Fine carbohydrate structure governs the structure and function of human gut microbiota independently of variation in glycosyl residue composition

Abstract

Increased dietary fiber consumption has been shown to increase human gut microbial diversity, but the mechanisms driving this effect remain unclear. One possible explanation is that microbes are able to divide metabolic labor in consumption of complex carbohydrates, which are composed of diverse glycosidic linkages that require specific cognate enzymes for degradation. However, as naturally derived fibers vary in both sugar composition and linkage structure, it is challenging to separate out the impact of each of these variables. We hypothesized that fine differences in carbohydrate linkage structure would govern microbial community structure and function independently of variation in glycosyl residue composition. To test this hypothesis, we fermented commercially available soluble resistant glucans, which are uniformly composed of glucose linked in different structural arrangements, in vitro with fecal inocula from each of three individuals. We measured metabolic outputs (pH, gas, and short-chain fatty acid production) and community structure via 16S rRNA amplicon sequencing. We determined that community metabolic outputs from identical glucans were highly individual, emerging from divergent initial microbiome structures. However, specific operational taxonomic units responded similarly in growth responses across individuals microbiota, though in context-dependent ways; these data suggested that certain taxa were more efficient in competing for some structures than others. Together, these data support the hypothesis that variation in linkage structure, independent of sugar composition, governs compositional and functional responses of microbiota. ImportancePrevious studies have reported how physical and chemical structures of carbohydrates influence the gut microbiota, however, variability across dietary fibers in monosaccharide composition and linkage structure obscures the relationship between fine polysaccharide linkage structure and microbial fitness. Revealing connections between subtle differences in glucan structure and microbial composition and metabolic responses, this study suggests much greater attention to substrate structure in the design of experiments to test fiber-microbiome responses in vitro and in vivo. Further, it underscores that, although microbiome responses to distinct fibers are individual and vary among specific glucans, similar carbohydrate structure-microbe relationships occur across individual donor communities. Together, these data may help explain why some individuals may respond (while others do not) to fiber treatments in human feeding trials and support the long-term goal of rational inclusion of specific fibers in dietary patterns to modulate the gut microbiome in support of health.