Previous Article | Next Article ![]()
Journal of Bacteriology, February 2008, p. 991-1010, Vol. 190, No. 3
0021-9193/08/$08.00+0 doi:10.1128/JB.01467-07
Copyright © 2008, American Society for Microbiology. All Rights Reserved.
,
Bundesinstitut für Risikobewertung, D-12277 Berlin, Germany,1 Robert Koch-Institut, D-13353 Berlin, Germany,2 Max-Planck-Institut für Molekulare Genetik, D-14195 Berlin, Germany3
Received 11 September 2007/ Accepted 20 November 2007
|
|
|---|
|
|
|---|
Here, we present the complete sequence analysis of plasmid pYE854 and data on the conjugative properties of the plasmid, including its capability to mobilize the Yersinia virulence plasmid pYV. The study was undertaken to assess the potential for pYV dissemination in Yersinia populations. Y. enterocolitica is a heterogenous species consisting of approximately 70 serogroups, of which only some (O:3, O:8, O:9, O:5,27) are known to be pathogenic in humans (4). Reports describing pYV in environmental Yersinia strains are yet not available, but some strains harbor plasmids displaying homologies to pYV (33). We therefore wanted to find out whether gene exchange by conjugation might be common between Yersinia strains.
|
|
|---|
(51) and Genehogs (Invitrogen) were used as hosts for cloning procedures. All strains were grown in Luria-Bertani (LB) broth. Apart from Yersinia strains, which were cultivated at 28°C, all strains were incubated at 37°C. Solid agar media contained 1.8% (wt/vol) agar. When required, ampicillin and kanamycin were supplemented at 100 µg ml–1 and chloramphenicol and tetracycline at 12.5 µg ml–1. |
View this table: [in a new window] |
TABLE 1. Bacterial strains and plasmids
|
In vivo mutagenesis of the conjugative plasmids.
Plasmids containing either a kanamycin (Kmr) or a chloramphenicol acetyltransferase (Cmr) resistance gene were created by in vivo transposon mutagenesis of the host strains using the mini-Tn5 transposon derivative pUT described by de Lorenzo et al. (13) and Herrero et al. (28). Y. enterocolitica strains harboring plasmid pYE854 or pYE966 were used as recipients for filter mating experiments with E. coli strain S17-1
pir containing pUTKm or mini-Tn5 Cm (Table 1). Transconjugants were selected on CIN agar (45) supplemented with the appropriate antibiotic. To detect plasmids in which the resistance gene had been inserted, plasmid DNA was isolated from the mutants and subjected to restriction analysis and hybridization with the transposon delivery vectors. The insertion sites of the resistance genes were determined by cloning DraI restriction fragments of the pYE854 mutants with the help of the vector pLitmus38 (Apr; New England Biolabs). Sequencing was performed using primers deduced from the marker genes.
Nucleotide sequence determination and analysis of plasmid pYE854. For shotgun sequencing of pYE854, a fragment library was constructed. Small (1.8-to 2.2-kb) DNA fragments of the plasmid were generated by ultrasonic treatment. After end repair with T4 polymerase (Roche), 10 µg DNA was loaded onto an agarose gel (0.9%). DNA fragments of the appropriate size were excised from the gel. The extracted DNA was cloned into the vector pUC19 cleaved with SmaI (Roche).
DNA sequencing reactions were set up using an Applied Biosystems BigDye Terminator version 3.1 cycle sequencing kit. We have sequenced the shotgun clones from both sides up to a sixfold genome coverage using an ABI 3730 XL sequencer. The data assembly was accomplished by using Staden Package software version 4.6 (Roger Staden, Cambridge, United Kingdom). Gap closure was performed by combinatorial PCR and sequencing of the generated PCR fragments. Sequence analyses were performed by using the MacVector 8.0 software of the Oxford Molecular Group. Blast searches were carried out with the NCBI database, and the GAP program of the GCG package (14) was used for calculating similarity and identity values.
To determine putative promoters, the 300-bp upstream regions of pYE854 open reading frames (ORFs) were analyzed for the existence of –35 and –10 consensus sequences (TTGACA-N15-20-TATAAT) described by Hawley and McClure (27) and Harley and Reynolds (26). To identify the origin of transfer (oriT) of pYE854, the complete nucleotide sequence was analyzed for the existence of consensus sequences for the oriT nic region families described by Zechner et al. (52).
Construction of replicative miniplasmid derivatives of pYE854. For all DNA manipulations, standard techniques were applied (44). To identify regions essential for plasmid replication, miniplasmid derivatives of pYE854 were created by cleavage of the plasmid DNA with various restriction endonucleases. Restriction fragments bearing overhanging ends were trimmed with T4 polymerase (New England Biolabs). The tetracycline resistance gene (tetA) of Tn5 was amplified by PCR using plasmid pBR327 as template and was ligated to the restriction fragments of pYE854.
Cloning of restriction fragments and of the oriT region of plasmid pYE854. Covalently closed circular plasmid DNA purified by cesium chloride gradient centrifugation (44) was digested with the restriction endonucleases ClaI and HindIII. The fragments were inserted into the corresponding sites of pIV2 (Kmr) (49). The constructs were introduced into the E. coli strain Genehogs. Recombinant plasmids were characterized by restriction analysis and sequencing. Plasmids containing ClaI restriction fragments resulted in the pJH800 series, and those harboring HindIII restriction fragments gave the pJH900 series (see Fig. 4).
![]() View larger version (26K): [in a new window] |
FIG. 4. Isolation of the pYE854 oriT region. The upper panel shows the linear gene map of pYE854 and a restriction fragment library (pJH800 series ClaI fragments and pJH900 series HindIII fragments) used to identify the oriT sequences present on pJH808 and pJH907. In the middle, the 4.2-kb HindIII restriction fragment of pJH907, which was further reduced in size, is depicted. The smallest DNA fragment conferring mobilization has a size of 309 bp (pJH907-F). +, transfer occurred; –, transfer did not occur. The lower panel illustrates an alignment of the oriT sequences of pYE854 and some other, related plasmids. A shaded background marks positions where two different types of nucleotides may occur. Nucleotides conserved within the oriT nic regions are depicted in bold letters. In cases where the cleavage site has been determined, it is indicated by an arrowhead. On the right, the sequence of the flanking palindrome, which is also essential for conjugation, is shown.
|
transformants containing one of the resulting plasmids, pJH907-C to -G, and either pYE854 or pYE966 were investigated in terms of mobilization of the constructs by the conjugative plasmids (see below). Nucleotide sequence accession number. The complete sequence of plasmid pYE854 has been submitted to the EMBL databank under the accession number AM905950.
|
|
|---|
|
View this table: [in a new window] |
TABLE 2. Characterization of pYE854 mutants
|
|
View this table: [in a new window] |
TABLE 3. Results of the mating experiments with various recipient strains
|
|
View this table: [in a new window] |
TABLE 4. Efficiencies of pYE854 and pYE966 conjugation and pYV mobilization
|
![]() View larger version (84K): [in a new window] |
FIG. 1. Plasmid analysis of donor strains, recipients, and transconjugants. All donor and recipient strains contained the plasmids pYE854/pYV and pBR327, respectively. (A) An 0.8% agarose gel showing DraI restriction patterns of the plasmids. Lane M, marker ( Eco130I); lane 1, pYE854; lane 2, pYV (p1340); lane 3, Y. pseudotuberculosis 1340 (donor); lane 4, E. coli DH5 (donor); lane 5, Y. enterocolitica 83/88 (recipient); lane 6, Y. enterocolitica 83/88 (transconjugant, donor strain was Y. pseudotuberculosis 1340); lane 7, Y. enterocolitica 83/88 (transconjugant, donor strain was E. coli DH5 ), lane 8, Y. enterocolitica 31080 (recipient); lane 9, Y. enterocolitica 31080 (transconjugant, donor strain was Y. pseudotuberculosis 1340); lane 10, Y. enterocolitica 31080 (transconjugant, donor strain was E. coli DH5 ); lane 11, Y. enterocolitica 12 (recipient); lane 12, Y. enterocolitica 12 (transconjugant, donor strain was Y. pseudotuberculosis 1340). (B) Southern hybridization of the plasmids shown in panel A to pYE854. Transconjugants obtained by using E. coli as donor were selected on agar containing chloramphenicol (pYV) and tetracycline (pBR327) and did not harbor pYE854 (lanes 7 and 10). (C) Southern hybridization of the plasmids shown in panel A to pYV (p1340). Arrows indicate restriction fragments that are different in the virulence plasmids of the recipients and the transconjugants.
|
Overall sequence analysis of pYE854. The complete nucleotide sequence (95,499 bp) of pYE854 has been determined. The plasmid has an average G+C content of 42.3%, slightly lower than the 48.5% ± 1.5% (mean ± standard deviation) reported for the host Y. enterocolitica (5). However, the G+C contents of different ORFs vary in the range from 28% to 60%. The bioinformatic analysis revealed 232 ORFs with potential coding capacity (minimum protein size, 50 amino acids) and an ATG start codon (see Table S2 in the supplemental material for details). Both DNA strands of the plasmid contain an equal number of ORFs. The database search gave matches for 80 deduced gene products with identities to other protein sequences of between 27% and 100% (Table 5). We did not detect genes obviously encoding known virulence factors or genes conferring antibiotic resistance. The strongest overall similarities found were to transfer proteins of Yersinia bercovieri and the F plasmid, but significant homologies also exist to proteins belonging to a phosphotransferase system (PTS) of Vibrio vulnificus.
|
View this table: [in a new window] |
TABLE 5. Selected ORFs of plasmid pYE854
|
![]() View larger version (29K): [in a new window] |
FIG. 2. Genetic map of pYE854. Black bars indicate DNA regions involved in plasmid transmission, replication, and maintenance, as well as genes that might belong to a PTS. The position of the oriT is marked by an arrowhead. DNA segments that have been identified and those that are missing in pYE966 are indicated by dashed lines and dotted lines, respectively.
|
![]() View larger version (13K): [in a new window] |
FIG. 3. Helicase motifs identified within TraI of pYE854 (according to Hall and Matson [25]). *, for the consensus sequence Ia, several possible motifs have been identified (data not shown). The position of the first amino acid residue of each motif of pYE854 is given. aa, amino acids. Amino acid residues identical in motifs of TraI of pYE854 and other helicases are indicated by bold letters. + and O represent a hydrophobic and a hydrophilic residue, respectively. Residues not restricted to being hydrophobic and hydrophilic are represented by an X.
|
The Tra2 region is approximately 20 kb in length and contains 17 ORFs whose products show homologies to known proteins (Fig. 2). Eleven of the deduced amino acid sequences are similar to transfer proteins. The majority of the transfer proteins are highly homologous to proteins of Y. bercovieri strain ATCC 43970, with identities between 94% and 100% (Table 5). Unfortunately, there is yet no information available about plasmids of this species. Nevertheless, a stretch of 33 kb (pYE854 positions 51500 to 84500) is nearly identical to a published whole-genome shotgun sequence (Y berA_01_29, NCBI accession no. NZ_AALC01000029) of the Y. bercovieri strain. This stretch comprises the whole Tra2 region and adjacent DNA sequences of pYE854 (see below). The ORF analysis suggests that, in the Tra2 region, mating-pair formation (Mpf) proteins are encoded. Mpf proteins form a membrane-associated apparatus that synthesizes and assembles mature conjugative pili on the cell surface. The conjugative pili facilitate the initial contact with the recipient cell, resulting in the formation of a stable mating pair. The pYE854 products are related to Mpf proteins of several transfer systems, e.g., F and R100 (IncF), R27 (IncH), RP4 (IncP), and R388 (IncW), and of the type IV secretion system of the Ti plasmid of Agrobacterium tumefaciens.
To determine which of the pYE854 tra genes are essential for conjugation, the plasmid was mutagenized by using an in vivo transposon mutagenesis system (see Materials and Methods). The analysis of the insertion sites of the marker gene disclosed that many insertions occurred hot spot-like within transposase genes (data not shown). However, insertions were also found in other ORFs (Table 2). All mutants were investigated in terms of their capability to conjugate and to mobilize pYV. Not surprisingly, insertions leading to a loss of conjugation and mobilization were mainly found in ORFs located in the Tra regions. All seven insertions within the putative traI gene (ORF 36) resulted in a deficit of transfer. Within the Tra2 region, several genes supposed to be important for conjugation were mutagenized (Table 2). Insertions in ORF 164 (TraN) only decreased mating ability slightly, consistent with the results of observations of plasmids F and R100-1 (37). All other insertions yielded defective mutants unable to conjugate and to mobilize the Yersinia virulence plasmid. In addition to tra, insertions resulting in a negative Tra phenotype were also observed in ORF 122. The function of this gene is not known. Its predicted product shows some similarity to a permease of Yersinia belonging to the major facilitator superfamily which comprises translocators exporting antibiotics and other small molecules.
The pYE854 replicon belongs to a new incompatibility group.
A search for ORFs encoding probable Rep proteins was undertaken to identify DNA regions on pYE854 which might be important for plasmid replication. Two genes (ORF 11 and ORF 187) possibly involved in replication have been detected. While the deduced ORF 11 product shows some similarity to a primase of Clostridium difficile, ORF 187, located upstream of the Tra2 region, encodes a protein that is similar (30% identity) to the Rep protein of the Clostridium perfringens plasmid pIP404 (Table 5). The rep gene of pIP404 is sufficient, in conjunction with some repetitive origin-like sequences located downstream of the gene, to drive plasmid replication (21, 22). We constructed a replicative miniplasmid derivative of pYE854 by digesting the plasmid with PstI and ligating the resulting restriction fragments with a tetracycline resistance gene. Upon transformation of E. coli DH5
, plasmid pJH1010, comprising the 50,189-bp PstI restriction fragment and tet (Fig. 5), was isolated. By using other restriction endonucleases, the size of the minimal replicon could be further reduced. The smallest miniplasmid, pJH1015, is composed of the 8,350-bp NdeI restriction fragment of pYE854 ligated to the tetracycline resistance gene. Besides ORF 187, pJH1015 contains the pYE854 ORFs 183, 185, 191, 194, 196, and 202 (Fig. 5). The relationship of the putative ORF 187 product to a replication protein of Clostridium prompted us to investigate whether pJH1015 may replicate in gram-positive bacteria. For that reason, several Bacillus strains belonging to various species (see Table S1 in the supplemental material) were transformed with the plasmid. We could not isolate any transformant containing pJH1015, while plasmid pUB110 (36) could be readily introduced into the bacteria, indicating that the pYE854 derivative does not replicate in Bacillus species.
![]() View larger version (37K): [in a new window] |
FIG. 5. Analysis of the pYE854 replication and partitioning (par) region. The upper panel shows miniplasmid derivatives of pYE854 retaining replicative competence. In the middle, the linear gene map of pYE854 is given. The locations of putative centromere sites are indicated. The lower panel shows the sequences of the predicted centromere sites and their positions on the plasmid. Lowercase letters show nucleotides diverging from the dyad symmetry of the palindromes. The vertical arrow and dotted line represent the center of dyad symmetry. Half sides of the palindrome are indicated by horizontal arrows.
|
Plasmid maintenance is accomplished by a partitioning and an addiction system. Next to the putative rep gene of pYE854, there are two ORFs (191 and 194) which apparently belong to a plasmid partitioning system (Fig. 2). Their predicted products are nearly identical to proteins of Y. bercovieri and also related to partitioning proteins (Par) of IncP plasmids (Table 5). While the ORF 191 product is similar to ParA-like ATPases, the deduced product of ORF 194 may represent the transcriptional repressor ParB. This protein binds to a centromere target site (parC) which consists of repetitive sequences often situated close to the par operon. A search for possible centromere sites revealed 14 nearly identical copies of a 16-bp palindromic sequence scattered on pYE854 (Fig. 5). Three copies of this sequence are located upstream of the parA gene. This region also contains a number of possible promoter sequences, which suggests that the detected repeats are binding sites for the partitioning proteins and probably involved in the autoregulation of par expression.
Stable maintenance of pYE854 might also be achieved by an addiction system (toxin-antitoxin system) responsible for postsegregational elimination of plasmidless cells from a bacterial population. ORFs 15 and 16 apparently encode a toxin and an antitoxin, respectively. Both products are similar to proteins of the bacteriophage N15 that were shown to stabilize heterologous replicons (15). Moreover, overexpression of the N15 toxin gene in E. coli resulted in a bacteriostatic effect discernible by elongation of bacterial cells and growth cessation. To find out whether the pYE854 ORF 15 also has adverse effects on bacterial growth, the gene was amplified by PCR and inserted into expression vector pMS470
8, which contains an inducible tac promoter (3). Upon transformation, the isolated E. coli clones were analyzed by sequencing. All plasmids revealed mutations within ORF 15 leading to amino acid exchanges which probably inactivated the toxin. As the cloned DNA fragment is very small (333 bp), it can be assumed that the intact gene has deleterious effects on the bacteria. This was corroborated by the fact that we could clone PCR fragments containing either the antitoxin gene or both the toxin and antitoxin genes.
Plasmid curing was examined by treating Y. enterocolitica strain 29854 with acridine orange at a 0.1 mM concentration. Using this concentration, the plasmid could easily be cured from its original host strain (data not shown).
pYE854 is a cryptic plasmid. As mentioned above, we identified a stretch of genes (ORFs 73, 75, 79, 82, 86, and 89) on pYE854 which obviously code for components belonging to a PTS (Fig. 2, Table 5). A similar cluster exists in V. vulnificus YJ016 and also in Psychromonas and Y. pseudotuberculosis. The products of this gene cluster might be important for the uptake of mannose and fructose. By using an API 50 CHB/E system (bioMerieux, Nürtingen, Germany), we tested if a Yersinia strain harboring pYE854 is more capable of utilizing sugars and other carbohydrates than the respective strain without the plasmid. We did not detect any differences between the strains. Thus, it is doubtful whether the gene cluster plays a role in bacterial growth. The same holds true for ORFs 97 and 219, whose predicted products show a relationship to proteins involved in heavy metal resistance (Table 5). A test on agar containing arsenic, cadmium, cobalt, or zinc revealed the sensitivity of the Yersinia strains to these elements, regardless of the plasmid content. Therefore, at the present level of knowledge, pYE854 has to be considered cryptic.
pYE966, a pYE854-related plasmid of a pathogenic Y. enterocolitica strain. Mating studies conducted with pYE854 showed that this plasmid is transmissible with high frequencies. To find out whether similar plasmids are prevalent in Yersinia, we used pYE854 as the probe for DNA hybridization experiments with plasmid preparations of approximately 50 strains of the BfR strain collection. One of the plasmids investigated, pYE966, isolated from pathogenic Y. enterocolitica serogroup O:5,27 strain 966/89 (Table 1) displayed strong hybridization signals (Fig. 6). According to the restriction patterns obtained, pYE966 has an approximate size of 70 kb. Two DNA regions of pYE854 are obviously missing in the smaller plasmid (Fig. 2). Using primers deduced from pYE854, genes of plasmid pYE966 coding for a number of transfer proteins, the putative replication protein (ORF 187 product), and the partitioning and addiction systems were amplified and sequenced (data not shown). In addition, the oriT region of pYE966 was analyzed. We detected only single mismatches to the pYE854 sequences that did not result in amino acid exchanges, confirming that the plasmids are closely related. The conjugative properties of pYE966 have been studied and compared to those of pYE854. Like its larger relative, pYE966 is a self-transmissible plasmid that is able to mobilize pYV, with slightly higher efficiencies even than pYE854 (Table 4). These findings document that the DNA regions of pYE854 which are missing in pYE966 are apparently not important for plasmid transfer.
![]() View larger version (46K): [in a new window] |
FIG. 6. Southern hybridization of pYE854 with pYE966. Several restriction digests of pYE854 were separated on an 0.8% agarose gel and hybridized to labeled pYE966 DNA. Lane M, marker ( Eco130I); lanes 1, 3, and 5, pYE854; lanes 2, 4, and 6, pYE966.
|
|
|
|---|
The sequence analysis of pYE854 and pYE966 demonstrated that these plasmids are very similar. They show a relationship to the F factor, especially in the transfer regions, which, unlike the tra operon of F, are separated on the Yersinia plasmids. In F, four proteins (TraD, TraI, TraM, and TraY) are involved in DNA metabolism during transfer (16). The sequence analysis of pYE854 revealed two products (TraD and TraI) that might be part of the relaxosome of this plasmid. Counterparts to TraM and TraY of the F factor have not been identified. Interestingly, the oriT of pYE854, which is also similar to those of the F group, was identified contiguous to the predicted traD gene, while the oriT of the F factor lies adjacent to the traM gene, approximately 23 kb away from traD (50). We have investigated whether F is able to mobilize a recombinant plasmid containing the oriT of pYE854 but could not detect any transmission of this plasmid (data not shown). Hence, although some proteins of the pYE854 relaxosome and the oriT sequence are similar to F, other transfer proteins and/or binding sites (e.g., for IHF, TraM, or TraY) are apparently too divergent to allow mobilization. We also searched for possible oriT sequences on the Yersinia virulence plasmid, especially within the lcr region. Several families of oriT nic regions and the oriT of pYE854 provided the basis for the search (52). Numerous sequences have been detected all over pYV, but none revealed strong similarity to one of the compared sequences. Therefore, the determination of the oriT of pYV requires further experiments, as already performed with pYE854.
Among the deduced pYE854 products belonging to the Mpf system, we found seven proteins showing similarity to pilus assembly proteins (TraB, TraC, TraE, TraF, TraH, TraU, and TrbC) of the F factor (19). Their predicted molecular masses are also in good agreement with those of the corresponding F proteins. The same holds true for two pYE854 products that are related to the mating-aggregate stabilization proteins TraG and TraN of F (17, 41). By contrast, a product similar to the pilus subunit pilin of F is not encoded by pYE854 (18). Taking the data on similarities of the transfer proteins and the oriT together, it can be reasoned that pYE854 and pYE966 are F-like plasmids, though this classification does not refer to the replicons of the Yersinia plasmids.
We could not allocate the plasmids to a known incompatibility group. Their replicons show some relationship to the Clostridium plasmid pIP404 but are obviously not functional in gram-positive bacilli. Stable replication in members of the Enterobacteriaceae apparently requires the partitioning system because we were unable to isolate a replicative miniplasmid deprived of the par operon. Since four identical copies of the predicted centromere sites are situated close to or within the par operon, it can be assumed that they are also of importance for miniplasmid maintenance. The stabilization of the native plasmids might be accomplished by additional centromere-like sites scattered on the plasmids and by an addiction system. Interestingly, the identified toxin-antitoxin system is related to that of the temperate E. coli phage N15, whose prophage is a linear plasmid (43). Besides pYE854, Y. enterocolitica strain 29854 harbors the linear plasmid prophage PY54, which is related to N15 but does not possess an addiction system. This observation raises the question of whether the addiction system of pYE854 was once acquired from PY54.
The isolated conjugative plasmids are large in size and contain numerous ORFs, but at present, they have to be regarded as cryptic. This might be caused by the fact that only 25% of the deduced gene products revealed similarities to known proteins. Apart from tra genes and those implicated in plasmid replication and maintenance, the functions of the ORFs could not be assigned. Although genes potentially encoding a PTS or conferring heavy metal resistance have been identified on pYE854, we were not able to demonstrate the respective functions. Perhaps additional genes essential for functionality are lacking on the plasmid, or we did not test for the right compounds (e.g., heavy metals). However, it has also been reported that PTSs are sometimes associated with bacterial virulence (53). The question of whether pYE854 plays a role in the virulence of Yersinia cannot be answered at the moment. Besides the PTS genes, the plasmid contains two genes (ORF 62 and ORF 73) whose products are similar to a hyaluronate lyase and a glucuronyl hydrolase, respectively, of Vibrio, proteins which are suggested to be virulence factors of some gram-positive pathogens (34, 40). Nevertheless, the finding that Y. enterocolitica 29854 can be cured from its conjugative plasmid without any consequences for growth in broth suggests that pYE854 and, presumably, also pYE966 are not pivotal for the bacteria under laboratory conditions.
This work was supported by a grant of the Deutsche Forschungsgemeinschaft to E.L.
Published ahead of print on 30 November 2007. ![]()
Supplemental material for this article may be found at http://jb.asm.org/. ![]()
|
|
|---|
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»