is a key autophagy-related molecule, and it is also an important regulator involved in inflammation, which can prevent unwanted inflammation and protect cells from oxidative stress by regulating the engulfment of apoptotic cells [13]. Also, it can regulate the levels of cytokines, such as IL-1 and interferon gamma (INF-), through affecting the signaling pathway of toll-like receptor (TLR) [14]. To date, several polymorphisms in IRGM gene are reported to be linked with different diseases, especially autoimmune diseases such as inflammatory bowel disease (IBD) and SLE [10]. However, it is unknown whether IRGM gene variations are associated with AITD. Therefore, we conducted a casecontrol study to clarify this important issue.

Patients and Controls.
In this study, a total of 780 patients with AITD (488 with GD and 292 with HT) were enrolled consecutively from the Outpatient Department of Endocrinology of Jinshan Hospital of Fudan University from November 1, 2012, to December 31, 2016, and 789 healthy controls were from the Health Check-Up Center of the same hospital during the same period. All subjects were south descendants of China. Individuals having other obvious chronic diseases were excluded from patient group. Individuals with thyroid disease or any other autoimmune disease were excluded from control group. Informed consent was obtained from all participants.
GD was diagnosed with both clinical findings of hyperthyroidism and laboratory data, such as decreased TSH value, elevated thyroid hormones or anti-thyroid-stimulating hormone receptor antibody (TRAb), anti-thyroid peroxide antibody (TPOAb), and anti-thyroglobulin antibody (TgAb). The diagnosis of HT was established by positive status of antithyroid peroxide antibody (TPOAb) or anti-thyroglobulin antibody (TgAb).

DNA Extraction.
We collected 1 ml of peripheral venous blood in a tube containing ethylene diamine tetraacetic acid (EDTA) from each participant. The genomic DNA from peripheral blood was isolated using RelaxGene Blood DNA System (Tiangen Biotech Company, Beijing, China) according to the manufacture's protocol. The purity and concentration of all DNA samples were then detected.

SNP Selection and
Genotyping. Three SNPs (rs10065172/rs4958847/rs13361189) of IRGM gene were selected from the Hapmap CHB data [15] with the following criteria: minor allele frequency (MAF) > 0.05, Hardy-Weinberg equilibrium (HWE) with > 0.001, and logarithm of odds (LOD) > 3.0. These three SNPs were genotyped by PCR-based ligase detection reaction (LDR) method used in our previous studies [16][17][18]. The PCR condition was initial denaturation at 95 ∘ C for 2 min, denaturation at 94 ∘ C for 30 s, annealing at 62 ∘ C for 90 s, extension at 72 ∘ C for 1 min with 40 cycles and final extension at 65 ∘ C for 10 min. Primer sequences of the targeted loci are as follows: The LDR was performed using denaturation at 95 ∘ C for 2 min, 40 cycles of annealing at 94 ∘ C for 15 s, and extension at 50 ∘ C for 25 s.

Statistical
Analysis. Data were expressed as mean ± SD. We used Chi-square ( 2) test to evaluate the differences in the frequencies of alleles, genotypes and haplotypes among different groups, and the odds ratios (OR) were also calculated out. HWE test, linkage disequilibrium (LD) test, and haplotype frequencies were analyzed using Haploview 4.2. Logistic regression analysis was also performed, and age and gender were used as confounding factors in the multivariate logistic regression analysis. SPSS (Version 17.0) was used for statistical analysis and value less than 0.05 suggested significant outcomes.    Table 2 showed the data of IRGM gene polymorphisms in AITD cases and controls. No distinguishing difference was found in both allele and genotype distributions of the three loci between cases and controls ( > 0.05).

Clinical
The distributions of genetic frequencies of IRGM polymorphisms in GD patients, HT patients, and controls were shown in Table 3. In GD patients, the genotype distributions of the rs10065172 and rs13361189 showed a marginally significant trend compared with those of the controls, and was 0.052 and 0.056, respectively. For rs4958847, no distinguishing difference in the genotype of IRGM polymorphisms was found between GD cases and controls. In GD patients, the minor T allele of rs10065172 was higher than controls (45.70% versus 41.06%, OR = 1.207, 95% CI = 1.027-1.418, = 0.022,). For rs4958847, the minor allele A frequency was significantly higher in GD groups than controls (65.57% versus 61.22%, OR = 1.207, 95% CI = 1.022-1.425, = 0.027). GD patients also had an increased frequency of allele C in rs13361189 compared with the controls (45.59% versus 41.13%, OR = 1.200, 95% CI = 1.021-1.409, = 0.027). As Table 4 showed, after adjusting for sex and age, rs10065172 and rs13361189 were still associated with GD under both the allele model and dominant model,   AITD, autoimmune thyroid disease; GD, Graves' disease; HT, Hashimoto's thyroiditis; OR, odds ratio; 95% CI, 95% confidence intervals. and the adjusted ORs for rs10065172 were 1.20 ( = 0.033) and 1.33 ( = 0.024), while the adjusted ORs for rs13361189 were 1.19 ( = 0.042) and 1.33 ( = 0.026), respectively. However, there was no distinguishing difference between HT patients and controls.

Genotype and Clinical Phenotype Correlations.
No correlation was found between the genotype or allele variants of the three tested SNPs and the clinical phenotypes, including family history, positive or negative antibodies (TGAb and TPOAb), thyroid size, and treatment outcomes ( > 0.05) (data not shown).

Discussion
IRGM plays an important role in autophagy [19]. IRGM gene can induce autophagy by encoding a GTP-binding protein [14]. IRGM inhibition can result in impaired autophagy [11]. IRGM genetic polymorphisms were confirmed to be related to many kinds of inflammatory and autoimmune diseases. Song et al. showed that IRGM rs10065172 was related to active tuberculosis in a Korean population [12]. Lu et al. reported that IRGM SNPs rs10065172 and rs13361189 were protective factors against latent TB progression in a Chinese population [20]. The polymorphism of IRGM rs13361189 was found to be related to leprosy by affecting the production of IL-4, IL-6, and INF- [21]. Our previous genetic studies also found that variants of inflammation and cytokine related genes were associated with AITD (Supplementary Table 1). Significant correlation between Crohn's disease (CD) and the C allele of rs13361189 (OR= 1.33) of the IRGM gene was noted in an Indian Population [22]. Besides, a recent study suggested IRGM polymorphisms were related to gastric cancer in Chinese [23]. In accordance with previous findings, IRGM rs13361189 and rs4958847 polymorphisms were confirmed to be crucial for CD susceptibility and phenotype modulation [24]. Interestingly, it was reported that the polymorphisms of IRGM gene were involved in susceptibility to SLE [25].
The present study is the first to investigate the relationship between the IRGM SNPs (rs10065172, rs4958847, and rs13361189) and susceptibility to AITD. Our allele and genotype results showed that T allele of rs10065172, A allele of rs4958847, and C allele of rs13361189 were higher in GD patients. After adjusting for sex and age, rs10065172 and rs13361189 were still associated with GD under both the allele model and dominant model, indicating that IRGM gene was associated with GD and might participate in the pathogenesis of GD and act as a risk factor. In GD, although the exact function of the IRGM SNPs is not fully elucidated, IRGM is possibly involved in the pathogenesis of GD because of its crucial role in autophagy, and autophagy has a vital function in innate and adaptive immunity. The polymorphism of IRGM gene may lead to the impaired autophagy, resulting in the disorder of immunity function, which needs to be explored by further studies. However, neither the allele nor genotype of rs10065172, rs4958847, and rs13361189 was related to HT, indicating that the three SNPs were not involved in the pathogenesis of HT. These different results of in the relationship of IRGM SNPs with GD and HT may be caused by their different etiology. Furthermore, haplotype analysis also suggested some roles of IRGM SNPs in AITD and GD.
In stratified analyses by clinical phenotype, no statistical significance of allele and genotype distributions was found between cases with each clinical phenotype and those without. The lack of significant findings may be because of the heterogeneity of environment or genetic factors in Chinese Han population and the relatively limited sample size caused by stratification. Further functional studies with large number of participants deserve continuing investigation.
In summary, our study is the first to suggest that IRGM rs10065172, rs4958847, and rs13361189 are related to GD in terms of both allele and haplotype frequencies, and TCG haplotype plays a protective role in AITD and GD. More studies with large number of participants are needed to further investigate the role of IRGM genetic polymorphisms in susceptibility to AITD.