The neural control of the serotonin content in mammalian enterochromaffin cells.

Serotonin (5-HT) in the mammalian gut is mainly stored in enterochromaffin cells (EC). Vagal or splanchnic nerve stimulation respectively increase portal blood levels of 5-HT and decrease 5-HT content in EC, suggesting a neural control of these cells. Previous studies have also suggested that the vagal effect is mediated via vagal adrenergic nerve fibers, since this effect could be blocked by sympathectomy, removal of the superior cervical sympathetic ganglia or by the administration of a beta-adrenoceptor blocking agent, d, 1-propranolol. In the present study the control of the 5-HT content in EC was studied by a cytofluorimetric method in specimens from rat duodenal mucosa after drug incubations and transmural field stimulation (TFS) in vitro: Incubations in a Krebs' solution containing adrenaline (A), noradrenaline (NA) or isoprenaline (IP) (10(-9)--10(-5) M) caused a decrease of the 5-HT content in EC. For NA and A this effect was concentration-dependent. The effect of A was antagonized by d, 1-propranolol but not by d-propranolol, metoprolol, phentolamine or phenoxybenzamine. This indicates that the effects observed were mediated via a true beta-adrenoceptor mechanism. Acetylcholine (ACh), in the presence of eserine, also decreased the 5-HT content in EC, but was less potent than the adrenergic substances. TFS decreased the 5-HT content of EC to about 75% of unstimulated control levels. The presence of tetrodotoxin or d, 1-propranolol in the stimulation bath antagonized the effect of TFS. In specimens from rats pretreated with 6-OH-dopamine, TFS had no effect. Thus, TFS appears to induce a release of 5-HT from EC mainly by activation of adrenergic nerves. The ultrastructural relationship between nerve terminals and EC was studied in the guinea-pig duodenum. Animals were pretreated with 5-OH-dopamine in order to visualize adrenergic nerves. Near the base of most of the EC examined, at least 3 different types of nerve terminals were observed with the appearance of adrenergic, cholinergic and peptidergic terminals. In addition, dendrite-like processes were demonstrated. No true synapse on EC was found, but the minimal distance between the nerve terminals and EC was 150-250 nm, well within the limits of the "autonomic gap", suggesting a functional neural influence on these cells. The importance of the demonstrated beta-adrenergic control of EC was further studied in vivo. Efferent vagal nerve stimulation in the cat caused increased 5-HT levels in the portal plasma to more than twice the normal values after 15 min stimulation. Administration of the beta-adrenoceptor antagonist d, 1-propranolol in various doses did not abolish, but significantly reduced the response, particularly during the final part of the stimulation period. These results not only confirm a beta-adrenoceptor mediated release of 5-HT, but also suggest that other mechanisms are involved in the total response to vagal nerve stimulation.