Abstract
There is strong genetic association between type 1A diabetes (T1D) and autoimmune thyroid disease (AITD). T1D and AITD frequently occur together in the same individual, a condition classified as a variant of the autoimmune polyglandular syndrome type 3 (APS3). Because T1D and AITD are individually strongly associated with different HLA class II sequences, we asked which HLA class II pocket sequence and structure confer joint susceptibility to both T1D and AITD in the same individual (APS3v). We sequenced the HLA-DR gene in 105 APS3v patients and 153 controls, and identified a pocket amino acid signature, DRβ-Tyr-26, DRβ-Leu-67, DRβ-Lys-71, and DRβ-Arg-74, that was strongly associated with APS3v (P = 5.4 × 10−14, odds ratio = 8.38). Logistic regression analysis demonstrated that DRβ-Leu-67 (P = 9.4 × 10−13) and DRβ-Arg-74 (P = 1.21 × 10−13) gave strong independent effects on disease susceptibility. Structural modeling studies demonstrated that pocket 4 was critical for the development of T1D+AITD; all disease-associated amino acids were linked to areas of the pocket that interact directly with the peptide and, therefore, influence peptide binding. The disease-susceptible HLA-DR pocket was more positively charged (Lys-71, Arg-74) compared with the protective pocket (Ala-71, Gln-74). We conclude that a specific pocket amino acid signature confers joint susceptibility to T1D+AITD in the same individual by causing significant structural changes in the MHC II peptide binding pocket and influencing peptide binding and presentation. Moreover, Arg-74 is a major amino acid position for the development of several autoimmune diseases. These findings suggest that blocking the critical Arg-74 pocket might offer a method for treating certain autoimmune conditions.
Keywords: autoimmunity, gene, structure
Type 1A (autoimmune) diabetes mellitus (T1D) and autoimmune thyroid disease (AITD) are the most common autoimmune endocrine disorders (1–3). Despite affecting different tissues, T1D and AITD share a common etiology. Both are organ-specific autoimmune diseases characterized by infiltration of the gland by autoreactive T and B cells and production of antibodies directed at the target organs (the pancreatic islets in T1D and the thyroid in AITD), resulting in their dysfunction or destruction. In fact, there is a well-known strong association between T1D and AITD (reviewed in ref. 1). They frequently occur within the same family (aggregation) and in the same individual (comorbidity) (1). When T1D and AITD occur in the same individual the phenotype is considered, according to a commonly accepted classification, as a variant of the autoimmune polyglandular syndrome type 3 (APS3) (4, 5) [in this manuscript, we use the nomenclature which defines APS3 as AITD + another autoimmune condition excluding Addison's disease (4) for better clarity, and we refer to the combination AITD+T1D in the same individual as APS3 variant (APS3v) to emphasize that this is not the only form of APS3]. However, it should be noted that the classification of APS is not uniform, and some scholars refer to all non-Mendelian forms of APS as APS2 (5, 6).
Epidemiological data show that 15–30% of patients with T1D have thyroid antibodies, and up to 50% of such patients progress to clinical AITD (7). Conversely, 2.3% of children with AITD have islet cell antibodies compared with 0% of controls (8). Although the exact mechanisms of this association are still evolving, there is growing evidence that genetic factors play a major role (9). However, so far few studies have examined the joint genetic etiology of T1D+AITD. Three loci have been shown to confer susceptibility to T1D+AITD in the same individual (APS3v): HLA-DR (10–12), CTLA-4 (11, 13), and PTPN22 (14).
To dissect the joint genetic etiology of T1D and AITD, we have previously performed a whole-genome linkage study in a unique cohort of families in which both T1D and AITD clustered (15). We identified three loci, HLA class II, CTLA-4, and FOXP3, that showed evidence for linkage when APS3v individuals (T1D+AITD) were classified as affected (15). Not unexpectedly, the locus showing the strongest linkage with APS3v was the HLA class II locus (15). These data are supported by other studies (10, 12, 16). Detailed analysis of the HLA class II locus demonstrated that the major HLA haplotype contributing to the shared susceptibility to T1D and AITD was DR3-DQB1*0201, with DR3 conferring most of the shared risk (11, 17). However, the specific HLA-DR3 sequence that confers joint susceptibility to T1D+AITD (comorbidity scenario, APS3v) and the mechanisms of the association are not known.
HLA class II molecules are heterodimeric molecules consisting of an α and a β chain. These two chains associate together to create a peptide binding pocket which is highly polymorphic (18). In several autoimmune diseases, notably type 1A diabetes (T1D) (19), there is persuasive evidence that the disease is associated with specific pocket amino acid sequences within the MHC class II genes. Moreover, we have recently identified specific HLA-DR pocket variants that were critical for the development of Graves’ disease (GD) (20) and Hashimoto's thyroiditis (HT) (21). The critical common amino acid for the development of both GD and HT was arginine at position 74 of the DRβ1 chain (20, 21). In view of these findings, we hypothesized that specific HLA-DR pocket amino acids are critical for the development of APS3v (T1D+AITD), in a similar manner to that shown for other autoimmune diseases (1). Therefore, the aims of our study were to determine the HLA-DR types, amino acid sequences, and pocket 3D structures in a cohort of well-characterized APS3v (T1D+AITD) patients (the comorbidity scenario) and compare them with the HLA-DR sequences in matched controls. Our results demonstrated that specific amino acid signatures within the HLA-DR peptide binding pockets conferred strong risk for APS3v, with leucine at position 67 and arginine at position 74 of the HLA-DRβ1 chain contributing most to disease risk. These signatures created a unique structure of the HLA-DR peptide binding pocket which likely influences peptide binding.
Results
Characteristics of the Dataset.
We analyzed 105 APS3v (T1D+AITD) patients (76 females and 29 males). Most of the patients had HT [93/105 (88.6%)], while 12/105 (11.4%) had GD. Sixty-eight patients (64.8%) had a documented family history of AITD, 16 (15.2%) confirmed a negative family history of AITD, and in 21 patients (20%), family history could not be confirmed. Moreover, 19 patients (18.1%) had a positive family history for other autoimmune diseases, 57 patients (54.3%) had a negative history for other autoimmune diseases, and in 29 patients (27.6%), the history was undetermined.
HLA-DR Typing.
Typing was performed in 105 patients and 153 controls. We found a significantly higher frequency of DR3 and DR4 in APS3v patients compared with controls. The DR3 allele was present in 33.3% of patients compared with 7.2% of controls [P = 2.49 × 10−14, odds ratio (OR) = 6.45]. DR4 was positive in 39.5% of patients and in 14.1% of controls (P = 3.6 × 10−11, OR = 4). Conversely, DR11, DR2, DR13, and DR1 alleles were significantly more expressed in the control group, suggesting a protective role for the development of APS3v. DR11 was positive in 2.4% of patients and 16.7% of controls (P = 3 × 10−7, OR = 0.12), DR2 in 2.4% of patients and 11.1% of controls (P = 2.2 × 10−4, OR = 0.2), DR13 in 4.8% of patients and 12.1% of controls (P = 4.4 × 10−3, OR = 0.36), and DR1 in 6.2% of patients and 14.4% of controls (P = 3.5 × 10−3, OR = 0.39; Table 1).
Table 1.
HLA-DR typing in APS3v patients and controls
| HLA-DR | APS3v (%) n = 210* | Controls (%) n = 306* | P value |
| 1 | 13 (6.2) | 44 (14.4) | 3.5 × 10−3 |
| 2 | 5 (2.4) | 34 (11.1) | 2.2 × 10−4 |
| 3 | 70 (33.3) | 22 (7.2) | 2.49 × 10−14 |
| 4 | 83 (39.5) | 43 (14.1) | 3.6 × 10−11 |
| 7 | 11 (5.2) | 28 (9.2) | NS |
| 8 | 6 (2.9) | 17 (5.6) | NS |
| 9 | 4 (1.9) | 4 (1.3) | NS |
| 10 | 0 (0.0) | 9 (2.9) | 0.01 |
| 11 | 5 (2.4) | 51 (16.7) | 3 × 10−7 |
| 12 | 0 (0.0) | 5 (1.6) | NS |
| 13 | 10 (4.8) | 37 (12.1) | 4.4 × 10−3 |
| 14 | 3 (1.4) | 12 (3.9) | NS |
NS, not significant.
*Number of chromosomes.
HLA-DR Sequence Analyses in APS3v (T1D+AITD) Patients and Controls.
Unequivocal sequences were obtained in all 105 APS3v (T1D+AITD) patients and in 149 controls. Sequencing analysis of the polymorphic exon 2 of the DRB1 gene showed 13 polymorphic amino acid positions. Significant differences in the frequencies of amino acids between APS3v patients and controls were observed at 11 of the 13 polymorphic amino acids, at positions 26, 28, 30, 37, 47, 67, 70, 71, 74, 77, and 86 (Table 2). Of these, five amino acid positions showed the strongest associations with APS3v: Tyr-26 (P = 1.25 × 10−13, using the Cochran–Armitage test; Table 2); Leu-67 (P = 9.84 × 10−13); Gln-70 (P = 8.97 × 10−13); Lys-71 (P = 4.59 × 10−17); and Arg-74 (P = 1.21 × 10−13; Table 2). Intriguingly, Arg-74 is also the critical amino acid for the development of GD (20) and HT (21). Seven amino acids at the same five positions showed the strongest protective effects (Table 3): Leu and Phe at position 26 (P = 2.5 × 10−3 and 3.8 × 10−3, respectively); Ile-67 (P = 4.2 × 10−7); Asp-70 (P = 1 × 10−8); Glu and Ala at position 71 (P = 5.6 × 10−3 and 2.9 × 10−5, respectively); and Gln at position 74 (P = 4.1 × 10−6). Other significant amino acids that confer either susceptibility to or protection from APS3v are shown in Tables 2 and 3.
Table 2.
Polymorphic HLA-DR amino acid residues in APS3v patients and controls
| Amino acid position | Amino acid variants | APS3v (%) n = 210 | Controls (%) n = 298 | P value* |
| 26 | Tyrosine (Y) | 74 (35.2) | 24 (8.1) | 1.25 × 10−13 |
| All others | 136 (64.8) | 274 (91.9) | ||
| 28 | Aspartic acid (D) | 162 (77.1) | 191 (64.1) | 1.25 × 10−3 |
| All others | 48 (22.9) | 107 (35.9) | ||
| 30 | Tyrosine (Y) | 181 (86.2) | 192 (64.4) | 1.12 × 10−7 |
| All others | 29 (13.8) | 106 (35.6) | ||
| 37 | Tyrosine (Y) | 115 (54.8) | 103 (34.6) | 2.37 × 10−5 |
| All others | 95 (45.2) | 195 (65.4) | ||
| 47 | Tyrosine (Y) | 170 (81) | 188 (63.1) | 3.79 × 10−7 |
| All others | 40 (19) | 110 (36.9) | ||
| 67 | Leucine (L) | 180 (85.7) | 169 (56.7) | 9.84 × 10−13 |
| All others | 30 (14.3) | 129 (43.3) | ||
| 70 | Glutamine (Q) | 177 (84.3) | 161 (54) | 8.97 × 10−13 |
| All others | 33 (15.7) | 137 (46) | ||
| 71 | Lysine (K) | 135 (64.3) | 69 (23.2) | 4.59 × 10−17 |
| All others | 75 (35.7) | 229 (76.8) | ||
| 74 | Arginine (R) | 65 (31) | 14 (4.7) | 1.21 × 10−13 |
| All others | 145 (69) | 284 (95.3) | ||
| 77 | Asparagine (N) | 74 (35.2) | 72 (24.2) | 3.15 × 10−3 |
| All others | 136 (64.8) | 226 (75.8) | ||
| 86 | Glycine (G) | 114 (54.3) | 126 (42.3) | 1.47 × 10−3 |
| All others | 96 (45.7) | 172 (57.7) |
*P values were calculated using the Cochran–Armitage test; n indicates number of chromosomes.
Table 3.
HLA-DR amino acid variants at DR positions that showed strong association with APS3v (T1D+AITD)
| Amino acid position | Amino acid variants | APS3v (%) n = 210 | Controls (%) n = 298 | P value | OR | CI |
| 26 | Phenylalanine (F) | 122 (58.1) | 210 (70.9) | 3.8 × 10−3 | 0.58 | 0.39–0.86 |
| Tyrosine (Y) | 74 (35.2) | 24 (8.1) | 2.06 × 10−14 | 6.21 | 3.65–10.63 | |
| Leucine (L) | 13 (6.2) | 44 (14.9) | 2.5 × 10−3 | 0.38 | 0.19–0.75 | |
| Asparagine (N) | 1 (0.5) | 0 | NS | |||
| Undetermined* | 0 | 20 (6.7) | NS | |||
| 67 | Leucine (L) | 180 (85.7) | 169 (56.7) | 3.8 × 10−12 | 4.58 | 2.86–7.37 |
| Isoleucine (I) | 20 (9.5) | 83 (27.9) | 4.2 × 10−7 | 0.27 | 0.16–0.47 | |
| Phenylalanine (F) | 10 (4.8) | 39 (13.1) | 1.7 × 10−3 | 0.33 | 0.15–0.71 | |
| Undetermined* | 0 | 7 (2.3) | ||||
| 70 | Glutamine (Q) | 177 (84.3) | 161 (54) | 1.1 × 10−12 | 4.56 | 2.89–7.23 |
| Aspartic acid (D) | 26 (12.4) | 104 (34.9) | 1 × 10−8 | 0.26 | 0.16–0.43 | |
| Arginine (R) | 6 (2.9) | 26 (8.7) | 7 × 10−3 | 0.31 | 0.11–0.80 | |
| Histidine (H) | 1 (0.5) | 0 | NS | |||
| Undetermined* | 0 | 7 (2.3) | ||||
| 71 | Lysine (K) | 135 (64.3) | 69 (23.2) | 1.25 × 10−20 | 5.97 | 3.97–9 |
| Glutamic acid (E) | 7 (3.3) | 29 (9.7) | 5.6 × 10−3 | 0.32 | 0.12–0.78 | |
| Alanine (A) | 1 (0.5) | 27 (9.1) | 2.9 × 10−5 | 0.05 | 0.00–0.33 | |
| Arginine (R) | 67 (31.9) | 166 (55.7) | 1.1 × 10−7 | 0.37 | 0.25–0.55 | |
| Undetermined* | 0 | 7 (2.3) | ||||
| 74 | Alanine (A) | 102 (48.6) | 144 (48.3) | NS | ||
| Arginine (R) | 65 (31) | 14 (4.7) | 8.9 × 10−16 | 9.09 | 4.77–17.6 | |
| Glutamic acid (E) | 18 (8.6) | 44 (14.8) | 0.03 | 0.54 | 0.29–1.00 | |
| Glutamine (Q) | 21 (10) | 79 (26.5) | 4.1 × 10−6 | 0.31 | 0.18–0.53 | |
| Leucine (L) | 4 (1.9) | 14 (4.7) | NS | |||
| Undetermined* | 0 | 3 (1) |
CI, confidence interval; NS, not significant; n indicates number of chromosomes.
*Unequivocal determination of amino acids was not possible.
Multiple Logistic Regression Analysis.
To determine which amino acids and haplotypes are causative, and to distinguish them from amino acids that are associated with APS3v only due to linkage disequilibrium (LD), we performed a logistic regression analysis. Eleven amino acid positions were polymorphic and showed association with APS3v (Table 2). However, using all of them in the logistic regression analysis would result in too many haplotypes, reducing the power of our analysis. Therefore, and in view of previous functional studies showing that the associated amino acids cause major changes in the structure of the peptide binding pocket (21, 22), we focused on the amino acid variants that created significant alterations to the HLA-DR peptide binding pocket. These included Tyr-26, Leu-67, Lys-71, and Arg-74. This haplotype showed significant association with APS3v (P = 5.4 × 10−14, OR = 8.38, confidence interval = 4.5–16.9). Moreover, these four amino acid positions showed significant differences between APS3v and Hashimoto's thyroiditis patients (Table S1). The logistic regression analysis showed a strong independent effect for Leu-67 (P = 1.6 × 10−3, when controlling for the effects of other amino acid residues) and Arg-74 (P = 5.1 × 10−3, when controlling for the effects of other amino acid residues). Therefore, our analysis for independent effects demonstrated that Leu-67 and Arg-74 gave the main contributions to risk, whereas the other two amino acid positions were associated only by LD. It is very interesting that Arg-74, the main contributor to risk for AITD, is also one of the two main amino acids contributing to the risk for APS3v.
Structural Modeling Studies.
Comparison between APS3v-susceptible and -resistant HLA-DR.
To analyze the structure of the APS3v-associated DR pocket, we substituted the genetically significant amino acid positions in the crystal structure of HLA-DR3 for those in the APS3v-susceptible HLA-DR (Table 3). In addition, the CLIP peptide was replaced with a representative peptide that was previously shown to bind to DR3 (Tg2098) (22). Fig. 1 shows the complex in which the Tg2098 peptide (purple) occupies the canonical pockets in the binding groove of the APS3v-associated protein. The residues that discriminate between APS3v-susceptible and -resistant types (yellow in Fig. 1) cluster in one region of the protein that is positioned near the central residues in the peptide P4, P5, and P7, which have been shown to be important in selective recognition. All of the substitutions in residues that convert resistant types to susceptible types are linked to areas of the pocket that interact directly with the peptide and, therefore, are expected to influence immunogenic peptide binding. The major difference in the binding pockets is that in APS3v-susceptible HLA-DR they are more positively charged (Lys-71, Arg-74) compared with the APS3v-resistant HLA-DR (Ala-71, Gln-74). An analysis of the pairwise residue interactions between the HLA-DR protein and the peptide shows that the substitutions from the susceptible to the resistant type reduce the strength of these interactions (Fig. S1). For example, the interaction of the peptide residue in position 4 (P4) is reduced by 1.2 kcal/mol due to replacement of R74Q and K71A, which also reduces the interaction with P5 by 3 kcal/mol. The substitution L67I reduced the interaction with P7 by 0.8 kcal/mol. These changes would be expected to affect the selectivity of binding of pathogenic peptides to the HLA-DR pocket, as shown by the increase in the strength of the interaction in the APS3v-susceptible pocket. Thus, the presence of specific residues in these positions could increase the risk for developing T1D+AITD (APS3v).
Fig. 1.
A structural model of the APS3v-susceptible DR3 in complex with a representative peptide (purple) which was previously shown to bind to DR3. The genetically relevant residues (yellow) are clustered near the peptide residues P4, P5, and P7 (shown by gray arrows). The three critical residues DRβ-L67, DRβ-K71, and DRβ-R74 are shown by yellow arrows. The residues in the binding groove of the HLA were selected by proximity (7 Å) to the peptide and are colored according to their type: blue, positively charged; red, negatively charged; green, polar; white, hydrophobic.
Discussion
Epidemiological studies across different geographic and ethnic populations have consistently shown a frequent co-occurrence of T1D and AITD within the same individual (designated APS3v) (1). However, the mechanisms underlying the strong association between T1D and AITD are poorly understood. In the current report, we show that a molecular signature of the HLA-DR pocket, determined by specific amino acids, confers a significant risk for the development of APS3v (T1D+AITD) that is higher than the risk conferred by any previously identified APS3v gene (1). The haplotype consisting of HLA-DR pocket amino acids Tyr-26, Leu-67, Lys-71, and Arg-74 was strongly associated with APS3v in our dataset. In contrast, seven pocket amino acids were protective: Leu-26, Phe-26, Ile-67, Asp-70, Glu-71, Ala-71, and Gln-74. Importantly, Arg-74, already shown to be the critical amino acid for the development of HT and GD, is also the main contributor to the susceptibility to APS3v, suggesting that Arg-74 is a key DR pocket amino acid for the development of several autoimmune diseases.
The likely mechanism by which HLA class II proteins confer susceptibility to autoimmunity is through alterations in pathogenic peptide presentation. T cells recognize and respond to an antigen by interacting with a complex between an antigenic peptide and an HLA class II molecule (1, 23). When autoantigenic peptides are presented by HLA class II and activate autoreactive T cells that have escaped tolerance, the result is autoimmunity. Therefore, it is clear that the MHC II pocket structure is critical to the etiology of autoimmunity, as different pocket variants have different affinities to autoantigenic peptides and therefore certain variants are more likely to present autoantigenic peptides to T cells than others (24). This could determine whether an autoimmune response to that autoantigen will develop.
For both T1D and AITD, the structural-functional correlations between HLA class II pocket variants and peptide binding have been established (21, 25). In T1D, the amino acid residue at position 57 of the DQβ chain plays a key role in the genetic susceptibility to disease (26). Lack of Asp at this position at both DQ alleles is strongly associated with T1D (27). Genetic-structural studies performed by us in AITD have shown that Arg-74 is the critical HLA-DR pocket amino acid associated with GD (20) and HT (21). Because T1D and AITD are associated with different HLA II pocket structures, we asked the question: Which HLA class II pocket structure predisposes to T1D+AITD in the same individual (APS3v)? Our previous studies showed that HLA-DR3 is the critical HLA class II molecule conferring joint susceptibility to T1D and AITD (11). In the current study, we defined the HLA-DR pocket that confers susceptibility to both T1D and AITD.
Our findings raise the question of how the same HLA class II pocket can confer joint susceptibility to T1D and AITD if the autoantigenic peptides causing each disease are distinct (islet cell antigens in T1D and thyroid antigens in AITD). There could be at least two potential explanations for this paradox: (i) Due to the weak fit of peptides inside the HLA class II pockets, the same HLA-DR pocket predisposes to both T1D and AITD by enabling the presentation of islet cell peptides as well as thyroidal peptides; and (ii) the same pocket variant predisposes to both diseases, but its influence is on anchoring the T-cell receptor to the peptide-HLA-DR complex and not on peptide binding. Supporting the latter mechanism are data demonstrating that HLA-DR3 is associated with several autoimmune diseases targeting different autoantigens. During the presentation of antigenic peptides by HLA class II molecules to the T-cell receptor, some of the HLA pocket amino acids serve as anchors for the T-cell receptor (28). It is possible that amino acids that show strong association with both T1D and AITD (e.g., arginine at position 74 of the DRβ1 chain) predispose to disease by increasing the contact between the HLA-DR-peptide complex and the T-cell receptor. Alternatively, it is possible that the alleles predisposing to T1D and AITD are distinct but in tight LD.
The logistic regression analysis showed that Leu-67 and Arg-74 gave strong and independent risk for disease. The finding that Arg-74 is a major pocket amino acid predisposing to APS3v is intriguing, as this amino acid was also associated with GD (20) and HT (21). Thus, it is possible that Arg-74 is a key HLA-DR pocket amino acid for the development of several autoimmune diseases. However, the mechanism by which Arg-74 confers susceptibility to various autoimmune conditions is unclear. Position 74 of the DRβ chain is located in pocket 4 of the DR peptide binding cleft. Structural modeling analysis demonstrated that the change at position 74, from the common neutral amino acids (Ala or Gln) to a positively charged hydrophilic amino acid (Arg), significantly modified the 3D structure of pocket 4 of the peptide binding cleft (Fig. 1) (20, 21). This could alter the peptide binding properties of the pocket, favoring autoantigenic peptides originating both from thyroid antigens and islet antigens (20, 29). Alternatively, Arg-74 could help anchor T-cell receptors to the peptide-MHC II complex.
In summary, we have identified an HLA-DR pocket amino acid signature that confers significant joint risk for T1D and AITD in the same individual (APS3v). These HLA-DR pocket signatures create specific pocket structures which could accommodate autoantigenic peptides that may initiate both T1D and AITD or may facilitate the anchoring of the T-cell receptor to the peptide-MHC II complex. These findings suggest that blocking the critical Arg-74 pocket might offer a method for treating certain autoimmune conditions.
Patients and Methods
Patients and Controls.
T1D-AITD patients.
The project was approved by the Mount Sinai institutional review board. One hundred and five Caucasian T1D-AITD patients were studied. Of the AITD patients, 93 had HT and 12 had GD. The diagnosis of HT was established on the basis of (i) documented clinical and biochemical hypothyroidism requiring thyroid hormone replacement and (ii) the presence of autoantibodies to thyroid peroxidase (TPO) and/or thyroglobulin (Tg). Anti-thyroglobulin and anti-TPO antibodies were measured by specific RIA (Kronus). The diagnosis of GD was based on (i) documented clinical and biochemical hyperthyroidism requiring medical, surgical, and/or radioactive iodine treatment, (ii) the presence of thyroid stimulating hormone (TSH) receptor antibodies and/or a diffusely increased 131I uptake on a thyroid scan, (iii) and diffuse goiter. TSH receptor antibodies were measured using a radioreceptor assay (Kronus). T1D was diagnosed based on American Diabetes Association criteria (30).
Controls.
One hundred and fifty-three healthy Caucasian subjects were used as controls. All controls had no personal or family history of thyroid disease and T1D; they had normal thyroid functions and were negative for thyroid autoantibodies (anti-Tg and anti-TPO antibodies).
DNA Preparation.
DNA was extracted from whole blood using the Puregene kit (Gentra Systems).
HLA-DR Typing.
All patients and controls were typed for HLA-DR. Molecular typing of HLA-DR was performed according to the requirements of the American Society for Histocompatibility (31). The alleles of HLA-DR were typed using the technique of group-specific PCR amplification followed by restriction enzyme digestion as previously described (20).
Direct Sequencing of Exon 2 of the Human HLA-DRB1 Gene.
To identify HLA-DR sequences that predispose to APS3v, we sequenced exon 2 of the HLA-DRB1 gene, which encodes the specific amino acids determining DR specificities. Unequivocal sequencing of exon 2 of the HLA-DRB1 gene was achieved in all patients (n = 105) and in 149 of the control subjects. Genomic DNA was amplified and sequenced using allele-specific primers as previously described (21). PCRs were performed in a 20-μL reaction mixture containing 50 ng genomic DNA; 5 pmol of each primer; PCR buffer containing 50 mmol/L KCl; 10 mmol/L Tris-HCl (pH 8.3); 2.25 mmol/L MgCl2; 200 mmol/L of each deoxy (d)ATP, dGTP, dTTP, and dCTP; and 1 U of AmpliTaq DNA polymerase (Applied Biosystems). Reaction mixtures were heated to 94 °C for 7 min, and then cycled 30 times as follows: 30 s at 94 °C, 30 s at 55 °C, and 30 s at 72 °C. The PCR products were sequenced using the same allele-specific primers used for amplification. Sequencing was performed using the ABI Big Dye DNA-sequencing kit (Applied Biosystems), and the sequencing products were separated on an ABI-3130 automated sequencer (Applied Biosystems).
Sequence Alignment and Analysis.
Sequence alignment analyses were performed using Sequencher version 3.1 (Gene Codes). The known DR1, DR2, DR3, DR4, DR7, DR8, DR9, DR10, DR11, DR12, DR13, and DR14 subtype sequences were obtained from the ImMunoGeneTics project (IMGT)/HLA Informatics Group (HIG) database (http://www.ebi.ac.uk/imgt/hla).
Statistical Analyses.
The analyses for association between APS3v and the HLA-DR variants were performed using the χ2 and the Cochran–Armitage trend tests. For the logistic regression analysis haplotypes were imputed based on multimarker predictors using the E-M algorithm embedded in a stepwise logistic regression framework. The logistic regression framework allowed us to determine which allele or combination of alleles may have a causal role in disease, and which show association with disease because of LD. Likelihood ratio tests were used to test nested models. Statistical analyses were performed using the statistical software Epi Info [version 6.03, Centers for Disease Control and Prevention (CDC)], Stata 10 (Stata), and PLINK (http://pngu.mgh.harvard.edu/purcell/plink) (32).
Power Calculations.
Power calculations were performed using CDC simulation software (Epi Info, version 3.3.2, CDC) (21). We assumed the population frequency of the susceptibility amino acid variants to be 10–50% based on the population frequencies observed in our controls for the 13 polymorphic HLA-DR amino acids we tested for association with APS3v. The power calculations demonstrated that our dataset of 105 APS3v patients and 149 controls would give the study 80% power to detect a difference between the patients and the controls resulting in odds ratios of ≥2.15 with an α of 0.05, and 90% power to detect a difference resulting in odds ratios of ≥2.37. Therefore, our study was sufficiently powered to detect functionally significant differences between patients and controls.
Construction of Model Structures.
The 3D structure of the APS3v-associated DR3 protein was constructed based on the crystallographic structure of the human HLA-DR3 class II molecule in complex with the invariant chain peptide (CLIP) (Protein Data Bank ID code 1A6A) (33). The genetically significant amino acids at positions β26, β67, β71, and β74 in the original crystal structure were substituted by the corresponding residues in the APS3v-susceptible HLA-DR. The substitutions were: Yβ26→F, Lβ67→I, Kβ71→A, and Rβ74→Q. The CLIP peptide in the complex was substituted by a representative peptide previously shown to bind to DR3 (Tg2098) (22). The sequence of the peptide is LSSVVVDPSIRHFDV. The side chains of the CLIP peptide were replaced with those of Tg2098 while keeping the backbone conformation of CLIP the same as in the X-ray structure.
Optimization and Molecular Dynamics Simulations of the APS3v-Tg2098 Structure.
The complex was placed in a periodic box that extended 10 Å beyond the boundaries of the molecule. The box was filled with waters and ions to neutralize the system. The entire system was minimized and heated to 300 K while keeping the protein complex positionally restrained with a force constant of 25 kcal/mol/Å2. Subsequently, the complex was relaxed in a series of equilibration steps gradually reducing the restraints to zero. Molecular dynamics simulation of the equilibrated system was conducted at constant temperature and pressure for 10 ns integrating the equations of motion every 2 fs. The trajectories were saved every 1 ps and used to analyze the structure and dynamics of the complex.
Supplementary Material
Acknowledgments
This work was supported in part by Grants DK61659, DK067555, and DK073681 (to Y.T.) and MH48858 and NS27941 (to D.A.G.) from the National Institutes of Health.
Footnotes
The authors declare no conflict of interest.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1009511107/-/DCSupplemental.
References
- 1.Huber A, Menconi F, Corathers S, Jacobson EM, Tomer Y. Joint genetic susceptibility to type 1 diabetes and autoimmune thyroiditis: From epidemiology to mechanisms. Endocr Rev. 2008;29:697–725. doi: 10.1210/er.2008-0015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Karvonen M, et al. Incidence of childhood type 1 diabetes worldwide. Diabetes Mondiale (DiaMond) Project Group. Diabetes Care. 2000;23:1516–1526. doi: 10.2337/diacare.23.10.1516. [DOI] [PubMed] [Google Scholar]
- 3.Jacobson DL, Gange SJ, Rose NR, Graham NM. Epidemiology and estimated population burden of selected autoimmune diseases in the United States. Clin Immunol Immunopathol. 1997;84:223–243. doi: 10.1006/clin.1997.4412. [DOI] [PubMed] [Google Scholar]
- 4.Betterle C, Zanchetta R. Update on autoimmune polyendocrine syndromes (APS) Acta Biomed. 2003;74:9–33. [PubMed] [Google Scholar]
- 5.Eisenbarth GS, Gottlieb PA. Autoimmune polyendocrine syndromes. N Engl J Med. 2004;350:2068–2079. doi: 10.1056/NEJMra030158. [DOI] [PubMed] [Google Scholar]
- 6.Michels AW, Gottlieb PA. Autoimmune polyglandular syndromes. Nat Rev Endocrinol. 2010;6:270–277. doi: 10.1038/nrendo.2010.40. [DOI] [PubMed] [Google Scholar]
- 7.Kordonouri O, et al. Thyroid autoimmunity in children and adolescents with type 1 diabetes: A multicenter survey. Diabetes Care. 2002;25:1346–1350. doi: 10.2337/diacare.25.8.1346. [DOI] [PubMed] [Google Scholar]
- 8.Bright GM, Blizzard RM, Kaiser DL, Clarke WL. Organ-specific autoantibodies in children with common endocrine diseases. J Pediatr. 1982;100:8–14. doi: 10.1016/s0022-3476(82)80227-8. [DOI] [PubMed] [Google Scholar]
- 9.Levin L, Tomer Y. The etiology of autoimmune diabetes and thyroiditis: Evidence for common genetic susceptibility. Autoimmun Rev. 2003;2:377–386. doi: 10.1016/s1568-9972(03)00080-6. [DOI] [PubMed] [Google Scholar]
- 10.Holl RW, et al. Thyroid autoimmunity in children and adolescents with type 1 diabetes mellitus. Effect of age, gender and HLA type. Horm Res. 1999;52:113–118. doi: 10.1159/000023446. [DOI] [PubMed] [Google Scholar]
- 11.Golden B, et al. Genetic analysis of families with autoimmune diabetes and thyroiditis: Evidence for common and unique genes. J Clin Endocrinol Metab. 2005;90:4904–4911. doi: 10.1210/jc.2004-2236. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Huang W, et al. Although DR3-DQB1*0201 may be associated with multiple component diseases of the autoimmune polyglandular syndromes, the human leukocyte antigen DR4-DQB1*0302 haplotype is implicated only in β-cell autoimmunity. J Clin Endocrinol Metab. 1996;81:2559–2563. doi: 10.1210/jcem.81.7.8675578. [DOI] [PubMed] [Google Scholar]
- 13.Howson JM, et al. A type 1 diabetes subgroup with a female bias is characterised by failure in tolerance to thyroid peroxidase at an early age and a strong association with the cytotoxic T-lymphocyte-associated antigen-4 gene. Diabetologia. 2007;50:741–746. doi: 10.1007/s00125-007-0603-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Dultz G, et al. The protein tyrosine phosphatase non-receptor type 22 C1858T polymorphism is a joint susceptibility locus for immunthyroiditis and autoimmune diabetes. Thyroid. 2009;19:143–148. doi: 10.1089/thy.2008.0301. [DOI] [PubMed] [Google Scholar]
- 15.Villano MJ, et al. Autoimmune thyroiditis and diabetes: Dissecting the joint genetic susceptibility in a large cohort of multiplex families. J Clin Endocrinol Metab. 2009;94:1458–1466. doi: 10.1210/jc.2008-2193. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Dittmar M, Ide M, Wurm M, Kahaly GJ. Early onset of polyglandular failure is associated with HLA-DRB1*03. Eur J Endocrinol. 2008;159:55–60. doi: 10.1530/EJE-08-0082. [DOI] [PubMed] [Google Scholar]
- 17.Levin L, et al. Analysis of HLA genes in families with autoimmune diabetes and thyroiditis. Hum Immunol. 2004;65:640–647. doi: 10.1016/j.humimm.2004.02.026. [DOI] [PubMed] [Google Scholar]
- 18.Tomer Y, Menconi F. Type 1 diabetes and autoimmune thyroiditis: The genetic connection. Thyroid. 2009;19:99–102. doi: 10.1089/thy.2008.1565. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Todd JA, Bell JI, McDevitt HO. HLA-DQ β gene contributes to susceptibility and resistance to insulin-dependent diabetes mellitus. Nature. 1987;329:599–604. doi: 10.1038/329599a0. [DOI] [PubMed] [Google Scholar]
- 20.Ban Y, et al. Arginine at position 74 of the HLA-DR β1 chain is associated with Graves’ disease. Genes Immun. 2004;5:203–208. doi: 10.1038/sj.gene.6364059. [DOI] [PubMed] [Google Scholar]
- 21.Menconi F, et al. Molecular amino acid signatures in the MHC class II peptide-binding pocket predispose to autoimmune thyroiditis in humans and in mice. Proc Natl Acad Sci USA. 2008;105:14034–14039. doi: 10.1073/pnas.0806584105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Jacobson EM, et al. Employing a recombinant HLA-DR3 expression system to dissect major histocompatibility complex II-thyroglobulin peptide dynamism: A genetic, biochemical, and reverse immunological perspective. J Biol Chem. 2009;284:34231–34243. doi: 10.1074/jbc.M109.041574. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Buus S, Sette A, Grey HM. The interaction between protein-derived immunogenic peptides and Ia. Immunol Rev. 1987;98:115–141. doi: 10.1111/j.1600-065x.1987.tb00522.x. [DOI] [PubMed] [Google Scholar]
- 24.Faas S, Trucco M. The genes influencing the susceptibility to IDDM in humans. J Endocrinol Invest. 1994;17:477–495. doi: 10.1007/BF03347743. [DOI] [PubMed] [Google Scholar]
- 25.Pugliese A. Genetics of type 1 diabetes. Endocrinol Metab Clin North Am. 2004;33:1–16. doi: 10.1016/S0889-8529(03)00082-3. vii. [DOI] [PubMed] [Google Scholar]
- 26.Aitman TJ, Todd JA. Molecular genetics of diabetes mellitus. Baillieres Clin Endocrinol Metab. 1995;9:631–656. doi: 10.1016/s0950-351x(95)80655-5. [DOI] [PubMed] [Google Scholar]
- 27.Morel PA, Dorman JS, Todd JA, McDevitt HO, Trucco M. Aspartic acid at position 57 of the HLA-DQ β chain protects against type I diabetes: A family study. Proc Natl Acad Sci USA. 1988;85:8111–8115. doi: 10.1073/pnas.85.21.8111. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Deng L, Mariuzza RA. Recognition of self-peptide-MHC complexes by autoimmune T-cell receptors. Trends Biochem Sci. 2007;32:500–508. doi: 10.1016/j.tibs.2007.08.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Sawai Y, DeGroot LJ. Binding of human thyrotropin receptor peptides to a Graves’ disease-predisposing human leukocyte antigen class II molecule. J Clin Endocrinol Metab. 2000;85:1176–1179. doi: 10.1210/jcem.85.3.6376. [DOI] [PubMed] [Google Scholar]
- 30.Anonymous. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care. 1997;20:1183–1197. doi: 10.2337/diacare.20.7.1183. [DOI] [PubMed] [Google Scholar]
- 31.Marsh SG, et al. Nomenclature for factors of the HLA system, 2000. Hum Immunol. 2001;62:419–468. doi: 10.1016/s0198-8859(01)00229-4. [DOI] [PubMed] [Google Scholar]
- 32.Purcell S, et al. PLINK: A tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007;81:559–575. doi: 10.1086/519795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Ghosh P, Amaya M, Mellins E, Wiley DC. The structure of an intermediate in class II MHC maturation: CLIP bound to HLA-DR3. Nature. 1995;378:457–462. doi: 10.1038/378457a0. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.