Fibronectin fragments have both catabolic and anabolic activities toward articular cartilage explants in vitro. Whereas a 1 nM concentration of an N-terminal 29 kDa fibronectin fragment (Fn-f) increases the proteoglycan (PG) content of cartilage without induction of matrix metalloproteinases (MMPs), 0.1Ő1 ƁM Fn-f temporarily suppresses PG synthesis and enhances MMP release. The higher concentrations cause an initially rapid PG depletion during the first week of culture, followed by much slower PG loss and gradually increasing rates of PG synthesis. To test for the involvement of mediators, human articular cartilage was cultured with Fn-f, and conditioned media were assayed for selected cytokines and factors. With 1 nM Fn-f, the release of the anabolic factors, insulin growth factor-I and transforming growth factor β1, from cultured cartilage was enhanced by 50Ő100% during the entire 28-day culture period and this was associated with both supernormal rates of PG synthesis and PG content. However, the higher concentrations of Fn-f additionally enhanced release, by at least 10-fold, of the cytokines, tumour necrosis factor α, interleukin-1α, interleukin-1β and interleukin-6 while causing depletion of cartilage PG. Release of tumour necrosis factor α, interleukin 1β and interleukin 1α peaked at days 2, 3 and 9 during or slightly after the period of maximal PG depletion and decreased to control levels by days 7, 7 and 21 respectively, whereas release of interleukin 6 was enhanced throughout the culture period. Neutralizing antibodies to the catabolic cytokines reduced Fn-f-mediated MMP-3 release and suppression of PG synthesis. The temporal aspects of this interplay between catabolic and anabolic factors are consistent with the kinetics of Fn-f-mediated cartilage damage and attempted repair and may be relevant to cartilage damage and repair in vivo.

These data demonstrate that although Fn-fs could be generated in vivo within synovial fluids and Fn-fs found in OA synovial fluid may contribute to cartilage damage in vivo, Fn-fs could also be generated within cartilage and amplify cartilage damage. Thus, Fn-fs may be both autocrine and paracrine regulators of cartilage metabolism.

A commercial preparation of 800-kDa hyaluronic acid (HA), (ARTZ from Seikagaku, Inc.), has been used as a therapeutic intervention in the treatment of osteoarthritis (OA). We tested the effect of this HA form, HA/800, in an in vitro cartilage chondrolytic system in which a specific amino-terminal 29-kDa fragment of fibronectin (Fn-f) penetrates cartilage tissue to activate chondrocytes to amplify two major chondrolytic activities: suppression of proteoglycan (PG) synthesis and induction of matrix metalloproteinases. We report that HA/800 did not block damage by Fn-f in serum free cartilage cultures. However, HA/800 was effective in blocking the ability of 100 nM Fn-f to cause the degradation and release of half of the total cartilage PG from cartilage in 10% serum/DMEM cultures. While the Fn-f caused a half-time for PG release of 3 days, continuous exposure to 0.1 or 1 mg/ml HA/800 slowed the half-time to 12 days. Further, a single 1 day pre-incubation with 0.1 or 1 mg/ml HA/800 was sufficient to decrease the half-time of 100 nM Fn-f mediated PG depletion to 7 and 12 days, respectively. HA/800 completely blocked the effect of 10 nM Fn-f. Blocking of Fn-f-mediated cartilage PG depletion was associated with a decreased concentration of Fn-f on the superficial cartilage surface and decreased penetration into the cultured cartilage tissue. Further, the two major chondrolytic activities of the Fn-f, suppression of synthesis of PG and enhanced release of stromelysin-1, were suppressed by HA/800. HA/800 also partially restored PG in cartilage first damaged with the Fn-F. We conclude that HA/800 slows Fn-f-mediated cartilage chondrolysis in vitro and has some reparative potential. The damage blocking activity appears to be associated with the ability of HA/800 to block penetration of the Fn-f, rather than with direct effects on cartilage tissue.