Prebiotics and Community Composition Influence Gas Production of the Human Gut Microbiota

Abstract

Prebiotics confer benefits to human health, often by promoting the growth of gut bacteria that produce metabolites valuable to the human body, such as short-chain fatty acids (SCFAs). While prebiotic selection has strongly focused on maximizing the production of SCFAs, less attention has been paid to gases, a by-product of SCFA production that also has physiological effects on the human body. Here, we investigate how the content and volume of gas production by human gut microbiota are affected by the chemical composition of the prebiotic and the community composition of the microbiota. We first constructed a linear system model based on mass and electron balance and compared the theoretical product ranges of two prebiotics, inulin and pectin. Modeling shows that pectin is more restricted in product space, with less potential for H₂ but more potential for CO₂ production. An ex vivo experimental system showed pectin degradation produced significantly less H₂ than inulin, but CO₂ production fell outside the theoretical product range, suggesting fermentation of fecal debris. Microbial community composition also impacted results: methane production was dependent on the presence of Methanobacteria, while interindividual differences in H₂ production during inulin degradation were driven by a Lachnospiraceae taxon. Overall, these results suggest that both the chemistry of the prebiotic and the composition of the microbiota are relevant to gas production. Metabolic processes that are relatively prevalent in the microbiome, such as H₂ production, will depend more on substrate, while rare metabolisms such as methanogenesis depend more strongly on microbiome composition.IMPORTANCE Prebiotic fermentation in the gut often leads to the coproduction of short-chain fatty acids (SCFAs) and gases. While excess gas production can be a potential problem for those with functional gut disorders, gas production is rarely considered during prebiotic design. In this study, we combined the use of theoretical models and an ex vivo experimental platform to illustrate that both the chemical composition of the prebiotic and the community composition of the human gut microbiota can affect the volume and content of gas production during prebiotic fermentation. Specifically, more prevalent metabolic processes such as hydrogen production were strongly affected by the oxidation state of the probiotic, while rare metabolisms such as methane production were less affected by the chemical nature of the substrate and entirely dependent on the presence of Methanobacteria in the microbiota.

Keywords: functional heterogeneity; gut microbiome; intestinal gas; prebiotics.

FIG 1

Modeling community production with mass and electron balance. (a) Detailed representation of the theoretical model Ms = 0, where M is a matrix with n chemicals (j inputs and k products). In M, inputs are represented in negative numbers, while outputs are represented in positive numbers. Each row in M represents an element (or electrons) that needs to be balanced. (b) The specific M matrix corresponding to our system of interest, fermentation of two different fibers, inulin and pectin. (c) Set of selected 2D projections of the feasible product space predicted from our theoretical model for the fermentation of 1 mol inulin or 1 mol pectin. See Fig. S1 in the supplemental material for the full set of 2D projections for all fermentation products.

FIG 2

Inulin fermentation produces more H₂ and less acetate than pectin in an ex vivo system. (a) Experimental scheme for studying fiber fermentation products in an ex vivo system. (b) Major product concentrations in the ex vivo system after 24 h measured as moles product production per mole of fiber. **, P < 0.01 for paired Kruskal-Wallis test; ns, not significant.

FIG 3

Pectin degradation requires the uptake of reducing agents. (a) Comparison of product measurements in the ex vivo system to theoretically predicted feasible product ranges. (b) Illustration of the two subsystems within the serum bottle and their material exchanges. (c) Comparison of product measurements in the ex vivo system to theoretically predicted feasible product ranges in a closed system and when allowing different inputs. Error bars in panels a and c represent standard deviations from biological replicates.

FIG 4

Principal-coordinates analysis (PCoA) of the Euclidian distance matrix of the fiber fermentation products from the ex vivo system. Each point corresponds to a community colored according to the fiber added to the system. Each point is labeled by the source (human subject [H]) from which the stool sample was collected.

FIG 5

Different gases are differentially influenced by substrate chemistry and gut microbiome composition. (a) Relationship between the observed amount of H₂ production per mole of inulin in the ex vivo system and the relative abundance of the Lachnospiraceae ASV selected by the Lasso regression. Error bars represent standard deviations from biological replicates (n = 3). (b) Relationship between the production of methane per mole of fiber in the ex vivo system and the relative abundance of Methanobacteria in the samples. Error bars represent standard deviations from biological replicates (n = 3). (c) Schematic of fiber degradation and production of gas and SCFAs. (d) Distributions of the abundances of the methanogenesis marker gene mcrA, hydrogenic hydrogenases, and CAZymes in the metagenomes of 160 different people in the HMP data set. All gene counts were increased by 10⁻⁵ so that the log-scaled x axis could accommodate samples with zero hits.