Linkage disequilibrium of HLA-DR3 and HLA-DR4 with HLA-B alleles in Mexican patients with rheumatoid arthritis.

Objective: We studied the gene frequencies of classes I, II and III antigens of the Major Histocompatibility Complex (MHC) in 32 Mexican mestizo patients with rheumatoid arthritis (RA) and compared them with those obtained from 110 of their first degree relatives and 100 Mexican mestizo controls. Furthermore, we analyzed the observed and expected frequencies of the haplotypes and calculated the delta values in the three groups.
Methods: The class I and class II MHC antigens were determined by the microlymphocytotoxicity test; class III MHC antigens were obtained by high voltage agarose gel electrophoresis and immunofixation. The significance of differences among the three groups was tested by chi-square analysis; linkage disequilibrium among the different alleles in each haplotype was estimated by computing the delta values (observed vs expected frequencies).
Results: Patients showed a significantly increased frequency of HLA-A1 (corrected p = 10(-5)), DR3 (corrected p = 0.04) and DQ2 (corrected p = 10(-4)) and a decreased frequency of A31 (corrected p = 0.003) as compared to the normal controls. First degree relatives compared to patients and controls showed a decreased frequency of HLA-DR4 (corrected p = 0.02 and 0.008 respectively); consequently, DQ3 was also diminished (corrected p = 10(-4)). Analysis of MHC haplotypes within families revealed in the patients seven MHC haplotypes with significant differences between the observed and expected frequencies (statistically significant delta values). These haplotypes were: [HLA-B8; DR3], [HLAB44; FC31], [HLA-B8; SC42], [HLA-DR4; SC31], [HLA-B35; SC32], [HLA-DR7; FC31] and [HLA-DR2; SC31]. On the other hand, the control haplotypes showed significant delta values in only one of these haplotypes ([HLA-DR4; SC31]), whereas first degree relatives showed none. Analysis of all the class I, II and III alleles, either alone or as part of specific haplotypes, showed two B-DR haplotypes with higher relative risks than their alleles alone. These haplotypes were: [B44; DR4] (RR = 6.0, p = 0.005), and [B8; DR3] (RR = 8.3, p = 0.010).
Conclusion: The results suggest that a specific combination of antigens in the same haplotype (for instance between HLA-B and HLA-DR) could contribute to increasing the genetic susceptibility to develop RA.