Characteristics of anti-type II collagen antibody binding to articular cartilage.

Abstract

Previous studies suggested that it was possible to characterize the intact surface of articular cartilage by probing it with antibodies against matrix macromolecules. The present studies were undertaken to investigate type II collagen (CII) epitope availability on the intact surface of articular cartilage. Normal bovine, rabbit, and human cartilage specimens were used to measure binding of anti-CII antibodies to the articular and cut surfaces of cartilage. Antisera were raised against the material obtained after brief extraction of the cartilage surface with 4M guanidine solution. Anti-CII did not bind to the intact surface of rabbit articular cartilage when compared with control rat sera, but did bind significantly to the cut surface. With normal human articular cartilage, the cut surfaces bound approximately 4 times as much anti-CII antibody as the intact articular surfaces. These findings suggested that the CII epitopes are normally protected by the superficial cartilage layer. In experiments carried out to characterize this layer, binding of anti-CII antibodies was unchanged after exhaustive washing of bovine or rabbit cartilage with phosphate buffered saline or 1M NaCl solution, whereas it was significantly increased after brief exposure to 4M guanidine solution or after incubation with neutrophil elastase. No restoration of the protection of CII epitopes in guanidine-treated cartilage could be achieved by incubation with undiluted normal bovine synovial fluid; however, 8-day culture of elastase-treated cartilage explants resulted in partial restoration of protection of the CII epitopes. Rat and rabbit antisera raised against the cartilage surface material extracted with 4M guanidine contained antibodies that bound to the cartilage surface. By Western blotting, rat antibodies identified 50-65-kd protein bands present in the guanidine extract, but not present either in the material obtained from brief digestion of cartilage with neutrophil elastase or in synovial fluid. The rabbit antisera identified a broad 70-95-kd band. Exposure of elastase-treated cartilage to the guanidine-extracted material resulted in a partial decrease of anti-CII antibody binding. These results suggest that CII on intact cartilage is protected from antibody binding, and that the protective material is at least partly proteinaceous in nature, is unlikely to be derived from synovial fluid, is noncovalently bound to the underlying intercellular matrix, and is synthesized by resident chondrocytes. Further characterization of the protective material may provide important information on the mechanisms of early cartilage damage in inflammatory arthritis.