Expression of matrix metalloproteinase 9 (96-kd gelatinase B) in human rheumatoid arthritis

Objective. To determine the expression of matrix metalloproteinase 9/gelatinase B (MMP-9) in synovial fluid (SF), plasma, and synovial tissue from individuals with rheumatoid arthritis (RA), inflammatory arthritis (IA), and osteoarthritis (OA), using specific monoclonal antibody reagents. Methods. Gelatinolytic activity in the SF and plasma of patients with RA, IA, and OA was assessed by gelatin zymography. A mouse monoclonal antiserum, 277.13, which selectively recognizes soluble latent forms of human MMP-9, was used to quantitate MMP-9 levels in patient synovial effusions, plasma, and synovial tissue with a capture sandwich enzyme-linked immunosorbent assay (ELISA). Fifty-one SF samples (31 RA, 9 OA, 11 IA) were analyzed. Immunolocalization of MMP-9 in RA, OA, and normal synovium was investigated using MMP-9-specific antisera. Results. MMP-9 antigen levels in synovial effusions were elevated 67-fold in RA samples compared with OA samples. In addition, although MMP-9 antigen levels in IA synovial effusions were 2.7- fold less than the values in RA samples, they were elevated 34-fold over the values in OA samples. These data indicate an association between increased MMP-9 levels and inflammatory arthritis. A predominant 92-kd gelatinolytic activity (specifically inhibited by EDTA) was evident in RA and IA samples, but no activity was observed in OA samples. Among 86 plasma samples (17 RA, 9 IA, 60 normal controls) analyzed for MMP-9 antigen levels by immunocapture ELISA, MMP-9 antigen levels were elevated 7-fold in RA plasma compared with normal plasma. RA synovial tissue extracts demonstrated elevated levels of MMP-9 antigen compared with OA synovial tissue. MMP-9 immunolocalization studies demonstrated expression in infiltrating leukocytes (neutrophils and macrophages), endothelial cells, and synovial fibroblasts in RA synovium. Conclusion. Latent MMP-9 and/or MMP-9 tissue inhibitor of metalloproteinases 1 (TIMP-1)complexes are elevated in RA and IA SF compared with OA SF. In addition, MMP-9 is increased in RA plasma versus normal control plasma. Synovial tissue levels of MMP-9 antigen are also elevated in RA versus OA. The tissue distribution of MMP-9 within RA synovium is localized to sites of inflammation comprising surface synovial lining cells, endothelium, and leukocytes. Taken together, these observations suggest that connective tissue turnover occurs as a result of excessive MMP activity over TIMP action in the invading pannus, periarticular tissue, or SF. Further studies such as those used in the present investigation will help elucidate the role of a number of different enzymes and inhibitors in the destructive arthropathies.