Abstract

It has been assumed that cyclooxygenase-2 (COX-2) is solely responsible for inflammatory processes. Recently, this view has been challenged because COX-2-selective agents caused a delay of gastric ulcer healing and exacerbation of inflammation in rats. To further characterize organ-specific toxic effects of selective and nonselective COX inhibitors, we assessed the eicosanoid release from different rat organs ex vivo after oral administration of the COX-2-selective inhibitor NS-398 and the unselective COX inhibitors diclofenac, meloxicam, and ketorolac. Prostanoid and leukotriene release from tissue fragments of the stomach, kidney, lung, and brain were determined after ex vivo incubation of tissue fragments in Tyrode's solution for 10 min at 37°C. Ketorolac (0.1, 0.3, and 0.9 mg/kg) inhibited prostanoid release from all organs most potently and led to a significant increase of leukotriene release from the lung. Effects of diclofenac and meloxicam (1, 3, and 9 mg/kg each) were similar for all organs tested. At 9 mg/kg, 6keto-prostaglandin F (PGF)_1αrelease from gastric mucosa was reduced by 79.1 ± 11.4 and 87.6 ± 7.7% and PGE₂ release from rat kidney was inhibited by 60.4 ± 6.8 and 78.6 ± 16.6% by diclofenac and meloxicam, respectively. NS-398 did not reduce prostanoid release from the lung. Consistent with the reported constitutive expression of COX-2, prostanoid release from kidney and brain was reduced by 20 to 30%. The release of 6keto-PGF_1α from gastric mucosa was reduced by 34.7 ± 22.2% at 3 mg/kg and by 86.9 ± 12.7% at 9 mg/kg. At these doses, NS-398 has been previously shown to be COX-2 selective. Because PGF_1α is the stable breakdown product of PGI₂, these results suggest that COX-2 contributes to PGI₂ synthesis in the rat stomach.