Members of the gut microbiota are thought to experience strong competition for nutrients. However, how such competition shapes their evolutionary dynamics and depends on intraand interspecies interactions is poorly known. Here we tested the hypothesis that Escherichia coli evolution in the mouse gut is more predictable across hosts in absence of interspecies competition than in the presence of other microbial species. Supporting this hypothesis, we observed a specific genetic adaptation in lrp, a gene encoding a global regulator of amino acid metabolism, predictably selected in germ-free mice two weeks after mono-colonization. Analysis of gut metabolites established that the lpr mutations increase E. coli ability to compete for amino acids and identified serine and threonine as the metabolites preferentially consumed by E. coli in the mono-colonized mouse gut. Preference for serine consumption was further demonstrated by testing a set of mutants in vitro and in vivo that showed loss of advantage of a lrp mutant impaired in serine metabolism. Remarkably, the presence of a single additional member of the microbiota (Blautia coccoides) was enough to alter the gut metabolic profile and consequently the evolutionary path of E. coli. In this environment, the lrp mutations did not conferred advantage to E. coli and genes involved in anaerobic respiration were selected instead, recapitulating the ecoevolutionary context from mice with a complex microbiota. Together, these results highlight the metabolic plasticity of E. coli and its extreme evolutionary versatility, tailored to the specific ecology it experiences in the gut.

No abstract

No abstract

SummaryDue to limitations on high-resolution strain tracking, selection dynamics during gut-microbiota colonization and transmission between hosts remain mostly mysterious. Here, we introduced hundreds of barcoded Escherichia coli strains into germ-free mice and quantified strain-level dynamics and metagenomic changes. Mutants involved in motility and utilization of abundant metabolites were reproducibly selected within days. Even with rapid selection, coprophagy enforced similar barcode distributions across co-housed mice. Whole-genome sequencing of hundreds of isolates quantified evolutionary dynamics and revealed linked alleles. A population-genetics model predicted substantial fitness advantages for certain mutants and that migration accounted for ~10% of the resident microbiota each day. Treatment with ciprofloxacin demonstrated the interplay between selection and transmission. While initial colonization was mostly uniform, in two mice a bottleneck reduced diversity and selected for ciprofloxacin resistance in the absence of drug. These findings highlight the interplay between environmental transmission and rapid, deterministic selection during evolution of the intestinal microbiota.

Many microbes colonise the gut establishing interactions with their host and their nutritional environment. Studying genetics and metabolism brought about the drive and potential to engineer communities to promote health and improve industrial processes. However, structuring artificial communities in predictable ways is underdeveloped. We studied Escherichia coli’s genetic targets and physiological mechanisms during gut colonisation and adaptation and how metabolic environment/microbiota complexity shape these processes. We introduced a tractable E. coli K-12 in mice Germ-free or with polymicrobial communities. Whole Genome Sequencing identified potential adaptive targets. Here, we established phenotypic assays as well characterising effects of key mutations and metabolomics was performed with 1H-NMR of intestinal contents. Genes for sugar alcohol metabolism (gat) was the only target common to both mouse models, evidencing specificity. Facing complex microbiota E. coli targeted use of sugar alcohols (srlR, kdgR) and anaerobic respiration (dcuB, focA) [1] whereas alone, we observed instead mutations pointing to increased ability for amino acid use (lrp, dtpB, alaA). Mutations selected correlated dinamically with metabolomics: our results fit the model whereby other microbiota members scavenge oxygen and breakdown complex sugars, limiting E. coli to anaerobically respire simple by-product carbon sources. In the opposing scenario (functional absence) improved amino acid use are favoured colonisation factors. Through experimental evolution we gained insight on shaping E. coli’s metabolic traits through genetic engineering to colonise specific host environments. This work also highlights the versatility of E. coli as potential biotic sensor. [1] Barroso-Batista, J. et al. The first steps of adaptation of Escherichia coli to the gut are dominated by soft sweeps, 2014.