Evidence is presented to show that nonsteroidal antiinflammatory drugs react with two sites on the cyclooxygenase (8,11,14-eicosatrienoate, hydrogen-donor:oxygen oxidoreductase, EC 1.14.99.1). Although the degree of interaction with the catalytic site determines the potency of such compounds, interaction with the supplementary site is also obligatory for efficacy as cyclooxygenase inhibitors and may explain the selectivity of such drugs in inhibiting the cyclooxygenase but not the lipoxygenase pathway. Drugs that interact more effectively with the supplementary site than with the catalytic site-i.e., those of weak to moderate activity as cyclooxygenase inhibitors-are shown to prevent inhibition ofthe enzyme by indomethacin. Compounds in this class are also capable of blocking the ulcerogenic action of indomethacin, which suggests that this antiulcerogenic property stems from a direct action at the level of the cyclooxygenase in the stomach.There are a number of reports indicating that the ulcerogenic action of high doses of nonsteroidal antiinflammatory drugs (NSAIDs) is secondary to a depression of prostaglandin (PG) synthesis in the stomach (1, 2). The most convincing evidence in support of this concept is the finding that coadministration of PGs prevents induction of lesions by such drugs (2). Oral administration of salicylic acid-a weak inhibitor of cyclooxygenase (8,11,14-eicosatrienoate, hydrogen-donor:oxygen oxidoreductase, EC 1.14.99.1), the initial step in PG biosynthesis and the target of most NSAIDs-has been found to diminish gastric lesions resulting from high levels of indomethacin and other NSAIDs in rats (3). A similar action of 5-(2,4-difluorophenyl)-salicylic acid (diflunisal), an inhibitor ofcyclooxygenase of intermediate potency, has been reported (4). In view of the requirement for endogenous synthesis of PGs for homeostasis ofthe stomach, these findings with both salicylate and diflunisal could be due to their ability, after oral administration, to prevent the local transport of indomethacin to the active site on stomach cyclooxygenase or due to direct competition with indomethacin for a common site on the enzyme. The recent finding that salicylic acid also blunts the ulcerogenic action of indomethacin when administered parenterally may be interpreted to support this latter possibility (5). To explore this question and its possible relation to ulcerogenicity, a number of NSAIDs were examined for their ability to counter the effect of indomethacin in inhibiting cyclooxygenase activity in vitro and to suppress indomethacin-induced gastric lesions in vivo.
MATERIALS AND METHODSPG Cyclooxygenase. The NSAIDs and surfactant S10-7 (6) were obtained from N. Jensen and R. W. Egan, respectively, of these laboratories. Arachidonic acid was purchased from Sigma. Other materials were obtained from standard suppliers.Microsomes containing cyclooxygenase activity were prepared from ram seminal vesicles as described (7). The NSAIDs were incubated for 2 min at 30'C with approximately 3 mg of micro...