Previous Article | Next Article ![]()
Journal of Bacteriology, September 2002, p. 4757-4766, Vol. 184, No. 17
0021-9193/02/$04.00+0 DOI: 10.1128/JB.184.17.4757-4766.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
Department of Environmental Life Sciences, Graduate School of Life Sciences, Tohoku University, 2-1-1 Katahira, Sendai 980-8577, Japan
Received 21 March 2002/ Accepted 5 June 2002
|
|
|---|
|
|
|---|
![]() View larger version (39K): [in a new window] |
FIG. 1. Structure of Tn4651. Restriction site abbreviations: A, AatII; Ap, ApaI; B, BamHI; E, EcoRI; H, HindIII; K, KpnI; N, NcoI; Sa, SacI; Sb, SacII; Sm, SmaI; St, StuI; and X, XbaI. Restriction sites derived from the vectors are in parentheses. (A) Schematic structures of Tn4651 (13, 35) and the related transposon Tn5041 (17). The following abbreviations and symbols are used: xyl, toluene-degrading genes; U, upper pathway operon; M, meta pathway operon; mer, genes for resistance to mercuric ion; TnpC, a regulator that controls negatively the TnpA expression at the posttranscriptional level (13); filled arrowhead, IR; and open arrowhead, IS1246 (26). The arrow indicates the transcriptional direction of the genes and operons. (B) Detailed structure of the tnpS-res-tnpT region. The numbers above the restriction sites are the base positions obtained by defining the upstream XbaI site as position 1. The short arrow below the map is the sequence in the primer used for amplification of DNA fragments or for primer extension analysis. Note that the tnpS and tnpT genes had been indicated to be located within the XbaI-SacII and AatII-SmaI fragments, respectively (34). (C) Structures of the pUC- or pBR322-based plasmids carrying tnpS, tnpT, and both genes. The detailed procedures for construction of these plasmids are described in Materials and Methods. The size of the Kmr gene on pGEN7K is arbitrary. While pGEN59 and pGEN77 carry the wild-type tnpS gene, pGEN95 and pGEN96 carry the Y293F mutation (triangle) in the tnpS gene.
|
In the present study, we analyzed the Tn4651-encoded site-specific resolution system under E. coli recA backgrounds. We show that (i) TnpS is a site-specific recombinase belonging to the integrase family and catalyzes both intra- and intermolecular site-specific recombinations and (ii) TnpT appears to have a role to enhance the former type of recombination.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Bacterial strains and plasmids used in this study
|
![]() View larger version (36K): [in a new window] |
FIG. 2. Construction of pMT299 and R388 derivatives carrying various parts of the res site. Only the relevant restriction sites are depicted, and their abbreviations are as described in the legend of Fig. 1 except for the following: Hc, HincII; P, PstI; and Sc, ScaI. The filled arrow indicates the res site. The white arrowheads above some plasmids indicate the primers used for amplification of parts of the res sites. The PCR product with primers M4 and RES1 was used to obtain res1 present on pGENKres1, pMT299res1, pGEN101, and pGEN388; that with M4 and RES2 was used to obtain res2 on pGENKres2, pMT299res2, pGEN102, and pGEN389; and that with M4 and RES3 was used to obtain res3 on pGENKres3, pMT299res3, and pGEN103. For construction of pGENres6, PCR was carried out by using pMT299res2 as the template with primers RES6 and 184E. See Fig. 3 for details. The bold line indicates the pMT299-derived region. The Kmr gene in all of the plasmids is transcribed in a leftward direction.
|
![]() View larger version (43K): [in a new window] |
FIG. 3. Structure of the Tn4651 res site. (A) Sequence of the res site. The numbers at the left are the base positions obtained by taking the upstream XbaI site (Fig. 1B) as position 1. The transcription start sites of tnpS and tnpT are indicated by closed circles. The putative -35 and -10 sequences for tnpS and tnpT and the putative SD (S.D.) sequence for tnpT are boxed, and the predicted translation start sites of the tnpS and tnpT genes are shown by bold arrows. The 9-bp IRs are indicated by a pair of thin arrows, and the putative IHF-binding site (9) is doubly overlined. The oligonucleotides for PCR amplification of various parts of the res site are indicated as white boxes above or below the sequence. (B) Summary of the Tn4651 site-specific recombination analysis. The structure of one of the two copies of the res region on each plasmid is schematically drawn. The number above the structure indicates the nucleotide position (Fig. 1B). Open and filled arrows indicate the 9-bp IRs and the putative IHF binding site, respectively. Ssr, detection of site-specific resolution of the plasmid on the agar plates; PCR, detection of the excised and nonreplicating DNA by PCR with primers 4Km1 and 4Km2 (see Fig. 6); Ssi, detection of site-specific integration between the res sites located at two different DNA molecules (Table 2); +, detected; -, not detected.
|
In order to change the tyrosine residue of TnpS to phenylalanine, the megaprimer PCR method (6) was employed for site-directed mutagenesis of tnpS by using KOD DNA Polymerase (Toyobo). The first PCR was carried out with the primers S1 and SA179T and pGEN59 as a template (Fig. 1C). The amplified fragment was used as the megaprimer for the second PCR with S2 as another primer. The amplicon containing the mutant tnpS gene was digested by EcoRI and HindIII and inserted into the corresponding sites of pKK223-3 to obtain pGEN95 (Fig. 1C). Excision of the BamHI-HindIII fragment containing the tac promoter and the mutant tnpS gene from pGEN95 and its subsequent insertion into the corresponding sites of pGEN21 gave rise to pGEN96.
Determination of transcription start site by primer extension. Plasmid pMT1405 (pWW0::kan) (35) was introduced into P. putida PT280 (Table 1) that was shown not to carry the Tn4651 tnpS-res-tnpT region (unpublished data). An RNeasy Mini Kit (Qiagen) was employed to prepare the total RNA from 1-ml samples of LB cultures of PT280 and PT280(pMT1405). A primer extension experiment in a 20-µl reaction volume was performed according to the protocol of Aloka, Ltd. (Tokyo, Japan), with the IRD-41-labeled oligonucleotide primer PEtnpS or PEtnpT (Fig. 1B). (The primer sequences are available upon request.) Gel electrophoresis was carried out by using a Li-Cor 4000 automated DNA sequencer.
Site-specific recombination assays. To analyze the site-specific resolution of Tn4651 in this study, a pACYC184-based plasmid, pMT299 (Table 1), was used to construct its derivative carrying the pUC4K-derived Kmr gene that was flanked by two copies of the res site or its part (Fig. 2). JM109 cells carrying the resulting plasmid were subjected to transformation of the pMB1-based plasmid that carried tnpS, tnpT, or both genes whose expression was under the control of the tac or trc promoter. The resulting transformant cells were briefly cultivated in LB medium with or without the addition of IPTG and then plated on LB agar plates containing appropriate antibiotics. The colonies thus obtained were examined for sensitivity to kanamycin and for the profiles of the residing plasmids. Excision of the circular DNA form consisting of the res site and the Kmr gene was also examined by PCR with primers 4Km1 and 4Km2 (Fig. 6).
![]() View larger version (37K): [in a new window] |
FIG. 6. PCR detection of site-specific resolution. (A) Schematic drawing for PCR detection of the excised and nonreplicating DNA consisting of the res site and Kmr gene. White arrowheads indicate the primers. (B) Agarose gel electrophoresis of the PCR-amplified excised DNA fragments. PCR with primers 4Km1 and 4Km2 was performed by using the cleared lysate prepared from the JM109 derivative carrying the following combinations of plasmids: pGEN101 and pGEN21, lane 1; pGEN101 and pGEN59, lane 2; pGEN101 and pGEN77, lane 3; pGEN101 and pTrc99A, lane 4; pGEN101 and pGEN95, lane 5; pGEN101 and pGEN96, lane 6; pGEN102 and pGEN77, lane 7; pGEN103 and pGEN77, lane 8; pGEN106 and pGEN77, lane 9. In lane M, size markers are indicated in kilobases.
|
cells that had contained the pMT299 derivative carrying one copy of the res site and the pMB1-based plasmid carrying tnpS, tnpT, or both genes. The strain harboring these three plasmids was employed as the donor to mate with HB101 on a membrane filter (34), and the TET resistance-streptomycin resistance (Tcr Smr) transconjugants were selected to detect the mobilization of the pMT299-based replicon. Such transconjugants were examined for the physical analysis of the plasmids. Determination and analysis of nucleotide sequence. The DNA fragments cloned on pUC18 and pUC19 were sequenced with an ABI 310 automated DNA sequencer (Applied Biosystems) according to the protocols recommended by the manufacturer. The nucleotide sequence reported here has been deposited in the DDBJ-EMBL-GenBank database under accession number AB077820. The computer analysis of the sequence was performed with the software programs GENETYX 10 (SDC, Inc., Tokyo, Japan) and BLAST 2 (National Institute of Genetics, Mishima, Japan).
|
|
|---|
BLAST searches revealed that TnpS had homology with the integrase family of site-specific recombinases (23), showing 36 to 38% identity and 53 to 55% similarity with the putative integrase from the atrazine catabolic plasmid pADP-1 of a Pseudomonas strain (U66917) (22), the integrase-like protein from Coxiella burnetii (Y15898), and the OrfI protein of Tn5041 (X98999) (17). The proteins in this family invariably have the conserved R-H-R-Y tetrad in domains I and II with proper spacing (23), and these structural characteristics were also found in TnpS (Fig. 4). In contrast, TnpT showed 29% identity and 47% similarity only with the OrfQ protein of Tn5041 (X98999) (17).
|
View larger version (14K): [in a new window] |
FIG. 4. Comparison of TnpS with other members in the integrase family of site-specific recombinases. The consensus sequences in this family are depicted at the top of the figure; the invariant R-H-R-Y tetrad residues in the family are indicated by the capital and boldface letters, and the highly conserved amino acid residue is indicated by the small letter (23). The number beside the amino acid sequence is the amino acid position. The mutagenized tyrosine residue in TnpS is indicated by an arrowhead. The accession numbers of the nucleotide sequences for the proteins are as follows: Tn4651 TnpS (AB077820); pADP-1 Int, putative integrase from plasmid pADP-1 (U66917); Coxiella Int, integrase-like protein from C. burnetii (Y15898); and Tn5041 OrfI, integrase-like protein from Tn5041 (X98999).
|
Site-specific resolution activity of tnpS-res-tnpT region. To analyze the Tn4651-encoded resolution system under an E. coli recA background, a pACYC184-based Tcr Kmr plasmid, pGEN101, was constructed in which a Kmr gene was flanked by the two copies of the 203-bp res site (res1) (Fig. 2 and 3B). We also constructed pGEN77 by insertion of the tac-tnpS region of pGEN59 into pGEN21 that had the trc-tnpT gene (Fig. 1C). We thereafter introduced pGEN21, pGEN59, or pGEN77 into JM109(pGEN101). Since the prolonged cultivation of JM109(pGEN101)(pGEN21) that overexpressed TnpT alone, but not the two other strains, on LB plates containing 0.1 mM IPTG led to the very severe growth defect, the three JM109 derivatives were cultivated only for 1 h at 37°C in AMP- and TET-containing LB medium with or without IPTG. After such treatment, the cells were plated on LB agar plates containing AMP and TET, and the 150 colonies obtained from each experiment were examined for resistance to kanamycin. Cultivation of JM109(pGEN101)(pGEN21) and JM109(pGEN101)(pGEN59), even in the presence of IPTG, generated no kanamycin-sensitive (Kms) colonies, and physical analysis of the residing plasmids (Fig. 5A, lanes 1 and 2) revealed that the structure of pGEN101 remained unaffected. In contrast, JM109(pGEN101)(pGEN77) led to formation of the Kms colonies at frequencies of 6 or 9% when the cells were cultivated in the absence or presence, respectively, of IPTG. The addition of IPTG did not lead to a drastic increase in the frequency of formation of the Kms colonies for unknown reasons. The leaky transcription from the trc and tac promoters in the absence of IPTG (20) might have allowed expression of TnpS and TnpT sufficient for the resolution. The Kms derivatives contained, in addition to pGEN77, a novel plasmid (pGEN101R) that was smaller than pGEN101 (Fig. 5A, lane 4). Analysis of pGEN101R by restriction digestion and sequencing confirmed that this plasmid was formed by site-specific resolution, indicating that both tnpS and tnpT were required for efficient resolution. pGEN101R was also detected in the cleared lysate prepared from the Kmr colonies of JM109(pGEN101)(pGEN77) after their prolonged cultivation in LB containing AMP and TET (Fig. 5A lane 3), indicating that pGEN101R was gradually generated from pGEN101 by site-specific resolution.
![]() View larger version (22K): [in a new window] |
FIG. 5. Resolution and cointegration of plasmids carrying Tn4651 res site. (A) Resolution of plasmid that carried the Kmr gene flanked by two directly repeated copies of the res site. The JM109 derivative carrying the substrate and helper plasmids was cultivated for 48 h in LB medium containing AMP and TET, and the cleared lysate prepared from each culture was analyzed by agarose gel electrophoresis. The substrate plasmids used were as follows: pGEN101, lanes 1 to 5; pGEN102, lane 6; pGEN103, lane 7; pGEN106, lane 8. The helper plasmids were as follows: pGEN21, lane 1; pGEN59, lane 2; pGEN77, lanes 3, 4, 6, 7, and 8; pGEN96, lane 5. The cleared lysates prepared from the LB cultures of the Kmr and Kms derivatives of JM109(pGEN101)(pGEN77) were loaded in lanes 3 and 4, respectively. In lanes 1 and 2, pGEN21 and pGEN59 overlap pGEN101. oc, open circular DNA; chr, chromosome. (B) Cointegration of two plasmids, each carrying the Tn4651 res site. The donor strain DH5 (pGEN101R)(pGEN388)(pGEN77) was mated with HB101, and the Tcr transconjugants were selected. Agarose gel electrophoresis of cleared lysates prepared from the donor strain (lane 1) and the transconjugant (lane 2) was done.
|
To investigate whether the excision of the Kmr gene was correlated with formation of the circular DNA form consisting of the Kmr gene and one copy of the res site, PCR detection of such excision events was performed with the primers 4Km1 and 4Km2 (Fig. 6A). Use of the cleared lysates from the Kmr cells of JM109(pGEN101)(pGEN77) and JM109(pGEN102)(pGEN77), but not from those of JM109(pGEN103)(pGEN77), led to successful amplification of the DNA molecules with the expected sizes (Fig. 6B, lanes 3, 7, and 8). These results also supported the conclusion that loss of the Kmr gene occurred by its site-specific excision. The excision events were also detected by PCR from the cleared lysates prepared from JM109(pGEN101)(pGNE59) (Fig. 6B, lane 2) and JM109(pGEN102)(pGNE59) (data not shown), although these two strains did not generate Kms derivatives (<0.6%). Therefore, TnpT was not certainly required for site-specific resolution but might play a role in enhancing the frequency of site-specific resolution.
The 95-bp sequence covering the rightmost portion of res2 (positions 1110 to 1204 in Fig. 3A) was designated res6, and we constructed a pGEN101-related plasmid, pGEN106, in which the Kmr gene was flanked by two copies of res6 (Fig. 2 and 3B). The JM109(pGEN106)(pGEN77) strain did not generate Kms derivatives (<0.6%), and the resolved plasmid derived from pGEN106 was not detected when the cleared lysate prepared from the Kmr cells was analyzed by agarose gel electrophoresis (Fig. 5A, lane 8). However, PCR analysis of the cleared lysate (Fig. 6, lane 9) supported the formation of the excised circular DNA consisting of the Kmr gene and a copy of res6. PCR analysis of the cleared lysate prepared from JM109(pGEN106)(pGEN59) led to the same observation (data not shown). The right end of res6 (Fig. 3B) had the IR of a 9-bp sequence with a 2-bp spacer and a putative consensus sequence for the binding site of the E. coli IHF protein (9).
Site-specific integration activity of Tn4651-specified recombination system.
Since TnpS had the motif commonly conserved in the integrase family of site-specific recombinases (Fig. 4), we anticipated that TnpS alone or in concert with TnpT might be able to mediate site-specific integration of a res-carrying plasmid into another plasmid that also carried the res site. We expected that such a recombination event could be detected effectively by conjugal mobilization of the nontransmissible plasmid by the coresiding transmissible plasmid only when both plasmids carried the functional res site. pGEN101R that carried res1 and the Tcr gene was employed as the former type of plasmid, and pGEN388 (Fig. 2) was constructed as the latter type of plasmid by insertion of res1 and the Kmr gene into transmissible plasmid R388 (4). The DH5
derivative carrying pGEN101R, pGEN388, and one of the pMB1-based plasmids (pGEN21, pGEN59, or pGEN77) was used as the donor to mate with HB101, and mobilization of pGEN101R was examined by selecting Tcr Smr transconjugants (Table 2). When we used pTrc99A that did not carry tnpS or tnpT, pGEN101R was mobilized at a frequency of ca. 10-6 per pGEN388 transfer, and each Tcr transconjugant carried a single species of plasmid (data not shown) that could revolve into the two parental plasmids in the presence of pGEN77. Restriction analysis of the fused plasmid and sequence analysis of its res1-containing regions confirmed that the plasmid was formed by reciprocal recombination between the two copies of res1. Since neither tnpS nor tnpT was involved in the unexpected recombination, we did not further investigate the mechanism for such a low-frequency recombination event. Supply of TnpT also resulted in such a low-frequency mobilization of pGEN101R. A >1,000-fold increase in mobilization of this plasmid was observed when TnpS alone and when TnpS and TnpT simultaneously were supplied. The transconjugants thus obtained carried a single species of plasmid (Fig. 5B) whose structure was the same as that residing in the transconjugants that were obtained by use of DH5
(pGEN101R)(pGEN388)(pTrc99A) as the donor. These results indicated that TnpS alone could efficiently promote site-specific integration without involvement of TnpT. Use of res2 and res6 on the two plasmids, pGEN389 and pMT299res6 (Fig. 2 and 3B), respectively, also generated the cointegrate (Table 2), indicating that res2, and most probably res6, was sufficient for site-specific integration.
|
View this table: [in a new window] |
TABLE 2. Site-specific integration mediated by Tn4651 res sitea
|
|
|
|---|
The res2 sequence, and most probably the res6 one, could act as the substrate for intramolecular resolution and intermolecular integration reactions, and analysis in this study revealed that (i) both reactions could be catalyzed by TnpS that had the structural properties characteristic of the integrase family and (ii) TnpT was most likely to enhance the former type of reactions. The site-specific recombination system of Tn4651 could therefore be considered to be apparently analogous to those encoded by integrons in which the integrases are able to catalyze both the excision and the integration reactions and, to a lesser extent, analogous to those encoded by lysogenic phages and some conjugative transposons (e.g., Tn916 and Tn5276) (18, 25, 30). With respect to the latter group of mobile elements, the small (<100 amino acid residues) accessory proteins collectively designated recombination direction factors (RDFs) (e.g., lambda Xis protein) (19) are, in addition to the integrases, required absolutely for site-specific excision, and some RDFs have been shown to bind to specific recognition sequences at the att sites so as to introduce sharp DNA bends (8). However, TnpT was much larger than RDFs and did not show homology to any RDFs. It is at present unknown whether TnpT also has the ability to bind to the Tn4651 res site since our attempt to overproduce the TnpT protein alone in E. coli led to a very toxic effect on the growth of the host cells. Our observation that such a toxic effect could most probably be suppressed by the chromosomal but uncharacterized mutation(s) indicated that TnpT might interact with the unknown host product(s). Identification of the mutated chromosomal gene(s) might provide us with some clues to elucidate the molecular mechanism of TnpT in the resolution. There is also a possibility that TnpT interacts directly with TnpS to enhance the resolution frequencies. Such possible titration of the free form of TnpT by TnpS might be able to account for our observation that the attempt to overexpress both TnpS and TnpT simultaneously in E. coli resulted in a less toxic effect on the host cells. Site-specific recombination between the two res sites, one on the pACYC184 replicon and the other on the R388 replicon, was detected at a low frequency without the involvement of TnpS or TnpT (Table 2). Construction of the latter R388 derivative was associated with disruption of the site-specific integrase gene of In3, an integron on R388 (10). Such a low-frequency integration event might therefore be most likely ascribed to the activity of an unknown integrase protein encoded by the host E. coli chromosome. However, BLAST searches of the protein databases did not clarify the candidate proteins that displayed similarity with TnpS.
A 15-kb mercury transposon, Tn5041 (Fig. 1A), from a Pseudomonas strain (17) is closely related to Tn4651 in that (i) these two transposons share >95% identity in the genes and sites involved in cointegration (i.e., tnpA and terminal IRs) and (ii) the Tn4651 tnpA product could catalyze the cointegration of Tn5041 (16, 17). The 2.3-kb region of Tn5041 probably responsible for the cointegrate resolution is separated by 9 kb from the tnpAC genes, and this 2.3-kb region carries the putative res site flanked by the two genes, orfI and orfQ, that are postulated to be transcribed divergently (17). Furthermore, the sizes of orfI, orfQ, and the res site of Tn5041 are similar to those of the tnpS, tnpT, and res site, respectively, of Tn4651, and OrfI that possesses the motifs conserved in the integrase family has been speculated to be the counterpart of TnpS. However, the resolution-related region of Tn5041 was not so homologous as to hybridize with the corresponding region of Tn4651 by Southern analysis (17). The Tn5041-encoded resolution system remains to be elucidated since the resolution activity of Tn5041 assayed by a system similar to ours was very weak in the E. coli cells (16). An analysis of the Tn5041-encoded resolution system in P. putida and further investigation of the functional exchangeability of TnpS and OrfI and of TnpT and OrfQ will uncover more clearly the degree of relatedness between the two transposons.
This work was supported by a grant-in-aid from the Ministry of Education, Science, Sports, and Culture of Japan.
|
|
|---|
S in starvation-induced transposition of Pseudomonas putida transposon Tn4652. J. Bacteriol. 183:5445-5448.
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»