3rd Annual BMRP Investigator Meeting - Abstract

Spatial Organization and Composition of the Mucosal Flora in Inflammatory Bowel Disease

Alexander Swidsinski^1,a and Laura P. Hale²

¹The University Hospital Charité of the Humboldt University of Berlin (Berlin, Germany); ²Department of Pathology, Duke University Medical Center (Durham, North Carolina, U.S.A.)

Background: Although intestinal bacteria have been strongly implicated in the pathogenesis of inflammatory bowel disease (IBD), the mechanisms involved are still poorly understood.

Materials and Methods: The spatial organization of intestinal flora was investigated using fluorescence in situ hybridization with over 60 bacterial probes that were applied simultaneously in different combinations. Biopsies from 80 patients with IBD and 160 controls were investigated. Additionally, whole intestines were examined for normal mice (N=10) and for three models of murine colitis (IL-10 knockout [n=10], DSS [n=10], RAG T-cell transfer [N=10]). To better understand mechanisms that contribute to establishment of mucosal biofilms, an in vitro model simulating mucus-gel/feces interactions was developed.

Results: Biofilms consisting of bacteria tightly adherent to the mucosal surface were not typically present in the intestine of healthy controls. Biopsies taken from ileum, ascending and sigmoid colon were covered with mucus that prevented contact with intestinal bacteria. In contrast, a dense tightly adherent biofilm was detected in all IBD patients not treated with antibiotics. Although less abundant, biofilm was also found in all patients with self limiting colitis and in 60% of patients with diverticulosis and irritable bowel syndrome (IBS). The composition of biofilm in controls and IBD differed. Bacteroides spp. were predominant in IBD, while Eubacterium rectale was predominant in IBS. In some IBD patients, specific biofilms with mainly E.coli or Serpulina were observed. Similarly to what was observed in human data, no bacterial adherence was found in normal mice. The small intestine was bacteria-free. The murine colonic flora was structurally organized in distinct differently composed compartments: crypt, interlaced, and fecal. Feces contained a homogeneous mix of different bacterial groups that were separated from the mucosa either by a mucus gap omitting bacteria in distal colonic segments or by an interlaced layer and cryptal bacteria in proximal colon. Colonic inflammation was accompanied by a depletion of bacteria within the fecal compartment, increased thickness and spread of the mucus gap, and massive increases of bacterial concentrations in the crypt and interlaced compartments. Adhesive and infiltrative bacteria were observed only in inflamed colon, with Bacteroides spp. dominating. In vitro studies showed that the velocity of bacterial movements through gels of different viscosity varied for different groups of bacteria. The mobility of bacteria was enhanced after addition of emulsifying agents like DSS.

Conclusions: Despite extremely high bacterial concentrations, the colon wall is normally effectively protected from a direct contact with potentially harmful bacterial groups like Bacteroides, Enterobacteriaceae, and Clostridium difficile. Colonic inflammation in murine models of IBD is accompanied by a gradual increase in thickness and spread of mucus and interlaced layers, migration of leucocytes into the fecal compartment, and their arraying at the borderline between mucus/feces or interlaced layer and feces. Bacterial adhesion and invasion of mucosa occur as these mechanisms progressively break down. This breakdown and propagation of adherent biofilm is also a prominent feature in biopsy material from human patients with IBD. Detergents and emulators counteract the separating function of the mucus layer and have potential impact on pathogenesis of IBD.

^aPrincipal Investigator