P118 Fecal content from IBD patients disturbs epithelial functions in two model systems

Abstract Background Inflammatory Bowel Diseases (IBD) result from abnormal interactions between the immune system, epithelial barrier, and gut microbiota. However, understanding the abnormal interaction between aerobic epithelia and mainly anaerobic microbiota is complicated and is limited in part by the lack of human-relevant models. Methods We developed a model system that is based on whole human fecal samples and human-derived epithelia to study the interactions between the gut bacteria and the epithelia. To capture patient heterogeneity, we prioritized and pooled fecal material from 10 Crohn disease (CD), 10 ulcerative colitis (UC), and 10 healthy subjects. Prioritization of samples was based on our published gut microbial health index (PMID: 35197084), matching gender and age. Monolayered epithelial cells (Caco2 cells and patient-derived colonoids) were used for the co-culturing. Measurable outputs included cell viability, epithelial integrity, and secretion of CXCL1. Results Fecal samples were processed by separating the supernatant, heat-killing the bacteria, and rejoining the samples (Fig. 1A). 16S sequencing showed that bacterial composition and alpha diversity (Faith) were overall preserved after the heat-killing (Fig. 1B), as seen in the original samples. The gut microbial health index was higher among the control samples and pool, compared to the IBD samples (Fig. 1C), preserving the IBD pathogenic signals. Because previous studies indicated the enrichment of salivary bacteria in IBD feces, salivary samples were also processed for comparison. We confirmed that similar amounts of bacterial DNA (as a proxy of bacterial mass) were used for coculturing as indicated by qPCR of the 16S-V4 (Fig. 1D). Co-culturing the fecal and saliva pools with Caco-2 cells revealed a significant increase in CXCL1 proinflammatory chemokine secretion to the media in comparison to no-treatment (Fig. 2A). Moreover, the UC and CD pools significantly decreased epithelial integrity [transepithelial electrical resistance (TEER)], without affecting cellular viability (Fig. 2B-D). Finally, incubation of the fecal content with colon-derived organoids (from 2 controls and 1 UC) resulted in a substantial reduction of cell viability, which was more pronounced with fecal content derived from UC patients (Fig. 2E), and this effect was noted most with the UC fecal supernatants. Conclusion Using a novel co-culturing system, we demonstrated the effects of fecal contents from IBD and control subjects on CXCL1 secretion and epithelial integrity in Caco2 cells, and on colonoids viability. These models will be used to identify fecal factors and interventions relevant to IBD epithelial functions.