Problems of identifying independent non-class II susceptibility loci within the HLA region for Graves disease
MJ Simmonds, JM Heward, JA Franklyn & SCL Gough
The HLA class II region, in particular DRB1/DQA1/DQB1, has been consistently associated with Graves disease (GD) for over thirty years. Only recently has work within our own group made progress in narrowing down the etiological variant(s) present within DRB1/DQA1/DQB1, by excluding DQB1, and by mapping association at DRB1 to nine amino acid positions present within the peptide binding domain, with position β74 being the most associated. Independent associations outside of the HLA class II region have been proposed although only recently have data been reported to support these associations. BTNL2, a co-stimulatory surface molecule involved in T cell activation, has been proposed as one such factor. Studies in patients with sarcoidosis have shown association of the A allele of the BTNL2 rs2076530 G-A single nucleotide polymorphism (SNP), with the A allele further associated with disruption of BTNL2 membrane localization. The aim of this study was to investigate rs2076530 within a large UK Caucasian GD dataset. In total 864 patients with GD and 864 control subjects were genotyped. All subjects gave informed written consent, and the project was approved by the local ethics committee. Association was found between GD and the A allele of rs2076530 (OR=1.32, 95% CI=1.141.52). When linkage disequilibrium (LD) was assessed between rs2076530 and DRB1 position β74, a D′=0.81 was generated, suggesting strong LD between these loci and lack of an independent effect. To be able to completely rule out an independent contribution to GD, mapping of all Tag SNPs within this region is needed, combined with logistic regression analysis. These data once again highlight the difficulties in detecting class II independent effects at HLA because of strong LD within this gene region.