INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

A class of temperate phages are capable of integrating their genomes into the host chromosome by site-specific recombination. The process is best known and has been described in detail for Escherichia coli phage  (28, 49, 55). However, there are some other well-characterized integrative systems, such as those of HP1, P22, L5, and pSAM2, where the array of the structural elements differs from the arrangement known from  phage (4, 17, 42). Those studies indicate that although the molecular mechanism of the process is basically the same, there are several different ways of accomplishing integration. The study of new integrative systems should provide an opportunity to reveal alternative pathways, widening our understanding of the mechanism of site-specific recombinations.

Site-specific recombination is one of the basic tools for basic research and biotechnology. Because of the specificities of required host factors, a particular integrative recombination system can be used in a limited field, so a new site-specific recombination system provides the opportunity to apply genome techniques to new species.

Phage 16-3 is able to integrate its genome into the chromosome of Rhizobium meliloti 41, forming lyzogens (35-37). The target site of this integration is between the cys-46 and met-5 genes of R. meliloti 41, and cys-46 may undergo specialized transduction with 16-3 (50). The 16-3 integrative recombination system has been partially characterized (34). Both the attachment regions, attB and attP, were localized (8, 50) and their nucleotide sequences were determined (10). The attB region contains a putative proline tRNA (tRNA^Pro) gene. A sequence of 51 bp, identical in the bacterial and the phage att regions overlapping the 3' end of the tRNA^Pro gene, was expected to contain the core region where strand exchanges take place during the recombination process. This sequence alone was sufficient to serve as a target site for phage integration. Due to the topology of the overlap, the nucleotide sequence of the tRNA^Pro gene is not altered by 16-3 integration. It was found that the putative tRNA^Pro gene of R. meliloti shows significant homology to the putative tRNA^Pro gene of Streptomyces ambofaciens, which serves as a target site for integration of pSAM2, a self-transmissible plasmid carrying integrative elements (39). The integrase (int) and excisionase (xis) functions of phage 16-3, previously localized on a 15-kb segment of the phage genome, are under the control of the C repressor protein, the domain structure and DNA binding specificity (related to the coliphage 434 cI repressor) of which are also known (6, 7, 11, 34, 35, 37, 38). Here we report the precise identification of the int and xis genes of phage 16-3. Amino acid sequence comparisons classify the 16-3 Int protein in the integrase family of tyrosine recombinases, analyzed recently in references 13 and 33. We found that the site-specific recombination system of phage 16-3 functions efficiently in R. meliloti 41 but is inactive in E. coli. Since site-specific recombination systems deriving from different sources play major roles in gene technology (46, 53), the development of a new integrative vector family based on the site-specific recombination system of phage 16-3 may serve as an appropriate tool in many applications.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial and phage strains, growth conditions, and triparental matings. E. coli DH5 (18) was used in all cloning experiments and served as the host of donor plasmids used for conjugation into the recipient R. meliloti strain, 41 (51) (the native host of phage 16-3), and as the host for the study of site-specific recombination. Growth conditions, media, and conditions for triparental matings were as described in reference 39.

DNA procedures. Basic DNA manipulations and molecular techniques were employed as described in reference 44. Extraction of DNA from agarose gel was done with a QIAEX II Gel Extraction Kit (Qiagen). Total bacterial DNA was prepared by the method described in reference 3. DNA was labelled by nick translation in the presence of [-³²P]dATP. Hybridization was performed as described previously (48). PCR primers are listed in Table 1. PCR-mediated DNA amplifications were carried out with Taq polymerase (Promega or Sigma) to generate DNA fragments for cloning. After 30 cycles of 1 min at 94°C, 1 min at 55°C, and 1 min at 72°C, the PCR products were extracted with phenol and precipitated in ethanol. Then the DNAs were resuspended in Tris-EDTA buffer and digested with the appropriate restriction enzyme(s) to generate the required ends of the fragments. The DNA fragments were purified before being cloned by isolating them from agarose gels. PCR mutagenesis was performed according to the method in reference 27. Nucleotide sequence determination was performed by the dideoxy chain termination method (45) by using a TaqTrack Sequencing Kit (Promega). Total protein samples were analyzed on a discontinuous sodium dodecyl sulfate-polyacrylamide gel electrophoresis system (26) and blotted to a polyvinylidene difluoride membrane. The protein in the bands representing the protein of interest was sequenced with an Applied Biosystems protein sequencer (model 471) with an Edman degradation sequenator program (21).

attL and attB diagnostic PCR assay. PCR was performed on total bacterial DNA preparations from R. meliloti. Each 50-µl reaction mixture contained 100 ng of the template and 30 pmol of each primer. PCR products were analyzed by electrophoresis through either a 2% agarose gel or a 6% nondenaturing polyacrylamide gel. Primer 1 and primer 2 (Table 1) were used to detect specifically attL, which was indicated by the appearance of a 186-bp-long PCR product. Use of primer 1 and primer 3 (Table 1) resulted in a PCR product of 199 bp, indicating specifically attB.

Sequence analysis. Sequence analyses were performed with the programs of the Wisconsin Package, version 9.1 (Genetics Computer Group, Madison, Wis.). BLAST (41) and FASTA (2) were used to search for similarity with sequences in the GenBank, EMBL, SwissProt, and PDB databases.

Construction of the pSEM91 expression vector. HincII digestion of pCU999 (40) generates three fragments. Two of the fragments were combined; one contained the kanamycin resistance gene and the other carried the replication region of plasmid pCU1 (24). The resulting plasmid was linearized by PvuII digestion and ligated to the HindIII-SalI fragment (both ends were made blunt by end filling) derived from pSUP201 (47) containing an RP4 mob region. The resulting plasmid was called pSEM64. The multicloning site (MCS) of the expression vector pKK223-3 (Pharmacia) was altered, i.e., pKK223-3 was digested with EcoRI and the ends were filled in, and then the DNA was further digested with HindIII. The NotI (blunt ended)-HindIII fragment of the MCS from pBluescript II KS (Stratagene) was inserted into the digested pKK223-3 vector. A HincII fragment of the resulting plasmid containing the tac promoter, the altered MCS, and rrnB T₁T₂ transcription terminators was inserted into the unique EcoRI site of pSEM64, the ends of which were made blunt by end filling. The resulting expression vector was called pSEM91 (Fig. 1).

View larger version (17K): [in this window] [in a new window]   FIG. 1.   Map of the pSEM91 expression vector constructed from the replicon of plasmid pCU1 (hatched bar), RP4 mob region (shaded bar), and kanamycin resistance gene (Km) (black arrow). The shaded triangle represents the tac promoter, while the open bar shows the location of transcription terminators. Restriction sites in bold letters are unique.

Plasmid constructs used to analyze site-specific integration. A detailed physical map of phage 16-3 facilitated building of the different plasmid constructs listed in Fig. 2. Restriction sites with numbers in parentheses refer to physical map positions as shown in reference 9. pSEM6 carries the EcoRI (48)-EcoRI(52) fragment of phage 16-3 at the EcoRI site of plasmid pLAFR1 (15).

View larger version (30K): [in this window] [in a new window]   FIG. 2.   Plasmid constructs and the locations of their 16-3 phage content. (A) The 16-3 genome is indicated at the top (the scale is given in kilobases). Bars represent the extent and topology of the 16-3-derived part of each cosmid or plasmid construct. The region present in pSEM35 is enlarged to show details. The shaded bar shows the region of the determined sequence (GenBank accession no. AJ131679; the scale is given in base pairs). Open bars indicate deletions. Black arrows indicate the gene of the phage repressor (c). Stippled boxes represent the attachment region of the phage (attP). Restriction sites with numbers in parentheses refer to physical map positions as shown in reference 9. The arrowhead marked pLOL indicates the promoter-operator unit to the left of the repressor gene. Binding of the repressor to the pLOL unit regulates not only the lytic-lysogenic decision but also influences the site-specific recombination process of phage 16-3. (B) ORFs of the potential candidates for encoding Int and Xis proteins (black arrows). Black bars indicate the regions carried by different plasmid constructs. Stippled boxes represent the attachment region of the phage (attP). Numbers in parentheses following the names of the plasmids indicate phage content by base positions (the scale is given in base pairs below the shaded bar).

Plasmid constructs used to study excision. To construct pSEM161, the StyI fragment (nt 242 to 2166) from pSEM35 was isolated and the ends were made blunt (Fig. 2B). The fragment was digested with EcoRI and ligated to XbaI (blunt ended)-EcoRI pSEM91 DNA. The recombinant plasmid containing the 619-bp region from nt 1558 to 2166 was called pSEM161. pSEM163 contains the EcoRI(52)-HinfI fragment of 366 bp (nt 1558 to 1923); the HinfI fragment (nt 1130 to 1923) was isolated from pSEM35 and the ends were made blunt. The EcoRI digest of the fragment was ligated to XbaI (blunt ended)-EcoRI pSEM91 DNA, and the recombinant plasmids were identified. To create pSEM208, primer 4 and primer 7 (Table 1) were used. The region containing open reading frame 111 (ORF-111) was amplified by PCR with pSEM25 DNA as the template. The product was digested with BspHI and EcoRI, and the fragment (nt 1558 to 1946) was inserted into NcoI- and EcoRI-digested pET23d (Novagen). From the resulting plasmid the XbaI-EcoRI fragment carrying a ribosome binding site in front of ORF-111 was isolated and cloned into XbaI- and EcoRI-cut pSEM91. pSEM231 was built to express ORF-140. With pSEM25 template DNA and primer 4 and primer 8 (Table 1), the region from nt 710 to 2046 was amplified by PCR. The product was digested with XbaI and EcoRI (nt 1558 to 2046) and inserted into XbaI- and EcoRI-cut pSEM91. pSEM249 derives from pSEM161. Its PstI(53) site was eliminated by T4 polymerase treatment following digestion of the plasmid DNA with PstI, and the treated DNA was self-ligated.

Nucleotide sequence accession number. The nucleotide sequence of the EcoRV(49)-EcoRI(61) region of the 16-3 phage has been deposited in GenBank under accession no. AJ131679.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Construction of pSEM91 suitable for expressing genes both in E. coli and in R. meliloti. Studying the site-specific recombination system of phage 16-3 required the construction of a plasmid which allowed functional analyses of different fragments of phage origin in R. meliloti 41, the native host of 16-3 phage. Basic components of the plasmid were selected to enable us to examine whether this recombination system can function in E. coli and, in addition, to create a vector overexpressing the desired protein in E. coli for purification purposes. The replicon of plasmid pCU1, providing a broad host range, was fused to a kanamycin resistance gene and to the RP4 mob region. The plasmid containing these three elements was further extended by inserting into it an expression panel which carried the tac promoter followed by MCS and transcription terminators. The resulting expression vector was called pSEM91 (Fig. 1). Expression of different genes inserted into pSEM91 was constitutive in both E. coli DH5 and R. meliloti 41 without the addition of any inducer, because neither strain contained the lac repressor gene.

Identification of the int gene of phage 16-3. Previous studies indicated that the site-specific recombination function is located in a 15-kb region present in two different cosmid clones, pDH79 and pDH114, which are able to perform autonomous, prophage-like integration into and excision from the chromosome of host R. meliloti 41 (34).

ORF-371 codes for 16-3 integrase belonging to the tyrosine recombinases. pSEM167 contains the expressible int gene and the attP region of phage 16-3 and readily integrates into the R. meliloti 41 chromosome (Fig. 3, lanes 1). To verify that the int function is coupled to ORF-371 in R. meliloti, we constructed pSEM223, in which the translation frame of ORF-371 was shifted by a 4-bp deletion while the frame of ORF-291 remained unharmed. Since this mutation abolished the ability of the attP-containing plasmid to be integrated, we concluded that ORF-371 represents the int gene of phage 16-3.

View larger version (28K): [in this window] [in a new window]   FIG. 3.   Site-specific integration of plasmids containing attP into the R. meliloti chromosome as detected by PCR (A) and by Southern blotting (B and C). (A) The appearance of a 186-bp-long PCR product indicating attL formation due to integration of pSEM167 (lane 1) and of pSEM6 in the presence of pSEM164 (lane 2) can be seen. The presence of pSEM6 alone (lane 3) or the pSEM91 expression vector (lane 4) in R. meliloti did not result in attL formation. attL is also not present in R. meliloti (lane 5). M indicates the AluI digest of pBluescript II KS used as a molecular size marker. (B) Southern hybridization with a 32P-labelled attP fragment. The order of the samples is the same as in panel A except that a mixture of known attP-containing fragments of different sizes was used for molecular size markers. The presence of attL and attR carrying EcoRI restriction fragments identified the site-specific recombination event (lanes 1 and 2). The attP-carrying fragments of pSEM6 indicate extrachromosomal copies of the plasmid (lanes 2 and 3). With the construct pSEM167 (lane 1) the extrachromosomal copies of the plasmid were lost, probably due to interference with plasmid replication. (C) Southern hybridization with the 32P-labelled attB fragment. The same filter as in panel B was used. attL and attR fragments were identified when site-specific recombination occurred (lanes 1 and 2), but only attB could be detected in the controls (lanes 3 to 5). Rm41, R. meliloti 41.

View larger version (20K): [in this window] [in a new window]   FIG. 4.   Comparison of 16-3 Int protein with the conserved regions of tyrosine recombinases. The open bar indicates schematically the sequence of a protein, and shaded boxes represent the locations of homology regions. The conserved residues and the ranges of spacings between them (as per reference 13) are indicated below the bar. Seven integrases were chosen to show the strong similarity of the homology regions found among the selected integrases and 16-3 Int protein. Filled and shaded circles indicate identical and similar residues, respectively. Boldface residues indicate the R-H-R-Y tetrad. Numbers between the sequences indicate actual spacings. The database sources for accession numbers are SwissProt (those starting with "P"), GenBank, and EMBL. Ref., reference.

Identification of the xis gene of phage 16-3. The system used to identify the xis gene consisted of two major components. One was R. meliloti 41 carrying the resident plasmid pSEM168, which can integrate specifically at the attB site of R. meliloti 41 due to the presence of the attP region and the active int gene, resulting in R. meliloti 41(pSEM168). In this bacterium the integration process converted the attB site to attR and attL. The advantage of our assay is that the extra chromosomal copies of the plasmid carrying attP do not interfere with monitoring of the attL (or attR)attB pathway (i.e., the reaction catalyzed by Xis). R. meliloti 41(pSEM168) was used as the recipient and conjugated with E. coli, which donated putative xis sequences (ORFs) (the second component of our assay system) carried by the plasmid derivatives of pSEM91. Different fragments were inserted into pSEM91 downstream of the tac promoter to identify the region which can provide xis function. The reappearance of attB (i.e., the activity of Xis) was indicated by the accumulation of a PCR-amplifiable fragment of 199 bp when an attB-specific primer pair was used. As expected, attB was detected in DNA derived from R. meliloti 41 but could not be seen in DNAs of strains that carried attL and attR sequences instead of attB, such as R. meliloti 41(pSEM168) and its derivatives into which plasmid pSEM91 or pSEM163 (both of which lack active Xis) was introduced by conjugation (Fig. 5, lanes 3 and 4). When pSEM161 was introduced into R. meliloti 41(pSEM168) cells by conjugation, the reappearance of the 199-bp fragment (representing attB) indicated the presence of plasmid-borne xis function (Fig. 5, lane 5).

View larger version (21K): [in this window] [in a new window]   FIG. 5.   (A) Schematic diagram of the assay used to identify the xis gene of phage 16-3. The presence of Xis protein expressed from the xis gene-containing derivative of pSEM91 results in the excision of pSEM168 from the R. meliloti 41 chromosome, regenerating the attB site from attL and attR, which can be identified by PCR with an attB-specific primer pair. A 23-kb fragment representing the integrated pSEM168 might have been amplified with the same primer pair, but the PCR conditions did not favor the accumulation of such a product. (B) Products of PCR amplification. Total DNA from R. meliloti 41 (Rm41) (lane 1), R. meliloti 41(pSEM168) (lane 2), R. meliloti 41(pSEM168) plus pSEM91 (lane 3), R. meliloti 41(pSEM168) plus pSEM163 (lane 4), and R. meliloti 41(pSEM168) plus pSEM161 (lane 5) were used for templates in attB-diagnostic PCR assays. M indicates the molecular size marker (AluI digest of pBluescript II KS).

The site-specific integrative system of 16-3 does not function in E. coli. The inability of pDH79 to integrate into the attB-carrying plasmid pGY1 (23) in E. coli suggested that the site-specific recombination system of phage 16-3 is inactive in E. coli. Inactivity might be explained by the inability to express the int gene due to a Rhizobium-specific promoter. This view was supported by the observation that the synthesis of the Int protein could not be demonstrated from pDH79 in E. coli. However, with pSEM167, due to the strong tac promoter it bears, expression of the Int protein in E. coli can be visualized on sodium dodecyl sulfate-polyacrylamide gel (data not shown). With plasmid pIP79 (39) as the attB target, formation of cointegrates between pSEM167 and pIP79 was detected in R. meliloti 41 but not in E. coli (Fig. 6), an apparent indication that the Int-catalyzed recombination between attP and attB takes place in R. meliloti 41 but not in E. coli. This result rules out the lack of expression of the int gene as the basis for our failure to observe integration in E. coli.

View larger version (45K): [in this window] [in a new window]   FIG. 6.   Detection of cointegrate formation between pSEM167 and pIP79. EcoRI digests of pSEM167 (lane 1), pIP79 (lane 2), and both plasmids derived from E. coli (lane 3) and R. meliloti 41 (Rm41) (lane 4) are presented. The  PstI digest served as the molecular size marker (lane M). The appearance of a 1,122-bp EcoRI fragment in lane 4 instead of the 276-bp EcoRI fragment of pIP79 in lanes 2 and 3 indicates cointegrate formation.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We have identified the int and xis genes of the temperate phage 16-3 of Rhizobium meliloti 41. The Int protein was classified as a member of the integrase family of tyrosine recombinases. The Int and Xis proteins consist of 371 and 140 residues, respectively. The two genes overlap over a 223-bp region; however, they are translated from different frames. In previous studies the two genes were not separated by independent mutations since the genetic analyses of the 16-3 site-specific recombination system were built on the deletion and insertion mutants tr4-2 and tr5-1 (transducing especially the cys-46 marker) (34). The overlapping topology of the int and xis genes is not unusual among the known site-specific recombination systems. The , P22 (20, 30), and pSAM2 (4) systems are examples. However, pSE211 (5) and L54a and 11 (56, 57) provide examples of nonoverlapping arrays.

Host factors may also be required for site-specific recombination. For phage  it was shown that IHF (integration host factor) is needed for recombination (32) but that FIS (factor for inversion stimulation) increases the efficiency of the process (52). The site-specific recombination system of 16-3 functions efficiently in R. meliloti 41 but not in E. coli. At least two simple explanations for these results can be put forward: either the 16-3 integration system requires a Rhizobium-specific host factor nonexistant in E. coli or the homologous host factors in the two species differ so much in structure or in DNA binding specificity that the site-specific recombination process cannot be cross-supported. In this sense, the 16-3 system differs from other site-specific recombination systems; for instance, systems of rather different origins, like those of CTX (Pseudomonas aeruginosa), AAU2 (Arthrobacter aureus), pSAM2 (Streptomyces ambofaciens), and pSE211 (Saccharopolyspora erythraea), are functional in vivo in E. coli (22, 29, 43, 54). Worth mentioning is that the target site (attB) of pSAM2 exhibits very extensive homology to the attB and attP sites of the 16-3 system (39).

Identification of the int and xis genes of 16-3 opens the field to their usage in biotechnological applications. Coupling of attP and int in plasmids creates a new class of vectors suitable for targeted gene insertions in microorganisms where compatible attB sites and the required host factors are available. The R. meliloti 41 attB site, which is the target of phage 16-3 integration, overlaps a tRNA^Pro gene (39). It can be expected that this target sequence may occur in many bacterial species because of the conservative nature of tRNAs. If the required host factor(s) can be supplied, the 16-3 integrative system can serve as a useful tool in gene technology.

We have constructed an expression vector called pSEM91 which can be used for functional analysis of different genes expressed not only in E. coli and R. meliloti but also in many other bacterial species. The useful host range of plasmid pSEM91 is determined by its different constituents. Among these elements the RP4 mob region has the widest range of hosts in which the plasmid can enter. Some of these potential hosts might not support the propagation of the plasmid, and in these cases the expression vector can be used as a suicide vector. If maintenance of the plasmid is required, the pCU1 replicon narrows the host range (25) within the set determined by RP4 mob. However, a drawback of a pSEM91 expression plasmid is that in some bacterial species the tac promoter may be repressed or may not function at all.

There have been several attempts to use the 16-3 integrative system for genetic modifications. Previously, we had developed a vector system containing the attP region of the 16-3 phage and the integrase function was provided in trans from helper phages (19). The disadvantage of that system was that it could be used only in strains within the host range of the helper phages. Progress has been achieved with a pRK290-derived plasmid carrying the attP region from phage 16-3 in combination with the integrase function provided in trans from a helper plasmid, pSEM102 (12). The weak link of this setup was that the function of xis was present, rendering the gene integrations unstable. This problem has now been eliminated by deleting the xis gene in plasmids pSEM167 and pSEM164; hence, they can be founders of a new integrative vector family based on the site-specific recombination system of phage 16-3.