identification of site-specific recombination genesint and xis of the

JOURNAL OF BACTERIOLOGY,0021-9193/99/$04.0010

July 1999, p. 4185–4192 Vol. 181, No. 14

Copyright © 1999, American Society for Microbiology. All Rights Reserved.

Identification of Site-Specific Recombination Genes int and xisof the Rhizobium Temperate Phage 16-3

SZABOLCS SEMSEY,1,2 ISTVAN PAPP,3 ZSUZSANNA BUZAS,4 ANDRAS PATTHY,4 LASZLO OROSZ,1,2

AND PETER P. PAPP1*

Institute for Molecular Genetics1 and Institute for Biochemistry and Protein Research,4 Agricultural BiotechnologyCenter, Godollo, H-2100 Hungary; Department of Biotechnology and Molecular Genetics, Godollo University of

Agricultural Sciences, Godollo, H-2105 Hungary2; and Max-Planck Institut fur Zuchtungsforschung,D-50829 Cologne, Germany3

Received 29 December 1998/Accepted 4 May 1999

Phage 16-3 is a temperate phage of Rhizobium meliloti 41 which integrates its genome with high efficiency intothe host chromosome by site-specific recombination through DNA sequences of attB and attP. Here we reportthe identification of two phage-encoded genes required for recombinations at these sites: int (phage integra-tion) and xis (prophage excision). We concluded that Int protein of phage 16-3 belongs to the integrase familyof tyrosine recombinases. Despite similarities to the cognate systems of the lambdoid phages, the 16-3 int xisatt system is not active in Escherichia coli, probably due to requirements for host factors that differ in Rhizobiummeliloti and E. coli. The application of the 16-3 site-specific recombination system in biotechnology is discussed.

A class of temperate phages are capable of integrating theirgenomes into the host chromosome by site-specific recombi-nation. The process is best known and has been described indetail for Escherichia coli phage l (28, 49, 55). However, thereare some other well-characterized integrative systems, such asthose of HP1, P22, L5, and pSAM2, where the array of thestructural elements differs from the arrangement known froml phage (4, 17, 42). Those studies indicate that although themolecular 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 oppor-tunity to reveal alternative pathways, widening our understand-ing of the mechanism of site-specific recombinations.

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

Phage 16-3 is able to integrate its genome into the chromo-some of Rhizobium meliloti 41, forming lyzogens (35–37). Thetarget site of this integration is between the cys-46 and met-5genes of R. meliloti 41, and cys-46 may undergo specializedtransduction with 16-3 (50). The 16-3 integrative recombina-tion system has been partially characterized (34). Both theattachment regions, attB and attP, were localized (8, 50) andtheir nucleotide sequences were determined (10). The attBregion contains a putative proline tRNA (tRNAPro) gene. Asequence of 51 bp, identical in the bacterial and the phage attregions overlapping the 39 end of the tRNAPro gene, was ex-pected to contain the core region where strand exchanges takeplace during the recombination process. This sequence alonewas sufficient to serve as a target site for phage integration.Due to the topology of the overlap, the nucleotide sequence ofthe tRNAPro gene is not altered by 16-3 integration. It wasfound that the putative tRNAPro gene of R. meliloti shows

significant homology to the putative tRNAPro gene of Strepto-myces ambofaciens, which serves as a target site for integrationof pSAM2, a self-transmissible plasmid carrying integrativeelements (39). The integrase (int) and excisionase (xis) func-tions of phage 16-3, previously localized on a 15-kb segment ofthe phage genome, are under the control of the C repressorprotein, the domain structure and DNA binding specificity(related to the coliphage 434 cI repressor) of which are alsoknown (6, 7, 11, 34, 35, 37, 38). Here we report the preciseidentification of the int and xis genes of phage 16-3. Amino acidsequence comparisons classify the 16-3 Int protein in the inte-grase family of tyrosine recombinases, analyzed recently inreferences 13 and 33. We found that the site-specific recom-bination system of phage 16-3 functions efficiently in R. meliloti41 but is inactive in E. coli. Since site-specific recombinationsystems deriving from different sources play major roles ingene technology (46, 53), the development of a new integrativevector family based on the site-specific recombination systemof phage 16-3 may serve as an appropriate tool in many appli-cations.

MATERIALS AND METHODS

Bacterial and phage strains, growth conditions, and triparental matings. E.coli DH5a (18) was used in all cloning experiments and served as the host ofdonor 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-specificrecombination. Growth conditions, media, and conditions for triparental matingswere as described in reference 39.

DNA procedures. Basic DNA manipulations and molecular techniques wereemployed as described in reference 44. Extraction of DNA from agarose gel wasdone with a QIAEX II Gel Extraction Kit (Qiagen). Total bacterial DNA wasprepared by the method described in reference 3. DNA was labelled by nicktranslation in the presence of [a-32P]dATP. Hybridization was performed asdescribed previously (48). PCR primers are listed in Table 1. PCR-mediatedDNA 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 minat 55°C, and 1 min at 72°C, the PCR products were extracted with phenol andprecipitated in ethanol. Then the DNAs were resuspended in Tris-EDTA bufferand digested with the appropriate restriction enzyme(s) to generate the requiredends of the fragments. The DNA fragments were purified before being cloned byisolating them from agarose gels. PCR mutagenesis was performed according tothe method in reference 27. Nucleotide sequence determination was performedby the dideoxy chain termination method (45) by using a TaqTrack SequencingKit (Promega). Total protein samples were analyzed on a discontinuous sodiumdodecyl sulfate-polyacrylamide gel electrophoresis system (26) and blotted to a

* Corresponding author. Mailing address: Institute for MolecularGenetics, Agricultural Biotechnology Center, Szent-Gyorgyi A. 4, Go-dollo, H-2100 Hungary. Phone: 36 (28) 430-600. Fax: 36 (28) 430-416.E-mail: [email protected].

4185

on March 31, 2018 by guest

http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

polyvinylidene difluoride membrane. The protein in the bands representing theprotein 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 bacterialDNA preparations from R. meliloti. Each 50-ml reaction mixture contained 100ng of the template and 30 pmol of each primer. PCR products were analyzed byelectrophoresis through either a 2% agarose gel or a 6% nondenaturing poly-acrylamide gel. Primer 1 and primer 2 (Table 1) were used to detect specificallyattL, which was indicated by the appearance of a 186-bp-long PCR product. Useof 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 ofthe Wisconsin Package, version 9.1 (Genetics Computer Group, Madison, Wis.).BLAST (41) and FASTA (2) were used to search for similarity with sequences inthe 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; onecontained the kanamycin resistance gene and the other carried the replicationregion of plasmid pCU1 (24). The resulting plasmid was linearized by PvuIIdigestion and ligated to the HindIII-SalI fragment (both ends were made bluntby end filling) derived from pSUP201 (47) containing an RP4 mob region. Theresulting plasmid was called pSEM64. The multicloning site (MCS) of the ex-pression vector pKK223-3 (Pharmacia) was altered, i.e., pKK223-3 was digestedwith EcoRI and the ends were filled in, and then the DNA was further digestedwith HindIII. The NotI (blunt ended)-HindIII fragment of the MCS from pBlue-script II KS (Stratagene) was inserted into the digested pKK223-3 vector. AHincII fragment of the resulting plasmid containing the tac promoter, the alteredMCS, and rrnB T1T2 transcription terminators was inserted into the uniqueEcoRI site of pSEM64, the ends of which were made blunt by end filling. Theresulting expression vector was called pSEM91 (Fig. 1).

Plasmid constructs used to analyze site-specific integration. A detailed phys-ical map of phage 16-3 facilitated building of the different plasmid constructs

listed in Fig. 2. Restriction sites with numbers in parentheses refer to physicalmap 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).

pSEM25 was created by deleting the KpnI(63)-KpnI(71) fragment frompDH79 (34). pSEM35 was constructed by inserting the isolated EcoRV(49)-KpnI(63) fragment of pDH79 into XbaI- and KpnI-digested pDH79, the XbaI-generated ends of which were filled in. To construct pSEM48, pSEM35 partiallydigested with PstI was religated and the recombinant plasmids were tested. InpSEM48 the PstI(51)-PstI(54) region (nucleotides [nt] 1212 to 3686) of pSEM35was deleted. pSEM62 was created by inserting the isolated EcoRV(49)-SalIfragment (nt 1 to 2413) of pSEM25 into XbaI- and XhoI-digested pSEM25. Theends generated by XbaI digestion of pSEM25 were filled in with Klenow enzymeprior to XhoI digestion. To construct pSEM102 (12), the EcoRV(49)-SalI (bluntended) fragment (nt 1 to 2413) of pSEM35 was inserted into the EcoRV site ofpBluescript II KS. In the resulting plasmid, pSEM80, the 59 end of the int geneis near the BamHI site of the MCS. PCR amplification of the int gene withprimer 4 (Table 1) and primer T7 (Stratagene) was performed, and the amplifiedfragment was digested with BamHI. The BamHI-generated ends were filled in,and the fragment was inserted into the EcoRV site of pSEM91. The correctDNA sequence of the region (nt 710 to 2413) was verified. The plasmid in whichthe orientation of the fragment allowed the transcription of the int gene wasdesignated pSEM102. pSEM164 was created by inserting the EcoRI(52)-KpnIfragment of pSEM102 into EcoRI- and KpnI-digested pSEM163 (see below).pSEM167 carries the StyI (blunt ended)-NaeI fragment (nt 242 to 1886) ofpSEM35 inserted into the EcoRV site of pSEM91 in the orientation such that theproduction of the Int protein can be driven by transcription from the tac pro-moter. The entire expression panel (tac promoter-int gene-terminators) wascloned by inserting the Acc65I-SalI fragment (both ends of which were madeblunt) of pSEM167 into the EcoRI-cut pLAFR1 vector, the ends of which werealso made blunt. The resulting plasmid was called pSEM168. The 16-3 phagecontent of pSEM223 is the same as that of pSEM167 except for a 4-bp deletionat the PstI(53) site. Construction of the plasmid required several steps. TheEcoRI(52)-NaeI fragment (nt 1558 to 1886) of pSEM35 was inserted into SmaI-and EcoRI-digested pBluescript II KS. In the resulting plasmid the PstI(53) sitewas eliminated by T4 polymerase (Stratagene) treatment following digestion ofthe plasmid DNA with PstI. The XbaI-EcoRI fragment from this plasmid wasused to replace the XbaI-EcoRI fragment of pSEM167.

Plasmid constructs used to study excision. To construct pSEM161, the StyIfragment (nt 242 to 2166) from pSEM35 was isolated and the ends were madeblunt (Fig. 2B). The fragment was digested with EcoRI and ligated to XbaI (bluntended)-EcoRI pSEM91 DNA. The recombinant plasmid containing the 619-bpregion from nt 1558 to 2166 was called pSEM161. pSEM163 contains theEcoRI(52)-HinfI fragment of 366 bp (nt 1558 to 1923); the HinfI fragment (nt1130 to 1923) was isolated from pSEM35 and the ends were made blunt. TheEcoRI digest of the fragment was ligated to XbaI (blunt ended)-EcoRI pSEM91DNA, and the recombinant plasmids were identified. To create pSEM208,primer 4 and primer 7 (Table 1) were used. The region containing open readingframe 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 to1946) was inserted into NcoI- and EcoRI-digested pET23d (Novagen). From theresulting plasmid the XbaI-EcoRI fragment carrying a ribosome binding site infront of ORF-111 was isolated and cloned into XbaI- and EcoRI-cut pSEM91.pSEM231 was built to express ORF-140. With pSEM25 template DNA andprimer 4 and primer 8 (Table 1), the region from nt 710 to 2046 was amplifiedby PCR. The product was digested with XbaI and EcoRI (nt 1558 to 2046) andinserted into XbaI- and EcoRI-cut pSEM91. pSEM249 derives from pSEM161.Its PstI(53) site was eliminated by T4 polymerase treatment following digestionof the plasmid DNA with PstI, and the treated DNA was self-ligated.

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

FIG. 1. Map of the pSEM91 expression vector constructed from the repliconof plasmid pCU1 (hatched bar), RP4 mob region (shaded bar), and kanamycinresistance gene (Km) (black arrow). The shaded triangle represents the tacpromoter, while the open bar shows the location of transcription terminators.Restriction sites in bold letters are unique.

TABLE 1. List of primers

Primer Sequencea nt positionb Use

1 GCAAGCTTGCGATAGGCGCTTGTGAAATC attL and attB detection2 gcgAATTCGACTAAAGCAAAAAGCTC 661 (2) attL detection3 GCGCGTCACCCGGCTGAG attB detection4 gatggatcCGGCAATGATTTACTTCT 728 (1) Cloning the int gene5 TGATGGTCCGGATGATGATGCCCaaATGTGCCTTCGAGCGTCTC 818 (1) Construction of intY346F6 AAACTGCGGGcTTTCTCGGAATGAC 824 (1) Construction of intY334F7 ggattcATGACAACGGCAGGGCTTA 1928 (2) Construction of ORF-1118 GATAAtctagaGGAGGTGGAGAAATG 2032 (2) Construction of ORF-140

a Restriction sites used in cloning procedures (BamHI, BspHI, and XbaI in primers 4, 7, and 8, respectively) are indicated with bold letters. Lowercase letters indicatebases not present in the original 16-3 sequence.

b Nucleotide positions refer to the base positions of the 39 bases of the primers. (1) and (2) indicate upper strand and lower strand, respectively.

4186 SEMSEY ET AL. J. BACTERIOL.


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

RESULTS

Construction of pSEM91 suitable for expressing genes bothin E. coli and in R. meliloti. Studying the site-specific recombi-nation system of phage 16-3 required the construction of aplasmid which allowed functional analyses of different frag-ments of phage origin in R. meliloti 41, the native host of 16-3phage. Basic components of the plasmid were selected to en-able us to examine whether this recombination system canfunction in E. coli and, in addition, to create a vector overex-pressing 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 fur-ther extended by inserting into it an expression panel whichcarried the tac promoter followed by MCS and transcriptionterminators. The resulting expression vector was calledpSEM91 (Fig. 1). Expression of different genes inserted intopSEM91 was constitutive in both E. coli DH5a and R. meliloti41 without the addition of any inducer, because neither straincontained the lac repressor gene.

Identification of the int gene of phage 16-3. Previous studiesindicated that the site-specific recombination function is lo-cated in a 15-kb region present in two different cosmid clones,pDH79 and pDH114, which are able to perform autonomous,

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 representthe 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 showsthe region of the determined sequence (GenBank accession no. AJ131679; the scale is given in base pairs). Open bars indicate deletions. Black arrows indicate the geneof the phage repressor (c). Stippled boxes represent the attachment region of the phage (attP). Restriction sites with numbers in parentheses refer to physical mappositions 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 thepLOL 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 candidatesfor encoding Int and Xis proteins (black arrows). Black bars indicate the regions carried by different plasmid constructs. Stippled boxes represent the attachment regionof 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 theshaded bar).

VOL. 181, 1999 SITE-SPECIFIC RECOMBINATION ELEMENTS OF PHAGE 16-3 4187


http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

prophage-like integration into and excision from the chromo-some of host R. meliloti 41 (34).

Cosmid pDH79 carries a 28-kb segment of the phage ge-nome (Fig. 2A). The region containing the elements requiredfor autonomous site-specific recombination was narrowed bydeletion derivatives of pDH79, pSEM25, and pSEM35. Thesequence of the EcoRV(49)-HindIII(55) region was deter-mined, and it was deposited together with the known sequenceof the HindIII(55)-EcoRI(61) region in GenBank.

Sequence analyses indicated the presence of a single ORFwith two possible start codons in the region (between the attPand c genes) where genetic analyses localized the determinantsof the int and xis functions, close to the attP region. The codingregions of the two possible transcripts were designated ORF-291 and ORF-371 (Fig. 2B). ORF-291 starts with an AUG,while ORF-371 starts with a GUG. Unlike ORF-291, ORF-371is preceded by a ribosome binding site.

We generated various plasmids in which sequences eitherneighboring or overlapping the ORFs were deleted (Fig. 2).pSEM62 was able to integrate specifically at the attB site intothe R. meliloti 41 chromosome, while pSEM48 was not. Theseresults indicated that the 1,201-bp region present in pSEM62but missing in pSEM48 was essential for the functioning of the16-3 site-specific recombination system.

To determine the length of the protein product and identifythe ORF representing the Int protein, pSEM167, in whichtranscription from the tac promoter allows the expression ofboth ORF-291 and ORF-371, was constructed. Only one pro-tein product was detected in E. coli (data not shown), and theamino acid sequence of its N terminus was determined. Theamino acid sequence corresponded to that of ORF-371.Hence, ORF-371 has become the candidate for the int gene ofphage 16-3.

ORF-371 codes for 16-3 integrase belonging to the tyrosinerecombinases. pSEM167 contains the expressible int gene andthe attP region of phage 16-3 and readily integrates into the R.meliloti 41 chromosome (Fig. 3, lanes 1). To verify that the intfunction is coupled to ORF-371 in R. meliloti, we constructedpSEM223, in which the translation frame of ORF-371 wasshifted by a 4-bp deletion while the frame of ORF-291 re-mained unharmed. Since this mutation abolished the ability ofthe attP-containing plasmid to be integrated, we concludedthat ORF-371 represents the int gene of phage 16-3.

It was shown that site-specific recombination also occurredwhen the Int protein was provided in trans to attP. Integrationof pSEM6 into the R. meliloti 41 chromosome could be de-tected only when pSEM164 was resident (Fig. 3, lanes 2).Neither pSEM6 nor pSEM91 alone was able to integrate intothe R. meliloti 41 genome (Fig. 3, lanes 3 and 4).

Comparisons of the amino acid sequence of 16-3 integrasewith sequences deposited in GenBank did not reveal signifi-cant homology with any known sequences. However, by inspec-tion of the 16-3 Int sequence, the R-H-R-Y tetrad (1), theconserved patterns (recognized from comparisons of manyknown integrases) can be found and the locations of the con-served residues fall into the intervals set by the known inte-grases. Figure 4 shows the conserved regions and their spacingsin seven matching integrases as well as their homologies to16-3 Int (i.e., ORF-371).

The catalytic tyrosines were localized in BoxC of the knownintegrases, and they are required to cleave the phosphodiesterbonds during strand exchanges. According to sequence align-ment, Tyr346 of the 16-3 Int protein was expected to have thesame function. To test the role of Tyr346, site-specific mutagen-esis was applied and an Int protein with Phe346 (IntY346F) wasconstructed. The one hydroxyl group difference between ty-

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) Theappearance 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) canbe 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. Theorder 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. Thepresence of attL and attR carrying EcoRI restriction fragments identified the site-specific recombination event (lanes 1 and 2). The attP-carrying fragments of pSEM6indicate extrachromosomal copies of the plasmid (lanes 2 and 3). With the construct pSEM167 (lane 1) the extrachromosomal copies of the plasmid were lost, probablydue 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 attRfragments 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. meliloti41.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

rosine and phenylalanine eliminated the integrase activity ofIntY346F. Tyr346 is the last tyrosine in the amino acid se-quence of the 16-3 Int protein. In contrast, when Tyr334, theclosest tyrosine to Tyr346, was similarly changed to Phe334, themutation had no detectable effect on the activity of 16-3 inte-grase.

Considering that Tyr346 is located in the 16-3 ortholog ofBoxC and that a mutation which affected only the hydroxylgroup of this moiety made the integrase inactive, we concludedthat Tyr346 is the catalytic tyrosine of the 16-3 integrase andhence that the 16-3 Int protein belongs to the integrase familyof tyrosine recombinases.

Identification of the xis gene of phage 16-3. The system usedto 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 41due to the presence of the attP region and the active int gene,resulting in R. meliloti 41(pSEM168). In this bacterium theintegration process converted the attB site to attR and attL.The advantage of our assay is that the extra chromosomalcopies of the plasmid carrying attP do not interfere with mon-itoring of the attL (or attR)3attB pathway (i.e., the reactioncatalyzed by Xis). R. meliloti 41(pSEM168) was used as therecipient and conjugated with E. coli, which donated putativexis sequences (ORFs) (the second component of our assaysystem) carried by the plasmid derivatives of pSEM91. Differ-ent fragments were inserted into pSEM91 downstream of thetac promoter to identify the region which can provide xis func-tion. The reappearance of attB (i.e., the activity of Xis) wasindicated by the accumulation of a PCR-amplifiable fragmentof 199 bp when an attB-specific primer pair was used. Asexpected, attB was detected in DNA derived from R. meliloti 41but could not be seen in DNAs of strains that carried attLand attR sequences instead of attB, such as R. meliloti41(pSEM168) and its derivatives into which plasmid pSEM91or pSEM163 (both of which lack active Xis) was introduced byconjugation (Fig. 5, lanes 3 and 4). When pSEM161 was in-troduced 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).

There are four possible AUG start codons for an ORFwithin the region carried by pSEM161. ORF-78, ORF-94,ORF-111, and ORF-140 (Fig. 2B) were the possible candidatesfor the encoding of Xis protein. ORF-78 and ORF-94 wereruled out by the lack of xis activity when pSEM163 was tested(Fig. 5, lane 4), while ORF-111 was ruled out by pSEM208.The xis function was identified when pSEM231 was introducedinto R. meliloti 41(pSEM168) (data not shown), suggesting thatthe protein encoded by ORF-140 is the Xis protein of phage16-3. This result was confirmed by the Xis2 phenotype ofpSEM249, which carried a frameshift mutation resulting in amutant protein of 108 residues.

The site-specific integrative system of 16-3 does not functionin 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 theint gene due to a Rhizobium-specific promoter. This view wassupported by the observation that the synthesis of the Intprotein could not be demonstrated from pDH79 in E. coli.However, with pSEM167, due to the strong tac promoter itbears, expression of the Int protein in E. coli can be visualizedon sodium dodecyl sulfate-polyacrylamide gel (data notshown). With plasmid pIP79 (39) as the attB target, formationof cointegrates between pSEM167 and pIP79 was detected inR. meliloti 41 but not in E. coli (Fig. 6), an apparent indicationthat the Int-catalyzed recombination between attP and attBtakes place in R. meliloti 41 but not in E. coli. This result rulesout the lack of expression of the int gene as the basis for ourfailure to observe integration in E. coli.

DISCUSSION

We have identified the int and xis genes of the temperatephage 16-3 of Rhizobium meliloti 41. The Int protein was clas-sified as a member of the integrase family of tyrosine recom-binases. The Int and Xis proteins consist of 371 and 140 resi-

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, andshaded boxes represent the locations of homology regions. The conserved residues and the ranges of spacings between them (as per reference 13) are indicated belowthe 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 shadedcircles 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.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

dues, respectively. The two genes overlap over a 223-bp region;however, they are translated from different frames. In previousstudies the two genes were not separated by independent mu-tations since the genetic analyses of the 16-3 site-specific re-combination system were built on the deletion and insertionmutants tr4-2 and tr5-1 (transducing especially the cys-46marker) (34). The overlapping topology of the int and xis genesis not unusual among the known site-specific recombinationsystems. The l, P22 (20, 30), and pSAM2 (4) systems areexamples. However, pSE211 (5) and L54a and f11 (56, 57)provide examples of nonoverlapping arrays.

Host factors may also be required for site-specific recombi-nation. For phage l it was shown that IHF (integration host

factor) is needed for recombination (32) but that FIS (factorfor inversion stimulation) increases the efficiency of the pro-cess (52). The site-specific recombination system of 16-3 func-tions efficiently in R. meliloti 41 but not in E. coli. At least twosimple explanations for these results can be put forward: eitherthe 16-3 integration system requires a Rhizobium-specific hostfactor nonexistant in E. coli or the homologous host factors inthe two species differ so much in structure or in DNA bindingspecificity that the site-specific recombination process cannotbe cross-supported. In this sense, the 16-3 system differs fromother site-specific recombination systems; for instance, systemsof rather different origins, like those of fCTX (Pseudomonasaeruginosa), fAAU2 (Arthrobacter aureus), pSAM2 (Strepto-myces ambofaciens), and pSE211 (Saccharopolyspora eryth-raea), are functional in vivo in E. coli (22, 29, 43, 54). Worthmentioning is that the target site (attB) of pSAM2 exhibits veryextensive 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 fieldto their usage in biotechnological applications. Coupling ofattP and int in plasmids creates a new class of vectors suitablefor targeted gene insertions in microorganisms where compat-ible attB sites and the required host factors are available. TheR. meliloti 41 attB site, which is the target of phage 16-3 inte-gration, overlaps a tRNAPro gene (39). It can be expected that

FIG. 5. (A) Schematic diagram of the assay used to identify the xis gene ofphage 16-3. The presence of Xis protein expressed from the xis gene-containingderivative of pSEM91 results in the excision of pSEM168 from the R. meliloti 41chromosome, regenerating the attB site from attL and attR, which can be iden-tified by PCR with an attB-specific primer pair. A 23-kb fragment representingthe 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 (lane3), R. meliloti 41(pSEM168) plus pSEM163 (lane 4), and R. meliloti 41(pSEM168) plus pSEM161 (lane 5) were used for templates in attB-diagnosticPCR assays. M indicates the molecular size marker (AluI digest of pBluescript IIKS).

FIG. 6. Detection of cointegrate formation between pSEM167 and pIP79.EcoRI digests of pSEM167 (lane 1), pIP79 (lane 2), and both plasmids derivedfrom E. coli (lane 3) and R. meliloti 41 (Rm41) (lane 4) are presented. The l PstIdigest served as the molecular size marker (lane M). The appearance of a1,122-bp EcoRI fragment in lane 4 instead of the 276-bp EcoRI fragment ofpIP79 in lanes 2 and 3 indicates cointegrate formation.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

this target sequence may occur in many bacterial species be-cause of the conservative nature of tRNAs. If the required hostfactor(s) can be supplied, the 16-3 integrative system can serveas a useful tool in gene technology.

We have constructed an expression vector called pSEM91which can be used for functional analysis of different genesexpressed not only in E. coli and R. meliloti but also in manyother bacterial species. The useful host range of plasmidpSEM91 is determined by its different constituents. Amongthese elements the RP4 mob region has the widest range ofhosts in which the plasmid can enter. Some of these potentialhosts might not support the propagation of the plasmid, and inthese cases the expression vector can be used as a suicidevector. If maintenance of the plasmid is required, the pCU1replicon narrows the host range (25) within the set determinedby RP4 mob. However, a drawback of a pSEM91 expressionplasmid is that in some bacterial species the tac promoter maybe repressed or may not function at all.

There have been several attempts to use the 16-3 integrativesystem for genetic modifications. Previously, we had developeda vector system containing the attP region of the 16-3 phageand the integrase function was provided in trans from helperphages (19). The disadvantage of that system was that it couldbe used only in strains within the host range of the helperphages. Progress has been achieved with a pRK290-derivedplasmid carrying the attP region from phage 16-3 in combina-tion with the integrase function provided in trans from a helperplasmid, pSEM102 (12). The weak link of this setup was thatthe function of xis was present, rendering the gene integrationsunstable. This problem has now been eliminated by deletingthe xis gene in plasmids pSEM167 and pSEM164; hence, theycan be founders of a new integrative vector family based on thesite-specific recombination system of phage 16-3.

ACKNOWLEDGMENTS

We thank Nelli S. Galne for excellent technical assistance, AndrasVaczi for help in plasmid characterization, and Tibor Sık and SusanGarges for discussion and helpful comments on the manuscript.

This work was supported by grants T 016092 and T 023695 from theHungarian Scientific Research Fund (OTKA), grant 0868/97 from theMKM Fund (FKFP), and grant 96-98 from the Academic Fund ofHungarian Academy of Sciences (MTA).

REFERENCES

1. Abremski, K. E., and R. H. Hoess. 1992. Evidence for a second conservedarginine residue in the integrase family of recombination proteins. ProteinEng. 5:87–91.

2. Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman. 1990.Basic local alignment search tool. J. Mol. Biol. 215:403–410.

3. Berman, M. L., L. W. Enquist, and T. J. Silhavy (ed.). 1982. Advencedbacterial genetics, p. 132. Cold Spring Harbor Laboratory, Cold SpringHarbor, N.Y.

4. Boccard, F., T. Smokvina, J.-L. Pernodet, A. Friedmann, and M. Guerineau.1989. The integrated conjugative plasmid pSAM2 of Streptomyces ambofa-ciens is related to temperate bacteriophages. EMBO J. 8:973–980.

5. Brown, D. P., K. B. Idler, and L. Katz. 1990. Characterization of the geneticelements required for site-specific integration of plasmid pSE211 in Saccha-ropolyspora erythrea. J. Bacteriol. 172:1877–1888.

6. Dallmann, G., P. Papp, and L. Orosz. 1987. Related specificity of unrelatedphages. Nature 330:398–401.

7. Dallmann, G., F. Marincs, P. Papp, M. Gaszner, and L. Orosz. 1991. Theisolated N-terminal DNA binding domain of the c repressor of bacterio-phage 16-3 is functional in DNA binding in vivo and in vitro. Mol. Gen.Genet. 227:106–112.

8. Dorgai, L., F. Olasz, M. Berenyi, G. Dallmann, A. Pay, and L. Orosz. 1981.Orientation of the genetic and physical map of Rhizobium meliloti temperatephage 16-3. Mol. Gen. Genet. 182:321–325.

9. Dorgai, L., G. Polner, E. Jonas, N. Garamszegi, Z. Ascher, A. Pay, G.Dallmann, and L. Orosz. 1983. The detailed physical map of the temperatephage 16-3 of Rhizobium meliloti. Mol. Gen. Genet. 191:430–433.

10. Dorgai, L., I. Papp, P. Papp, M. Kalman, and L. Orosz. 1993. Nucleotide

sequence of the sites involved in the integration of phage 16-3 of Rhizobiummeliloti 41. Nucleic Acids Res. 21:1671.

11. Dudas, B., and L. Orosz. 1980. Correlation between map position and phe-notype of Cti mutants in the C cistron of Rhizobium meliloti phage 16-3.Genetics 96:321–329.

12. Elo, P., S. Semsey, A. Kereszt, T. Nagy, P. Papp, and L. Orosz. 1998.Integrative promoter cloning plasmid vectors for Rhizobium meliloti. FEMSMicrobiol. Lett. 159:7–13.

13. Esposito, D., and J. J. Scocca. 1997. The integrase family of tyrosine recom-binases: evolution of a conserved active site domain. Nucleic Acids Res.25:3605–3614.

14. Freiberg, C., R. Fellay, A. Bairoch, W. J. Broughton, A. Rosenthal, and X.Perret. 1997. Molecular basis of symbiosis between Rhizobium and legumes.Nature 387:394–401.

15. Friedman, A., S. Long, S. Brown, W. Buikema, and F. Ausubel. 1982. Con-struction of a broad host range cosmid cloning vector and its use in thegenetic analysis of Rhizobium mutants. Gene 18:289–296.

16. Goodman, S. D., and J. J. Scocca. 1989. Nucleotide sequence and expressionof the gene for site-specific integration protein from bacteriophage HP1 ofHaemophilus influenzae. J. Bacteriol. 171:4232–4240.

17. Hakimi, J. M., and J. J. Scocca. 1994. Binding sites for bacteriophage HP1integrase on its DNA substrates. J. Biol. Chem. 269:21340–21345.

18. Hanahan, D. 1983. Studies on transformation of Escherichia coli with plas-mids. J. Mol. Biol. 166:557–580.

19. Hermesz, E., F. Olasz, L. Dorgai, and L. Orosz. 1992. Stable incorporationof genetic material into the chromosome of Rhizobium meliloti 41: construc-tion of an integrative vector system. Gene 119:9–15.

20. Hoess, R. H., C. Foeller, K. Bidwell, and A. Landy. 1980. Site-specific re-combination functions of bacteriophage l: DNA sequence of regulatoryregions and overlapping structural genes for Int and Xis. Proc. Natl. Acad.Sci. USA 77:2482–2486.

21. Hunkapiller, M. W., R. M. Hewick, W. J. Dreyer, and L. E. Hood. 1983.High-sensitivity sequencing with a gas-phase sequenator. Methods Enzymol.91:399–413.

22. Katz, L., D. P. Brown, and S. Donadio. 1991. Site-specific recombination inEscherichia coli between the att sites of plasmid pSE211 from Saccharopoly-spora erythraea. Mol. Gen. Genet. 227:155–159.

23. Kiss, G., K. Dobo, I. Dusha, A. Breznovits, L. Orosz, E. Vincze, and A.Kondorosi. 1980. Isolation and characterization of an R-prime plasmid fromRhizobium meliloti. J. Bacteriol. 141:121–128.

24. Kozlowski, M., V. Thatte, P. C. K. Lau, L. P. Visentin, and V. N. Iyer. 1987.Isolation and structure of the replicon of the promiscuous plasmid pCU1.Gene 58:217–228.

25. Krishnan, B. R., and V. N. Iyer. 1988. Host ranges of the IncN group plasmidpCU1 and its minireplicon in gram-negative purple bacteria. Appl. Environ.Microbiol. 54:2273–2276.

26. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly ofthe head of bacteriophage T4. Nature 227:680–685.

27. Landt, O., H.-P. Grunert, and U. Hahn. 1990. A general method for rapidsite-directed mutagenesis using the polymerase chain reaction. Gene 96:125–128.

28. Landy, A. 1989. Dynamic, structural, and regulatory aspects of l site-specificrecombination. Annu. Rev. Biochem. 58:913–949.

29. Le Marrec, C., S. Moreau, S. Loury, C. Blanco, and A. Trautwetter. 1996.Genetic characterization of site-specific integration function of fAAU2 in-fecting “Arthrobacter aureus” C70. J. Bacteriol. 178:1996–2004.

30. Leong, J. M., S. E. Nunes-Duby, A. B. Oser, C. F. Lesser, P. Youderian,M. M. Susskind, and A. Landy. 1986. Structural and regulatory divergenceamong site-specific recombination genes of lambdoid phage. J. Mol. Biol.189:603–616.

31. Lillehaug, D., and N.-K. Birkeland. 1993. Characterization of genetic ele-ments required for site-specific integration of the temperate lactococcalbacteriophage fLC3 and construction of integration-negative fLC3 mu-tants. J. Bacteriol. 175:1745–1755.

32. Miller, H. I., and D. I. Friedman. 1980. An E. coli gene product required forl site-specific recombination. Cell 20:711–719.

33. Nunes-Duby, S. E., H. J. Kwon, R. S. Tirumalai, T. Ellenberger, and A.Landy. 1998. Similarities and differences among 105 members of the Intfamily of site-specific recombinases. Nucleic Acids Res. 26:391–406.

34. Olasz, F., L. Dorgai, P. Papp, E. Hermesz, E. Kosa, and L. Orosz. 1985. Onthe site-specific recombination of phage 16-3 of Rhizobium meliloti: identi-fication of genetic elements and att recombinations. Mol. Gen. Genet. 201:289–295.

35. Orosz, L. 1980. Methods for analysis of the C cistron of temperate phage16-3 of Rhizobium meliloti. Genetics 94:265–276.

36. Orosz, L., and T. Sik. 1970. Genetic mapping of rhizobiophage 16-3. ActaMicrobiol. Acad. Sci. Hung. 17:185–194.

37. Orosz, L., Z. Svab, A. Kondorosi, and T. Sik. 1973. Genetic studies onrhizobiophage 16-3. I. Genes and functions on the chromosome. Mol. Gen.Genet. 125:341–350.

38. Orosz, L., K. Rostas, and R. Hotchkiss. 1980. A comparison of two-point,three-point and deletion mapping in the C cistron of rhizobiophage 16-3,



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

with an explanation for the recombination pattern. Genetics 94:249–263.39. Papp, I., L. Dorgai, P. Papp, E. Jonas, F. Olasz, and L. Orosz. 1993. The

bacterial attachment site of the temperate Rhizobium phage 16-3 overlapsthe 39 end of a putative proline tRNA gene. Mol. Gen. Genet. 240:258–264.

40. Papp, P. P., and V. N. Iyer. 1995. Determination of the binding sites ofRepA, a replication initiator protein of the basic replicon of the IncN groupplasmid pCU1. J. Mol. Biol. 246:595–608.

41. Pearson, W. R., and D. J. Lipman. 1988. Improved tools for biologicalsequence comparison. Proc. Natl. Acad. Sci. USA 85:2444–2448.

42. Pena, C. E. A., M. H. Lee, M. L. Pedulla, and G. F. Hatfull. 1997. Charac-terization of the mycobacteriophage L5 attachment site, attP. J. Mol. Biol.266:76–92.

43. Raynal, A., K. Tuphile, C. Gerbaud, T. Luther, M. Guerineau, and J.-L.Pernodet. 1998. Structure of the chromosomal insertion site for pSAM2:functional analysis in Escherichia coli. Mol. Microbiol. 28:333–342.

44. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold SpringHarbor, N.Y.

45. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing withchain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 81:654–658.

46. Sezonov, G., V. Blanc, N. Bamas-Jacques, A. Friedmann, J.-L. Pernodet, andM. Guerineau. 1997. Complete conversion of antibiotic precursor to pristi-namycin IIA by overexpression of Streptomyces pristinaespiralis biosyntheticgenes. Nat. Biotechnol. 15:349–353.

47. Simon, R., U. B. Priefer, and A. Puhler. 1982. A broad host range mobili-zation system for in vivo genetic engineering: transposon mutagenesis ingram-negative bacteria. Biotechnology 1:784–791.

48. Southern, E. M. 1975. Detection of specific sequencing among DNA frag-

ments separated by gel electrophoresis. J. Mol. Biol. 98:503–517.49. Stark, W. M., M. R. Boocock, and D. J. Sherratt. 1992. Catalysis by site-

specific recombinases. Trends Genet. 8:432–439.50. Svab, Z., A. Kondorosi, and L. Orosz. 1978. Specialised transduction of a

cystein marker by Rhizobium meliloti phage 16-3. J. Gen. Microbiol. 106:321–327.

51. Szende, K., and F. Ordogh. 1960. Die Lysogenie von Rhizobium meliloti.Naturwissenschaften 47:404–405.

52. Thompson, J. F., L. M. de Vargas, C. Koch, R. Kahmann, and A. Landy.1987. Cellular factors couple recombination with growth phase: character-ization of a new component in the l site-specific recombination pathway.Cell 50:901–908.

53. Vergunst, A. C., L. E. Jansen, and P. J. Hooykaas. 1998. Site-specific inte-gration of Agrobacterium T-DNA in Arabidopsis thaliana mediated by Crerecombinase. Nucleic Acids Res. 26:2729–2734.

54. Wang, Z., G. Xiong, and F. Lutz. 1995. Site-specific integration of phagefCTX genome into the Pseudomonas aeruginosa chromosome: characteriza-tion of the functional integrase gene located close to and upstream of attP.Mol. Gen. Genet. 246:72–79.

55. Weisberg, R., and A. Landy. 1983. Site specific recombination in phage l, p.211–250. In R. Hendrix, J. Roberts, F. Stahl, and R. Weisberg (ed.), LambdaII. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

56. Ye, Z.-H., and C. Y. Lee. 1989. Nucleotide sequence and genetic character-ization of staphylococcal bacteriophages L54a int and xis genes. J. Bacteriol.171:4146–4153.

57. Ye, Z.-H., S. L. Buranen, and C. Y. Lee. 1990. Sequence analysis and com-parison of int and xis genes from the staphylococcal bacteriophages L54a andf11. J. Bacteriol. 172:2568–2575.



http://jb.asm.org/

Dow

nloaded from

http://jb.asm.org/

identification of site-specific recombination genesint and xis of the

Documents