The influence of human leucocyte antigen (HLA) genes on autoimmune thyroid disease (AITD): Results of studies in HLA-DR3 positive AITD families

OBJECTIVE: Population-based, case-control studies have consistently shown association of Graves' disease (GD) with human leucocyte antigen (HLA)-DR3 in Caucasian populations. HLA association studies in Hashimoto's thyroiditis (HT) have also suggested an association with DR3, as well as with other HLA alleles. In contrast, HLA linkage studies in autoimmune thyroid disease (AITD) have been largely negative. The aim of the present study was to investigate the role of HLA in AITD and to explain the observed associations, but lack of linkage, by examining only AITD families with the associated allele, DR3. PATIENTS: We studied 99 probands (60 with GD and 39 with HT) from 99 multiplex, multigenerational Caucasian AITD families, and 135 age- and sex-matched Caucasian controls in association studies. In addition, a dataset of 34 Caucasian AITD families (out of the 99 families) with HLA-DR3 positive probands were analysed in linkage studies. DESIGN: HLA typing was performed using the technique of group-specific polymerase chain reaction-amplification with restriction enzyme digestion. Whole genome screening was performed using the ABI microsatellite panels. For fine mapping of the HLA region, we used the following markers: D6S276, D6S464, D6S439, D6S273, tumour necrosis factor α and D6S1610. LOD scores were calculated using the LIPED and GeneHunter programs. RESULTS: Case-control association analyses using the probands from our 99 Caucasian families showed an association of GD with DRB1*03 [P=0.00032, relative risk (RR)=3.4]. Linkage analysis for the HLA region in the 34 DR3 positive AITD families showed negative LOD scores throughout the region. The two-point LOD score at marker D6S273 (the closest to HLA-DRB1) was -3.0, and the multipoint LOD score was -7.6, demonstrating strong evidence against linkage to the HLA region in the subset of DR3 positive families. Whole genome screening in the subset of 34 DR3 positive families revealed one locus showing evidence for linkage to AITD: D3S1580 on chromosome 3q27 with a maximum two-point LOD score of 2.1. CONCLUSIONS: The HLA locus did not cosegregate with disease in DR3 positive families, suggesting that HLA genes are not major genes for AITD expression even within DR3 positive families; Hence, although HLA-DR3 was associated with GD in the probands, it was most likely a modulating gene and not causative; and, as the DR3 positive families showed evidence for linkage with D3S1580, it may imply that the DR3 gene modulated the effect of a susceptibility gene within the D3S1580 locus.