Journal of Bacteriology, February 2003, p . 809-822, Vol . 185, No . 3

Quorum Sensing in Rhizobium sp . Strain NGR234 Regulates Conjugal Transfer (tra) Gene Expression and Influences Growth Rate

Xuesong He, William Chang, Deanne L . Pierce, Laura Ort Seib, Jennifer Wagner, and Clay Fuqua^*

Department of Biology, Indiana University, Bloomington, Indiana 47405

Received 26 July 2002/ Accepted 31 October 2002

Several different genera of the rhizobia, members of the family Rhizobiaceae, cause the formation of symbiotic, nitrogen-fixing nodules on leguminous plants . The complex developmental process of nodule formation involves sequential bacterial-plant signal exchange, requiring specific nodulation (nod) and nitrogen fixation (nif and fix) genes (see references 25, 38, and 50 for extensive reviews) . Among the Rhizobiaceae, acyl-HSL quorum sensing is quite prevalent . The best studied of these systems in a member of the Rhizobiaceae is found in the plant pathogen Agrobacterium tumefaciens, which causes crown gall disease (71) . The A . tumefaciens quorum sensor requires the acyl-HSL synthase TraI, which directs synthesis of N-3-oxo-octanoyl-L-homoserine lactone (3-oxo-C8-HSL) . The acyl-HSL-responsive regulator TraR associates with 3-oxo-C8-HSL, binds to DNA sequences called tra boxes upstream of regulated promoters, and activates transcription (Fig . 1A) (19, 20, 29, 51, 73) . A third regulatory protein, TraM, functions as an antiactivator, preventing TraR from activating target genes under noninducing conditions (13, 28, 62) . TraR, TraI, and TraM are encoded on the tumor-inducing (Ti) virulence plasmid and regulate the conjugal transfer (tra) and vegetative replication (rep) functions of this plasmid (19, 29, 36) .

 
Acyl-HSL quorum sensing in the symbiotic rhizobia is generally more complex than in A . tumefaciens . There are often multiple LuxR-type and LuxI-type proteins active within the same strain, organized in a regulatory hierarchy . Rhizobium leguminosarum bv . viciae synthesizes four to six different acyl-HSLs via the activity of four LuxI-type proteins and has at least five LuxR-type proteins (6, 37) . Rhizobium etli CNPAF512 and Sinorhizobium meliloti AK631 synthesize seven and nine different acyl-HSLs, respectively, and encode multiple LuxR-LuxI type regulators (2, 7, 42, 56) . In R . leguminosarum the quorum sensors are arranged in a hierarchal network that coordinates conjugal transfer gene expression on the pRL1JI symbiotic plasmid, rhizosphere-specific gene expression, and other, as-yet-undefined functions (69) . Rhizobial mutants deficient in quorum sensing exhibit a variety of defects during the symbiotic nodulation of plants, ranging from modest to severe (7) .

The wide-host-range nodulator Rhizobium sp . strain NGR234 can nodulate over 120 different legume species and at least one nonlegume (54) . The NGR234 genome is comprised of two large episomes, pNGR234a (536,165 bp) and pNGR234b (approximately 2 Mb) and a chromosome (approximately 3.5 Mb), all of which can exist independently or as cointegrates in every possible combination (11, 44) . The majority of symbiotic functions are encoded on pNGR234a . The complete pNGR234a sequence revealed the presence of (i) nod, fix, and nif genes that direct the plant interaction; (ii) genes with presumptive functions in the symbiotic process; and (iii) a cluster of genes homologous to Ti plasmid rep and tra genes of A . tumefaciens (12) . The pNGR234a tra-rep cluster is homologous and syntenous to its counterparts on the Ti plasmids over a discrete region including a pair of divergent operons encoding predicted DNA transfer and replication (Dtr) functions (including the origin of transfer, oriT) and a separate but linked region with a pair of divergent operons encoding predicted rep and mating pair formation (Mpf) functions (Fig . 1B) . Regulatory genes within this region are also conserved with those on the A . tumefaciens Ti plasmids, including the traR, traI, and traM regulatory genes .

The pNGR234a quorum-sensing mechanism was investigated in this study . Although there are interesting differences, the pNGR234a quorum-sensing mechanism is similar to that of A . tumefaciens Ti plasmids, and expression of traR limits expression of pNGR234a tra genes . Additionally, the pNGR234a-encoded quorum sensor significantly influences the growth rate of NGR234, providing a mechanism by which quorum sensing and population growth may be integrated .

 
NGR234 and derivatives were cultivated at 28°C in TY complex medium (5 g of Bacto Tryptone per liter, 3 g of yeast extract per liter, 1.2 g of CaCl₂ · 2H₂O per liter [pH 6.9]) and a modified M9 minimal salts medium (Rhizobium minimal medium [RMM]) supplemented with 15 mM glucose and with thiamine, calcium pantothenate, and biotin at 1 mg/ml each (3) . A . tumefaciens strains were grown at 30°C in either Luria-Bertani medium or AT minimal salts medium with 0.5% glucose and 15 mM (NH₄)₂SO₄ (ATGN) (63) . Antibiotics were used at the following concentrations (micrograms per milliliter): (i) for Escherichia coli, ampicillin (100), gentamicin (25), kanamycin (25), spectinomycin (100), and tetracycline (10); (ii) for A . tumefaciens, gentamicin (800), spectinomycin (50), rifampin (50), streptomycin (1,000), and tetracycline (5); and (iii) for NGR234, gentamicin (50), kanamycin (25), rifampin (150), streptomycin (1,000), spectinomycin (25), and tetracycline (10) .

Plasmids were electroporated into A . tumefaciens by using a standard method (46) . Introduction of plasmids into NGR234 was performed by conjugation or by electroporation with the A . tumefaciens protocol modified by washing the cells in 10% glycerol rather than water . Following electroporation, the NGR234 cell suspension was diluted in TY medium and usually incubated for 1 to 2 h at 28°C prior to plating on selective medium .

Isolation and manipulation of NGR234 quorum-sensing genes. The complete coding sequences of traR and traM and several segments of traI from pNGR234a were isolated by PCR with NGR234 genomic DNA and primers designed by using the pNGR234a DNA sequence (12) .

To generate a broad-host-range (BHR) derivative expressing the pNGR234a traR gene from the P_lac promoter on the pBBR1-MCS5 vector, the traR coding sequence was amplified from total genomic DNA by PCR with two oligonucleotides, 5'-cgcggtaaccctgataaCGGCAAAAGATAGGAGAGATGTCC and 5'-gcgtctagaTCAGACCAGGCCGCGATCCTTG (for all oligonucleotides, the pNGR234a sequence is in uppercase) . The first oligonucleotide incorporates (i) a KpnI site (underlined) upstream of the coding sequence, (ii) a tandem pair of stop codons (in boldface) to prevent readthrough from the lacZ peptide of pBBR1-MCS5, and (iii) the first two traR codons plus associated translational signals (in boldface) . The second, downstream oligonucleotide includes the native stop codon (in boldface) flanked by an engineered XbaI site (underlined) . PCR amplification generated a 755-bp product that was purified, digested with KpnI and XbaI, and ligated with similarly digested pBBR1-MCS5 to generate pBDJ6 . The traR coding sequence of the pBDJ6 fusion construct was verified as correct by DNA sequencing .

The traM gene from pNGR234a was PCR amplified from genomic DNA by using oligonucleotides 5'-ggcggtaccctgataaacaggaaacagctATGAACGATATGGGCTCATCCGAGG and 5'-ggcctgcagCTAGTTGTCATCGTCGAAAGGC . The first oligonucleotide carries an engineered KpnI site (underlined), followed by two stop codons (in boldface), the Shine-Dalgarno sequence from the Escherichia coli lacZ gene (underlined), and the first seven codons of traM . The second oligonucleotide includes the native stop codon (boldface) flanked with a PstI site (underlined) . The purified 361-bp amplification product was cloned by using the PCR-Script Amp cloning system (Stratagene Corp.), blunt-end ligating the product with linearized pCR-Script vector to generate pCF422 . The traM insert in pCF422 was sequenced and found to be identical to the published sequence (12) . pCF422 was digested with KpnI and PstI and ligated with similarly digested pBBR1-MCS5 to produce pCF424, a BHR derivative with a P_lac-traM fusion . To generate a single plasmid carrying traR and traM from pNGR234a, pCF422 was linearized at its single EagI site, the 5' extensions were filled in by using the Klenow fragment of DNA polymerase I, and the entire plasmid was ligated into the SmaI site of pBDJ6 immediately downstream of the traR coding sequence . This BHR plasmid, designated pDP101, expresses a P_lac-traR-traM fusion . Lastly, to generate a recombinational insertion plasmid, the traM insert in pCF422 was excised as a KpnI fragment and ligated with KpnI-cleaved pVIK112, an R6K-derived suicide vector that can be maintained only in E . coli hosts that express the trans-acting R6K replication factor gene pir (31), to generate pCF429 .

The traI promoter region from pNGR234a was isolated by using oligonucleotides 5'-ccggaattcTGAATATATTCGAGTTCTGAGTTGC and 5'-ggctctagaCTGCATGAATTTCTCCGTCGTTGTTGG . The first oligonucleotide contains an EcoRI site (underlined) and is positioned 495 bp upstream of the predicted traI start codon . The second oligonucleotide carries the native translational signals of traI (in boldface) flanked by an XbaI cleavage site (underlined) . The 513-bp amplification product was purified, digested with EcoRI and XbaI, and ligated with similarly digested pRA301, an IncP BHR plasmid, generating a BHR₃ PtraI-lacZ fusion plasmid designated pBDJ3 (1) . In order to test the role of the presumptive tra box proximal to traI, pBDJ3 was cleaved at its single BamHI site located in the exact center of the tra box (Fig . 1B), the 5' overhangs were filled in with the Klenow fragment of DNA polymerase I, and the plasmid was recircularized to generate pBDJ7 . Two fragments of the traI gene were PCR amplified from NGR234 genomic DNA for recombinational mutagenesis . An internal fragment truncated on both ends of the traI coding sequence was isolated by using oligonucleotides 5'-cgctctagaTTTGCCGGCAGAGTGCCCGC and 5'-cgcgaattcGAAGTCAATGTCGTCGGAGG . The first oligonucleotide anneals upstream of the predicted traI stop codon and is flanked by an XbaI site (underlined) . The second oligonucleotide contains an EcoRI site (underlined) and anneals 105 bp downstream of the predicted traI start codon . The predicted PCR product is 464 bp . A larger fragment with all of the traI upstream sequence, including the promoter region, but truncated in the carboxy-terminal coding region was PCR amplified by using the oligonucleotide 5'-ccggaattcTGAATATATTCGAGTTCTGAGTTGC and the same second oligonucleotide as used for the doubly truncated product . The first oligonucleotide for this larger product also has an EcoRI site (underlined) but is positioned 495 bp upstream of the predicted traI start codon . The expected product of this fragment is 1,064 bp . The two amplification products were purified, digested with EcoRI and XbaI, and ligated with similarly digested pVIK112 integration vector, generating pXH100 and pXH101 . These plasmids also create traI::lacZ transcriptional fusions at the identical position . The same PCR fragments used to generate pXH100 and pXH101 were also used to create the pBSK(+) derivatives pXH102 and pXH103, respectively .

Recombinational integration of plasmids into pNGR234a. To integrate plasmids into pNGR234a, the plasmids carrying segments of pNGR234a were conjugated from the BHR conjugal donor E . coli S17-1/pir into NGR234-ACF1, a spontaneous Rif^r derivative of NGR234 . Conjugal matings were performed by mixing concentrated suspensions of the donor and recipient strains, spotting onto a 0.2-µm-pore-size cellulose acetate filter on a TY agar plate, and incubating for 18 h at 28°C . Following incubation, cells were removed from the filter and plated in dilution series on RMM plates supplemented with rifampin and kanamycin . Prototrophic Rif^r Km^r transconjugants were streak purified, and the integration of the plasmid into pNGR234a was verified by PCR with primers flanking the site of insertion .

Conjugal transfer assays. NGR234 with pCF429 carrying a Km^r marker integrated into the traM gene of pNGR234a was used as a conjugal donor . A plasmid expressing lacI^q, pSW213, was introduced into this background to allow subsequent introduction of the P_lac-traR plasmid pBDJ6 (see Results) . The recipient was NGR234-ANU265, a derivative cured for pNGR234a . Likewise, A . tumefaciens C58 C1RS was also employed as a heterologous recipient . Cultures of donors and recipients were grown to an optical density at 600 nm (OD₆₀₀) of 1.0 . Cells were collected by centrifugation at 18,000 x g and concentrated 20- to 100-fold in RMM . The donor cell suspensions were mixed with an equal volume of recipient cells, spotted onto 0.2-µm-pore-size cellulose acetate filters on unsupplemented RMM agar and on RMM agar supplemented with 0.1 mM IPTG (isopropyl-ß-D-thiogalactopyranoside), and incubated at 28°C for 18 h . Cells were resuspended, and viable counts were performed for (i) transconjugants (Km^r Sm^r transconjugants for NGR234-ANU265 recipients and Km^r Sm^r Rif^r transconjugants for A . tumefaciens C58 C1RS recipients) and (ii) Km^r conjugal donors . Conjugation efficiencies were calculated as the ratio of transconjugants per recovered donor . Donor and transconjugant identities were further verified by randomly amplified polymorphic DNA PCR analysis with the short primers 5'-GTAGACCCGT and 5'-GGTGCGGGAA to generate genomic fingerprints (70) .

Preparation of NGR234 culture extracts and bioassays for acyl HSLs. Crude preparations of acyl-HSLs were generated from whole cultures as described by Marketon et al . (43; M . M . Marketon and J . E . González, personal communication) . NGR234 derivatives were cultivated to an OD₆₀₀ of 1.5 in 5 ml of TY medium . An equal volume of dichloromethane was added to the cultures, and the mixtures were vortexed for 10 min with occasional venting . The vortexed extractions were centrifuged at 12,000 x g for 5 min to separate the aqueous and organic phases . The dichloromethane phase was collected and evaporated in a fume hood . The dried extract was dissolved in 200 µl of fresh dichloromethane . A control with autoclaved TY medium alone was performed in parallel .

Concentrated dichloromethane extracts (10 µl) were fractionated by reverse-phase thin-layer chromatography (TLC) with C₁₈ silica plates (J.T . Baker, Phillipsburg, N.J.) in a mobile phase of 60% methanol as described previously (63) . Following the TLC fractionation, the plates were air dried and acyl-HSLs were detected by overlaying with soft agar containing either Chromobacterium violaceum CV026 or A . tumefaciens(pCF218)(pCF372), both of which cannot produce but respond to acyl-HSLs (17, 45) . A . tumefaciens(pCF218)(pCF372) activates expression of a traI-lacZ fusion (derived from the A . tumefaciens Ti plasmid traI gene and activated by the Ti plasmid TraR) in the presence of exogenous acyl-HSLs of variable chemistries ranging from C₆ to C₁₂ in acyl chain length (74) . Provision of the chromogenic substrate X-Gal (5-bromo-4-chloro-3-indolyl-ß-D-galactopyranoside) in the overlay agar allows visualization of acyl-HSLs as spots of blue coloration after overnight incubation at 28°C . Likewise, C . violaceum CV026 cultured in Luria-Bertani agar produces the pigment violacein in response to acyl-HSLs ranging from C₄ to C₆, generating similar spots over the areas of the TLC plate with acyl-HSL fractions after incubation at 37°C (45) . Standards of organically synthesized acyl-HSLs were fractionated alongside the extracts .

RNA isolation and primer extension assays. Total RNA was isolated from NGR234 cultures grown in RMM by using the technique of Summers (61) . Primer extension reactions were performed with an estimated 30 µg of RNA, using avian myeloblastosis virus reverse transcriptase and [-³²P]-end-labeled oligonucleotides as described previously (17) . The oligonucleotides 5'-TGGGTTCGTCGGGAGCGGTGG, 5'-GAGGCAGGTTTCTTCATTCCG, and 5'-ATGGCTGCGGAGAAGTTGGG were used for traA, traC, and traI transcripts, respectively . An oligonucleotide complementary to the tetA transcript encoded on pSW213 harbored in all of the NGR234 strains tested, 5'-CGATCAGGGGTATGTTGGGT, was used for control reactions . Reaction products were fractionated on 6% polyacrylamide denaturing gels containing 7 M urea, and dried gels were analyzed with a PhosphorImager (Molecular Dynamics, Sunnyvale, Calif.) . A DNA sequencing ladder was obtained with the traI-specific oligonucleotide and pXH103 carrying the traI promoter region, using -³⁵S-dATP, Sequenase enzyme, and the reaction conditions recommended by the enzyme supplier (United States Biochemical Corp.) .

To examine the acyl-HSLs produced by NGR234, dichloromethane extracts were prepared from whole NGR234 cultures grown in TY medium . Aliquots of the extracts were fractionated on reverse-phase TLC, and acyl HSLs were detected by overlaying the plates with acyl-HSL-responsive reporter strains . A single spot of activity, which comigrated with the synthetic 3-oxo-C8-HSL standard and also gave a typical tadpole conformation of the 3-oxo-acyl-HSLs, was detected using the A . tumefaciens reporter strain (Fig . 2, lane 2) (59) . An additional faint, slowly migrating spot of activity was observed very close to the point of sample application . A very weak, rapidly migrating spot was observed in all samples, including the medium-only control (Fig . 2, lane 7) . No activity was detected using the C . violaceum reporter with either the direct activation or inhibition assay (data not shown) (assay described in reference 45) .

 
The traR gene was PCR amplified from NGR234 genomic DNA and used to construct a BHR P_lac-traR fusion plasmid (pBDJ6) . Likewise, the traI promoter region was isolated and fused to lacZ carried on a separate, compatible BHR plasmid (pBDJ3) . These two constructs were introduced into A . tumefaciens NTL4, a strain cured of the Ti plasmid and hence lacking the Ti-encoded quorum-sensing regulators . The addition of cell-free culture fluids containing 3-oxo-C8-HSL (from an overproducing A . tumefaciens derivative) induced traI-lacZ expression by over 3,000-fold in the A . tumefaciens derivative harboring P_lac-traR (Table 2) . Expression in the absence of the acyl-HSL or the traR-expressing plasmid remained very low . Dose-response experiments comparing NGR234 traI-lacZ expression activated by NGR234 traR in the A . tumefaciens NTL4 background revealed that the NGR234-based system was very sensitive to 3-oxo-C8-HSL and was activated to high levels of expression (Fig . 3) .

 
TraM inhibits TraR-dependent transcriptional activation. In A . tumefaciens the antiactivator TraM associates with TraR to form an inactive complex (30, 40, 62) . Artificially elevated expression of traM results in a reduction of TraR-dependent activation of target promoters (13, 28) . To determine whether the traM gene carried on pNGR234a has a similar inhibitory function, pDP101, in which traR and traM are expressed in tandem under the control of the P_lac promoter (P_lac-traR-traM), was constructed . An A . tumefaciens NTL4 derivative harboring the P_lac-traR-traM pDP101 plasmid demonstrated a 50-fold reduction in AAI-dependent induction of the traI-lacZ fusion relative to pBDJ6 expressing traR alone (Table 2) .

Ectopic expression of traR elevates traI expression on pNGR234a. To evaluate the expression of the traI-lacZ fusion in its parental background, pBDJ3 was introduced into NGR234 . Only weak constitutive expression which was unaffected by the addition of saturating amounts of AAI was detectable (data not shown) . In A . tumefaciens a similar pattern of expression is observed for TraR target genes due to tight regulation of traR expression by crown gall tumor-released opines (Fig . 1A) (18, 52) . Therefore, it seemed likely that the activation of the plasmid-borne traI-lacZ fusion was similarly limited by the expression level of the native copy of traR on pNGR234a .

To test this hypothesis, a plasmid carrying an internal fragment of the traI coding sequence (pXH100) fused to a promoterless lacZ (traI::lacZ) was integrated by single crossover into pNGR234a to generate NGR234-XH100, resulting in a traI null mutation . Likewise, a plasmid carrying a larger fragment of traI with a complete 5' end (pXH101) but fused to lacZ at the same point as in pXH100 was used to generate NGR234-XH101, retaining an intact, functional copy of traI . NGR234-XH100 expressed the integrated traI::lacZ weakly, and this was unaffected by addition of saturating AAI (Table 3) . Surprisingly, introduction of the P_lac-traR plasmid pBDJ6 into NGR234-XH100 resulted in modest activation of the traI::lacZ fusion (basal level of 2 U elevated to 55 U) . Addition of 0.1% (vol/vol) AAI to this culture resulted in strong induction of the traI::lacZ fusion . The P_lac-traR can be introduced stably into the NGR234-XH101 derivative (retaining the functional copy of traI) only when the strain harbors a lacI^q repressor-expressing plasmid, pSW213 (lacI^q is required to obtain transformants with the P_lac-traR plasmid in most NGR234 derivatives [see below]) . The presence of lacI^q also provides for regulated expression of traR through IPTG . Introduction of the P_lac-traR plasmid into NGR234-XH101(pSW213) in the absence of IPTG resulted in very weak elevation of traI::lacZ expression, which was only marginally enhanced by addition of exogenous AAI (Table 3) . However, addition of IPTG alone strongly induced traI::lacZ expression, and this was enhanced by exogenous AAI . These observations suggest that traR expression is limiting in wild-type NGR234 and in the strain NGR234-XH100, in which lacI^q represses expression of the pBDJ6-borne P_lac-traR .

 
Analysis of culture extracts reveals additional acyl-HSLs in NGR234. The presence of the P_lac-traR plasmid strongly elevated expression of the pNGR234a traI gene . Consistent with this observation, the presence of the P_lac-traR plasmid in an otherwise wild-type NGR234 harboring the lacI^q plasmid pSW213 and induced with IPTG resulted in strong overproduction of an acyl-HSL activity as assessed by cross-feeding assays with the acyl-HSL-responsive A . tumefaciens reporter strain (data not shown) . TLC bioassay analysis of dichloromethane extracts from cultures of the NGR234 strain revealed overproduction of a compound that comigrated with the 3-oxo-C8-HSL standard and with the compound from untransformed NGR234 (Fig . 2, lanes 1 and 2 [the sample in lane 1 was 200-fold more dilute than the other samples]) . In addition, a spot of activity that did not migrate far from the point of sample application was readily visible . Surprisingly, extracts from the traI null mutant (NGR234-XH100), although significantly reduced for activity, possessed two weak inducing activities, one that comigrated with the 3-oxo-C8-HSL standard and a second that migrated only slightly from the application site (Fig . 2, lane 3) . Extracts from NGR234-XH100 harboring the P_lac-traR plasmid contained these two activities, both notably stronger than that from NGR234-XH100 alone (Fig . 2, lanes 3 and 4) . This observation suggested that in addition to the signal molecules specified by traI carried on pNGR234a, a second acyl-HSL activity that comigrates with 3-oxo-C8-HSL is synthesized by NGR234 . Consistent with this, extracts from NGR234-ANU265, a strain that is cured of pNGR234a, contained a barely detectable activity comigrating with 3-oxo-C8-HSL . The slowly migrating activity appeared to be slightly more abundant in NGR234-ANU265 than in strains harboring pNGR234a (Fig . 2, lane 5) . Introduction of the P_lac-traR into the NGR234-ANU265 strain resulted in extracts that were again significantly elevated for both activities (Fig . 2, lane 6) .

TraR activates pNGR234a promoters associated with putative tra box-type elements. Our findings suggested that NGR234 P_traI is regulated by its cognate TraR and acyl-HSL . Careful inspection of the NGR234 sequence revealed potential regulatory sites in positions highly analogous to those of the Ti plasmid tra boxes (Fig . 1B) . Directly upstream of the NGR234 traI translational start site is an 18-bp inverted repeat, tentatively designated NGR tra box II, with similarity to tra box elements identified in A . tumefaciens (Fig . 1B) . A sequence 542 bp further upstream of traI has significant similarity to NGR tra box II and is tentatively designated NGR tra box III . Within the pNGR234a traA-traC intergenic region in the Dtr-oriT cluster, a single possible tra box promoter element, presumptively designated NGR tra box I, is positioned identically to tra box I on the A . tumefaciens Ti plasmids (17) . The primary sequence conservation between the presumptive pNGR234a sequence elements and the Ti plasmid tra boxes is modest, and there is surprising sequence degeneracy among the NGR234 elements (averaging 55% identity) compared to the similarity between Ti plasmid tra boxes (averaging 88%) .

A derivative of pBDJ3 (carrying the traI-lacZ fusion) designated pBDJ7 carries a 4-bp insertion in the exact center of the presumptive NGR tra box II sequence (underlined in Fig . 1B) . A . tumefaciens NTL4 harboring pBDJ7 and the P_lac-traR plasmid pBDJ6 was 500-fold reduced for traI-lacZ expression in the presence of 3-oxo-C8-HSL relative to pBDJ3 with the intact tra box (Table 2) .

Primer extension assays to detect the presumptive transcripts initiating from P_traI, P_traA, and P_traC promoters were performed on total RNA preparations of NGR234 harboring the pSW213 plasmid to allow lacI^q-dependent control of P_lac and either pBDJ6 (P_lac-traR) or the pBBR1-MCS5 vector control . Reverse transcripts specified by the traI and traC promoters were detected in the RNA preparation from NGR234(pBDJ6)(pSW213) grown in the presence of IPTG to induce P_lac-traR expression but were absent in all other RNA preparations (Fig . 4) . No detectable primer extension products were specified by the traA primer for any of the RNA preparations . Three different traA antisense primers were tested, and none yielded a detectable reverse transcript . A primer extension product for the tetA gene carried on pSW213 was detectable in all RNA preparations and served as a control . These results suggested that although the tra operons initiated with traI and traC (Mpf and Dtr, respectively) are strongly activated by TraR, the Dtr operon initiated by traA may not be significantly expressed under these conditions .

 
The primer extension product for traI was used to map its transcription start point (TSP) by comparison to a sequencing reaction with the traI-specific oligonucleotide (Fig . 4) . The traI TSP was identified 24 bp upstream of the predicted traI start codon, with NGR tra box II centered at position -44 (Fig . 1B) . The size of the traC reverse transcript, based on the traI sequencing ladder, predicted a TSP 34 bp upstream of the traC start codon, with the presumptive NGR tra box I centered at position -46 (Fig . 1B) . Although no primer extension product was observed for the transcript initiating with traA, the position of NGR tra box I would predict a TSP 129 bp upstream of the traA start codon and upstream of the predicted oriT sequence (Fig . 1B) .

Quorum-sensing regulators of A . tumefaciens and NGR234 do not cross-function. The TraR and TraI proteins encoded on pNGR234a share significantly higher similarity with TraR and TraI from the Ti plasmids (56 and 27% identity with regulators from the octopine-type Ti plasmids, respectively) than they do with other LuxR and LuxI homologues . Additionally, the organization of the pNGR234a tra-rep cluster clearly indicates that these plasmid-borne systems are orthologous, perhaps reflecting a relatively recent ancestral plasmid replicon . Given these observations, it seemed plausible that the quorum-sensing components might cross-function . This is apparently correct for the TraI orthologues, where the primary reaction product of NGR234 TraI is likely to be 3-oxo-C8-HSL . To test for TraR protein cross-functionality, the pBDJ6 (P_lac-traR) plasmid was introduced into A . tumefaciens NTL4 harboring pCF372, carrying the Ti plasmid traI-lacZ fusion . Conversely, pCF218, strongly expressing the Ti plasmid TraR, was introduced into A . tumefaciens NTL4 harboring pBDJ3, carrying the NGR234 traI-lacZ fusion . Neither traI-lacZ fusion was activated by the noncognate TraR, although each was strongly activated by its cognate regulator in the presence of AAI (Table 4) .

 
The NGR234 traM gene was tested for its ability to inhibit transcriptional activation by the Ti plasmid TraR, as it does with the NGR234 TraR . A . tumefaciens harboring the Ti plasmid traR expression plasmid pCF218 and the traI-lacZ fusion on pCF372 was transformed with pCF424, carrying the NGR234 traM expressed as a P_lac-traM fusion . A control strain harboring pGB200, expressing the Ti plasmid TraM but otherwise identical to pCF424, was also analyzed . As expected, when its expression was induced with IPTG, the Ti plasmid traM strongly inhibited Ti plasmid traR activation of traI-lacZ expression (Table 5) . However, expression of the NGR234 traM in this background had no effect on the activity of the Ti plasmid traR and was identical to that of a strain lacking either P_lac-traM plasmid .

 
Low-efficiency conjugal transfer of pNGR234a is not improved by ectopic expression of traR. The effect of TraR activation on conjugal transfer of pNGR234a to the NGR234-ANU265 recipient (pNGR234a cured) was examined . The NGR234 strain harboring the P_lac-traR derivative repressed by lacI^q can be grown in the presence of IPTG, and although the inducer slows growth (see below), it also strongly elevates pNGR234a traI expression and acyl-HSL synthesis . A pNGR234a derivative carrying a Km^r marker (NGR234-BDJ3) conjugated at an extremely low efficiency (approximately 10^-9), and this remained constant regardless of the presence of the P_lac-traR plasmid and the IPTG inducer (Table 6) . Analysis of the putative transconjugants with pNGR234a-specific primers revealed the presence of the plasmid (data not shown) . As NGR234-ANU265 transconjugants harboring pNGR234a would be virtually indistinguishable from the donor strain, it was possible that the infrequent Km^r colonies were spontaneous Km^r donors . Conjugation assays with A . tumefaciens C58 C1RS, a strain that lacks the Ti plasmid, were performed (8) . Efficiency of transfer of the Km^r marker to C58 C1RS from NGR234-BDJ3 was equivalent to that when NGR234-ANU265 was used as a recipient (Table 6) . Randomly amplified polymorphic DNA PCR analysis of the C58 C1RS transconjugants demonstrated that the transconjugants were C58 C1RS derivatives that had received the pNGR234a plasmid (data not shown) .

 
Subsequent use of the C58 C1RS harboring the Km^r-marked pNGR234a as a donor with the plasmid-cured NGR234-ANU265 as a recipient also resulted in similarly low conjugal transfer efficiencies, suggesting that the weak transfer observed is a characteristic intrinsic to pNGR234a and is not unique to the NGR234 genetic background (data not shown) .

Ectopic expression of traR reduces the growth rate of NGR234. We repeatedly were unable to introduce pBDJ6 (P_lac-traR) into a wild-type NGR234 strain by electroporation . In related rhizobia, quorum sensing significantly affects growth rate (24, 58) . We reasoned that the introduction of the strongly expressed P_lac-traR might be growth inhibitory, preventing establishment of the plasmid . Two derivatives of NGR234 were constructed to examine this possibility: one harboring pSW213, expressing the lacI^q repressor of P_lac, and the other harboring pSW172, which is identical to pSW213 but without the lacI^q gene (4) . The NGR234 strain harboring pSW213 would have relatively high levels of LacI, repressing P_lac-traR expression . The efficiency of electroporation of pBDJ6 (P_lac-traR) into the NGR234(pSW213) lacI^q-expressing strain and the control recipient without lacI^q [NGR234(pSW172)] was compared with that of a pBBR1-MCS5 vector control . Efficient transformation (4.3 x 10⁴ transformants/µg plasmid) into the derivative harboring lacI^q was observed for pBDJ6, but, consistent with previous results, no transformants were observed for transformation of pBDJ6 into the control background without lacI^q (Table 7) . In contrast, the pBBR1-MCS5 vector control was transformed with equal efficiency into both recipient strains (Table 7) . Null mutation of the pNGR234a traI gene (NGR234-XH100) resulted in a strain that was transformed with the P_lac-traR plasmid at an efficiency similar to that of the vector control (Table 7) . Addition of synthetic 3-oxo-C8-HSL to 100 nM prior to or immediately following electroporation prevented transformation of pBDJ6 into the traI null mutant but did not affect the vector control . Collectively, these results indicate that transformation inhibition requires expression of the traR gene and the presence of its inducing ligand . We designated this phenotype TraR-mediated transformation inhibition (Tmi) .

 
The lacI^q-expressing NGR234 strain harboring the P_lac-traR expression plasmid pBDJ6 grew as normal-size colonies and had a cellular morphology typical for NGR234 (data not shown) . However, NGR234(pSW213 lacI^q) harboring pBDJ6 had a generation time of 3.5 h/doubling in ATGN broth cultures, compared to that of 2 h for the same strain with the pBBR1-MCS5 vector control (Fig . 5) . Addition of IPTG to induce expression of P_lac-traR increased the generation time to 6 h/doubling for the strain harboring pBDJ6 but did not affect the pBBR1-MCS5 vector control . The final culture density of the IPTG-induced, growth rate-inhibited culture did eventually reach an optical density similar to that of the others, indicating a reduction in growth rate but not early entry into stationary phase . The traI null mutant NGR234-XH100 harboring pBDJ6 (P_lac-traR) was observed to grow similarly to the wild-type strain (2 h/doubling) . However, addition of synthetic 3-oxo-C8-HSL substantially reduced the growth rate of this strain, to 14 h/doubling at high concentrations of the inducer (100 nM), while the growth rate of the traI null mutant with the pBBR1-MCS5 control was unaffected . These data suggest that ectopic expression of traR in the presence of its inducing ligand results in a slowing of growth in NGR234 .

 
The TraM and TraR proteins of A . tumefaciens form an antiactivation complex that prevents TraR from activating transcription of Ti plasmid target genes (30, 40, 62) . Reasoning that the growth inhibition was likely due to the activation of TraR target gene expression, we examined transformation with a plasmid expressing the NGR234 traR and traM genes from the same promoter (pDP101, P_lac-traR-traM) . Transformation of pDP101 was observed in both the lacI^q and control backgrounds, although the efficiency was improved in the lacI^q-harboring strain (Table 7) . This result suggests that inhibition of TraR, either at the level of gene expression, as with the lacI^q recipient, or posttranslationally by antiactivation with TraM effectively suppresses the Tmi phenotype .

Growth rate inhibition requires traR and traI from pNGR234a and functions encoded elsewhere in the genome. Introduction of the P_lac-traR plasmid pBDJ6 was examined in NGR234-ANU265, which lacks pNGR234a . Transformation of pBDJ6 was equally efficient to that of the vector control when NGR234-ANU265 was used as a recipient (Table 7) . However, this strain lacks the pNGR234a traI gene and therefore synthesizes only trace levels of what is presumptively 3-oxo-C8-HSL (Fig . 2) . Preincubation of recipient NGR234-ANU265 with 100 nM synthetic 3-oxo-C8-HSL prevented transformation with the P_lac-traR plasmid, while transformation with a vector control was unaffected (Table 7) . Furthermore, the growth rate of NGR234-ANU265 harboring pBDJ6 was significantly inhibited by addition of 3-oxo-C8-HSL, with 20 nM 3-oxo-C8-HSL increasing the generation time to 13 to 14 h/doubling .

A derivative of pNGR234a carrying a Km^r marker (pNGR234a::pCF429) was conjugally transferred into the Ti plasmid-cured A . tumefaciens derivative C58 C1RS . Subsequent introduction of the P_lac-traR plasmid (pBDJ6) was as efficient as that of the vector control (Table 7), and the growth rate of the pBDJ6 transformant was normal, suggesting that growth inhibition requires additional functions encoded specifically within the NGR234 genome .

 
A more nonpolar activity, migrating as slowly as our C₁₂/C₁₄-HSL standards, was detected in all extracts analyzed . The same activity was also detectable in extracts of cell-free culture fluids (data not shown) but was notably more abundant in whole-cell extracts, consistent with a long-chain acyl-HSL (43) . Synthesis of this putative long-chain acyl-HSL was elevated in the absence of pNGR234a and further induced by pNGR234a TraR (Fig . 2) . It is possible that this compound(s) is synthesized by the same enzyme responsible for the pNGR234a-independent synthesis of the more polar activity . However, traI orthologues have not been reported to synthesize long-chain acyl-HSLs, and it is equally plausible that this activity is synthesized by a different enzyme encoded elsewhere in the NGR234 genome (Fig . 6) . Several other rhizobia synthesize long-chain acyl-HSLs via the activity of LuxI homologues (7, 37, 42) . In R . leguminosarum the CinI protein directs synthesis of N-(3-hydroxy-7-cis-tetradecenoyl)-L-homoserine lactone (3-OH-C_14:1-HSL), and this compound has a growth-inhibitory activity (24, 58) . Furthermore, the presence of the symbiotic plasmid pRL1JI represses but does not abolish 3-OH-C_14:1-HSL synthesis, consistent with our observations that the pNGR234a-cured strain produces higher levels of the slowly migrating activity than the strain harboring pNGR234a .

Expression of traR is limiting in NGR234. The pNGR234a TraR protein responds to 3-oxo-C8-HSL to activate at least two of the tra-trb structural gene operons (Dtr and Mpf) on the symbiotic plasmid (Fig . 6) . Activation of the promoter upstream of traI is highly sensitive to 3-oxo-C8-HSL . However, it is striking that activation of tra genes on pNGR234a occurs only when traR is expressed ectopically . Our results suggest that expression of the pNGR234a traR gene is limiting, similar to the case for the A . tumefaciens Ti plasmid-borne traR genes (18, 52) . Expression of the Ti plasmid traR gene, and hence conjugal transfer of the Ti plasmid, is strictly regulated by the presence of a specific conjugation-inducing subset of opines, compounds released from the crown gall tumors of infected plants (22, 33) . We speculate that an analogous upstream signal cascade regulates pNGR234a traR expression under a specific set of conditions, perhaps in response to plant metabolites . Although legumes nodulated by rhizobia are not reported to produce opines per se, strains of the species R . leguminosarum and S . meliloti synthesize a class of inositol derivatives called rhizopines within nodules, and these are thought to feed free-living rhizobia in the rhizosphere (48) . Rhizopine catabolism is relatively common among soil bacteria, and NGR234 has a cluster of genes homologous to the moc rhizopine catabolic genes of S . meliloti, suggesting that NGR234 may utilize these compounds (21, 65, 67) . It is plausible that rhizopines regulate traR expression, similar to opines in A . tumefaciens . Alternatively, additional plant-released compounds or other aspects of the environment may control pNGR234a traR expression .

Regulation of pNGR234a genes by TraR and TraM. TraR activation of P_traI requires the NGR tra box II . Other putative NGR tra box sequences similar to this element but only weakly similar to Ti plasmid tra boxes were identified in positions almost identical to the positions of tra boxes on Ti plasmids (17) . We provide direct evidence that TraR activates expression of the Dtr transcript initiating with traC, suggesting that tra box I is at least partially functional . However, despite repeated efforts we could not detect a transcript for the operon initiating with traA, suggesting that this operon is not activated by TraR or also requires additional factors .

Weak expression of the traA Dtr operon may explain why conjugal transfer of the pNGR234a plasmid is not enhanced by ectopic expression of traR in NGR234 (Table 6), although the traI and traC operons are strongly induced . The traAFBH operon encodes homologues of the nickase-helicase TraA and the peptidase TraF, both of which are essential for conjugal transfer of Ti plasmids (10) . The rare transfer events that we observe may be due to weak basal expression of the pNGR234a traAFBH operon, or the plasmid may be mobilized in trans by other, unidentified NGR234 Dtr functions . The dynamic structure of the NGR243 genome, with reversibly integrated pNGR234a and pNGR234b episomes in equilibrium, could also contribute to the low conjugal transfer efficiency we observe for pNGR234a (11) .

In A . tumefaciens TraR elevates plasmid copy number through activation of the repABC genes, which is mediated in part by interactions with tra box III immediately upstream of the rep operon and divergent from traI (36; S . C . Winans, personal communication) . We have identified a sequence with weak similarity to NGR tra box II in a similar position relative to the pNGR234a repABC genes (a putative NGR tra box III), and we speculate that TraR influences the copy number of this symbiotic plasmid as well .

Ti plasmid-encoded TraR is strongly inhibited by formation of an antiactivation complex with TraM, and this inhibition is required for cell density responsiveness (13, 53, 55, 62) . Our results suggest that the pNGR234a traM gene product is orthologous and inhibits the activity of TraR . Repeated attempts to generate a traM null mutant were unsuccessful, perhaps due to the growth-inhibitory affect of unregulated TraR activity (B . D . Jones and C . Fuqua, unpublished data) . Therefore, we do not know whether this gene is required to maintain quorum-sensing functions suppressed under noninducing conditions, as is the Ti plasmid traM .

Functional divergence of the pNGR234a and Ti plasmid quorum-sensing regulators. Despite significant sequence conservation and similar mechanisms, the TraR and TraM quorum-sensing regulators from Ti plasmids and from pNGR234a do not cross-function when coexpressed . The overall levels of identity between these proteins are 27 and 19%, for TraR and TraM, respectively . However, alanine-scanning mutagenesis of traM carried on the octopine-type Ti plasmid, based on conservation with the pNGR234a TraM, revealed that 13 of the 18 conserved residues were required for TraM activity (62) . Clearly, these quorum-sensing proteins are orthologous and share similar mechanisms but must recognize specific attributes of their cognate TraR proteins . The pNGR234a TraR responds to the same inducing ligand as the Ti plasmid TraR, but these proteins cannot recognize the DNA binding sites of the other protein, reflecting the divergence of their tra box sequences . LasR of Pseudomonas aeruginosa and LuxR of Vibrio fischeri are more divergent and productively recognize each other's DNA binding sites (23) .

Growth inhibition and quorum sensing in NGR234. We observe a significant effect of TraR activity on the growth rate of NGR234, manifested most dramatically by the Tmi phenotype, i.e., the inability to transform wild-type NGR234 with a plasmid that strongly expresses traR . The Tmi phenotype is consistently associated with growth rate reduction, and we believe that these phenomena are manifestations of the same regulatory mechanism . Freshly electroporated cells provided with an extended incubation period in the absence of antibiotic selection produce slow-growing P_lac-traR transformants (data not shown) . This suggests that it is the combination of electroporation, antibiotic selection, and the growth-inhibitory effect of elevated TraR activity that prevents successful establishment of the introduced plasmid .

Effects of quorum sensing on growth rate have been reported for R . leguminosarum bv . viciae and R . etli CNPAF512 (7, 24, 58) . Before its identification as an acyl-HSL, 3-OH-C_14:1-HSL produced by several biovars of R . leguminosarum was named small bacteriocin for its growth inhibitory effects (26) . Subsequent studies demonstrated that small was an acyl-HSL synthesized via the activity of the LuxI-type protein CinI (37) . CinI and its paired LuxR-type regulator CinR govern the complex hierarchy of multiple quorum-sensing systems in R . leguminosarum . The presence of the symbiotic plasmid pRL1JI partially suppresses production of 3-OH-C_14:1-HSL through inhibition of cinI expression, although this signal is still required for synthesis of plasmid-encoded acyl-HSLs . Addition of crude extracts or 3-OH-C_14:1-HSL to R . leguminosarum can arrest growth but does not cause a loss of viability (24) . The mechanism of growth arrest requires genes carried on pRL1JI . R . etli CNPAF512 produces a saturated-long-chain (slc) acyl-HSL compound similar to 3-OH-C_14:1-HSL, referred to as 3-OH-(slc)-HSL, also via a protein called CinI (7) . The R . etli CNPAF512 signal molecule inhibits growth of R . leguminosarum strains sensitive to 3-OH-C_14:1-HSL . The cinI gene also influences R . etli CNPAF512 growth rate along with a second quorum-sensing system encoded by the raiI and raiR genes (56) . R . etli CNPAF512 raiR and cinI single mutants and cinI raiR double mutants exhibit significantly longer lag times than the wild type upon inoculation and grow 1.5- to 2-fold more slowly, eventually reaching full density . These results suggest that 3-OH-(slc)-HSL is required for the normal growth rate of R . etli, in contrast to the case for R . leguminosarum . NGR234 has attributes of both systems; elevated expression of traR and addition of 3-oxo-C8-HSL lead to a reduction in growth rate, but rather than causing growth arrest, the growth rate is simply slowed .

Our findings indicate that the only genes on pNGR234a required for growth inhibition are traR and traI . Additionally, A . tumefaciens strains harboring pNGR234a are readily transformed with the traR expression plasmid, suggesting that pNGR234a is not sufficient for growth inhibition . A model in which growth inhibition is mediated by an interaction between TraR and functions encoded elsewhere in the NGR234 genome, perhaps including the genes responsible for synthesis of the pNGR234a-independent acyl-HSLs, is consistent with our findings (Fig . 6) . We have isolated several NGR234 mutations that specifically suppress the growth-inhibitory effect of traR expression, none of which are located on pNGR234a (X . He and C . Fuqua, unpublished results) .

TraR and TraM in other rhizobia. Marketon and González have identified genes encoding homologues of TraR and TraM with 80% identity to the pNGR234a gene products on the pRme41a plasmid from the alfalfa symbiont S . meliloti AK631 (42) . In contrast to pNGR234a, the pRme41a plasmid from AK631 is not required for plant nodulation (27) . Preliminary sequencing around the pRme41a traR and traM genes identified Mpf homologues of trbH and trbI, suggesting a gene organization similar to that of the Mpf cluster on pNGR234a (12) . S . meliloti AK631 synthesizes a remarkable range of acyl-HSLs, and similar to our findings with NGR234, a portion of these are encoded by the pRme41a plasmid, while other long-chain acyl-HSLs are encoded within the remainder of the AK631 genome (42, 43) . Mutation of traR significantly reduces synthesis of these acyl-HSLs in laboratory culture, suggesting that quorum sensing is active under these conditions (42) . In contrast, our findings suggest that traR on pNGR234a is tightly controlled, keeping acyl-HSL levels low in laboratory cultures . Tight regulation of traR expression in NGR234 may be related to its role in growth rate control, as high-level activity would hamper growth . No effect of quorum sensing on growth rate has been reported for derivatives of S . meliloti .

Following submission of this work for publication, Wilkinson et al . published findings examining the regulation of conjugal transfer genes and growth inhibition in R . leguminosarum bv . viciae, which are in many ways consistent with our results with NGR234 (69) . However, in R . leguminosarum, the traR homologue triR is activated by an adjacently encoded LuxR-type protein called BisR, which also inhibits 3-OH-C_14:1-HSL synthesis . There is no BisR homologue encoded adjacent to traR on the pNGR234a plasmid, but a second LuxR homologue, designated y4qH, is 250 kb away from the tra-rep cluster and could conceivably influence traR expression (12) . Wilkinson et al . also found that the soluble translation elongation factor Ts (EF-Ts) is present at high levels in R . leguminosarum inhibited by 3-OH-C_14:1-HSL (69) . It remains unclear whether this is a cause or an effect of growth inhibition . The EF-Ts levels have not been tested in growth-inhibited NGR234, but the protein is a possible target for TraR-dependent growth inhibition .

Complex quorum-sensing signaling pathways. Several different quorum-sensing microorganisms employ complex cascades with multiple acyl-HSLs and associated regulatory proteins . The best studied of these is P . aeruginosa, in which two quorum-sensing systems (Las and Rhl) function in a hierarchal, overlapping cascade that is hypothesized to provide staged expression of virulence factors for this pathogen in association with its hosts (5, 68) . The rhizobia possess extremely complex quorum-sensing systems, and given the highly developed rhizobial-legume interactions, it is not surprising that quorum sensing contributes . Quorum sensing can influence different aspects of interaction with the host, ranging from modulatory to direct roles during nodulation (7, 37, 41, 56) . We speculate that quorum sensing in NGR234 affects population dynamics in association with host plants . The growth regulation via TraR might function to control population expansion of NGR234 under crowded conditions, either within nodules or in the rhizospheres of nodulated plants .

This study was supported by National Science Foundation grant MCB-9974863 (to C.F.) .

Present address: Department of Microbiology, Cornell University, Ithaca, NY 14853 .

Present address: U.S . Plant Soil Nutrition Lab, USDA-Agricultural Research Station, Cornell University, Ithaca, NY 14853 .

1. Akakura, R., and S . C . Winans. 2002 . Constitutive mutations of the OccR regulatory protein affect DNA bending in response to metabolites released from plant tumors . J . Biol . Chem . 277:5866-5874.
2. Cha, C., P . Gao, Y.-C . Chen, P . D . Shaw, and S . K . Farrand. 1998 . Production of acyl-homoserine lactone quorum-sensing signals by gram-negative plant-associated bacteria . Mol . Plant-Microb . Interact . 11:1119-1129.
3. Charles, T . C., and T . M . Finan. 1991 . Analysis of a 1600-kilobase Rhizobium meliloti megaplasmid using defined deletions generated in vivo . Genetics 127:5-20.
4. Chen, C.-Y., and S . C . Winans. 1991 . Controlled expression of the transcriptional activator gene virG in Agrobacterium tumefaciens by using the Escherichia coli lac promoter . J . Bacteriol . 173:1139-1144.
5. Chugani, S . A., M . Whiteley, K . M . Lee, D . D'Argenio, C . Manoil, and E . P . Greenberg. 2001 . QscR, a modulator of quorum-sensing signal synthesis and virulence in Pseudomonas aeruginosa . Proc . Natl . Acad . Sci . USA 98:2752-2757.
6. Cubo, M . T., A . Economou, G . Murphy, A . W . B . Johnston, and J . A . Downie. 1992 . Molecular characterization and regulation of the rhizosphere-expressed genes rhiABCR that can influence nodulation by Rhizobium leguminosarum biovar viciae . J . Bacteriol . 174:4026-4035.
7. Daniels, R., D . E . De Vos, J . Desair, G . Raedschelders, E . Luyten, V . Rosemeyer, C . Verreth, E . Schoeters, J . Vanderleyden, and J . Michiels. 2002 . Quorum-sensing in Rhizobium etli CNPAF512 affects growth and symbiotic nitrogen fixation . J . Biol . Chem . 277:462-468.
8. Ellis, J . G., A . Kerr, A . Petit, and J . Tempé. 1982 . Conjugal transfer of nopaline and agropine Ti-plasmids—the role of agrocinopines . Mol . Gen . Genet . 186:269-273.
9. Evans, K., L . Passador, R . Srikumar, E . Tsang, J . Nezezon, and K . Poole. 1998 . Influence of the MexAB-OprM multidrug efflux system on quorum sensing in Pseudomonas aeruginosa . J . Bacteriol . 180:5443-5447.
10. Farrand, S . K., I . Hwang, and D . M . Cook. 1996 . The tra region of the nopaline-type Ti plasmid is a chimera with elements related to the transfer systems of RSF1010, RP4, and F . J . Bacteriol . 178:4233-4247.
11. Flores, M., P . Mavingui, L . Girard, X . Perret, W . J . Broughton, E . Martinez-Romero, G . Davila, and R . Palacios. 1998 . Three replicons of Rhizobium sp . strain NGR234 harbor symbiotic gene sequences . J . Bacteriol . 180:6052-6053.
12. Freiberg, C., R . Fellay, A . Balroch, W . J . Broughton, A . Rosenthal, and X . Perret. 1997 . Molecular basis of symbiosis between Rhizobium and legumes . Nature 387:394-401.
13. Fuqua, C., M . Burbea, and S . C . Winans. 1995 . Activity of the Agrobacterium Ti plasmid conjugal transfer regulator TraR is inhibited by the product of the traM gene . J . Bacteriol . 177:1367-1373.
14. Fuqua, C., and A . Eberhard. 1999 . Signal generation in autoinduction systems: synthesis of acylated homoserine lactones by LuxI-type proteins, p . 211-230 . In G . M . Dunny and S . C . Winans (ed.), Cell-cell signaling in bacteria . ASM Press, Washington, D.C.
15. Fuqua, C., and E . P . Greenberg. 2002 . Listening in on bacteria: acylhomoserine lactone signalling . Nat . Rev . Mol . Cell . Biol . 3:685-695.
16. Fuqua, C., M . Parsek, and E . P . Greenberg. 2001 . Regulation of gene expression by cell-to-cell communication: acyl-homoserine lactone quorum sensing . Annu . Rev . Genet . 35:439-468.
17. Fuqua, C., and S . C . Winans. 1996 . Conserved cis-acting promoter elements are required for density-dependent transcription of Agrobacterium tumefaciens conjugal transfer genes . J . Bacteriol . 178:435-440.
18. Fuqua, C., and S . C . Winans. 1996 . Localization of the OccR-activated and TraR-activated promoters that express two ABC-type permeases and the traR gene of the Ti plasmid pTiR10 . Mol . Microbiol . 120:1199-1210.
19. Fuqua, W . C., and S . C . Winans. 1994 . A LuxR-LuxI-type regulatory system activates Agrobacterium Ti plasmid conjugal transfer in the presence of a plant tumor metabolite . J . Bacteriol . 176:2796-2806.
20. Fuqua, C., S . C . Winans, and E . P . Greenberg. 1996 . Census and consensus in bacterial ecosystems: the LuxR-LuxI family of quorum-sensing transcriptional regulators . Annu . Rev . Microbiol . 50:727-751.
21. Gardener, B . B . M., and F . J . de Bruijn. 1998 . Detection and isolation of novel rhizopine-catabolizing bacteria from the environment . Appl . Environ . Microbiol . 64:4944-4949.
22. Genetello, C., N . Van Larebeke, M . Holsters, A . De Picker, M . Van Montagu, and J . Schell. 1977 . Ti plasmids of Agrobacterium as conjugative plasmids . Nature 265:561-563.
23. Gray, K . M., L . Passador, B . H . Iglewski, and E . P . Greenberg. 1994 . Interchangeability and specificity of components from the quorum-sensing regulatory systems of Vibrio fischeri and Pseudomonas aeruginosa . J . Bacteriol . 176:3076-3080.
24. Gray, K . M., J . P . Pearson, J . A . Downie, B . E . A . Boboye, and E . P . Greenberg. 1996 . Cell-to-cell signalling in the symbiotic nitrogen-fixing bacterium Rhizobium leguminosarum: autoinduction of a stationary phase and rhizosphere-expressed genes . J . Bacteriol . 178:372-376.
25. Hirsch, A . M., M . R . Lum, and J . A . Downie. 2001 . What makes the rhizobia-legume symbiosis so special? Plant Physiol . 127:1484-1492.
26. Hirsch, P . R. 1979 . Plasmid-determined bacteriocin production by Rhizobium leguminosarum . J . Gen . Microbiol . 113:219-228.
27. Huguet, T., C . Rosenberg, F . Casse-Delbart, P . de Lajudie, L . Jouanin, J . Batut, B . P., J.-S . Julliot, and J . Denarie. 1983 . Studies on Rhizobium meliloti plasmids and on their role in the control of nodule formation and nitrogen fixation: the pSym megaplasmids and the other large plasmids, p . 35-45 . In A . Puhler (ed.), Molecular genetics of the bacteria-plant interaction . Springer-Verlag, New York, N.Y.
28. Hwang, I., D . M . Cook, and S . K . Farrand. 1995 . A new regulatory element modulates homoserine lactone-mediated autoinduction of Ti plasmid conjugal transfer . J . Bacteriol . 177:449-458.
29. Hwang, I., P.-L . Li, L . Zhang, K . R . Piper, D . M . Cook, M . E . Tate, and S . K . Farrand. 1994 . TraI, a LuxI homologue, is responsible for production of conjugation factor, the Ti plasmid N-acylhomoserine lactone autoinducer . Proc . Natl . Acad . Sci . USA 91:4639-4643.
30. Hwang, I., A . J . Smyth, Z.-Q . Luo, and S . K . Farrand. 1999 . Modulating quorum sensing by antiactivation: TraM interacts with TraR to inhibit activation of Ti plasmid conjugal transfer genes . Mol . Microbiol . 34:282-294.
31. Kalogeraki, V . S., and S . C . Winans. 1997 . Suicide plasmids containing promoterless reporter genes can simultaneously disrupt and create fusions to target genes of diverse bacteria . Gene 188:69-75.
32. Kaplan, H . B., and E . P . Greenberg. 1985 . Diffusion of autoinducer is involved in regulation of the Vibrio fischeri luminescence system . J . Bacteriol . 163:1210-1214.
33. Kerr, A., P . Manigault, and J . Tempé. 1977 . Transfer of virulence in vivo and in vitro in Agrobacterium . Nature 265:560-561.
34. Kovach, M . E., P .