Development of a cell-based in vitro gut model to study the impact of microbiome-derived metabolites : MetaboGut: An in vitro intestinal epithelial model to study microbiome-derived metabolites

Abstract

Gut microbiome-derived metabolites are key regulators of host-microbe crosstalk’s. Several classes of microbial metabolites, such as microbiome-conjugated bile acids and N-acyl lipids, have emerged as active biomolecules that may influence intestinal physiology. However, we still do not how these emerging microbial metabolites impact host cells. The aim of this study was to develop an in vitro intestinal epithelial model using Transwell to study the effects of type 1 diabetes (T1D)-associated fecal metabolite extracts and bile acids on epithelial differentiation, tight junction organization, and apical and basolateral metabolite profiles from the Transwell. Caco-2 and Caco-2:HT29 2D models were developed and treated with fecal metabolite extracts from T1D, healthy control, ursodeoxycholic acid, and ursodeoxycholic acid–asparagine and vehicle control. Immunofluorescent staining was used to assess the tight junction organization and epithelial differentiation using zonula occludens-1 (ZO-1) and villin-1 (VIL1) markers, while apical and basolateral samples were analysed using untargeted liquid chromatography–tandem mass spectrometry (LC-MS/MS). Mucin-2 (MUC2) PCR was also performed to assess mucus presence in the co-culture model. The cell models formed epithelial monolayers, and strong VIL1 staining indicates epithelial differentiation. However, ZO-1 staining was weak or inconsistent. Untargeted LC–MS/MS analysis revealed a trend of lower abundance of basolateral lipid annotated features across conditions, except for certain lipid-related metabolites, including diacylglycerol (16:0/18:0/0:0), which were elevated in the vehicle control and T1D conditions. However, no statistically significant annotated features were identified by untargeted metabolomics in faecal and biletreated cell models. MUC2 PCR and gel electrophoresis did not show a clear expected band. In conclusion, this study established a preliminary in vitro intestinal model to study microbiomederived metabolites, and metabolites from T1D-associated fecal metabolite extracts that may influence the intestinal epithelial barrier. However, further optimization and functional validation of developed model are required to study the impact of metabolites on intestinal barrier integrity.