Evolution of the core and pan-genome of Streptococcus: positive selection, recombination, and genome composition

Abstract

Background: The genus Streptococcus is one of the most diverse and important human and agricultural pathogens. This study employs comparative evolutionary analyses of 26 Streptococcus genomes to yield an improved understanding of the relative roles of recombination and positive selection in pathogen adaptation to their hosts.

Results: Streptococcus genomes exhibit extreme levels of evolutionary plasticity, with high levels of gene gain and loss during species and strain evolution. S. agalactiae has a large pan-genome, with little recombination in its core-genome, while S. pyogenes has a smaller pan-genome and much more recombination of its core-genome, perhaps reflecting the greater habitat, and gene pool, diversity for S. agalactiae compared to S. pyogenes. Core-genome recombination was evident in all lineages (18% to 37% of the core-genome judged to be recombinant), while positive selection was mainly observed during species differentiation (from 11% to 34% of the core-genome). Positive selection pressure was unevenly distributed across lineages and biochemical main role categories. S. suis was the lineage with the greatest level of positive selection pressure, the largest number of unique loci selected, and the largest amount of gene gain and loss.

Conclusion: Recombination is an important evolutionary force in shaping Streptococcus genomes, not only in the acquisition of significant portions of the genome as lineage specific loci, but also in facilitating rapid evolution of the core-genome. Positive selection, although undoubtedly a slower process, has nonetheless played an important role in adaptation of the core-genome of different Streptococcus species to different hosts.

Figure 1

Venn diagram for six sets of three taxa. Above are taxa of the same species and below are taxa of different species. The surfaces are approximately proportional to the number of genes.

Figure 2

Accumulation curves for the total number of genes (left) or the number of genes in common (right) given a number of genomes analyzed for the different species of Streptococcus (in blue), the different strains of S. agalactiae (in red) and S. pyogenes (in green). The vertical bars correspond to standard deviations after repeating one hundred random input orders of the genomes.

Figure 3

Frequency of genes within the 26 genomes included in this analysis. Genes present in a single genome represent lineage specific genes, while at the opposite end of the scale, genes found in all 26 genomes represent the Streptoccocus core-genome.

Figure 4

Gene gain, loss and duplication, and positive selection. Core-genome phylogenies of Streptococcus (left), S. agalactiae (middle), and S. pyogenes (right) based on concatenated genes. Dashed lines correspond to unresolved branches. Numbers adjacent to angle brackets facing the branch refer to genes gained, opposite direction - genes lost, and '×' refers to duplicated loci. Values correspond to the most parsimonious unambiguous changes, following an equally penalized model (that is, gain, loss and duplication events cost the same numbers of changes). Numbers adjacent to the red dot correspond to the number of genes under positive selection within the core-genome, on a particular lineage.

Figure 5

Streptococcus recombination heatmaps. Heatmaps of the (a) AU test, (b) bipartitions bootstrap scores and (c) well supported conflicting bipartitions on the core-genome of Streptococcus. Topologies are ordered from the less rejected (on the left) to the most rejected (on the right). Bipartitions are ordered from the less supported (on the left) to the most supported (on the right), and only bipartitions supported by at least a 70% bootstrap score are represented. Genes are ordered from the less conflicting (left and top) to the most conflicting (right and bottom). The well supported conflicting bipartitions heatmap represents a symmetrical distances matrix, where each cell corresponds to the number of well supported (that is, bootstrap ≥90) conflicting bipartitions between two genes. A color key is given on the right side, and gradations correspond to p values, bootstrap percentages, and number of conflicting bipartitions, left to the right respectively. The arrow locates the concatenated tree.

Figure 6

S. agalactiae recombination heatmaps. The layout is the same as Figure 5 but for the core-genome of S. agalactiae.

Figure 7

S. pyogenes recombination heatmaps. The layout is the same as Figure 5 but for the core-genome of S. pyogenes.

Figure 8

Frequency of positive selection. Numbers of genes showing evidence of positive selection in 1-7 lineages.

Figure 9

Positive selection occurrence per genes and lineages. A black dot indicates positive selection. The genes and lineages were ordered following a correspondence analysis.