Predicting butyrate- and propionate-forming bacteria of gut microbiota from sequencing data

Abstract

BackgroundThe bacteria-derived short chain fatty acids (SCFAs) butyrate and propionate play important (distinct) roles in health and disease and understanding the ecology of respective bacteria on a community-wide level is a top priority in microbiome research. The aim of this study was to reveal members harboring main pathways for the production of those metabolites and assess the applicability of sequence data (metagenomics and 16S rRNA gene) to predict SCFAs production in vitro and in vivo. ResultsA clear split between butyrate- and propionate-forming bacteria was detected with only very few taxa exhibiting pathways for the production of both SCFAs. After in vitro growth of fecal communities from distinct donors (n=8) on different substrates (n=7) abundances of bacteria exhibiting pathways correlated with respective SCFA concentrations, in particular in the case of butyrate. While final growth differed markedly between cultures, communities showed high functional redundancies with comparable yields, i.e., concentration of metabolite per grown bacterium exhibiting pathway(s), irrespective of the donor and substrate used. For propionate, correlations were weaker indicating that its production is less imprinted into the core metabolism compared with butyrate-forming bacteria. Longitudinal measurements in vivo (five samples derived from 20 subjects) also revealed a correlation between abundances of pathway-carrying bacteria and concentrations of the two SCFAs. Additionally, lower bacterial cell concentrations, together with higher stool moisture, promoted overall bacterial activity (measured by flow cytometry and coverage patterns of metagenome-assembled genomes) that led to elevated SCFAs concentrations with over-proportional levels of butyrate. Butyrate concentrations displayed lower temporal stability than propionate, however, abundances of bacteria exhibiting the butyrate-forming pathway were more stable than those carrying pathways for propionate production. Predictions on pathway abundances based on 16S rRNA gene data using our in-house database worked well yielding similar results as metagenomic-based analyses. ConclusionsWe demonstrated that pathway abundances enable predictions on concentrations of SCFAs indicating that stimulating bacterial growth directly leads to more production of those compounds. The strong separation of gut microbiota into two functional communities facilitates the development of precision intervention strategies targeting either metabolite.