A cryptic plasmid of Yersinia enterocolitica encodes a conjugative transfer system related to the regions of CloDF13 Mob and IncX Pil

Eckhard Strauch1, Greta Goelz1, Dorothea Knabner1, Antje Konietzny1, Erich Lanka2 and Bernd Appel1

1 Robert Koch-Institut, Projekt Horizontaler Gentransfer, Nordufer 20, D-13353 Berlin, Germany
2 Max-Planck-Institut für Molekulare Genetik, Dahlem D-14195 Berlin, Germany

Correspondence
Eckhard Strauch
strauche{at}rki.de


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Yersinia enterocolitica 29930 (biotype 1A; O : 7,8), the producing strain of the phage-tail-like bacteriocin enterocoliticin, possesses a plasmid-encoded conjugative type IV transfer system. The genes of the conjugative system were found by screening of a cosmid library constructed from total DNA of strain 29930. The cosmid Cos100 consists of the vector SuperCos1 and an insert DNA of 40 303 bp derived from a cryptic plasmid of strain 29930. The conjugative transfer system consists of genes encoding a DNA transfer and replication system (Dtr) with close relationship to the mob region of the mobilizable plasmid CloDF13 and a gene cluster encoding a mating pair formation system (Mpf) closely related to the Mpf system of the IncX plasmid R6K. However, a gene encoding a homologue of TaxB, the coupling protein of the IncX system, is missing. The whole transfer region has a size of approximately 17 kb. The recombinant plasmid Cos100 was shown to be transferable between Escherichia coli and Yersinia with transfer frequencies up to 0·1 transconjugants per donor. Mutations generated by inserting a tetracycline cassette into putative tri genes yielded a transfer-deficient phenotype. Conjugative transfer of the cryptic plasmid could not be demonstrated in the original host Y. enterocolitica 29930. However, a kanamycin-resistance-conferring derivative of the plasmid was successfully introduced into E. coli K-12 by transformation and was shown to be self-transmissible. Furthermore, Southern blot hybridization and PCR experiments were carried out to elucidate the distribution of the conjugative transfer system in Yersinia. In total, six Y. enterocolitica biotype 1A strains harbouring closely related systems on endogenous plasmids were identified.


Abbreviations: T4SS, type IV secretion/transfer system

The GenBank accession number for the sequence reported in this paper is AJ519722.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
The genus Yersinia is composed of 11 species including the three pathogenic species Y. pestis, Y. pseudotuberculosis and Y. enterocolitica (Carniel, 2002). All the pathogenic strains share a common virulence plasmid pYV of about 70 kb which encodes a type III secretion/transfer system delivering Yop (Yersinia outer protein) virulence effector proteins into eukaryotic cells (Cornelis, 2002). The type III secretion system of Yersinia has been intensively studied and is regarded as an archetype of the type III secretion systems. Additional type III systems have been found in other pathogenic bacteria delivering effectors or toxins in a contact-dependent mechanism to animals and plants (Hueck, 1998; Page & Parsot, 2002). Type III secretion systems are built from components that share a high degree of homology to subunits of the flagellar machine (Young et al., 1999).

Another type of bacterial secretion/transfer system delivering virulence factors by secretion or cell-to-cell transfer is the type IV secretion/transfer system (T4SS). While a number of Gram-negative pathogenic bacteria use T4SSs for transfer of virulence factors, e.g. Agrobacterium tumefaciens, Bartonella tribocorum, Brucella spp., Helicobacter pylori, Bordetella pertussis, Legionella pneumophila, Rickettsia prowazekii (Christie, 2001; Baron et al., 2002), so far in Yersinia the occurrence of a T4SS has not been reported. The prototype of T4SSs is the VirB system of A. tumefaciens that exports the oncogenic T-DNA across the bacterial membrane into susceptible plants, which leads to the formation of crown galls (Lai & Kado, 2000). The virB locus of A. tumefaciens consists of 11 genes (virB1 through virB11) encoding a pilus-like mating system and a mating channel that spans the bacterial cell wall. The family of T4SSs is defined on the basis of sequence homologies between the components (Cao & Saier, 2001) and includes conjugative transfer systems that mediate the spread of broad-host-range plasmids between Gram-negative bacteria (Christie & Vogel, 2000). Bacterial conjugation requires two functional classes of proteins. One group of proteins builds the mating pair formation (Mpf) complex which is composed of a conjugative pilus for establishing contact with a recipient cell and a mating channel through which the DNA–protein intermediate is translocated across the bacterial cell wall. Another set of proteins carries out DNA transfer and replication reactions (Dtr) by forming the relaxosome for the preparation of the DNA single strand to be transferred (Willetts & Wilkins, 1984; Pansegrau & Lanka, 1996). Conjugation can be regarded as a rolling-circle-type DNA replication system linked to a T4SS. The two processes are connected by an additional component, the coupling protein, which links the relaxosome to the membrane transporter (Llosa et al., 2002). Coupling proteins are found in all conjugative plasmids, e.g. the TrwB of IncW plasmid R388 (Gomis-Rüth et al., 2002), TraG of IncP plasmids, VirD4 in the Agrobacterium Ti plasmids, TraD of IncF plasmids (Pansegrau & Lanka, 1996) and TaxB in IncX plasmid R6K (Núñez et al., 1997), but are missing in most mobilizable plasmids (Llosa et al., 2002).

Here we report the characterization and distribution of a conjugative transfer system in strains of Y. enterocolitica. The species Y. enterocolitica is heterogeneous, comprising pathogenic strains belonging to the biotypes IB and II–V, carrying the Yop virulon plasmid, and a large number of environmental strains which belong to the biotype 1A. The latter are regarded as nonpathogenic as they do not possess the virulence plasmid. However, the pathogenic potential of Y. enterocolitica biotype 1A strains is controversial, as in some reports the isolation of biotype 1A strains from clinical samples was described and a significant association of Y. enterocolitica biotype 1A strains with gastrointestinal diseases was postulated (Burnens et al., 1996; Bottone, 1997; Grant et al., 1999).

The T4SS described in this study was discovered in Y. enterocolitica 29930, a biotype 1A strain that releases the phage-tail-like particle enterocoliticin into the medium, which acts as a bacteriocin on a number of pathogenic and apathogenic Yersinia strains and rapidly kills sensitive strains, probably by forming channels in the membrane (Strauch et al., 2001). The Yersinia conjugative system is encoded by a cryptic plasmid present in the bacteriocin producer.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Strains, plasmids and culture conditions.
Strains and plasmids used in this study are described in Table 1. E. coli, Y. enterocolitica and Y. pseudotuberculosis were grown in Luria broth (LB) and on LB agar (Sambrook et al., 1989).


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids used in this study

 
Mating experiments.
Mating experiments were performed on solid media with late-exponential-phase cultures of donor and recipient strains according to standard procedures (Simon et al., 1983). Matings were carried out by mixing donor and recipient strains in a ratio of 1 to 10 per filter. After incubation of the filters at 28 °C for 2 h on LB agar the bacteria were plated on selection plates. Transfer rates were calculated as the number of transconjugants per donor during the 2 h period. Transconjugants were confirmed by randomly picking clones from the selection plates and were analysed by plasmid isolation, subjected to PCR-reactions with appropriate primers (see Table 2), and/or in some cases serotyping with standard sera against Yersinia O antigen. Selective media used in mating experiments were LB agar and Yersinia selective agar containing cefsulodin-irgasan-novobiocin (CIN, Oxoid) plus appropriate antibiotic combinations (kanamycin, 100 µg ml-1; tetracycline, 10 µg ml-1).


View this table:
[in this window]
[in a new window]
 
Table 2. PCR primers

 
Construction of a cosmid library.
A cosmid library was constructed using the SuperCos1 Cosmid Vector Kit following the recommendations of the manufacturer (Stratagene). SuperCos1 carries the Tn5 kanamycin-resistance (KmR) gene. The cosmid vector was digested with XbaI, dephosphorylated with alkaline phosphatase and digested with BamHI. Total DNA of Y. enterocolitica strain 29930 was partially digested with Sau3A and ligated with SuperCos1. Using a lambda packaging mix (Gigapack III Gold, Stratagene) the ligation mixture was introduced into E. coli XL-1 Blue MR cells, which were plated on LB agar containing 100 µg ampicillin ml-1. Cosmids used for further analysis were introduced into E. coli DH5{alpha} by transformation (Sambrook et al., 1989).

DNA manipulations, hybridization, PCR, sequence analysis.
Genomic DNA of Yersinia strains was prepared using the CTAB method of DNA extraction (Ausubel et al., 1987). Plasmids and cosmids were isolated using the NucleoSpin Plus Miniprep Kits (Clontech). In the case of preparations of large plasmids from Yersinia strains an alkaline lysis method followed by phenol/chloroform extraction was applied (Birnboim & Doly, 1979). Plasmid profiles of wild-type strains and transconjugants were obtained according to the method of Kado & Liu (1981). Restriction enzyme analysis, ligations, etc., were performed with commercially available enzymes according to the manufacturer's recommendations. Southern blot hybridizations were carried out following standard procedures (Sambrook et al., 1989) using fluorescein-labelled DNA probes. Amplification of DNA by PCR was performed with 30 cycles [30 s 96 °C, 30 s annealing temperature (see Table 2) and 30 s 72 °C for each cycle].

The insert DNA of cosmid Cos100 was sequenced by subcloning restriction fragments in standard cloning vectors (pLitmus28, New England Biolabs) and by primer walking. Sequencing reactions were carried out using dye terminator chemistry (PE Applied Biosystems) and separated on an automated DNA sequencer (ABI PRISM 3100 Genetic Analyser). The sequences were analysed using the Lasergene software (DNASTAR) and the Mac Vector software (Oxford Molecular Group) to assemble, to align and to determine putative ORFs. Sequence similarity searching of the current version of GenBank (National Center for Biotechnology Information; http://www.ncbi.nlm.nih.gov/BLAST/) was accomplished with the BLASTN, BLASTP, or BLASTX algorithm (Altschul et al., 1997). Promoter searches were performed with the Neural Network Promoter Prediction (NNPP) program (available at http://www.fruitfly.org/seq_tools/promoter.html).

Transposon mutagenesis.
In vitro mutagenesis of Cos100 DNA was performed with the EZ : : TN<TET-1>insertion Kit according to the manufacturer's protocol (EPICENTRE). The insertion locus of the tetracycline cassette in the target DNA was determined by sequencing using the TET-1 FP-1 and TET-1 RP-1 primers delivered with the kit. The cryptic plasmid of strain 29930 was tagged with a KmR gene using the suicide vector pUTKm of the mini-Tn5 transposon delivery system carrying the KmR gene from Tn903 (Herrero et al., 1990; de Lorenzo et al., 1990). Matings were performed on LB agar by mixing the E. coli S17-1({lambda}pir)(pUTKm) donor strain with the recipient Y. enterocolitica 29930. KmR Yersinia recipients were selected on Yersinia agar containing cefsulodin-irgasan-novobiocin (CIN, Oxoid) and 100 µg kanamycin ml-1. The insertion of the KmR gene into the large cryptic plasmid of strain 29930 was identified by hybridization of plasmid preparations to a 556 bp fluorescein-labelled probe, which was PCR amplified from the KmR gene using the primers Tn903S1 (5'-CGAGGCCGCGATTAAATTCCAAC-3') and Tn903AS1 (5'-TGAGTGACGACTGAATCCGGTGAG-3'). The sequencing primers PutneoR3 (5'-TGGCTCATAACACCCCTT-3') and PutneoRf (5'-ATTACGCTGACTTGACGGGACG-3') were used to determine the insertion of the KmR gene derived from the pUTKm vector. The absence of the mob region of the pUTKm vector was tested by PCR using the primers R6Kmob2 (5'-ATTGTCACGCTCAAGCCCGTAG-3') and R6Kmob3 (5'-CTTCTTCACTGTCCCTTATTCG-3') (Strauch et al., 2000b).


   RESULTS AND DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Cosmid Cos100 consists of the vector SuperCos1 and a 40·3 kb insert of genomic DNA of Y. enterocolitica 29930. The recombinant plasmid was identified by hybridization studies aimed at isolating the biosynthetic genes of enterocoliticin, a phage-tail-like bacteriocin produced by strain 29930 (Strauch et al., 2001; Goelz et al., 2003). Sequencing of a 3·2 kb EcoRI fragment of the insert DNA of Cos100 revealed the presence of ORFs encoding putative products with significant homology to proteins of the mating pair formation complex of T4SSs. Mating experiments were carried out and revealed that the cosmid Cos100 was self-transmissible (see below); therefore we decided to determine the complete sequence of the insert DNA.

Nucleotide sequence analysis
The entire sequence of the insert DNA of Cos100 is exactly 40 303 bp (GenBank accession no. AJ519722). Significant homologies to database entries were identified for 44 ORFs of the entire sequence. The positions and transcriptional orientations of these ORFs are depicted in Fig. 1. Table 3 lists the putative function, molecular mass, number of amino acids and closest relative for the predicted product of each ORF. The sequences between the annotated ORFs are below 350 bp with the exception of four regions (between ORFs 4/5, ORFs 11/12, ORFs 14/15, ORFs 31/32) which have a length between 450 and 900 bp. For ORFs in these regions (cutoff 120 bp) no database entries with significant homology were found. The overall GC content of the DNA is 50·4 mol%, which is slightly higher than the 48·5±1·5 mol% reported for Y. enterocolitica (Bercovier & Mollaret, 1984).



View larger version (28K):
[in this window]
[in a new window]
 
Fig. 1. Map of Cos100. ORFs are represented as arrows (see Table 3). The fill patterns indicate different postulated functions: stippling for mating pore formation, cross-hatching for DNA transfer processing functions, white for replication and maintenance. Other ORFs encoding proteins with predicted functions are black and ORFs encoding genes of unknown function are grey. Vertical hatching indicates the vector part of Cos100 derived from SuperCos1. The short radial lines mark the positions of insertions of the tetracycline cassette yielding Tra- derivatives.

 

View this table:
[in this window]
[in a new window]
 
Table 3. ORFs of the 40 303 bp insert DNA of Cos100

 
The sequence analysis made it likely that the insert DNA was derived from a cryptic plasmid present in Y. enterocolitica 29930 (see below). Genes involved in plasmid functions were identified, such as replication genes and partitioning genes. Additionally, genes involved in conjugative DNA transfer, some putative housekeeping genes and a number of sequences derived from mobile genetic elements (transposons, IS elements) commonly found in plasmids were discovered.

Conjugative plasmid transfer
The genes of the conjugative transfer region lie within a DNA segment of 17 kb (gene nos 18–34) and were divided into two functional classes: one class encoding proteins of the Mpf system and the second class encoding proteins performing DNA transfer and replication (Dtr system). Proteins encoded by the latter class of genes were found to be homologous to the topoisomerase TraE of RP4 and to the MobB/C proteins of the mobilizable plasmid CloDF13 (Fig. 1, Table 3).

The Mpf system encoded by a gene cluster (genes nos 18–28) in the insert DNA of Cos100 is a member of the T4SSs. Plasmid-specified T4SSs of Gram-negative bacteria are involved in DNA transfer during conjugation, while chromosomal systems of this type found in pathogenic bacteria are implicated in the transfer of proteinaceous virulence factors (Christie & Vogel, 2000). Based on the conjugative properties of Cos100 and the presence of a replicon of the I-complex plasmids in the insert DNA derived from Y. enterocolitica 29930 (see below), we designated all putative genes found in this region tri (transfer genes of I-complex plasmid). We identified in total 10 ORFs (triA–triJ) which may encode a set of products responsible for the mating pore formation (Mpf), which is involved in pilus biogenesis and initial contact with the recipient cell. The order and arrangement of all putative tri genes is collinear to the virB operon of A. tumefaciens, which is regarded as the reference point for T4SSs (Christie, 2001). The best homologies of the tri gene products are shown in Table 3 and indicate a close relationship of the tri operon to the pilX genes of the IncX plasmid R6K (Núñez et al., 1997).

The triA product, containing a transglycosylase domain, is likely to be a VirB1-like protein. VirB1 analogues are thought to cause local lysis of the peptidoglycan layer, thus opening the cell wall temporarily during conjugative DNA transfer (Bayer et al., 1995). The next ORF of the putative tri operon, triB, may encode the pre-pilin with a length of 97 aa. Conservation in pilins manifests itself in structural but not in sequence details; structural features are the presence of an N-terminal peptidase I cleavage site, and three peaks of mean hydrophobicity (Cao & Saier, 2001). The predicted pre-pilin of the triB gene possesses a putative peptidase cleavage site at position 29–30 (27AFA{downarrow}DQ31) (Paetzel et al., 2000), and three transmembrane segments, two of which are located in the putative mature pilin peptide. Also a cyclization motif 89AEIG92 is found at the C-terminus of the putative pilin protein (Kalkum et al., 2002). A number of homologues to the putative TriB protein were identified within the VirB2 family, with the TraC protein of the environmental plasmid pSB102 showing the highest similarity (Schneiker et al., 2001).

Downstream from the triB gene the next ORF may encode a protein of the tri cluster with a size of 915 aa. The predicted protein was designated TriC. Alignment of the N-terminal part of the TriC protein (aa residues 1–65) showed homology to proteins of the VirB3 family, while the remaining amino acid chain of the putative gene product is significantly similar to VirB4 homologues. Proteins of the VirB3 family play a role in the pilus assembly and interact with VirB4 and VirB6 proteins by stabilizing the channel (Hapfelmeier et al., 2000; Christie, 2001). VirB4 proteins contain NTP-binding sites and are supposed to energize the transport of macromolecules. The putative TriC protein possesses all four motifs recently described for proteins of the VirB4 family: motif A Walker A box (541GTAGSGKT548); motif B Walker B Box (754VIFMDEFW761), motif C (705DNDLD709) and motif D (781RKLNGVNIPGTQ792) (Cao & Saier, 2001; Rabel et al., 2003). In most T4SSs VirB3 and VirB4 proteins are encoded by separate genes; however, the putative magB03 gene of plasmid pVT745 of Actinobacillus actinomycetemcomitans also encodes a putative product encompassing VirB3 and VirB4 homologous proteins in a single polypeptide chain (Galli et al., 2001).

The triD and triE products, related to VirB5 and VirB6 proteins, show the highest similarity to two putative Y. pestis proteins which are present on a small, cryptic plasmid with a size of 5919 bp found in some plaque isolates from China (Dong et al., 2000).

In the triDtriE intergenic region a short ORF (orf22) is found which may encode a protein of 97 aa. The predicted product shows similarity to the Eex protein of the IncX plasmid R6K (Table 3) and contains a putative conserved lipoprotein modification/processing site (20LGG{downarrow}C23). As the eex gene of plasmid R6K is also positioned in the corresponding region of the IncX Mpf system (between the pilX5 and pilX6 genes), it seems likely that the orf22 product may produce the entry exclusion phenotype for the tri-encoded Mpf system.

The putative products of the triF, triG, triH, triI and triJ genes are closely related to the corresponding products of the pilX genes (pilX7–pilX11) of the mating pore formation system of the R6K plasmid. The predicted TriF peptide contains a putative conserved lipoprotein modification/processing site (13LAG{downarrow}C16). In the A. tumefaciens T4SS the corresponding proteins VirB7, VirB8, VirB9 and VirB10 (and VirB6) participate in forming the mating channel across the periplasmic space, while VirB11 may form the cytoplasmic pore (Yeo et al., 2000) possessing additional chaperone function and ATPase activity, which could serve to supply the energy for the transfer process (Christie, 2001; Burns, 1999). The putative TriJ product of the tri gene cluster possesses Walker A and Walker B boxes (167GGTGSGKT174; 233RMNPDRILLAEVRG246), an Asp box and a His box typically found in VirB11 family transport NTPases (Krause et al., 2000) and may be in addition to the TriC protein also be involved in energizing the macromolecular transfer.

Downstream from the triJ gene three ORFs were found which may play a role in conjugative transfer; however their role remains hypothetical in the context of our study. The putative product of orf29, whose coding region overlaps with the last five codons of triJ, shows similarity to phage integrases, a family of proteins that cleave DNA substrates by a series of staggered cuts, during which the protein becomes covalently linked to the DNA (http://www.ncbi.nlm.nih.gov/Structure/cdd/cdd.shtml). orf30 is sequence-related to the traE gene of IncP1 plasmids, which encodes DNA topoisomerases of type III. TraE proteins are nonessential for the conjugation process, as chromosomally encoded topoisomerases can substitute their function (Li et al., 1997). The putative product of orf31 is related to proteins of the H-NS (histone-like nucleoid-structuring) family. Such proteins bind DNA and are involved in transcriptional regulation of gene expression by silencing (Yarmolinsky, 2000). One insertion of the tetracycline cassette into the coding region of orf31 led to a Tra- phenotype, suggesting a role of the putative product of orf31 in the transfer process.

Conjugative DNA transport requires another set of proteins that perform DNA-processing reactions (cleavage to provide the ssDNA destined to be transferred) and form the relaxosome during the DNA transfer (Pansegrau & Lanka, 1996). Genes encoding protein homologues of the MobB and MobC protein of the mobilizable plasmid CloDF13 were identified downstream from orf31 and were designated triK and triL. In mobilization of plasmid CloDF13 the proteins MobB and MobC act together in relaxing the plasmid DNA and cleaving the plasmid at the nic site. Additionally, MobB seems to function as the coupling protein which connects the relaxosome to the membrane-associated transport apparatus, thus allowing the plasmid CloDF13 to be mobilized efficiently by conjugative plasmids from many incompatibility groups (Núñez & de la Cruz, 2001). Remarkably, the sequence of Cos100 insert DNA does not contain another gene which may encode a putative coupling protein. Therefore, it seems likely that the predicted TriK protein, the homologue of MobB of plasmid CloDF13, acts as a coupling and Dtr protein in the conjugative system of Cos100.

The potential oriT of plasmid CloDF13 lies 347 bp upstream from the mobB gene on the (-) strand and is defined by the sequence 5'-GGGTG{downarrow}GTCGGG-3' (Núñez & de la Cruz, 2001). At 367 bp upstream from the putative triK gene of the Cos100 insert DNA a similar sequence 5'-GGGTGGTCACG-3' (9 out of 11 identical) is present on the (-) strand. In conjugative plasmids the oriT maps asymmetrically to the transfer genes, such that the tra genes normally enter the recipient cell last. Under the assumption that the deduced sequence upstream from triK serves as the oriT in the Cos100 conjugative system, the triK and triL genes would enter the recipient cells – in analogy to the CloDF13 system (Núñez & de la Cruz, 2001) – last. However, the majority of the transfer genes (triAtriJ) belonging to the Mpf system as well as orf29–orf31 would be transferred early in the conjugation process. It is noteworthy that the 11 bp sequence 5'-GGGTGGTCACG-3' is also present on the (+) strand at a distance of 281 bp upstream from the triK gene. It remains to be analysed if this sequence may play a role in the conjugative transfer.

Two more ORFs are present downstream from triKtriL that have been found in other conjugative plasmids in the vicinity of transfer genes. The putative product of orf34 is related to periplasmic endonucleases that cleave ss and dsDNA; however, a role of these enzymes in the biology of their respective plasmids remained unclear (Pohlman et al., 1993). The following ORF downstream (orf35) shows significant similarity to ardC genes encoding an antirestriction function that is specific for type I restriction and modification systems. The ArdC protein of the conjugative IncW plasmid pSA was shown to bind to ssDNA and is believed to be transported together with the transferred DNA into the recipient bacterial cell, thus protecting the incoming DNA against nucleases of the recipient cell (Belogurov et al., 2000). The putative ArdC protein of Cos100 shows a significant similarity to a number of ArdC proteins and the similarity to ArdC of plasmid pSA is 59·7 % (179 aa in a stretch of 295 aa); however, the antirestriction motif described for the ArdC protein of plasmid pSA (Belogurov et al., 2000) is only poorly conserved in the potential ArdC protein of Cos100.

As conjugative transfer systems are composed of building blocks, we investigated the GC contents of the region encoding the Mpf complex (genes triAtriJ) that are related to the IncX transfer genes of plasmid R6K and the region encoding the Dtr genes (triK, triL) with homology to the mobA/mobB genes of plasmid CloDF13. Remarkably, the Mpf region of Cos100 shows an overall GC content of 52·1 mol%, while the Dtr region of Cos100 has a GC content of only 45·7 mol%. The differences in GC contents are reflected in the codon usage; the Mpf-encoding genes possess in the third letter of the codon a GC content above 60 mol%, whereas in the same position of the Dtr-encoding genes the GC contents is significantly lower (46·2 mol% for triK and 51·0 mol% for triL). The observed differences between the two transfer regions of Cos100 indicate that the regions may have a divergent evolutionary origin and have not always existed as a transfer unit. We compared the two regions of Cos100 additionally with the Mpf region (pilX genes) of the IncX plasmid R6K (Núñez et al., 1997) and the mobA/mobB genes of plasmid CloDF13 (Núñez & de la Cruz, 2001). However, the discrepancies concerning the GC contents between the corresponding regions are striking and do not suggest a close evolutionary relationship of the different transfer regions. The GC content of the pilX cluster is 42·4 mol%, while that of the Dtr-encoding genes mobA/mobB of CloDF13 is 64·2 mol%.

Plasmid replication and partitioning
The insert DNA of Cos100 contains several ORFs encoding putative products with significant homologies to plasmid genes, thus indicating its origin from a cryptic plasmid of Y. enterocolitica 29930.

The putative replication region is located approximately 1·5 kb upstream from the triA gene. A putative repA gene was identified whose product showed the closest relationship to the putative RepA protein of the virulence plasmid pYVe8081 of Y. enterocolitica strain 8081, which belongs to the pathogenic serotype O : 8 strains (Snellings et al., 2001). A number of replication initiation proteins of plasmids belonging to the I-complex also show significant similarities to the deduced RepA protein, which suggests that the cryptic plasmid of strain 29930, the source of Cos100 insert DNA, contains a replicon of this type. I-complex plasmids comprise plasmids of the IncL/M, IncB, IncI{alpha} and IncK groups (Praszkier et al., 1991), which were grouped together based on DNA and protein homologies and the regulation by countertranscript RNAs. The analysis of the 5' region of the repA gene of the Cos100 insert DNA revealed the presence of a short ORF that partially overlapped with the repA coding sequence. In several plasmids of the I-complex such ORFs are found in the corresponding region of the repA gene and it was shown in some cases that the translation of the ORF is a prerequisite for the translation of the replication initiation protein. The putative RepB peptide of the insert DNA of Cos100 shows a significant similarity to RepB peptide of the replicon of plasmid pSW800 from Pantoea stewartii (Wu et al., 2001). Upstream from repA/repB we found additional elements that are essential for the replication of I-complex plasmids.

Incompatibility determinants of these plasmids are small countertranscript RNA molecules, which inhibit the translation of a target RNA (RNAII) whose product, the RepA protein, is essential for replication. The putative countertranscript RNA (RNAI)-encoding gene was easily identifiable by comparison with the corresponding sequences of the plasmids pYVe8081 (Snellings et al., 2001) and pSW800. The RNAI-encoding region stretches from nucleotide position 13330 to 13409 of the Cos100 insert DNA upstream from the start codon of the putatve repA and repB genes. The predicted RNAI molecule possesses the secondary structure that has been described for the countertranscript RNAs of the I-complex plasmids. Using an RNA folding computer program (GeneQuest; DNASTAR) a small stem–loop structure is found at the 5' end followed by a major stem possessing five conserved GC pairings and a loop containing the conserved sequence 5'-CGCCAA-3' (data not shown).

Using the CLUSTALW program an alignment between replication regions of the plasmids pSW800 and pYVe8081 and the Cos100 upstream from the repA sequence was performed and revealed stretches of identity between the sequences. Based on the results of pSW800, which has been studied experimentally in detail (Wu et al., 2001), nucleotide position 13462 of the Cos100 sequence can be deduced as the likely start point of RNAII transcription, and likely promoters for RNAI and RNAII (Table 4) can be identified in silico. Approximately 150 bp downstream from repA lies the sequence 5'-TTATGCACA-3', which may function as binding site for a DnaA protein. The sequence matches the DnaA box in eight of nine positions (Praszkier et al., 1991) and the region surrounding this site shows a higher AT content (about 60 mol%) than the overall AT content of the Cos100 insert. Immediately downstream from the DnaA box are three short sequences that may serve as RepA binding sites (starting 3, 27 and 48 nt downstream from the DnaA box). The consensus sequence 5'-tANCNGCAA-3' matches the RepA box with the exception of the leading T (Betteridge et al., 2003), thus indicating that an origin of replication of the cryptic plasmid of strain 29930 might be located in this region. It should be pointed out, however, that in our opinion the ColE1-derived origin of replication of the vector part of Cos100 (SuperCos1) probably serves as the main replication origin of Cos100, as comparison of the intensity of the plasmid band of Cos100 preparations was always significantly higher than the intensity of the band attributed to the large cryptic plasmid band of Y. enterocolitica strain 29930 (not shown).


View this table:
[in this window]
[in a new window]
 
Table 4. Putative promoters identified in the Cos100 sequence

 
Adjacent to the replication region (approx. 1·2 kb downstream from repA) a putative partition region is located. The region encompasses two ORFs encoding two putative proteins showing strong homologies to the ParF and ParG proteins of the Salmonella enterica plasmid TP228 (Hayes, 2000) and Erwinia amylovora plasmid pEA29 (McGhee & Jones, 2000). Partitioning proteins are active in the maintenance and distribution of low-copy plasmids to daughter cells at bacterial cell division (Bignell & Thomas, 2001). The putative parG gene of the insert of Cos100 lies separated by a short intergenic region of 43 bp downstream from the parF ORF. The intergenic region contains a potential ribosome-binding site (5'-AAGAGG-3') separated by 7 bp from the putative start codon of the parB ORF, suggesting that the two genes are transcriptionally coupled as in the case of plasmid TP228 (Hayes, 2000). The partition site of TP228 is not yet fully determined and probably lies in a 518 bp fragment encompassing the 5' upstream region of parF and the first 37 codons of parF (Hayes, 2000).

Miscellaneous genetic elements present in the insert DNA of Cos100
The sequence analysis of the insert DNA of Cos100 revealed a number of additional genetic elements, which are frequently encountered on plasmids. One group of such ORFs is likely to be derived from mobile genetic elements like IS elements or transposons. orf1 and orf2 are probably part of a transposon, as their two putative products are highly homologous to transposase TnpA and resolvase TnpR of Tn3-like transposons. A 38 bp sequence identical to a terminal inverted repeat of Tn3 is present at position 6731–6768, indicating that orf3, encoding a putative methyltransferase, and orf4, encoding a protein highly similar to a hypothetical protein of Klebsiella pneumoniae, may be part of a transposon. orf36 and orf37, encoding putative transposases, probably belong to an IS1-like insertion element. The next three ORFs (orf38–40) are partially related to transposase genes; however, the more detailed sequence analysis of the deduced gene products makes it unlikely that they are functional (Table 3). The deduced proteins of orf41, orf 42, orf 43 and orf 44 show a variable degree of similarity to other bacterial proteins (see Table 3).

The entire sequence was analysed for putative {sigma}70-dependent promoters with the NNPP program (see Methods) using a threshold score of 0·8, which recognizes 60 % of all promoters, with a rate of false positives of 0·4 %. Eight putative promoters of this analysis were selected by the logic of the gene organization (Table 4, Fig. 1). In three more cases promoters could be deduced by directly comparing published sequences with the analysed genes (orf1, orf2, repA/B; see Table 4).

Conjugative exchange of Cos100 in E. coli and Yersinia
To corroborate the presence of a functional conjugation system, matings between E. coli K-12 DH5{alpha} harbouring Cos100 and different Yersinia strains were performed on solid media using appropriate selective conditions. Donor and recipient cells were mixed on filters and transfer rates were calculated as the number of transconjugants per donor. Y. enterocolitica strains from the major pathogenic serotypes O : 3, O : 5,27, O : 8 and O : 9 (for strain designations see Table 1) were chosen as recipients. Furthermore the nonpathogenic Y. enterocolitica strain 29807/6 of biotype 1A and two Y. pseudotuberculosis strains (Table 1) were used in mating experiments. Transfer frequencies of Cos100 from E. coli to Yersinia ranged from 10-1 to 10-7, thus confirming the functionality of the transfer system. The highest transfer rates were achieved using the Y. enterocolitica strain 83/88 (serotype O : 5,27) and its derivatives as recipients.

Additional matings were performed to find out if Cos100 is also self-transmissible between Yersinia strains. Recipient Yersinia strains were marked by introduction of small plasmids (pACYC184 or pBR329) by electroporation (Chang & Cohen, 1978; Covarrubias et al., 1981). The matings revealed transfer frequencies between 10-1 and 10-7, thus demonstrating that the conjugative transfer system is also functional in Yersinia. Again, the highest transfer frequencies were obtained using derivatives of Y. enterocolitica 83/88 (serotype O : 5,27).

To identify transfer genes an in vitro mutagenesis of Cos100 DNA was performed by randomly inserting a tetracycline-resistance cassette into the cosmid DNA. The mutagenized plasmids were introduced into E. coli K-12 by transformation. Resulting tetracycline-resistant transformants (approx. 70) were used as donors in mating experiments with Y. enterocolitica 83/88/2. In these experiments 11 insertions of the tetracycline cassette were identified that abolished conjugative DNA transfer completely (detection limit was 10-7 transconjugants; the transfer rate with Cos100 in parallel experiments was up to 10-1). Table 5 summarizes the positions of the 11 insertions of the tetracycline cassette in the Cos100 sequence. All the insertions were found in coding regions proposed to be involved in plasmid transfer based on the sequence comparison. Nine of the insertions mapped in tri genes, which are likely to encode the mating pore formation system. Insertion 10 was in orf31, which encodes the HNS-like DNA-binding protein. Insertion 11 was in the putative triL gene, whose product has been implicated in DNA processing.


View this table:
[in this window]
[in a new window]
 
Table 5. Positions of insertion of the tetracycline cassette in Cos100 DNA abolishing conjugal transfer

Matings between E. coli K-12 derivatives (donor) and Y. enterocolitica 83/88/2 (recipient); transfer rate <10-7 transconjugants per donor.

 
To confirm the correct order of the genetic elements found in the Y. enterocolitica 29930 insert DNA of Cos100 we searched the cosmid library for additional cosmids carrying the transfer system, by hybridization with the 3·26 kb EcoRI fragment of Cos100. We were able to identify two more cosmids that were self-transmissible from E. coli to Y. enterocolitica 83/88/2 with approximately the same transfer frequency as Cos100. The two cosmids, Cos96 and Cos81, possessed a restriction pattern partially identical to Cos100. Cos96 corresponds to an internal fragment of Cos100 (positions 2602–38682). Cos81 is larger than Cos100 and comprises the sequence from nucleotide position 4403 to 40 303 of Cos100 followed by an additional sequence of Y. enterocolitica 29930 DNA.

Conjugation experiments with the cryptic plasmid of Y. enterocolitica strain 29930
Cos100 was produced by inserting partially digested SauIIIA genomic DNA of the enterocoliticin-producing strain Y. enterocolitica 29930 into BamHI-cleaved SuperCos1 vector. The sequence data (see above) made it likely that the insert DNA was derived from a cryptic plasmid and not from the chromosomal DNA of strain 29930. Southern blot hybridization using a gene probe derived from the triC region (Table 2) clearly confirmed that a plasmid visible above the chromosomal band gave a positive signal (Fig. 2, lane 5). To find out if the large plasmid is self-transmissible, we used the mini-Tn5 suicide vector pUTKm to tag the cryptic plasmid with a KmR gene (Herrero et al., 1990; de Lorenzo et al., 1990). After screening of several hundred transconjugants by hybridization with a probe (see Methods), we finally obtained eight stable transconjugants which had integrated the KmR gene into the large plasmid (Fig. 2). Fig. 2 also shows that Y. enterocolitica 29930 harbours at least one additional but smaller cryptic plasmid with an estimated size of 5–6 kb. We performed matings with the eight KmR strains of Y. enterocolitica 29930 with Y. enterocolitica 83/88/2 harbouring plasmid pBR329 as recipient; however, we did not detect transfer of the KmR plasmids (detection level for mating experiments was approx. 10-8). These experiments suggested that the conjugative transfer system is repressed in strain 29930.



View larger version (44K):
[in this window]
[in a new window]
 
Fig. 2. Insertion of a KmR gene into the large cryptic plasmid of Y. enterocolitica 29930. Agarose gel electrophoresis analysis (right panel) and Southern blots (left panels) of plasmid preparations (‘Kado’ lysis) of Y. enterocolitica 29930 wild-type (lane 5) and kanamycin-resistant derivative strains (lanes 1–4), and of E. coli K-12(Cos100) (lane 6). The blots were incubated with a triC probe (top left) and a Tn903-derived probe detecting the KmR gene of pUTKm (bottom left; the KmR gene of Cos100 is derived from Tn5). The KmR-conferring plasmids in lanes 1–3 are three out of eight derivatives of strain 29930 (see text). The extra band in lane 2 (right panel) is probably due to the alkaline preparation procedure. The plasmid preparation shown in lane 4 stems from a spontaneous KmR mutant of strain 29930. The arrow in the right panel indicates the plasmids carrying the tri region; each preparation of strain 29930 (lanes 1–5) contains a second small plasmid (band comigrating with the marker at 4254 bp in lane M).

 
By transformation of commercially available competent E. coli K-12 cells (‘Gene Hogs’; Table 1) we introduced the KmR gene carrying derivatives of the cryptic plasmids of Y. enterocolitica 29930 into E. coli. The plasmids of the E. coli transformants were reisolated and compared by EcoRV digestion to the restriction pattern of a plasmid preparation of strain 29930. EcoRV was chosen as restriction enzyme, as only fragments up to approximately 15 kb are generated. The EcoRV restriction patterns of most transformed plasmids revealed, however, that a number of bands were missing in comparison to the EcoRV restriction patterns of the original Yersinia plasmids (data not shown). Only in one case did a KmR-marked plasmid, designated p29930 : : Km5, show a restriction pattern nearly identical to the original Yersinia plasmid (Fig. 3, lanes 4 and 5). The plasmid p29930 : : Km5 was verified for the absence of the mob region of the suicide vector pUTKm and the presence of the KmR gene delivered by pUTKm (see Methods). The EcoRV fragment containing the latter gene is indicated by the arrow in Fig. 3(a).



View larger version (77K):
[in this window]
[in a new window]
 
Fig. 3. Conjugative transfer of the large cryptic plasmid of strain 29930 marked with a KmR gene. Agarose gel electrophoresis analysis of EcoRV digests of plasmids prepared from Y. enterocolitica 29930, Y. enterocolitica 83/88/2 and E. coli K-12 derivatives. (a) Lane 1, E. coli K-12(Cos100); lane 2, Y. enterocolitica 29930 wild-type; lane 3, Y. enterocolitica 29930 derivative carrying the KmR gene (integrated into large cryptic plasmid); lanes 4 and 5, transformants of E. coli K-12 containing plasmid p29930 : : Km5; lanes 6 and 7, transconjugants of Y. enterocolitica 83/88/2 carrying p29930 : : Km5.tra. The arrow indicates the fragment carrying the KmR gene of pUTKm. (b) Lane 1, E. coli K-12(Cos100); lane 2, Y. enterocolitica 29930 wild-type; lanes 3–6, four transconjugants of Y. enterocolitica 83/88/2(pBR329) carrying p29930 : : Km5.tra. The arrow indicates pBR329 DNA linearized with EcoRV.

 
Despite the differences in the restriction patterns we used all the E. coli transformants (two colonies from each transformation experiment, giving 16 donors in total) in a further round for mating experiments with Y. enterocolitica 83/88/2 as recipient. Only in the case of two donors harbouring plasmid p29930 : : Km5 was a conjugative transfer of the KmR gene to Y. enterocolitica, with a frequency of 10-4 transconjugants per donor, detected. The transferred plasmids (designated p29930 : : Km5.tra) were analysed by restriction digestion with EcoRV (Fig. 3a, lanes 6 and 7) and compared to the plasmids of the respective E. coli K-12 donors and the Y. enterocolitica 29930 parents. It turned out that the plasmids of the Y. enterocolitica 83/88/2 transconjugants showed a restriction pattern in which a number of bands were missing when compared to the pattern of the two E. coli K-12(p29930 : : Km5) donors. In separate experiments we checked the transferred plasmids by PCR for the presence of repA, triC, triH and triK sequences (see Table 2). From six randomly chosen transconjugants the desired products were successfully amplified, suggesting that all putative conjugative transfer functions and the putative plasmid origin of the IncI type are present in the transferred plasmids.

In subsequent matings the transconjugants of Y. enterocolitica 83/88/2 harbouring p29930 : : Km5.tra were used as donors and Y. enterocolitica 83/88/2(pBR329) as recipient. All donors transferred the plasmids with a frequency of about 10-3, thus confirming the transfer-proficient phenotype (Fig. 3b, lanes 3–6). We tried to determine the insertion point of the KmR gene in plasmid p29930 : : Km5 by direct sequencing of the plasmid using the primers R6Kneo and R6Kneo3 (see Methods). The sequences obtained showed the left end and right end of the mini-Tn5 ends of the pUTkm vector; however, the sequences outside the mini Tn5 cassette did not show any similarity to the Cos100 sequence. This suggests that the KmR gene cassette is inserted into a part of the cryptic plasmid outside the analysed Cos100 sequence.

The experiments were designed to investigate the conjugative system in its original host Y. enterocolitica 29930. As none of the KmR-tagged large cryptic plasmids were self-transmissible when present in the genetic background of Y. enterocolitica 29930, it appears that the conjugative system is repressed in strain 29930 under our experimental conditions. We have not investigated this phenomenon any further. Based on the EcoRV restriction digests (see Fig. 3a) we calculated the following plasmid sizes: the derivative plasmid p29930 : : Km5, introduced into E. coli K-12 by transformation and used as donors in the first matings, has a size of approximately 107±10 kb (lanes 4 and 5) and the Y. enterocolitica transconjugants harbour a KmR-conferring plasmid of 56±5 kb, designated p29930 : : Km5.tra (lanes 6 and 7).

Distribution of the T4SS in Yersinia
We investigated several Yersinia strains to elucidate the distribution of the T4SS in Yersinia. As the tri region is encoded by a cryptic plasmid, we first studied plasmid-harbouring strains from previous investigations (Lewin et al., 1996; Hoffmann et al., 1998). Plasmid preparations of 22 nonpathogenic Y. enterocolitica biotype 1A strains belonging to different serotypes, one Y. intermedia and one Y. frederiksenii strain were subjected to Southern blot hybridizations with probes generated by PCR amplification of Cos100 template using the primer pairs TriCfw/TriCbw and TriHfw/TriHbw (Table 2). Among these 24 strains the plasmids of five Y. enterocolitica biotype 1A strains gave a positive hybridization signal with the probes (Fig. 4). All the hybridization signals were derived from plasmid bands which were visible above the chromosomal DNA band in standard 0·8 % agarose gels. The Y. enterocolitica strains 29968 and 29984 each contain only one plasmid, while all other strains (including 29930) have more than one plasmid. We performed restriction digests of plasmids of strain 29968 and 29984 using EcoRV, EcoRI, HindIII and PstI. These digests did not indicate similarity of the restriction patterns of p29968 and p29984 and also no similarity to the patterns of plasmid preparations from strain 29930 (data not shown). The estimated sizes were 53±5 kb for the plasmid of strain 29984 and 60±5 kb for the plasmid of strain 29968.



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 4. Detection of Y. enterocolitica biotype 1A strains harbouring cryptic plasmids with the tri transfer system. Agarose gel electrophoresis analysis (left panel) and Southern blot (right panel) of plasmid preparations (‘Kado’ lysis) of E. coli K-12(Cos100) (lane 1), Y. enterocolitica 29930 (lane 3), Y. enterocolitica 29931 (lane 4), Y. enterocolitica 29968 (lane 5), Y. enterocolitica 29984 (lane 6), Y. enterocolitica 29971 (lane 7) and Y. enterocolitica 56 (lane 8). Hybridization was performed with a triC probe.

 
We additionally investigated several pathogenic Y. enterocolitica and Y. pseudotuberculosis strains from which genomic DNA was prepared. These experiments were performed to find out (i) if a chromosomally encoded T4SS may be present and (ii) if plasmid-specified T4SS may be present, as some of the pathogenic strains harbour cryptic plasmids in addition to the virulence plasmid pYV (S. Hertwig, personal communication). DNA of 12 serotype O : 3,10 serotype O : 9, 4 serotype O : 5,27 and 5 serotype O : 8 Y. enterocolitica strains and 12 Y. pseudotuberculosis strains (different serotypes) from the strain collection of the Robert Koch-Institut were tested, but did not hybridize.

To confirm the presence of a tri region in those five Y. enterocolitica biotype 1A strains that gave a hybridization signal with the two tri probes, we performed a number of PCR reactions with primers derived from the sequenced tri region of Cos100 (Table 2) using plasmid preparations of the five strains as template. All primers gave PCR products of the expected size, and sequencing of the products showed that they were all either identical or highly similar (Table 6).


View this table:
[in this window]
[in a new window]
 
Table 6. Distribution of conjugative system and repA homologues in Yersinia

++, 100 % identical PCR products to Cos100 sequence; +, PCR products with sequence deviation up to 3·9 % maximal to Cos100 sequence; h, detection by hybridization.

 
We also tested the plasmids for the presence of sequences related to the triK probe and repA probe by hybridization using probes derived from Cos100 template. In some cases also PCR products could be obtained (Table 6). All plasmids carrying the tri region indeed contained sequences closely related to the triK and repA probes, suggesting that all six plasmids (including p29930) possess the same combination of these genetic elements.

Conclusion
The T4SS which was found after shotgun cloning of genomic DNA of the enterocoliticin-producing strain Y. enterocolitica 29930 is encoded on a cryptic plasmid in the original strain. In contrast to the intensively studied type III secretion system of the Yop virulon (Carniel, 2002; Cornelis, 2002), which is specified by a plasmid present in all pathogenic Yersinia strains, the function of the Yersinia T4SS seems to be the translocation of DNA–protein complexes across bacterial cell walls as part of a conjugation process. This means that the T4SS probably does not contribute to pathogenicity. However, in view of the controversial opinion about the potential pathogenicity of Y. enterocolitica biotype 1A strains (Bottone, 1997), it may be worth investigating if the Yersinia T4SS is also able to transport proteins through the bacterial cell wall. It was shown for the prototypical T4SS of A. tumefaciens that besides the nucleoprotein complex, the VirD2-T complex, also the virulence proteins VirE2 and VirF are transported by the system (Schrammeijer et al., 2003).

The genes of the conjugative system of Cos100 span a region of 17 kb, which is longer than the transfer region (14·8 kb) of the IncX plasmid R6K (Núñez et al., 1997). In the IncP conjugative system of plasmid RP4, however, the genes of the Mpf system and Dtr system are organized in physically separated gene clusters with a size of nearly 20 kb each (Pansegrau & Lanka, 1996). It was proposed to view conjugation as a widespread DNA transport mechanism of bacteria, which links a rolling-circle-type DNA replication system to a membrane export system by a coupling factor (Llosa et al., 2002). A new feature of the Yersinia conjugative system is the likely dual function of the TriK protein, which is a homologue of MobB of the mobilizable plasmid CloDF13 (Núñez & de la Cruz, 2001): TriK exerts Dtr functions and serves at the same time as a coupling factor. The coupling factor TaxB of the IncX system (Núñez et al., 1997) does not have a counterpart in the Yersinia Tri system. Conjugation systems are composed of different building blocks and the Yersinia conjugation system described represents a new combination of a T4SS with a DNA-processing region, which so far has only been found in the mobilizable plasmid CloDF13. The differences in the GC contents of the different transfer regions of Cos100 suggest that the two regions have a different evolutionary origin.

It appears that the conjugative system of the cryptic plasmid p29930 is repressed in the genetic background of the original host – at least under laboratory conditions – while under the same conditions the recombinant construct Cos100 efficiently mediates conjugative transfer in Yersinia and E. coli. However, bearing in mind that horizontal gene transfer, especially conjugative transfer of plasmids, is a major force driving variation in prokaryotes, it seems likely that under other circumstances, possibly in different ecological environments, a genetic exchange based on the described T4SS takes place.


   ACKNOWLEDGEMENTS
 
We thank the students Britta Kraushaar and David Mandel for experimental help and Dr Stefan Hertwig and Dr Petra Dersch for strains.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS AND DISCUSSION
REFERENCES
 
Altschul, S. F., Madden, T. L., Schäffer, A. A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D. J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25, 3389–3402.[Abstract/Free Full Text]

Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A. & Struhl, K. (1987). Preparation and analysis of DNA. In Current Protocols in Molecular Biology, pp. 2.0.1–2.14.8. New York: Wiley.

Baron, C., O'Callaghan, D. & Lanka, E. (2002). Bacterial secrets of secretion: EuroConference on the biology of type IV secretion processes. Mol Microbiol 43, 1359–1365.[CrossRef][Medline]

Bayer, M., Eferl, R., Zellnig, G., Teferle, K., Dijkstra, A., Koraimann, G. & Högenauer, G. (1995). Gene 19 of plasmid R1 is required for both efficient conjugative DNA transfer and bacteriophage R17 infection. J Bacteriol 177, 4279–4288.[Abstract]

Belogurov, A. A., Delver, E. P., Agafonova, O. V., Belogurova, N. G., Lee, L. Y. & Kado, C. I. (2000). Antirestriction protein Ard (type C) encoded by IncW plasmid pSa has a high similarity to the "protein transport" domain of TraC1 primase of promiscuous plasmid RP4. J Mol Biol 296, 969–977.[CrossRef][Medline]

Bercovier, H. & Mollaret, H. H. (1984). Yersinia. In Bergey's Manual of Systematic Bacteriology, pp. 498–506. Edited by N. R. Krieg and others. Baltimore: Williams & Wilkins.

Betteridge, T. J., Yang, J., Pittard, A. J. & Praszkier, J. (2003). Interaction of the initiator protein of an IncB plasmid with its origin of DNA replication. J Bacteriol 185, 2210–2218.[Abstract/Free Full Text]

Bignell, C. & Thomas, C. M. (2001). The bacterial ParA-ParB partitioning proteins. J Biotechnol 91, 1–34.[CrossRef][Medline]

Birnboim, H. C. & Doly, J. (1979). A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res 7, 1513–1523.[Abstract]

Bottone, E. J. (1997). Yersinia enterocolitica: the charisma continues. Clin Microbiol Rev 10, 257–276.[Abstract]

Burnens, A. P., Frey, A. & Nicolet, J. (1996). Association between clinical presentation, biogroups and virulence attributes of Yersinia enterocolitica strains in human diarrhoeal disease. Epidemiol Infect 116, 27–34.[Medline]

Burns, D. L. (1999). Biochemistry of type IV secretion. Curr Opin Microbiol 2, 25–29.[CrossRef][Medline]

Cao, T. B. & Saier, M. H., Jr (2001). Conjugal type IV macromolecular transfer systems of Gram-negative bacteria: organismal distribution, structural constraints and evolutionary conclusions. Microbiology 147, 3201–3214.[Free Full Text]

Carniel, E. (2002). Plasmids and pathogenicity islands of Yersinia. Curr Top Microbiol Immunol 264, 89–108.[Medline]

Chang, A. C. & Cohen, S. N. (1978). Construction and characterization of amplifiable multicopy DNA cloning vehicles derived from the P15A cryptic miniplasmid. J Bacteriol 134, 1141–1156.[Medline]

Christie, P. J. (2001). Type IV secretion: intercellular transfer of macromolecules by systems ancestrally related to conjugation machines. Mol Microbiol 40, 294–305.[CrossRef][Medline]

Christie, P. J. & Vogel, J. P. (2000). Bacterial type IV secretion: conjugation systems adapted to deliver effector molecules to host cells. Trends Microbiol 8, 354–360.[CrossRef][Medline]

Cornelis, G. R. (2002). Yersinia type III secretion: send in the effectors. J Cell Biol 158, 401–408.[Abstract/Free Full Text]

Covarrubias, L., Cervantes, L., Covarrubias, A., Soberon, X., Vichido, I., Blanco, A., Kupersztoch-Portnoy, Y. M. & Bolivar, F. (1981). Construction and characterization of new cloning vehicles. V. Mobilization and coding properties of pBR322 and several deletion derivatives including pBR327 and pBR328. Gene 13, 25–35.[CrossRef][Medline]

de Lorenzo, V., Herrero, M., Jakubzik, U. & Timmis, K. N. (1990). Mini-Tn5 transposon derivatives for insertion mutagenesis, promoter probing, and chromosomal insertion of cloned DNA in gram-negative eubacteria. J Bacteriol 172, 6568–6572.[Medline]

Dong, X. Q., Lindler, L. E. & Chu, M. C. (2000). Complete DNA sequence and analysis of an emerging cryptic plasmid isolated from Yersinia pestis. Plasmid 43, 144–148.[CrossRef][Medline]

Galli, D. M., Chen, J., Novak, K. F. & Leblanc, D. J. (2001). Nucleotide sequence and analysis of conjugative plasmid pVT745. J Bacteriol 183, 1585–1594.[Abstract/Free Full Text]

Goelz, G., Knabner, D., Appel, B. & Strauch, E. (2003). A conjugal type IV transfer system in Yersinia enterocolitica strains. Adv Exp Med Biol 529, 51–53.[Medline]

Gomis-Rüth, F. X., Moncalian, G., de la Cruz, F. & Coll, M. (2002). Conjugative plasmid protein TrwB, an integral membrane type IV secretion system coupling protein. Detailed structural features and mapping of the active site cleft. J Biol Chem 277, 7556–7566.[Abstract/Free Full Text]

Grant, T., Bennett-Wood, V. & Robins-Browne, R. M. (1999). Characterization of the interaction between Yersinia enterocolitica biotype 1A and phagocytes and epithelial cells in vitro. Infect Immun 67, 4367–4375.[Abstract/Free Full Text]

Hapfelmeier, S., Domke, N., Zambryski, P. C. & Baron, C. (2000). VirB6 is required for stabilization of VirB5 and VirB3 and formation of VirB7 homodimers in Agrobacterium tumefaciens. J Bacteriol 182, 4505–4511.[Abstract/Free Full Text]

Hayes, F. (2000). The partition system of multidrug resistance plasmid TP228 includes a novel protein that epitomizes an evolutionarily distinct subgroup of the ParA superfamily. Mol Microbiol 37, 528–541.[CrossRef][Medline]

Herrero, M., de Lorenzo, V. & Timmis, K. N. (1990). Transposon vectors containing non-antibiotic resistance selection markers for cloning and stable chromosomal insertion of foreign genes in gram-negative bacteria. J Bacteriol 172, 6557–6567.[Medline]

Hoffmann, B., Strauch, E., Gewinner, C., Nattermann, H. & Appel, B. (1998). Characterization of plasmid regions of foodborne Yersinia enterocolitica biogroup 1A strains hybridizing to the Yersinia enterocolitica virulence plasmid. Syst Appl Microbiol 21, 201–211.[Medline]

Hueck, C. J. (1998). Type III protein secretion systems in bacterial pathogens of animals and plants. Microbiol Mol Biol Rev 62, 379–433.[Abstract/Free Full Text]

Kado, C. I. & Liu, S. T. (1981). Rapid procedure for detection and isolation of large and small plasmids. J Bacteriol 145, 1365–1373.[Medline]

Kalkum, M., Eisenbrandt, R., Lurz, R. & Lanka, E. (2002). Tying rings for sex. Trends Microbiol 10, 382–387.[CrossRef][Medline]

Krause, S., Bárcena, M., Pansegrau, W., Lurz, R., Carazo, J. M. & Lanka, E. (2000). Sequence-related protein export NTPases encoded by the conjugative transfer region of RP4 and by the cag pathogenicity island of Helicobacter pylori share similar hexameric ring structures. Proc Natl Acad Sci U S A 97, 3067–3072.[Abstract/Free Full Text]

Lai, E. M. & Kado, C. I. (2000). The T-pilus of Agrobacterium tumefaciens. Trends Microbiol 8, 361–369.[CrossRef][Medline]

Lewin, A., Strauch, E., Hertwig, S., Hoffmann, B., Nattermann, H. & Appel, B. (1996). Comparison of plasmids of strains of Yersinia enterocolitica biovar 1A with the virulence plasmid of a pathogenic Y. enterocolitica strain. Zentbl Bakteriol 285, 52–63.

Li, Z., Hiasa, H., Kumar, U. & DiGate, R. J. (1997). The traE gene of plasmid RP4 encodes a homologue of Escherichia coli DNA topoisomerase III. J Biol Chem 272, 19582–19587.[Abstract/Free Full Text]

Llosa, M., Gomis-Rüth, F. X., Coll, M. & de la Cruz, F. (2002). Bacterial conjugation: a two-step mechanism for DNA transport. Mol Microbiol 45, 1–8.[CrossRef][Medline]

McGhee, G. C. & Jones, A. L. (2000). Complete nucleotide sequence of ubiquitous plasmid pEA29 from Erwinia amylovora strain Ea88: gene organization and intraspecies variation. Appl Environ Microbiol 66, 4897–4907.[Abstract/Free Full Text]

Núñez, B. & de la Cruz, F. (2001). Two atypical mobilization proteins are involved in plasmid CloDF13 relaxation. Mol Microbiol 39, 1088–1099.[CrossRef][Medline]

Núñez, B., Avila, P. & de la Cruz, F. (1997). Genes involved in conjugative DNA processing of plasmid R6K. Mol Microbiol 24, 1157–1168.[Medline]

Paetzel, M., Dalbey, R. E. & Strynadka, N. C. (2000). The structure and mechanism of bacterial type I signal peptidases. A novel antibiotic target. Pharmacol Ther 87, 27–49.[CrossRef][Medline]

Page, A. L. & Parsot, C. (2002). Chaperones of the type III secretion pathway: jacks of all trades. Mol Microbiol 46, 1–11.[CrossRef][Medline]

Pansegrau, W. & Lanka, E. (1996). Enzymology of DNA transfer by conjugative mechanisms. Prog Nucleic Acid Res Mol Biol 54, 197–251.[Medline]

Pohlman, R. F., Liu, F., Wang, L., More, M. I. & Winans, S. C. (1993). Genetic and biochemical analysis of an endonuclease encoded by the IncN plasmid pKM101. Nucleic Acids Res 21, 4867–4872.[Abstract]

Praszkier, J., Wei, T., Siemering, K. & Pittard, J. (1991). Comparative analysis of the replication regions of IncB, IncK, and IncZ plasmids. J Bacteriol 173, 2393–2397.[Medline]

Rabel, C., Grahn, A. M., Lurz, R. & Lanka, E. (2003). The VirB4 family of proposed traffic nucleoside triphosphatases: common motifs in plasmid RP4 TrbE are essential for conjugation and phage adsorption. J Bacteriol 185, 1045–1058.[Abstract/Free Full Text]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Schneiker, S., Keller, M., Dröge, M., Lanka, E., Pühler, A. & Selbitschka, W. (2001). The genetic organization and evolution of the broad host range mercury resistance plasmid pSB102 isolated from a microbial population residing in the rhizosphere of alfalfa. Nucleic Acids Res 29, 5169–5181.[Abstract/Free Full Text]

Schrammeijer, B., den Dulk-Ras, A., Vergunst, A. C., Jurado, J. E. & Hooykaas, P. J. J. (2003). Analysis of Vir protein translocation from Agrobacterium tumefaciens using Saccharomyces cerevisiae as a model: evidence for transport of a novel effector protein VirE3. Nucleic Acids Res 31, 860–868.[Abstract/Free Full Text]

Sherrat, D. (1989). Tn3 and related transposable elements: site-specific recombination and transposition. In Mobile DNA, pp. 163–184. Edited by D. E. Berg & M. M. Howe. Washington, DC: American Society for Microbiology.

Simon, R., Priefer, U. & Pühler, A. (1983). A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in Gram-negative bacteria. Biotechnology 1, 784–791.

Snellings, N. J., Popek, M. & Lindler, L. E. (2001). Complete DNA sequence of Yersinia enterocolitica serotype O : 8 low-calcium-response plasmid reveals a new virulence plasmid-associated replicon. Infect Immun 69, 4627–4638.[Abstract/Free Full Text]

Strauch, E., Voigt, I., Broll, H. & Appel, B. (2000a). Use of a plasmid of a Yersinia enterocolitica biogroup 1A strain for the construction of cloning vectors. J Biotechnol 79, 63–72.[CrossRef][Medline]

Strauch, E., Hoffmann, B., Heins, G. & Appel, B. (2000b). Isolation of a new insertion element of Yersinia intermedia closely related to remnants of mobile genetic elements present on Yersinia plasmids harboring the Yop virulon. FEMS Microbiol Lett 193, 37–44.[CrossRef][Medline]

Strauch, E., Kaspar, H., Schaudinn, C., Dersch, P., Madela, K., Gewinner, C., Hertwig, S., Wecke, J. & Appel, B. (2001). Characterization of enterocoliticin, a phage tail-like bacteriocin, and its effect on pathogenic Yersinia enterocolitica strains. Appl Environ Microbiol 67, 5634–5642.[Abstract/Free Full Text]

Willetts, N. & Wilkins, B. (1984). Processing of plasmid DNA during conjugation. Microbiol Rev 48, 24–41.[Medline]

Wu, C. Y., Fu, J. F. & Liu, S. T. (2001). The replicon of pSW800 from Pantoea stewartii. Microbiology 147, 2757–2767.[Abstract/Free Full Text]

Yarmolinsky, M. (2000). Transcriptional silencing in bacteria. Curr Opin Microbiol 3, 138–143.[CrossRef][Medline]

Yeo, H. J., Savvides, S. N., Herr, A. B., Lanka, E. & Waksman, G. (2000). Crystal structure of the hexameric traffic ATPase of the Helicobacter pylori type IV secretion system. Mol Cell 6, 1461–1472.[Medline]

Young, G. M., Schmiel, D. H. & Miller, V. L. (1999). A new pathway for the secretion of virulence factors by bacteria: the flagellar export apparatus functions as a protein-secretion system. Proc Natl Acad Sci U S A 96, 6456–6461.[Abstract/Free Full Text]

Received 11 April 2003; revised 16 June 2003; accepted 25 June 2003.