Skip to main content
NCBI home page
The Journal of Clinical Endocrinology and Metabolism logoLink to The Journal of Clinical Endocrinology and Metabolism
. 2012 Nov 1;98(1):E144–E152. doi: 10.1210/jc.2012-2408

Fine Mapping of Loci Linked to Autoimmune Thyroid Disease Identifies Novel Susceptibility Genes

Yaron Tomer 1,✉, Alia Hasham 1, Terry F Davies 1, Mihaela Stefan 1, Erlinda Concepcion 1, Mehdi Keddache 1, David A Greenberg 1
PMCID: PMC3537111  PMID: 23118423

Abstract

Context:

Genetic factors play a major role in the etiology of autoimmune thyroid disease (AITD) including Graves' disease (GD) and Hashimoto's thyroiditis (HT). We have previously identified three loci on chromosomes 10q, 12q, and 14q that showed strong linkage with AITD, HT, and GD, respectively.

Objectives:

The objective of the study was to identify the AITD susceptibility genes at the 10q, 12q, and 14q loci.

Design and Participants:

Three hundred forty North American Caucasian AITD patients and 183 healthy controls were studied. The 10q, 12q, and 14q loci were fine mapped by genotyping densely spaced single-nucleotide polymorphisms (SNPs) using the Illumina GoldenGate genotyping platform. Case control association analyses were performed using the UNPHASED computer package. Associated SNPs were reanalyzed in a replication set consisting of 238 AITD patients and 276 controls.

Results:

Fine mapping of the AITD locus, 10q, showed replicated association of the AITD phenotype (both GD and HT) with SNP rs6479778. This SNP was located within the ARID5B gene recently reported to be associated with rheumatoid arthritis and GD in Japanese. Fine mapping of the GD locus, 14q, revealed replicated association of the GD phenotype with two markers, rs12147587 and rs2284720, located within the NRXN3 and TSHR genes, respectively.

Conclusions:

Fine mapping of three linked loci identified novel susceptibility genes for AITD. The discoveries of new AITD susceptibility genes will engender a new understanding of AITD etiology.

Autoimmune thyroid diseases (AITD), including Graves' disease (GD) and Hashimoto's thyroiditis (HT), are common autoimmune endocrine disorders. Clinical AITD affects 1% of the population in the United States, and subclinical AITD affects up to 5% of the population (reviewed in Ref. 1). GD is characterized by hyperthyroidism, whereas HT can manifest by clinical hypothyroidism or by the presence of thyroid autoantibodies only (2, 3). Despite their contrasting clinical presentations, GD and HT share similar pathogenetic mechanisms. In both diseases autoreactive T cells escape tolerance and infiltrate the thyroid gland, leading to apoptosis of thyroid cells and hypothyroidism in HT (3), and to generation of TSH receptor (TSHR)-stimulating antibodies, causing proliferation of thyroid cells and excess hormone production in GD (2).

GD and HT are complex diseases caused by interaction between genetic and environmental factors. Indeed, genetic factors play a major role in the etiology of AITD, and several AITD susceptibility genes have been identified (reviewed in Ref. 1). Previously we performed a whole-genome linkage study in a large cohort of multiplex, multigenerational families, leading to the identification of eight new AITD loci (4, 5): three loci (on 6p, 8q, and 10q) were linked with both GD and HT (AITD loci); three loci (on 7q, 14q, and 20q) were linked with GD; one locus on 12q was linked only with HT (4); and one locus on 2q was linked with thyroid antibody production (5). Moreover, at three loci AITD susceptibility genes were identified: thyroglobulin on 8q (6), CD40 on 20q (7), and CTLA-4 on 2q (8). In the current study, we focused on three of the previously identified AITD loci on 10q, 12q, and 14q in which the causative genes have not yet been identified. We report here the detailed fine mapping of these loci and identification of novel AITD susceptibility genes located in them.

Patients and Methods

Patients and controls

Cases and controls

The project was approved by the Mount Sinai School of Medicine Institutional Review Board. Our discovery set included 340 Caucasian AITD patients, 225 with GD [182 (80.9%) females and 43 (19.1%) males; average age 48.3 (10–80) yr] and 115 with HT [93 (80.9%) females, and 22 (19.1%) males, average age 46.7 (9–89)]. Our replication set included 238 Caucasian AITD patients, 73 with GD [59 (80.8%) females and 14 (19.2%) males; average age 50.4 (20–81) yr] and 165 with HT (153 (92.7%) females and 12 (7.3%) males; average age 55.1 (19–92) yr]. The discovery set was recruited by one group of investigators (led by Y.T.), and the replication set was recruited by a separate group of investigators (led by Dr. Peter K. Gregersen, the Feinstein Institute for Medical Research, North Shore-Long Island Jewish Health System, Manhasset, NY). Diagnosis of GD was based on the following: 1) documented clinical and biochemical hyperthyroidism requiring treatment with or without palpable goiter and 2) presence of TSHR antibodies (Abs) and/or diffuse thyroid scan. Both patients with and without Graves' ophthalmopathy were included in the study because the loci linked with GD (10q and 14q) showed linkage to GD with or without Graves' ophthalmopathy. HT was diagnosed by the following: 1) the presence of clinical and biochemical hypothyroidism requiring thyroid hormone replacement with or without goiter and 2) the presence of antithyroperoxidase Abs, with or without antithyroglobulin (Tg) Abs. Control discovery set [n = 183; 124 (67.8%) females and 59 (32.2%) males] and control replication set [n = 276; 209 (75.7%) females and 67 (24.3%) males] included Caucasian individuals with no personal or family history of thyroid disease.

AITD families

For the linkage analyses, we studied 102 families (540 individuals; for a full description of the families; see Ref. 4). All families enrolled in the study were multiplex for AITD (more than one affected) and/or multigenerational. Families were ascertained through a patient with AITD, who confirmed having at least one other first-degree relative with AITD. On the average our families had 5.3 members.

Genotyping

DNA purification

DNA was extracted from whole blood using the Puregene kit (Gentra Systems, Minneapolis, MN).

Methods for microsatellite genotyping and single-nucleotide polymorphism (SNP) genotyping are detailed in Supplemental Information 1, published on The Endocrine Society's Journals Online web site at http://jcem.endojournals.org.

Statistical analyses

Linkage analysis

Linkage analysis was performed using maximum likelihood-based [logarithm of odds (LOD) score] methods of linkage analysis. Multipoint linkage analysis was performed assuming heterogeneity (HLOD). The marker positions were based on the University of California, Santa Cruz, genome browser (http://genome.ucsc.edu/cgi-bin/hgGateway). HLOD scores were computed by the GeneHunter program (9) using all the markers at each locus. Multipoint linkage analysis yields the maximum marker information for the area of interest. Using GeneHunter, we set the inheritance parameters to the values that gave the maximum LOD scores in the two-point analyses of these loci (4). We assumed a population prevalence of 1% for GD and HT and adjusted the gene frequency accordingly, although the analysis is highly robust to the gene frequency assumption (10).

Case control association analyses

Case-control association analyses were performed using the UNPHASED computer package (http://unphased.sourceforge.net/). UNPHASED (11) is a suite of programs for association analysis of multilocus haplotypes from unphased genotype data. We used the Cocaphase program (within the UNPHASED package) for case-control association analyses performing the χ2 test on large numbers of SNPs simultaneously. The odds ratio (OR) was calculated by the method of Woolf (12). Linkage disequilibrium (LD) testing was performed using the Haploview program (http://www.broad.mit.edu/mpg/haploview/) (13). The number of SNPs tested at each locus is provided in Results. SNPs were chosen to cover the main LD blocks in each locus around coding regions. Note that because each of the three loci has been previously confirmed to harbor an AITD susceptibility gene, there was no need to correct for testing three loci. Within each locus, correction was done for the number of independent tests performed. Markers in LD are not independent of each other (i.e. their alleles are transmitted together as haplotypes). However, correction is appropriate for the number of LD blocks tested. Because at the three linked loci our SNPs covered 10 or fewer LD blocks we used a Bonferoni correction of 10 and considered a P < 5 × 10−3 as significant in the discovery set. No correction for multiple testing was performed in the replication set since only one SNP was tested in each associated peak.

Results

Fine mapping of loci, 10q, 12q, and 14q using linkage

The 10q, 12q (designated HT-2), and 14q (designated GD-1) loci have been previously shown to be linked with AITD, HT, and GD, respectively (4), based on the Lander and Kruglyak criteria (14); moreover, these linkage results have been replicated in an independent data set of multiplex-multigenerational families (4). Each of these three replicated loci spanned greater than 25 cm (∼25 Mb). These loci were first fine mapped by the method of multipoint linkage analysis using additional closely spaced microsatellite markers at these loci. The peak of linkage in the multipoint linkage analysis indicated the region most likely to harbor the susceptibility gene at each locus. Using this strategy, we narrowed the 12q locus to the 12 cm region between markers D12S88 and D12S346, thus approaching the practical resolution limit of linkage analysis. Similarly, the 10q locus was narrowed to an 8 cm interval between D10S1659 and D10S1743, and 14q was narrowed to a 9-cm interval between D14S251 and D14S1000. Linkage analysis can fine map a locus to only about 5- to 10-cm intervals (∼5–10 Mb). This is because as the linked interval is narrowed, all markers in the region will show linkage (15). To further map and identify the AITD susceptibility genes at these loci, we next performed SNP-based association studies.

Fine mapping of the 12q (HT-2), 14q (GD-1), and 10q loci using case-control association studies

Fine mapping the chromosome 12q (HT-2) locus

The 12q locus (designated HT-2) was linked with HT but showed evidence against linkage with GD (4). Linkage analysis narrowed the gene location to a 12-cm region (∼12 Mb). To further fine map this locus, we performed an association analysis in a dataset of 115 Caucasian HT patients, comparing them with 183 Caucasian controls. We used 241 SNPs spanning this locus. SNPs were selected to cover the HapMap LD blocks at this locus enriched for coding regions. Association analysis, using the program UNPHASED, revealed three peaks of association (Fig. 1). Marker rs1628799 showed the strongest statistical evidence of association in peak no. 1. The associated C allele was present in 15.3% of patients and 8.2% of controls (P = 9 × 10−3, OR 2.0); marker rs28399538 showed the strongest evidence of association in peak no. 2. The associated T allele was present in 5.9% of patients and in 0.9% of controls (P = 8 × 10−4, OR 6.9). This high OR may reflect inflation of the OR by the low minor allele frequency of the SNP in the controls. Peak no. 3 consisted of only one marker (rs17227744, A allele frequency 80.3 and 69% in patients and controls, respectively, P = 9 × 10−3, OR 1.8). The markers with the most significant P values in each peak are shown in Table 1. Using the Haploview program, we identified a 12-SNP haplotype in peak no. 1, spanning 298 kb, which was strongly associated with HT (P = 4 × 10−4, Supplemental Table A). In peak no. 2, we identified a five-SNP haplotype (92 kb), also showing a strong association (P = 8 × 10−4, Supplemental Table A). In contrast, peak no. 3 consisted of only one SNP (see above), suggesting that this region is a less optimal region to search further for an HT-related gene. Thus, either peak no. 1 or peak no. 2, or both, may point to the same HT susceptibility gene or to independent susceptibility alleles at the same locus. Because there was no LD between peak no. 1 and peak no. 2, it suggested that two independent HT susceptibility genes are located within this linked region. The SNP showing the strongest association in peak no. 1, rs1628799, was located within the MGAT4C (GnT-IV-H) gene, and the most strongly associated SNP in peak no. 2, rs28399538, was located in the BTG1 gene.

Fig. 1.

Fig. 1.

Open in a new tab

Fine mapping of the 12q (HT-2) locus. SNPs were genotyped using the Illumina GoldenGate custom made assays (San Diego, CA) and association analyses were performed using the program UNPHASED. Association analysis revealed three peaks of association. Peak no. 1 spanned 298 kb, and marker rs1628799 showed the strongest association in this peak; peak no. 2 spanned 92 kb, and marker rs28399538 showed the strongest association in this peak. Peak no. 3 consisted of only one marker, rs17227744. The LD map for each peak derived from the program Haploview is shown above each peak. Shown is the region surrounding the associated peaks.

Table 1.

SNPs showing the strongest associations in three loci linked with AITD

Locus SNP Allele Patients (%) Controls (%) P value (OR)
12q rs1628799 C 31 (15.3%) 30 (8.2%) 9 × 10−3 (2.0)
A 171 (84.7%) 334 (91.8%)
rs28399538 T 11 (5.9%) 3 (0.9%) 8 × 10−4 (6.9)
C 177 (94.1%) 331 (99.1%)
rs17227744 A 144 (80.4%) 165 (68.8%) 9 × 10−3 (1.8)
C 35 (19.6%) 75 (31.2%)
14q rs12147587 A 37 (8.9%) 3 (0.9%) 1.9 × 10−7 (10.3)
C 381 (91.1%) 319 (99.1%)
rs2284720 G 121 (29.1%) 62 (19.1%) 1.8 × 10−3 (1.7)
A 295 (70.9%) 262 (80.9%)
10q rs6479778 G 460 (86.8%) 291 (79.9%) 6 × 10−3 (1.7)
A 70 (13.2%) 73 (20.1%)
Open in a new tab

Fine mapping of the chromosome 14q (GD-1) locus

The 14q locus (designated GD-1) was linked with GD and showed evidence against linkage with HT (4). Linkage analysis narrowed GD-1 to a region of approximately 9 cm, which included the TSHR gene, known to be associated with GD (16, 17). We further fine mapped this region by association analysis using 163 SNP markers selected to cover the LD blocks in this locus surrounding coding regions. Association analysis revealed two association peaks spanning 207 and 128 kb, respectively (Fig. 2). The most strongly associated SNP in peak no. 1 was rs12147587. The associated A allele was present in 8.9% of patients and 0.9% of controls (P = 1.9 × 10−7, OR 10.3). This high OR may reflect inflation of the OR by the low minor allele frequency of the SNP in the controls. The most strongly associated SNP in peak no. 2 was rs2284720. The associated G allele was present in 29.1% of patients and 19.1% of controls (P = 1.8 × 10−3, OR 1.7). This SNP is located in intron 1 of the TSHR gene (20,973 bp, downstream exon 1). The most strongly associated markers in each peak are shown in Table 1. Haplotype analysis using Haploview identified an eight-SNP haplotype in peak no. 1, spanning 142 kb, which was strongly associated with GD (P = 8 × 10−4). In peak no. 2, we identified a five-SNP haplotype (128.5 kb), also showing a strong association with GD (P = 4 × 10−4, Supplemental Table A). Thus, either peak no. 1 or peak no. 2, or both, may point to the same GD susceptibility gene or to independent susceptibility alleles at the same locus. Because there was no LD between peak no. 1 and peak no. 2, it suggested that two independent GD susceptibility genes are located within this linked region. The most strongly associated SNP in peak no. 1, rs12147587, was located within the neurexin 3 (NRXN3) gene, highly expressed in neuronal tissues and believed to be a synaptic adhesion molecule (18). However, it has also been shown to be involved in adhesion of other cell types (19).

Fig. 2.

Fig. 2.

Open in a new tab

Fine mapping of the 14q (GD-1) locus using association analysis. SNPs were genotyped using the Illumina GoldenGate custom-made assays, and association analyses were performed using the program UNPHASED. Association analysis revealed two association peaks spanning 207 and 128 kb, respectively. The most strongly associated SNP in peak no. 1 was rs12147587. The most strongly associated SNP in peak no. 2 was rs2284720. The LD map for each peak is shown above each peak. Shown is the region surrounding the associated peaks.

The most strongly associated SNP in peak no. 2, rs2284720, was located within intron 1 of the TSHR gene, 20 kb downstream of exon 1. This SNP is close to several TSHR intron 1 SNPs that were previously reported to be associated with GD (16, 17, 20).

Fine mapping of the chromosome 10q locus

The 10q locus was linked with both GD and HT and thus may harbor a general thyroid autoimmunity gene (4). Linkage analysis narrowed it to a region of approximately 8 Mb. Our fine-mapping association analysis for this locus was done in a data set of 340 Caucasian AITD patients. We used 181 SNPs spanning this locus enriching for coding regions. Association analysis revealed one peak of association (Fig. 3). Marker rs6479778 showed the strongest association in this peak. The associated G allele was present in 86.8% of patients and 79.7% of controls (P = 6 × 10−3, OR 1.7, Table 1). Using the Haploview program, we identified a five-SNP haplotype in the peak (194 kb) that was strongly associated with AITD (P = 1.7 × 10−3, Supplemental Table A). The SNP showing the strongest association in the peak, rs6479778, was located within the ARID5B gene that encodes a member of the AT-rich interaction domain family of DNA binding proteins (21).

Fig. 3.

Fig. 3.

Open in a new tab

Fine mapping of the 10q locus. SNPs were genotyped using the Illumina GoldenGate custom-made assays, and association analyses were performed using the program UNPHASED. Association analysis revealed one peak of association spanning 194 kb. Marker rs6479778 showed the strongest association in this peak. The LD map for the peak is shown above it. Shown is the region surrounding the associated peak.

Replication studies

To confirm the associations identified at the three linked loci, we performed a replication study using another data set of 238 Caucasian AITD patients, 73 with GD and 165 with HT and a control replication set (n = 276). Analysis of SNP rs6479778, the most strongly associated SNP at the 10q locus (ARID5B) in the discovery set, continued to show significant association with AITD in the replication set (P = 0.033, OR 0.7; Table 2). Interestingly, in the replication set, the associated allele was the A allele, whereas in the discovery set, it was the G allele. This phenomenon called Flip Flop has been shown to reflect the effects of other interacting genes (22), as we have previously shown for the CTLA-4 gene (23).

Table 2.

SNPs that showed significant associations in the discovery set were reanalyzed in the replication set

Locus (Gene) SNP Allele Patients (%) Controls (%) P value (OR)
12q (MGAT4C) rs1628799 C 44 (13.4%) 66 (12.0%) NS
A 284 (86.6%) 486 (88.0%)
12q (BTG1) rs28399538 C 311 (97.8%) 527 (95.5%) 0.079 (2.1)
T 7 (2.2%) 25 (4.5%)
14q (NRXN3) rs12147587 A 11 (7.6%) 19 (3.4%) 0.027 (2.3)
C 133 (92.4%) 533 (96.6%)
14q (TSHR) rs2284720 G 46 (31.5%) 126 (22.8%) 0.03 (1.6)
A 100 (68.5%) 426 (77.2%)
10q (ARID5B) rs6479778 G 317 (80.5%) 473 (85.7%) 0.033 (0.7)
A 77 (19.5%) 79 (14.3%)
Open in a new tab

NS, Not significant.

Another possible explanation is that the difference in associated alleles between the discovery and replication sets may have been caused by the fact that the percentage of HT and GD patients was significantly different between the discovery and replication sets. Analysis of SNP rs1628799, the most strongly associated SNP at peak no. 1 of the 12q locus (MGAT4C), showed no evidence of association with HT in the replication data set (Table 2). Thus, the association with the MGAT4C gene was not replicated. Analysis of the rs28399538 SNP, the most strongly associated SNP at peak no. 2 of the 12q locus (BTG1), showed a borderline P value (P = 0.079, OR 2.1, Table 2) not confirming this association. Analysis of SNP rs12147587, the most strongly associated SNP in the peak no. 1 of the 14q locus (NRXN3), continued to show significant association with GD in the replication set (P = 0.027, OR 2.3, Table 2). SNP rs2284720, the most strongly associated SNP in peak no. 2 of GD-1 (TSHR), also continued to show significant association with GD in the replication set (P = 0.03, OR 1.6, Table 2).

Discussion

The etiology of AITD is complex involving multiple genes, environmental factors (24, 25), and epigenetic effects (26). Moreover, the genetic relationship between GD and HT is not fully understood and even within GD and HT, there exists significant phenotypic heterogeneity (27); therefore, in this study we analyzed only the subset of HT patients that were clinically hypothyroid. Previously we (4) and others (28, 29) have identified several loci as potential genetic regions harboring AITD susceptibility genes (reviewed in Ref. 1). In some of these loci, the AITD susceptibility genes have been identified [e.g. CD40 (7) and Tg (6, 30)]. Moreover, functional studies unraveled the mechanisms by which Tg and CD40 variants predispose to disease. In the case of the Tg gene, a specific Tg promoter variant was shown to interact epigenetically with interferon-α (which is produced during infections) to up-regulate Tg gene expression (26). This provided an attractive potential mechanism for triggering AITD by infection. Functional studies on the CD40 gene demonstrated a Kozak SNP that up-regulated CD40 gene expression in the thyroid, thereby contributing to the thyroid's autoimmune response (31).

In several confirmed loci the genes predisposing to AITD have not been identified yet (4). In this manuscript we report fine mapping of three confirmed loci on chromosomes 10q, 12q (HT-2), and 14q (GD-1). Our analyses identified two new putative AITD susceptibility genes: ARID5B at the 10q locus (associated with both GD and HT) and neurexin 3 (NRXN3) at the 14q locus (associated with GD). Intriguingly, at the 14q locus, we identified two independently associated peaks, at the NRXN3 and TSHR genes. NRXN3 and TSHR are separated by approximately 2.5 Mb (Fig. 2) and are not in LD with each other, suggesting independent genetic influences, but we have not ruled out that the signals represent two different disease-related alleles at some locus.

Examining the function of these candidate AITD susceptibility genes may suggest new mechanistic pathways involved in the pathogenesis of AITD. The AT-rich interaction domain 5B (ARID5B) gene is a transcription factor that binds to the AATA(C/T) sequence to activates target genes (21). ARID5B is believed to induce cell differentiation in adipose tissues (32). Intriguingly, a recent meta-analysis of genome-wide association studies reported a significant association of the ARID5B gene with rheumatoid arthritis as well as with Graves' disease in Japanese patients (33). These data are consistent with our results showing association of the ARID5B gene with AITD. The association with three autoimmune diseases (rheumatoid arthritis, GD, and HT) suggests that ARID5B may be a general autoimmunity gene. Functional studies are needed to better characterize the function of ARID5B and how polymorphisms in it contribute to autoimmunity.

NRXN3 (located at 14q) is a neuronal cell surface protein that is involved in cell recognition and cell adhesion. NRXN3 was shown to be associated with obesity (34), and this may provide a mechanism by which a chronic inflammatory state, possibly induced by leptin, could contribute to thyroid autoimmunity. Interestingly, both ARID5B and NRXN3 are genes that regulate adiposity, and both were found to play a role in thyroid autoimmunity. Using quantitative PCR, we have shown expression of NRXN3 in the thyroid (data not shown), suggesting a role in thyroid physiology. Further detailed functional description of these candidate genes is provided in Supplemental Table 2.

The hallmark of GD is the presence of stimulating TSHR antibodies (2). Not surprisingly, the TSHR was the first non-MHC gene to be tested for association with GD. Recently it was found that SNPs in intron 1 of the TSHR gene are associated with GD (16, 35). The SNP we identified to be associated with GD, rs2284720, is located between the two most strongly associated SNPs reported by Brand and Gough (35) (rs179247 and rs12101255) (Fig. 4). All these SNPs are in LD, and it is unclear which one is the causative SNP. Mechanistically, SNPs in intron 1 of the TSHR can alter its splicing (36, 37).

Fig. 4.

Fig. 4.

Open in a new tab

Map of intron 1 of the TSHR gene showing four SNPs that are associated with GD.

At least two possible mechanisms can explain the contrasting clinical features in GD and HT despite the similar immunogenetic mechanisms. One possibility is that GD and HT share common genetic susceptibility, whereas differing environmental and epigenetic factors determine the final phenotype (38). Alternatively, GD and HT may be influenced by distinct genes. Recently it is becoming clear that GD and HT are associated with distinct susceptibility genes. However, these mechanisms are not mutually exclusive, and environmental and epigenetic factors likely also play a role in determining the AITD phenotype. Indeed, previous studies and the current study identified both common genes for GD and AITD (HLA-DR, thyroglobulin, CTLA-4, PTPN22, ARID5B) as well as unique genes for GD (TSHR, CD40, CD25, NRXN3) and HT (BTG1).

Because the AITD genes identified so far do not explain the majority of the genetic liability to AITD (39), additional genes must contribute to the etiology of AITD. Recently it has been suggested that to identify the additional genetic contribution to complex diseases such as AITD, we must look for rare variants that cosegregate with disease in families, but are rare in the general population (40). The best way to identify rare variants causing AITD will be through next-generation sequencing in families in which AITD cluster, such as our own cohort of multiplex families, because in these families rare variants will be found frequently in the affected members (40). Moreover, targeting regions of linkage will most likely be the best approach.

In conclusion, previous studies have demonstrated the power of investigations to discover AITD susceptibility genes. The discoveries of new AITD susceptibility genes engendered new understanding of AITD etiology. Here we report the fine identification of three new putative AITD genes in linked loci. Although the exact mechanisms by which ARID5B, NRXN3, and BTG1 predispose to AITD remain to be discovered, it seems plausible that they modulate immune responses.

Acknowledgments

We thank all the AITD patients and families who graciously agreed to participate in the study. We also thank Dr. Peter Gregersen, who gave us access to the Multiple Autoimmune Disease Genetics Consortium (MADGC) data set for our replication sets.

This work was supported in part by Grants DK061659, DK067555, and DK073681 from the National Institute of Diabetes and Digestive and Kidney Diseases (to Y.T.); Grants DK069713, DK052464, and DK080459 from the National Institute of Diabetes and Digestive and Kidney Diseases (to T.F.D.); Grants NS27941, NA61829, and MH48858 (to D.A.G.); MADGC grant National Institutes of Health/National Institute of Allergy and Infectious Diseases Grant RO1-AI-68759 (to Peter Gregersen); and the Veterans Affairs Merit Award Program (to Y.T. and T.F.D.).

Disclosure Summary: T.F.D. is a member of the Board of Kronus Inc., distributors of thyroid diagnostics. All other authors have nothing to disclose.

Footnotes

Abbreviations:
Ab
Antibody
AITD
autoimmune thyroid diseases
GD
Graves' disease
HT
Hashimoto's thyroiditis
LD
linkage disequilibrium
LOD
logarithm of odds
OR
odds ratio
SNP
single-nucleotide polymorphism
Tg
thyroglobulin
TSHR
TSH receptor.

References

  • 1. Huber A, Menconi F, Corathers S, Jacobson EM, Tomer Y. 2008. Joint genetic susceptibility to type 1 diabetes and autoimmune thyroiditis: from epidemiology to mechanisms. Endocr Rev 29:697–725 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Menconi F, Oppenheim YL, Tomer Y. 2008. Graves' disease. In: Shoenfeld Y, Cervera R, Gershwin ME, eds. Diagnostic criteria in autoimmune diseases. Totowa, NJ: Humana Press; 231–235 [Google Scholar]
  • 3. Weetman AP. 2000. Chronic autoimmune thyroiditis. In: Braverman LE, Utiger RD, eds. Werner and Ingbar's the thyroid. Philadelphia: Lippincott Williams and Wilkins; 721–732 [Google Scholar]
  • 4. Tomer Y, Ban Y, Concepcion E, Barbesino G, Villanueva R, Greenberg DA, Davies TF. 2003. Common and unique susceptibility loci in Graves and Hashimoto diseases: results of whole-genome screening in a data set of 102 multiplex families. Am J Hum Genet 73:736–747 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Ban Y, Greenberg DA, Davies TF, Jacobson E, Concepcion E, Tomer Y. 2008. Linkage analysis of thyroid antibody production: evidence for shared susceptibility to clinical autoimmune thyroid disease. J Clin Endocrinol Metab 93:3589–3596 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Ban Y, Greenberg DA, Concepcion E, Skrabanek L, Villanueva R, Tomer Y. 2003. Amino acid substitutions in the thyroglobulin gene are associated with susceptibility to human and murine autoimmune thyroid disease. Proc Natl Acad Sci USA 100:15119–15124 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Jacobson EM, Concepcion E, Oashi T, Tomer Y. 2005. A Graves' disease-associated Kozak sequence single-nucleotide polymorphism enhances the efficiency of CD40 gene translation: a case for translational pathophysiology. Endocrinology 146:2684–2691 [DOI] [PubMed] [Google Scholar]
  • 8. Ueda H, Howson JM, Esposito L, Heward J, Snook H, Chamberlain G, Rainbow DB, Hunter KM, Smith AN, Di Genova G, Herr MH, Dahlman I, Payne F, Smyth D, Lowe C, Twells RC, Howlett S, Healy B, Nutland S, Rance HE, Everett V, Smink LJ, Lam AC, Cordell HJ, Walker NM, et al. 2003. Association of the T-cell regulatory gene CTLA4 with susceptibility to autoimmune disease. Nature 423:506–511 [DOI] [PubMed] [Google Scholar]
  • 9. Kruglyak L, Daly MJ, Reeve-Daly MP, Lander ES. 1996. Parametric and nonparametric linkage analysis: a unified multipoint approach. Am J Hum Genet 58:1347–1363 [PMC free article] [PubMed] [Google Scholar]
  • 10. Pal DK, Durner M, Greenberg DA. 2001. Effect of misspecification of gene frequency on the two-point LOD score. Eur J Hum Genet 9:855–859 [DOI] [PubMed] [Google Scholar]
  • 11. Dudbridge F. 2008. Likelihood-based association analysis for nuclear families and unrelated subjects with missing genotype data. Hum Hered 66:87–98 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12. Woolf B. 1955. On estimating the relation between blood group and disease. Ann Hum Genet 19:251–253 [DOI] [PubMed] [Google Scholar]
  • 13. Barrett JC, Fry B, Maller J, Daly MJ. 2005. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics 21:263–265 [DOI] [PubMed] [Google Scholar]
  • 14. Lander E, Kruglyak L. 1995. Genetic dissection of complex traits: guidelines for interpreting and reporting linkage results. Nat Genet 11:241–247 [DOI] [PubMed] [Google Scholar]
  • 15. Greenberg DA. 1993. Linkage analysis of “necessary” loci versus “susceptibility” loci. Am J Hum Genet 52:135–143 [PMC free article] [PubMed] [Google Scholar]
  • 16. Dechairo BM, Zabaneh D, Collins J, Brand O, Dawson GJ, Green AP, Mackay I, Franklyn JA, Connell JM, Wass JA, Wiersinga WM, Hegedus L, Brix T, Robinson BG, Hunt PJ, Weetman AP, Carey AH, Gough SC. 2005. Association of the TSHR gene with Graves' disease: the first disease specific locus. Eur J Hum Genet 13:1223–1230 [DOI] [PubMed] [Google Scholar]
  • 17. Yin X, Latif R, Bahn R, Tomer Y, Davies TF. 2008. Influence of the TSH receptor gene on susceptibility to Graves' disease and Graves' ophthalmopathy. Thyroid 18:1201–1206 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18. Graf ER, Kang Y, Hauner AM, Craig AM. 2006. Structure function and splice site analysis of the synaptogenic activity of the neurexin-1β LNS domain. J Neurosci 26:4256–4265 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Occhi G, Rampazzo A, Beffagna G, Antonio Danielli G. 2002. Identification and characterization of heart-specific splicing of human neurexin 3 mRNA (NRXN3). Biochem Biophys Res Commun 298:151–155 [DOI] [PubMed] [Google Scholar]
  • 20. Chu X, Pan CM, Zhao SX, Liang J, Gao GQ, Zhang XM, Yuan GY, Li CG, Xue LQ, Shen M, Liu W, Xie F, Yang SY, Wang HF, Shi JY, Sun WW, Du WH, Zuo CL, Shi JX, Liu BL, Guo CC, Zhan M, Gu ZH, Zhang XN, Sun F, Wang ZQ, Song ZY, Zou CY, Sun WH, Guo T, Cao HM, Ma JH, Han B, Li P, Jiang H, Huang QH, Liang L, Liu LB, Chen G, Su Q, Peng YD, Zhao JJ, Ning G, Chen Z, Chen JL, Chen SJ, Huang W, Song HD. 2011. A genome-wide association study identifies two new risk loci for Graves' disease. Nat Genet 43:897–901 [DOI] [PubMed] [Google Scholar]
  • 21. Baba A, Ohtake F, Okuno Y, Yokota K, Okada M, Imai Y, Ni M, Meyer CA, Igarashi K, Kanno J, Brown M, Kato S. 2011. PKA-dependent regulation of the histone lysine demethylase complex PHF2-ARID5B. Nat Cell Biol 13:668–675 [DOI] [PubMed] [Google Scholar]
  • 22. Lin PI, Vance JM, Pericak-Vance MA, Martin ER. 2007. No gene is an island: the flip-flop phenomenon. Am J Hum Genet 80:531–538 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23. Vieland VJ, Huang Y, Bartlett C, Davies TF, Tomer Y. 2008. A multilocus model of the genetic architecture of autoimmune thyroid disorder, with clinical implications. Am J Hum Genet 82:1349–1356 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Akeno N, Smith EP, Stefan M, Huber AK, Zhang W, Keddache M, Tomer Y. 2011. Interferon-α mediates the development of autoimmunity both by direct tissue toxicity and through immune-cell recruitment mechanisms. J Immunol 186:4693–4706 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Tomer Y, Huber A. 2009. The etiology of autoimmune thyroid disease: a story of genes and environment. J Autoimmun 32:231–239 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26. Stefan M, Jacobson EM, Huber AK, Greenberg DA, Li CW, Skrabanek L, Conception E, Fadlalla M, Ho K, Tomer Y. 2011. Novel variant of thyroglobulin promoter triggers thyroid autoimmunity through an epigenetic interferon α-modulated mechanism. J Biol Chem 286:31168–31179 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27. Tomer Y, Menconi F, Davies TF, Barbesino G, Rocchi R, Pinchera A, Concepcion E, Greenberg DA. 2007. Dissecting genetic heterogeneity in autoimmune thyroid diseases by subset analysis. J Autoimmun 29:69–77 [DOI] [PubMed] [Google Scholar]
  • 28. Sakai K, Shirasawa S, Ishikawa N, Ito K, Tamai H, Kuma K, Akamizu T, Tanimura M, Furugaki K, Yamamoto K, Sasazuki T. 2001. Identification of susceptibility loci for autoimmune thyroid disease to 5q31-q33 and Hashimoto's thyroiditis to 8q23-q24 by multipoint affected sib-pair linkage analysis in Japanese. Hum Mol Genet 10:1379–1386 [DOI] [PubMed] [Google Scholar]
  • 29. Allen EM, Hsueh WC, Sabra MM, Pollin TI, Ladenson PW, Silver KD, Mitchell BD, Shuldiner AR. 2003. A genome-wide scan for autoimmune thyroiditis in the Old Order Amish: replication of genetic linkage on chromosome 5q11.2-q14.3. J Clin Endocrinol Metab 88:1292–1296 [DOI] [PubMed] [Google Scholar]
  • 30. Hsiao JY, Hsieh MC, Tien KJ, Hsu SC, Shin SJ, Lin SR. 2007. Association between a C/T polymorphism in exon 33 of the thyroglobulin gene is associated with relapse of Graves' hyperthyroidism after antithyroid withdrawal in Taiwanese. J Clin Endocrinol Metab 92:3197–3201 [DOI] [PubMed] [Google Scholar]
  • 31. Huber AK, Finkelman FD, Li CW, Concepcion E, Smith E, Jacobson E, Latif R, Keddache M, Zhang W, Tomer Y. 2012. Genetically driven target tissue overexpression of CD40: a novel mechanism in autoimmune disease. J Immunol 189:3043–3053 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32. Whitson RH, Tsark W, Huang TH, Itakura K. 2003. Neonatal mortality and leanness in mice lacking the ARID transcription factor Mrf-2. Biochem Biophys Res Commun 312:997–1004 [DOI] [PubMed] [Google Scholar]
  • 33. Okada Y, Terao C, Ikari K, Kochi Y, Ohmura K, Suzuki A, Kawaguchi T, Stahl EA, Kurreeman FA, Nishida N, Ohmiya H, Myouzen K, Takahashi M, Sawada T, Nishioka Y, Yukioka M, Matsubara T, Wakitani S, Teshima R, Tohma S, Takasugi K, Shimada K, Murasawa A, Honjo S, Matsuo K, Tanaka H, Tajima K, Suzuki T, Iwamoto T, Kawamura Y, Tanii H, Okazaki Y, Sasaki T, Gregersen PK, Padyukov L, Worthington J, Siminovitch KA, Lathrop M, Taniguchi A, Takahashi A, Tokunaga K, Kubo M, Nakamura Y, Kamatani N, Mimori T, Plenge RM, Yamanaka H, Momohara S, Yamada R, Matsuda F, Yamamoto K. 2012. Meta-analysis identifies nine new loci associated with rheumatoid arthritis in the Japanese population. Nat Genet 44:511–516 [DOI] [PubMed] [Google Scholar]
  • 34. Heard-Costa NL, Zillikens MC, Monda KL, Johansson A, Harris TB, Fu M, Haritunians T, Feitosa MF, Aspelund T, Eiriksdottir G, Garcia M, Launer LJ, Smith AV, Mitchell BD, McArdle PF, Shuldiner AR, Bielinski SJ, Boerwinkle E, Brancati F, Demerath EW, Pankow JS, Arnold AM, Chen YD, Glazer NL, McKnight B, et al. 2009. NRXN3 is a novel locus for waist circumference: a genome-wide association study from the CHARGE Consortium. PLoS Genet 5:e1000539. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35. Brand OJ, Gough SC. 2010. Genetics of thyroid autoimmunity and the role of the TSHR. Mol Cell Endocrinol 322:135–143 [DOI] [PubMed] [Google Scholar]
  • 36. Graves PN, Tomer Y, Davies TF. 1992. Cloning and sequencing of a 1.3 kb variant of human thyrotropin receptor mRNA lacking the transmembrane domain. Biochem Biophys Res Commun 187:1135–1143 [DOI] [PubMed] [Google Scholar]
  • 37. Kakinuma A, Nagayama Y. 2002. Multiple messenger ribonucleic acid transcripts and revised gene organization of the human TSH receptor. Endocr J 49:175–180 [DOI] [PubMed] [Google Scholar]
  • 38. Phillips D, McLachlan S, Stephenson A, Roberts D, Moffitt S, McDonald D, Ad'Hiah A, Stratton A, Young E, Clark F. 1990. Autosomal dominant transmission of autoantibodies to thyroglobulin and thyroid peroxidase. J Clin Endocrinol Metab 70:742–746 [DOI] [PubMed] [Google Scholar]
  • 39. Brix TH, Kyvik KO, Christensen K, Hegedüs L. 2001. Evidence for a major role of heredity in Graves' disease: a population-based study of two Danish twin cohorts. J Clin Endocrinol Metab 86:930–934 [DOI] [PubMed] [Google Scholar]
  • 40. Vieland VJ, Devoto M. 2011. Next-generation linkage analysis. Hum Hered 72:227. [DOI] [PubMed] [Google Scholar]

Articles from The Journal of Clinical Endocrinology and Metabolism are provided here courtesy of The Endocrine Society

RESOURCES