Consequences of Citrobacter rodentium infection on enteroendocrine cells and the enteric nervous system in the mouse colon.

We tested the hypothesis that Citrobacter rodentium infection leads to changes in the mucosal enteroendocrine signalling and the enteric nervous system and that the host's immune response contributes to these changes. Enteroendocrine cells, serotonin (5-HT) reuptake transporter (SERT), 5-HT release, and inducible nitric oxide synthase (iNOS) expression were assessed in the colon of infected wild-type or severe combined immunodeficient (SCID) mice. Immunoreactivity for iNOS and neuropeptides were examined in the submucosal and myenteric plexuses. Mice were orogastrically infected with C. rodentium and experiments were conducted during the injury phase (10 days) and the recovery phase (30 days). 5-HT and somatostatin enteroendocrine cells and SERT were significantly reduced 10 days after infection, with numbers returning to control values at 30 days. 5-HT release was increased at 10 days. Changes to the mucosal serotonin signalling system were not observed in SCID mice. iNOS immunoreactivity was increased in the submucosa and mucosa at 10 days and returned to baseline levels by 30 days. No differences were observed in neuropeptide or iNOS immunoreactivity in the enteric plexuses following infection. The host's immune response underlies changes to enteroendocrine cells, SERT expression and 5-HT release in C. rodentium infection. These changes could contribute to disturbances in gut function arising from enteric infection.

Citrobacter rodentium infection causes iNOS-independent intestinal epithelial dysfunction in mice.

Attaching-effacing bacteria are major causes of infectious diarrhea in humans worldwide. Citrobacter rodentium is an attaching-effacing enteric pathogen that causes transmissible murine colonic mucosal hyperplasia. We characterized colonic inflammation and ion transport at 3, 7, 10, 30, and 60 d after infection of C57Bl/6 mice with C. rodentium. Macroscopic damage score was significantly increased 7 and 10 d after infection. Colonic wall thickness was increased at 7, 10, 30, and 60 d. Myeloperoxidase (MPO) activity was significantly increased at 3, 7, and 10 d and returned to control levels by days 30 and 60. The expressions of inducible nitric oxide synthase and cyclooxygenase-2 were increased by C. rodentium infection. Significant reductions in the epithelial secretory response to carbachol, but not to electrical field stimulation or forskolin, were observed at 3 and 10 d of infection. Translocation of enteric bacteria into the mesenteric lymph nodes was observed 10 d following infection. There was no difference in response to infection between animals deficient in inducible nitric oxide synthase and wild-type controls. The COX-2 inhibitor rofecoxib caused decreased wall thickness and MPO activity at day 10. However, COX-2 inhibition did not alter infection-induced changes in ion transport. Citrobacter rodentium infection causes colonic inflammation, mucosal hyperplasia, and nitric-oxide-independent epithelial dysfunction in association with increased permeability to luminal bacteria.

Nitric oxide inhibits cAMP-dependent CFTR trafficking in intestinal epithelial cells.

Nitric oxide is produced during intestinal inflammation and inhibits the epithelial responsiveness to cAMP-dependent secretagogues. The effect is presumably due to inhibition of activation of the CFTR. However, because insertion of CFTR into the epithelial apical membrane is also a cAMP-dependent process, we tested the hypothesis that NO could inhibit cAMP-dependent CFTR trafficking. SCBN intestinal epithelial cells were treated with forskolin to activate adenylate cyclase activity. The cells were fixed at various times and immunostained for CFTR. Some cells were pretreated with the nitric oxide donor PAPA-NONOate, the protein kinase A inhibitor H89, or the microtubule blocker nocodazole. Cross sections of epithelial monolayers were then studied under fluorescence, and the ratio of apical to basolateral CFTR immunoreactivity was determined. Stimulation of adenylate cyclase activity caused an increase in the apical-to-basolateral ratio of CFTR within 30 s. This effect was transient and preceded changes in short-circuit current in SCBN monolayers mounted in Ussing chambers. PAPA-NONOate, H89, and nocodazole all reduced forskolin-stimulated CFTR trafficking. The inhibitory effect of the NO donor was not affected by pretreatment with the guanylate cyclase inhibitor 1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one. PAPA-NONOate reduced forskolin-stimulated increases in intracellular cAMP. The data suggest that a portion of the inhibitory effect of nitric oxide donors on cAMP-dependent chloride secretion is through the inhibition of cAMP-dependent insertion of CFTR into the apical plasma membrane. These data provide insight into the mechanism of secretory dysfunction in inflammatory diseases of the gut where mucosal nitric oxide is elevated.