Search
Close
Search
Advanced Search
Search Menu

In human type 1 diabetes (T1D) and in its murine model, the major histocompatibility complex (MHC) class II molecules, human leukocyte antigens (HLA)-DQ and -DR and their murine orthologues, IA and IE, are the major genetic determinants. In this report, we have ranked HLA class II molecule-associated T1D risk in a two-sided gradient from very high to very low. Very low risk corresponded to dominant protection from T1D. We predicted the protein structure of DQ by using the published crystal structures of different allotypes of the murine orthologue of DQ, IA. We discovered marked similarities both within, and cross species between T1D protective class II molecules. Likewise, the T1D predisposing molecules showed conserved similarities that contrasted with the shared patterns observed between the protective molecules. We also found striking inter-isotypic conservation between protective DQ, IA allotypes and protective DR4 subtypes. The data provide evidence for a joint action of the class II peptide-binding pockets P1, P4 and P9 in disease susceptibility and resistance with a main role for P9 in DQ/IA and for P1 and P4 in DR/IE. Overall, these results suggest shared epitope(s) in the target autoantigen(s), and common pathways in human and murine T1D.

INTRODUCTION

Most cases of type 1 diabetes (T1D) result from a T lymphocyte-dependent, selective destruction of the insulin-producing pancreatic β-cells and subsequent irreversible insulin deficiency. The disease is caused by both genetic and environmental factors. The major histocompatibility complex (MHC) human leukocyte antigen (HLA) region on human chromosome 6p21 contains the major disease locus insulin dependent diabetes mellitus 1 (IDDM1) (1–3). A large number of case-control and family-based association studies from different populations have indicated that common allelic variants at the class II HLA-DRB1, -DQA1 and -DQB1 loci are primarily and jointly associated with the disease. The three loci act as a complex superlocus, with both haplotype- and genotype-specific effects (4). Certain HLA-DR3 (DRB1*0301-DQA1-0501-DQB1*0201) and DR4 (DRB1*04-DQA1-0301-DQB1*0302) haplotypes are positively associated with the disease and represent the main predisposing molecules. In contrast, another haplotype, HLA-DR2 (DRB1*1501-DQA1-0102-DQB1*0602), is negatively associated with the disease in all populations.

The generation of the T cell repertoire and its autoimmune potential is controlled directly by HLA class II molecule-mediated editing of thymocyte antigen receptor specificities and affinities for class II-peptide (5–7). Class II molecules also regulate target antigen presentation at the sites of inflammation, as best illustrated by the DQ-regulated anti-gliadin response in coeliac disease (8). Within the peptide-binding sites of DR and DQ molecules the presence or absence of critical residues has been proposed to play a crucial role in conferring predisposition to, and protection from, the disease (3,5,7,9–12). Certain residues, in particular residue 57 of the DQ β chain and its charged/non-charged amino acid dimorphism, correlate well, though not completely, with T1D resistance and susceptibility. However, there is still uncertainty about relative associations of different class II alleles and haplotypes. Namely, whether common lower risk haplotypes such as DRB1*1501-DQB1*0602 are dominantly protective or simply reduced in frequency in cases versus controls owing to the increase in frequency of common susceptibility haplotypes such as DRB1*04-DQB1*0302 and DRB1*0301-DQB1*0201. Furthermore, no one has yet comprehensively applied recently acquired knowledge of the crystal structures of class II molecules in the context of the true relative predisposition of each class II molecule. In the present report we have addressed both these questions, identifying significant correlates not only across allotypes but also across isotypes and species. In mouse models of T1D and immune tolerance, the mechanisms distal to the class II-peptide-T cell receptor (TCR) interaction are being elucidated (13–15). Given the striking human–mouse conservation in class II sequence and structure correlates with T1D observed here, we predict conservation of autoantigenic β-cell peptide sequences and downstream mechanisms.

RESULTS

The relative association of DR and DQ

We ranked the three-locus DRB1-DQA1-DQB1 haplotypes into three main groups based on their transmission from parents to affected children (Table 1): (i) very frequently transmitted, >65% or very high risk (positively associated with T1D); (ii) between 65 and 10% transmitted or intermediate risk; and (iii) never/hardly ever transmitted or very low risk (highly negatively associated).

By analysis of the relative association of the three-locus haplotypes using the transmission disequilibrium test (TDT) and pairwise-TDT (PW-TDT) (Table 1) we can see which alleles constitute the contributions of the class II loci to disease association. A large part of the DQB1 association, independent of DRB1 and DQA1, is accounted for by the DQB1*0301/*0302 split of the DR4 haplotypes. The two most predisposing haplotypes, DRB1*0405-DQA1*0301-DQB1*0302 and DRB1*0401-DQA1*0301-DQB1*0302 [odds ratios for transmissions (ORTs) = 10.8 and 7.2, respectively], are reduced to an intermediate risk status by having the DQB1*0301 allele present rather than the DQB1*0302 allele (ORTs = 0.4 and 0.8, respectively). As expected in these European populations there are no informative splits of the HLA-DRB1-DQB1 haplotypes defined by DQA1 alleles. Hence, this sample set does not provide power to evaluate the individual contribution of DQA1.

The genetic contribution of the DRB1 locus is largely accounted for in these samples by the DR4 subtypes, in particular, by one allele, DRB1*0403. Because DRB1*0403 is relatively common in the Sardinian population, we were able to show here that it significantly reduces the very high risk of DQB1*0302 to an intermediate risk (Table 1; ORT = 0.5). The DR4 subtypes DRB1*0405 and DRB1*0401 were transmitted 18.6 and 18 times, respectively, more than the DRB1*0403 subtype on the same DQA1*03-DQB1*0302 background haplotype (Table 2: P = 5.2 × 10–8 and 1.7 × 10–9). Also, the DRB1*0404 and DRB1*0402 subtypes were transmitted 9.1 and 7.2 times, respectively, more than DRB1*0403 (Table 2: P = 5.4 × 10–5 and 5.3 × 10–3).

Genotype analysis

The analysis of transmission of alleles from one parent individually to a diabetic child as in Table 1 does not take into account the allele transmitted from the other parent, which together form the genotype of the offspring. Genotype-dependent effects are highly significant in T1D (4). In general, the presence of only one never/hardly ever transmitted haplotype results in a genotype that is neutrally or negatively associated with T1D (Table 3 and unpublished data). More specifically, genotypes positive for haplotypes such as DR2 are negatively associated with the disease, no matter what the other haplotype is (Table 3). Other haplotypes, such as DR5 or DR7-DQB1*0201, which by transmission analysis conferred less protection than DR2 (Table 1), generated genotypes that are only negatively associated with disease in the absence of the predisposing DR4-DQB1*0302 haplotype (Table 3). Both DR5 and DR7 are neutrally associated with T1D when a highly predisposing DR4-DQB1*0302 haplotype is present on the other chromosome. DR5 and DR7 haplotypes, therefore, reduce the risk of DR4-DQB1*0302 but not as much as DR2.

Structural comparisons of DQ and IA

At the time of this analysis the crystal structure of a DQ molecule was not published. However, in a straightforward alignment of the orthologous murine IA sequences, for which the crystal structures of three allotypes are known (IAk, IAd, IAg7) (16–19), we were able to predict the structures of the P1, P4, P6 and P9 peptide-binding pockets of the various predisposing and protective DQ molecules (Fig. 1).

We first considered the peptide-binding pockets of the murine IAg7, which is carried by the non-obese diabetes (NOD) mouse model of T1D, and of the two main DQ predisposing molecules observed in humans: DQ (α1*0301, β1*0302) and DQ (α1*0501, β1*0201), respectively (Table 4). Overall, IAg7 and DQ (α1*0301, β1*0302) have very similar pocket structures, with the exception of P1. P1 of IAg7 has a broad specificity for amino acids in peptide ligands with a general preference for hydrophobic and basic residues while disfavouring acidic residues (18,19). P1 of DQ (α1*0301, β1*0302) has several positive charges (Hisα24, Argα52, Argα53) predicting a high affinity for acid residues, although all other non-positively charged residues might be accepted. P1 of DQ (α1*0501, β1*0201) might have a preference for hydrophobic residues although it might also bind negatively charged amino acids as well.

Note that the DQα chain Arg52 residue, proposed previously as primary risk factor (20), corresponds in the predicted peptide-binding motifs reported in Figure 1B to position 49, which is located outside any peptide binding pockets. As such, this position is unlikely to be a primary aetiological amino acid variant in T1D.

DQ (α1*0301, β1*0302) and IAg7 show remarkable similarities in P4. At P4, these predisposing molecules could accommodate with high affinity, medium to large sized hydrophobic or non-charged residues (such as Ala, Val, Leu, Ile, Phe, Ser, Thr, Tyr). This class of amino acids might also be accepted at P4 of DQ (α1*0501, β1*0201) although acid residues (Asp, Glu) are preferred (21). At P6, the three predisposing molecules may bind preferentially medium sized hydrophobic or polar residues (Ala, Val, Thr, Asn), but acid residues (Asp, Glu) could also be accepted (21,22). The three predisposing molecules are similar in P9, which overall appears slightly more capacious than in the protective molecules (see below). DQ (α1*0501, β1*0201) might have an even larger P9 because it carries an Ile instead of the usual Tyr at position β37, which in the X-ray structure of the IAk allotype seems to play a role in the steric configuration of P9 (16). In P9, all the predisposing molecules might show a preference for acidic residues (Asp, Glu) although they could bind with lower affinity also medium to large sized hydrophobic residues. The lack of Asp at position β57 is the key feature in determining the specificity of this pocket (18,19,23). Also residues flanking β57 might contribute to the overall conformation of P9 and hence to peptide binding and T1D susceptibility. For instance, it has been proposed that a replacement of Pro with His at position β56 further contributes to T1D in the NOD mouse (24). Similarly, in the human DQβ1*0402 allele the presence of a Leu, instead of the usual Pro at position β56, could alter the conformation of P9, thus explaining why this allotype has an intermediate TID risk (Table 1) despite the presence of Asp at position β57.

We next considered the peptide-binding pockets of the diabetes protective murine IAb molecule and of the very low risk human DQ (α1*0501, β1*0301) and DQ (α1*0102, β1*0602) molecules (Table 5). These three molecules also show several similarities in their predicted peptide motif specificities. P1 is relatively deep and could accommodate medium to large sized hydrophobic residues with high affinity. Acidic residues, but not basic residues, could be also accepted in P1. IAb and DQ (α1*0501, β1*0301) are very similar in P4, which could bind, with high affinity, small hydrophobic residues (Gly, Ala). Medium-sized hydrophobic or non-charged residues (such as Val, Leu, Ile, Ser, Thr, Tyr) are also accepted in the slightly more capacious P4 of DQ (α1*0102, β1*0602). Note that the Leuβ26→Tyrβ26 polymorphism is the critical polymorphism, making the P4 pocket of IAb and DQ (α1*0501, β1*0301) smaller. These three protective molecules are also very similar in P6 where they would bind with high affinity small to medium sized hydrophobic residues. However, contrasting the structure of the predisposing (Table 4) and the protective molecules (Table 5), no obvious correlation was observed between the physical properties of P6 and the degree of T1D susceptibility or protection. Finally, the three protective molecules are also very similar in P9. It is likely that in this position all the protective molecules bind preferentially medium sized, non-charged residues such as Ser or Ala but also Gly, Thr, Gln. DQ Aspβ57 is the critical residue in P9. Its carboxylate group forms a salt bridge with Argα76 that stabilizes the heterodimer and determines the affinity of the peptide binding.

Taken together, these data suggest that the predisposing molecules, in particular the human DQ (α1*0301, β1*0302) and the mouse IAg7, are very similar in their active site and in their predicted peptide binding properties. Structurally, these T1D predisposing pockets are very different from those of the protective DQ molecules, which in turn are very similar to each other. This similarity is most striking for the trans-species pair, human DQ (α1*0501, β1*0301) and mouse IAb.

We next investigated how the overall size of P1, P4 and P9, resulting from the summation of the mass of the individual constituents, correlates with the disease risk in a larger number of DQ molecules (Table 6). A common trend is apparent: while the sum of the net mass of the amino acids forming P1 of the very low/intermediate risk allotypes tends to be smaller than those of the high risk molecules, the sum of the net mass of the amino acids forming P9 of the very low/intermediate risk allotypes tends to be larger than those of the high risk molecules. Consequently, P1 and P9 of the very low/intermediate risk molecules will be more capacious and shallower, respectively, than those of the high risk allotypes. It should be noted that in P4 the protective mouse IAb and the human DQ (α1*0501, β1*0301), DQ (α1*0301, β1*0301) molecules have a shallow pocket, which would predict a higher affinity for small residues. Overall, the best correlation between size of the pocket and disease risk is observed in P9.

DR4 subtypes

Analysis of the DR4 subtypes provides the most informative approach to pinpoint which DR β chain residues could have a role in T1D predisposition and protection. Indeed, DR4 subtypes aside, we failed to find any structural motif shared between alleles present on very high/intermediate risk haplotypes and differentiating them from alleles on very low risk haplotypes. The combined presence of Asp, Glu and Val at positions 57 (P9), 74 (P4) and 86 (P1), respectively, differentiate the intermediate risk DRβ1*0403 from the very high risk DRβ1*0405, carrying at the same positions, Ser, Ala and Gly (Fig. 1C). Position β74 is a key determinant of the binding of the peptide in P4. It is predicted that in this position the protective DRβ1*0403 cannot bind negatively charged residues with high affinity. In contrast, the P4 of the permissive DRβ1*0401 allotype will bind optimally Asp residue (Asp in P4 was indeed found in the X-ray crystal structure of a DRβ1*0401 molecule complexed with a collagen peptide) (25). It is predicted that also the predisposing DRB1*0405 would bind with high affinity Asp in P4.

Polymorphisms in DRβ86 (P1) are also important to explain the different degree of disease association of different DR4 subtypes. The Gly/Val dimorphism at DRβ86 influences side chain specificity at P1. In the P1 of DRβ1*0403 the presence of Val at β86 would create a preference for small to medium sized hydrophobic residues. In the same position, the smaller side chain of Gly in the permissive DRβ1*0405 and DRB1*0401 molecules allows the high affinity binding of large aromatic side chains. Experimental evidence indicates that substitutions of DRβ86 in the DRβ1*0401 allele modiy peptide binding and affect T cell recognition (26).

These predictions are consistent with experimental data from the mouse. IEg7, which is protective when expressed in Eα transgenic mice (27), carries Phe, which is large and non-polar, at position β86. Interestingly, another mouse allotype, IEk, which was able to confer some degree of protection in a transgenic mouse model (13) carries Glu at position 74, and a large residue (Phe) at position β86 in a manner reminiscent of the human DRβ1*0403 protective allele.

Taken together these data suggest that different degrees of association with T1D of different DR4 subtypes correlate well with differences in physical properties of key residues in P1 (position β86) and P4 (position β74). In turn, these structural features most likely reflect differences in peptide binding. Importantly, from the predicted structure of the DR (α1*0101, β*0403) heterodimer it is evident that this protective molecule is structurally similar to the IAb and DQ (α1*0501, β1*0301) protective molecules (Table 7). Note that for the inter-isotypic comparison shown in Table 6 the alignment has been performed using the X-ray structure of the IA molecules. However, similar results were also obtained aligning the molecules based on the known X-ray structure of the DRβ1*0401 subtype (data not shown).

DISCUSSION

Since 1987 Bain et al. (28), Cucca et al. (29) and Lernmark et al. (30) have been accumulating large, homogenous family and case-control data sets and carrying out HLA typing in order to more reliably analyse the association of HLA with T1D. In 1987 it was recognized that the DQβ57 residue correlated with susceptibility and resistance to T1D, the first ‘single nucleotide polymorphism’ identified in a common, multifactorial disease (9). In the same year, the HLA class I A2 allotype crystal structure predicted the central structural role of β57 in the peptide-binding site (31). However, the DQβ57 correlation was not complete (32), and a role for DR was also indicated (33), and confirmed (29,34,35). We have now brought together the available HLA typing data, combining this genetic analysis with the class II crystal structures solved over recent years, permitting a ‘second generation’ analysis of the correlation of the genetic association of T1D with the structure.

We have demonstrated that haplotypes that are never or hardly ever transmitted to T1D children are dominantly protective. We grouped DQ and DR (that is the DRB1*04 subtypes) alleles/haplotypes into very high risk, intermediate and very low risk categories to compare and contrast their predicted protein structures, including the known murine class II protein structures. The association of these class II molecules with T1D is the joint result of the structure and the action of the peptide-binding pockets, P1, P4 and P9. Most strikingly, protective molecules IAb, DR (α1*0101, β*0403) and DQ (α1*0501, β1*0301) showed significant similarities, particularly in P4 and P9. After completion of this analysis, the crystal structure of the DQ (α1*0301, β1*0302) molecule (HLA-DQ8) was reported (36). The results confirm our predictions of the DQ protein structure and suggest, as we have proposed here, that humans and NOD mice share similar peptide-presentation event(s) in T1D and that other features of the P9 pocket in addition to β57 play a role in T1D.

This conservation of structure indicates that not only is it likely that similar peptides are involved in T1D etiology in mice and humans but also that the mechanisms distal to class II-peptide-TCR interaction are shared. Owing to the difficulty of studying mechanisms in the immune system in humans, particularly in the generation and maintenance of tolerance, murine model systems have been developed and provided the first insights into these distal mechanisms. In murine models of T1D, low risk class II allotypes have been shown experimentally to be dominantly protective (13,14,24,27). The NOD mouse possesses, in homozygosity, a unique MHC class II haplotype (H2g7) (37,38). This haplotype carries a null IEa gene (IE is the murine orthologue of DR), and encodes the IA (IA is the murine orthologue of DQ) β chain, IAg7, carrying Ser at position 57 instead of the more common Asp. Studies of congenic NOD mice expressing non-NOD MHC haplotypes, and of NOD mice expressing IEαd, modified IAβg7, IAαk/IAβk or IAβd transgenes, have proved that both isotypes of class II molecules are directly involved in disease predisposition and protection (27,39–42). Moreover, some TCR transgenic mice models have been developed to study the mechanisms underlying disease resistance. Depending on the TCR studied, the protective effect of MHC class II allotypes such as IAb, is due to negative selection of the diabetogenic TCR in the thymus (13). The requirement of homozygous MHC class II expression in susceptibility to T1D in NOD mice is not simply dependent on the expression of a minimum number of predisposing IAg7 class II molecules on the cell surface (14). Other peptide-TCR combinations are involved in the class II-mediated generation of protective T regulatory cells with no evidence for deletion (15,43). Several mechanisms, therefore, are responsible for the generation of self-tolerance in the thymus, and perhaps also in the periphery (44), all of which depend on the structure of the class II peptide binding site.

Taken together, these results indicate that negative selection processes in the thymus, mediated by high affinity class II-peptide interactions, account in part for the dominant protection against T1D provided by IAb and, by extrapolation based on our results, certain DQ and DR alleles. Anti β-cell T cells could be deleted in the thymus more frequently in individuals carrying IAb, DQ (α1*0501, β1*0301), DR (α1*0101, β*0403) than in those individuals carrying very high risk alleles such as IAg7, DQ (α1*0301, β1*0302) or DQ (α1*0501, β1*0201). This model is not exclusive of other possible effects of class II structure on T1D predisposition, such as an overall instability of the high risk DQ and IAg7 molecules owing to the absence of βAsp57 and promiscuous peptide binding (18,45).

As highlighted by many studies, the affinity of the class II-peptide interaction is only one of many factors in the role of the thymus in ensuring unresponsiveness to the host’s own tissues. The overall avidity of the interaction involves several accessory molecules such as CD80, CD86, CD54, CTLA-4 and CD28. The levels of these molecules, the class II molecules and the autoantigenic peptides themselves in the thymus [and in the periphery (44)] will together determine the predisposition of an individual to T1D. Results from Hanahan et al. (46) and Jordan et al. (15) illustrate the importance of the level of autoantigen in the thymus. In particular, genetic analyses and expression of the insulin gene (INS) in humans has shown that promoter variants of INS (the IDDM2 locus on chromosome 11p15), which are known to be protective for T1D, cause higher levels of preoproinsulin (PPI) expression in the thymus (47,48). By analogy with results from murine systems (15,46), the protection provided by certain INS alleles may be via increased thymic tolerance owing to higher levels of autoantigen affecting the degree of negative selection or selection of T regulatory cells to key peptides from PPI. This model provides a direct link between the IDDM1/HLA class II locus and the IDDM2/INS promoter locus in T1D etiology. Insulin and its precursors are certainly among the most favoured candidates for the primary autoantigens in T1D. Our molecular modelling data suggest that the following epitopes from PPI (defined by the anchor positions from P1 to P9) might bind the main protective class II molecules with high affinity, triggering immune tolerance and non-responsiveness: LVEALYLVC (35–43 PPI), LQVGQVELG (61–69 PPI), VELGGGPGA (66–74 PPI), LGGGPGAGS (68–76 PPI), QKRGIVEQC (87–95 PPI). Note that the PPI peptides 35–43 overlap the β chain 9–23 (33–47 PPI) immunodominant epitope, which has been associated with diabetogenic T cell specificities in many studies (49–51). Moreover, another epitope LPLLALLAL (8–16 PPI) located in the peptide leader sequence, could bind with high affinity to the protective DQB1*0602-DQA1*0102 molecule.

The protective activity of class II molecules in T1D via presentation of critical autoantigens in the thymus is only one feature of the mechanisms underlying the association of HLA with T1D. Previous studies have attempted to analyse the association of the various HLA class II haplotypes with the human T1D after removing the main predisposing DR3/DR4 alleles (52). In this sub-group of patients, the intermediate risk alleles were apparently the most predisposing. These analyses, while confirming the continuum of association from the most to the least predisposing molecules, which we observed in the present study, mirror what happens in populations such as the Japanese, in which the highly predisposing European haplotypes are absent and the disease is extremely rare. Haplotypes that appear predisposing in Asian populations provide only intermediate risk in European populations, consistent with the overall risk they bestow in Asian populations (5,53). Importantly, these data suggest that an HLA class II antigen-presenting molecule is a necessary factor for disease occurrence. This molecule is most likely encoded by the alleles most positively associated with the disease but in their absence other molecules, with a lower probability, can invoke the same response, consistent with a continuum of risk, not an all-or-none association. Indeed, the risk of developing the disease will be different in the different HLA categories, as the disease prevalence is different in diverse populations. Hence, the continuum of risk from very high to very low reflects at least two HLA class II mechanisms in T1D, active protection and active predisposition (54). The latter most likely reflects active autoantigen presentation in the pancreas and its draining lymph nodes by the predisposing molecules as a key step in the initiation and maintenance of the inflammation in the target organ. The genetic evidence, therefore, supports a mixed model for the mechanisms of HLA class II molecules in T1D encompassing many of the features of past models: a thymic tolerance model (5,12) and a peripheral antigen presentation model (10).

Moreover, the genetics of HLA and T1D is even more complex than described here, with unexplained genotype effects such as the synergistic effect of DR3/4 heterozygosity in many, but not all populations, and the loss of the protective effect of DQB1*0301 in DR1/4 heterozygous individuals (4 and unpublished data). These genotype/trans effects in the formation of variable levels of certain DQ heterodimers could create or reduce peptide binding specificities and/or change the number of active molecules on the cell surface, both of which could influence class II molecule-mediated signals to T cells.

Further experiments, including peptide-binding measurements to molecules encoded in cis and trans by different class II alleles, and characterization of T cell responses to peptides early in the etiology of T1D, are required to substantiate the proposed model and clarify the sequence of molecular events that ultimately results in destruction of the pancreatic β-cells.

MATERIALS AND METHODS

Subjects

The family data set consisted of 255 (Sardinian), 277 (UK) and 180 (USA) T1D families (total no. of affected families = 712 and total no. of affected children = 1181; 267 from Sardinia, 554 from the UK and 360 from the USA). We have also considered 345 additional Sardinian sporadic patients (total no. of independent Sardinian patients = 600). 120 of the Sardinian sporadic patients have already been reported in a previous study (29). The average age at disease onset of the UK patients (probands) analysed in this study was 8.6 years, SD ± 6.1 years (minimum age = 6 months; maximum age = 26 years). The average age of patients from the USA was 9.3, SD ± 6.0 years (minimum age = 1 year; maximum age = 25 years). The average age of the Sardinian patients was 8.3 years SD ± 4.7 (minimum age = 10 months; maximum age = 23 years). All the UK families were part of the Diabetes UK (formerly the British Diabetic Association) Warren I Repository (28). The 180 USA multiplex families were from the Human Biological Data Interchange (HBDI) (30). These USA families overlap with those previously studied by Noble et al. (4).

HLA typing

The Sardinian sample set was typed through PCR amplification of the polymorphic second exon of the HLA-DRB1, -DQA1, -DQB1 genes and dot blot analysis of amplified DNA with sequence specific oligonucleotide (SSO) probes (55–58). DR4-specific amplification and SSO probe hybridization for the various DRB1*04 alleles were performed in the DR4 positive individuals (59). Typing data for the 180 USA families were obtained from the HBDI, from which DNA samples from family members were purchased. The 277 UK families were typed for the DRB1, DQB1 loci using a combination of serological and PCR-SSP methods by the Transplantation Unit in Oxford (60). In the UK, the DQA1 alleles were inferred based on the established linkage disequilibrium patterns with the DQB1 and DRB1 loci.

Statistical analyses

The association of the individual DRB1-DQA1-DQB1 haplotypes with T1D, was evaluated using the TDT (61). We also performed a hierarchical analysis of the association of haplotypes with the disease, around an arbitrarily chosen reference haplotype, the PW-TDT. When more than one haplotype is associated to a disease, the association of a test haplotype might be influenced by the other associated haplotypes. To minimize this problem and to provide a more accurate estimate of the strength of the association, and thus of the relative risk for disease, the test haplotype is analysed relative to a reference haplotype. The reference haplotype has been selected to be relatively frequent and similarly distributed in the populations considered. Moreover, we selected the reference haplotype to be not among the haplotypes in the very high or very low risk categories. Specifically, in the pairwise comparison of haplotypes, PW-TDT, the transmission and non-transmission counts for the test (A) and the reference (B) haplotypes are evaluated by TDT and the resulting data points deriving from the A/X and B/X parental meioses (where X is non-A, non-B) are arranged in a 2 × 2 contingency Table and tested by Fisher’s exact or Pearson’s χ2 test. To maintain independence of these data we excluded parents having both the alleles or haplotypes being considered (e.g. A/B), but the transmission data for those parents have been analysed by standard TDT and the statistic added to that of the 2 × 2 Table to give an overall χ2 test with 2 d.f. (62). The heterogeneity in transmission between the two haplotypes can be quantified by the ORTs. The transmission data from the individual parents carrying both the haplotypes being compared have been also discarded in computing the ORTs. Haplotypes were established following the co-segregation of alleles within families using computer program TDTPHASE written by F.Dudbridge and available through http://www-gene.cimr.cam.ac.uk/tdt/. Only haplotypes certain from parental genotype data, and in the absence of inter-crosses (that is when both parents are heterozygous for the same alleles), are considered in the analyses shown in this paper. Only probands from families with more than one affected sibling were evaluated. When the two compared haplotypes are identical at one locus (variant A) but different at another closely linked locus (variant B), the PW-TDT could be used as a conditional/stratification test, namely to assess whether there is heterogeneity in the transmission of variant B conditional on variant A. In this case we refer to the method as the haplotype method-TDT (HM-TDT) (62,63).

The TDT does not take into account the transmission (or the lack of transmission) of a haplotype from one parent relative to the haplotype transmitted from the other parent, that is genotype effects. Therefore, we carried out a case-control study in the homogeneous population of Sardinia, for which we have previously shown that the issue of genetic mismatching of the cases and controls is not a problem (64). In the UK/USA data set we used the AFBAC method assuming Hardy–Weinberg equilibrium (4,65). Note that in the Sardinian sets, the AFBAC genotype frequencies calculated from the families were virtually identical to those from 617 new-borns (not shown). The frequencies of the HLA class II genotypes obtained in patients and those observed in controls, were compared using a 2 × 2 χ2 test, or the Fisher exact test.

Protein analyses

We confined the structural analyses to the amino acid sequences encoded by the second exons of DQ and DR alleles with a clear association with the disease. These human sequences and the mouse orthologues were aligned. Determination of the residues forming the peptide binding pockets was performed based on molecules having a known X-ray crystal structure (DRB1*0401 and IEk for DR/IE and IAk, IAd and IAg7 for DQ/IA) (16–19,25,66). This analysis was facilitated by the high homology between the different class II alleles observed even in distantly related species. The physical properties, mass, polarity and charge of the amino acids in the peptide-binding pockets were tabulated and the optimal peptide motifs were predicted.

ACKNOWLEDGEMENTS

We wish to thank the Medical Research Council of UK, Eli Lilly UK, the Wellcome Trust, Diabetes UK, Assessorato Sanita’ Regione Autonoma Sardegna, the Italian Telethon and the Juvenile Diabetes Research Foundation International for financial support; A.Cao, M.Whalen, M.G.Marrosu, M.Silvetti, W.Kwok, P.Reed and P.Zavattari for help and advice; F.Dudbridge and H.Cordell for statistical advice; M.Chessa, P.Frongia and R.Ricciardi for DNA collection from Sardinian patients; the HBDI for USA family DNA samples and HLA typing data; H.Stevens, P.Carr and Diabetes UK for provision of UK families and DNA. F.C. and J.A.T. are recipients of a Wellcome Trust Biomedical Research Collaboration Grant, and J.A.T. was a Wellcome Trust Principal Research Fellow.

+

To whom correspondence should be addressed. Tel: +39 070 6095681; Fax: +39 070 6095558; Email: [email protected] Correspondence may also be addressed to John Todd at: Wellcome Trust/MRC Building, Addenbrooke’s Hospital, Hills Road, Cambridge CB2 2XY, UK. Tel: +44 1223 762101; Fax: +44 1223 762102; Email: [email protected]

Figure 1. Alignment of the amino acid sequences of DQ/IA and DR/IE allotypes of interest with molecules with a known X-ray crystal structure. Protein sequence alignment of corresponding DQβ-IAβ- (A), DQα-IAα,DRα (B, opposite) and DRβ-IEβ (C, opposite) human and mouse class II molecules. The amino acid positions refer to the mouse sequence. Positions of interest are labelled. Peptide-binding pocket number and amino acid residue number are indicated. ^, class II alleles with a known crystallographic structure. Dashes indicate missing residues. The established and predicted peptide-binding pockets are highlighted with different colours: P1, green; P4, blue; P6, pink; P7, grey; P9, yellow.

Figure 1. Alignment of the amino acid sequences of DQ/IA and DR/IE allotypes of interest with molecules with a known X-ray crystal structure. Protein sequence alignment of corresponding DQβ-IAβ- (A), DQα-IAα,DRα (B, opposite) and DRβ-IEβ (C, opposite) human and mouse class II molecules. The amino acid positions refer to the mouse sequence. Positions of interest are labelled. Peptide-binding pocket number and amino acid residue number are indicated. ^, class II alleles with a known crystallographic structure. Dashes indicate missing residues. The established and predicted peptide-binding pockets are highlighted with different colours: P1, green; P4, blue; P6, pink; P7, grey; P9, yellow.

Open in new tabDownload slide
Table 1.
Open in new tab

Association analysis of the HLA-DRB1-DQA1-DQB1 haplotypes in a mixed data set of 712 T1D families from UK, USA and Sardiniaa

SardinianUKUSATotal95% CI
T1D riskDRB1DQA1DQB1TNTTNTTNTTNT%TP TDT <0.05ORT95% CIP PW-TDT <0.05
Very high
DR4040503010302701311280891585.64.0 × 10–1310.85.6–20.62.3 × 10–14
DR4040103010302201251981212084083.91.4 × 10–267.24.6–11.54.2 × 10–20
DR30301050102012204515755992847612878.83.5 × 10–434.32.9–6.31.1 × 10–16
DR4 0404030103021031163212642869.61.7 × 10–44.12.3–7.17.3 × 10–6
DR40402030103021462164221166.73.11.4–6.88.0 × 10–3
Intermediate
DR40405030102016710108753.32.00.7–5.6
DR8080401040213815105192345.21.60.8–3.1
DR613020102060430371114172144.71.30.6–2.7
DR9090103010303024105191340.91.20.5–2.9
DR101010105011744466719308214136.89.6 × 10–51
DR2AZH16010102050247900482559636.41.3 × 10–30.80.5–1.2
DR40401030103011625215184031.03.8 × 10–30.80.4–1.5
DR707010201020152012431027279023.17.6 × 10–90.50.3–0.91.9 × 10–2
DR404050501030141641620.07.2 × 10–30.40.1–1.3
DR4040303010302313120141620.07.3 × 10–30.50.1–1.4
DR6130101030603042841762917.11.0 × 10–40.40.2–1.12.2 × 10–2
DR10100101010501214021231814.31.0 × 10–30.20.05–0.9
DR61303–13050501030114051621511.81.6 × 10–30.30.1–1.2
Very low
DR511–1205010301486533536141558.31.1 × 10–250.20.1–0.33.4 × 10–9
DR21501010206020226114331062.81.4 × 10–220.040.01–0.11.0 × 10–10
DR707010201030302010090210.04.6 × 10–60.10.01–0.64.8 × 10–4
DR61401010503011090100300.04.3 × 10–80.10.01–0.41.7 × 10–4
SardinianUKUSATotal95% CI
T1D riskDRB1DQA1DQB1TNTTNTTNTTNT%TP TDT <0.05ORT95% CIP PW-TDT <0.05
Very high
DR4040503010302701311280891585.64.0 × 10–1310.85.6–20.62.3 × 10–14
DR4040103010302201251981212084083.91.4 × 10–267.24.6–11.54.2 × 10–20
DR30301050102012204515755992847612878.83.5 × 10–434.32.9–6.31.1 × 10–16
DR4 0404030103021031163212642869.61.7 × 10–44.12.3–7.17.3 × 10–6
DR40402030103021462164221166.73.11.4–6.88.0 × 10–3
Intermediate
DR40405030102016710108753.32.00.7–5.6
DR8080401040213815105192345.21.60.8–3.1
DR613020102060430371114172144.71.30.6–2.7
DR9090103010303024105191340.91.20.5–2.9
DR101010105011744466719308214136.89.6 × 10–51
DR2AZH16010102050247900482559636.41.3 × 10–30.80.5–1.2
DR40401030103011625215184031.03.8 × 10–30.80.4–1.5
DR707010201020152012431027279023.17.6 × 10–90.50.3–0.91.9 × 10–2
DR404050501030141641620.07.2 × 10–30.40.1–1.3
DR4040303010302313120141620.07.3 × 10–30.50.1–1.4
DR6130101030603042841762917.11.0 × 10–40.40.2–1.12.2 × 10–2
DR10100101010501214021231814.31.0 × 10–30.20.05–0.9
DR61303–13050501030114051621511.81.6 × 10–30.30.1–1.2
Very low
DR511–1205010301486533536141558.31.1 × 10–250.20.1–0.33.4 × 10–9
DR21501010206020226114331062.81.4 × 10–220.040.01–0.11.0 × 10–10
DR707010201030302010090210.04.6 × 10–60.10.01–0.64.8 × 10–4
DR61401010503011090100300.04.3 × 10–80.10.01–0.41.7 × 10–4

The PW-TDT P-values as well as the ORT were calculated with pair wise comparisons using DRB1*01-DQA1*0101-DQB1*0501 as the reference haplotype (bold type). See Methods regarding the choice of the reference haplotype. T, transmitted to affected children using the TDT; NT, not transmitted to affected children using the TDT.

aOnly transmissions from at least 15 informative meioses in the total data set are considered in this Table.

Table 1.
Open in new tab

Association analysis of the HLA-DRB1-DQA1-DQB1 haplotypes in a mixed data set of 712 T1D families from UK, USA and Sardiniaa

SardinianUKUSATotal95% CI
T1D riskDRB1DQA1DQB1TNTTNTTNTTNT%TP TDT <0.05ORT95% CIP PW-TDT <0.05
Very high
DR4040503010302701311280891585.64.0 × 10–1310.85.6–20.62.3 × 10–14
DR4040103010302201251981212084083.91.4 × 10–267.24.6–11.54.2 × 10–20
DR30301050102012204515755992847612878.83.5 × 10–434.32.9–6.31.1 × 10–16
DR4 0404030103021031163212642869.61.7 × 10–44.12.3–7.17.3 × 10–6
DR40402030103021462164221166.73.11.4–6.88.0 × 10–3
Intermediate
DR40405030102016710108753.32.00.7–5.6
DR8080401040213815105192345.21.60.8–3.1
DR613020102060430371114172144.71.30.6–2.7
DR9090103010303024105191340.91.20.5–2.9
DR101010105011744466719308214136.89.6 × 10–51
DR2AZH16010102050247900482559636.41.3 × 10–30.80.5–1.2
DR40401030103011625215184031.03.8 × 10–30.80.4–1.5
DR707010201020152012431027279023.17.6 × 10–90.50.3–0.91.9 × 10–2
DR404050501030141641620.07.2 × 10–30.40.1–1.3
DR4040303010302313120141620.07.3 × 10–30.50.1–1.4
DR6130101030603042841762917.11.0 × 10–40.40.2–1.12.2 × 10–2
DR10100101010501214021231814.31.0 × 10–30.20.05–0.9
DR61303–13050501030114051621511.81.6 × 10–30.30.1–1.2
Very low
DR511–1205010301486533536141558.31.1 × 10–250.20.1–0.33.4 × 10–9
DR21501010206020226114331062.81.4 × 10–220.040.01–0.11.0 × 10–10
DR707010201030302010090210.04.6 × 10–60.10.01–0.64.8 × 10–4
DR61401010503011090100300.04.3 × 10–80.10.01–0.41.7 × 10–4
SardinianUKUSATotal95% CI
T1D riskDRB1DQA1DQB1TNTTNTTNTTNT%TP TDT <0.05ORT95% CIP PW-TDT <0.05
Very high
DR4040503010302701311280891585.64.0 × 10–1310.85.6–20.62.3 × 10–14
DR4040103010302201251981212084083.91.4 × 10–267.24.6–11.54.2 × 10–20
DR30301050102012204515755992847612878.83.5 × 10–434.32.9–6.31.1 × 10–16
DR4 0404030103021031163212642869.61.7 × 10–44.12.3–7.17.3 × 10–6
DR40402030103021462164221166.73.11.4–6.88.0 × 10–3
Intermediate
DR40405030102016710108753.32.00.7–5.6
DR8080401040213815105192345.21.60.8–3.1
DR613020102060430371114172144.71.30.6–2.7
DR9090103010303024105191340.91.20.5–2.9
DR101010105011744466719308214136.89.6 × 10–51
DR2AZH16010102050247900482559636.41.3 × 10–30.80.5–1.2
DR40401030103011625215184031.03.8 × 10–30.80.4–1.5
DR707010201020152012431027279023.17.6 × 10–90.50.3–0.91.9 × 10–2
DR404050501030141641620.07.2 × 10–30.40.1–1.3
DR4040303010302313120141620.07.3 × 10–30.50.1–1.4
DR6130101030603042841762917.11.0 × 10–40.40.2–1.12.2 × 10–2
DR10100101010501214021231814.31.0 × 10–30.20.05–0.9
DR61303–13050501030114051621511.81.6 × 10–30.30.1–1.2
Very low
DR511–1205010301486533536141558.31.1 × 10–250.20.1–0.33.4 × 10–9
DR21501010206020226114331062.81.4 × 10–220.040.01–0.11.0 × 10–10
DR707010201030302010090210.04.6 × 10–60.10.01–0.64.8 × 10–4
DR61401010503011090100300.04.3 × 10–80.10.01–0.41.7 × 10–4

The PW-TDT P-values as well as the ORT were calculated with pair wise comparisons using DRB1*01-DQA1*0101-DQB1*0501 as the reference haplotype (bold type). See Methods regarding the choice of the reference haplotype. T, transmitted to affected children using the TDT; NT, not transmitted to affected children using the TDT.

aOnly transmissions from at least 15 informative meioses in the total data set are considered in this Table.

Table 2.
Open in new tab

Conditional analysis of the DRB1*04 subtypes on fixed DQA1*0301-DQB1*0302 haplotypes

DRB1DQA1DQB1TNTORT95% CIP HM-TDT < 0.05
040503010302891518.65.4–64.95.2 × 10–8
0401030103022084018.05.6–57.61.7 × 10–9
04040301030264289.12.8–29.85.4 × 10–5
04020301030222117.21.9–26.95.3 × 10–3
0403030103024161.01
DRB1DQA1DQB1TNTORT95% CIP HM-TDT < 0.05
040503010302891518.65.4–64.95.2 × 10–8
0401030103022084018.05.6–57.61.7 × 10–9
04040301030264289.12.8–29.85.4 × 10–5
04020301030222117.21.9–26.95.3 × 10–3
0403030103024161.01

P-HM-TDT-values and ORTs value have been calculated with pairwise comparisons using DRB1*0403-DQA1*0301-DQB1*0302 as the reference haplotype (bold type).

T, transmitted to affected children using the TDT; NT, not transmitted to affected children using the TDT.

Table 2.
Open in new tab

Conditional analysis of the DRB1*04 subtypes on fixed DQA1*0301-DQB1*0302 haplotypes

DRB1DQA1DQB1TNTORT95% CIP HM-TDT < 0.05
040503010302891518.65.4–64.95.2 × 10–8
0401030103022084018.05.6–57.61.7 × 10–9
04040301030264289.12.8–29.85.4 × 10–5
04020301030222117.21.9–26.95.3 × 10–3
0403030103024161.01
DRB1DQA1DQB1TNTORT95% CIP HM-TDT < 0.05
040503010302891518.65.4–64.95.2 × 10–8
0401030103022084018.05.6–57.61.7 × 10–9
04040301030264289.12.8–29.85.4 × 10–5
04020301030222117.21.9–26.95.3 × 10–3
0403030103024161.01

P-HM-TDT-values and ORTs value have been calculated with pairwise comparisons using DRB1*0403-DQA1*0301-DQB1*0302 as the reference haplotype (bold type).

T, transmitted to affected children using the TDT; NT, not transmitted to affected children using the TDT.

Table 3.
Open in new tab

Distribution of (DR2) DRB1*1501-DQB1*0602, (DR5) DRB1*11–12-DQB1*0301 and (DR7) DRB1*0701-DQB1*0201 positive genotypes in the Sardinian, UK and USA patients and controlsa

PatientsControls
GenotypeUK/USA (n = 445)Sardinia (n = 600)Total (n = 1045)UK/USAa (n = 325)Sardinia (n = 617)Total (n = 942)P < 0.05bORb95% CI
DRB1DQB1DRB1DQB1n%n%n%n%n%n%
15010602040302c20.40020.26.82.130.59.811.4 × 10–20.20.04–0.8
1501060203010201000000144.430.5171.81.3 × 10–50.050.01–0.4
15010602Xd10.20010.16620111.8778.12.1 × 10–200.010.00–0.08
11–120301040302c61.3111.8171.63.91.2111.8151.61.0 0.5–2.1
11–1203010301020120.450.870.78.12.5447.1525.52.0 × 10–100.1 0.1–0.3
11–120301Xd10.20010.1391213121170184.7 × 10–460.000.00–0.03
07010201040302c112.500111.14.31.361101.11.0 0.4–2.3
070102010301020161.350.8111.192.8152.4242.51.1 × 10–20.4 0.2–0.8
07010201Xd81.830.5111.14313508.1939.91.2 × 10–180.10.05–0.18
PatientsControls
GenotypeUK/USA (n = 445)Sardinia (n = 600)Total (n = 1045)UK/USAa (n = 325)Sardinia (n = 617)Total (n = 942)P < 0.05bORb95% CI
DRB1DQB1DRB1DQB1n%n%n%n%n%n%
15010602040302c20.40020.26.82.130.59.811.4 × 10–20.20.04–0.8
1501060203010201000000144.430.5171.81.3 × 10–50.050.01–0.4
15010602Xd10.20010.16620111.8778.12.1 × 10–200.010.00–0.08
11–120301040302c61.3111.8171.63.91.2111.8151.61.0 0.5–2.1
11–1203010301020120.450.870.78.12.5447.1525.52.0 × 10–100.1 0.1–0.3
11–120301Xd10.20010.1391213121170184.7 × 10–460.000.00–0.03
07010201040302c112.500111.14.31.361101.11.0 0.4–2.3
070102010301020161.350.8111.192.8152.4242.51.1 × 10–20.4 0.2–0.8
07010201Xd81.830.5111.14313508.1939.91.2 × 10–180.10.05–0.18

aIn the UK/USA sample set we considered expected control genotype frequencies evaluated on the basis of the AFBAC haplotype frequencies using Hardy–Weinberg expectations on the assumption of sample size of 325 which is equal to half of the observed AFBAC haplotypes.

bOdds ratio and P-values have been calculated on the total data set comparing patients versus the control genotypes.

cDRB1*04-DQB1*0302 represents the sum of DRB1*0401-DQB1*0302, DRB1*0405-DQB1*0302, DRB1*0404-DQB1*0302 and DRB1*0402-DQB1*0302 haplotypes.

dX represents non-DR3 or DR4-DQB1*0302 haplotypes.

Table 3.
Open in new tab

Distribution of (DR2) DRB1*1501-DQB1*0602, (DR5) DRB1*11–12-DQB1*0301 and (DR7) DRB1*0701-DQB1*0201 positive genotypes in the Sardinian, UK and USA patients and controlsa

PatientsControls
GenotypeUK/USA (n = 445)Sardinia (n = 600)Total (n = 1045)UK/USAa (n = 325)Sardinia (n = 617)Total (n = 942)P < 0.05bORb95% CI
DRB1DQB1DRB1DQB1n%n%n%n%n%n%
15010602040302c20.40020.26.82.130.59.811.4 × 10–20.20.04–0.8
1501060203010201000000144.430.5171.81.3 × 10–50.050.01–0.4
15010602Xd10.20010.16620111.8778.12.1 × 10–200.010.00–0.08
11–120301040302c61.3111.8171.63.91.2111.8151.61.0 0.5–2.1
11–1203010301020120.450.870.78.12.5447.1525.52.0 × 10–100.1 0.1–0.3
11–120301Xd10.20010.1391213121170184.7 × 10–460.000.00–0.03
07010201040302c112.500111.14.31.361101.11.0 0.4–2.3
070102010301020161.350.8111.192.8152.4242.51.1 × 10–20.4 0.2–0.8
07010201Xd81.830.5111.14313508.1939.91.2 × 10–180.10.05–0.18
PatientsControls
GenotypeUK/USA (n = 445)Sardinia (n = 600)Total (n = 1045)UK/USAa (n = 325)Sardinia (n = 617)Total (n = 942)P < 0.05bORb95% CI
DRB1DQB1DRB1DQB1n%n%n%n%n%n%
15010602040302c20.40020.26.82.130.59.811.4 × 10–20.20.04–0.8
1501060203010201000000144.430.5171.81.3 × 10–50.050.01–0.4
15010602Xd10.20010.16620111.8778.12.1 × 10–200.010.00–0.08
11–120301040302c61.3111.8171.63.91.2111.8151.61.0 0.5–2.1
11–1203010301020120.450.870.78.12.5447.1525.52.0 × 10–100.1 0.1–0.3
11–120301Xd10.20010.1391213121170184.7 × 10–460.000.00–0.03
07010201040302c112.500111.14.31.361101.11.0 0.4–2.3
070102010301020161.350.8111.192.8152.4242.51.1 × 10–20.4 0.2–0.8
07010201Xd81.830.5111.14313508.1939.91.2 × 10–180.10.05–0.18

aIn the UK/USA sample set we considered expected control genotype frequencies evaluated on the basis of the AFBAC haplotype frequencies using Hardy–Weinberg expectations on the assumption of sample size of 325 which is equal to half of the observed AFBAC haplotypes.

bOdds ratio and P-values have been calculated on the total data set comparing patients versus the control genotypes.

cDRB1*04-DQB1*0302 represents the sum of DRB1*0401-DQB1*0302, DRB1*0405-DQB1*0302, DRB1*0404-DQB1*0302 and DRB1*0402-DQB1*0302 haplotypes.

dX represents non-DR3 or DR4-DQB1*0302 haplotypes.

Table 4.
Open in new tab

Comparison of DA/IA allotypes conferring predisposition to T1D in humans and mouse

DQ(α1*0301, β1*0302) (in DR4)DQ(α1*0501, β1*0201) (in DR3)IAg7 (NOD)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNEHRRFNEHRFNTHILF
Mass114129137156156147840114129137156147684114101137113113147726
pKa4.2612.512.535.24.2612.522.766
Charge=–+++==–++===+===
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Total
Amino acidFGLTEVYGNFGLSAVYGNFGLTEAYGN
Mass1475711310112999163571149811475711387719916357114909147571131011297116357114953
pKa4.210.114.310.110.14.210.114.3
Charge====–=================–====
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
Amino acidYFYYNNVLNYFSINNVLNHFYYTNAEN
Mass16314716316311411499113114119116314787113114114991131141065137147163163101114711291141140
pKa10.110.110.130.310.110.1610.110.14.230.4
Charge==================+======–=
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
Amino acidAWHNIRAWHNSRSYHNIR
Mass711861371141131567787118613711411315677887163137114113156771
pKa12.518.5612.518.510612.528.6
Charge==+==+==+==+==+==+
DQ(α1*0301, β1*0302) (in DR4)DQ(α1*0501, β1*0201) (in DR3)IAg7 (NOD)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNEHRRFNEHRFNTHILF
Mass114129137156156147840114129137156147684114101137113113147726
pKa4.2612.512.535.24.2612.522.766
Charge=–+++==–++===+===
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Total
Amino acidFGLTEVYGNFGLSAVYGNFGLTEAYGN
Mass1475711310112999163571149811475711387719916357114909147571131011297116357114953
pKa4.210.114.310.110.14.210.114.3
Charge====–=================–====
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
Amino acidYFYYNNVLNYFSINNVLNHFYYTNAEN
Mass16314716316311411499113114119116314787113114114991131141065137147163163101114711291141140
pKa10.110.110.130.310.110.1610.110.14.230.4
Charge==================+======–=
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
Amino acidAWHNIRAWHNSRSYHNIR
Mass711861371141131567787118613711411315677887163137114113156771
pKa12.518.5612.518.510612.528.6
Charge==+==+==+==+==+==+

Amino acid position, composition, mass, pKa and charge are shown for the individual constituents of pockets 1, 4, 6 and 9.The total mass and the sum of PKa of the residues forming each pocket are shown in bold type.

Table 4.
Open in new tab

Comparison of DA/IA allotypes conferring predisposition to T1D in humans and mouse

DQ(α1*0301, β1*0302) (in DR4)DQ(α1*0501, β1*0201) (in DR3)IAg7 (NOD)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNEHRRFNEHRFNTHILF
Mass114129137156156147840114129137156147684114101137113113147726
pKa4.2612.512.535.24.2612.522.766
Charge=–+++==–++===+===
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Total
Amino acidFGLTEVYGNFGLSAVYGNFGLTEAYGN
Mass1475711310112999163571149811475711387719916357114909147571131011297116357114953
pKa4.210.114.310.110.14.210.114.3
Charge====–=================–====
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
Amino acidYFYYNNVLNYFSINNVLNHFYYTNAEN
Mass16314716316311411499113114119116314787113114114991131141065137147163163101114711291141140
pKa10.110.110.130.310.110.1610.110.14.230.4
Charge==================+======–=
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
Amino acidAWHNIRAWHNSRSYHNIR
Mass711861371141131567787118613711411315677887163137114113156771
pKa12.518.5612.518.510612.528.6
Charge==+==+==+==+==+==+
DQ(α1*0301, β1*0302) (in DR4)DQ(α1*0501, β1*0201) (in DR3)IAg7 (NOD)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNEHRRFNEHRFNTHILF
Mass114129137156156147840114129137156147684114101137113113147726
pKa4.2612.512.535.24.2612.522.766
Charge=–+++==–++===+===
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Total
Amino acidFGLTEVYGNFGLSAVYGNFGLTEAYGN
Mass1475711310112999163571149811475711387719916357114909147571131011297116357114953
pKa4.210.114.310.110.14.210.114.3
Charge====–=================–====
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
Amino acidYFYYNNVLNYFSINNVLNHFYYTNAEN
Mass16314716316311411499113114119116314787113114114991131141065137147163163101114711291141140
pKa10.110.110.130.310.110.1610.110.14.230.4
Charge==================+======–=
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
Amino acidAWHNIRAWHNSRSYHNIR
Mass711861371141131567787118613711411315677887163137114113156771
pKa12.518.5612.518.510612.528.6
Charge==+==+==+==+==+==+

Amino acid position, composition, mass, pKa and charge are shown for the individual constituents of pockets 1, 4, 6 and 9.The total mass and the sum of PKa of the residues forming each pocket are shown in bold type.

Table 5.
Open in new tab

Comparison of DQ/IA allotypes conferring protection from T1D in humans and mouse

DQ(α1*0501, β1*0301) (IN DR5)DQ(α1*0102, β1*0602) (IN DR2)IAb (mouse)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNEHRFNAHGGFNPFASF
Mass114129137156147684114711375757147584114971477187147664
pKa4.2612.522.766
Charge=–++===+=========
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Total
Amino acidFAYTEVYGNFGLTEVCGNFGYTEVYGN
Size14771163101129991635711410441475711310112999103571149211475716310112999163571141030
pKa10.14.210.124.44.28.312.510.14.210.124.4
Charge====–========–========–====
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
Amino acidYFYYNNVLNFFYYNNVANYFYYSNVVN
Mass163147163163114114991131141191147147163163114114997111411331631471631638711499991141150
pKa10.110.110.130.310.110.120.210.110.110.130.3
Charge===========================
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
AADWHNSRDWHNIRDWHNVR
Size1151861371148715679611518613711411315682211518613711499156808
pKa3.9612.522.43.9612.522.43.9612.522.4
Charge–=+==+–=+==+–=+==+
DQ(α1*0501, β1*0301) (IN DR5)DQ(α1*0102, β1*0602) (IN DR2)IAb (mouse)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNEHRFNAHGGFNPFASF
Mass114129137156147684114711375757147584114971477187147664
pKa4.2612.522.766
Charge=–++===+=========
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Total
Amino acidFAYTEVYGNFGLTEVCGNFGYTEVYGN
Size14771163101129991635711410441475711310112999103571149211475716310112999163571141030
pKa10.14.210.124.44.28.312.510.14.210.124.4
Charge====–========–========–====
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
Amino acidYFYYNNVLNFFYYNNVANYFYYSNVVN
Mass163147163163114114991131141191147147163163114114997111411331631471631638711499991141150
pKa10.110.110.130.310.110.120.210.110.110.130.3
Charge===========================
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
AADWHNSRDWHNIRDWHNVR
Size1151861371148715679611518613711411315682211518613711499156808
pKa3.9612.522.43.9612.522.43.9612.522.4
Charge–=+==+–=+==+–=+==+

Amino acid position, composition, mass, pKa and charge are shown for the individual constituents of pockets 1, 4, 6 and 9. The total mass and the sum of PKa of the residues forming each pocket are shown in bold type.

Table 5.
Open in new tab

Comparison of DQ/IA allotypes conferring protection from T1D in humans and mouse

DQ(α1*0501, β1*0301) (IN DR5)DQ(α1*0102, β1*0602) (IN DR2)IAb (mouse)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNEHRFNAHGGFNPFASF
Mass114129137156147684114711375757147584114971477187147664
pKa4.2612.522.766
Charge=–++===+=========
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Total
Amino acidFAYTEVYGNFGLTEVCGNFGYTEVYGN
Size14771163101129991635711410441475711310112999103571149211475716310112999163571141030
pKa10.14.210.124.44.28.312.510.14.210.124.4
Charge====–========–========–====
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
Amino acidYFYYNNVLNFFYYNNVANYFYYSNVVN
Mass163147163163114114991131141191147147163163114114997111411331631471631638711499991141150
pKa10.110.110.130.310.110.120.210.110.110.130.3
Charge===========================
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
AADWHNSRDWHNIRDWHNVR
Size1151861371148715679611518613711411315682211518613711499156808
pKa3.9612.522.43.9612.522.43.9612.522.4
Charge–=+==+–=+==+–=+==+
DQ(α1*0501, β1*0301) (IN DR5)DQ(α1*0102, β1*0602) (IN DR2)IAb (mouse)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNEHRFNAHGGFNPFASF
Mass114129137156147684114711375757147584114971477187147664
pKa4.2612.522.766
Charge=–++===+=========
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α9aα62Total
Amino acidFAYTEVYGNFGLTEVCGNFGYTEVYGN
Size14771163101129991635711410441475711310112999103571149211475716310112999163571141030
pKa10.14.210.124.44.28.312.510.14.210.124.4
Charge====–========–========–====
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
Amino acidYFYYNNVLNFFYYNNVANYFYYSNVVN
Mass163147163163114114991131141191147147163163114114997111411331631471631638711499991141150
pKa10.110.110.130.310.110.120.210.110.110.130.3
Charge===========================
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
AADWHNSRDWHNIRDWHNVR
Size1151861371148715679611518613711411315682211518613711499156808
pKa3.9612.522.43.9612.522.43.9612.522.4
Charge–=+==+–=+==+–=+==+

Amino acid position, composition, mass, pKa and charge are shown for the individual constituents of pockets 1, 4, 6 and 9. The total mass and the sum of PKa of the residues forming each pocket are shown in bold type.

Table 6.
Open in new tab

Total mass of the residue forming P1, P4 and P9 pockets in T1D risk

DQ moleculesT1D riskP1 sizeDQ moleculesT1D riskP4 sizeDQ moleculesT1D riskP9 size
DQ (α1*0301, β1*0302) (in DR4)(+)839.96DQ (α1*0301, β1*0301) (in DR4)(±)1045.16DQ (α1*0401, β1*0402) (in DR8)(±)821.93
DQ(α1*0301, β1*0301) (in DR4)(±)839.96DQ (α1*0501, β1*0301) (in DR11)(–)1045.16DQ (α1*0301, β1*0301) (in DR4)(±)821.92
IAk(±)821.98IAk(±)1033.25IAk(±)821.92
IAg7 (NOD)(+)725.89IAb(–)1031.14DQ (α1*0102,. β1*0602) (in DR2)(–)821.92
DQ (α1*0501, β1*0201) (in DR3)(+)683.76DQ (α1*0301, β1*0302) (in DR4)(+)981.13DQ (α1*0104, β1*0503) (in DR14)(–)821.92
DQ(α1*0401, β1*0402) (in DR8)(±)683.76IAg7 (NOD)(+)953.07IAb(–)807.89
DQ (α1*0501, β1*0301) (in DR11)(–)683.76DQ (α1*0102, β1*0602) (in DR2)(–)921.09DQ (α1*0101, β1*0501) (in DR1)(±)805.98
IAb(–)663.75DQ (α1*0501, β1*0201) (in DR3)(+)909.06DQ (α1*0501, β1*0301) (in DR11)(–)795.83
DQ (α1*0101, β1*0501) (in DR1)(±)583.64DQ (α1*0401, β1*0402) (in DR8)(±)897.01DQ (α1*0301, β1*0302) (in DR4)(+)777.92
DQ (α1*0104, β1*0503) (in DR14)(–)583.64DQ (α1*0101, β1*0501) (in DR1)(±)836.97IAg7 (NOD)(+)770.89
DQ (α1*0102, β1*0602) (in DR2)(–)583.64DQ (α1*0104, β1*0503) (in DR14)(–)836.97DQ (α1*0501, β1*0201) (in DR3)(+)751.83
DQ moleculesT1D riskP1 sizeDQ moleculesT1D riskP4 sizeDQ moleculesT1D riskP9 size
DQ (α1*0301, β1*0302) (in DR4)(+)839.96DQ (α1*0301, β1*0301) (in DR4)(±)1045.16DQ (α1*0401, β1*0402) (in DR8)(±)821.93
DQ(α1*0301, β1*0301) (in DR4)(±)839.96DQ (α1*0501, β1*0301) (in DR11)(–)1045.16DQ (α1*0301, β1*0301) (in DR4)(±)821.92
IAk(±)821.98IAk(±)1033.25IAk(±)821.92
IAg7 (NOD)(+)725.89IAb(–)1031.14DQ (α1*0102,. β1*0602) (in DR2)(–)821.92
DQ (α1*0501, β1*0201) (in DR3)(+)683.76DQ (α1*0301, β1*0302) (in DR4)(+)981.13DQ (α1*0104, β1*0503) (in DR14)(–)821.92
DQ(α1*0401, β1*0402) (in DR8)(±)683.76IAg7 (NOD)(+)953.07IAb(–)807.89
DQ (α1*0501, β1*0301) (in DR11)(–)683.76DQ (α1*0102, β1*0602) (in DR2)(–)921.09DQ (α1*0101, β1*0501) (in DR1)(±)805.98
IAb(–)663.75DQ (α1*0501, β1*0201) (in DR3)(+)909.06DQ (α1*0501, β1*0301) (in DR11)(–)795.83
DQ (α1*0101, β1*0501) (in DR1)(±)583.64DQ (α1*0401, β1*0402) (in DR8)(±)897.01DQ (α1*0301, β1*0302) (in DR4)(+)777.92
DQ (α1*0104, β1*0503) (in DR14)(–)583.64DQ (α1*0101, β1*0501) (in DR1)(±)836.97IAg7 (NOD)(+)770.89
DQ (α1*0102, β1*0602) (in DR2)(–)583.64DQ (α1*0104, β1*0503) (in DR14)(–)836.97DQ (α1*0501, β1*0201) (in DR3)(+)751.83

Sizes of the pockets are indicated as the sum of the mass of the individual constituents.

(+), Very high risk; (–), very low risk; (±), intermediate risk for T1D.

The human and mouse susceptibility allotypes are indicated in bold type.

Table 6.
Open in new tab

Total mass of the residue forming P1, P4 and P9 pockets in T1D risk

DQ moleculesT1D riskP1 sizeDQ moleculesT1D riskP4 sizeDQ moleculesT1D riskP9 size
DQ (α1*0301, β1*0302) (in DR4)(+)839.96DQ (α1*0301, β1*0301) (in DR4)(±)1045.16DQ (α1*0401, β1*0402) (in DR8)(±)821.93
DQ(α1*0301, β1*0301) (in DR4)(±)839.96DQ (α1*0501, β1*0301) (in DR11)(–)1045.16DQ (α1*0301, β1*0301) (in DR4)(±)821.92
IAk(±)821.98IAk(±)1033.25IAk(±)821.92
IAg7 (NOD)(+)725.89IAb(–)1031.14DQ (α1*0102,. β1*0602) (in DR2)(–)821.92
DQ (α1*0501, β1*0201) (in DR3)(+)683.76DQ (α1*0301, β1*0302) (in DR4)(+)981.13DQ (α1*0104, β1*0503) (in DR14)(–)821.92
DQ(α1*0401, β1*0402) (in DR8)(±)683.76IAg7 (NOD)(+)953.07IAb(–)807.89
DQ (α1*0501, β1*0301) (in DR11)(–)683.76DQ (α1*0102, β1*0602) (in DR2)(–)921.09DQ (α1*0101, β1*0501) (in DR1)(±)805.98
IAb(–)663.75DQ (α1*0501, β1*0201) (in DR3)(+)909.06DQ (α1*0501, β1*0301) (in DR11)(–)795.83
DQ (α1*0101, β1*0501) (in DR1)(±)583.64DQ (α1*0401, β1*0402) (in DR8)(±)897.01DQ (α1*0301, β1*0302) (in DR4)(+)777.92
DQ (α1*0104, β1*0503) (in DR14)(–)583.64DQ (α1*0101, β1*0501) (in DR1)(±)836.97IAg7 (NOD)(+)770.89
DQ (α1*0102, β1*0602) (in DR2)(–)583.64DQ (α1*0104, β1*0503) (in DR14)(–)836.97DQ (α1*0501, β1*0201) (in DR3)(+)751.83
DQ moleculesT1D riskP1 sizeDQ moleculesT1D riskP4 sizeDQ moleculesT1D riskP9 size
DQ (α1*0301, β1*0302) (in DR4)(+)839.96DQ (α1*0301, β1*0301) (in DR4)(±)1045.16DQ (α1*0401, β1*0402) (in DR8)(±)821.93
DQ(α1*0301, β1*0301) (in DR4)(±)839.96DQ (α1*0501, β1*0301) (in DR11)(–)1045.16DQ (α1*0301, β1*0301) (in DR4)(±)821.92
IAk(±)821.98IAk(±)1033.25IAk(±)821.92
IAg7 (NOD)(+)725.89IAb(–)1031.14DQ (α1*0102,. β1*0602) (in DR2)(–)821.92
DQ (α1*0501, β1*0201) (in DR3)(+)683.76DQ (α1*0301, β1*0302) (in DR4)(+)981.13DQ (α1*0104, β1*0503) (in DR14)(–)821.92
DQ(α1*0401, β1*0402) (in DR8)(±)683.76IAg7 (NOD)(+)953.07IAb(–)807.89
DQ (α1*0501, β1*0301) (in DR11)(–)683.76DQ (α1*0102, β1*0602) (in DR2)(–)921.09DQ (α1*0101, β1*0501) (in DR1)(±)805.98
IAb(–)663.75DQ (α1*0501, β1*0201) (in DR3)(+)909.06DQ (α1*0501, β1*0301) (in DR11)(–)795.83
DQ (α1*0101, β1*0501) (in DR1)(±)583.64DQ (α1*0401, β1*0402) (in DR8)(±)897.01DQ (α1*0301, β1*0302) (in DR4)(+)777.92
DQ (α1*0104, β1*0503) (in DR14)(–)583.64DQ (α1*0101, β1*0501) (in DR1)(±)836.97IAg7 (NOD)(+)770.89
DQ (α1*0102, β1*0602) (in DR2)(–)583.64DQ (α1*0104, β1*0503) (in DR14)(–)836.97DQ (α1*0501, β1*0201) (in DR3)(+)751.83

Sizes of the pockets are indicated as the sum of the mass of the individual constituents.

(+), Very high risk; (–), very low risk; (±), intermediate risk for T1D.

The human and mouse susceptibility allotypes are indicated in bold type.

Table 7.
Open in new tab

Comparison of MHC class II alleles conferring protection from T1D in humans and mouse

IAbDR(α1*0101, β1*0403)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNPFASFNVFASF
Mass114971477187147664114991477187147666
pKa
Charge============
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α62Total
Amino acidFGYTEVYGNVHFDEYQN
Mass1475716310112999163571141030991371471151291631141141018
pKa10.14.210.124.463.94.210.124.2
Charge====–=====+=––===
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
AminoacidYFYYSNVVNEVYYENVDN
Mass163147163163871149999114115012999163163129114991151141126
pKa10.110.110.130.34.210.110.14.23.932.5
Charge=========–===–==–=
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
AminoacidDWHNVRDWANIR
Mass1151861371149915680811518671114113156756
pKa3.9612.522.43.912.516.4
Charge–=+==+–====+
IAbDR(α1*0101, β1*0403)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNPFASFNVFASF
Mass114971477187147664114991477187147666
pKa
Charge============
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α62Total
Amino acidFGYTEVYGNVHFDEYQN
Mass1475716310112999163571141030991371471151291631141141018
pKa10.14.210.124.463.94.210.124.2
Charge====–=====+=––===
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
AminoacidYFYYSNVVNEVYYENVDN
Mass163147163163871149999114115012999163163129114991151141126
pKa10.110.110.130.34.210.110.14.23.932.5
Charge=========–===–==–=
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
AminoacidDWHNVRDWANIR
Mass1151861371149915680811518671114113156756
pKa3.9612.522.43.912.516.4
Charge–=+==+–====+

Amino acid position, composition, mass, pKa and charge are shown for the individual constituents of pockets 1, 4, 6 and 9. The total mass and the sum of PKa of the residues forming each pocket are shown in bold type.

Table 7.
Open in new tab

Comparison of MHC class II alleles conferring protection from T1D in humans and mouse

IAbDR(α1*0101, β1*0403)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNPFASFNVFASF
Mass114971477187147664114991477187147666
pKa
Charge============
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α62Total
Amino acidFGYTEVYGNVHFDEYQN
Mass1475716310112999163571141030991371471151291631141141018
pKa10.14.210.124.463.94.210.124.2
Charge====–=====+=––===
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
AminoacidYFYYSNVVNEVYYENVDN
Mass163147163163871149999114115012999163163129114991151141126
pKa10.110.110.130.34.210.110.14.23.932.5
Charge=========–===–==–=
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
AminoacidDWHNVRDWANIR
Mass1151861371149915680811518671114113156756
pKa3.9612.522.43.912.516.4
Charge–=+==+–====+
IAbDR(α1*0101, β1*0403)
P1β82β86α24α52α53α54Totalβ82β86α24α52α53α54Total
Amino acidNPFASFNVFASF
Mass114971477187147664114991477187147666
pKa
Charge============
P4β11β13β26β28β74β78α9α9aα62Totalβ11β13β26β28β74β78α9α62Total
Amino acidFGYTEVYGNVHFDEYQN
Mass1475716310112999163571141030991371471151291631141141018
pKa10.14.210.124.463.94.210.124.2
Charge====–=====+=––===
P6β9β11β30β37α11α62α65α66α69Totalβ9β11β30β37α11α62α65α66α69Total
AminoacidYFYYSNVVNEVYYENVDN
Mass163147163163871149999114115012999163163129114991151141126
pKa10.110.110.130.34.210.110.14.23.932.5
Charge=========–===–==–=
P9β57β61α68α69α72α76Totalβ57β61α68α69α72α76Total
AminoacidDWHNVRDWANIR
Mass1151861371149915680811518671114113156756
pKa3.9612.522.43.912.516.4
Charge–=+==+–====+

Amino acid position, composition, mass, pKa and charge are shown for the individual constituents of pockets 1, 4, 6 and 9. The total mass and the sum of PKa of the residues forming each pocket are shown in bold type.

1 Risch, N. (

1987
) Assessing the role of HLA-linked and unlinked determinants of disease.
Am. J. Hum. Genet.
,
40
,
1
–14.

2 Todd, J.A. (

1995
) Genetic analysis of type 1 diabetes using whole genome approaches.
Proc. Natl Acad. Sci. USA
,
92
,
8560
–8565.

3 Cucca, F. and Todd, J. (

1996
) HLA/MHC: Genes, Molecules and Function. BIOS Scientific Publishers, Oxford, UK.

4 Noble, A.J., Valdes, A.M., Cook, M., Klitz, W., Thomson, G. and Erlich, H.A. (

1996
) The role of HLA class II genes in insulin-dependent diabetes mellitus: molecular analysis of 180 Caucasian, multiplex families.
Am. J. Hum. Genet.
,
59
,
1134
–1148.

5 Sheehy, M.J. (

1992
) HLA and insulin-dependent diabetes. A protective perspective.
Diabetes
,
41
,
123
–129.

6 Tisch, R. and McDevitt, H. (

1996
) Insulin-dependent diabetes mellitus.
Cell
,
85
,
291
–297.

7 Nepom, G.T. and Kwok, W.W. (

1998
) Molecular basis for HLA-DQ associations with IDDM.
Diabetes
,
47
,
1177
–1184.

8 Sollid, L.M. (

2000
) Molecular basis of celiac disease.
Annu. Rev. Immunol.
,
18
,
53
–81.

9 Todd, J.A., Bell, J.I. and McDevitt, H.O. (

1987
) HLA-DQ beta gene contributes to susceptibility and resistance to insulin-dependent diabetes mellitus.
Nature
,
329
,
599
–604.

10 Nepom, G.T. (

1990
) A unified hypothesis for the complex genetics of HLA associations with IDDM.
Diabetes
,
39
,
1153
–1157.

11 Sanjeevi, C.B., Lybrand, T.P., DeWeese, C., Landin-Olsson, M., Kockum, I., Dahlquist, G., Sundkvist, G., Stenger, D. and Lernmark, A. (

1995
) Polymorphic amino acid variations in HLA-DQ are associated with systematic physical property changes and occurrence of IDDM. Members of the Swedish Childhood Diabetes Study.
Diabetes
,
44
,
125
–131.

12 Hoover, M.L. and Marta, R.T. (

1997
) Molecular modelling of HLA-DQ suggests a mechanism of resistance in type 1 diabetes.
Scand.
J. Immunol
.,
45
,
193
–202.

13 Schmidt, D., Verdaguer, J., Averill, N. and Santamaria, P. (

1997
) A mechanism for the major histocompatibility complex-linked resistance to autoimmunity.
J. Exp. Med.
,
186
,
1059
–1075.

14 Ridgway, W.M., Ito, H., Fasso, M., Yu, C. and Fathman, C.G. (

1998
) Analysis of the role of variation of major histocompatibility complex class II expression on nonobese diabetic (NOD) peripheral T cell response.
J. Exp. Med.
,
188
,
2267
–2275.

15 Jordan, M.S., Boesteanu, A., Reed, A.J., Petrone, A.L., Holenbeck, A.E., Lerman, M.A., Naji, A. and Caton, A.J. (

2001
) Thymic selection of CD4+CD25+ regulatory T cells induced by an agonist self-peptide.
Nat. Immunol.
,
2
,
301
–306.

16 Fremont, D.H., Monnaie, D., Nelson, C.A., Hendrickson, W.A. and Unanue, E.R. (

1998
) Crystal structure of I-Ak in complex with a dominant epitope of lysozyme.
Immunity
,
8
,
305
–317.

17 Scott, C.A., Peterson, P.A., Teyton, L. and Wilson, I.A. (

1998
) Crystal structures of two I-Ad-peptide complexes reveal that high affinity can be achieved without large anchor residues.
Immunity
,
8
,
319
–329.

18 Corper, A.L., Stratmann, T., Apostolopoulos, V., Scott, C.A., Garcia, K.C., Kang, A.S., Wilson, I.A. and Teyton, L. (

2000
) A structural framework for deciphering the link between I-Ag7 and autoimmune diabetes.
Science
,
288
,
505
–511.

19 Latek, R.R., Suri, A., Petzold, S.J., Nelson, C.A., Kanagawa, O., Unanue, E.R. and Fremont, D.H. (

2000
) Structural basis of peptide binding and presentation by the type I diabetes-associated MHC class II molecule of NOD mice.
Immunity
,
12
,
699
–710.

20 Khalil, I., d′Auriol, L., Gobet, M., Morin, L., Lepage, V., Deschamps, I., Park, M.S., Degos, L., Galibert, F. and Hors, J. (

1990
) A combination of HLA-DQ beta Asp57-negative and HLA DQ alpha Arg52 confers susceptibility to insulin-dependent diabetes mellitus.
J. Clin. Invest.
,
85
,
1315
–1319.

21 Arentz-Hansen, H., Korner, R., Molberg, O., Quarsten, H., Vader, W., Kooy, Y.M., Lundin, K.E., Koning, F., Roepstorff, P., Sollid, L.M. and McAdam, S.N. (

2000
) The intestinal T cell response to alpha-gliadin in adult celiac disease is focused on a single deamidated glutamine targeted by tissue transglutaminase.
J. Exp. Med.
,
191
,
603
–612.

22 Vartdal, F., Johansen, B.H., Friede, T., Thorpe, C.J., Stevanovic, S., Eriksen, J.E., Sletten, K., Thorsby, E., Rammensee, H.G. and Sollid, L.M. (

1996
) The peptide binding motif of the disease associated HLA-DQ (alpha 1* 0501, beta 1* 0201) molecule.
Eur. J. Immunol.
,
26
,
2764
–2772.

23 Kwok, W.W., Domeier, M.L., Raymond, F.C., Byers, P. and Nepom, G.T. (

1996
) Allele-specific motifs characterize HLA-DQ interactions with a diabetes-associated peptide derived from glutamic acid decarboxylase.
J. Immunol
.,
156
,
2171
–2177.

24 Singer, S.M., Tisch, R., Yang, X.D., Sytwu, H.K., Liblau, R. and McDevitt, H.O. (

1998
) Prevention of diabetes in NOD mice by a mutated I-Ab transgene.
Diabetes
,
47
,
1570
–1577.

25 Dessen, A., Lawrence, C.M., Cupo, S., Zaller, D.M. and Wiley, D.C. (

1997
) X-ray crystal structure of HLA-DR4 (DRA*0101, DRB1*0401) complexed with a peptide from human collagen II.
Immunity
,
7
,
473
–481.

26 Busch, R., Hill, C.M., Hayball, J.D., Lamb, J.R. and Rothbard, J.B. (

1991
) Effect of natural polymorphism at residue 86 of the HLA-DR beta chain on peptide binding.
J. Immunol
.,
147
,
1292
–1298.

27 Nishimoto, H., Kikutani, H., Yamamura, K. and Kishimoto, T. (

1987
) Prevention of autoimmune insulitis by expression of I-E molecules in NOD mice.
Nature
,
328
,
432
–434.

28 Bain, S.C., Todd, J.A. and Barnett, A.H. (

1990
) The British Diabetic Association – Warren Repository.
Autoimmunity
,
7
,
83
–85.

29 Cucca, F., Muntoni, F., Lampis, R., Frau, F., Argiolas, L., Silvetti, M., Angius, E., Cao, A., De Virgiliis, S. and Congia, M. (

1993
) Combinations of specific DRB1, DQA1, DQB1 haplotypes are associated with insulin-dependent diabetes mellitus in Sardinia.
Hum. Immunol.
,
37
,
85
–94.

30 Lernmark, A., Ducat, L., Eisenbarth, G., Ott, J., Permutt, M.A., Rubenstein, P. and Spielman, R. (

1990
) Family cell lines available for research.
Am. J. Hum. Genet.
,
47
,
1028
–1030.

31 Bjorkman, P.J., Saper, M.A., Samraoui, B., Bennett, W.S., Strominger, J.L. and Wiley, D.C. (

1987
) Structure of the human class I histocompatibility antigen, HLA-A2.
Nature
,
329
,
506
–512.

32 Erlich, H.A., Bugawan, T.L., Scharf, S., Nepom, G.T., Tait, B. and Griffith, R.L. (

1990
) HLA-DQ beta sequence polymorphism and genetic susceptibility to IDDM.
Diabetes
,
39
,
96
–103.

33 Sheehy, M.J., Scharf, S.J., Rowe, J.R., Neme de Gimenez, M.H., Meske, L.M., Erlich, H.A. and Nepom, B.S. (

1989
) A diabetes- susceptible HLA haplotype is best defined by a combination of HLA-DR and -DQ alleles.
J. Clin. Invest.
,
83
,
830
–835.

34 Erlich, H.A., Zeidler, A., Chang, J., Shaw, S., Raffel, L.J., Klitz, W., Beshkov, Y., Costin, G., Pressman, S. and Bugawan, T. (

1993
) HLA class II alleles and susceptibility and resistance to insulin dependent diabetes mellitus in Mexican-American families.
Nat. Genet.
,
3
,
358
–364.

35 Cucca, F., Lampis, R., Frau, F., Macis, D., Angius, E., Masile, P., Chessa, M., Frongia, P., Silvetti, M., Cao, A. et al. (

1995
) The distribution of DR4 haplotypes in Sardinia suggests a primary association of type I diabetes with DRB1 and DQB1 loci.
Hum. Immunol.
,
43
,
301
–308.

36 Lee, K.H., Wucherpfennig, K.W. and Wiley, D.C. (

2001
) Structure of a human insulin peptide-HLA-DQ8 complex and susceptibility to type 1 diabetes.
Nat. Immunol.
,
2
,
501
–507.

37 Hattori, M., Buse, J.B., Jackson, R.A., Glimcher, L., Dorf, M.E., Minami, M., Makino, S., Moriwaki, K., Kuzuya, H., Imura, H. et al. (

1986
) The NOD mouse: recessive diabetogenic gene in the major histocompatibility complex.
Science
,
231
,
733
–735.

38 Acha-Orbea, H. and McDevitt, H.O. (

1987
) The first external domain of the nonobese diabetic mouse class II I-A beta chain is unique.
Proc. Natl Acad. Sci. USA
,
84
,
2435
–2439.

39 Lund, T., O′Reilly, L., Hutchings, P., Kanagawa, O., Simpson, E., Gravely, R., Chandler, P., Dyson, J., Picard, J.K., Edwards, A. et al. (

1990
) Prevention of insulin-dependent diabetes mellitus in non-obese diabetic mice by transgenes encoding modified I-A beta-chain or normal I-E alpha-chain.
Nature
,
345
,
727
–729.

40 Miyazaki, T., Uno, M., Uehira, M., Kikutani, H., Kishimoto, T., Kimoto, M., Nishimoto, H., Miyazaki, J. and Yamamura, K. (

1990
) Direct evidence for the contribution of the unique I-A NOD to the development of insulitis in non-obese diabetic mice.
Nature
,
345
,
722
–724.

41 Slattery, R.M., Kjer-Nielsen, L., Allison, J., Charlton, B., Mandel, T.E. and Miller, J.F. (

1990
) Prevention of diabetes in non-obese diabetic I-Ak transgenic mice.
Nature
,
345
,
724
–726.

42 Singer, S.M., Tisch, R., Yang, X.D. and McDevitt, H.O. (

1993
) An Abd transgene prevents diabetes in nonobese diabetic mice by inducing regulatory T cells.
Proc. Natl Acad. Sci. USA
,
90
,
9566
–9570.

43 Luhder, F., Katz, J., Benoist, C. and Mathis, D. (

1998
) Major histocompatibility complex class II molecules can protect from diabetes by positively selecting T cells with additional specificities.
J. Exp. Med.
,
187
,
379
–387.

44 Pugliese, A., Brown, D., Garza, D., Murchison, D., Zeller, M., Redondo, M., Diez, J., Eisenbarth, G.S., Patel, D.D. and Ricordi, C. (

2001
) Self-antigen-presenting cells expressing diabetes-associated autoantigens exist in both thymus and peripheral lymphoid organs.
J. Clin. Invest.
,
107
,
555
–564.

45 Kanagawa, O., Martin, S.M., Vaupel, B.A., Carrasco-Marin, E. and Unanue, E.R. (

1998
) Autoreactivity of T cells from nonobese diabetic mice: an I-Ag7-dependent reaction.
Proc. Natl Acad. Sci. USA
,
95
,
1721
–1724.

46 Hanahan, D. (

1998
) Peripheral-antigen-expressing cells in thymic medulla: factors in self-tolerance and autoimmunity.
Curr. Opin. Immunol.
,
10
,
656
–662.

47 Pugliese, A., Zeller, M., Fernandez, A.,Jr, Zalcberg, L.J., Bartlett, R.J., Ricordi, C., Pietropaolo, M., Eisenbarth, G.S., Bennett, S.T. and Patel, D.D. (

1997
) The insulin gene is transcribed in the human thymus and transcription levels correlated with allelic variation at the INS VNTR-IDDM2 susceptibility locus for type 1 diabetes.
Nat. Genet.
,
15
,
293
–297.

48 Vafiadis, P., Bennett, S.T., Todd, J.A., Nadeau, J., Grabs, R., Goodyer, C.G., Wickramasinghe, S., Colle, E. and Polychronakos, C. (

1997
) Insulin expression in human thymus is modulated by INS VNTR alleles at the IDDM2 locus.
Nat. Genet.
,
15
,
289
–292.

49 Alleva, D.G., Crowe, P.D., Jin, L., Kwok, W.W., Ling, N., Gottschalk, M., Conlon, P.J., Gottlieb, P.A., Putnam, A.L. and Gaur, A. (

2001
) A disease-associated cellular immune response in type 1 diabetics to an immunodominant epitope of insulin.
J. Clin. Invest.
,
107
,
173
–180.

50 Daniel, D. and Wegmann, D.R. (

1996
) Protection of nonobese diabetic mice from diabetes by intranasal or subcutaneous administration of insulin peptide B-(9–23).
Proc. Natl Acad. Sci. USA
,
93
,
956
–960.

51 Wong, F.S., Karttunen, J., Dumont, C., Wen, L., Visintin, I., Pilip, I.M., Shastri, N., Pamer, E.G. and Janeway, C.A.,Jr (

1999
) Identification of an MHC class I-restricted autoantigen in type 1 diabetes by screening an organ-specific cDNA library.
Nat. Med.
,
5
,
1026
–1031.

52 Undlien, D.E., Kockum, I., Ronningen, K.S., Lowe, R., Saanjeevi, C.B., Graham, J., Lie, B.A., Akselsen, H.E., Lernmark, A. and Thorsby, E. (

1999
) HLA associations in type 1 diabetes among patients not carrying high risk DR3-DQ2 or DR4-DQ8 haplotypes.
Tissue Antigens
,
54
,
543
–551.

53 Todd, J.A., Fukui, Y., Kitagawa, T. and Sasazuki, T. (

1990
) The A3 allele of the HLA-DQA1 locus is associated with susceptibility to type 1 diabetes in Japanese.
Proc. Natl Acad. Sci. USA
,
87
,
1094
–1098.

54 Risch, N. (

1989
) Genetics of IDDM: evidence for complex inheritance with HLA.
Genet. Epidemiol.
,
6
,
143
–148.

55 Horn, G.T., Bugawan, T.L., Long, C.M. and Erlich, H.A. (

1988
) Allelic sequence variation of the HLA-DQ loci: relationship to serology and to insulin-dependent diabetes susceptibility.
Proc. Natl Acad. Sci. USA
,
85
,
6012
–6016.

56 Wordsworth, B.P., Allsopp, C.E., Young, R.P. and Bell, J.I. (

1990
) HLA-DR typing using DNA amplification by the polymerase chain reaction and sequential hybridization to sequence-specific oligonucleotide probes.
Immunogenetics
,
32
,
413
–418.

57 Bugawan, T.L. and Erlich, H.A. (

1991
) Rapid typing of HLA-DQB1 DNA polymorphism using nonradioactive oligonucleotide probes and amplified DNA.
Immunogenetics
,
33
,
163
–170.

58 Cucca, F., Frau, F., Lampis, R., Floris, M., Argiolas, L., Macis, D., Cao, A., De Virgiliis, S. and Congia, M. (

1994
) HLA-DQB1*0305 and -DQB1*0304 alleles among Sardinians. Evolutionary and practical implications for oligotyping.
Hum. Immunol.
,
40
,
143
–149.

59 Petersdorf, E.W., Smith, A.G., Mickelson, E.M., Martin, P.J. and Hansen, J.A. (

1991
) Ten HLA-DR4 alleles defined by sequence polymorphisms within the DRB1 first domain.
Immunogenetics
,
33
,
267
–275.

60 Bunce, M., Barnardo, M.C. and Welsh, K.I. (

1998
) The PCR-SSP Manager computer program: a tool for maintaining sequence alignments and automatically updating the specificities of PCR-SSP primers and primer mixes.
Tissue Antigens
,
52
,
158
–174.

61 Spielman, R.S., McGinnis, R.E. and Ewens, W.J. (

1993
) Transmission test for linkage disequilibrium: the insulin gene region and insulin-dependent diabetes mellitus (IDDM).
Am. J. Hum. Genet.
,
52
,
506
–516.

62 Cucca, F., Dudbridge, F., Loddo, M., Mulargia, A.P., Lampis, R., Angius, E., De Virgiliis, S., Koeleman, B.P., Bain, S.C., Barnett, A.H. et al. (

2001
) The HLA-DPB1–associated component of the IDDM1 and its relationship to the major loci HLA-DQB1, -DQA1, and -DRB1.
Diabetes
,
50
,
1200
–1205.

63 Valdes, A.M. and Thomson, G. (

1997
) Detecting disease-predisposing variants: the haplotype method.
Am. J. Hum. Genet.
,
60
,
703
–716.

64 Lampis, R., Morelli, L., Congia, M., Macis, M.D., Mulargia, A., Loddo, M., De Virgiliis, S., Marrosu, M.G., Todd, J.A. and Cucca, F. (

2000
) The inter-regional distribution of HLA class II haplotypes indicates the suitability of the Sardinian population for case-control association studies in complex diseases.
Hum. Mol. Genet.
,
9
,
2959
–2965.

65 Thomson, G. (

1995
) Mapping disease genes: family-based association studies.
Am. J. Hum. Genet.
,
57
,
487
–498.

66 Fremont, D.H., Hendrickson, W.A., Marrack, P. and Kappler, J. (

1996
) Structures of an MHC class II molecule with covalently bound single peptides.
Science
,
272
,
1001
–1004.

Topic:
Issue Section:
Article
Download all slides