Soluble HLA-G in rheumatoid arthritis.

We investigated potential correlations between soluble HLA-G (sHLA-G) and soluble HLA class I (sHLA-I) levels, respectively, and parameters of disease activity or genetic factors determined by HLA-DRB1 and HLA-DQB1 in patients with rheumatoid arthritis (RA). SHLA-G plasma concentrations from 106 RA patients (mean age 59.8 years, 80 women) were assessed by a sensitive enzyme-linked immunosorbent assay format. The mean sHLA-G levels were lower and sHLA-I levels higher in the RA patients than in healthy controls. Correlation coefficients of 0.248 to 0.344 (p < 0.01) between sHLA-G and rheumatoid factor, CRP, and EULAR joint swelling score were found. Patients with disease-associated HLA epitopes had higher sHLA-G levels than those without. Significantly lower sHLA-G was observed in groups of patients having HLA-DRB1*03 or HLA-DQB1*02 compared to groups without these genotypes. In contrast, HLA-DQB1*03 or disease-associated epitopes combined with HLA-DQB1*03 were associated with higher sHLA-G levels, whereas the inverse was observed in the combined presence of HLA-DRB1*03 and HLA-DQB1*02. SHLA-G as a percentage of sHLA-I was lower in patients positive for HLA-DQB1*02 and higher in patients positive for HLA-DQB1*03 and in its combined presence with disease-associated epitopes or with HLA-DRB1*07. As especially sHLA-G strongly inhibits T and natural killer (NK) cell functions, low sHLA-G suggests that T and NK cell activities are not efficiently restricted by sHLA-G molecules in rheumatoid arthritis. The sHLA-G levels, however, increase in correlation with parameters of disease activity and appear to be affected by the presence of disease-predisposing epitopes and other HLA-DRB1, DQB1 genotypes.