FixJ-regulated genes evolved through promoter duplication in Sinorhizobium meliloti

Lionel Ferrières{dagger}, Anne Francez-Charlot{ddagger}, Jérôme Gouzy, Stéphane Rouillé and Daniel Kahn

Laboratoire des Interactions Plantes-Microorganismes, UMR 2594 INRA-CNRS, Chemin de Borde-Rouge, BP 27, 31326 Castanet-Tolosan Cedex, France

Correspondence
Daniel Kahn
dkahn{at}toulouse.inra.fr


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The FixLJ two-component system of Sinorhizobium meliloti is a global regulator, turning on nitrogen-fixation genes in microaerobiosis. Up to now, nifA and fixK were the only genes known to be directly regulated by FixJ. We used a genomic SELEX approach in order to isolate new FixJ targets in the genome. This led to the identification of 22 FixJ binding sites, including the known sites in the fixK1 and fixK2 promoters. FixJ binding sites are unevenly distributed among the three replicons constituting the S. meliloti genome: a majority are carried either by pSymA or by a short chromosomal region of non-chromosomal origin. Thus FixJ binding sites appear to be preferentially associated with the pSymA replicon, which carries the fixJ gene. Functional analysis of FixJ targets led to the discovery of two new FixJ-regulated genes, smc03253 and proB2. This FixJ-dependent regulation appears to be mediated by a duplication of the whole fixK promoter region, including the beginning of the fixK gene. Similar duplications were previously reported for the nifH promoter. By systematic comparison of all promoter regions we found 17 such duplications throughout the genome, indicating that promoter duplication is a common mechanism for the evolution of regulatory pathways in S. meliloti.


Tables showing duplications of promoter regions located upstream of non-homologous genes, and giving details of nine intergenic elements appearing to be duplicated within the coding regions of putative genes are available as supplementary data with the online version of this paper at http://mic.sgmjournals.org.

{dagger}Present address: Molecular Microbiology Laboratory, Department of Biology and Biochemistry, University of Bath, Bath BA2 7AY, UK.

{ddagger}Present address: Laboratoire de Microbiologie et de Génétique Moléculaires, UMR5100 CNRS-UPS, 118 route de Narbonne, 31062 Toulouse Cedex, France.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The soil bacterium Sinorhizobium meliloti interacts symbiotically with alfalfa (Medicago sativa) by inducing the formation of nodules on plant roots. Within these organs, bacterial cells differentiate into N2-fixing bacteroids and provide ammonia to the host plant. In S. meliloti, expression of nitrogen-fixation genes is under the control of a cascade which responds to the low oxygen concentration prevailing inside nodules (David et al., 1988). The FixLJ two-component system acts at the very beginning of this cascade. Under microoxic conditions, FixL autophosphorylates and transmits phosphate to the FixJ response regulator. Once phosphorylated, FixJ~P activates transcription of the nifA and fixK genes. NifA and FixK are themselves transcriptional regulators which in turn switch on expression of genes involved in nitrogen fixation and bacteroid respiration, respectively (Fischer, 1994).

The FixLJ, FixK and NifA regulatory elements are conserved among rhizobia, but they differ in connectivity, targets and function. As a consequence, different Rhizobium species possess distinct, species-specific regulatory networks for symbiotic nitrogen fixation (Fischer, 1994). In S. meliloti the nifA gene is directly regulated by FixJ, in Azorhizobium caulinodans nifA is under the direct control of FixK, while in Bradyrhizobium japonicum nifA expression is fully independent of FixJ. Moreover, two copies of fixK are present in the latter species, with FixJ activating fixK2, which in turn activates fixK1 (Nellen-Anthamatten et al., 1998). In Rhizobium etli, fixL is not even involved in the regulation of nitrogen fixation (D'Hooghe et al., 1995, 1998). Therefore, there appears to be a lot of plasticity in the regulatory pathways controlling nitrogen fixation in rhizobia, making generalization across species difficult.

So far, nifA and fixK have been the only genes known to be directly regulated by FixJ in S. meliloti (Fischer, 1994). The position of the FixLJ system at the top of the regulatory cascade led us to hypothesize the existence of other FixJ targets in the S. meliloti genome. Here we used an in vitro selection approach, genomic SELEX (Singer et al., 1997), in order to systematically identify FixJ binding sites in the S. meliloti genome. Twenty-two FixJ binding sites were thus isolated. Functional analysis of five of these sites led to the identification of two novel FixJ-regulated genes. This FixJ-dependent expression appeared to be mediated by tandem duplication of the entire pfixK promoter region. A systematic search throughout the genome indicated that promoter duplication is a frequent event underlying the evolution of regulatory pathways in S. meliloti.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, plasmids and growth conditions.
Bacterial strains and plasmids used in this work are listed in Table 1. Escherichia coli was grown in Luria broth at 37 °C. S. meliloti was grown in TY medium at 30 °C. Microaerobic conditions were achieved in the stoppered-tube assay (Ditta et al., 1987), with 2 % oxygen (initial concentration). Antibiotics and other selective compounds were added to the culture medium at the following concentrations: gentamicin, 20 µg ml–1; neomycin, 30 µg ml–1; sucrose 5 %; spectinomycin, 150 µg ml–1; streptomycin, 400 µg ml–1 (Rm1021); X-Gluc (5-bromo-4-chloro-3-indolyl-{beta}-D-glucuronic acid), 40 µg ml–1.


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids used in this study

 
Genomic SELEX.
A pool of DNA fragments covering the entire genome was prepared as described by Singer et al. (1997). Total genomic DNA (0·17 mg ml–1) was sheared and incubated for 3 min at 93 °C with the OLBran primer randomized over nine 3'-terminal nucleotides (12 µM) (Table 2). The annealing mixture was chilled on ice and the elongation reaction was initiated by adding 10 mM Tris/HCl, pH 7·5, 5 mM MgCl2, 7·5 mM DTT, 300 µM dNTPs and 0·5 U µl–1 DNA polymerase I Klenow fragment (BioLabs). The reaction mixture was left on ice for 5 min and successively incubated at 25 °C for 25 min and at 50 °C for 5 min. The reaction was stopped with 10 mM EDTA, and the enzyme was inactivated by heating at 75 °C for 10 min. Primers in excess were removed through a micro-spin S400 column (Pharmacia). Second strand synthesis was carried out by the same procedure using the OLAran 3'-randomized primer, and the reaction products were separated on a 5 % polyacrylamide denaturing gel. Fragments with a size ranging from 150 bp to 200 bp were excised, electro-eluted and amplified by PCR using OLA and OLB as primers (Table 2). After ethanol precipitation, 100 ng of this PCR product was used for in vitro selection.


View this table:
[in this window]
[in a new window]
 
Table 2. Primers used in this study

Sequences matching the ends of the uidA-{Omega}sp cassette are italicized. XbaI sites are underlined.

 
The SELEX procedure exploited a fusion protein between glutathione-S-transferase (GST) and the C-terminal domain of FixJ (FixJC), as previously described (Ferrières & Kahn, 2002). GST-FixJC (10 µM) was incubated with 100 ng DNA for 15 min at 30 °C in 100 µl binding buffer (60 mM Tris/acetate, pH 8, 60 mM KCl, 30 mM potassium acetate, 27 mM ammonium acetate, 1 mM DTT, 0·1 mM EDTA) containing 25 µg poly(dI-dC) ml–1, 100 µg BSA ml–1, 4 % PEG 6000 and 10 % glycerol (v/v). Glutathione–Sepharose beads were added, and the mixture was left for a further 10 min at 30 °C to allow protein–DNA complexes to adsorb. Beads were recovered and quickly washed with 1 ml of ice-cold binding buffer. Bound DNA was eluted in 100 µl binding buffer by heating for 10 min at 100 °C. Selected fragments were subsequently amplified by PCR, using OLA and OLB as primers (30 s at 94 °C, 30 s at 52 °C and 30 s at 72 °C, in an Eppendorf MasterCyclerGradient). The resulting products were selected again, and the same procedure was repeated until enrichment was sufficient to allow the detection of FixJ~P-DNA complexes by gel-retardation analysis. DNA fragments were finally cloned in pGEM-T (Promega) and sequenced using the T7 universal primer. Gel retardation assays and DNase I footprint experiments were carried out on selected molecules, as described previously (Ferrières & Kahn, 2002).

Construction of the uidA-{Omega}sp cassette.
A 1·9 kb fragment containing the uidA reporter gene and a preceding Shine–Dalgarno sequence was extracted from pRG960SD-32 (Van den Eede et al., 1992) by digestion with BamHI and EcoRI, then inserted into pBlueScript-II KS+ (Promega) to generate pGMI2191. A truncated {Omega}sp cassette, lacking the T4 transcriptional terminator located upstream of aadA, was amplified by PCR using pHP45{Omega} (Prentki & Krisch, 1984) as a template, with {Omega}4 and {Omega}5 as primers (Table 2). To avoid the introduction of mutations, the amplification was carried out with the Pfu polymerase (2 min at 94 °C; 20 cycles of 30 s at 94 °C, 30 s at 60 °C, 4 min at 72 °C; 5 min at 72 °C). The PCR product was purified, digested with EcoRI and integrated into the EcoRI site of pGMI2191, leading to pGMI2192. In pGMI2192, aadA is positioned downstream and in the same orientation as uidA.

Construction of S. meliloti mutant strains.
Genomic fragments (200–300 bp) corresponding to the beginning of each gene to be inactivated were amplified by PCR, using the gene-specific A and B primers (Table 2), 10 ng of total genomic DNA from Rm1021 and the Expand Long Template PCR System Kit (Boehringer Mannheim). Fragments corresponding to the end of each gene were similarly amplified using the gene-specific C and D primers (Table 2). The full uidA-{Omega}sp cassette was also amplified from pGMI2192 with the Pfu polymerase and primers uidA-1 and {Omega}3. This cassette was then fused with the two amplified gene fragments by primerless elongation with the Expand Long Template PCR System Kit polymerase (2 min at 94 °C; 10 cycles of 30 s at 94 °C, 30 s at 60 °C, 3 min at 68 °C; 5 min at 68 °C). The gene-specific B and C primers were engineered so that they overlapped the uidA-{Omega}sp cassette, allowing for the integration of uidA-{Omega}sp between the ends of the target gene. In order to amplify the resulting 3·8 kb fusion product, 10 µl of the reaction mixture was added to a 50 µl PCR reaction mixture containing the gene-specific A and D primers. The long range PCR was performed using the same polymerase as above, with the following amplification conditions: 2 min at 94 °C; 10 cycles of 30 s at 94 °C, 30 s at 65 °C, 3 min at 68 °C; 15 cycles of 30 s at 94 °C, 30 s at 65 °C, 3 min plus 20 s/cycle at 68 °C; 7 min at 68 °C. The PCR product was digested with XbaI and cloned into the S. meliloti pJQ200KS suicide plasmid (Quandt & Hynes, 1993). The functionality of the uidA and aadA genes was subsequently verified by introducing the resulting plasmid into DH5{alpha} and plating with X-Gluc and streptomycin. This plasmid was then introduced into Rm1021 by tri-parental mating, using GMI3442 as helper strain. E. coli strains were counter-selected on TY medium containing gentamicin, spectinomycin and streptomycin (400 µg ml–1). GMI516, GMI518, GMI520, GMI522, GMI524 and GMI526 S. meliloti mutants strains were finally isolated as Spr, Sucr and Gms colonies. The GMI517, GMI519, GMI521, GMI523, GMI525, GMI527 and GMI528 fixJ strains were obtained by transducing the fixJ2·3 : : Tn5 mutation from GMI347 (David et al., 1988) into GMI516, GMI518, GMI520, GMI522, GMI524, GMI52 and Rm1021, respectively, using bacteriophage N3 (Martin & Long, 1984).

{beta}-Glucuronidase assays.
{beta}-Glucuronidase activity was assayed as described by Jefferson (1987). Mid-log culture cells (OD600=0·5) were harvested and resuspended in lysis buffer (50 mM sodium phosphate, pH 7·0, 10 mM {beta}-mercaptoethanol, 10 mM EDTA, 0·1 % sodium lauryl sarcosine and 0·1 % Triton X-100). MUG, 4-methyl-umbelliferyl-{beta}-D-glucuronide (1 mM; Sigma M-9130), was added to initiate the enzymic reaction. After 15 min incubation at 37 °C, the reaction was stopped by 0·2 M Na2CO3. Samples were kept on ice and protected from light until analysis. The concentration of methyl-umbelliferone (MU) was determined by fluorescence at 455 nm (excitation at 365 nm) on a Kontron SFM 25 spectrofluorimeter, using 10 µM MU (Sigma M-1508) as a standard.

RT-PCR.
RNA was extracted from a 15 ml culture grown at 30 °C under aerobic or microoxic conditions, as described by Cabanes et al. (2000). Briefly, cells were harvested at OD600=0·5, lysed for 10 min at 65 °C in 600 µl pre-warmed lysis solution [1·4 % SDS, 4 mM EDTA and 420 µg proteinase K ml–1 (Boehringer Mannheim)]. Proteins were precipitated on ice by addition of 150 µl 5 M NaCl solution. The supernatant was recovered, and nucleic acids were precipitated with ethanol, resuspended and digested at 37 °C for 1 h with 80 U DNase I (FPLC pure, Amersham Pharmacia Biotech) in 0·1 M sodium acetate, pH 5, containing 5 mM MgSO4. After phenol/chloroform extraction, RNA was precipitated with ethanol, dissolved in DEPC-treated water and verified on a 1 % agarose gel.

First-strand cDNA synthesis was achieved in a 20 µl volume. RNA (100 ng) was incubated for 10 min at 70 °C in the presence of the appropriate reverse primer (500 nM final concentration), then quickly chilled on ice. Reaction buffer was added (50 mM Tris/HCl, pH 8·3, 75 mM KCl, 3 mM MgCl2, 10 mM DTT, 500 µM each dNTP), the mix was kept at 42 °C for 2 min, and 200 U SuperScript II (Life Technologies) was added for cDNA synthesis (50 min at 42 °C). Reverse transcriptase was inactivated for 5 min at 95 °C. A portion (2 µl) of this cDNA was then introduced into a 50 µl PCR reaction mix (10 mM Tris/HCl, pH 8·3, 50 mM KCl, 1·5 mM MgCl2, 200 µM each dNTP, 0·02 U Taq polymerase µl–1, 800 nM reverse and forward primers) and amplified as follows: 2 min at 94 °C; cycle of 30 s at 94 °C, 30 s at 55 °C and 30 s at 72 °C; 5 min at 72 °C. We used 25 cycles for hemA, fixK and proB2 and 40 cycles for smc03253. The resulting PCR products were analysed on a 2 % agarose gel. The following forward and reverse oligonucleotides were used: HemAS and HemAR for hemA; FixK-PF and FixK-PR for fixK; PROHYD1 and PROHYD2 for smc03253; ProB2-A and ProB2-B for proB2 (see Table 2).

Plant assays.
Seeds of Medicago sativa cv europe were surface sterilized and grown in tubes containing nitrogen-free Fahraeus medium slants (growth conditions: 22 °C, 60 % relative humidity, 16 h of light/8 h of darkness alternations). Three weeks after root inoculation with S. meliloti, nitrogen-fixing activity was assayed by acetylene reduction (Turner & Gibson, 1980).


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Identification of new FixJ binding sites in the S. meliloti genome
An in vitro selection procedure (SELEX), using a fusion protein between glutathione S-transferase (GST) and the C-terminal DNA-binding domain of FixJ (FixJC), had previously been developed and applied to a pool of randomized synthetic oligonucleotides in order to define consensus sequences for FixJ binding sites (Ferrières & Kahn, 2002). Two classes of artificial FixJ binding sites were thus defined, Class I and Class II, characterized by two distinct consensus sequences (Ferrières & Kahn, 2002). Here we used the same SELEX approach on S. meliloti genomic fragments in order to systematically identify natural FixJ targets.

DNA fragments covering the entire genome were generated by random priming and size-selected as described by Singer et al. (1997). Genomic fragments ranging from 150 to 200 bp were incubated with the GST–FixJC fusion protein, and protein–DNA complexes were adsorbed on glutathione–Sepharose beads. Bound DNA was eluted, amplified by PCR and protein–DNA complexes were selected again. After three selection cycles, the mix appeared highly enriched in FixJ binding fragments. These fragments were cloned and sequenced, generating 71 independent sequences. The sequences were compared with one another and clustered into 32 groups, corresponding to different regions in the genome. Among these, 22 regions showed affinity for FixJ~P, as determined by gel retardation assays (listed in Fig. 1). Importantly, many of them were isolated on distinct overlapping fragments, indicating that they contain genuine FixJ binding sites, independently of the flanking sequences used in the SELEX procedure.



View larger version (35K):
[in this window]
[in a new window]
 
Fig. 1. FixJ binding sites in the S. meliloti genome. Genomic segments containing FixJ binding sites (diamonds) are depicted with respect to their local gene environment. Sites showing a DNase I footprint in the presence of FixJ~P are underlined. Sequence numbering is as published by Barnett et al. (2001), and can be found on the S. meliloti genome server (http://sequence.toulouse.inra.fr/meliloti.html). Chrom., chromosome.

 
The 22 regions were further tested by DNase I footprint analysis in the presence of FixJ~P (data not shown). Eight regions generated unambiguous footprints, allowing the confirmation and accurate definition of the corresponding FixJ binding sites, including the sites in the previously known fixK1 and fixK2 promoters (Waelkens et al., 1992; Galinier et al., 1994) and two other sites which are similar to the FixJ binding site in pfixK (sites Ic and Id, Fig. 2a). Consistent with this similarity, FixJ~P protection patterns on pfixK1, pfixK2, Ic and Id are similar. The region protected by FixJ~P extends over 40 bp and lies at the same location within pfixK1, pfixK2, Ic and Id. Furthermore, a DNase I-hypersensitive site was found at a matching position on the top strands, corresponding to position G–62 of pfixK1. This defines sites Ic and Id as belonging to the previously defined Class I FixJ binding sites, like pfixK1 and pfixK2 (Ferrières & Kahn, 2002). The remaining four sites, II, XX, VII and XXIII, share a conserved TACGTAG motif located within the protected region (Fig. 2b), as evidenced by analysis with the MEME program (Bailey & Elkan, 1994). This motif exhibits similarity with Consensus 2, thereby assigning these sites to Class II FixJ binding sites (Ferrières & Kahn, 2002). Five of the new FixJ binding sites (Ic, II, VII, XX and XXII) were chosen for further analysis. These sites appeared to be good candidates for mediating transcriptional regulation because they are located upstream of genes and lie within intergenic regions (Fig. 1). Furthermore, four of them (Ic, II, VII and XX) formed complexes with FixJ~P stable enough to allow DNase I footprinting (Fig. 2).



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 2. Alignment of new FixJ binding sites. (a) Alignment of the pfixK1 and pfixK2 promoters (Waelkens et al., 1992) with the Ic and Id FixJ binding sites identified by genomic SELEX. Sequences shown correspond to the extents of DNase I footprints in the presence of FixJ~P. Sequence numbering relates to the fixK transcription start (Batut et al., 1989). (b) Alignment of four genomic FixJ binding sites with FixJ Consensus 2 previously defined (Ferrières & Kahn, 2002). Sequences shown correspond to the extents of DNase I footprints in the presence of FixJ~P.

 
FixJ activates expression of smc03253 and proB2 under microaerobic conditions
In order to test the biological relevance of the selected FixJ binding sites, we inactivated the downstream genes smc03253, sma1447, smc03247, smc03149 and proB2 by insertion of a uidA-{Omega}sp cassette. This cassette creates a transcriptional fusion with the uidA reporter gene encoding {beta}-glucuronidase (Jefferson et al., 1986). Expression of the mutated genes was followed in a wild-type or a fixJ background, under aerobic or microoxic conditions. Only two of the five inactivated genes, smc03253 and proB2, appeared to be regulated by FixJ (Table 3). These genes showed a strong FixJ-dependent induction under low oxygen pressure and were also expressed in nodules. FixJ-dependent expression of smc03253 and proB2 was further confirmed by RT-PCR, using hemA and fixK as controls for constitutive and FixJ-regulated genes, respectively. Therefore, smc03253 and proB2 are clearly new members of the fixJ regulon. It remains, however, to be determined whether all three FixJ binding sites upstream of smc03253 and proB2 are functional. Two other genes, smc03247 and sma1447, appeared to be induced under microoxic conditions, but surprisingly this induction occurred independently of FixJ (Table 3).


View this table:
[in this window]
[in a new window]
 
Table 3. {beta}-Glucuronidase (GUS) activities of reporter strains

Cells were grown in TY medium until mid-exponential phase under aerobic (O2) or microaerobic (µO2; ~2 % O2) conditions. GUS activities were measured (Jefferson, 1987) and reported as nmol of methyl-umbelliferone (MU) produced per min per mg protein.

 
We further investigated the functions of these genes by inoculating Medicago sativa roots with each of the five mutant strains. All of them led to the formation of fully nitrogen-fixing nodules (data not shown). There was no delay in nodule formation, and plants did not display any symptom of nitrogen starvation. Therefore, the smc03253, sma1447, smc03247, smc03149 and proB2 genes are not essential for symbiosis with alfalfa.

FixJ-dependent regulation of smc03253 is mediated by duplication of the fixK promoter
The FixJ-regulated smc03253 gene is positioned in a particular region of the chromosome, proposed to have been acquired by horizontal transfer (Capela et al., 2001). smc03253 follows the two Ic and Id FixJ binding sites, which share extensive homology with the pfixK region located on pSymA (Fig. 3). Besides the FixJ binding site of pfixK, the Ic and Id regions include 124 bp from the fixK coding region. These constitute the beginning of two short identical ORFs, {Psi}fixK1 and {Psi}fixK2. The corresponding protein sequences share homology with only the first 21 residues of FixK, which suggests that {Psi}fixK1 and {Psi}fixK2 are pseudogenes resulting from the duplication of the fixK promoter region. As in the pfixK region, a copy of the fixT gene, fixT3, precedes the Id FixJ binding site (Foussard et al., 1997), with, however, a frameshift occurring 52 bp before the fixT3 stop codon. Moreover, only 12 bp are conserved between the upstream sequences of fixT1 and fixT3, suggesting that the two genes might be regulated differently. This is consistent with previously published genetic data, which indicate that fixT3 is not involved in the negative regulation of the FixLJ cascade (Foussard et al., 1997). A 384 bp portion of the Id region, containing the FixJ binding site, the {Psi}fixK1 pseudogene and the 3' end of fixT3, is tandemly duplicated and forms the Ic site. The Ic and Id duplications differ by a single nucleotide, which is deleted in the Ic copy (Fig. 3), indicating that this is a recent tandem duplication. We therefore propose that the chromosome acquired the Id region first by duplication of the pfixK region from pSymA, followed by tandem duplication to form the Ic site. Expression of smc03253 is thus regulated by FixJ via a functional duplication of the fixK promoter.



View larger version (43K):
[in this window]
[in a new window]
 
Fig. 3. Duplication of the fixK promoter upstream of smc03253. (a) Sequence and structure. fixT3, {Psi}fixK1, {Psi}fixK2 and smc03253 coding regions are indicated. The tandemly duplicated copies of the pfixK region are shown by square brackets. Nucleotides shown in white type on black correspond to positions protected by FixJ~P from DNase I attack in sites Ic and Id (lower and upper sites, respectively). The DNA sequence matching the fixK gene is double underlined. Amino acid residues conserved between {Psi}FixK and FixK are shown in bold. Arrowheads indicate nucleotide deletions in fixT3 and {Psi}fixK duplications. The single nucleotide that differentiates region Id from Ic is in bold and underlined. (b) Schematic representation of the smc03253 upstream sequence. Both copies of pfixK are in brackets. The Ic and Id FixJ~P binding sites are shown as black boxes. smc03253 is not drawn to scale.

 
Promoter duplications are widespread in the S. meliloti genome
Two examples of functional promoter duplications were reported earlier for S. meliloti. Better and co-workers identified four copies of the nifH promoter in Rm102F34, three of which (P1, P2 and P3) are located in the nif region (Better et al., 1983). Another duplication of pnifH was found upstream of the rhizopine locus in S. meliloti strain L5-30, allowing nifA-dependent expression of the mosABC genes (Murphy et al., 1988, 1993). The identification of a new promoter duplication in the S. meliloti Rm1021 chromosome led us to hypothesize that promoter duplication could be a common feature in S. meliloti. In order to test this hypothesis, we looked systematically for other promoter duplications in the recently sequenced Rm1021 genome. Genomic sequences containing the first 100 bp of each gene and the corresponding upstream intergenic regions (no longer than 500 bp) were systematically compared by BLASTN. Matching sequences were selected which displayed more than 70 % identity over at least 50 bp. In order to exclude repeated intergenic elements, the functions of which might differ, one of the matching sequences was required to overlap the downstream initiation codon. Furthermore, insertion sequences (ISRm), ABC motifs (Osteras et al., 1998) and RIME elements (Osteras et al., 1995) were masked before comparison. Finally, duplications extending over less than 40 bp upstream of the initiation codon were discarded, because they are unlikely to contain a full promoter region.

In this way, 49 distinct families of duplicated promoters were identified. Twenty-three promoter duplications were found upstream of whole-gene duplications. These include promoter regions of fixN1/fixN2, fixT1/fixT2 and fixK1/fixK2, belonging to the fix-1/fix-2 duplicated clusters, respectively (Renalier et al., 1987). The three previously reported syrB-like genes, syrB1, syrB2 and sma0806 (Barnett et al., 2001), also display strong similarity in their upstream regions. traA1/traA2, repC1/repC2 and groES1/groES2 are other examples of gene duplications that include upstream promoter regions (refer to http://sequence.toulouse.inra.fr/rhime/Promoter_Duplication for complete data). Furthermore, we identified 17 duplications of promoter regions located upstream of non-homologous genes, potentially bringing them under heterologous control (Table S1, available as supplementary data with the online version of this paper at http://mic.sgmjournals.org). These include the P3 copy of pnifH (cluster C02) (Better et al., 1983), which is located 939 bp upstream of the original nifH promoter, and the pfixK-like site described above (cluster C16). Curiously, a duplication of the rpoN 5' region (rpoN') was found between P3 and the downstream sma0824 gene (cluster C21). Further analysis indicates that rpoN' results from a duplication of the full rpoN gene followed by a 1280 bp internal deletion (Fig. 4). Indeed, sma0824 shares homology over 190 bp with the 3' end of rpoN, which is consistent with the strong similarity observed between the N-terminal end of SMa0824 and the C-terminal end of {sigma}54 (Barnett et al., 2001). It thus appears that this region has been extensively rearranged. It does not contain essential symbiotic elements, as evidenced by Tn5 insertion analysis (Hirsch et al., 1983). A third duplication of the nifH promoter region was identified 320 bp upstream of the nodD2 gene (cluster C02). However, the homology between the two sequences extends over 52 bp only, including the first 48 bp of nifH, but excluding promoter elements. The –6 to –89 region including the nifH promoter was found in the opposite orientation, suggesting that the nifH promoter was inverted after duplication. Finally, nine intergenic elements appeared to be duplicated within the coding regions of putative genes (Table S2, available as supplementary data with the online version of this paper at http://mic.sgmjournals.org).



View larger version (13K):
[in this window]
[in a new window]
 
Fig. 4. Mosaic structure of the sma0824 region on pSymA. The sma0824 and nifH genes are preceded by copies of the nifH promoter (diagonal hatching) P3 (position 452 385–452 678) and pnifH (position 453 324–453 605), respectively. The 5' and 3' ends of the rpoN duplications, rpoN' (position 452 679–452 854) and rpoN'' (position 452 855–453 047), are represented by a white box and waved hatching, respectively. The arrowhead indicates the 1280 bp deletion in the rpoN duplication. The circled S corresponds to the rpoN start codon (position 452 738).

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Evolution of the FixJ regulon
The S. meliloti genome has a composite structure, made of three replicons with distinct structural and functional properties: the chromosome, pSymA and pSymB (Galibert et al., 2001). The chromosome is the largest replicon and carries genes needed for basic metabolism, including the three rRNA operons (Capela et al., 2001). pSymB contains many genes involved in solute transport and polysaccharide synthesis. It extends the capacity of the bacterium to adapt to different environmental conditions, and also contains a few essential genes (Finan et al., 2001). Genome sequence analysis suggests that pSymA was acquired relatively recently by S. meliloti, bringing the ability to live at low oxygen concentration, to colonize legume roots and to fix nitrogen (Barnett et al., 2001). The pSymA-borne fixLJ system is essential for microaerobic growth and for nitrogen fixation, because it controls expression of both the high-affinity cbb3 oxidase and the nif genes (Fischer, 1994; David et al., 1988). The FixJ binding sites identified here appear to be unequally distributed among the three replicons. pSymA contains ten sites evenly spread throughout the replicon, while only three sites are located on pSymB. Moreover, all three sites on pSymB are located within putative genes, making them unlikely candidates for functional sites. Similarly, the chromosome contains very few FixJ~P binding sites, with the exception of an 80 kb region (region 1) hitherto recognized to be of non-chromosomal origin because of its low GC content and because of a high number of transposon remnants (Capela et al., 2001). Thus, both the fixJ gene and the majority of its targets appear to be carried by pSymA, which is consistent with the proposal that pSymA was acquired recently, together with the fixJ regulon. In this scenario, the cluster of FixJ binding sites on the chromosome would mostly derive by transfer from pSymA, as was found to be the case for the duplication of pfixK.

Non-functional FixJ binding sites?
Most of the FixJ binding sites isolated by genomic SELEX do not seem to be directly functional for transcriptional activation. Indeed 11 of the 22 newly identified FixJ targets are located downstream from or inside genes, making them unlikely candidates as regulatory targets. Furthermore, three of the five candidate sites chosen for further analysis were found not to mediate fixJ-dependent regulation of downstream genes. The determination of the function of these sites will require further experimentation. One hypothesis is that they might serve as reservoir sites facilitating FixJ~P diffusion and target location (Gowers & Halford, 2003; von Hippel & Berg, 1989). Indeed it has been shown that topologically linked non-specific protein–DNA interactions facilitate target location through 3D space (von Hippel & Berg, 1989), accounting for the very fast forward rates typically found (Gowers & Halford, 2003; von Hippel & Berg, 1989). This hypothesis would also rationalize the strong bias in the distribution of ‘non-functional’ FixJ sites, which places them in the vicinity of functional FixJ targets (Gowers & Halford, 2003). The prevalence of seemingly non-functional FixJ sites in the S. meliloti genome can also be paralleled with earlier work by Berg and von Hippel, who predicted the existence of a large number of CRP pseudo-sites in the E. coli genome (Berg & von Hippel, 1988).

Novel fixJ-regulated genes
In this work, two novel FixJ-regulated genes have been identified in S. meliloti: smc03253 and proB2. Like nifA and fixK, these genes are induced microaerobically under FixJ control. However, unlike nifA and fixK, they are not essential for symbiotic nitrogen fixation, and their role remains elusive. Both genes share homology with genes involved in proline metabolism. SMc03253 is homologous to L-proline cis-3-hydroxylase isolated from Streptomyces. This enzyme modifies free L-proline by stereospecific hydroxylation (Mori et al., 1997). ProB2 is homologous to the ProB {gamma}-glutamate kinase of E. coli, which acts at the first step in the proline biosynthetic pathway. However, proB2 is not the likely proB orthologue in S. meliloti. Indeed, the genome contains another copy, proB1, which displays a stronger similarity with proB and forms an operon with proA. Furthermore, proB2 was not required for proline prototrophy in S. meliloti (data not shown). We therefore suggest that proB2 and smc03253 are involved in a facultative proline-related pathway in bacteroids. In this context, it is also worth mentioning the importance of proline catabolism for symbiotic nitrogen fixation (King et al., 2000; Jimenez-Zurdo et al., 1997). The possible function of hydroxyproline in S. meliloti is unclear. In eukaryotes, hydroxyproline has been identified in collagen and in many plant cell-wall proteins in which L-proline is hydroxylated post-translationally (Adams & Frank, 1980). In prokaryotes, hydroxyproline has been found in some peptide antibiotics such as etamycin (Sheehan et al., 1958), plusbacin (Shoji et al., 1992) and telomycin (Irreverre et al., 1962). Detection of cis-3-proline hydroxylase activity in the Streptomyces canus strain producing telomycin suggests that hydroxyproline is incorporated as a free residue into peptide antibiotics (Mori et al., 1996). By analogy, smc03253 might be involved in secondary metabolism in S. meliloti. Although we have shown that proB2 and smc03253 are not essential for symbiosis, we cannot rule out that they might provide an advantage to the bacterium in a natural environment, as was shown to be the case for the mos genes involved in rhizopine synthesis (Murphy et al., 1988). Orthologues of smc03253 and proB2 can be found in the symbiosis island of Mesorhizobium loti strain MAFF303099: mlr6283 and mlr6298, respectively (Kaneko et al., 2000). However, they are missing in the symbiosis island of another strain of M. loti, R7A (Sullivan et al., 2002), and in the Bradyrhizobium japonicum genome (Kaneko et al., 2002).

FixJ-independent microaerobic gene induction
During the course of this study we identified two microaerobically induced genes, smc03247 and sma1447, which unexpectedly escaped FixJ-dependent regulation, despite an adequate FixJ binding site. Other genes have been shown to be regulated by oxygen in a FixJ-independent fashion, such as the asnO gene, which encodes an asparagine synthetase homologue (Bergès et al., 2001), and the six loe-2, loe-3, loe-5, loe-6, loe-8 and loe-9 loci (Trzebiatowski et al., 2001). Furthermore, microaerobic induction of nifA appears not to be solely regulated by the FixLJ system (Kahn & Ditta, 1991). These observations point to the existence of other regulatory systems in S. meliloti that respond to low oxygen concentration.

Promoter duplication and genome adaptation
The FixJ-dependent expression of smc03253 is mediated by duplication of the fixK promoter region. This is not unprecedented; indeed, earlier studies have identified several duplications of the pnifH promoter, which controls expression of nitrogenase in S. meliloti. pnifH is a {sigma}54-dependent promoter which requires the NifA transcriptional regulator for activation. Four copies of pnifH (P1, P2, P3 and P4) have been identified in S. meliloti strain 102F34 (Better et al., 1983). P1, P2 and P3 are located in the nif region of pSymA (Better et al., 1983). P1 is the original nifH promoter, while P2 controls expression of the fixABCX genes involved in electron transport to nitrogenase (Earl et al., 1987; Better et al., 1985). The function of P3 remains unknown, as it is not essential for symbiotic nitrogen fixation (Hirsch et al., 1983). Intriguingly, this region contains duplications of parts of the {sigma}54 gene rpoN, a gene required for pnifH activity. Another copy of pnifH has been identified upstream of the mos operon in S. meliloti strains L5-30 (Murphy et al., 1988) and Rm220-3 (Rao et al., 1995). mos genes are involved in rhizopine production in these strains, but are absent from strain Rm1021. Rhizopines have been proposed to be used as growth substrates by nodule-inducing bacteria, providing a selective advantage for the bacterial partner. However, they are not required for symbiotic nitrogen fixation. Two key symbiotic promoters, pnifH and pfixK, have thus been recruited for the control of other genes which are not essential for symbiosis. A systematic search for other such promoter duplications in the genome revealed the existence of 15 additional candidates, suggesting that promoter duplication is a common phenomenon that extends the regulatory repertoire of S. meliloti. It thus appears that bacteria can generate new regulatory pathways by the combinatorial assortment of pre-existing cis-regulatory elements.


   ACKNOWLEDGEMENTS
 
We wish to thank Britta Singer and Larry Gold for the genomic SELEX protocol, and Joëlle Fourment for technical assistance. L. F. was the recipient of a ‘Bourse de Docteur Ingénieur’ from CNRS. This work was supported by grants from the European Union (BIO4 CT 97-2143), the Ministère de l'Education Nationale, de la Recherche et de la Technologie (Programme de Recherche Fondamentale en Microbiologie et Maladies Infectieuses et Parasitaires) and INRA (Programme prioritaire ‘Microbiologie’).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Adams, E. & Frank, L. (1980). Metabolism of proline and the hydroxyprolines. Annu Rev Biochem 49, 1005–1061.[CrossRef][Medline]

Bailey, T. L. & Elkan, C. (1994). Fitting a mixture model by expectation maximization to discover motifs in biopolymers. Proc Int Conf Intell Syst Mol Biol 2, 28–36.[Medline]

Barnett, M. J., Fisher, R. F., Jones, T. & 23 other authors (2001). Nucleotide sequence and predicted functions of the entire Sinorhizobium meliloti pSymA megaplasmid. Proc Natl Acad Sci U S A 98, 9883–9888.[Abstract/Free Full Text]

Batut, J., Daveran-Mingot, M. L., David, M., Jacobs, J., Garnerone, A. M. & Kahn, D. (1989). fixK, a gene homologous with fnr and crp from Escherichia coli, regulates nitrogen fixation genes both positively and negatively in Rhizobium meliloti. EMBO J 8, 1279–1286.[Abstract]

Berg, O. G. & von Hippel, P. H. (1988). Selection of DNA binding sites by regulatory proteins. Trends Biochem Sci 13, 207–211.[CrossRef][Medline]

Bergès, H., Checroun, C., Guiral, S., Garnerone, A. M., Boistard, P. & Batut, J. (2001). A glutamine-amidotransferase-like protein modulates FixT anti-kinase activity in Sinorhizobium meliloti. BMC Microbiol 1, 6.[CrossRef][Medline]

Better, M., Lewis, B., Corbin, D., Ditta, G. & Helinski, D. R. (1983). Structural relationships among Rhizobium meliloti symbiotic promoters. Cell 35, 479–485.[Medline]

Better, M., Ditta, G. & Helinski, D. R. (1985). Deletion analysis of Rhizobium meliloti symbiotic promoters. EMBO J 4, 2419–2424.

Boyer, H. W. & Roulland-Dussoix, D. (1969). A complementation analysis of the restriction and modification of DNA in Escherichia coli. J Mol Biol 41, 459–472.[Medline]

Cabanes, D., Boistard, P. & Batut, J. (2000). Symbiotic induction of pyruvate dehydrogenase genes from Sinorhizobium meliloti. Mol Plant-Microbe Interact 13, 483–493.[Medline]

Capela, D., Barloy-Hubler, F., Gouzy, J. & 25 other authors (2001). Analysis of the chromosome sequence of the legume symbiont Sinorhizobium meliloti strain 1021. Proc Natl Acad Sci U S A 98, 9877–9882.[Abstract/Free Full Text]

David, M., Daveran, M. L., Batut, J., Dedieu, A., Domergue, O., Ghai, J., Hertig, C., Boistard, P. & Kahn, D. (1988). Cascade regulation of nif gene expression in Rhizobium meliloti. Cell 54, 671–683.[Medline]

D'Hooghe, I., Michiels, J., Vlassak, K., Verreth, C., Waelkens, F. & Vanderleyden, J. (1995). Structural and functional analysis of the fixLJ genes of Rhizobium leguminosarum biovar phaseoli CNPF512. Mol Gen Genet 249, 117–126.[Medline]

D'Hooghe, I., Michiels, J. & Vanderleyden, J. (1998). The Rhizobium etli FixL protein differs in structure from other known FixL proteins. Mol Gen Genet 257, 576–580.[CrossRef][Medline]

Ditta, G., Virts, E., Palomares, A. & Kim, C. H. (1987). The nifA gene of Rhizobium meliloti is oxygen regulated. J Bacteriol 169, 3217–3223.[Medline]

Earl, C. D., Ronson, C. W. & Ausubel, F. M. (1987). Genetic and structural analysis of the Rhizobium meliloti fixA, fixB, fixC, and fixX genes. J Bacteriol 169, 1127–1136.[Medline]

Ferrières, L. & Kahn, D. (2002). Two distinct classes of FixJ binding sites defined by in vitro selection. FEBS Lett 517, 185–189.[CrossRef][Medline]

Figurski, D. H. & Helinski, D. R. (1979). Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc Natl Acad Sci U S A 76, 1648–1652.[Abstract]

Finan, T. M., Weidner, S., Wong, K. & 9 other authors (2001). The complete sequence of the 1,683-kb pSymB megaplasmid from the N2-fixing endosymbiont Sinorhizobium meliloti. Proc Natl Acad Sci U S A 98, 9889–9894.[Abstract/Free Full Text]

Fischer, H. M. (1994). Genetic regulation of nitrogen fixation in rhizobia. Microbiol Rev 58, 352–386.[Medline]

Foussard, M., Garnerone, A. M., Ni, F., Soupène, E., Boistard, P. & Batut, J. (1997). Negative autoregulation of the Rhizobium meliloti fixK gene is indirect and requires a newly identified regulator, FixT. Mol Microbiol 25, 27–37.[Medline]

Galibert, F., Finan, T. M., Long, S. R. & 53 other authors (2001). The composite genome of the legume symbiont Sinorhizobium meliloti. Science 293, 668–672.[Abstract/Free Full Text]

Galinier, A., Garnerone, A. M., Reyrat, J. M., Kahn, D., Batut, J. & Boistard, P. (1994). Phosphorylation of the Rhizobium meliloti FixJ protein induces its binding to a compound regulatory region at the fixK promoter. J Biol Chem 269, 23784–23789.[Abstract/Free Full Text]

Gowers, D. M. & Halford, S. E. (2003). Protein motion from non-specific to specific DNA by three-dimensional routes aided by supercoiling. EMBO J 22, 1410–1418.[Abstract/Free Full Text]

Hanahan, D. (1983). Studies on transformation of Escherichia coli with plasmids. J Mol Biol 166, 557–580.[Medline]

Hirsch, A. M., Bang, M. & Ausubel, F. M. (1983). Ultrastructural analysis of ineffective alfalfa nodules formed by nif : : Tn5 mutants of Rhizobium meliloti. J Bacteriol 155, 367–380.[Medline]

Irreverre, F., Morita, K., Ishii, S. & Witkop, B. (1962). Occurrence of cis- and trans-3-hydroxy-L-proline in acid hydrolyzate of telomycin. Biochem Biophys Res Commun 9, 69–71.

Jefferson, R. A. (1987). Assaying chimeric genes in plants : the GUS gene fusion system. Plant Mol Biol Report 5, 387–405.

Jefferson, R. A., Burgess, S. M. & Hirsh, D. (1986). Beta-glucuronidase from Escherichia coli as a gene-fusion marker. Proc Natl Acad Sci U S A 83, 8447–8451.[Abstract]

Jimenez-Zurdo, J. I., Garcia-Rodriguez, F. M. & Toro, N. (1997). The Rhizobium meliloti putA gene: its role in the establishment of the symbiotic interaction with alfalfa. Mol Microbiol 23, 85–93.[CrossRef][Medline]

Kahn, D. & Ditta, G. (1991). Modular structure of FixJ: homology of the transcriptional activator domain with the –35 binding domain of sigma factors. Mol Microbiol 5, 987–997.[Medline]

Kaneko, T., Nakamura, Y., Sato, S. & 21 other authors (2000). Complete genome structure of the nitrogen-fixing symbiotic bacterium Mesorhizobium loti. DNA Res 7, 331–338.[Medline]

Kaneko, T., Nakamura, Y., Sato, S. & 14 other authors (2002). Complete genomic sequence of nitrogen-fixing symbiotic bacterium Bradyrhizobium japonicum USDA110. DNA Res 9, 189–197.[Medline]

King, N. D., Hojnacki, D. & O'Brian, M. R. (2000). The Bradyrhizobium japonicum proline biosynthesis gene proC is essential for symbiosis. Appl Environ Microbiol 66, 5469–5471.[Abstract/Free Full Text]

Martin, M. O. & Long, S. R. (1984). Generalized transduction in Rhizobium meliloti. J Bacteriol 159, 125–129.[Medline]

Meade, H. M., Long, S. R., Ruvkun, G. B., Brown, S. E. & Ausubel, F. M. (1982). Physical and genetic characterization of symbiotic and auxotrophic mutants of Rhizobium meliloti induced by transposon Tn5 mutagenesis. J Bacteriol 149, 114–122.[Medline]

Mori, H., Shibasaki, T., Uozaki, Y., Ochiai, K. & Ozaki, A. (1996). Detection of novel proline 3-hydroxylase activities in Streptomyces and Bacillus spp. by regio- and stereospecific hydroxylation of L-proline. Appl Environ Microbiol 62, 1903–1907.[Abstract]

Mori, H., Shibasaki, T., Yano, K. & Ozaki, A. (1997). Purification and cloning of a proline 3-hydroxylase, a novel enzyme which hydroxylates free L-proline to cis-3-hydroxy-L-proline. J Bacteriol 179, 5677–5683.[Abstract]

Murphy, P. J., Heycke, N., Trenz, S. P., Ratet, P., de Bruijn, F. J. & Schell, J. (1988). Synthesis of an opine-like compound, a rhizopine, in alfalfa nodules is symbiotically regulated. Proc Natl Acad Sci U S A 85, 9133–9137.[Abstract]

Murphy, P. J., Trenz, S. P., Grzemski, W., de Bruijn, F. J. & Schell, J. (1993). The Rhizobium meliloti rhizopine mos locus is a mosaic structure facilitating its symbiotic regulation. J Bacteriol 175, 5193–5204.[Abstract]

Nellen-Anthamatten, D., Rossi, P., Preisig, O., Kullik, I., Babst, M., Fischer, H. M. & Hennecke, H. (1998). Bradyrhizobium japonicum FixK2, a crucial distributor in the FixLJ-dependent regulatory cascade for control of genes inducible by low oxygen levels. J Bacteriol 180, 5251–5255.[Abstract/Free Full Text]

Osteras, M., Stanley, J. & Finan, T. M. (1995). Identification of Rhizobium-specific intergenic mosaic elements within an essential two-component regulatory system of Rhizobium species. J Bacteriol 177, 5485–5494.[Abstract]

Osteras, M., Boncompagni, E., Vincent, N., Poggi, M. C. & Le Rudulier, D. (1998). Presence of a gene encoding choline sulfatase in Sinorhizobium meliloti bet operon: choline-O-sulfate is metabolized into glycine betaine. Proc Natl Acad Sci U S A 95, 11394–11399.[Abstract/Free Full Text]

Prentki, P. & Krisch, H. M. (1984). In vitro insertional mutagenesis with a selectable DNA fragment. Gene 29, 303–313.[CrossRef][Medline]

Quandt, J. & Hynes, M. F. (1993). Versatile suicide vectors which allow direct selection for gene replacement in Gram-negative bacteria. Gene 127, 15–21.[CrossRef][Medline]

Rao, J. P., Grzemski, W. & Murphy, P. J. (1995). Rhizobium meliloti lacking mosA synthesizes the rhizopine scyllo-inosamine in place of 3-O-methyl-scyllo-inosamine. Microbiology 141, 1683–1690.[Abstract]

Renalier, M. H., Batut, J., Ghai, J. & 8 other authors (1987). A new symbiotic cluster on the pSym megaplasmid of Rhizobium meliloti 2011 carries a functional fix gene repeat and a nod locus. J Bacteriol 169, 2231–2238.[Medline]

Sheehan, J. C., Zachau, H. G. & Lawson, W. B. (1958). The structure of etamycin. J Am Chem Soc 80, 3349–3355.

Shoji, J., Hinoo, H., Katayama, T., Nakagawa, Y., Ikenishi, Y., Iwatani, K. & Yoshida, T. (1992). Structures of new peptide antibiotics, plusbacins A1-A4 and B1-B4. J Antibiot 45, 824–831.[Medline]

Singer, B. S., Shtatland, T., Brown, D. & Gold, L. (1997). Libraries for genomic SELEX. Nucleic Acids Res 25, 781–786.[Abstract/Free Full Text]

Sullivan, J. T., Trzebiatowski, J. R., Cruickshank, R. W. & 11 other authors (2002). Comparative sequence analysis of the symbiosis island of Mesorhizobium loti strain R7A. J Bacteriol 184, 3086–3095.[Abstract/Free Full Text]

Trzebiatowski, J. R., Ragatz, D. M. & de Bruijn, F. J. (2001). Isolation and regulation of Sinorhizobium meliloti 1021 loci induced by oxygen limitation. Appl Environ Microbiol 67, 3728–3731.[Abstract/Free Full Text]

Turner, G. L. & Gibson, A. H. (1980). Measurement of nitrogen fixation by indirect means. In Methods for Evaluating Biological Nitrogen Fixation, pp. 111–138. Edited by F. J. Bergersen. Chichester: Wiley.

Van den Eede, G., Deblaere, R., Goethals, K., van Montagu, M. & Holsters, M. (1992). Broad host range and promoter selection vectors for bacteria that interact with plants. Mol Plant-Microbe Interact 5, 228–234.[Medline]

von Hippel, P. H. & Berg, O. G. (1989). Facilitated target location in biological systems. J Biol Chem 264, 675–678.[Abstract/Free Full Text]

Waelkens, F., Foglia, A., Morel, J. B., Fourment, J., Batut, J. & Boistard, P. (1992). Molecular genetic analysis of the Rhizobium meliloti fixK promoter: identification of sequences involved in positive and negative regulation. Mol Microbiol 6, 1447–1456.[Medline]

Received 4 February 2004; revised 5 April 2004; accepted 6 April 2004.



This Article
Abstract
Full Text (PDF)
Supplementary data
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Ferrières, L.
Articles by Kahn, D.
Articles citing this Article
PubMed
PubMed Citation
Articles by Ferrières, L.
Articles by Kahn, D.
Agricola
Articles by Ferrières, L.
Articles by Kahn, D.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2004 Society for General Microbiology.