Objective. To measure levels of cytokines, proteases, and glycosaminoglycans (GAG) in synovial fluid (SF) From the knees of animals with experimental osteoarthritis (OA) and from their contralateral (control) knees, and to compare and correlate these values with each other as well as with measures of proteoglycan synthesis in the corresponding articular cartilage. This study will help to identify cytokines of potential importance in the early stages of the development of OA.Methods. OA was induced in 12 mature animals by sectioning the anterior cruciate ligament. After 3 months, SF From the operated and contralateral (control) knee joints was assayed for interleukind (IL-6), tumor necrosis Factor (TNF), IL-1, latent metalloproteinase, and sulfated GAG. Proteoglycan synthesis in the corresponding articular cartilage was also measured.Results. IL-6 levels in SF from the operated joint compared with the control joint were significantly ele- Submitted for publication July 9, 1992; accepted in revised form January 6, 1993. vated in 11 of 12 animals. TNF levels were also elevated in 10 of 11 SF samples from operated joints, but to a lesser extent than those of IL-6. IL-1 and IL-1 inhibitors were undetectable in either the operated or control joint SF. The GAG concentration was elevated in SF from experimental OA joints. This elevation correlated with that of TNF, but not IL-6. There was no significant difference in the concentration of APMA-activatable metalloproteinase. The rate of proteoglycan synthesis was higher in the cartilage from the operated joint in 8 of 12 animals, and the mean rate of synthesis was significantly higher than in the control joint. There was a positive correlation between this increase in cartilage proteoglycan synthesis (operated versus control) and the increase in SF IL-6, but there was no correlation with the levels of TNF or GAG.Conclusion. This is the first study of SF levels of cytokines in early experimental OA. Our results show surprisingly high levels of IL-6 in operated joints, where the cytokine could act directly on the chondrocytes, and thus play a role in mediating their responses to cartilage injury.

The absence of keratan sulphate synthesis from skeletal tissues of young and mature mice and rats has been confirmed by (1) analysis of specific enzyme degradation products of newly synthesized glycosaminoglycans, and (2) immunohistochemistry and radioimmunoassay using a monoclonal antibody directed against keratan sulphate. Approx. 98% of the [35S]glycosaminoglycans synthesized in vivo by mouse and rat costal cartilage, and all of those of lumbar disc, are chondroitin sulphate. The remainder in costal cartilage were identified as heparan sulphate in mature rats. In contrast, [35S]glycosaminoglycans synthesized by cornea of both species comprised both chondroitin sulphate and keratan sulphate. In mice keratan sulphate accounted for 12-25% and in rats 40-50% of the total [35S]glycosaminoglycans, depending on the age of the animal. Experiments in vitro with organ culture of cartilage and cornea confirm these results. Absence of keratan sulphate from mouse costal cartilage and lumbar disc D1-proteoglycans was corroborated by inhibition radioimmunoassay with the monoclonal antibody MZ15 and by lack of staining for keratan sulphate in indirect immunofluorescence studies using the same antibody.

Objective. To determine whether chondrocytes in early experimental osteoarthritic (OA) cartilage continue to show increased synthesis and turnover of proteoglycans (PGs) during explant culture. A comparison was also made between the responsiveness of experimental OA and control cartilage to interleukin-lf3 (IL-1p) and tumor necrosis factor a (TNFa) after 1 day and 3 days in culture.Methods. OA was induced in mature animals by sectioning of the anterior cruciate ligament followed by 3 months of normal exercise. PG synthesis in the articular cartilage was determined by measuring "S-sulfate incorporation during explant culture over 1-3 days. Inhibition of PG synthesis was also determined with various concentrations of IL-lfl and TNFa after 1 and 3 days in culture. PGs extracted from the articular cartilage over 1-3 days in culture were examined by agarosepolyacrylamide gel electrophoresis.Results. Up to 24 hours after excision from the joint, PG synthesis was higher in experimental OA cartilage than in control cartilage. It was also less sensitive to inhibition by TNFa. These differences were no longer detected after 48-72 hours in culture. There were no changes in the relative proportions of aggrecan and decoridbiglycan extracted from and synthesized by control and experimental OA cartilage over the 3 days in culture.Conclusion. Previous results indicated that PG synthesis and turnover in articular cartilage was increased for many months after induction of experimen- ____tal OA. Our present results show that the enhanced rate of PG synthesis and turnover were evident in freshly explanted tissue, but the differences were lost over 3 days in culture. A decreased responsiveness to TNFa was also lost. The hypermetabolic activity of experimental OA chondrocytes was thus reversible and not a permanent change in chondrocyte phenotype.The canine model of experimental osteoarthritis (OA) described by Pond and Nuki ( l ) , in which OA changes are induced by resection of the anterior cruciate ligament of the stifle joint, has features which resemble those of naturally occurring canine OA and can progress to exhibit many features of advanced OA in humans (2,3). Damage and eventual loss of the articular cartilage is a frequent outcome of natural and experimental OA, and this experimental model provides an opportunity to investigate the early changes in articular cartilage that precede its damage.Work from our laboratory and others has shown that some of the earliest changes in the articular cartilage are both increased rates of matrix protein synthesis (4-7) and increased rates of degradation of matrix components (8-1 l), which suggest that there is hyperactivity of the chondrocytes. The newly synthesized proteoglycans (PGs) exhibit qualitative differences from those of control cartilage as do those released from experimental OA cartilage (9). The factors which cause these changes are not known. It is also not established if these cellular changes are reversible or reflect the emergence of the "OA chondrocyte." In experimenta...

The synthesis and turnover in vivo of 35S-labelled proteoglycans in mouse cervical, thoracic and lumbar intervertebral discs, and in costal cartilage, was investigated after intraperitoneal injection of [35S]sulphate. Intervertebral discs and costal cartilage synthesize similar amounts of 35S-labelled proteoglycans per microgram of DNA. Discs and cartilage all synthesize a major proteoglycan species (approx. 85%) of large hydrodynamic size and a minor species (approx. 15%) of small size. Both proteoglycans carry chondroitin sulphate chains. Keratan sulphate was not found associated with either species. The total 35S-labelled proteoglycan pool had a metabolic half-life (t1/2) of 10-12 days in discs, and 17 days in cartilage. The extractable major and minor species turned over at similar rates. Those proteoglycans left in the tissue after 29 days turn over very slowly. Approx. 50% of the major 35S-labelled proteoglycan species formed mixed aggregates with hyaluronic acid and rat chondrosarcoma proteoglycan. The ability to form aggregates did not decrease up to 45 days after synthesis. Of the heterogeneous population of proteoglycans comprising the major species, those remaining in the tissue 9 days after synthesis were of smaller average hydrodynamic size and had shorter chondroitin sulphate side chains than the average size at the time of synthesis. With increasing time after synthesis, proteoglycans were less readily extracted from the tissue by 4.0 M-guanidinium chloride than at the time of synthesis.