The diarrhea of cholera is considered to rely solely on a cyclic adenosine monophosphate-mediated secretory mechanism. However, both 5-hydroxytryptamine and prostaglandin E2 have been proposed to be involved in the pathogenesis of cholera. In vivo experiments were performed, therefore, in the rat jejunum to investigate the influence of purified cholera toxin on fluid secretion, luminal release of 5-hydroxytryptamine and prostaglandin E2, and formation of mucosal cyclic adenosine monophosphate. Also the effects of ketanserin, indomethacin, verapamil, and nifedipine on the named parameters were studied. Cholera toxin dose-dependently (0.1-0.5 microgram/ml) and time-dependently (1-5 h) increased mean net fluid secretion with a maximum response at 4 h. It also caused a significant (p less than 0.01) rise in release of 5-hydroxytryptamine and prostaglandin E2, in addition to formation of cyclic adenosine monophosphate. The dose-response curve for cholera toxin-induced fluid secretion was shifted to the right by indomethacin (10 mg/kg s.c.) and ketanserin (200 micrograms/kg s.c.), none of which caused a change in cholera toxin-induced release of 5-hydroxytryptamine. However, both agents significantly decreased the release of prostaglandin E2. Verapamil (0.2-9.5 micrograms/min i.a.) and nifedipine (0.05-0.5 microgram/min i.a.) dose-dependently reduced cholera toxin-induced fluid secretion. The estimated local concentrations at half-maximal inhibition were 5 x 10(-7) M verapamil and 5 x 10(-8) M nifedipine, respectively. The cholera toxin-induced increase in release of 5-hydroxytryptamine and prostaglandin E2 and formation of cyclic adenosine monophosphate was unaffected by verapamil. These results support the concept that cholera toxin-induced fluid secretion in vivo is caused, in part, by release of 5-hydroxytryptamine, which in turn stimulates formation of prostaglandin E2.