Luminal hypertonic solutions stimulate concentration-dependent duodenal serotonin release.

Abstract

The serotonin (5-HT) response to varying concentrations and types of luminal hyperosmolar solutions was examined in rabbit duodenum to define the mechanisms and direction (luminal or vascular) of release. Segments of duodenum were studied in either the Ussing chamber or an isolated, vascularly perfused model. Adjacent mucosal sheets devoid of muscularis were studied in parallel Ussing chambers with exposure of the mucosal surface to 0 to 50 gm/dl dextrose or 0 to 20 gm/dl mannitol. The 5-HT response to hyperosmolar dextrose was measured in the presence of autonomic receptor blockers and tetrodotoxin. In the isolated, perfused duodenum, venous and luminal drainage were sampled during basal, dextrose-infusion, and recovery periods. 5-HT content was measured by high-performance liquid chromatography. In chambered mucosal sheets the percentage of change in mucosal 5-HT release was a linear function of luminal dextrose (r = 0.96; p less than 0.05) or mannitol (r = 0.98; p less than 0.02) concentrations. A significant 65.8 +/- 19% inhibition of 5-HT release was observed for tetrodotoxin but not for atropine, hexamethonium, phentolamine, or propranolol. In the isolated, perfused duodenum, 50 gm/dl dextrose caused an integrated 5-HT venous response of 3.0 +/- 0.7 micrograms/100 gm (p less than 0.05 vs luminal or control venous release). It was therefore concluded that duodenal 5-HT release is concentration dependent for intraluminal glucose and mannitol, is partially neurally mediated by a nonadrenergic, noncholinergic pathway, and is predominantly into the venous drainage.