Search

In the gut microbiota, resident bacteria prevent pathogens infection by producing specific metabolites. Among bacteria belonging to phylum Bacteroidota, we have previously shown that Bacteroides fragilis or its cell-free supernatant inhibited in vitro Salmonella Heidelberg translocation. In the present study, we have analyzed this supernatant to identify bioactive molecules after extraction and subsequent fractionation using a semi-preparative reversed-phase Liquid Chromatography High-Resolution Tandem Mass Spectrometry (LC-HRMS/MS). The results indicated that only two fractions (F3 and F4) strongly inhibited S. Heidelberg translocation in a model mimicking the intestinal epithelium. The efficiency of the bioactive fractions was evaluated in BALB/c mice, and the results showed a decrease of S. Heidelberg in Peyer’s patches and spleen, associated with a decrease in inflammatory cytokines and neutrophils infiltration. The reduction of the genus Alistipes in mice receiving the fractions could be related to the anti-inflammatory effects of bioactive fractions. Furthermore, these bioactive fractions did not alter the gut microbiota diversity in mice. To further characterize the compounds present in these bioactive fractions, Liquid Chromatography High-Resolution Tandem Mass Spectrometry (LC-HRMS/MS) data were analyzed through molecular networking, highlighting cholic acid (CA) and deoxycholic acid. In vitro, CA had inhibitory activity against the translocation of S. Heidelberg by significantly decreasing the expression of Salmonella virulence genes such as sipA. The bioactive fractions also significantly downregulated the flagellar gene fliC, suggesting the involvement of other active molecules. This study showed the interest to characterize better the metabolites produced by B. fragilis to make them means of fighting pathogenic bacteria by targeting their virulence factor without modifying the gut microbiota.

The gut microbiota contributes to human health and disease; however, the mechanisms by which commensal bacteria interact with the host are still unclear. To date, a number of in vitro systems have been designed to investigate the host–microbe interactions. In most of the intestinal models, the enteroendocrine cells, considered as a potential link between gut bacteria and several human diseases, were missing. In the present study, we have generated a new model by adding enteroendocrine cells (ECC) of L-type (NCI-H716) to the one that we have previously described including enterocytes, mucus, and M cells. After 21 days of culture with the other cells, enteroendocrine-differentiated NCI-H716 cells showed neuropods at their basolateral side and expressed their specific genes encoding proglucagon (GCG) and chromogranin A (CHGA). We showed that this model could be stimulated by commensal bacteria playing a key role in health, Roseburia intestinalis and Bacteroides fragilis, but also by a pathogenic strain such as Salmonella Heidelberg. Moreover, using cell-free supernatants of B. fragilis and R. intestinalis, we have shown that R. intestinalis supernatant induced a significant increase in IL-8 and PYY but not in GCG gene expression, while B. fragilis had no impact. Our data indicated that R. intestinalis produced short chain fatty acids (SCFAs) such as butyrate whereas B. fragilis produced more propionate. However, these SCFAs were probably not the only metabolites implicated in PYY expression since butyrate alone had no effect. In conclusion, our new quadricellular model of gut epithelium could be an effective tool to highlight potential beneficial effects of bacteria or their metabolites, in order to develop new classes of probiotics.