Genome sequencing of environmental Escherichia coli expands understanding of the ecology and speciation of the model bacterial species

Abstract

Defining bacterial species remains a challenging problem even for the model bacterium Escherichia coli and has major practical consequences for reliable diagnosis of infectious disease agents and regulations for transport and possession of organisms of economic importance. E. coli traditionally is thought to live within the gastrointestinal tract of humans and other warm-blooded animals and not to survive for extended periods outside its host; this understanding is the basis for its widespread use as a fecal contamination indicator. Here, we report the genome sequences of nine environmentally adapted strains that are phenotypically and taxonomically indistinguishable from typical E. coli (commensal or pathogenic). We find, however, that the commensal genomes encode for more functions that are important for fitness in the human gut, do not exchange genetic material with their environmental counterparts, and hence do not evolve according to the recently proposed fragmented speciation model. These findings are consistent with a more stringent and ecologic definition for bacterial species than the current definition and provide means to start replacing traditional approaches of defining distinctive phenotypes for new species with omics-based procedures. They also have important implications for reliable diagnosis and regulation of pathogenic E. coli and for the coliform cell-counting test.

Fig. 1.

Whole-genome phylogeny of the Escherichia genomes used in the study. The phylogenetic network shown was constructed with the SplitsTree software (27), using as input the concatenated alignment of 1,910 single-copy core genes. (Inset) The graph represents the amount of recent horizontal transfer of core genes between the genomes of the clades. The thickness of the line is proportional to the number of genes transferred (scale at upper left in figure).

Fig. 2.

Gene-content signatures of Escherichia clades. Heatmap of gene presence (yellow) and absence (blue) in 20 selected genomes, using all nonredundant genes that were found in at least two of the genomes as reference. (A) Genomes were clustered based on the presence/absence of genes; values in red represent bootstrap support from Jackknifing resampling with 100 replicates. (B) Genes and pathways distinguishing enteric and environmental genomes were expanded (underlying data are provided in Table S2). 1, acetylglucosamine transporter; 2, fructose transporter; 3, diol utilization operon; 4, lysozyme production. (C) Occurrence of the genes composing the Escherichia pangenome in the 20 genomes ranges from one (a genome-specific gene) to 20 (a core gene).

Fig. 3.

Lack of evidence in support of the fragmented speciation model. A representative example of the nucleotide divergence patterns, measured as the number of SNPs (y axis) observed in the flanking regions of a genomic island (x axis) that differentiates environmental from enteric genomes. Note the difference between the SNP level expected for the environmental genomes according to the fragmented speciation model (ecologically distinct population) relative to the observed level (for the enteric group, the genome average SNP level represents the expected SNP level according to the model). The island shown encodes the genes for utilization of fucose, a sugar commonly found in the glycan structures of the cell wall of animals.