Impact of Homologous Recombination on the Evolution of Prokaryotic Core Genomes

Abstract

Homologous recombination (HR) enables the exchange of genetic material between and within species. Recent studies suggest that this process plays a major role in the microevolution of microbial genomes, contributing to core genome homogenization and to the maintenance of cohesive population structures. However, we still have a very poor understanding of the possible adaptive roles of intraspecific HR and of the factors that determine its differential impact across clades and lifestyles. Here we used a unified methodological framework to assess HR in 338 complete genomes from 54 phylogenetically diverse and representative prokaryotic species, encompassing different lifestyles and a broad phylogenetic distribution. Our results indicate that lifestyle and presence of restriction-modification (RM) machineries are among the main factors shaping HR patterns, with symbionts and intracellular pathogens having the lowest HR levels. Similarly, the size of exchanged genomic fragments correlated with the presence of RM and competence machineries. Finally, genes exchanged by HR showed functional enrichments which could be related to adaptations to different environments and ecological strategies. Taken together, our results clarify the factors underlying HR impact and suggest important adaptive roles of genes exchanged through this mechanism. Our results also revealed that the extent of genetic exchange correlated with lifestyle and some genomic features. Moreover, the genes in exchanged regions were enriched for functions that reflected specific adaptations, supporting identification of HR as one of the main evolutionary mechanisms shaping prokaryotic core genomes.IMPORTANCE Microbial populations exchange genetic material through a process called homologous recombination. Although this process has been studied in particular organisms, we lack an understanding of its differential impact over the genome and across microbes with different life-styles. We used a common analytical framework to assess this process in a representative set of microorganisms. Our results uncovered important trends. First, microbes with different lifestyles are differentially impacted, with endosymbionts and obligate pathogens being those less prone to undergo this process. Second, certain genetic elements such as restriction-modification systems seem to be associated with higher rates of recombination. Most importantly, recombined genomes show the footprints of natural selection in which recombined regions preferentially contain genes that can be related to specific ecological adaptations. Taken together, our results clarify the relative contributions of factors modulating homologous recombination and show evidence for a clear a role of this process in shaping microbial genomes and driving ecological adaptations.

Keywords: comparative genomics; core genomes; genome evolution; homologous recombination; intraspecific diversity.

FIG 1

Data set composition. A 16S rRNA phylogenetic tree of the 54 prokaryotic species included in this study is shown. The tree was drawn using itool (https://itol.embl.de/). The innermost circle layer shows the species nape and associated clade. Analysis of HR events was performed. The innermost layer indicates the competence capability. The second (i.e., contiguous) layer, third layer, and fourth layer correspond to the number of HR events per strain, the proportion (%) of recombined genome, and the size distribution (%) of HR events.

FIG 2

HR characteristics and lifestyle effect. Four lifestyles are represented in all the figures by the same color code: black, endosymbionts and intracellular pathogens; blue, opportunistic pathogens; red, commensal and free-living pathogens; green, obligate pathogens (green). (A) GC content among 54 species included in this study distributed in 4 lifestyles. (B) Box plot comparing average levels of GC content in recombinant events (solid color; right paired boxes) and whole genomes (grayed-out color; left paired boxes). (C and D) HR distribution (events/strain) (C) and proportion of genome recombined (D) based on lifestyle distributions (both P < 0.05 [Kruskal-Wallis and Jonkheere-Tepstra tests]).

FIG 3

Effect of genomic variables on HR distribution. (A) Proportion (%) of genome recombined based on competence capabilities (Com+/Com−). (B) HR event fragment size distribution based on competence capabilities: competent (Com+) (solid color, right paired boxes) and noncompetent (Com−) (grayed out color, left paired boxes). (C and D) Genomic island (GI) distributions (%), com gene content (C), and type I restriction modification system (RM-I) gene content distribution (D) based on the different lifestyles considered. Four lifestyles are represented with the following color code: black, endosymbionts and intracellular pathogens; blue, opportunistic pathogens; red, commensal and free-living pathogens; green, obligate pathogens (green).

FIG 4

Influence of lifestyle and role of HR in population structure and evolution. (A) Correlation of r/m and rho/theta ratios (r/m of >1 [blue] or <1 [green]) for 54 species and (B) distribution based on core genome length and core genome identity (ANIb).

FIG 5

General model. A scheme is presented of a path analysis model proposed for the analysis of the influence of lifestyles, phylogeny, barrier/motility genomics variables, and genomics characteristics on HR levels detected among 54 species analyzed. Significant r values (*, P < 0.05; **, P < 0.01; ***, P < 0.001) for partial comparisons carried out during Mantel test are indicated. Arrows indicate relationships between variables, with the thickness of the arrows being proportional to the correlation between the connected variables.

FIG 6

Gene flow and adaptive implications. (A) Heat map for similarity matrix representing the grouping of 54 species analyzed based on their profile and on those with significantly enriched (green) or underrepresented (red) genes (P < 0.05 [Fisher’s test; FDR correction, <0.1]). The x axis shows the layout of the functional categories and the y axis the species analyzed. (B to D) Distribution (%) of the most abundant GO terms among HR events and associated with (B) “Information, processing and storage and cellular processes” COG categories (red), (C) “Cellular processes and signaling” COG categories (dark blue), and (D) “Metabolism” COG categories (red). GO terms that presented significant enrichment or underrepresentation are marked with an asterisk (*) (P < 0.05, pFDR < 0.05 [Fisher’s exact test]).