Quantifying adaptive landscapes of commensal gut bacteria using high-resolution lineage tracking

Gut microbiota can adapt to their host environment by rapidly acquiring new mutations. However, the evolutionary dynamics of process are difficult to characterize in dominant gut species in their complex in vivo environment. In this talk, I will show how the fine-scale dynamics of genome-wide transposon libraries can enable quantitative inferences of these in vivo evolutionary forces. By modeling the collective behavior of >400,000 lineages across four human gut bacteria in germ-free mice, we detected positive selection on thousands of previously hidden mutations – most of which were unrelated to their original gene knockouts. We found that the spectrum of fitness benefits varied between closely related gut species, and displayed diverse tradeoffs over time and in different dietary conditions. However, our broader characterization revealed few global tradeoffs in the underlying fitness landscapes. This suggests that long-term fitness tradeoffs under different diets may not reflect a fundamental physiological constraint, but simply the entropic sampling of the adaptive landscape.