Dichotomy in the evolution of pathogenicity island and bacteriophage encoded integrases from pathogenic Escherichia coli strains

摘要

The evolutionary histories of mobile and integrative genetic elements (MIGEs), key factors in the emergence of pathogenic bacteria, remain obscure due to their widespread distribution and diversity in gene content. Pathogenicity islands (PAIs) are large chromosomal regions that encode bacterial virulence factors present in pathogenic isolates and absent from non-pathogenic isolates. We examined the utility of PAL-encoded integrases as a marker to determine PAL phylogeny and evolutionary relationships to other MIGEs particularly bacteriophages. We examined 68 tyrosine recombinase (TR) integrases from 27 PAIs and 39 phages from five fully sequenced pathogenic Escherichia coli strains in the database. In general, all PAI regions identified were integrated adjacent to tRNA loci and had a percent GC content that differed from the host chromosome, which was not the case for most phages examined. Our phylogenetic analysis demonstrated that PAI-encoded integrases clustered within a distinct lineage, separate from phage-encoded integrases, which suggested a discrete evolutionary history. The tree branch lengths among PAI integrases were shorter than those among phage integrases suggesting that PAIs may be a more recent addition to E. coil. Further phylogenetic comparisons of these 68 integrases with 53 TR integrases from other characterized MIGEs also demonstrated that PAI-encode integrases form a distinct lineage, suggesting that these PAIs in E. coli are not remnants of other MIGEs. We identified recombination directionality factors/excisionases in the proximity of many of the E. coil PAI integrases, factors that were previously not shown to be present in E. coil PAIs. Overall this work is the first to demonstrate a dichotomy in the evolution of integrases encoded on PAIs and phages from pathogenic E. coil suggesting that PAIs are an evolutionary distinct genetic element