Synovial membrane consists of fibroblasts and macrophages forming the synovial lining supported by vascularized subsynovium. Each of these components may specifically react to a particular stimulus. Thus, reactions of isolated synovial cells may not correspond to that of intact tissue. We characterized the production of hyaluronan (HA) by rat synovial membrane exposed in vitro to pro- and anti-inflammatory cytokines and compared it with previous results obtained with isolated fibroblasts. Synovial membrane dissected from one knee joint served as a control to that from the opposite knee exposed to IL-1 beta, TGF-beta1, TNF-alpha, IFN-gamma or IL-4 for 24 h. The HA content was determined by ELISA, and hyaluronan synthase (HAS) mRNA by real-time PCR. The size distribution of the HA chain was evaluated by agarose gel electrophoresis. The HA content in the freshly dissected synovial membrane was approximately 1 microg and decreased to 0.1 microg after incubation, while in the medium it increased from 0 to 3 to 5 microg. All cytokines stimulated production of HA. The strongest effect was observed in the case of TNF-alpha. The level of HAS1 and HAS2 mRNA increased 2-fold during a 12-h incubation while that of HAS3 decreased. The distribution of the HA chain length did not differ in the medium from the control and stimulated membranes. Transfer of the synovial membrane from the HA-rich synovial fluid into the medium stimulated release of HA from the membrane and increased HAS expression and HA production. Thus, the synovial membrane acts as a sensor reacting to changes in HA concentration in its environment. Pro-inflammatory cytokines stimulate production of HA in intact synovial membranes similarly as in cultures of rheumatoid fibroblasts. In contrast, our results suggest that the response to anti-inflammatory cytokines (TGF-beta1 and IL-4) of the whole synovial membrane differs from that of isolated fibroblasts.