Vous êtes sur la page 1sur 11

Plasmid 48 (2002) 202–212 www.academicpress.


When phage, plasmids, and transposons collide: genomic islands, and conjugative- and mobilizable-transposons as a mosaic continuum
€ltner A. Mark Osborn* and Dietmar Bo
Department of Biological Sciences, University of Essex, Colchester, CO4 3SQ, UK Received 29 August 2002

Abstract Plasmids and bacteriophage represent the classical vectors for gene transfer within the horizontal gene pool. However, the more recent discovery of an increasing array of other mobile genetic elements (MGE) including genomic islands (GIs), conjugative transposons (CTns), and mobilizable transposons (MTns) which each integrate within the chromosome, offer an increasingly diverse assemblage contributing to bacterial adaptation and evolution. Molecular characterisation of these elements has revealed that they are comprised of functional modules derived from phage, plasmids, and transposons, and further that these modules are combined to generate a continuum of mosaic MGE. In particular, they are comprised of any one of three distinct types of recombinase, together with plasmid-derived transfer and mobilisation gene functions. This review highlights both the similarities and distinctions between these integrating transferable elements resulting from combination of the MGE toolbox. Ó 2002 Elsevier Science (USA). All rights reserved.
Keywords: Gene transfer; Genomic islands; Conjugative transposons; Conjugation; Tyrosine recombinase; Resolvase; Transposase

1. Mobile genetic elements and the horizontal gene pool The prokaryotic horizontal gene pool (HGP) represents a rich tapestry of adaptive phenotypes conveyed within and between bacterial (and archaeal) cells, by an increasingly diverse assemblage of mobile genetic elements (MGE). Fifty years on from the discovery of transduction by

Corresponding author. Fax: +44-1206-872592. E-mail address: osborn@essex.ac.uk (A. Mark Osborn).


Salmonella bacteriophage (Zinder and Lederberg, 1952) and the introduction of the term plasmid (Lederberg, 1952), following the earlier identification of bacterial conjugation, research has revealed numerous combinations of genetic modules derived from phage, plasmid (and transposons) to generate an array of MGE that include conjugative and mobilizable transposons, genomic islands, and integrons (Toussaint and Merlin, 2002). Whilst plasmids and bacteriophage have wellestablished genetic and phenotypic identities, the emerging families of mosaic MGE pose a considerable challenge for the systematicist as to how

0147-619X/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved. PII: S 0 1 4 7 - 6 1 9 X ( 0 2 ) 0 0 1 1 7 - 8

D. which are found in some bacterial strains but are absent from otherwise very closely related strains. and transposons) that contribute key functional components are circled (dotted lines). which vary in size between 10 and 500 kb. 2002). 1. (ii) serine Scocca. MGE (plasmids.A. Thus this review will focus also on the emerging groups of mobilizable transposons comprised of different combinations of recombinases and plasmid-related mobilisation functions. the situation is complicated further by the additional inclusion on MGE of genes encoding plasmid-related mobilisation and transfer functions. were first identified as chromosomally . and second. GIs. B€ oltner / Plasmid 48 (2002) 202–212 203 Fig. are now recognised as important contributors to bacterial adaptation and evolution. phage. the genomic islands (GIs) and conjugative transposons (CTns). Genomic islands Genomic islands.. MGE functional modules are as given in the key. named following the conserved aspartate and glutamate residues (Polard and Chandler. Such complications arise not least from the utilisation of three distinct enzyme families catalysing the vital process of recombination between DNA molecules (Toussaint and Merlin. and which are being discovered increasingly on GIs. 1997. The mosaic continuum of mobilizable and conjugative genetic elements. 2. which from their names at least suggest vastly differing entities. 1995). 1) that are either self-transmissible or mobilizable between bacterial cells. Black triangles inserted within the functional modules of the Anchored Genomic Island indicate that molecular evidence exists for remnants or mutant forms of such modules (see text for further details). These families are named after key conserved residues in their active site: (i) tyrosine recombinases (reviewed in Esposito and €by et al. However. best to classify such elements. Mark Osborn. Yet molecular comparison of their backbone modules identifies key similarities. 1998). it is apparent on further inspection that they represent just part of a continuum of mosaic MGE (Fig. namely the presence of phage-related integrases belonging to the tyrosine recombinase family. Moreover. 2002). Nunes-Du recombinases (resolvase/invertase) (reviewed in Smith and Thorpe. and (iii) DD-E recombinases (transposases). the occurrence of plasmid-related transfer genes on CTns. whilst comparative analysis of these elements reveals fascinating evolutionary insights. In this review we focus in particular on two groups of MGE.

the clc element encoding chlorocatechol degradation (Ravatn et al... 2001). 2000) has shown this to be a mosaic containing a tyrosine recombinase related to that of the CTns Tn5276 and Tn5252. Significantly. that includes an integrase most closely related to that from Mesorhizobium loti GIs. Xanthomonas (No€ el et al. a 153 kb pathogenicity island has been identified in Enterococcus faecalis MMH594 PAI (Shankar et al. and subsequently found to integrate using a tyrosine recombinase into either of two tRNA gene attachment sites. and Neisseria) and Gram-positive genera (Staphylococcus. 2002). Other integrases. 2000). Recent comparison of tRNA attachment sites for the broader group of elements utilising tyrosine recombinases. resulting in permanently anchored islands (see below). 2002). B€ oltner / Plasmid 48 (2002) 202–212 located virulence genes in uropathogenic Escherichia coli. In addition. Salmonella. 1998) was initially identified following conjugative transfer from Pseudomonas strain B13 to P. In some GIs the integrase gene may be deleted or non-functional. given that the phenotype of bacteria receiving such islands is often altered dramatically (e. the SRL PAI from Shigella flexneri (Turner et al. PAIs carry genes encoding a variety of phenotypes including adhesins. 1983). D... especially if the island also carries conjugal transfer genes.. 1998). 1996).g. and also carries transfer genes related to the Enterococcus plasmids pAD1 and pAM373. or near to.and halo-aromatic compounds is also believed to integrate within the chromosome (Hickey et al.. and thus. 1999). together with plasmid pSK41. Vibrio. 2002). For example the clc element (Ravatn et al. and Pseudomonas syringae (Jackson et al. Whilst the presence of plasmid-related transfer genes on GIs offers conjugation as a dissemination . this phenomenon has been described as Ôevolution in quantum leapsÕ (Groisman and Ochman. Integration of this element has been shown to be site-specific within its host. 2000). Other likely members of this group of Ôconjugative genomic islandsÕ include the related conjugative integrating elements R391 and SXT that carry metal and/or antibiotic resistance genes (see below). and Clostridium) (Hacker and Kaper. 2001). Helicobacter. has revealed three sub-locations for integration within these genes.. 2000).g.. 2002).204 A. Although mobilisation of the MMH594 PAI has been demonstrated. and subsequently termed pathogenicity islands (PAIs) by Hacker et al. putida Fl. Similarly. (1990). it is not currently known whether the element is also self-transmissible. 1998). In addition to the virulence genes they typically carry an integrase gene related to that of phage lambda and belonging to the larger family of tyrosine recombinases. whilst at present it is not known whether this element is capable of independent transfer. Accordingly. tRNA genes and are flanked by short direct repeats resembling phage attachment sites (Hacker and Kaper. loti R7A and MAFF303099 (Sullivan et al.and Tn916-related transfer genes. and subsequently in the related islands from M. More recently. partial sequence analysis of the integrative element ICEStl (Burrus et al. have been demonstrated to be functional. from Pseudomonas putida KF715 also transfers by conjugation and is believed to integrate within a specific insertion hot-spot (Nishi et al. Conjugative transfer has also been demonstrated for the 500 kb Sym island from Mesorhizobium loti (Sullivan and Ronson. whilst a third conjugative element from Pseudomonas aeruginosa JB2 encoding degradation of hydroxy. the bph-sal element. however. As a consequence such islands have been termed ÔecologicalÕ or ÔfitnessÕ islands (Hacker and Carniel. 1998) and symbiosis (Sym islands) (Sullivan and Ronson. potentially capable of horizontal gene transfer. Acquisition of such traits as functional units allows bacteria to respond rapidly to environmental challenges and explore new ecological niches.g... degradation of xenobiotic compounds e. from non-pathogenic to pathogenic. or from non-symbiotic to symbiotic). which is increasingly observed (see below). Phenotypes carried by genomic islands are not limited to those encoding virulence but also include antibiotic resistance e. a phylogeny of integrase sequences has also identified three groups that are consistent with the tRNA sub-location classification (Williams. In Gram-positive bacteria an increasing number of potentially transmissible GIs are identified.. In Streptococcus thermophilus. Listeria.. 2001).. 2000). A number of the ecological islands have been demonstrated to be conjugative. encoding biphenyl and salicylate metabolism functions.. More recently PAIs have been identified in a diverse range of animal pathogens from both Gram-negative (Yersinia. Mark Osborn. secretion systems and iron uptake systems critical to pathogenicity (reviewed in Hacker and Kaper. PAIs have also been identified in a number of plant pathogens including Erwinia. differing in G + C content and codon usage from the surrounding DNA (Hacker et al. Typically PAIs are integrated into.

with at least 95% identify to each other over 65 kb. The recent deter€ltner mination of the DNA sequences of R391 (Bo et al. Following encapsulation of SaPI1 and the subsequent transfer to another cell.e. 2000) and the inability to demonstrate transfer of the multi-drug resistances (Threlfall et al. B€ oltner / Plasmid 48 (2002) 202–212 205 mechanism. 1991). GIs can also utilise other MGE and mechanisms to facilitate their dispersion. and in the case of SaPI1 this is a direct consequence of SaPI1 replication interfering with that of the phage. albeit a self-transmissable one. 1995). being variously described as a transmissible resistance factor. 1).. the nature of R391. had represented something of an enigma.. coli (Hochhut and Waldor. UV-inducible excision has not been demonstrated. in a semi-parasitic manner. i. However. Whilst these may facilitate integration into the chromosome. 2002) and SXT (Beaber et al. both R391 and SXT are capable of excision and subsequent conjugative transfer to recipient cells. 1999). the DNA sequence of SGI1 suggests a possible previous existence as an integrative conjugative plasmid. referred to as IncJ elements (following plasmid incompatibility nomenclature). the PAIs can be excised and circularised by certain Staphylococcal phage (80a and /13) to form circular forms of the PAIs that can then re-integrate in a site-specific manner. The absence of any identifiable plasmid €ltner et al. as indicated by both the presence of a number of ORFs related to genes from the IncHI1 plasmid R27 encoding mating pair stabilisation and pilus ... enterica DT104—see below). classically. 1998) both contain a putative integrase (tyrosine recombinase) and are flanked by direct repeats. are readily transferred intact via phage transduction.. 1). this proposed classification of R391 and SXT as conjugative transposons might at first seem appropriate. offering another example of the numerous mechanistic variations found in the HGP. an integrating plasmid. Anchored genomic islands and integrative plasmids Analysis of the 43 kb genomic island (SGI1) carrying multi-drug resistances. in Salmonella enterica serovar Typhimurium DT104 (Boyd et al.. 1979) and pMERPH (Peters et al. R391. and the IncJ elements Two notable recent examples of the combination of conjugative plasmid transfer genes with phage-related integration systems are offered by two related elements. where these elements differ from archetypal conjugative transposons. Mark Osborn. R391. on the basis of sequence data and phenotypic and molecular characterisation. may in fact have been the first genomic island to be isolated. recA-independent chromosomal integration combined with conjugative transfer.. at least at the level of phenotype. SaPI1 uses its own integrase to integrate into the recipient hostÕs chromosome (Ruzin et al. such as Tn916. together with a conjugative transfer system related to that from the plasmid R27 (seen also in SGI1 from S. The 15. SXT. However. though now inappropriately. isolated 30 years ago. 2002) is consistent with replicon (Bo repeated failures to isolate ccc DNA and now confirms that these elements are not plasmids. They form part of a larger series of elements. 4.2 kb PAI SaPI1 and the related SaPI2 from Staphylococcus aureus which both carry the tst gene for toxic shock syndrome toxin-1 (Lindsay et al. with R391 and SXT only integrating into a single site on the E. R391 from a South African isolate of Providencia rettgeri (Coetzee et al. this island would now appear to be permanently anchored within the chromosome (Fig. Alternatively.. 1994). including R997 (Matthew et al. 2001). 2002) shows them to share conserved backbone regions. 3. whilst lacking their own transfer system. In particular. is in the absence of random chromosomal integration (see below). in contrast to SGI1.. they share greater similarities with the GIs that similarly integrate into just one or two sites (typically tRNA genes) on the chromosome. However. 1972) and the SXT element from Vibrio cholerae. In retrospect. With a failure to detect excision of the SGI1 island (Boyd et al. and subsequently as a conjugative transposon (Murphy and Pembroke... However. coli chromosome (see Fig.A. replication of the circular form can lead to transduction of the PAIs at high frequencies. Prior to the determination of the sequences of these elements and the earlier demonstration of the site-specific integration of SXT and R391 into the prfC gene of E.. suggesting an absence of an excisionase on these elements. In this respect. isolated in India (Waldor et al. Thus. 1996). 2001) indicates a complex evolutionary history. these backbone regions include a phage-related integrase and associated regulatory genes. Thus these elements. D..

206 A. we speculate that these plasmids may represent the first examples of integrative plasmids in the archaea. 1981). in particular the PAI on the Shigella flexneri virulence plasmid pWR201 (Venkatesan et al. In contrast. Chromosomal integration of plasmids has long been recognised. ORF439 in Bo the 41. 1994) more akin to the relatively random integration of CTns.. We have also identified ORFs encoding potential tyrosine recombinases using the PFAM database within archaeal plasmids (Osborn and €ltner. D. Similarly. the similarities between plasmid conjugative transfer systems and the type IV protein secretion systems including those carried by the cag PAI of Helicobacter pylori or. Whilst the function of their gene products is unknown. 1). 5.2 kb conjugative circular plasmid pSAM2 (Boccard et al.. Past traces of a conjugative lifestyle may also be indicated in the Neisseria gonorrhoeae Gonococcal Genetic Island (GGI) by the presence of traG and traH homologues (Dillard and Seifert. which encode a DD-E type recombinase (transposase) at the flanking ends of the 44. 2001) but also notably in the discovery of a PAI carrying the three toxin genes on the plasmid pXO1 from Bacillus anthracis (Okinaka et al. 1999). it is likely that its insertion into pXO1 is mediated by insertion sequences.. Plasmid-located genomic islands Localisation of GIs is not limited to the chromosome. such similarities between DNA and protein transporters may represent divergent evolution of ancestral macromolecule secretion systems. B€ oltner / Plasmid 48 (2002) 202–212 assembly proteins. and the related ORF457 in the pING plasmid from Sulfolobus islandicus (Stedman et al. a lambda-like integrase mediates integration of the 11.3 kb plasmid pSE101 into a threonine tRNA gene in its host Saccharoployspora erythraea. 1998). 1980). Two such elements were originally identified: Tn916 from Enterococcus faecalis (Franke and Clewell. the Streptomyces ambofaciens 11...2 kb conjugative plasmid pNOB8 from Sulfolobus sp. Whilst integration by homologous recombination enables plasmids to form stable cointegrates. Whereas this PAI includes an integrase gene. it is tempting to speculate on the possibility for divergence and subsequent evolution over time of these genes to fulfil new functional roles. without requiring autolysis (Hamilton et al.. 2001). with reversible integration of the plasmid pKLK106 into two lysine tRNA genes in Pseudomonas aeruginosa (Kiewitz et al. Mark Osborn.. though subsequent insertion mutagenesis suggests the fascinating possibility that these genes function instead as a Type IV secretion system acting as a potential DNA donation system to facilitate DNA transformation. Integration of F typically occurs by homologous recombination between IS elements (IS2 or IS3a and 3b) present on the F plasmid and chromosome. introduced into Streptomyces lividans the element demonstrates integration into multiple sites (Brown et al. 2000) These examples of PAIs that utilise transposons for insertion represent yet another intriguing illustration of the versatility of mosaic MGE. most notably of the F plasmid in the formation of E. with a number of elements found located on plasmids. responsible for extracellular transport of the pertussis toxin in Bordetella pertussis.. 2001). sequence analysis of a 27. For example. Given that the incorporation of plasmid-derived conjugative transfer genes into the chromosome is readily demonstrated.. NOB8H2 (She et al. 1).. coli Hfr strains. Conjugative transposons Conjugative transposons were first identified as chromosomally-borne transposons that were capable of conjugative transfer. Alternatively. as opposed to a tyrosine recombinase (Fig. which is now the accepted archetype of the CTns.5 kb PAI found on the 80 kb plasmid p33071 in Rhodococcus equi again suggest involvement of transposons in PAI insertion. 2001). Similarly. 2000). 2000). together with an ORF related to the replication gene (repA) from the Rhodopseudomonas plasmid pMG101. suggest common evolutionary origins for these systems (Christie.3 kb element. 1989) carries a tyrosine recombinase in addition to a rolling circle type replicon and conjugative transfer genes (Fig. when pSE101 is. Whilst numerous CTns have . unpublished data) namely. 6. as evidenced by the presence of two transposon-related resolvases at either end of the PAI (Takai et al. and Tn5253 from Streptococcus pneumoniae (Shoemaker et al. In Gram-negative bacteria tyrosine recombinase-mediated integration of plasmids has also been reported. as evidenced by the presence of copies of IS1627. an increasing number of plasmids are reported to carry a lambda-related integrase that facilitates recA-independent chromosomal integration and excision. Indeed.

A functional origin of transfer has been located between ORFs 20 and 21. Moreover. 2000) and the rep gene from the Bacillus plasmid pGI3 (Hoflack et al. 2000) and the increasing number of SpoIIIE homologues suggest that this may be a widespread mechanism for DNA transport. which is involved in pilus assembly. whilst the TraG protein is homologous to the BctA protein involved in conjugative transfer of the Bacteroides pBF4 plasmid (Morgan and Macrina. 2001) and pSAM2 (Hag ege et al. although no individual base within this generic sequence is conserved amongst the numerous integration sites identified.. 1997) and the M. D.. Thus it is probable that the CTnDOT transfer operon represents a distant relative of Gram-negative conjugative plasmid transfer systems. the putative protein encoded by ORF1 is homologous to relaxase proteins from both conjugative plasmids and mobilizable transposons. we refer readers to the excellent reviews by Salyers et al. In contrast. 2001) and pAM373 (De Boever et al. (1995) and Scott and Churchward (1995). yet many of these elements would appear to have more in common with genomic islands. on the basis of their preference for specific integration sites... SpoIIIE genes are believed to drive DNA transport from the mother cell into the prespore during sporulation in Bacillus (Bath et al. whilst Tn4371 carries RP4/Ti-related transfer genes. notwithstanding the more detailed characterisation of the Enterococcus pheromone-sensing plasmids pAD1 (Francia et al.. B€ oltner / Plasmid 48 (2002) 202–212 207 subsequently been identified. conjugative . CTnscr94 from Salmonella senftenberg (Hochhut et al. as opposed to the more random integration exhibited by classical CTns. The transfer system of CTnDOT is unrelated to that of Tn916-like elements. the putative TraP protein shows homology to DNA primases. 1995). in particular tRNA genes. Classical CTns such as Tn916 demonstrate semi-random integration in a manner more redolent of classical transposons. and also to a number of proteins related to the TraC protein from the F plasmid. offer a better candidate for a conjugative transposon within the proteobacteriaceae. we again found that many of the putative ORFs described show no relationship to known transfer genes. 1993). Tn916 does demonstrate distinct preferences for AT-rich regions with a consensus target of 50 -TT/ ATTTT(N6 )AAAAAA/TA-30 (Lu and Churchward. 1995). and is thought to represent a new class of conjugal transfer system (Bonheyo et al. Similarly the TraJ protein from CTnDOT is related to the TrwI protein from the transfer operon of R388. For both Tn916 and CTnDOT.. A full review of the CTns is beyond the scope of this current paper. much of our molecular understanding of CTns stems from analyses of Tn916 and the Bacteroides element CTnDOT. 1997).. This element that utilises a tyrosine recombinase demonstrates semirandom integration into a number of sites present on both the CH34 chromosome and the co-resident pMOL plasmid. Over recent years a number of elements isolated from members of the proteobacteriaceae have been proposed as CTns. Mark Osborn. as is similarly the case for most GIs. However. including R391. 2000). The 55 kb catabolic biphenyl degradation transposon Tn4371 from Ralstonia eutropha CH34 does. and in particular from Bacteroides. however. CTnDOT shows greater specificity and is found to integrate into only about seven sites in the Bacteroides chromosome (Bedzyk et al. SpoIIIE homologues are also responsible for conjugative transfer of the Streptomyces plasmids pIJ101 (Pettis and Cohen.. suggesting a possible role for this ORF for DNA nicking prior to transfer. The CTnDOT TraA protein is related to a number of ParA proteins involved in plasmid partitioning. Of two additional ORFS upstream of the putative tra genes on the CTnDOT sequence (accession number: AF289050). However. 1992). 2002). albeit predominantly from Gram-positive bacteria.A. loti strain R7A symbiosis island (Sullivan et al. unpublished data) of ORF21 of Tn916 Bo shows it to be a member of the FtsK/SpoIIIE family. this may identify a possible closer linkage between the transfer systems of classical CTns and those from conjugative plasmids. although FASTA analysis (Osborn and €ltner. 2001).. and includes sequences related to both IncP and F plasmid nic sites (Jaworski and Clewell. Thus. 1997). it is possible that the difficulties in identifying relationships to well-known plasmid conjugation systems may be a consequence that these CTns have been isolated from organisms for which our knowledge of plasmid biology is relatively sparse.. Integration of the majority of CTns utilises a tyrosine recombinase. FASTA analysis of ORF20 €ltner. however. as more plasmids are sequenced from these organisms. Less is known about the transfer system of Tn916. unpublished data) identifies (Osborn and Bo relationships to certain Vibrio phage replication proteins (Nasu et al. Following our own FASTA analysis. albeit conjugative ones. However. and SXT (see above).

have been reported.g. whether the element integrates into multiple sites (as is the case for Tn916).g. 2000) to mediate excision and insertion. 1997)]. the M. 2000. when transferred into E.. historical precedence in the naming of individual elements will lead to future researchers trying to classify newly discovered elements as part of existing groupings. Firstly transposons such as Tn3 and cd utilise a serine recombinase (resolvase/invertase). These transposon-encoded mob determinants permit mobilisation of Tn4551—carrying plasmids lacking their own mob system by a helper plasmid.208 A. yet an increasing number of mobilizable transposons (MTns). 1996a. Tn4451 and Tn4453 from Clostridium spp. demonstrates semi-random integration (Shoemaker et al. and R391). can be delineated. they clearly share a tyrosine recombination-type integration mechanism. and the IS3 family utilise the so-called DD-E recombinase (transposase). Mark Osborn. NBU1. loti R7A symbiosis island. sites. Mobilisation of these elements utilises oriT sequences and mob genes resembling those from Gram-positive plasmids (Shoemaker et al. namely. carry a putative oriT site related to that of the conjugative transposon Tn916 (Fig. although the transfer genes have been demonstrated to be functional via mobilisation of a second transposable element Tn-bph (Merlin et al. to those from conjugative plasmids. Tn5398 from C.. however. enabling mobilisation by a co-resident conjugative plasmid. this is most apparent with respect to the conjugative transposons (e. albeit to differing degrees. Upon comparison of such classical transposons with their namesakes. D. Perhaps on this basis alone CTns and a second group.. 2000). Wang et al. On molecular comparison of these elements. being mobilizable and capable of semi-random integration upon transfer to Bacillus subtilis. The majority of these have been identified in Bacteroides isolates. the clc element. 8. 1996b). (Tribble et al. as well as potential mobilisation genes. required for integration. Arguably the clearest distinction between CTns and GIs is with respect to integration site preference. However. consisting of a transposon that includes an oriT site. conjugative MGE. 2001). Tn10. 1) to integrate into either tRNA genes (NBU1 and NBU2) (Shoemaker et al.. Smith and Parker.. or semi-randomly (Tn4555) [as is the case for Tn916. represent a second type of MTn that instead utilise a serine recombinase (the TnpX resolvase) (Lyras and Rood. suggesting that site-specificity of tyrosine recombinase-based systems is a host-specific phenomenon. In contrast. and additionally carry a plasmid-related mob gene (Fig. the conjugative transposons (that utilise a tyrosine recombinase). with indistinct boundaries between a number of groups of elements. 1). It does. with the Non-replicating Bacteroides Units (NBUs). as is typical for the GIs. and both utilise conjugative transfer systems that are related. or occasionally two.. Obviously. Arguably. We would suggest that CTn is retained for those classical CTns exhibiting random or . 2000). 1998.. Tn5398 was found to lack both a gene encoding any recognisable recombinase. coli. it is attractive to conjecture upon the existence of truly randomly integrating. Intriguingly. At present such systems await discovery. perhaps combining a DD-E or serine recombinase and a complete plasmid-related conjugative transfer system. B€ oltner / Plasmid 48 (2002) 202–212 transfer of the element has not been demonstrated.. and the mobilizable transposon Tn4555 all identified as utilising a tyrosine recombinase-based integration system (Fig. 1). a second group including Tn7. 1999). Conjugative transposons or conjugative genomic islands—that is the question The state of MGE nomenclature is a direct consequence of their highly mosaic composition. such as RP4 (Crellin and Rood. as described above for pSE 101.. Alternatively. Tn916 and CTn-DOT) and the increasing number of GIs that encode their own conjugation systems (e. 1992). for which we propose the name Conjugative Genomic Islands. difficile 630 represents yet another fascinating variation on this theme. 7. yet following analysis of the DNA sequence of this element.. whilst others have led to highly similar MGE being classified with very different names. as new DNA sequences are generated and molecular mechanisms unravelled. NBU1 and NBU2. Wang et al. A third type of mobilizable element. Mobilizable transposons Classical transposons utilise one of two types of recombinase. It is postulated that Tn5398 utilises both resolvases and mob gene functions carried on the conjugative transposon Tn5397 which is co-resident in the same host (Farrow et al. or into one. it is clear that some earlier classifications are inappropriate.

the desire to also assign generic names to new and existing elements remains compelling. an integrating antibiotic resistance gene transfer element derived from Vibrio cholerae. Friedmann. Rood. 1327– 1341. Young. G. Crellin.. As a lead. Salyers.. P. Microbiol. Boccard. 1994.M. Christie. 543–552. Peters. A. J. or the integrative plasmid pSE101 (see above) that show single-site specificity in one host. Graham.A. G. Genomic and functional analyses of SXT...A..Õ classification remains an important discipline within biology. would clearly fall within this group. L.... 183. K. Ng.T. 184..A. However...B. Bedzyk. L. Mark Osborn. Gu edon. 1749– 1753. R factors from Proteus rettgeri. C.. € ltner. D. 1992. Tn4451 from Clostridium perfringens is a mobilizable transposon that encodes the functional Mob protein. and will increasingly further benefit from ongoing developments of a number of MGE-module specific websites.. L.. M. 2000. Bacteriol. Microbiol. Imberechts.M. Mulvey. Brown.. Backer. 294–305. K. are termed conjugative genomic islands (a sub-family of the GIs). Katz. 1972. Idler. Bacteriol. Enterococcus faecalis conjugative plasmid pAM373: complete nucleotide sequence and genetic analyses of sex pheromone response. Datta. This term has been adopted already for a number of MGE (see above). as needed. F. Waldor. N. 631–642. D. N.. 2001.. 37.. R391: a conjugative integrating mosaic comprised of phage. Characterization of the genes and attachment sites for site-specific integration of plasmid pSE101 in Saccharopolyspora erythraea and Streptomyces lividans.. CTnDOT. M. Beaber. C. and can be mobilised via conjugation to other cells. to cover MGE that integrate (irrespective of their type of recombinase) into the chromosome. J. Bo Osborn.. Gen. Whilst some may ask ÔWhat’s in a name? That which we call a rose by any other name would smell as sweet. whilst those conjugative integrating elements which have distinct sitepreference.J. EMBO J. 41–51. D. Y. Hedges. Complete nucleotide sequence of a 43kilobase genomic island associated with the multidrug resistance region of Salmonella enterica serovar Typhimurium DT104 and its identification in phage type DT120 and serovar Agona. L. D. A. Bonheyo. T. S.. E.. 2000. Type IV secretion: intracellular transfer of macromolecules by systems ancestrally related to conjugation machines... A.A. Microbiol. Role of Bacillus subtilis SpoIIIE in DNA transport across the mother cell-prespore division septum. G.. one approach is to accept their mosaic nature.. Chaslus-Dancla. Strike.. J. but lack a conjugation system of their own.. but integration into multiple sites in other recipients.. Roussel.C. even here the waters are muddied by elements such as the mobilizable transposons NBU1 and NBU2. Clewell.H.. 40..M.-K. 174. Salyers. Shoemaker. and classify all MGE on the presence of key functional modules. Burrus. D..... B. A. which in many cases will be for a tRNA gene. .A. Mol.S.K. Pernodet. Bacteriol. ICEStl..J. 2002. K. 5725–5732. FEMS Microbiol. Mulvey. 1992. Errington.. Gen. Partial characterization of a genomic island associated with the multidrug resistance region of Salmonella enterica Typhimurium DT104. Bacteriol. H... J. Acknowledgments The authors wish to thank Professor Julian Rood for an introduction to the serine-recombinase-based mobilizable transposons. Science 290. 4259–4269. D. Boumedine. J. Microbiol. References Bath. 2001.R. J. Appl. E. 2000. 72. The integrated conjugative plasmid pSAM2 of Streptomyces ambofaciens is related to temperate bacteriophages. Decaris.E.R.. N. 166–172.. J. Whilst such an approach is extremely valuable (Toussaint and Merlin. MacMahon.. We hope that the MGE research community will face these challenges and be willing to embrace appropriate changes to nomenclature. R.. A. G. V.A. Environ. 66. B€ oltner / Plasmid 48 (2002) 202–212 209 semi-random integration. 184. 189. J. Mol. in the lactic acid bacterium Streptococcus thermophilus. Wang. Mol.B. Genet. 5158–5169. Characterization of a novel integrative element. Fraser. Micro. plasmid and transposon elements. Plasmid 45.. J. J. P. Pembroke. Insertion and excision of Bacteroides conjugative chromosomal elements. P. Peters. M.I.N.. we point towards the numerous examples of reclassification of bacterial genera and species that have resulted from ribosomal RNA-based phylogenetic studies over the last 15 years. or in the case of R391 and SXT for prfC. 285–291. J. but a number of additional elements. De Boever. M. D.. 8.W. As a third group we propose the adoption of the generic term mobilizable transposon. Cloeckaert. 973–980. B. 995–997. 28. TnpZ. J. 2002). Hochhut. Boyd.. Lett. Shoemaker. 2001.P.A.. Coetzee.. Donadio. D. 1989. Transfer region of a Bacteroides conjugative transposon..W.L. Guerineau..B.. For MGE. 2002. Wu.K.. J. Smokvina. 242.B. Mol. 185– 193. Boyd. 2000..

.. C. 791–794. (USA) 96. Hochhut. E. H. 5000–5008.A. Acad. 1979. Acids Res. D.P. 4718– 4726. nucleotide sequencing. 1997. Transfer functions of the conjugative integrating element pSAM2 from Streptomyces ambofaciens: characterization of a kil-kor system associated with transfer.210 A. 1997.. Clewell.. 1999. J. 376–381. 1997. Mol. Friedmann. Rood. 99–110. Lyras.. Lindsay.. EMBO Reports 2. Arnold.... 25. 2365–2373. Rev. Sezonov. B€ oltner / Plasmid 48 (2002) 202–212 osa strain JB2. 1997. D. E. Dillard.. 154. Hamilton. A. Schwartz. 134. 8. W. Monitoring genome evolution ex Tu vivo: reversible chromosomal integration of a 106 kb plasmid at two tRNALys gene loci in sequential Pseudomonas aeruginosa airway isolates. Goebel. Esposito. A. FEMS Microbiol. Ike. A. 1998. 1999. A. A. Sesma. 41. 641–679. Bacteriol. J.S. Pembroke. L. Hoflack. 177. Bacteriol. 1145–1152. Tsiamis. Mahillon.. 213– 225. G.. Lyras. Bacteriol. Farrow.L. Clewell. Nucleotide sequence and characterization of the cryptic Bacillus thuringiensis plasmid pG13 reveals a new type of rolling circle replicon. S. Gu erineau. R. 111–127. Toussaint. The integrase family of tyrosine recombinases: evolution of a conserved active site domain. Marre. Mol. 2000.. a conjugative transposon found in enterobacteria. J. Bacteriol. D. 1995.. The gene for toxic shock toxin is carried by a family of mobile pathogenicity islands in Staphylococcus aureus... F. Seifert.. J. J.B. 2001.J. 179. Kiewitz.. D. J. Plasmid 46. 2001.. Microbiol. Rev. 2097–2102. Ott. Microbiology 147. Pathog. Groisman. Pernodet. J... Microb. Muscholl-Silberhorn. Samberger. Proc. R.. E. 2000.J. 2001. Weinel.. Spontaneous deletions and flanking regions of the chromosomal inherited hemolysin determinant of an Escherichia coli O6 strain. €mmler... Smith. Pathogenicity islands: bacterial evolution in quantum leaps. Microbiol. C. J.. D... A functional origin of transfer (oriT) on the conjugative transposon Tn916. 1995. Hag ege. 494–502. K. Mol. Churchward. J. Bacteriol. Deletions of chromosomal regions coding for fimbriae and hemolysins occur in vivo and in vitro in various extra-intestinal Escherichia coli isolates. Taylor. Dillard... Knapp. J. D.V. Merlin. 403–430. 10875–10880. Gilmore. Y.. R. 4603– 4609. K.K.. Arment. 2001. Bacteriol. Haas.. Klockgether. Clewell.I. Novick. D. G. 1993. Mansfield. Hacker... 40–54.. J. 588–601. Lengeler. 527–543. J.. Transfer of the IncJ plasmid R391 to recombination deficient Escherichia coli K12: evidence that R391 behaves as a conjugal transposon.. J. 3605–3614. A variable genetic island specific for Neisseria gonorrhoeas is involved in providing DNA for natural transformation and is found more often in disseminated infection isolates.D. 29. which contains genes for virulence and avirulence.. M. W.. Insertion-duplication mutagenesis of Neisseria: use in characterization of DNA transfer genes in the Gonococcal Genetic Island..D. A. Morgan.. Francia. J. B. 1990.P.. 2001.. Scocca.J. J. Kurepina. D. P erez-Lesher. J.. Springael. F. 1981. C.. Micro.S. Pathogenicity islands and the evolution of microbes. Tn916 target DNA sequences bind the C-terminal domain of integrase protein with different affinities that correlate with transposon insertion frequency. J. Gerbaud. Yuroff. Tn4371: a modular structure encoding a phage-like integrase. G. G.F. H. Hickey. N. 5529–5538..R. bctA: A novel pBF4 gene necessary for conjugal transfer in Bacteroides spp..A. Microbiol. Lu. Environ. 657–662. Natl. a Pseudomonas-like catabolic pathway..L.. 145. Weaver.. Evidence for a chromosome-borne resistance transposon (Tn916) in Streptococcus faecalis that is capable of ‘‘conjugal’’ transfer in the absence of a conjugative plasmid. Ross.L. Matthew. M. 2717–2728. Rood.. 1938–1946. J.P.. Site-specific integration of the conjugal Vibrio cholerae SXT element into prfC. 67.M. An.. A. W. Hacker.. D.W.T. Ann.I. mobile cluster of biodegradation genes from Pseudomonas aerugin- . J. C. Murphy. Murillo. J.. 263–277. J. J.B. Larbig. Cell 87. 54.W. J. Lett. Bacteriol.L. Bacteriol... Completion of the nucleotide sequence of the Enterococcus faecalis conjugative virulence plasmid pAD1 and identification of a second transfer origin. Gibbon. Mark Osborn. 32.. Seurinck.B. R.B. CTnscr94. Ochman.. J. Types of beta-lactamase determined by plasmids in Gramnegative bacteria. J. Sci. J.. 138.. Wingender. B.E. Schmid. K.Y. Ruzin. W.. Transposition of Tn4451 and Tn4453 involves a circular intermediate that forms a promoter for the large resolvase. M. Hochhut.. Lederberg. E. J. 183. Bacteriol. Vivian. M.. Microbiol. Micro. J.... K. Hedges.. Macrina.W. Genomic analysis of the erythromycin resistance element Tn5398 from Clostridium difficile. Jahreis. 2001. 2155–2165. J. Plasmid 41. 1999.. K. A.. R. Cell genetics and hereditary symbioses. 1996. Kaper. M. and RP4/Tilike transfer functions...J. F. Waldor. J..B. Jaworski.. R. Lund. Sabat. 6644–6651. TnpX. 153–158. Franke. Appl. 38. Carniel. Goebel. K... Cloning.. genomic islands and bacterial pathogenicity. Bender. Microbiology 146.W. Hacker. A Darwinian view of the evolution of microbes. J. on a large native plasmid in the bean pathogen Pseudomonas syringae pathovar phaseolicola. Jackson. H.. Athanassopoulos.. Identification of a pathogenicity island.. Mol.... 1952.. Physiol. J. J.. B.A.. L.E. A. Hacker. M. Wirth. 177. M... 1983.. J. D. 175. Nucl. Microbiology 143.S.... D. B. Ecological fitness. 2000. H. 32. 179.. 1995. and functional analysis of a novel.T..

.. 38. Shankar. Rowe.....W. Garrett. J.A. 2002..B. Infect. N.. J.. Mobile elements as a combination of functional modules.T. Brown. Bacteriol..R. S. M. Tominaga. Y. 1998. Zillig. D. Iida.. R.. 928– 936. R.M.. 211 Shoemaker. Bacteriol. Bacteriol. Kwon.D.. 181. N. Furukawa.. Novick. 182. Q. 6840– 6847.R.. Ritchie. 1998. J.Y. Nishi. Two novel type III-secreted proteins of Xanthomonas campestris pv.. R. J..A..P.. Elliot..G. D.. Thorpe.A. Gen. J. I... Bacteriol. Koehler. Immun. Park. Gen.. S. G. J. 6509–6515. H... Nature 417. J... 2001. Yokoyama. A. Prescott. 579–590. 1996b. H. H. Churchward.D.. Hines. Prangishvili. Plasmid 3. Hampton. Okinaka.. Mark Osborn..F. Singh.R. Ogino. 1980.S. D. Smith. 26– 35. Mol. Rec. 435–439. Bacteriol. A.A. Mol. Fleetwood. Salyers.C.. Sekizaki. 2000. van der Meer.M.. H... Evolution of rhizobia by acquisition of a 500 kb symbiosis island that integrates into a phe-tRNA gene. Bacteriol.. 49.. Martinez.V. P. K.. Salyers. D. A.. 2002..B.J.. Gilmore. H.. 367–397.. integrates into the 30 end of a Leu-tRNA gene and has an integrase that is a member of the lambda integrase family. G. V. Unravelling the essential role in conjugation of the Tra protein of Streptomyces lividans plasmid pIJ101 Anton.A. Stedman... R. 4360–4369. Landy. DNA sequence and comparison of virulence plasmids from Rhodococcus equi ATCC 33701 and 103. J. Ann. G.. A. D. 365–377. G. G. Toussaint. Tribble. Thieme.J.. Bacterial transposases and retroviral integrases. C. 137.E. C. 41. Mol. Manter.. J. €by. A. Stevens. W. S. Shoemaker.. Molecular genetics of SaPI1—a mobile pathogenicity island in Staphylococcus aureus.. A. 2001. Yamaichi.W. Nennstiel.A..M. Ravatn. F.. T... T. H. Cloud. M.G.. Peters.. Microbiol. Nakamura.N. Springael. Vet. Pettis. Microbiol. Alperin. Guild.. Wang. Holz..W. S. Comparative sequence analysis of the symbiosis island of Mesorhizobium loti strain R7A. 178. Sequence and organization of pX01..B.C. Plasmid 47. J. D. C. F. Merlin. Hoffmaster... Shoemaker. M. Similarities and differences among 105 members of the Int family of site-specific recombinases.M. Trzebiatowski. N. Suzuki. Smith. 2000. Mol... J. 1340–1348. Y. C. She. Modulation of virulence within a pathogenicity island in vancomycin-resistant Enterococcus faecalis. Salyers. 79. G. Ruzin.A. 5145–5149. N.B.. Stedman..J. 228. S. P..T..E. K. J. 182. Wang.. O. 68.-R. 80–87. Kumano. P. The transfer origin for Bacteroides mobilizable transposon Tn4555 is related to a plasmid family from Gram-positive bacteria. Phan..J. L.. Mol. 44. Ward...R. Microbiol. Bacteriol. Microbiol. Studer. Rev. L. A. 2002.. A. 1996a.. 180. J... Hobman. No€ el..... The Bacteroides mobilizable insertion element. 417–425. Takai. Extremophiles 2. 294–299. A.. Jackson. Cruickshank. D.. A filamentous phage associated with recent pandemic Vibrio parahaemolyticus O3:K6 strains. S. L.T. V. Parker. Clin. Sullivan.M. Genet. vesicatoria are encoded within the hrp pathogenicity island. 3086–3095.. 178.....A.. Int.S.. R.. the large Bacillus anthracis plasmid harboring the anthrax toxin genes. Zillig. Makino. Ronson. tandem amplification. Rossbach. McCallum. 3601–3607.L. Genetic profile of pNOB8 from Sulfolobus: the first conjugative plasmid from an archaeon... Chromosomal integration. A. Chandler.. Shinagawa. K.A. Li. can integrate nonspecifically in Escherichia coli.. Zehnder. N. N. DNAse-resistant transfer of chromosomal cat and let insertions by filter mating in pneumococcus. Frost.J. Rev. Diversity in the serine recombinases.S. C. Albers. K. 1949–1955. Weaver. 746–750. 2000. A 90kilobase conjugative chromosomal element coding for biphenyl and salicylate catabolism in Pseudomonas putida KF715. Microbiol.. Takamatsu. J.. Multiple gene products and sequences required for the excision of the mobilizable integrated Bacteroides element NBU1. S.R. Microbiol.S..M... 299–307. Stuart..J.B. 1998.. Cohen. H. Novel mercury resistance determinants carried by IncJ plasmids pMERPH and R391.M. Gouzy. Honda.. T. G.. Sullivan. Leeuwen. Shoemaker.. Scott.. E. Nucl.. K.. Lindsay. 1998. 1998. 26. D. 59. 184. B€ oltner / Plasmid 48 (2002) 202–212 Nasu. Keim. G.... Bonas. NBU1. Nicholson. Conjugative transposons: an unusual and diverse set of integrated gene transfer elements. 1995. Bacteriol.-R.J.. Baghdayan. 1995.. Webby. H. Lamke. NBU1. Ronson. M. Parker. Bacteriol.K. M. She. 184. B. J.. 247–250.. 2002.D. 2000. and deamplification in Pseudomonas putida F1 of a 105-kilobase genetic element containing the chlorocatechol degradative genes from Pseudomonas sp. D. (USA) 95. J. Tirumalai. 1994. Shoemaker. J. Garrett. de Bruijn. Epidemic in cattle and humans of Salmonella typhimurium DT104 with chromosomally integrated multiple drug resistance.. 1999. A.M.E. 391– 406. Microbiol. The Bacteroides mobilizable transposon Tn4555 integrates by a site-specific recombination mechanism .S.. T. W.. J. Ricke. N. Strike. Acad. pING family of conjugative plasmids from the extremely thermophilic archaeon Sulfolobus islandicus: insights into recombination and conjugation in Crenarchaeota.. A. J. K. 1997. D. 577... Conjugative transposition. J. W. T. 7014–7020. K.. U. R. Hill. J. Threlfall. Sugahara. J. Kakuda. K. Acids Res. T. K. J. 1991. Svensson.... 1995. 182. strain B13. Salyers. Wang.. C.. Osaki. 3594– 3600. Proc. R. A. 2156–2161. K. Dan. EllenNunes-Du berger. Smith..J. R.. 13–23. Smith. N. P. Sci. Bacteriol. 180. S.S. 2000. Mahillon. R. R. 15. 2002.A. S.J. Phan. U. J. a mobilizable site-specific integrated element from Bacteroides spp. T.. Natl. Q.R. A. Polard.

G. Rose. Bacteriol. J.K. N. M. Bacteriol. 2731–2739. 183.. Nested deletions of the SRL Pathogenicity Island of Shigella flexneri 2a. J. Luck. Bacteriol. 4157–4165.J. 30. B. Infect. B€ oltner / Plasmid 48 (2002) 202–212 tance to sulfamethoxazole. Goldberg..... Nucl. Immun. F.A. Adler.. Burland.. J. 3271–3285.R. similar to that of the Gram-positive bacterial element Tn916. 679–699. 1996. Rajakumar. M..D. Lederberg. Mark Osborn. trimethoprim. 5535–5543.. Zinder. J.. Venkatesan. Bacteriol.. Williams.B.. 64. 2000. H. K. 2001. Sakellaris. NBU2.. 1952.M. A. Waldor.J. Mekalanos. J. A new type of conjugative transposon encodes resis- .. H. Genetic exchange in Salmonella.-R. N. Turner.. 866–875. V.A. Integration sites for genetic elements in prokaryotic tRNA and tmRNA genes: sublocation preference of integrase subfamilies. 2002.P. S. M. Wang. 3559–3571.J.. K.. Tschape.. and streptomycin in Vibrio cholerae O139. 69. Wang. E. which carries a functional lincomycin resistance gene... D. Grotbeck.. 182. Acids Res. Complete DNA sequence and analysis of the large virulence plasmid of Shigella flexneri. 178. Salyers. Shoemaker.N.212 A. J. J. J. D. Blattner. 2001. Bacteriol. 179. S. Characterization of a Bacteroides mobilizable transposon.