Distributions of HLA class II alleles in autoantibody subsets among Norwegian patients with systemic lupus erythematosus.

In order to find potential correlations between HLA class II alleles and anti-SS-A, -SS-B, -Sm and anti-snRNP responses among Norwegian patients with systemic lupus erythematosus (SLE), HLA-DRB1, -DRB3*0101, -DQA1 and -DQB1 alleles were determined by DNA typing 50 patients and 108 controls. HLA distributions were analysed in the following autoantibody subgroups: anti-SS-A with -SS-B, anti-SS-A without -SS-B, anti-snRNP without -Sm, anti-SS-A without -snRNP and anti-snRNP without -SS-A. The autoantibodies were detected by EIA (enzyme immunuassay). Patients with anti-SS-A and -SS-B had significantly increased frequencies of DRB1*03, DRB3*0101, DQA1*0501, DQB1*0201 (in linkage disequilibrium) versus controls and versus patients without anti-SS-A and -SS-B. No differences in HLA distribution were found when the group with anti-SS-A alone was compared to the group with anti-SS-A and concomitant -SS-B. Comparing the groups with and without anti-SS-A and -SS-B, the highest RR were found for the alleles DRB1*03, DRB3*0101, DQB1*0501, DQB1*0201 (in linkage disequilibrium) with RR: 16.8, 5.0, 19.6, 10.3, respectively, P < 0.05). RR for DQw2/DQw6 heterozygotes was 3.5 (Ns.), and RR for cases having DQ alpha molecules with glutamine in position 34 and DQ beta molecules with leucine in position 26 on both chains was 6.3 (P < 0.05). No HLA associations were observed in the group with anti-snRNP without concomitant -Sm or without concomitant -SS-A. These results show that production of anti-SS-A and -SS-B is associated to the HLA alleles DRB1*03, DRB3*0101, DQA1*0501, DQB1*0201, and that this haplotype shows stronger correlation to these responses than DQw2/DQw6 heterozygosity or HLA molecules having glutamine in position 34 (DQ alpha) and leucine in position 26 (DQ beta). The failure to observe any correlation with DRBI*15,16 (DR2) in the group with anti-SS-A alone may demonstrate ethnic differences concerning this response. The failure to identify any HLA associations for the anti-snRNP response may reflect the heterogeneity of the molecules that constitute this antigen.