Structure, diversity, and mobility of the Salmonella pathogenicity island 7 family of integrative and conjugative elements within Enterobacteriaceae

Abstract

Integrative and conjugative elements (ICEs) are self-mobile genetic elements found in the genomes of some bacteria. These elements may confer a fitness advantage upon their host bacteria through the cargo genes that they carry. Salmonella pathogenicity island 7 (SPI-7), found within some pathogenic strains of Salmonella enterica, possesses features indicative of an ICE and carries genes implicated in virulence. We aimed to identify and fully analyze ICEs related to SPI-7 within the genus Salmonella and other Enterobacteriaceae. We report the sequence of two novel SPI-7-like elements, found within strains of Salmonella bongori, which share 97% nucleotide identity over conserved regions with SPI-7 and with each other. Although SPI-7 within Salmonella enterica serovar Typhi appears to be fixed within the chromosome, we present evidence that these novel elements are capable of excision and self-mobility. Phylogenetic analyses show that these Salmonella mobile elements share an ancestor which existed approximately 3.6 to 15.8 million years ago. Additionally, we identified more distantly related ICEs, with distinct cargo regions, within other strains of Salmonella as well as within Citrobacter, Erwinia, Escherichia, Photorhabdus, and Yersinia species. In total, we report on a collection of 17 SPI-7 related ICEs within enterobacterial species, of which six are novel. Using comparative and mutational studies, we have defined a core of 27 genes essential for conjugation. We present a growing family of SPI-7-related ICEs whose mobility, abundance, and cargo variability indicate that these elements may have had a large impact on the evolution of the Enterobacteriaceae.

Fig 1

Distribution of conserved and cargo CDSs in ICEs. The presence of conserved CDSs is indicated by green boxes; the location and number of cargo CDSs are shown in orange boxes. White boxes indicate the absence of a CDS in that ICE. Some CDSs are shared between a subset of ICEs (yellow), pseudogenes (Ψ) are shown in brown, and translocations/inversions are indicated in blue. The reference strains used are S. Typhi strain CT18 and S. bongori CEIM46082, and CDS numbers refer to the STY and Sb1 locus numbers, respectively. Of the putative core CDSs used in knockout studies (*), 27 were used in phylogenetic analysis (†). Where the specific integration site (tRNA) could be determined, it is indicated. ICEHin1056 and PAPI-1, previously characterized ICEs from Haemophilus influenzae and Pseudomonas aeruginosa, respectively, which display synteny and a low level of amino acid identity to SPI-7, were used in the analysis as nonenterobacterial outgroups.

Fig 2

Conjugation frequencies of gene knockouts compared to the wild-type ICESb1. Values are given as a percentage of the wild-type frequency, with error bars given as a percentage of the total from triplicate samples. W.T. refers to ICESb1-cat.

Fig 3

Comparative phylogenetic analyses of concatenated MLST (top), SPI-1 (middle), and ICE (bottom) nucleotide sequences for SPI-7, ICESb1, and ICESb2 and their host strains. Trees were produced by maximum likelihood analysis. Dashed lines indicate the scale of 0.01 substitutions per site, and dotted lines indicate 0.0025 substitutions per site. Bootstrapping values are 78% to 100% for the ICEs and are 100% on all branches for SPI-1 and MLST, except the S. Paratyphi C and S. Dublin branches, which are 66% and 65% for the SPI-1 and MLST trees, respectively.

Fig 4

Phylogenetic analysis of ICE core CDSs compared to strain phylogeny. Comparison of tree topologies produced by maximum likelihood analysis of concatenated ICE core CDS amino acid sequences (left) and 16S rRNA gene sequences to indicate host phylogenetic affiliations (right). Dotted gray lines join the ICE to the relevant 16S rRNA gene sequence of the host. Bootstrapping of the ICE phylogeny gave values shown on branches, with bootstrap values between the SPI-7 variants of 94% and 97%. The scale bars indicate the number of substitutions per amino acid/nucleotide position. This family contains closely related groups of ICEs as well as some highly divergent family members. The root of the tree is defined by the ICEHin1056 and PAPI-1 outgroups.