Sodium salicylate inhibits prostaglandin formation without affecting the induction of cyclooxygenase-2 by bacterial lipopolysaccharide in vivo.

The mechanisms underlying the anti-inflammatory properties of salicylate are not well understood. In particular, while salicylate inhibits prostaglandin production in vivo it only weakly inhibits cyclooxygenase (COX)-1 or -2 activity in vitro. Thus, it has often been suggested that in vivo salicylate may inhibit the expression rather than the activity of COX, particularly COX-2. Using a model of acute COX-2 expression in the rat, we show that salicylate inhibits COX-2 activity in vivo without affecting COX-2 expression. In anesthetized rats LPS (6 mg kg(-1), i.p.) increased the expression of COX-2 as evidenced by increased circulating levels of 6-keto-prostaglandin F(1alpha) (6-keto-PGF(1alpha), a stable breakdown product of PGI(2)), greatly exaggerated formation of 6-keto-PGF(1alpha) following arachidonic acid (AA) challenge (3 mg kg(-1), i.v.), and increased expression of COX-2, but not COX-1, protein. Diclofenac (3 mg kg(-1), i.p.) or the COX-2 selective agent diisopropyl fluorophosphate (10 mg kg(-1), i.p.) decreased the LPS-induced increase in circulating 6-keto-PGF(1alpha) and the exaggerated 6-keto-PGF(1alpha) production following AA challenge. Sodium salicylate (20 or 120 mg kg(-1), i.p.) (administered either 1 h prior, or once per day for 3 days prior, to LPS injection) reduced only the LPS-induced increase in circulating 6-keto-PGF(1alpha), but not the exaggerated 6-keto-PGF(1alpha) production following AA challenge or the expression of COX-2. Thus, salicylate inhibits LPS-induced COX-2 activity in a manner that is overcome by provision of excess substrate and independent of effects on COX-2 expression. In conclusion, our results exclude mechanisms other than direct enzyme inhibition as responsible for the anti-COX effects of salicylate.