• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of plosgenPLoS GeneticsSubmit to PLoSGet E-mail AlertsContact UsPublic Library of Science (PLoS)View this Article
PLoS Genet. Dec 2009; 5(12): e1000786.
Published online Dec 24, 2009. doi:  10.1371/journal.pgen.1000786
PMCID: PMC2791158

Comparative ICE Genomics: Insights into the Evolution of the SXT/R391 Family of ICEs

Ivan Matic, Editor

Abstract

Integrating and conjugative elements (ICEs) are one of the three principal types of self-transmissible mobile genetic elements in bacteria. ICEs, like plasmids, transfer via conjugation; but unlike plasmids and similar to many phages, these elements integrate into and replicate along with the host chromosome. Members of the SXT/R391 family of ICEs have been isolated from several species of gram-negative bacteria, including Vibrio cholerae, the cause of cholera, where they have been important vectors for disseminating genes conferring resistance to antibiotics. Here we developed a plasmid-based system to capture and isolate SXT/R391 ICEs for sequencing. Comparative analyses of the genomes of 13 SXT/R391 ICEs derived from diverse hosts and locations revealed that they contain 52 perfectly syntenic and nearly identical core genes that serve as a scaffold capable of mobilizing an array of variable DNA. Furthermore, selection pressure to maintain ICE mobility appears to have restricted insertions of variable DNA into intergenic sites that do not interrupt core functions. The variable genes confer diverse element-specific phenotypes, such as resistance to antibiotics. Functional analysis of a set of deletion mutants revealed that less than half of the conserved core genes are required for ICE mobility; the functions of most of the dispensable core genes are unknown. Several lines of evidence suggest that there has been extensive recombination between SXT/R391 ICEs, resulting in re-assortment of their respective variable gene content. Furthermore, our analyses suggest that there may be a network of phylogenetic relationships among sequences found in all types of mobile genetic elements.

Author Summary

Integrative and conjugative elements (ICEs) are a class of mobile genetic elements that are key mediators of horizontal gene flow in bacteria. These elements integrate into the host chromosome, yet are able to excise and transfer via conjugation. Our understanding of ICE evolution is rudimentary. Here, we developed a method to capture ICEs on plasmids, thus facilitating their sequencing. Comparative analyses of the DNA sequences of ICEs from the same family revealed that they have an identical genetic structure consisting of syntenous, highly conserved core genes that are interrupted by clusters of diverse variable genes. Unexpectedly, many genes in the core backbone proved non-essential for ICE transfer. Comparisons of the variable gene content in the ICEs analyzed revealed that these elements are mosaics whose genomes have been shaped by inter–ICE recombination. Finally, our work suggests that ICEs contribute to a larger gene pool that connects all types of mobile elements.

Introduction

There are three types of self-transmissible mobile genetic elements: plasmids, bacteriophages and integrative conjugative elements (ICEs). All three classes of elements enable horizontal transmission of genetic information and all have had major impacts on bacterial evolution [1][4]. ICEs, (aka conjugation transposons), like plasmids, are transmitted via conjugation; however, unlike plasmids, ICEs integrate into and replicate along with the chromosome. Following integration, ICEs can excise from the chromosome and form circular molecules that are intermediates in ICE transfer. Plasmids and phages have been the subject of more extensive study than ICEs and while there is growing understanding of the molecular aspects of several ICEs [5][10], to date there have been few reports of comparative ICE genomics [11],[12] and consequently understanding of ICE evolution is only beginning to be unraveled.

Diverse ICEs have been identified in a variety of gram-positive and gram–negative organisms [13]. These elements utilize a variety of genes to mediate the core ICE functions of chromosome integration, excision and conjugation. In addition to a core gene set, ICEs routinely contain genes that confer specific phenotypes upon their hosts, such as resistance to antibiotics and heavy metals [14][18], aromatic compound degradation [19] or nitrogen fixation [20].

SXT is an ~100 Kb ICE that was originally discovered in Vibrio cholerae O139 [16], the first non-O1 serogroup to cause epidemic cholera [21]. SXT encodes resistances to several antibiotics, including sulfamethoxazole and trimethoprim (which together are often abbreviated as SXT) that had previously been useful in the treatment of cholera. Since the emergence of V. cholerae O139 on the Indian subcontinent in 1992, SXT or a similar ICE has been found in most clinical isolates of V. cholerae, including V. cholerae serogroup O1, from both Asia and Africa. Other vibrio species besides V. cholerae have also been found to harbor SXT-related ICEs [22]. Furthermore, SXT-like ICEs are not restricted to vibrio species, as such ICEs have been detected in Photobacterium damselae, Shewanella putrefaciens and Providencia alcalifaciens [23][25]. Moreover, Hochhut et al [26] found that SXT is genetically and functionally related to the so-called ‘Inc J’ element R391, which was derived from a South African Providencia rettgeri strain isolated in 1967 [27]. It is now clear that Inc J elements are SXT-related ICEs that were originally misclassified as plasmids. In the laboratory, SXT has a fairly broad host range and can be transmitted between a variety of gram-negative organisms [16].

The SXT/R391 family of ICEs is now known to include more than 30 elements that have been detected in clinical and environmental isolates of several species of γ- proteobacteria from disparate locations around the globe [28]. SXT/R391 ICEs are grouped together as an ICE family because they all encode a nearly identical integrase, Int. Int, a tyrosine recombinase, is considered a defining feature of these elements because it enables their site-specific integration into the 5′ end of prfC, a conserved chromosomal gene that encodes peptide chain release factor 3 [29]. Int mediates recombination between nearly identical element and chromosome sequences, attP and attB respectively [29]. When an SXT/R391 ICE excises from the chromosome, Int, aided by Xis, a recombination directionality factor, mediates the reverse reaction - recombination between the extreme right and left ends (attR and attL) of the integrated element - thereby reconstituting attP and attB [6],[29]. The excised circular SXT form is thought to be the principal substrate for its conjugative transfer. The genes that encode activities required for SXT transfer (tra genes) were originally found to be distantly related to certain plasmid tra genes [30][32]. The tra genes encode proteins important for processing DNA for transfer, mating pair formation and generating the conjugation machinery. Regulation of SXT excision and transfer is at least in part governed by a pathway that resembles the pathway governing the lytic development of the phage lambda. Agents that damage DNA and induce the bacterial SOS response are thought to stimulate the cleavage and inactivation of SetR, an SXT encoded λ cI-related repressor, which represses expression of setD and setC, transcription activators that promote expression of int and tra genes [5].

The complete nucleotide sequences of SXT (99.5kb) and R391 (89kb) were the first SXT/R391 ICE family genomes to be reported [14],[32]. Comparative [33] and functional genomic analyses [5],[32] revealed that these 2 ICEs share a set of conserved core genes that mediate their integration/excision (int and xis), conjugative transfer (various tra genes), and regulation (setR, setCD). In addition to the conserved genes, these 2 ICEs contain element specific genes that confer element specific properties such as resistance to antibiotics or heavy metals. Interestingly, many of these genes were found in identical locations in SXT and R391, leading Beaber et al [33] to propose that there are ‘hotspots’ where SXT/R391 ICEs can acquire new DNA. The genomes of two additional SXT/R391 ICEs, ICEPdaSpa1, isolated from Photobacterium damselae [23], and ICESpuPO1, derived from an environmental isolate of Shewanella putrefaciens [24] are now also known. These two genomes also share most of the conserved set of core genes present in SXT and R391 and contain element specific DNA.

Determination of the sequences of SXT/R391 family ICE genomes was a fairly arduous task due to their size and predominantly chromosomal localization. Here, we developed a method to capture and then sequence complete SXT/R391 ICE genomes. In addition, we identified 3 as yet unannotated SXT/R391 ICE genomes in the database of completed bacterial genomes. Comparative analyses of the 13 SXT/R391 genomes now available allowed us to greatly refine our understanding of the organization and conservation of the core genes that are present in all members of this ICE family. Comparative and functional analyses also facilitated our proposal of the minimal functional SXT/R391 ICE genome. Furthermore, this work provides new knowledge of the considerable diversity of genes and potential accessory functions encoded by the variable DNA found in these mobile elements. Finally, this comparative genomics approach has allowed us to garner clues regarding the evolution of this class of mobile elements.

Results/Discussion

An ICE capture system

To date, ICE sequencing has been cumbersome because it has typically required construction of chromosome-derived cosmid libraries and screening for sequences that hybridize to ICE probes [23],[32]. We constructed a vector (pIceCap) that enables capture of complete SXT/R391 ICE genomes on a low-copy plasmid to simplify the protocol for ICE sequencing. This plasmid is a derivative of the single-copy modified F plasmid pXX704 [34],[35], which contains a minimal set of genes for F replication and segregation but lacks genes enabling conjugation. We modified pXX704 to include an ~400bp fragment that encompasses the SXT/R391 attachment site (attB) and thereby enabled Int-catalyzed site-specific recombination between attB on pIceCap and attP on an excised and transferred ICE to drive ICE capture (Figure 1). Conjugations between an SXT/R391 ICE-bearing donor strain and an E. coli recipient deleted for prfC (and thus chromosomal attB) and harboring pIceCap yielded exconjugants containing the transferred ICE integrated into pIceCap (Figure 1). We used the ΔprfC recipient to bias integration of the transferred ICE into pIceCap rather than the chromosome. In these experiments, we selected for exconjugants containing the transferred ICE integrated into pIceCap, using an antibiotic marker present on the ICE as well as a marker present in pIceCap. The low copy IceCap::ICE plasmid was then isolated and used as a substrate for shotgun sequencing. We also found that the IceCap::ICE plasmids were transmissible. Thus, in principle this technique should facilitate capture of ICEs that do not harbor genes conferring resistance to antibiotics, by mating out the IceCap::ICE plasmid into a new recipient and selecting for the marker on pIceCap.

Figure 1
Schematic of the ICE capture system.

SXT/R391 ICEs included in this analysis

A list of the 13 SXT/R391 ICEs whose genomes were analyzed and compared in this study is shown in Table 1. All of the ICEs included in our analyses contain an int gene that was amplifiable using PCR primers for intsxt [29]. They were isolated on 4 continents and from the Pacific Ocean during a span of more than 4 decades. They are derived from 7 different genera of γ-proteobacteria and the ICEs derived from V. cholerae strains are from both clinical and environmental isolates of 3 different V. cholerae serogroups.

Table 1
SXT/R391 ICE family members analyzed in this study.

Five of these ICE genome sequences were determined at the J. Craig Venter Institute (JCVI) using the ICE capture system described above (Table 1, rows 1–5). In addition, we sequenced ICEVflInd1, also at the JCVI, by isolating cosmids that encompassed this V. fluvialis derived ICE prior to developing the ICE capture technique (Table 1, row 6). Table 1 (rows 7–10) also includes 4 previously unannotated ICE genomes that we found in BLAST searches of the NCBI database of completed but as yet unannotated genomes; 3 of these ICEs are clearly members of SXT/R391 ICE family since they are integrated into their respective host's prfC locus and contain int genes that are predicted to encode Int proteins that are 99% identical to Intsxt. The fourth element, ICEVchBan8 does not encode an Intsxt orthologue; however, this element contains nearly identical homologues of most of the known conserved core SXT/R391 ICE family genes. ICEVchBan8 will be discussed in more detail below but since it does not contain an Intsxt orthologue it is not considered a member of the SXT/R391 family of ICEs and thus not included in our comparative study. Finally, Table 1 also includes the 4 SXT/R391 ICEs that were previously sequenced (Table 1, rows 11–14).

Despite the diversity of our sources for SXT/R391 ICEs, the genomes of two pairs of ICEs that we analyzed proved to be very similar. SXTMO10 and ICEVchInd4 only differed by 13 SNPs in 7 genes and by the absence from ICEVchInd4 of dfr18, a gene conferring trimethoprim resistance. These ICEs were derived from V. cholerae O139 strains isolated in India from different cities at different times: SXTMO10 from Chennai in 1992 and ICEVchInd4 from Kolkata in 1997. The high degree of similarity of these two ICE genomes suggests that ICEs can be fairly stable over time. ICEVchBan9 and ICEVchMoz10 were also extremely similar although ICEVchMoz10 lacks dfrA1, another allele for trimethoprim resistance. These two ICEs were derived from V. cholerae O1 strains from Bangladesh (1994) and Mozambique (2004) respectively. The great similarity of these ICEs suggests that there has been spread of SXT-related ICEs between Asia and Africa in recent times. Studies of CTX prophage genomes have also suggested the spread of V. cholerae strains between these continents [36].

General structure and sizes of SXT/R391 genomes

The ICEs listed in Table 1 were initially compared using MAUVE [37] and LAGAN [38], programs that enable visualization of conserved and variable regions on a global scale. All of the SXT/R391 ICEs we analyzed share a common structure and have sizes ranging from 79,733 bp to 108,623 bp (Table 1 and Figure 2). They contain syntenous sets of 52 conserved core genes (Figure 2A) that total approximately 47kb and encode proteins with an average of 97% identity to those encoded by SXT. All of the individual ICEs also contain DNA that is relatively specific for individual elements (Figure 2B); the differences in the sizes of the variable regions accounts for the range in ICE sizes.

Figure 2
Structure of the genomes of 13 SXT/R391 ICEs.

Five sites within the conserved SXT/R391 ICE structure have variable DNA present in all of the ICEs in Figure 2. Four of these sites were previously termed ‘hotspots’ for ICE acquisition of new DNA [33]. Due to similarities between SXT and R391, the fifth hotspot only became apparent through our comparison of the 13 ICEs examined here. Each of these hotspots (HS1 to HS5 in Figure 2B) is found in an intergenic region (see below), suggesting that the acquisition of these variable DNA regions has not interrupted core ICE gene functions. In addition, some of the ICEs have variable DNA inserted in additional intergenic locations or in rumB (labeled I–IV in Figure 2B). Previous analyses [32] indicated that the insertion in rumB, did not impair SXT transmissibility. Overall, comparison of these 13 SXT/R391 ICE genomes suggests that: 1) these elements consist of the same perfectly syntenous and nearly identical 52 core genes that serve as a scaffold (see below) capable of mobilizing a large range of variable DNA; and 2) selection pressure to maintain ICE mobility has restricted insertions of variable DNA into sites that do not interrupt core functions.

The SXT/R391 ICE core genes

The 52 core genes present in all the SXT/R391 ICEs analyzed include sets of genes that are known to be required for the key ICE functions of integration/excision, conjugative transfer and regulation [32] as well as many genes of unknown function. Most genes of known or putative (based on homology) function (coded by gray shading or hatch marks in Figure 2A) are clustered with genes that have related functions. For example, int and xis, genes required for integration and excision, are adjacent and setR, and setC/D, the key SXT regulators are near each other at the extreme 3′ end of the elements, although separated by 4 conserved genes of unknown function. Each ICE also has four gene clusters implicated in conjugative DNA processing and transfer (shown in light gray in Figure 2A). Finally, each of the ICEs has a nearly identical origin of transfer (oriT), a cis-acting DNA site that is thought to be nicked to initiate DNA processing events during conjugative transfer [39], in the same relative location.

The conserved core genes include approximately as many genes of unknown function as genes of known function. Some of the genes of unknown function are found either interspersed amongst gene clusters that likely comprise functional modules (e.g s091 between traD and s043) while others are grouped together (e.g. most genes between traN and traF). In several cases, the interspersed genes appear to be part of operons with genes of known function (e.g. s086-s082 maybe in an operon with setDC).

Variable ICE DNA

In addition to sharing 52 core genes, all of the ICE genomes analyzed contain variable DNA regions, ranging in size from 676 to 29,210 bp. Most of the variable DNA sequences are found in 5 intergenic hotspots (Figure 2B). However, some ICEs contain additional variable DNA inserts outside the 5 hotspots. For example, SXT and five other ICEs in Figure 2 have variable DNA segments, corresponding to related ISCR2 elements, disrupting rumB (Figure 2B, site III). ISCR2 elements are IS91-like transposable elements that tend to accumulate antibiotic resistance genes [40]. Interestingly, it is unusual for the contents of the hotspots and other variable regions to be found in only one ICE. Instead, the variable gene content of most of the ICEs shown in Figure 2B is found in more than one ICE. For example, ICESpuPO1, ICEPalBan1, and ICEVflInd1, all have identical contents in hotspot 5 (lavender genes in hotspot 5 in Figure 2B); however, the contents of the other hotspots in these 3 elements are almost entirely different. Thus, the variable gene content of the SXT/R391 ICEs reveals that these elements are mosaics. The overlapping distribution of variable DNA segments seen in the ICEs in Figure 2B suggests that recombination among this family of mobile elements may be extensive. In addition, in some instances, the variable regions appear subject to additional genetic modifications. For example, ICEPdaSpa1 and ICEVchBan9 contain ICE-specific DNA nested within the shared sequences inserted at hotspot 5 DNA (the green and pink genes in hotspot 5 in these elements, Figure 2B).

The variable genes encode a large array of functions and only a few will be discussed here. A complete list of the diverse genes found in the hotspots is found in Table S1. Although we cannot predict functions for many genes found in the hotspots, since they lack homology to genes of known function, at least a subset of the known genes seem likely to confer an adaptive advantage upon their hosts. Most of the ICE antibiotic resistance genes are found within transposon-like structures (e.g., the ISCR2 elements noted above) but four ICEs contain a dfrA1 cassette, which confers resistance to trimethoprim [25], in a class IV integron located in hotspot 3. A disproportionate number of variable genes are likely involved in DNA modification, recombination or repair, as they are predicted to encode diverse putative restriction-modification systems, helicases and endonucleases. Such genes may provide the host with barriers to invasion by foreign DNA including phage infection and/or promote the integrity of the ICE genome during its transfer between hosts. Three ICEs contain genes that encode diguanylate cyclases [41] in hotspot 3. These enzymes catalyze the formation of cyclic-diguanosine monophosphate (c-di-GMP), a second messenger molecule that regulates biofilm formation, motility and virulence in several organisms including V. cholerae [42],[43]. Most SXT/R391 ICEs contain mosA and mosT in hotspot 2. These two genes encode a novel toxin-antitoxin pair that promotes SXT maintenance by killing or severely inhibiting the growth of cells that have lost this element [44]. Not all ICEs in the SXT/R391 family contain mosAT; however, those lacking these genes may encode similar systems to prevent ICE loss. For instance, R391 and ICEVchMex1 contain two genes (orf2 and orf3) encoding a predicted HipA-like toxin and a predicted transcriptional repressor distantly related to the antitoxin HipB.

Locations of the ICE variable genes

The variable regions found in the 5 hotspots are found exclusively in intergenic regions, punctuating the conserved ICE backbone (Figure 2). The boundaries between the conserved and variable sequences were mapped on the nucleotide level and compared (Figure 3A–3E). Each hotspot had a distinct boundary. Remarkably, even though the contents of the variable regions markedly differ, with few exceptions the left and right boundaries between conserved and variable DNA for each hotspot was identical among all the ICEs (Figure 3). For example, the left junctions of the inserts in hotspot 2 immediately follow the stop codon of traA and the right junctions are exactly 79 bp upstream of the start of s054 (Figure 3B), despite the fact that the DNA contents within these borders greatly differ. In hotspot 2, the right junction appears to begin with a 15 bp sequence that has two variants (Figure 3B, brown & light brown sequence). These sequences may reflect the presence of earlier insertions that have since been partially replaced. A similar pattern was found adjacent to the left boundary of hotspot 4 in several ICEs (Figure 3D, lines 3–6). Once an insertion is acquired, the number of permissive sites for the addition of new variable DNA likely increases.

Figure 3
The boundaries of the 5 hotspots.

There are two exceptions to the precise boundaries between variable and conserved DNA. Hotspot 1 and hotspot 3 in ICEVchMex1 and ICEPdaSpa1, respectively, contain variable DNA that extends beyond the boundary exhibited by all the other ICEs in these locations (Figure 3A, line 3, and Figure 3C, line 7). The only boundary that could not be identified was the left border of hotspot 5, the region containing genes between s026 and traI. As discussed below, s026 is the least conserved core gene and its variability obscured any consensus sequence abutting the variable DNA. Perhaps this border has eroded because s026 is not required for ICE mobility [32].

The relative precision of most boundaries between conserved and variable DNA sequences in all the ICEs analyzed suggests that a particular recombination mechanism, such as bet/exo-mediated recombination, may explain the acquisition of the variable regions. However, at this point, we cannot exclude the possibility that the precise locations for variable DNA insertions simply reflects selection for optimal ICE fitness; i.e., ICEs can optimally accommodate variable DNA in these locations while preserving their essential functions.

Similarity of SXT/R391 ICE and IncA/C plasmid core genes

Unexpectedly, BLAST analyses revealed that most of the conserved core SXT/R391 genes are also present in IncA/C conjugative plasmids. These multidrug resistance plasmids are widely distributed among Salmonella and other enterobacterial isolates from agricultural sources [45],[46]. Recently, members of this family of plasmids have also been identified in Yersinia pestis, including from a patient with bubonic plague [47], and in aquatic γ-proteobacteria [48], including Vibrio cholerae [49],[50]. To date, the closest known relatives of the SXT/R391 transfer proteins are found in the IncA/C plasmids. Every predicted SXT transfer protein is encoded by the IncA/C plasmid pIP1202 isolated from Y. pestis [50] and the identities of these predicted protein sequences vary from 34 to 78% (Figure 4A). Furthermore, there is perfect synteny between the four gene clusters encoding the respective conjugative machineries of these two mobile elements (yellow and orange genes in Figure 4A). Despite the extensive similarity of the SXT and IncA/C conjugative transfer systems, these plasmids lack homologues of setR and setD/C as well as int/xis, suggesting that regulation of conjugative transfer differs between these elements.

Figure 4
Comparison of the SXT/R391 core genome with the genome of pIP1202 and defining the minimal functional SXT/R391 gene set.

The similarity of IncA/C plasmids and SXT/R391 ICEs is not limited to genes important for conjugal DNA transfer. Ten genes of unknown function (shown in black in Figure 4A), some of which are interspersed within likely tra gene operons and some of which are clustered together between traN and traF, are similar in the two elements. Furthermore, most of these ten genes are in identical locations in the two elements. Both elements also contain homologs of bet and exo (shown in green in Figure 4A); these are the only known homologs of the λ Red recombination genes found outside of bacteriophages. Together, the similarity of DNA sequences and organization of SXT/R391 ICEs and IncA/C plasmids suggests that these elements have a common ancestor. The fact that the contents of the hotspots in the two classes of elements are entirely distinct suggests that their evolutionary paths diverged prior to acquisition of these variable DNA segments.

The minimal functional SXT/R391 ICE gene set

The conservation of the 52 core genes in all 13 SXT/R391 ICEs analyzed suggested that many or even all of these genes would be required for key ICE functions of excision/integration, conjugative transfer and regulation. The presence of ten ICE core genes of unknown function in IncA/C plasmids (black genes in Figure 4A) is also consistent with the hypothesis that these genes might be required for ICE transfer. However, our previous work demonstrated that not all genes recognized here as part of the conserved core gene set are required for SXT transfer. Beaber et al showed that deletion of rumB – s026 (which includes 5 cores genes) from SXT had no detectable influence on SXT excision or transfer [32]. Therefore, we systematically deleted all of the core ICE genes whose contributions had not previously been assessed, in order to explore the hypothesis that these genes (especially those also present in IncA/C plasmids) would be essential for ICE transfer and to define the minimum functional SXT/R391 gene set.

Surprisingly, deletion of most of the ICE core genes of unknown function, including genes with homologues in IncA/C plasmids, did not alter SXT transfer efficiency. Deletion of s002 or s003, which are located downstream of int in all SXT/R391 ICEs, did not alter the frequency of SXT transfer; similarly, deletion of s082, s083, and s084, core genes of unknown function that are found near the opposite end of SXT/R391 ICEs but not in IncA/C plasmids, also did not influence SXT transfer frequency (Figure 4B). Furthermore, deletion of s091, which is found between traD and s043 in ICEs and IncA/C plasmids, did not reduce SXT transfer (Figure 4B). In contrast, deletion of s043, which has weak homology to traJ in the F plasmid (a gene important in DNA processing) and is located in a transfer cluster containing traI and traD, abolished transfer (Figure 4B, Δd), suggesting that s043, here re-named traJ is required for SXT transfer. It is unlikely that the transfer defect of SXTΔtraJ can be explained by polar effects of the deletion on downstream genes, since traJ appears to be the last gene of an operon found immediately upstream of hotspot 1. Similarly, deletion of s054, which is found immediately 5′ of traC and is homologous to a disulfide-bond isomerase dsbC, also abolished transfer (Figure 4B, Δe). Interestingly, disulfide bond-isomerases are present in several other conjugative systems [51]. However, it is not clear at this point if the deletion of s054 from SXT accounts for the transfer defect of SXTΔs054, since we could not restore transfer by complementation.

Additionally, Beaber et al found that deletion of s060 through s073 in SXT, which includes 7 genes that are also found in IncA/C plasmids reduced SXT transfer more than 100-fold [32]. We constructed several smaller deletions in this region and found that deletion of s063, which is also found in pIP1202, reduced the transfer frequency of SXT by ~100-fold, nearly the same amount as deleting the entire region (Figure 4B). Complementation analyses revealed that the absence of s063 accounted for the transfer defect of SXTΔs063 (data not shown). Even though SXTΔs063 was still capable of transfer, in our view, the drastic reduction in the transfer frequency of this mutant warrants inclusion of s063 into the minimum functional SXT ICE genome (shown in Figure 4C). Other deletions in this region, including deletions of bet, exo, s067, s068 and s070, which have orthologues in IncA/C plasmids, resulted in ≤10-fold reductions in transfer frequency. We therefore did not include these genes in the minimal functional core SXT/R391 genome (Figure 4C).

The findings from our experiments testing the transfer frequencies of SXT derivatives harboring core gene deletions (shown in Figure 4B), coupled with our previous work demonstrating the requirements for the predicted SXT tra genes in the element's transfer [32], suggest a minimal functional SXT/R391 ICE structure as shown in Figure 4C. This minimum element is ~29.7 kb and consists of 25 genes. Genes with related functions, which in some cases encode proteins that likely form large functional complexes (such as the conjugation apparatus), are grouped together in the minimal genome. At the left end of the minimum ICE genomes are xis and int, the integration/excision module of SXT/R391 ICEs. In the minimal ICE genome, the ICE oriT and mobI, which encodes a protein required for SXT transfer [39], are no longer separated from the other genes (traIDJ) that are also thought to play roles in the DNA processing events required for conjugative DNA transfer. The genes required for formation of the conjugation machinery, including the pilus, and mating pair formation and stabilization [32],[39] are divided between three clusters (denoted mpf1-3 in Figure 4C). Finally, at the right end of the minimal functional genome are the genes that regulate ICE transfer (setC/D and setR). Thus, the minimal functional SXT/R391 ICE is relatively small and organized into 3 discrete functional modules that mediate excision/integration, conjugation, and regulation.

Even though deletion of 27 out of 52 SXT/R391 ICE core genes proved to have little or no effect on SXT transfer frequency, and hence these genes were not included in Figure 4C, it is reasonable to presume that these genes encode functions that enhance ICE fitness given their conservation. For example, the presence of highly conserved bet and exo genes in all SXT/R391 ICEs suggests that there has been selection pressure to maintain this ICE-encoded recombination system that promotes ICE diversity by facilitating inter ICE recombination (G Garriss, MK Waldor, V Burrus, in press). A key challenge for future studies will be to determine how core genes of unknown function promote ICE fitness.

Variations in the similarity of core genes

To identify genes in the SXT/R391 core genome that may be subject to different selection pressures, we compared the percent identity of each ICE's core genes to the corresponding SXT gene (Figure 5). Most of the ICEs' core genes exhibited 94% to 98% identity on the nucleotide level to SXT's core genes. There was no discernable difference in the degree of conservation of most core genes that were or were not part of the minimal ICE, suggesting that there are equal selective pressures on essential and non-essential genes. However, we identified 8 genes (s026, traI, orfZ, s073, traF, eex, s086, and setR) that exhibit significantly different degrees of conservation (Figure 5 and Figure S1). Three of these showed unusually high conservation, while the other 5 had below average conservation. Two of the highly conserved genes, setR and s086, are found at the extreme 3′ end of the elements. The conservation of setR may reflect the key role of this gene in controlling SXT gene expression. S086 may also play a role in regulating SXT transfer [52]. The other highly conserved gene, orfZ, is found between bet and exo and has no known function.

Figure 5
Variations in the nucleotide conservation of core ICE genes.

s026 and s073 are the most divergent of all the genes in the backbone. s026 encodes a hypothetical protein with homologues in many gram negative organisms. Although S026 is predicted to contain a conserved domain, COG2378, which has a putative role in transcription regulation, this protein is not required for SXT transfer [32]. The significant divergence of s026 along with its lack of essentiality suggests that this gene could become a pseudogene. A similar argument could be made for s073, which encodes a hypothetical protein that is also not required for ICE transfer. However, this argument does not hold for traI or traF, two genes which are essential for ICE transfer. Although the reasons which account for the different degrees of conservation of these 8 core genes are hard to ascertain at this point, the data in Figure 5 suggests that individual core genes are subject to different evolutionary pressures.

Comparisons of core gene phylogenies

We created phylogenetic trees for each core gene based on their respective nucleotide sequences to further explore the evolution of the conserved backbone of SXT/R391 ICEs. Since we found such a high degree of conservation for most of the core genes, the bootstrap values for most of these trees were relatively low. Thus, we concentrated on the most polymorphic genes found in Figure 5, s026, s073, traI, and eex, for phylogenetic analyses. As shown in Figure 6A, the trees for s026, traI and s073 exhibit 3 distinct branching patterns. The lack of similarity in these phylogenetic trees suggests that either individual core genes have evolved independently or that high degrees of recombination mask their common evolutionary history. The latter hypothesis seems more likely since experimental findings have revealed that SXT/R391 ICEs can co-exist in a host chromosome in tandem [26] and recombination between tandem elements can yield novel hybrid ICEs with considerable frequency [53] (G Garriss, MK Waldor, V Burrus, in press). Also, as noted above, the distributions of variable genes among the ICEs shown in Figure 2 also supports the idea that inter-ICE recombination is commonplace.

Figure 6
Phylogenetic analysis of several core ICE genes.

Unlike most core genes, the trees for traG and eex were similar. In these two trees, the ICEs segregate into two evolutionarily distinct groups (Figure 6B), confirming and extending previous observations that revealed that there are two groups of eex and traG sequences in SXT/R391 ICEs [54]. These two groups correspond to the two functional SXT/R391 ICE exclusion groups. Interactions between traG and eex of the same group mediate ICE exclusion [55]. Thus, the identical 2 clusters of traG and eex sequences observed in their respective trees reveals the co-evolution of the traG/eex functional unit. The two groups of eex sequences can also be observed in Figure 5 where the bifurcating pattern reveals the 2 exclusion groups. This pattern is difficult to discern for traG, perhaps because of the large size of this multi-functional gene.

ICEVchBan8, an SXT-like ICE that lacks Intsxt

The sequence of ICEVchBan8, which was derived from a non-O1, non-O139 V. cholerae strain, is incomplete but it appears to contain 49 out of 52 SXT/R391 core genes. However, since this strain lacks Intsxt it was not included in our comparative analyses above. It is not known if ICEVchBan8 is capable of excision or transmission; however, it contains a P4-like integrase and a putative xis. It is tempting to speculate that the genome of ICEVchBan8 provides an illustration of how acquisition (presumably via recombination) of a new integration/excision module may generate a novel ICE family.

Perspectives

Comparative analysis of the genomes of the 13 SXT/R391 ICEs studied here has greatly refined our understanding of this group of mobile genetic elements. These elements, which have been isolated from 4 continents and the depths of the Pacific Ocean, all have an identical genetic structure, consisting of the same syntenous set of 52 conserved core genes that are interrupted by clusters of diverse variable genes. All the elements have insertions of variable DNA segments in the same five intergenic hotspots that interrupt the conserved backbone. Furthermore, some of the elements have additional insertions outside the hotspots; however, in all cases the acquisition of variable DNA has not compromised the integrity of the core genes required for ICE mobility. Functional analyses revealed that less than half of the conserved genes are necessary for ICE transmissibility and the contributions of the 27 core genes of unknown function to ICE fitness remains an open question. Finally, several observations presented here suggest that recombination between SXT/R391 ICEs has been a major force in shaping the genomes of this widespread family of mobile elements.

Although comparisons of the 13 ICE genomes analyzed here strongly suggest that these mobile elements have undergone extensive recombination during their evolutionary histories, there is a remarkable degree of similarity among the SXT/R391 ICEs. All of these ICEs consist of the same syntenous and nearly identical 52 genes. In contrast, other families of closely related mobile elements, such as lambdoid or T4-like phages for example, exhibit greater diversity [56],[57]. Since the elements that we sequenced were isolated from several different host species and from diverse locations, the great degree of similarity of the SXT/R391 ICE family does not likely reflect bias in the elements that we sequenced. It is possible that this family of mobile elements is a relatively recent creation of evolution and has yet to undergo significant diversification.

To date, relatively few formal comparative genomic analyses of other ICE families have been reported. Mohd-Zain et al [11] identified several diverse ICEs and genomic islands that shared a largely syntenous set of core genes with ICEHin1056, an ICE originally identified in Haemophilus influenzae. However, even though these elements share a similar genomic organization, they exhibit far greater variability in the sites of insertion of variable DNA and in the degree of conservation in their core genes compared to SXT/R391 ICEs. Thus, although this group of elements appears to share a common ancestor, they seem to have diverged earlier in evolutionary history than the SXT/R391 ICEs. However, when comparative genomic analyses were restricted to ICEHin1056-related ICEs found in only two Haemophilus sp., Juhas et al found that, like the SXT/R391 family of ICEs, these 7 ICEHin1056-related ICEs share greater than 90% similarity at the DNA level in their nearly syntenous set of core genes [12]. It will be interesting to learn the extent of conservation of genetic structure and DNA sequence in additional ICE families to obtain a wider perspective on ICE evolution.

Comparative genomic studies of bacteriophages have led to the idea that the full range of phage sequences are part of common but extremely diverse gene pool [58],[59]. The SXT/R391 ICE genomes suggest that there may be an even larger network of phylogenetic relationships linking sequences found in all types of mobile genetic elements including phages, plasmids, ICEs and transposons. The genomes of SXT/R391 ICEs appear to be amalgams of genes commonly associated with other types of mobile elements. Many of the ICE core genes are usually associated with phages, such as int, bet, exo and setR, or with plasmids, such as the tra genes. Additionally, the SXT/R391 ICEs and IncA/C plasmids clearly have a common ancestor, as we found that the entire set of SXT/R391 tra genes are also present in IncA/C plasmids. Thus, the genes present in all types of mobile genetic elements appear to contribute to a common gene pool from which novel variants of particular elements (such as ICEVchBan8) or perhaps even novel types of mobile genetic elements can arise.

Materials and Methods

ICE Sequencing

ICEPalBan1, ICEVchMex1, ICEVchInd4, ICEVchInd5 and ICEVchBan5 were isolated using the plasmid capture system described in Figure 1. The SXT chromosomal attachment sequence, attB, was introduced into the modified F plasmid pXX704 [34] to create pIceCap. This plasmid was then introduced into a ΔprfC derivative of the TcR E. coli strain CAG18439. Exconjugants derived from matings between this strain and those harboring the 5 ICEs listed above resulted in strains carrying a pIceCap::ICE plasmid. Once captured, the plasmids were isolated using the Qiagen plasmid midi kit for low-copy plasmids (Qiagen). Isolated pIceCap::ICE plasmids were then sequenced.

ICEVflInd genome was determined by sequencing several overlapping cosmids that encompassed this ICE's genome. Briefly, genomic DNA from a Vibrio fluvialis strain carrying ICEVflInd was prepared using the GNome DNA kit (QBIOgene). Sau3A1 restricted genomic DNA was used to create a SuperCos1 (Stratagene)-based cosmid library according the manufacture's instructions. The library was subsequently screened for cosmids containing ICE-specific sequences using PCR with primers to conserved core ICE sequences. Four cosmids containing overlapping ICEVflInd sequences were identified and sequenced.

The genomes of 6 ICEs were sequenced by the Sanger random shotgun method [60]. Briefly, small insert plasmid libraries (2–3 kb) were constructed by random nebulization and cloning of pIceCap::ICE DNA or of cosmid DNA for ICEVflInd. In the initial random sequencing phase, 8–12 fold sequence coverage was achieved. The sequences of either pIceCap or pSuperCos were subtracted and the remaining sequences were assembled using the Celera Assembler [61]. An initial set of open reading frames (ORFs) that likely encode proteins was identified using GLIMMER [62], and those shorter than 90 base pairs (bp) as well as some of those with overlaps eliminated.

Bioinformatics

Nucleotide and amino acid conservation were assessed with the appropriate BLAST algorithms. ICEs were aligned using clustalW with default settings [63]. MAUVE [37] and LAGAN [38] were used to identify core genes in Figure 2. To map the boundaries of the hotspots, sequence comparisons were made using MAUVE and then manually compared to find boundaries between conserved and variable DNA as shown in Figure 3.

Phylogenetic trees were generated from alignments of nucleotide sequences using the neighbor-joining method as implemented by ClustalX software, version 2.011 [64]. The reliability of each tree was subjected to a bootstrap test with 1000 replications. Trees were edited using FigTree 1.22 (http://tree.bio.ed.ac.uk/software/figtree/).

Generation and testing of SXT deletion mutants

CAG81439 harboring SXT was used as the host strain to create the SXT deletion mutants shown in Figure 3; the deletions were constructed using one-step gene inactivation as previously described [44],[65]. The primers used to create the deletion mutants are available upon request. Matings were conducted as previously described [16],[44] using deletion mutants and a KnR E. coli recipient, CAG18420. Exconjugants were selected on LB agar plates containing chloramphenicol, 20µg/ml (for SXT selection) and kanamycin, 50 µg/ml. The frequency of exconjugant formation was calculated by dividing the number of exconjugants by the number of donors.

Supporting Information

Figure S1

Variations in the conservation of individual core ICE genes. The percent identity of the nucleotide sequence of each core gene and oriT versus the corresponding sequence in SXT was calculated for all ICEs studied. The average values for each gene (as shown in the inset of Figure 5) were then used in one-way ANOVA comparisons to determine genes that exhibit significantly more or less conservation compared to other core genes. p-values of one-way ANOVA comparisons of each core ICE gene are shown. The grid represents all pair-wise comparisons, and the color indicates the level of significance as follows: red: p<.001, orange: p<.01, and yellow: p<.05. Genes that exhibited a p-value<.05 when compared with at least 50% of all other core genes are discussed in the text.

(0.56 MB TIF)

Table S1

Contents of the hotspots.

(0.14 MB DOC)

Acknowledgments

We thank Brigid Davis and Frédérique Le Roux for helpful comments on the manuscript. We thank Robert Hall for facilitating the work described here and Yoshiharu Yamaichi for suggesting the design of pIceCap.

Footnotes

The authors have declared that no competing interests exist.

MKW acknowledges support from NIH (AI R37-42347) and HHMI. GG is the recipient of a PhD research scholarship from FQRNT. VB holds a Canada Research Chair in molecular biology, impact and evolution of bacterial mobile elements and is grateful for support from the Natural Sciences and Engineering Research Council of Canada (Discovery Grant Program). MS was supported by a grant from PRIN 2007 - Italy, and DC was supported by a fellowship from Cenci Bolognetti-Institut Pasteur Foundation, Italy. Sequencing at the JCVI was funded by the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Department of Health and Human Services under contract number N01-AI-30071. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1. Frost LS, Leplae R, Summers AO, Toussaint A. Mobile genetic elements: the agents of open source evolution. Nat Rev Microbiol. 2005;3:722–732. [PubMed]
2. Gogarten JP, Townsend JP. Horizontal gene transfer, genome innovation and evolution. Nat Rev Microbiol. 2005;3:679–687. [PubMed]
3. Lawrence JG, Hendrickson H. Genome evolution in bacteria: order beneath chaos. Curr Opin Microbiol. 2005;8:572–578. [PubMed]
4. Ochman H, Lerat E, Daubin V. Examining bacterial species under the specter of gene transfer and exchange. Proc Natl Acad Sci U S A. 2005;102(Suppl 1):6595–6599. [PMC free article] [PubMed]
5. Beaber JW, Hochhut B, Waldor MK. SOS response promotes horizontal dissemination of antibiotic resistance genes. Nature. 2004;427:72–74. [PubMed]
6. Burrus V, Waldor MK. Control of SXT integration and excision. J Bacteriol. 2003;185:5045–5054. [PMC free article] [PubMed]
7. Auchtung JM, Lee CA, Monson RE, Lehman AP, Grossman AD. Regulation of a Bacillus subtilis mobile genetic element by intercellular signaling and the global DNA damage response. Proc Natl Acad Sci U S A. 2005;102:12554–12559. [PMC free article] [PubMed]
8. Caparon MG, Scott JR. Excision and insertion of the conjugative transposon Tn916 involves a novel recombination mechanism. Cell. 1989;59:1027–1034. [PubMed]
9. Cheng Q, Paszkiet BJ, Shoemaker NB, Gardner JF, Salyers AA. Integration and excision of a Bacteroides conjugative transposon, CTnDOT. J Bacteriol. 2000;182:4035–4043. [PMC free article] [PubMed]
10. Hagege J, Pernodet JL, Friedmann A, Guerineau M. Mode and origin of replication of pSAM2, a conjugative integrating element of Streptomyces ambofaciens. Mol Microbiol. 1993;10:799–812. [PubMed]
11. Mohd-Zain Z, Turner SL, Cerdeno-Tarraga AM, Lilley AK, Inzana TJ, et al. Transferable antibiotic resistance elements in Haemophilus influenzae share a common evolutionary origin with a diverse family of syntenic genomic islands. J Bacteriol. 2004;186:8114–8122. [PMC free article] [PubMed]
12. Juhas M, Power PM, Harding RM, Ferguson DJ, Dimopoulou ID, et al. Sequence and functional analyses of Haemophilus spp. genomic islands. Genome Biol. 2007;8:R237. [PMC free article] [PubMed]
13. Burrus V, Waldor MK. Shaping bacterial genomes with integrative and conjugative elements. Res Microbiol. 2004;155:376–386. [PubMed]
14. Boltner D, MacMahon C, Pembroke JT, Strike P, Osborn AM. R391: a conjugative integrating mosaic comprised of phage, plasmid, and transposon elements. J Bacteriol. 2002;184:5158–5169. [PMC free article] [PubMed]
15. Whittle G, Shoemaker NB, Salyers AA. The role of Bacteroides conjugative transposons in the dissemination of antibiotic resistance genes. Cell Mol Life Sci. 2002;59:2044–2054. [PubMed]
16. Waldor MK, Tschape H, Mekalanos JJ. A new type of conjugative transposon encodes resistance to sulfamethoxazole, trimethoprim, and streptomycin in Vibrio cholerae O139. J Bacteriol. 1996;178:4157–4165. [PMC free article] [PubMed]
17. Shoemaker NB, Barber RD, Salyers AA. Cloning and characterization of a Bacteroides conjugal tetracycline-erythromycin resistance element by using a shuttle cosmid vector. J Bacteriol. 1989;171:1294–1302. [PMC free article] [PubMed]
18. Franke AE, Clewell DB. Evidence for a chromosome-borne resistance transposon (Tn916) in Streptococcus faecalis that is capable of “conjugal” transfer in the absence of a conjugative plasmid. J Bacteriol. 1981;145:494–502. [PMC free article] [PubMed]
19. Ravatn R, Studer S, Springael D, Zehnder AJ, van der Meer JR. Chromosomal integration, tandem amplification, and deamplification in Pseudomonas putida F1 of a 105-kilobase genetic element containing the chlorocatechol degradative genes from Pseudomonas sp. Strain B13. J Bacteriol. 1998;180:4360–4369. [PMC free article] [PubMed]
20. Sullivan JT, Ronson CW. Evolution of rhizobia by acquisition of a 500-kb symbiosis island that integrates into a phe-tRNA gene. Proc Natl Acad Sci U S A. 1998;95:5145–5149. [PMC free article] [PubMed]
21. Cholera Working Group. Large epidemic of cholera-like disease in Bangladesh caused by Vibrio cholerae O139 synonym Bengal. Lancet. 1993;342:387–390. [PubMed]
22. Ahmed AM, Shinoda S, Shimamoto T. A variant type of Vibrio cholerae SXT element in a multidrug-resistant strain of Vibrio fluvialis. FEMS Microbiol Lett. 2005;242:241–247. [PubMed]
23. Osorio CR, Marrero J, Wozniak RA, Lemos ML, Burrus V, et al. Genomic and functional analysis of ICEPdaSpa1, a fish-pathogen-derived SXT-related integrating conjugative element that can mobilize a virulence plasmid. J Bacteriol. 2008;190:3353–3361. [PMC free article] [PubMed]
24. Pembroke JT, Piterina AV. A novel ICE in the genome of Shewanella putrefaciens W3-18-1: comparison with the SXT/R391 ICE-like elements. FEMS Microbiol Lett. 2006;264:80–88. [PubMed]
25. Hochhut B, Lotfi Y, Mazel D, Faruque SM, Woodgate R, et al. Molecular analysis of antibiotic resistance gene clusters in vibrio cholerae O139 and O1 SXT constins. Antimicrob Agents Chemother. 2001;45:2991–3000. [PMC free article] [PubMed]
26. Hochhut B, Beaber JW, Woodgate R, Waldor MK. Formation of chromosomal tandem arrays of the SXT element and R391, two conjugative chromosomally integrating elements that share an attachment site. J Bacteriol. 2001;183:1124–1132. [PMC free article] [PubMed]
27. Coetzee JN, Datta N, Hedges RW. R factors from Proteus rettgeri. J Gen Microbiol. 1972;72:543–552. [PubMed]
28. Burrus V, Marrero J, Waldor MK. The current ICE age: biology and evolution of SXT-related integrating conjugative elements. Plasmid. 2006;55:173–183. [PubMed]
29. Hochhut B, Waldor MK. Site-specific integration of the conjugal Vibrio cholerae SXT element into prfC. Mol Microbiol. 1999;32:99–110. [PubMed]
30. Maeda K, Nojiri H, Shintani M, Yoshida T, Habe H, et al. Complete nucleotide sequence of carbazole/dioxin-degrading plasmid pCAR1 in Pseudomonas resinovorans strain CA10 indicates its mosaicity and the presence of large catabolic transposon Tn4676. J Mol Biol. 2003;326:21–33. [PubMed]
31. Murata T, Ohnishi M, Ara T, Kaneko J, Han CG, et al. Complete nucleotide sequence of plasmid Rts1: implications for evolution of large plasmid genomes. J Bacteriol. 2002;184:3194–3202. [PMC free article] [PubMed]
32. Beaber JW, Hochhut B, Waldor MK. Genomic and functional analyses of SXT, an integrating antibiotic resistance gene transfer element derived from Vibrio cholerae. J Bacteriol. 2002;184:4259–4269. [PMC free article] [PubMed]
33. Beaber JW, Burrus V, Hochhut B, Waldor MK. Comparison of SXT and R391, two conjugative integrating elements: definition of a genetic backbone for the mobilization of resistance determinants. Cell Mol Life Sci. 2002;59:2065–2070. [PubMed]
34. Niki H, Hiraga S. Subcellular distribution of actively partitioning F plasmid during the cell division cycle in E. coli. Cell. 1997;90:951–957. [PubMed]
35. Yamaichi Y, Niki H. migS, a cis-acting site that affects bipolar positioning of oriC on the Escherichia coli chromosome. Embo J. 2004;23:221–233. [PMC free article] [PubMed]
36. Faruque SM, Tam VC, Chowdhury N, Diraphat P, Dziejman M, et al. Genomic analysis of the Mozambique strain of Vibrio cholerae O1 reveals the origin of El Tor strains carrying classical CTX prophage. Proc Natl Acad Sci U S A. 2007;104:5151–5156. [PMC free article] [PubMed]
37. Darling AC, Mau B, Blattner FR, Perna NT. Mauve: multiple alignment of conserved genomic sequence with rearrangements. Genome Res. 2004;14:1394–1403. [PMC free article] [PubMed]
38. Brudno M, Do CB, Cooper GM, Kim MF, Davydov E, et al. LAGAN and Multi-LAGAN: efficient tools for large-scale multiple alignment of genomic DNA. Genome Res. 2003;13:721–731. [PMC free article] [PubMed]
39. Ceccarelli D, Daccord A, Rene M, Burrus V. Identification of the origin of transfer (oriT) and a new gene required for mobilization of the SXT/R391 family of integrating conjugative elements. J Bacteriol. 2008;190:5328–5338. [PMC free article] [PubMed]
40. Toleman MA, Bennett PM, Walsh TR. ISCR elements: novel gene-capturing systems of the 21st century? Microbiol Mol Biol Rev. 2006;70:296–316. [PMC free article] [PubMed]
41. Bordeleau E, Brouillette E, Robichaud N, Burrus V. Beyond antibiotic resistance: integrating conjugative elements of the SXT/R391 family that encode novel diguanylate cyclases participate to c-di-GMP signalling in Vibrio cholerae. Environ Microbiol 2009 [PubMed]
42. Jenal U, Malone J. Mechanisms of cyclic-di-GMP signaling in bacteria. Annu Rev Genet. 2006;40:385–407. [PubMed]
43. Romling U, Gomelsky M, Galperin MY. C-di-GMP: the dawning of a novel bacterial signalling system. Mol Microbiol. 2005;57:629–639. [PubMed]
44. Wozniak RA, Waldor MK. A toxin-antitoxin system promotes the maintenance of an integrative conjugative element. PLoS Genet. 2009;5:e1000439. doi: 10.1371/journal.pgen.1000439. [PMC free article] [PubMed]
45. Lindsey RL, Fedorka-Cray PJ, Frye JG, Meinersmann RJ. Inc A/C plasmids are prevalent in multidrug-resistant Salmonella enterica isolates. Appl Environ Microbiol. 2009;75:1908–1915. [PMC free article] [PubMed]
46. Llanes C, Gabant P, Couturier M, Bayer L, Plesiat P. Molecular analysis of the replication elements of the broad-host-range RepA/C replicon. Plasmid. 1996;36:26–35. [PubMed]
47. Galimand M, Guiyoule A, Gerbaud G, Rasoamanana B, Chanteau S, et al. Multidrug resistance in Yersinia pestis mediated by a transferable plasmid. N Engl J Med. 1997;337:677–680. [PubMed]
48. Welch TJ, Fricke WF, McDermott PF, White DG, Rosso ML, et al. Multiple antimicrobial resistance in plague: an emerging public health risk. PLoS One. 2007;2:e309. doi: 10.1371/journal.pone.0000309. [PMC free article] [PubMed]
49. Pan JC, Ye R, Wang HQ, Xiang HQ, Zhang W, et al. Vibrio cholerae O139 multiple-drug resistance mediated by Yersinia pestis pIP1202-like conjugative plasmids. Antimicrob Agents Chemother. 2008;52:3829–3836. [PMC free article] [PubMed]
50. Fricke WF, Welch TJ, McDermott PF, Mammel MK, LeClerc JE, et al. Comparative genomics of the IncA/C multidrug resistance plasmid family. J Bacteriol. 2009;191:4750–4757. [PMC free article] [PubMed]
51. Elton TC, Holland SJ, Frost LS, Hazes B. F-like type IV secretion systems encode proteins with thioredoxin folds that are putative DsbC homologues. J Bacteriol. 2005;187:8267–8277. [PMC free article] [PubMed]
52. Beaber JW, Waldor MK. Identification of operators and promoters that control SXT conjugative transfer. J Bacteriol. 2004;186:5945–5949. [PMC free article] [PubMed]
53. Burrus V, Waldor MK. Formation of SXT tandem arrays and SXT-R391 hybrids. J Bacteriol. 2004;186:2636–2645. [PMC free article] [PubMed]
54. Marrero J, Waldor MK. The SXT/R391 family of integrative conjugative elements is composed of two exclusion groups. J Bacteriol. 2007;189:3302–3305. [PMC free article] [PubMed]
55. Marrero J, Waldor MK. Interactions between inner membrane proteins in donor and recipient cells limit conjugal DNA transfer. Dev Cell. 2005;8:963–970. [PubMed]
56. Juhala RJ, Ford ME, Duda RL, Youlton A, Hatfull GF, et al. Genomic sequences of bacteriophages HK97 and HK022: pervasive genetic mosaicism in the lambdoid bacteriophages. J Mol Biol. 2000;299:27–51. [PubMed]
57. Filee J, Bapteste E, Susko E, Krisch HM. A selective barrier to horizontal gene transfer in the T4-type bacteriophages that has preserved a core genome with the viral replication and structural genes. Mol Biol Evol. 2006;23:1688–1696. [PubMed]
58. Hendrix RW, Hatfull GF, Smith MC. Bacteriophages with tails: chasing their origins and evolution. Res Microbiol. 2003;154:253–257. [PubMed]
59. Hendrix RW, Smith MC, Burns RN, Ford ME, Hatfull GF. Evolutionary relationships among diverse bacteriophages and prophages: all the world's a phage. Proc Natl Acad Sci U S A. 1999;96:2192–2197. [PMC free article] [PubMed]
60. Fouts DE, Mongodin EF, Mandrell RE, Miller WG, Rasko DA, et al. Major structural differences and novel potential virulence mechanisms from the genomes of multiple campylobacter species. PLoS Biol. 2005;3:e15. doi: 10.1371/journal.pbio.0030015. [PMC free article] [PubMed]
61. Myers EW, Sutton GG, Delcher AL, Dew IM, Fasulo DP, et al. A whole-genome assembly of Drosophila. Science. 2000;287:2196–2204. [PubMed]
62. Delcher AL, Harmon D, Kasif S, White O, Salzberg SL. Improved microbial gene identification with GLIMMER. Nucleic Acids Res. 1999;27:4636–4641. [PMC free article] [PubMed]
63. Thompson JD, Higgins DG, Gibson TJ. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 1994;22:4673–4680. [PMC free article] [PubMed]
64. Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, et al. Clustal W and Clustal X version 2.0. Bioinformatics. 2007;23:2947–2948. [PubMed]
65. Datsenko KA, Wanner BL. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A. 2000;97:6640–6645. [PMC free article] [PubMed]
66. Burrus V, Quezada-Calvillo R, Marrero J, Waldor MK. SXT-related integrating conjugative element in New World Vibrio cholerae. Appl Environ Microbiol. 2006;72:3054–3057. [PMC free article] [PubMed]
67. Das B, Halder K, Pal P, Bhadra RK. Small chromosomal integration site of classical CTX prophage in Mozambique Vibrio cholerae O1 biotype El Tor strain. Arch Microbiol. 2007;188:677–683. [PubMed]
68. Pearson MM, Sebaihia M, Churcher C, Quail MA, Seshasayee AS, et al. Complete genome sequence of uropathogenic Proteus mirabilis, a master of both adherence and motility. J Bacteriol. 2008;190:4027–4037. [PMC free article] [PubMed]
69. Nair GB, Faruque SM, Bhuiyan NA, Kamruzzaman M, Siddique AK, et al. New variants of Vibrio cholerae O1 biotype El Tor with attributes of the classical biotype from hospitalized patients with acute diarrhea in Bangladesh. J Clin Microbiol. 2002;40:3296–3299. [PMC free article] [PubMed]

Articles from PLoS Genetics are provided here courtesy of Public Library of Science
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

    Your browsing activity is empty.

    Activity recording is turned off.

    Turn recording back on

    See more...