Dynamics of bacterial recombination in the human gut microbiome

Abstract

Horizontal gene transfer is a ubiquitous force in microbial evolution. Previous studies have shown that the human gut is a hot spot for gene transfer between species, but the more subtle exchange of variation within species — collectively known as recombination — remains poorly characterized in this ecosystem. Here, we show that the genetic structure of the human gut microbiome provides unique opportunities to measure individual recombination events directly, enabling quantitative comparisons of recombination across a diverse range of species that inhabit a common environment. By analyzing a large collection of recent recombination events in the core genomes of 29 gut commensals, we uncovered systematic heterogeneities in the rates and lengths of transferred segments, which are difficult to explain by existing models of ecological isolation or reduced efficiencies of recombination. We also find that natural selection plays a role in facilitating the spread of genetic variants onto different strain backgrounds, both within individual hosts and across the broader global population. These results shed light on the dynamics of in situ recombination, which place important constraints on the adaptability of gut microbial communities.

Linkage equilibrium between rare mutations

“…9; Appendix H). Previous estimates of LD in this species have suggested that E. rectale strains in different hosts experience high rates of homologous recombination (Good, 2022; Liu and Good, 2024), which provides a natural opportunity for exploring the homoplasy metrics discussed above. The core genomes of these strains contained a total of 338,669 synonymous single-nucleotide variants (SNVs), the vast majority of which were rare (median minor allele frequency of 0.6%).…”

“…We reasoned that at larger coordinate distances, the gene synteny of individual strains might start to differ from the E. rectale reference genome, which could make it harder to identify the precise distance between sites. To include pairs of sites with larger recombination rates, we recorded analogous haplotype counts between 10 3 pairs of randomly selected genes. These randomly sampled pair of genes are typically separated by hundreds of kilobases in the reference genome, which is longer than the typical recombination length in bacteria (Liu and Good, 2024). This suggests that their recombination rate should approach a constant value, R max = lim 𝓁→∞ R (𝓁).…”

“…To include pairs of sites with larger recombination rates, we recorded analogous haplotype counts between 10 3 pairs of randomly selected genes. These randomly sampled pair of genes are typically separated by hundreds of kilobases in the reference genome, which is longer than the typical recombination length in bacteria (Liu and Good, 2024). This suggests that their recombination rate should approach a constant value, R max = lim 𝓁→∞ R (𝓁).…”

“…Chief among these is recombination, which breaks up genetic linkage by reshuffling existing variants onto new genetic backgrounds. Linkage disequilibrium has played a central role in illuminating the recombination dynamics of natural populations, from fine-scale recombination maps in sexual organisms (Chan et al ., 2012; Coop et al ., 2008; McVean et al ., 2004; Myers et al ., 2005; Spence and Song, 2019) to the rates of horizontal gene transfer in bacteria (Didelot and Falush, 2007; Didelot and Wilson, 2015; Garud et al ., 2019; Lin and Kussell, 2017; Liu and Good, 2024; Rosen et al ., 2015), viruses (Neher and Leitner, 2010; Romero and Feder, 2024; Turakhia et al ., 2022; Zanini et al ., 2015), and other microbes (Lynch et al ., 2022; Vakhru-sheva et al ., 2020). In addition to recombination, LD also encodes important information about the demographic history of a population (Li and Durbin, 2011; Ragsdale and Gravel, 2019; Ragsdale et al ., 2023; Santiago et al ., 2020) and the action of positive (Garud et al ., 2015; Sabeti et al ., 2002; Stephan et al ., 2006; Wolff and Garud, 2023) or negative (Corbett-Detig et al ., 2013; Garcia and Lohmueller, 2021; Ragsdale, 2022; Sohail et al ., 2017) selection.…”

Inference of the demographic histories and selective effects of human gut commensal microbiota over the course of human history

“…One important assumption of applying an SFS-based analysis is quasi-independence between sites, i.e., free recombination between sites (Bustamante et al 2001, Genetics). While bacteria reproduce asexually, most commensal gut bacteria experience extensive recombination (Vos and Didelot 2009, ISME; Liu and Good 2023, bioRxiv; Garud and Good et al 2019, PLoS Biology; Lin and Kussel 2019, Nature Methods; Smith et al 1993, PNAS). This recombination plays a critical role in shuffling genetic material and breaking up correlations between loci in the genome on long timescales exceeding within-host evolution.…”