Correspondence: Masataka Tsuda, Department of Environmental Life Sciences, Graduate School of Life Sciences, Tohoku University, 2-1-1 Katahira, Sendai 980-8577, Japan. Tel./fax: +81 22 217 5699; e-mail: email@example.com
Conjugative plasmid transfer systems have been well studied, but very little is known about the recipient factors that control horizontal transmission. A self-transmissible IncP-9 naphthalene catabolic plasmid, NAH7, carries the traF gene, whose product is considered to be a host-range modifier of NAH7, because its traF deletion mutant (NAH7dF) is transmissible from Pseudomonas putida to P. putida and Escherichia coli and from E. coli to E. coli, but not from E. coli to P. putida. In this study, transposon mutagenesis of P. putida KT2440 was performed to isolate the mutants that could receive NAH7dF from E. coli. The mutants had the transposon insertions in ptsP or ptsO, encoding two of three components of the nitrogen-related phosphotransferase system (PTSNtr). The KT2440 derivative lacking ptsN, encoding the remaining component of PTSNtr, was also able to receive NAH7dF. These results indicated that the PTSNtr in P. putida is involved in inhibition of conjugative transfer of NAH7dF from E. coli. Our further experiments using site-directed mutants suggested the indirect involvement of the phosphorylated form of PtsO in the inhibition of the conjugative transfer. Conjugative transfer of NAH7 and another IncP-9 plasmid, pWW0, from E. coli was partially inhibited by the PtsO function in KT2440.
Horizontal transfer of plasmids by conjugation contributes greatly to the rapid adaptation and evolution of host bacteria, because the plasmids often carry various genetic traits (Springael & Top, 2004). Each self-transmissible plasmid in gram-negative bacteria has two well-studied systems for conjugative transfer: DNA transfer and replication (Dtr) and mating pair formation (Mpf) systems (Lawley & Wilkins, 2004). The former system governs the formation of the relaxase–single-stranded DNA (ssDNA) complex at the origin of transfer (oriT) for initiating the conjugation-specific DNA replication, while the latter system is involved in the formation of the apparatus for the transfer of the relaxase–ssDNA complex into the recipient cell. In contrast, only limited information has accumulated on the recipient factors that control the conjugation. Principally, it is known that (1) efficient conjugative transfer of F plasmid and R64 in Escherichia coli requires specific recipient cell surface components (e.g. intact LPS and an outer membrane protein), probably for the stability of Mpf (Manoil & Rosenbusch, 1982; Ishiwa & Komano, 2004), and (2) the recipient-encoded CRISPR system and restriction endonucleases digest ds-DNA of plasmids (Purdy et al., 2002; Wiedenheft et al., 2012).
An 82-kb naphthalene catabolic plasmid, NAH7, from Pseudomonas putida is self-transmissible and belongs to the IncP-9 incompatibility group (Sota et al., 2006). As is the case in another well-characterized IncP-9 toluene catabolic plasmid, pWW0, NAH7 is transmissible between Pseudomonas and E. coli and is maintained in both hosts (Tsuda & Iino, 1990). NAH7 carries the traDEF operon, which is transcribed from the putative oriT-containing region (Miyazaki et al., 2008). Our previous analysis of the operon revealed that the three gene products, mainly located at the cytoplasm, periplasm, and membrane fractions, respectively, are not essential to the conjugation but are required in the donor cell for efficient conjugation (Miyazaki et al., 2008). We further found that the NAH7 mutant deleting the last gene (traF) in the operon (NAH7dF) is proficient in conjugative transfer from P. putida to P. putida and E. coli and from E. coli to E. coli, but is deficient in transfer from E. coli to P. putida, suggesting that TraF acts as a host-range modifier in the conjugative transfer of NAH7. This phenotype associated with NAH7dF suggested the presence of P. putida-encoded factor(s) that restrict the efficient conjugative transfer of NAH7dF from E. coli.
In this study, our transposon mutagenesis of P. putida KT2440 led to the successful isolation of mutants able to receive NAH7dF from E. coli. The mutants had transposon insertions in genes encoding the members of the nitrogen-related phosphotransferase system (PTSNtr) (Fig. 1a) (Pfluger-Grau & Gorke, 2010). Further conjugation experiments revealed that PtsO, a component of PTSNtr, regulates the conjugative transfer of NAH7dF, its wild type, and pWW0 from E. coli.
Materials and methods
Strains, plasmids, and growth conditions
The bacterial strains and plasmids used in this study are listed in Table 1 and Table S1 in Supporting Information. Escherichia coli cells were grown at 37 °C in Luria–Bertani (LB) broth (Sambrook et al., 1989), and P. putida cells, at 30 °C in 1/3LB (0.33% tryptone, 0.16% yeast extract, 0.5% NaCl) broth. When the E. coli cells carried IncP-9 plasmids, the cells were cultivated at 30 °C to avoid the unstable maintenance of these plasmids in E. coli at 37 °C (Sota et al., 2006). The solid media were prepared by the addition of 1.5% agar. Antibiotics were used at final concentrations of 100 μg mL−1 for ampicillin, 30 μg mL−1 for chloramphenicol (Cm), 25 μg mL−1 for kanamycin (Km), 30 μg mL−1 for gentamicin (Gm), and 20 μg mL−1 for tetracycline (Tc).
Table 1. Bacterial strains and plasmids used in this study
DNA and RNA manipulation and construction of plasmids
Standard methods were used for extraction of plasmid DNA, DNA digestion with restriction endonucleases, DNA ligation, and transformation of E. coli cells (Sambrook et al., 1989). Electrocompetent cells of P. putida were prepared according to the 10-min method, and electroporation was carried out as described previously (Choi et al., 2006). Escherichia coli DH5α or JM109 was used for the construction of plasmids. PCR was carried out using KOD-plus DNA polymerase (Toyobo) or Ex Taq polymerase (Takara). Sequence determination was performed with an ABI Prism model 3130xl sequencer (Applied Biosystems). The primers used in this study are listed in Table S2 in Supporting Information. Total RNA from P. putida cells in mid-log-phase liquid culture was prepared using RNeasy Mini Kit (Qiagen). Reverse transcription was performed using ReverTra Ace Kit (Toyobo), and the resulting cDNA sample was used to carry out quantitative reverse transcription-polymerase chain reaction (qRT-PCR) analysis using an Opticon 2 System (Bio-Rad). Quantification was performed using SYBR Premix Ex Taq and the primer set of ptsO-F and ptsO-R or 16S-F and 16S-R in a reaction volume of 50 μL. 16S rRNA was used as an internal standard.
Donor and recipient cells separately grown overnight were harvested by centrifugation, washed with 1/3LB broth, and resuspended in fresh 1/3LB broth. They were mixed at a ratio of 1 : 1 and subsequently spotted on a 0.45-μm membrane filter (Advantec) that had been placed on a 1/3LB agar plate. If required, IPTG was added to this 1/3LB agar plate at a final concentration of 1 mM. After incubation at 30 °C for 24 h, the cells were suspended in PBS, diluted, and spread on selective agar plates.
Transposon mutagenesis of P. putida KT2440 for screening of its mutants able to receive NAH7dF from E. coli
A TnMod plasposon, pTnMod-OGm (Dennis & Zylstra, 1998), carries the pMB1-derived replication origin and Gm resistance (Gmr) gene in the transposable region. This plasmid was introduced into P. putida KT2440 by electroporation to obtain c. 3000 independent Gmr transformants, and every 300–500 pooled transformants were used as the recipients in the conjugation experiment with E. coli MV1190(NAH7dF) as the donor to select KmrGmr transconjugants. The detailed procedures to determine the genomic insertion sites of TnMod-OGm are described in Supporting Information, Data S1.
Construction of KT2440 derivatives
The detailed procedures for allelic mutagenesis and construction of strains carrying chromosomal mini-Tn7 derivatives with the tac-driven wild-type and mutant pts genes are described in Data S1 in Supporting Information.
Isolation of P. putida KT2440 mutants able to receive NAH7dF from E. coli
Our previous finding that NAH7dF (NAH7ΔtraF) exhibited its transferability from P. putida to P. putida and E. coli and from E. coli to E. coli, but not from E. coli to P. putida (Miyazaki et al., 2008) raised a possibility that unknown factor(s) present in P. putida inhibit the conjugative transfer of NAH7dF from E. coli. We attempted to identify such factors by TnMod-OGm mutagenesis of P. putida KT2440 to isolate mutants that were able to receive NAH7dF from E. coli MV1190. Screening of c. 3000 independent mutants led to the isolation of three candidates. The insertion sites of TnMod-OGm in these candidates were determined; two candidates (KTDptsP1 and KTDptsP2) had the insert in ptsP, and the remaining one (KTDptsO) in ptsO (Fig. 1b). To investigate whether the insertion of TnMod-OGm in the two genes was indeed responsible for the ability to receive NAH7dF from E. coli, ptsP and ptsO genes in the wild-type KT2440 were directly disrupted by allelic exchange mutagenesis using the TnMod-OGm-containing genomic fragments from KTDptsP1 and KTDptsO. Both disruptants exhibited phenotypes indistinguishable from those of their respective parental mutants with respect to the ability to receive NAH7dF from E. coli, supporting the actual involvement of these TnMod-OGm inserts. The transfer frequencies of NAH7dF from E. coli to KTDptsP1 and KTDptsO were 3000- and 60-fold, respectively, lower than those of NAH7K2 (Ono et al., 2007), a NAH7 derivative carrying a Kmr gene (Fig. 2a).
Involvement of P. putida PTSNtr in inhibition of NAH7dF conjugation from E. coli
The ptsO and ptsP products of P. putida KT2440, together with its ptsN product, constitute PTSNtr, which is highly conserved in many proteobacterial species. In KT2440, ptsO and ptsN are the members of an operon that includes rpoN for nitrogen-related sigma factor of RNA polymerase (Fig. 1b), whereas ptsP is far apart from the rpoN operon (Nelson et al., 2002). The high-energy phosphate moiety (~P) of phosphoenolpyruvate is transferred to PtsP (EINtr), relayed to PtsO (NPr) and then to PtsN (EIIANtr) (Fig. 1a) (Deutscher et al., 2006). Analysis in various proteobacteria, including P. putida KT2440, has revealed that PTSNtr controls directly or indirectly various microbial processes, and approximately half of such PTSNtr-controlled processes have been clarified to be regulated directly by PtsN or PtsN~P at the transcriptional or post-transcriptional level (Pfluger-Grau & Gorke, 2010).
Because the mutations in two of the three pts genes for the KT2440-specified PTSNtr allowed their host cells to receive NAH7dF from E. coli, we examined whether the KT2440 ptsN deletion mutant constructed by substitution with a Gmr gene (KTDptsN) exhibits a similar phenotype. This mutant strain was also able to receive NAH7dF from E. coli at a frequency sixfold higher and ninefold lower than the ptsP and ptsO mutants, respectively (Fig. 2), also supporting the importance of the KT2440-specified PTSNtr for the conjugative transfer of NAH7dF.
Our use of multicopy vector plasmids for the complementation of chromosomal pts mutations was unsuitable, because even the conjugative transfer of NAH7K2 from E. coli to the wild-type KT2440 derivative carrying the vectors was detected at a frequency drastically lower than that to KT2440 (data not shown), for unknown reasons. We therefore employed a mini-Tn7-based delivery system (Choi et al., 2005) to introduce a wild-type and promoterless but tac-driven pts gene into a specific chromosomal attTn7 site in the KT2440 background. Although repeated attempts to introduce the ptsP-containing mini-Tn7 into KTDptsP were unsuccessful, the tac-driven ptsO and ptsN genes were inserted into the attTn7 sites of the respective mutants. The resulting strains, KTDptsOchr::ptsO and KTDptsNchr::ptsN, were still able to receive NAH7dF from E. coli in the absence of ITPG (Fig. 2). However, the former recipient was, in the presence of IPTG, unable to receive the plasmid by conjugation (Fig. 2a). In contrast, the latter recipient strain was still able to receive NAH7dF from E. coli under this condition, at the frequency statistically indistinguishable from that under the conditions without IPTG (Fig. 2b). One plausible reason for this noncomplementation of the KTDptsN phenotype was that there was no transcription of the ptsO-containing downstream region due to the polar effect caused by insertion of the Gmr gene in ptsN. However, our qRT-PCR analysis showed that ptsO was transcribed in KTDptsN at a level 3.4-fold higher than in KT2440, clearly excluding the possibility of no transcription of ptsO in the former strain. Instead, the IPTG-induced expression of ptsO in KTDptsN led to an absence of conjugative acceptance of NAH7dF from E. coli (Fig. 2b). These results strongly suggested that PtsO in recipient cells played an important role in inhibition of the conjugative transfer of NAH7dF.
Involvement of phosphoryl acceptor/donor residue in PtsO in regulation of NAH7dF conjugative transfer
The phosphorylation sites in the KT2440 PtsO and PtsN proteins are the His residues at the 15th and 68th positions, respectively (Postma et al., 1993; Cases et al., 1999). To investigate whether nonphosphorylated or phosphorylated forms of the two proteins affected the conjugation, site-directed mutagenesis of the promoterless ptsO and ptsN genes was performed so that each mutant gene was expressed in only one of the two forms: the conversion of the His codon to the Ala one for the former form and to the negatively charged Asp one to give rise to the mutant protein that mimics the latter form. Each of the resulting ptsOH15A and ptsOH15D mutant genes was cloned just downstream of the tac promoter in the mini-Tn7 and inserted into the chromosomal attTn7 site of KTDptsO. The KTDptsO derivatives expressing the PtsOH15A and PtsOH15D proteins were both able to receive NAH7dF from E. coli at frequencies not drastically (< 10-fold) different from that when KTDptsO was used as the recipient (Fig. 2a). The tac-driven ptsNH68A and ptsNH68D mutant genes were also introduced into the chromosomal attTn7 site of KTDptsN, and the resulting two strains were, similarly to KTDptsN, able to receive NAH7dF from E. coli (Fig. 2b). These results strongly suggested the important role of phosphoryl acceptor/donor residue of PtsO in the regulation of conjugative transfer of NAH7dF from E. coli.
PtsO affects the conjugation of wild-type form of NAH7 and another IncP-9 plasmid from E. coli
In the experiments described above, the conjugation period of donor and recipient cells was 24 h, which was expected to be sufficient for the saturation of plasmid transfer frequency. To investigate whether the ptsO mutation in the KT2440 recipient affected the transfer frequency of NAH7 at much shorter conjugation periods, the transfer frequencies of NAH7K2 from E. coli to KT2440 and KTDptsO were measured using the 2- and 7-hour periods. As shown in Fig. 3a, the transfer frequencies of NAH7K2 to KT2440 were 30- and 5-fold lower than those to KTDptsO in the respective periods. Such difference in frequencies was not observed when NAH7K2 was transferred from KT2440 (data not shown). These shorter-period conjugation experiments strongly suggested that the PtsO function in recipient cells plays an inhibitory role in the transfer of the wild-type NAH7 plasmid.
To determine whether the effect of PtsO function on the conjugation was limited to NAH7, the transfer frequencies of another IncP-9 plasmid, pWW0, from E. coli to KT2440 and KTDptsO were measured using the same conjugation periods. The transfer of pMT1405 (a pWW0 derivative carrying a Kmr gene) (Tsuda & Iino, 1988) to KTDptsO occurred at a frequency more than 10-fold higher than that to KT2440 (Fig. 3b). The difference in the frequencies was not observed when pMT1405 was transferred from KT2440. These results also suggested the involvement of the recipient PtsO function in conjugative transfer of pWW0 from E. coli.
This study indicated that the defect of NAH7dF in its conjugative transfer from E. coli to P. putida KT2440 could be suppressed by the loss of PTSNtr function in the recipient. In P. putida, the involvement of PtsNtr has been reported in the three metabolisms: (1) PtsN~P modulates the pWW0-specified toluene catabolism by regulating the Pu promoter for the catabolic operon (Cases et al., 1999; Aranda-Olmedo et al., 2006), (2) intracellular accumulation of polyhydroxyalkanoates as carbon storage compounds is significantly decreased by the loss of PstP or PtsO and is increased by the loss of PtsN (Velazquez et al., 2007), and (3) PtsP is required for transcription of the gene for a monooxygenase involved in dimethyl sulfone catabolism (Kouzuma et al., 2007). Our finding in this study provides an additional role of PTSNtr function, which is unique in governing horizontal gene transfer at the intergenus level.
The inhibition of NAH7dF conjugation was further investigated by the complementation analysis using wilt-type ptsO and ptsN genes and their site-directed mutant genes. Although the results were complicated (Fig. 2), the most straightforward results were (1) the successful complementation of the ptsO mutant phenotype by supply of the wild-type ptsO gene and (2) the unsuccessful complementation of the phenotype by supply of the ptsOH15A and ptsOH15D mutant genes. PtsOH15A remains constitutively in its nonphosphorylated form, and PtsOH15D cannot be phosphorylated by PtsP~P. Therefore, it is unlikely that PtsO~P per se is directly involved in the inhibition of conjugation, but it is likely that PtsN or another unknown protein able to receive ~P from PtsO~P is directly or indirectly involved in the inhibition of conjugation. A previous report showing that almost all of the PtsN proteins in LB-grown stationary-phase cells of KT2440 are in the phosphorylated form (Pfluger & de Lorenzo, 2007) led us to predict that PtsN~P directly or indirectly inhibits the NAH7dF conjugation from E. coli, because KTDptsN was able to receive the plasmid due to the absence of PtsN production. However, this prediction was inconsistent with the results that (1) the ptsN mutant phenotype was not complemented by supply of the wild-type ptsN or ptsNH68D gene and (2) the ptsN mutant phenotype was successfully complemented by supply of the wild-type ptsO gene (Fig. 2b). Therefore, a possible alternative mechanism is that an unknown host protein (hereafter designated protein X) that is able to directly receive ~P from PtsO is involved in the direct or indirect inhibition of the conjugation; the PtsN protein overproduced by the IPTG induction might receive ~P from PtsO, leading to the relatively high amounts of the nonphosphorylated form of protein X, while the absence of PtsN in the cells might result in the preferential transfer of ~P to protein X from the overproduced PtsO protein. Further genetical and biochemical analyses including the identification and characterization of protein X will provide clues regarding the possible mechanism. It would also be of interest to know which step in the conjugative transfer of NAH7dF from E. coli to wild-type KT2440 is defective. Detailed investigations of this type will help us to understand the mechanism that links the PTSNtr function to the conjugation phenomenon.
The inhibitory effect of PtsO in the recipient was also observed in the conjugative transfer of NAH7 itself and pWW0 from E. coli. pWW0 carries the traF homolog (orf176) whose product shows 77% amino acid identity with the NAH7 traF product (Fig. S1 in Supporting Information) (Miyazaki et al., 2008). It would be of interest to determine whether the orf176 mutant of pWW0 exhibits a phenotype similar to that of NAH7dF. The inhibition of the conjugative transfer of NAH7 was only c. 30-fold, in contrast with the much higher-level inhibitory effect when the NAH7dF was used for the conjugation. Therefore, the inhibitory effect of PtsO in the recipient P. putida cells appeared to be compensated for by the expression of the traF product in the donor E. coli cells. It will also be of interest to clarify the molecular mechanisms governing this phenomenon, and the elucidation of the TraF function will contribute to clarifying these mechanisms. In any case, the present study clearly indicated that the conjugative transfer of plasmids is controlled by a previously unidentified recipient system, PtsNtr, which is involved in various and diverse cellular processes in bacteria.
This work was supported by Grants-in-Aids from the Ministry of Education, Culture, Sports, Science and Technology, Japan.