[RA-specific autoantibodies against a 68k antigen]

Despite commonly applied clinical criteria, the early diagnosis of rheumatoid arthritis (RA) often remains difficult, thus delaying on suitable early treatment. In search for a test furthering the early and reliable diagnosis of RA, we have screened for novel disease specific autoantibodies. To this end proteins were isolated from synovial membranes and other tissues following a special protein purification protocol, and these were separated electrophoretically. Western blots were then used to screen sera of RA patients and of individuals suffering from other rheumatic diseases for antibodies to any of these proteins. The most prominent RA specific immunoreaction was with a 68k antigen, occurring in 110 of 167 RA patients (sensitivity is 66%). The antibody could also be identified in seronegative RA patients but not in healthy individuals (55 tested), in only 1 SLE patient of a group of 98 patients with other rheumatic diseases and in 1 out of 22 HIV patients, resulting in a specificity of 99%. Moreover, the anti-68k antibody could be correlated with a more severe course of RA. 13 out of 20 anti-68k positive RA patients (58%) had subcutaneous nodules, while only 2 out of 11 anti-68k negative (20%) did. The mean sedimentation rate of these antibody positive patients was 51 mm/h and 26 mm/h for the negative respectively. The 68k antigen was shown to be present in all human tissues investigated and is probably ubiquitously expressed. It is either located in the endoplasmatic reticulum or cytoplasm or both. Its isoelectric point is 5.1. It proved to be O-glycosylated and contains only one or a few sugar residues as the untreated and the deglycosylated antigen identical electrophoretical mobilities. The patient derived anti-68k antibodies were directed against the sugar residue: deglycosylation of the antigen completely abolished its immunoreactivity. N-acetylglucosamine competes with the antibody for binding the 68k antigen. The antigen physicochemical data of the 68k antigen argue against identity with one of the autoantigens in this molecular mass range already known to be associated with RA or other autoimmune diseases. It is neither identical to the 62k human antigen (EBNA-1) nor to RA33 (A2hnRNP), the 50k Sa antigen or the Hsp70 class of heats-hock proteins. It is argued that the particular method of protein purification applied in combination with separation via SDS-PAGE in the presence of urea, made it possible to detect a hitherto unidentified antigen. Considering the striking disease specificity of the anti-68k antibody it is now worthwhile to look for corresponding autoreactive T cells in order to analyse its role in the pathogenesis of RA.