Genetic variants of the HLA class II region (DRB1, DQB1 and DQA1) and CTLA-4 contribute to susceptibility to Graves' disease (GD). Whilst disease susceptibility has been mapped to a non-coding 6.1kb 3' region of CTLA-4, the primary aetiological variants within the HLA class II region remain unknown. The aims of this study were (i) determine which of the three HLA class II loci account for the primary association with GD and (ii) examine disease associations of specific polymorphic amino acids encoded by the most associated locus. In total, 905 patients with GD and 864 controls were genotyped at the DRB1, DQB1 and DQA1 loci. All subjects gave informed written consent, and the project approved by the local ethics committee. A significant increase of the DRB1*03-DQB1*02-DQA1*0501 haplotype was observed in GD patients compared to controls (P=7.2x10-12, OR=1.9 [95% CI = 1.58-2.28]). Stepwise logistic regression analysis showed the DRB1 locus alone (P=9.1x10-14) was sufficient to explain the association of the HLA class II region, as neither DQB1 nor DQA1 improved the model (P=0.149 and P=0.222, respectively). An amino acid map of exon two domain of DRB1 for each subject was analysed using stepwise logistic regression (Cordell & Clayton, 2002). Of 102 amino acids, 70 were non-polymorphic and 20 were not associated with disease, leaving 12 associated with GD (P=0.0003-9.3x10-10). Seven of the twelve associated amino acids (11, 13, 26, 28, 30, 47, 71) are within the P4-P7 peptide residue positional environments of the binding pockets with four occurring within P5 (11, 13, 30, 71), which appears to play a central role in T cell receptor recognition. Changes in amino acid charge and size within the peptide binding pockets of DRB1, could lead to increased auto-antigen presentation and binding, perhaps changing the autoimmune T cell repertoire, potentially increasing the risk of developing GD.
22 - 24 Mar 2004
British Endocrine Societies