Gut microbial activity as influenced by fiber digestion: dynamic metabolomics in an in vitro colon simulator

Understanding the interaction between the gut microbial activity and the host is essential, and in vitro models are being used to test and develop hypotheses regarding the impact of food components/drugs on the human gut ecosystem. However, while in vitro models provide excellent possibilities for dynamic investigations, studies have commonly been restricted to analyses of few, targeted metabolites. In the present study, we employed NMR-based metabolomics combined with multilevel data analysis as a tool to characterize the impact of polydextrose (PDX) fiber on the in vitro derived fecal metabolome. This approach enabled us to identify and quantify the fiber-induced response on several fecal metabolites; we observed higher levels of butyrate, acetate, propionate, succinate, N -acetyl compound and a lower level of amino acids (leucine, valine, isoleucine, phenylalanine, and lysine), valerate, formate, isovalerate and trimethylamine among the PDX-treated sample compared to the control samples. In addition, by the application of multilevel data analysis we were able to examine the specific inter-individual variations, and caprylic acid was identified to be the main marker of distinct microbial compositions among the subjects. Our work is expected to provide a useful approach to understand the metabolic impact of potential prebiotic compounds and get deeper insight into the molecular regulation of gut-microbe activities in the complex gut system.