Abstract 1332 Poster Session II, Sunday, 5/2 (poster 53)

Prematurity, ischemia, bacterial colonization, and substrate have all been implicated in the pathogenesis of Necrotizing Enterocolitis (NEC). Cytokines and other inflammatory mediators have been examined in babies with NEC, in animal models, and have been linked to injury. We have previously reported NEC-like injury in a weanling rabbit ileal loop model using adherent E. coli and blockage of injury upon co-infection with a candidate Gram positive partner (Enterococcus faecium).

In the current study, we examined the changes induced by each bacterial strain alone and in co-infection (E. faecium+E. coli). An attempt was also made to evaluate the role of a general immunosuppressive agent (dexamethasone) in this bacteria-induced disease model. Using our previously reported protocols, 4 loops per rabbit (1 control receiving PBS, 3 inoculated with bacteria) were made in multiple rabbits. Blood was collected at 3 and 6 hours post-surgery and at necropsy for culture. Rabbits in the immunosuppressant group received 1.5mg of dexamethasone/kg body weight IM, 12 hr before and at surgery. The E. coli strain consistently produced diffuse, severe, ulcerative, necrohemorrhagic ileitis. Specific epithelium-bacterial association was consistently seen. In several loops, heterophils and/or eosinophils predominated. Very few (6-95 CFU/ml) E. coli were cultured from blood at 6 hr that reached about 104 CFU/ml by 12-14 hr. We were unable to demonstrate any necrosis in E. faecium (alone) and co-infected loops. However, there was mild to moderate inflammation, submucosal edema, lacteal dilatation, and there was almost total blockage of bacterial translocation (≈0-10 CFU/ml at necropsy). In the dexamethasone-treated, E. coli-infected loops, similar diffuse, severe ulcerative, necrohemorrhagic changes were seen without eosinophilic predominance. Interestingly, the loops co-infected with E. faecium did not show protection against injury. Necrotic changes with severity similar to loops in non-dexamethasone-treated animals were noted transmurally. Bacterial translocation was observed in E. coli and co-infections, that started at 6 hr(3-25 CFU/ml) and was 10-fold higher (>105 CFU/ml) at necropsy compared to non-dexamethasone animals.

Dexamethasone, having a broad range of anti-inflammatory efficacy, could not ameliorate the inflammation and injury process in our system; rather, the protective effects of Gram (+) bacteria (E. faecium) against E. coli were abrogated. These results suggest that the Gram (+) organisms in our model are capable of inducing specific responses in the intestine and may provide protection against injury and bacterial translocation via these protective host responses. Identification and characterization of specific response(s) evoked by different components and secretions from Gram (+) organisms warrant further investigation.

Supported by a grant (HD 29735) from NICHD.