Autoantibodies to GPI Found in Patients with Rheumatoid Arthritis
A recently described mouse model of spontaneous inflammatory arthritis (the K/BxNT cell receptor-transgenic mouse) shares many features with human rheumatoid arthritis (RA). In the mice, the arthritis is sustained by autoantibodies to glucose-6-phosphate isomerase (GPI) and the arthritis can be induced in nave mice by transfer of serum. In this study, Ditzes, et al have assessed whether antibodies to GPI play a role in human RA.
Sera from 69 RA patients, 107 normal controls were analyzed by ELISA for anti-GPI IgG. 64% (44 of 69) of the RA sera were positive (A405 = 1.8 + 0.84) where as only 3% (3 of 107) of the controls were weakly positive (A405 0.59 + 0.37). Additionally, the sera were analyzed for the presences of GPI. The mean concentration of GPI in the RA patients was significantly higher than in the normal controls, 0.210 + 0.139 U/ml and 0.069 + 0.048 U/ml, respectively. A significant positive correlation (R=0.79) was found between the serum concentrations of GPI and anti-GPI antibodies in the RA patients.
Synovial fluid was also analyzed. Antibodies to GPI were found in the synovial fluid in 33% (8 of 24) of the RA patients, but in none of the osteoarthritis (OA) patients or normal controls. The mean GPI concentration in the RA patients (0.431 + 0.049 U/ml) was also significantly higher compared to the concentration found in both OA patient and normal controls, which was the same in both groups (0.060 + 0.052 U/ml). Chromatography of the synovial fluid confirmed the presence of anti-GPI/GPI immune complexes.
Immunohistochemical analysis of the synovial tissue from patients with RA showed intense staining of endothelial cells of synovial arterioles and some capillaries, and keratinocytes outside the synovium. In contrast, synovial tissue from controls and patients with OA showed minimal staining for GPI, as expected with a protein that is ubiquitously expressed.
This study confirms the presence of both GPI and anti-GPI antibodies in the sera and synovial fluid, as well as high levels of GPI staining in the synovial tissue of patients with RA. The findings in this study suggest further work is needed to determine if an autoimmune reaction to GPI is pathogenic in human RA.
Editorial Comment: A number of Autoantibodies has been described in RA, including rheumatoid factor (anti-Ig), anti-collagen, and anti-filaggrin. However, it remains unclear whether these antigen-antibody complexes represent primary disease pathology or are “innocent bystanders” resulting from tissue damage. The role of GPI in this process is similarly unclear and entirely unexpected, given its ubiquitous expression. Nonetheless, the parallels between this mouse model and human disease are striking and the role of GPI in RA deserves further investigation.