Fine-Tuning of Galactoglucan Biosynthesis in Sinorhizobium meliloti by Differential WggR (ExpG)-, PhoB-, and MucR-Dependent Regulation of Two Promoters

Abstract

Depending on the phosphate concentration encountered in the environment Sinorhizobium meliloti 2011 synthesizes two different exopolysaccharides (EPS). Galactoglucan (EPS II) is produced under phosphate starvation but also in the presence of extra copies of the transcriptional regulator WggR (ExpG) or as a consequence of a mutation in mucR. The galactoglucan biosynthesis gene cluster contains the operons wga (expA), wge (expE), wgd (expD), and wggR (expG). Two promoters, differentially controlled by WggR, PhoB, and MucR, were identified upstream of each of these operons. The proximal promoters of the wga, wge, and wgd transcription units were constitutively active when separated from the upstream regulatory sequences. Promoter activity studies and the positions of predicted PhoB and WggR binding sites suggested that the proximal promoters are cooperatively induced by PhoB and WggR. MucR was shown to strongly inhibit the distal promoters and bound to the DNA in the vicinity of the distal transcription start sites. An additional inhibitory effect on the distal promoter of the structural galactoglucan biosynthesis genes was identified as a new feature of WggR in a mucR mutant. A regulatory model of the fine-tuning of galactoglucan production is proposed.

Sinorhizobium meliloti associates symbiotically with roots of the leguminous plant Medicago sativa (alfalfa). This symbiosis requires specific recognition that relies on a complex exchange of small signal molecules between both partners (24, 29, 55). The bacteria associate with the plant root and induce the formation of nodules that become colonized by the bacteria via infection threads (23). Inside the nodule, the bacteria differentiate into nitrogen-fixing bacteroids (19). Bacterial exopolysaccharides (EPS) play a key role in the establishment of symbiosis. Low-molecular-mass EPS are required for elongation of the infection thread (22, 28, 35) and for protection against plant defense responses (43). Depending on the phosphate concentration encountered in its environment, S. meliloti produces two different EPS that are both able to promote symbiosis. Low-phosphate conditions (0.1 to 10 μΜ), typically found in soil, stimulate the production of galactoglucan (EPS II), whereas high-phosphate conditions (10 to 20 mM), found in nodules, block EPS II synthesis and induce the production of succinoglycan (EPS I) (15). Succinoglycan is composed of octasaccharide repeating units, with each containing one galactose and seven glucose residues (1, 46). It is decorated by acetyl, succinyl, and pyruvyl groups. Biosynthesis of EPS I is directed by the exo gene cluster located on the pSymB megaplasmid (6-8, 17, 26, 27). EPS II is composed of alternating glucose and galactose residues that are acetylated and pyruvylated, respectively (25). In addition to low-phosphate conditions (58), production of EPS II was observed in the presence of a mutation in mucR (31) or the expR101 mutation (25). The galactoglucan biosynthesis gene cluster, located 200 kb apart from the exo gene region, is responsible for biosynthesis of EPS II (10, 25). This cluster was suggested to be organized into four transcription units, wga (expA), wgcA (expC), wgd (expD), and wge (expE), which contain structural genes required for biosynthesis, and a further transcription unit that contains the regulatory gene wggR (expG) (10, 25). The EPS II biosynthesis genes were recently renamed according to the nomenclature for polysaccharide biosynthesis genes suggested by Reeves et al. (45).

Regulation of EPS II synthesis has been the subject of numerous studies (2, 4, 5, 9, 47). It is assumed that under low-phosphate conditions the induction of EPS II is mediated by the two-component regulatory system PhoR/PhoB, where PhoR is the sensor kinase and PhoB is the response regulator (10, 40, 47). In Escherichia coli, PhoB was found to regulate transcription of target genes by binding to the so-called PHO box (16, 56). In S. meliloti, it affects the transcription of a considerable number of genes (33, 56). Sequences with similarity to the E. coli PHO box were identified in the promoter regions of the wga, wgd, wggR, and wge operons (47). Recently, the presence of a PHO box-like sequence in the promoter region of wggR has been confirmed in a large-scale analysis of PhoB-dependent promoters (56). Moreover, the wgeA (expE1), wgdB (expD2), and wggR genes were shown to be significantly induced in the presence of active PhoB in microarray-based transcriptome studies (33). However, in S. meliloti an interaction between PhoB and predicted PHO boxes was only demonstrated in the pstS promoter region (57), and the role of the different PHO boxes localized in the promoter regions of the EPS II biosynthesis genes has not yet been elucidated.

EPS II production is also affected by the transcriptional regulator WggR. This protein is thought to be an activator since extra copies of wggR weakly stimulated expression of the galactoglucan biosynthesis genes (2, 47). In the promoter regions of these genes the binding site of WggR has previously been identified as a conserved palindrome with two associated boxes (4, 5). Interestingly, this palindrome partially overlaps with one PHO box-like sequence. Moreover, the stimulating effect of phosphate deficiency on EPS II biosynthesis gene expression is strongly reduced in a wggR mutant (47).

In this context, MucR is another regulator of interest. This protein plays a key role in the control of EPS I and EPS II biosynthesis. The mucR gene codes for a small protein with a C₂H₂ zinc finger motif, sharing 80% identity with the Ros protein from Agrobacterium tumefaciens (21). Previously, we have demonstrated that MucR was able to bind to a short DNA region upstream of exoH and exoY, whereby exerting a positive effect on the production of EPS I (13). Moreover, a mutation in mucR increases dramatically the biosynthesis of EPS II (31, 47). However, the regulatory mechanism is still unknown and particularly direct binding of MucR to the promoter regions of the EPS II biosynthesis genes has not been demonstrated yet.

Another protein controlling biosynthesis of galactoglucan is the LuxR-type regulator ExpR involved in quorum sensing governed by SinI and SinR in S. meliloti (38, 44). Due to insertion of an insertion element, the reference strain S. meliloti 1021 and the closely related strain 2011 lack a functional expR gene (44). This allowed us to study the regulatory function of the other regulators independent of an activated Sin quorum-sensing system.

We focus here on the molecular basis of the interwoven regulation paths that control the production of EPS II by S. meliloti 2011 in phosphate-sufficient and phosphate-deficient environments. Identification of promoters and studies of their activities affected by the transcriptional regulators WggR, PhoB, and MucR suggest a model that allows fine-tuning of EPS II biosynthesis gene transcription by means of interactions between the different regulators.

MATERIALS AND METHODS

Bacterial strains and plasmids.

The strains and plasmids used in the present study are listed in Table 1.

TABLE 1.

Strains and plasmids

Strain or plasmid	Relevant characteristicsa	Source or reference
Strains
S. meliloti
Rm2011	Wild-type, Nxr Smr	20
Rm101	Rm2011, Spcr cassette of pHP45Ω inserted into the PmacI site of mucR (mucR101-spc)	10
SmSRΔG	Rm2011, deletion of an wggR fragment comprising 490 nucleotides of the 3′ terminus of the wggR coding region and 17 nucleotides downstream of wggR (ΔwggR)	47
RmH406	ΩphoB3::Tn5-132	T. M. Finan
SmBBΔG101	SmSRΔG (ΔwggR), Spcr cassette of pHP45Ω inserted into the PmacI site of mucR (mucR101-spc)	This study
E. coli
XL1-Blue	recA1lac [F′ proABlacIqZΔM15 Tn10(Tetr)] thi	18
S17-1	E. coli 294, thi, RP4-2-Tc::Mu-Km::Tn7 integrated into the chromosome	51
BL21	F−dcm ompT [lon] hsdSB (rB− mB−) gal	52
Plasmids
pK18mob	pUC18 derivate, lacZα Kmr, mob site	49
pSpB35	pUC19 containing a 3.465-kb SphI-BamHI fragment with exoP	7
pUC57	Sequencing vector, lacZα, Apr	Stratagene
pBCSK	Sequencing vector, lacZα, Cmr	Stratagene
pT7/T3α19	Plasmid for overexpression with the T7 promoter (T7 φ10) from pUC19, lacZα, Apr, pMB1-ori	Amersham Biosciences
pFLAGMucR	pPT7/T3α19 plasmid carrying the gene coding for the Flag-MucR protein, including the native RBS of mucR, inserted in the HindIII/EcoRI restriction sites	This study
pPHU231_FMucR	pPHU231 plasmid carrying the gene coding for the Flag-MucR protein, under the control of the lacZ promoter, inserted into the EcoRI/BamHI restriction site	This study
pPHU231	Broad-host-range expression vector	30
pHisG4032	pWH844 plasmid carrying the gene coding for the His6-WggR protein, inserted in the BamHI/HindIII restriction sites	4
pGP1-2	T7 gene 1, λcI857ts, Kmr	54
pAB1002	pUC6S derivate containing a promoterless lacZ gene	11
pSRPP18	Promoter probe vector for integration into the S. meliloti genome	This study
pARIIa	pUC19 containing a 4.249-kb EcoRI-BglII fragment of the EPS II biosynthesis gene cluster	10

^aCm^r, chloramphenicol resistance; Ap^r, ampicillin resistance; Tet^r, tetracycline resistance; Ap^r, ampicillin resistance; Spc^r, spectinomycin resistance; Km^r, kanamycin resistance; Nx^r, nalidixic acid resistance.

Media and growth conditions.

E. coli strains were grown in Penassay broth (Difco Laboratories, Augsburg, Germany) or in LB medium (48) at 37°C, while S. meliloti strains were grown in TY (12) or LB medium (48) at 30°C. To study the effect of limiting phosphate conditions on expression of EPS II biosynthesis genes, S. meliloti strains were grown in a morpholinepropanesulfonic acid (MOPS)-buffered medium as described by Zhan et al. (58). For β-galactosidase assays, strains grown for 48 h in LB or TY liquid medium were washed twice with MOPS-buffered medium containing either 2 mM phosphate (high phosphate) or 0.1 mM phosphate (low phosphate). A volume of washed cells corresponding to 200 μl of stationary culture was used to inoculate 10 ml of MOPS-buffered medium. Antibiotics were supplied as required: for S. meliloti, streptomycin (600 μg/ml), oxytetracycline (0.5 μg/ml), spectinomycin (200 μg/ml), and ampicillin (100 μg/ml), and for E. coli, chloramphenicol (50 μg/ml) and kanamycin (50 μg/ml).

DNA biochemistry.

Preparation of plasmid DNA, DNA restriction, agarose gel electrophoresis, cloning procedures, and transformation of E. coli cells were carried out according to established protocols (48). Southern hybridizations were performed according to the method of Kessler (32), and rhizobial genomic DNA was isolated according to the method of Meade et al. (39).

Protein purification.

The WggR protein was purified as a His₆ tag fusion protein, applying Ni²⁺ affinity chromatography as described elsewhere (5). The mucR gene was cloned downstream of the T7 promoter into vector pT7/T3α19. E. coli BL21, containing the pGP1-2 plasmid carrying a gene encoding a heat-inducible T7 polymerase, was transformed with the resulting plasmid. Expression of mucR by this strain resulted in a MucR protein carrying an N-terminal Flag peptide. For mucR expression cells were grown in LB medium at 30°C up to an optical density at 600 nm (OD₆₀₀) of 0.4, followed by incubation at 42°C for 30 min and 2 h at 37°C. Cells were harvested by centrifugation and broken by using a French press at 30,000 lb/in². The clear supernatant obtained by ultracentrifugation of the cell lysate (27,000 × g, 4°C, 15 min) was applied to an anti-Flag M2 affinity gel (Sigma, Schnelldorf, Germany) according to the manufacturer's instructions. The Flag-MucR fusion protein was then eluted by using a Flag peptide, concentrated by using an Amicon Ultra15 centrifugal device (Millipore, Schwalbach, Germany) and stored at 4°C until usage. The protein concentration was determined by using a standard Bradford assay.

Plasmid pPHU231_FMucR (Table 1) was constructed for expression of Flag-MucR in S. meliloti.

EMSA.

Electrophoretic mobility shift assays (EMSAs) were performed as previously reported by Baumgarth et al. (5). PCRs were carried out with the primers indicated in Fig. 2. Upstream and downstream primers were equipped with KpnI and BamHI restriction sites, respectively. Plasmids pARIIa and pARIV were used as templates to amplify specific DNA fragments comprising complete or partial promoter regions. Amplified DNA fragments were cloned into pUC57. PCR with Cy3-labeled universal and reverse primers was performed to obtain Cy3-labeled PCR products. These fragments (0.25 ng/μl [final concentration]), purified protein, and buffer (50 mM Tris-HCl [pH 8.0], 100 mM NaCl, 0.1 mM Mg₂SO₄, 4.5% glycerol, 0.05 mg of sonicated herring testis DNA/ml, and 0.5 mg of bovine serum/ml) were mixed to give a final volume of 20 μl. His₆-WggR and Flag-MucR proteins were added to final concentrations of 13 and 170 ng/μl, respectively. After 15 min of incubation at 20°C, the mixture was loaded onto a 2% nondenaturing agarose gel prepared in gel buffer (40 mM Tris base, 10 mM sodium acetate, and 1 mM EDTA adjusted to pH 7.8 with acetic acid) and electrophoresed at 4.5 V cm² for 3 h at 4°C in gel buffer. Gel images were acquired by using a Typhoon 8600 variable mode imager (Amersham Bioscience, Freiburg, Germany).

FIG. 2.

Galactoglucan biosynthesis gene cluster (top) and DNA sequence of the intergenic regions between wgcA and wgaA (A), wgdB and wgeA (B), wggR and wgdA (C), and wgdA and wggR (D). Start codons are indicated by a short dashed line associated with the gene name. Stop codons are marked by three asterisks. Black and gray boxes represent potential PHO boxes and reverse-oriented PHO boxes, respectively. Open boxes mark the 21-bp conserved region of the WggR binding site (5). Thin arrows represent the start of the primers used to generate DNA fragments for insertion into pSRPP18. These primers anneal to 18 bp of the intergenic region starting from the base of the arrow. Transcription starts are indicated by +1 and thick arrows. The dashed lines represent the predicted −35 boxes, whereas the bold lines denote the predicted −10 boxes.

5′RACE.

Rapid amplification of 5′ cDNA ends (5′RACE) was essentially carried out by using a 3′/5′RACE kit (Roche, Mannheim, Germany) according to the manufacturer's instructions. Cells were grown in MOPS-buffered medium up to an OD₆₀₀ of 0.8. After the cells were harvested by centrifugation, the pellets were quickly frozen in liquid nitrogen. RNA was purified by using an RNeasy purification kit (Qiagen, Hilden, Germany). RNA was then reverse transcribed at 55°C for 1 h using the specific primers listed in Table 2. A homopolymeric A tail was then added to the 3′ end of the cDNA. Finally, the dA-tailed cDNA was amplified by the Taq DNA polymerase using a second specific primer (SP2, Table 2) and an oligo(dT) primer. The resulting PCR products were cloned by using a TOPO/TA cloning kit (Invitrogen, Karlsruhe, Germany). The inserted DNA fragments were sequenced, and the transcription starts were mapped. Each start was confirmed by at least four independent clones.

TABLE 2.

Primers used in this study

Primer	Sequence (5′→3′)	Use
SP1_wgaA	CTT CAC CCT CAT CCG AGA AG	wgaA cDNA synthesis
SP2_wgaA	CAA GCG CGA TAA GAG GAA AG	PCR of A-tailed cDNA for wgaA
SP1_wggR	ATA TCT TCG ACC CCG AGC TT	wggR cDNA synthesis
SP2_wggR	TAG GTG ACG AGC GAA TCA TC	PCR of A-tailed cDNA for wggR
SP1_wgeA	CCA GGA TAT CGT TGC CAT TT	wgeA cDNA synthesis
SP2_wgeA	AGA TGA TGT CGT CCC CAT TG	PCR of A-tailed cDNA for wgeA
SP1_wgdA	TAT CCA TGC TCT GGC TGT TC	wgdA cDNA synthesis
SP2_wgdA	ATG GTG AGC TGC AGG AGG TT	PCR of A-tailed cDNA for wgdA
SP1-lacZ	AAGCGCCCATTCGCCATTCAG	cDNA synthesis of fragments inserted into pSRPP18
Oligo(dT)	GACCACGCGTATCGATGTCGAC-poly(T)	PCR of A-tailed cDNA for wgaA, wgeA, wgdA, and wggR
lacZGm2	TTGAGGGGACGACGACAGTAT	Sequencing of fragments inserted into pSRPP18
mucR up1	GGGGAAGCTTAGGAGAAAGAAATGGACTACAAGGACGACGAT	PCR for cloning of mucR
MucR down	AAAAGAATTCTCACTTGCCGCGACGCTT	PCR for cloning of mucR
MucR up2	GACTACAAGGACGACGATGACAAAACAGAGACTTCGCTC	Addition of the Flag-encoding sequence before cloning

Construction of promoter probe vector pSRPP18.

To allow quantitative determination of the activities of single copies of the promoters, the promoter probe vector pSRPP18 was constructed. The basic replicon of pSRPP18 is a pK18mob derivative that is characterized by the deletion of a HindIII-NheI fragment containing 174 bp of the lacZ′ gene. The resulting plasmid was designated pK18mobΔH,N. A 505-bp StuI-NruI fragment from plasmid pSpB35 carrying the 3′ terminus of the exoP gene, and the downstream terminator was equipped with flanking multiple cloning sites derived from pUC57 and pBCSK. Subsequently, it was cloned as an EcoRI-KpnI fragment into the multiple cloning site of pK18mobΔH,N. A promoterless lacZ gene preceded by a ribosome binding site, which was recovered on a BamHI-BglII fragment from plasmid pAB1002, was inserted into the BamHI restriction site of the resulting plasmid.

The promoter fragments to be tested can be inserted into the resulting pSRPP18 vector by using the NruI, KpnI, and BamHI restriction sites located upstream of the promoterless lacZ gene. Plasmid pSRPP18 integrates into the intergenic region between exoP (SMb20961) and thiD (SMb20962) by homologous recombination (Fig. 1). Readthrough of exoP transcription into the promoter-probe construct is blocked by the terminator located immediately downstream of exoP.

FIG. 1.

Genomic structure of the exoP downstream region after integration of the promoter probe vector pSRPP18 carrying a promoter fragment inserted into the KpnI and BamHI restriction sites. The stem-loop symbol indicates the position of the terminator downstream of exoP. The restriction sites available to clone the promoter fragments into this vector are marked as N for NruI, K for KpnI, and B for BamHI. The thick line denotes the vector part of the integrated plasmid, whereas the incomplete exoP gene is shown in parentheses. The scheme is not drawn to scale.

Construction of plasmids used for analysis of regulatory sequences in the promoter regions.

KpnI- and BamHI-restricted PCR products, which were also used for EMSA experiments, were inserted into the KpnI- and BamHI-restricted promoter probe vector pSRPP18. Resulting plasmids were transferred from the broad-host-range-mobilizing strain E. coli S17-1 (51) to S. meliloti as described by Simon (50). The vector-encoded neomycin antibiotic resistance was used to select transconjugants carrying the plasmid inserted into the genome by homologous recombination (Fig. 1). Plasmid insertion was verified by Southern hybridization or PCR. Transconjugants were assayed for β-galactosidase activities.

DNA sequencing.

DNA sequencing was carried out by the IIT Biotech sequencing service (Bielefeld, Germany) using ABI 3130 XL/3730 XL devices. PCR fragments carried by pUC57 or pT7/3α19 were sequenced using the universal and reverse M13 sequencing primers, whereas sequences of fragments inserted into pSRPP18 were determined by using the lacZGm2 primer (Table 2).

β-Galactosidase assays of S. meliloti strains carrying the integrated promoter probe vector pSRPP18 with promoter fragment inserts.

S. meliloti strains were grown to an OD₆₀₀ of 0.6 to 0.8 in MOPS-buffered medium supplemented with 0.1 mM or 2 mM phosphate. The assay for β-galactosidase activity was carried out according to the protocol devised by Miller (41). The specific β-galactosidase activity, relative to the cell biomass, was calculated according to the method of Miller (41).

RESULTS

Two promoters drive transcription of each EPS II biosynthesis operon.

To derive the positions of the promoters of the EPS II biosynthesis operons, transcription start sites (TSS) were determined by 5′RACE. TSS upstream of wgaA (expA1), wgdA (expD1), and wgeA (expE1) were identified in the ΔwggR-mucR double mutant (strain SmBBΔG101, Table 1) cultured in high phosphate MOPS-buffered medium to obtain a high transcription rate. Two TSS were identified upstream of each of these operons (Fig. 2). These TSS were confirmed in medium with a low phosphate concentration, as well as in the wild type and the mucR mutant in media with both high and low phosphate concentrations. Two TSS upstream of wggR (expG) (Fig. 2) were determined in the wild-type and the mucR mutant, since deletion of wggR in the ΔwggR-mucR double mutant did not allow identification of these TSS in this strain.

The presence of two TSS upstream of the EPS II biosynthesis operons suggests that transcription of each of these operons is directed by a proximal promoter and a distal promoter. Alignment of 50 bases preceding the TSS suggested the consensus sequence 5′-CTGTAT∼N14-16∼CAATAT-3′ for the promoters of the EPS II biosynthesis gene cluster (Fig. 3). Although the predicted promoter sequences are poorly conserved and the distances between the TSS and the predicted −10 boxes vary considerably, the consensus sequence shows some similarities to the consensus of S. meliloti promoters proposed by MacLellan et al. (36).

FIG. 3.

Alignment of 50 bp downstream of the different transcription starts found by 5′RACE upstream of the wge, wga, wgd, and wggR operons determined by using CLUSTALW (34). The distance of the transcription start to the ATG of the first gene of the operon is given on the right. The bases are colored as a function of the shared similarities, where dark boxes represent the highest homology between the different sequences. At the bottom, a consensus sequence is proposed. The −35 and −10 boxes are marked by open boxes. The TSS are indicated in boldface.

Previous studies have shown that EPS II biosynthesis is controlled by the transcriptional regulators WggR, MucR, and PhoB (47). We therefore looked at the positions of binding sites of these regulators in relation to the promoter sequences (Fig. 2). Putative binding sites for the phosphate-dependent regulator PhoB were previously found in the upstream regions of the EPS II biosynthesis operons (47). Two PHO box-like sequences were suggested upstream of the wga and wge operons, while one sense and one reverse oriented putative PHO box were identified upstream of wggR and wgdA (Fig. 2). Compared to the consensus sequence the box preceding wggR and the proximal box upstream of wgeA are well conserved, whereas the other PHO box-like sequences are only weakly conserved (53). Previously identified binding sites of WggR (4, 5) map to the distal promoter regions of the wga and wge operons and overlap the predicted distal PHO boxes (Fig. 2). In contrast, a WggR binding site overlapping a putative PHO box is located downstream of the proximal TSS preceding wggR. Upstream of the wgd operon, the reverse-oriented PHO box-like sequence overlaps the WggR binding site between both TSS (Fig. 2).

Transcription is negatively controlled by the binding of MucR to the promoter regions of the EPS II biosynthesis genes.

Sequences similar to previously identified binding sites of MucR (13) could not be found in or close to the promoter regions (Fig. 2). In the present study, MucR was purified as a Flag fusion protein to determine whether the inhibitory effect of MucR on transcription (47) is exerted by binding to the promoter regions of the EPS II biosynthesis genes. The activity of the Flag-MucR protein in vivo was confirmed by expression in the mucR mutant, which resulted in repression of the EPS II biosynthesis genes.

EMSAs were performed with purified Flag-MucR protein and DNA fragments corresponding to the complete or parts of the intergenic regions upstream of the EPS II biosynthesis operons. To investigate the relationship between WggR and MucR, EMSAs were also carried out with His₆-WggR, as well as with combined His₆-WggR and Flag-MucR protein. EMSAs showed a reduced migration of the entire fragments Ef-A1A4, Ef-E1E5, Ef-D1D6, and Ef-G1G5 in the presence of His₆-WggR only, Flag-MucR only, or both proteins (Fig. 4). Reduced mobility was not observed for several heterologous DNA fragments and control fragments from S. meliloti (data not shown). This indicates that MucR is able to specifically bind to the upstream regions of the EPS II biosynthesis operons. However, such a mobility shift required a rather high Flag-MucR protein concentration of at least 170 ng/μl.

FIG. 4.

On the left, a schematic representation of the investigated intergenic region is shown for wgcA-wgaA (A), wgdB-wgeA (B), wggR-wgdA (C), and wgdA-wggR (D). Black and gray boxes represent potential PHO boxes and reverse-oriented PHO boxes, respectively. Open boxes indicate the 21-bp conserved region of the WggR binding site (5). Transcription starts are marked by +1 and arrows. In the center, β-galactosidase activities mediated by the DNA fragments in different mutant backgrounds are shown. β-Galactosidase activities were determined in the wild-type (WT), the phoB mutant (phoB), the mucR101-spc mutant (mucR), the wggR deletion mutant SmSRΔG (wggR), and the double mutant mucR101-spc/wggR (wggR/mucR) in low-phosphate medium (gray bars) and in high-phosphate medium (black bars). β-Galactosidase activity values are the average of at least five independent assays. The background β-galactosidase activity of pSRPP18 in the wild-type strain Sm2011 was 15 ± 1 Miller units. Error bars denote the standard deviation. On the right, results of EMSAs with His₆-WggR and Flag-MucR are shown using the DNA fragments indicated on the left.

Flag-MucR has a calculated molecular mass of 16.7 kDa and reduced the mobility of the DNA fragments, except of the wgaA promoter fragments, at least two times more than WggR did. Since His₆-WggR has a calculated molecular mass of 22.6 kDa and is assumed to bind the DNA as a dimer (4, 5), MucR was probably associated to the DNA as a multimer. The mobility shift obtained with both proteins was similar to that observed with Flag-MucR only. This implies that Flag-MucR might have inhibited binding of His₆-WggR or that some Flag-MucR proteins dissociated from the complex and were replaced by His₆-WggR. A slightly higher mobility was observed if the truncated fragments Df-D5D6 and Pf-G1G4 were subjected to both proteins compared to incubation with Flag-MucR only, which may indicate destabilization of protein binding if both proteins are present. Both proteins seem to bind the DNA in close vicinity and binding may involve protein-protein interactions. It can therefore not be excluded that the terminal protein modifications affected these interactions in vitro.

Regulation of the wga (expA) and wge (expE) operons by WggR (ExpG), PhoB, and MucR.

We hypothesize that transcription of each EPS II biosynthesis operon is directed by two promoters that are subject to differential control by the regulators WggR, PhoB, and MucR. The influence of these regulators on the activities of the promoters of the EPS II biosynthesis gene cluster was studied individually and in combination. Regulation of the wgcA (expC) gene was not included since the intergenic region preceding wgcA displayed neither constitutive nor WggR-, PhoB-, or MucR-controlled promoter activity (data not shown). Expression of a lacZ insertion in the coding region of wgcA followed the pattern of expression of a lacZ insertion in wggR, implying that wgcA may be controlled by the promoter upstream of wggR (47).

Each intergenic region was investigated as a whole as well as divided into a distal fragment (Df) and a proximal fragment (Pf) according to the positions of the TSS (Fig. 2). Promoter activities of these DNA fragments were studied in different mutant backgrounds applying the promoter probe vector pSRPP18 which integrates in single copy into the genome (Fig. 1). Promoter activities were derived from the activity of the lacZ reporter gene product driven by the promoter fragments. β-Galactosidase activities conferred by the promoter probe constructs were assayed in the S. meliloti 2011 wild type, as well as in phoB (RmH406), wggR (SmSRΔG), mucR (Rm101), and wggR/mucR (SmBBΔG101) mutants. Figure 4 summarizes the results of these experiments.

The positions of TSS and predicted binding sites in the upstream regions of the wga and wge operons are quite similar except for the position of the proximal weakly conserved PHO box-like sequence in the wgaA upstream region. In contrast to the well-conserved predicted proximal PHO box upstream of wgeA, this sequence element is situated downstream of the proximal TSS of the wga operon (Fig. 2). The patterns of promoter activities of the entire intergenic regions Ef-A1A4 (wgcA-wgaA) and Ef-E1E5 (wgdB-wgeA) were quite similar and corresponded well to previous reports based on lacZ reporter gene fusions inserted into the EPS II biosynthesis gene coding regions (10, 47). β-Galactosidase activities exceeding the control level were only found under low-phosphate (P_i) conditions in the wild type. The activity was absent or strongly reduced by a mutation in phoB or wggR, suggesting that PhoB and WggR are components of a regulatory cascade or cooperatively mediate phosphate-dependent activation of EPS II biosynthesis genes. The entire wge upstream fragment showed a weak induction of promoter activities under P_i starvation in a wggR mutant background, implying that PhoB is able to partially mediate activation of this operon in the absence of WggR.

Compared to the wild type, a mucR mutation resulted in increased activities under P_i-sufficient conditions. P_i starvation in a mucR mutant or wggR/mucR double mutant did not further increase the activities of the reporter gene fusions. These results confirm the previously suggested role of WggR and PhoB as activator of the wga and wge operons under low-P_i conditions and of MucR as an inhibitor of these genes independent of the P_i concentration (2, 4, 5, 47).

The activities of the proximal and distal promoters preceding wgaA and wgeA were investigated separately. The proximal fragments contained the proximal TSS and the proximal PHO box, while the distal fragments carried the distal TSS and the distal PHO box, as well as binding sites of WggR and MucR. High β-galactosidase activities were conferred by the proximal promoters on fragments Pf-A1A2 and Pf-E1E6 in the wild type and in all mutant backgrounds tested in high and low P_i conditions. Comparison of the different mutant backgrounds showed no clear difference in the activities mediated by these fragments. This indicates a constitutive activity of the proximal promoters if the upstream sequences are missing. These findings are in agreement with the EMSA experiments, which showed that the proximal fragments did not bind to either His₆-WggR or Flag-MucR protein (Fig. 4).

In the phoB and wggR mutants, fusion of the distal fragments Df-A3A4 and Df-E3E5 to the lacZ reporter gene conferred only low β-galactosidase activities. In the wild type, the induction of β-galactosidase activity under P_i starvation was lost, which indicates that the distal promoter is not activated under low-P_i conditions or that the predicted distal PHO box is not sufficient for such an activation. Fragment Df-A4A8 (Fig. 2) carrying both promoters preceding wgaA but missing the PHO box-like sequence downstream of the proximal TSS did not exhibit induction of promoter activities under P_i starvation (data not shown). This suggests that a sequence motif downstream of the proximal TSS, probably the putative PHO box, is required for PhoB-controlled activation of wga transcription.

In both P_i-sufficient and P_i-limiting conditions, the mucR mutation enhanced expression of lacZ governed by the distal fragments. In agreement with the binding of Flag-MucR to the distal fragments (Fig. 4), this implies that MucR has an inhibitory effect on the activity of the distal promoters. In the mucR mutant, the distal wgaA promoter fragment Df-A3A4 displayed a stronger promoter activity under high-P_i conditions than under low-P_i conditions, implying that this promoter was negatively affected under P_i-limiting conditions in this genetic background. The wggR/mucR double mutation further enhanced the activities of the distal promoters, suggesting that WggR may have an inhibitory effect on the activity of these promoters in the absence of MucR.

Regulation of the wgd (expD) operon.

The intergenic region between wggR and wgdA differs from the upstream regions of the wga and wge operons (Fig. 2). Instead of two sense-oriented putative PHO boxes, one sense- and one reverse-oriented PHO box-like sequence precede wgdA, and the distal TSS is not located downstream but upstream of the WggR binding site. Nevertheless, the entire fragment Ef-D1D6 (wggR-wgdA) displayed an expression pattern corresponding well to those of the entire wga and wge promoter regions (Fig. 4). Maximal PhoB-dependent induction under P_i-limiting conditions was only observed for the entire fragment and required an active wggR gene.

In agreement with the promoter activities of the proximal fragments of the wgaA and wgeA promoter regions, the proximal fragment Pf-D1D2 also showed constitutive promoter activity (Fig. 4). Although the WggR binding site is located downstream of the distal TSS, the distal fragment Df-D5D6 exhibited expression patterns similar to those of the distal promoter fragments of wgaA and wgeA. MucR inhibited the promoter activity and, in the absence of MucR, WggR had a negative effect (Fig. 4).

Regulation of wggR.

In the wggR upstream region, the WggR binding site and the overlapping putative PHO box are located downstream of the proximal TSS (Fig. 2). A putative reverse-oriented PHO box overlaps the proximal promoter (Fig. 2). It was therefore not possible to separate the proximal promoter from this second PHO box-like sequence.

The entire fragment Pf-G1G5 (wgdA-wggR) and the shorter fragment Pf-G1G4 that lacks the distal TSS displayed patterns of promoter activities similar to those of the entire promoter regions of the structural EPS II biosynthesis genes, but the Pf-G1G4 fragment generally conferred lower expression levels (Fig. 4). Both WggR and PhoB were required for maximal transcriptional induction under P_i limitation if MucR was present, but a low level of induction was still visible in a phoB mutant. The mucR mutation increased the promoter activity in low- and high-P_i conditions but P_i-dependent regulation still controlled expression. Compared to the mucR mutant, the double wggR/mucR mutant in P_i-sufficient conditions showed a similar promoter activity that was further enhanced by a low P_i concentration. This suggests that WggR did not act as a repressor of the proximal promoter in absence of MucR, as was the case for the distal promoters of the other operons. Transcriptional control of the proximal promoter by WggR and MucR is in agreement with the binding of Flag-MucR and His₆-WggR to the proximal and entire fragments found in the EMSA (Fig. 4).

The distal fragment Df-G5G6 conferred only low levels of β-galactosidase activity that was not significantly affected by the presence of the regulators WggR, MucR, or PhoB, a finding which is in agreement with the lack of a PHO box-like sequence and binding sites for WggR and MucR on this fragment. Corresponding results were obtained for fragment Df-G5G8 (Fig. 2) that included the reverse-oriented PHO box-like sequence downstream of the TSS (data not shown).

In conclusion, MucR seems to inhibit the proximal promoter directing transcription of wggR by DNA binding. It appears as an antagonist of WggR and PhoB that are probably collaborating to activate this promoter, although PhoB alone may be able to activate wggR transcription to a low extent under P_i starvation.

DISCUSSION

Phosphate-dependent regulation of the EPS II biosynthesis gene cluster is cooperatively mediated by WggR and PhoB.

Zhan et al. (58) first reported the induction of EPS II biosynthesis under P_i limitation in S. meliloti. PhoB controlling the Pho regulon by binding to PHO boxes was found to be required for this induction, and deletion of the regulatory gene wggR reduced stimulation under P_i starvation (47). Based on PHO boxes, known in S. meliloti and in E. coli, Yuan et al. (56) identified 96 putative Pho regulon members in S. meliloti, including wggR. Furthermore, Krol and Becker (33) found wggR, wgdB, and wgeA among 98 genes that were significantly induced in a phoB-dependent manner in P_i-limited cells, and two PHO box-like sequences were reported upstream of four EPS II biosynthesis operons (47). In the present study, we sought to extend the knowledge about the regulatory function of PhoB and WggR in the control of EPS II biosynthesis. Since the sequenced reference strain S. meliloti 1021 carries a mutation in pstC leading to a partially constitutively active PhoB (33, 56), the closely related strain S. meliloti 2011 was selected as wild type for this investigation.

Our findings suggest that transcription of each EPS II biosynthesis operon is directed by two promoters and that PhoB-mediated regulation requires the complete intergenic regions, including both promoters. Interestingly, both wggR and phoB were required for maximal induction of transcription under P_i starvation in the presence of MucR. This corresponds well to a previous study based on insertion of a lacZ reporter gene in the wgaA, wgdA, and wgeB (expE2) coding regions in a wggR mutant background (47). The requirement of both proteins may be explained by activation of wggR transcription by PhoB under P_i limitation. However, a remaining weak induction of the wge operon in the absence of WggR under P_i starvation, the presence of WggR and putative PhoB DNA-binding sites in the promoter regions, and the requirement of PhoB and WggR for maximal induction of transcription in P_i-starved cells imply that both proteins function cooperatively in transcriptional activation. Maximal transcriptional induction of wggR in P_i-starved cells was also dependent on the presence of functional wggR and phoB genes. The assumption that both regulators cooperate is further supported by the effect of multiple copies of wggR driven by a constitutive promoter (47). Multiple copies resulted in a weak induction of expression of the EPS II biosynthesis genes that was further enhanced under P_i limitation. This further stimulation was also observed in a strain that was deleted for the native copy of wggR implying that positive regulation of wggR by activated PhoB could not exclusively be responsible for this effect. WggR may be required to compensate for the low degree of conservation of the PHO box-like sequences upstream of the structural EPS II biosynthesis genes.

By analogy to E. coli, we assume that phosphorylated PhoB binds as a dimer to each PHO box, forms a complex, and activates transcription by direct interaction with the σ subunit of the RNA polymerase (RNAP) (37). Thus, two plausible mechanisms may be envisaged (3). Following the first mechanism, WggR may function as a class I activator by recruiting the α-subunit of the RNAP, whereas PhoB functions independently as a class II activator recruiting the σ-subunit. In the second mechanism, the partially bounded PhoB requires the integration of WggR into the protein complex to initiate transcription. In contrast to the first mechanism, the second relies on protein-protein interactions.

Phosphate-independent inhibition of the distal promoter by MucR and WggR.

The zinc finger motif protein MucR was found to positively affect biosynthesis of EPS I and to interact with DNA sequences upstream of exoH and exoY (13). We demonstrated here that MucR is also able to specifically bind to the DNA region upstream of wgaA, wgeA, wgdA, and wggR. Although the binding site was localized in the distal part of the intergenic regions upstream of the EPS II biosynthesis operons and three other binding sites are known in S. meliloti (13, 14), a consensus sequence could not be established. Binding was found in the vicinity of the distal TSS of the EPS II biosynthesis operons that correlates well with the position of a binding site of an inhibitor.

A high MucR protein concentration was required for efficient reduction of DNA mobility in EMSA experiments. This may be explained by only partially active purified protein or by the formation of multimers prior or upon binding. The latter possibility correlates with the size of the observed mobility shift. Eukaryotic zinc finger proteins contain several zinc finger motifs. Each finger binds to 3 to 4 bp, and only the association of several fingers results in specific DNA binding (42). These regulators are present in large amount and cause significant alteration of the DNA conformation. In contrast, MucR contains only a single zinc finger motif. It is therefore tempting to speculate that MucR may inhibit transcription by binding to several short DNA targets as a multimer.

Although extra copies of wggR had a modest activating effect on transcription of the EPS II biosynthesis operons (47) and WggR in concert with PhoB induced transcription, a negative effect of WggR on the activity of the distal promoter of the structural EPS II biosynthesis genes was observed in the absence of MucR. This implies that WggR rather functions as an activator of the proximal than the distal promoter of the complete promoter regions.

Model of the transcriptional regulation of EPS II biosynthesis operons.

Our studies imply that transcriptional regulation of the operons comprising the structural EPS II biosynthesis genes is governed by two promoters that are differentially controlled by PhoB, WggR, and MucR. PhoB and WggR seem to cooperatively activate at least the proximal promoter under P_i-limiting conditions. Separated from the distal part of the intergenic regions, this promoter was constitutively active, indicating that it underlies repression mediated by the binding of MucR to the distal region. This binding appears to also inhibit the activity of the distal promoter. Moreover, a supplementary inhibitory effect of WggR on the distal promoter was found. These inhibitory effects were observed in both P_i-sufficient and P_i-limiting conditions.

Based on these observations, we propose a regulatory model for the transcriptional regulation of the structural genes grouped in the wga, wge, and wgd operons (Fig. 5). In this model, these genes are transcribed at a very low level under P_i-sufficient conditions in the presence of MucR that inhibits the activity of both promoters. Consequently, only traces of EPS II are synthesized. Under the same condition, but in a mucR mutant, the inhibitory effect of MucR is missing and WggR binds to its target, slightly inhibiting the activity of the distal promoter. Both promoters are active due to the lack of the inhibitory binding of MucR. This leads to a strong stimulation of transcription, which is further enhanced in the absence of WggR. The modest activation caused by extra copies of wggR (47) points to the possibility of binding of WggR in the presence of MucR if WggR is present in excess and may point to a weak activation of the proximal promoter by WggR even if phosphorylated PhoB is not available or only available at a very low level.

FIG. 5.

Model for transcriptional regulation of the structural EPS II biosynthesis genes by MucR, WggR, and PhoB. Black boxes represent potential PHO boxes, the open box marks the WggR binding site, and the hatched box represents the putative MucR binding site. Arrows indicate the transcription starts. Small gray boxes represent the −10 and −35 boxes for each promoter. The thickness of the arrow beginning at the transcription start marked by +1 indicates the amount of transcript in each situation. The circled P indicates the phosphorylation status of PhoB.

If the cells encounter P_i starvation, phosphorylated PhoB and WggR bind to the promoter regions. This may result in an altered DNA conformation and recruitment of the RNAP that initiates transcription at the proximal start site. The resulting transcription remains, however, lower than in the mucR mutant, indicating that MucR still exerts its inhibitory effect on the activity of the distal promoter only or on both promoters.

The promoter region and regulation of wggR differs from that of the structural genes. The main differences were the position of the WggR binding site and an overlapping PHO box-like sequence downstream of the proximal TSS, strong further enhancement of transcription in a mucR mutant under P_i limitation compared to P_i-sufficient conditions, and the weak constitutive activity of the distal promoter that was not affected by MucR. Instead, MucR bound to the proximal region and inhibited the proximal promoter. In spite of these differences, the requirement of WggR for maximal PhoB-dependent induction of transcription and a remaining weak induction of wggR promoter activity in absence of WggR are common features of the regulation of wggR and the structural genes.

The activities of the regulators WggR, PhoB, and MucR may allow fine-tuning of EPS II biosynthesis gene expression and adaptation to different environmental conditions by affecting the composition of extracellular polysaccharides. In addition, at least one other regulator controls transcription. This is the LuxR-type regulator ExpR that is involved in quorum sensing in S. meliloti (38, 44). Since S. meliloti 2011 lacks an active expR gene, we have studied the regulatory function of the other regulators independent of an activated quorum-sensing system. ExpR binding sites were identified in the promoter regions of EPS II biosynthesis genes, and it was shown that transcriptional activation by ExpR overrules the inhibitory effect of MucR but also requires a functional wggR gene for this action (M. McIntosch, E. Krol, and A. Becker, unpublished data). This suggests that WggR is an important player that mediates regulation by ExpR and PhoB.