Diet drives convergence in gut microbiome functions across mammalian phylogeny and within humans

Abstract

Coevolution of mammals and their gut microbiota has profoundly affected their radiation into myriad habitats. We used shotgun sequencing of microbial community DNA and targeted sequencing of bacterial 16S ribosomal RNA genes to gain an understanding of how microbial communities adapt to extremes of diet. We sampled fecal DNA from 33 mammalian species and 18 humans who kept detailed diet records, and we found that the adaptation of the microbiota to diet is similar across different mammalian lineages. Functional repertoires of microbiome genes, such as those encoding carbohydrate-active enzymes and proteases, can be predicted from bacterial species assemblages. These results illustrate the value of characterizing vertebrate gut microbiomes to understand host evolutionary histories at a supraorganismal level.

Figure 1. Procrustes analysis shows that mammalian bacterial lineages and metagenomic gene content give similar clustering patterns

(A) Cartoon illustrating Procrustes analysis. The Procrustes transformation of the blue and red data types (i) results in a good fit, while the transformation of the green and red data (ii) yields a worse fit with large distances separating data from samples B and C. (B–E) Procrustes analysis of 16S rRNA sequences (weighted UniFrac) against KEGG Orthology (KO) groups, CAZymes (glycoside hydrolases), peptidases, and E.C.s. Every sphere represents a single mammalian fecal community and is colored by host diet. The black end of each line connects to the 16S data for the sample, while the orange end is connected to the functional annotation data. The fit of each Procrustes transformation over the first three dimensions, is reported as the M² value.

Figure 2. Mammalian gut bacterial communities share a functional core

(A–B) Bipartite network diagrams of evenly sampled bacterial 16S rRNA-derived OTUs (A) or KOs (B). Edges connecting mammalian nodes (circles) to species-level OTUs or KOs found in that sample are colored by host diet. Sample labels are removed from the KO diagram for legibility (high-resolution image of removed labels presented in Fig. S2). (C) Mammalian gut communities share a core suite of KOs. Using evenly subsampled OTU or KO datasets, the distribution of counts is plotted as a function of the number of mammalian host microbiomes where the KO or phylotype was detected. The results demonstrate exponential decay for the 16S rRNA data, with no OTU or bacterial species found in all samples, although a “core” set of KOs is detectable in all fecal communities sampled.

Figure 3. Differences in metabolic features encoded in fecal microbiomes among herbivores versus carnivores

(A) Carnivorous and herbivorous microbiomes indicate opposing directionality for amino acid metabolism. Colored arrows denote enzyme functions significantly (p<0.001) enriched in the fecal microbiomes of herbivores (green) or carnivores (red). (B) Carnivorous and herbivorous microbiomes suggest opposing directionality at the central PEP-Pyruvate-Oxaloacetate node. Coloring scheme as in panel A. Abbreviations: 2-OG, alpha-ketoglutarate; GABA, γ–aminobutyrate; DH, Dehydrogenase; OAA, Oxaloacetate; Dx, Decarboxylase; Ck, Carboxykinase; Cx, Carboxykinase; PEP, Phosphoenolpyruvate; Pyr, Pyruvate; SSA, Succinate-Semialdehyde.