Binding Site Determinants for the LysR-Type Transcriptional Regulator PcaQ in the Legume Endosymbiont Sinorhizobium meliloti

Abstract

LysR-type transcriptional regulators represent one of the largest groups of prokaryotic regulators described to date. In the gram-negative legume endosymbiont Sinorhizobium meliloti, enzymes involved in the protocatechuate branch of the β-ketoadipate pathway are encoded within the pcaDCHGB operon, which is subject to regulation by the LysR-type protein PcaQ. In this work, purified PcaQ was shown to bind strongly (equilibrium dissociation constant, 0.54 nM) to a region at positions −78 to −45 upstream of the pcaD transcriptional start site. Within this region, we defined a PcaQ binding site with dyad symmetry that is required for regulation of pcaD expression in vivo and for binding of PcaQ in vitro. We also demonstrated that PcaQ participates in negative autoregulation by monitoring expression of pcaQ via a transcriptional fusion to lacZ. Although pcaQ homologues are present in many α-proteobacteria, this work describes the first reported purification of this regulator, as well as characterization of its binding site, which is conserved in Agrobacterium tumefaciens, Rhizobium leguminosarum, Rhizobium etli, and Mesorhizobium loti.

In a soil environment, plant-derived aromatic acids represent significant carbon and energy sources. The first step in the metabolism of these compounds involves their conversion into either protocatechuate or catechol, which is subsequently metabolized to tricarboxylic acid intermediates via the β-ketoadipate pathway (18). This metabolic pathway has been documented in many members of the family Rhizobiaceae (23, 35, 38-40), suggesting that the β-ketoadipate pathway is important for the survival of these soil-dwelling microorganisms.

Sinorhizobium meliloti is a gram-negative, soil-dwelling bacterium that participates in a symbiotic relationship with the legume alfalfa through the establishment of nitrogen-fixing root nodules. Enzymes involved in the protocatechuate branch of the β-ketoadipate pathway in S. meliloti are encoded within the pcaDCHGB and pcaIJF operons, which are subject to regulation by products encoded by pcaQ and pcaR, respectively (23). The regulator encoded by pcaQ is a member of the LysR-type transcriptional regulator (LTTR) superfamily, and PcaQ homologues are present in many species of α-proteobacteria (3, 4, 7, 23, 34, 36, 37).

LysR-type regulators comprise one of the largest groups of prokaryotic transcriptional regulators characterized to date; these proteins regulate a diverse range of regulons, including genes whose products are involved in nitrogen and carbon fixation, biofilm formation, the oxidative stress response, bacterial virulence, and the catabolism of various compounds, including aromatic acids (10, 16, 19, 22, 23, 28, 34, 46, 47, 51, 55, 57). LTTR proteins consist of a conserved helix-turn-helix DNA binding motif located in the N-terminal portion of the polypeptide, whereas the C terminus includes an inducer binding site. As a general rule, LTTRs act as transcriptional activators by inducing expression of a target gene(s) upon interaction with a coeffector molecule, although there are also reports of these proteins acting as repressors (9, 19). As transcriptional activators, LTTRs typically associate with two distinct binding sites (47). A recognition binding site (RBS) is often located upstream of the target gene's promoter and may be sufficient to elicit regulator binding even in the absence of a coinducer. Interaction of an LTTR with an activation binding site (ABS), located near the −35 regulatory region of the target gene, generally occurs in the presence of a coeffector and is required to induce target gene expression via interaction with RNA polymerase.

In this report, we describe data from electrophoretic mobility shift assays which indicate that PcaQ binds with high affinity to a sequence upstream of the pcaD promoter. In DNase I footprinting assays, a region protected by PcaQ was located at positions −78 to −45 upstream of the pcaD transcriptional start site. Employing site-directed mutagenesis, we demonstrated that PcaQ recognition and binding of the pcaD regulatory region involve a sequence with partial dyad symmetry (5′-ATAACC-N₄-GGTTAA-3′), as determined by both in vitro binding assays and in vivo expression analyses. By measuring expression of a pcaQ::lacZ fusion in S. meliloti, we also showed that PcaQ participates in negative autoregulation, possibly through interaction with the same binding site necessary for regulation of pcaD expression.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The bacterial strains and plasmids used throughout this study are described in Table 1. Escherichia coli strains were grown aerobically at 37°C in LB broth; S. meliloti strains were grown aerobically at 30°C in LB broth supplemented with 2.5 mM MgSO₄ and 2.5 mM CaCl₂ (LBmc) or in M9 minimal medium (Difco). M9 minimal medium was supplemented with 1.0 mM MgSO₄, 0.25 mM CaCl₂, 1 μg/ml d-biotin, and 10 ng/ml CoCl₂. The following carbon sources were added to M9 minimal medium: 5 mM protocatechuate (Sigma-Aldrich) or 0.5% (vol/vol) glycerol. For E. coli, antibiotics were added at the following concentrations: ampicillin, 50 μg/ml; chloramphenicol, 20 μg/ml; and gentamicin, 10 μg/ml. For S. meliloti, the following concentrations of antibiotics were used: streptomycin, 200 μg/ml; spectinomycin, 200 μg/ml; gentamicin, 60 μg/ml; and rifampin, 20 μg/ml.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant characteristics	Reference or source
S. meliloti strains
Rm1021	Smr derivative of wild-type strain SU47	25
RmG212	Rm1021 lac; Smr	Strain collection
RmP110	Rm1021 with wild-type pstC; Smr	58
RmP134	RmG212 pcaQ::Ω; Smr Gmr	23
RmP137	RmG212(pTH467); Smr Tcr	This study
RmP138	RmP134(pTH467); Smr Gmr Tcr	This study
RmP1676	RmP110 pcaQ::Ω; Spr Smr	This study
Rm5000	SU47 rif-5	13
Plasmids
pUCP30T	Cloning vector; Gmr	GenBank accession no. U33752
pOT1	Broad-host-range gfpuv transcriptional reporter; Gmr	2
pTH467	434-bp EcoRI-XbaI PCR product encompassing pcaDQ intergenic region in pMP220 (pcaQ::lacZ); Tcr	This study
pTH1522	Reporter vector used in construction of S. meliloti reporter gene fusion library; Gmr	8
pFL2211	1.6-kb insert encompassing 781 bp of pcaD and 740 bp of pcaQ in pTH1522; clone obtained from S. meliloti reporter gene fusion library; Gmr	8
pTH2209	194-bp XbaI PCR product extending across pcaD/pcaQ intergenic region in pUCP30T; Gmr	This study
pTH2273	pTH2209 site-directed mutagenesis; T(−63)→G and A(−51)→G; Gmra	This study
pTH2276	184-bp PCR-amplified insert from pTH2209 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study
pTH2282	184-bp PCR-amplified insert from pTH2273 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study
pTH2294	pTH2209 site-directed mutagenesis; A(−58)→G; Gmr	This study
pTH2295	pTH2209 site-directed mutagenesis; A(−57)→G: Gmr	This study
pTH2298	184-bp PCR-amplified insert from pTH2294 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study
pTH2299	184-bp PCR-amplified insert from pTH2295 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study
pTH2336	pTH2209 site-directed mutagenesis; T(−60)→G; Gmr	This study
pTH2337	184-bp PCR-amplified insert from pTH2336 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study
pTH2387	pTH2209 site-directed mutagenesis; T(−59)→G; Gmr	This study
pTH2388	pTH2209 site-directed mutagenesis; A(−72)→G; Gmr	This study
pTH2389	pTH2209 site-directed mutagenesis; T(−71)→G; Gmr	This study
pTH2390	pTH2209 site-directed mutagenesis; A(−70)→G; Gmr	This study
pTH2391	pTH2209 site-directed mutagenesis; A(−69)→G; Gmr	This study
pTH2392	pTH2209 site-directed mutagenesis; G(−61)→C; Gmr	This study
pTH2393	pTH2209 site-directed mutagenesis; G(−62)→C; Gmr	This study
pTH2395	184-bp PCR-amplified insert from pTH2387 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study
pTH2396	184-bp PCR-amplified insert from pTH2388 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study
pTH2397	184-bp PCR-amplified insert from pTH2389 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study
pTH2398	184-bp PCR-amplified insert from pTH2390 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study
pTH2399	184-bp PCR-amplified insert from pTH2391 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study
pTH2400	184-bp PCR-amplified insert from pTH2392 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study
pTH2401	184-bp PCR-amplified insert from pTH2393 transferred into pOT1 to generate pcaD::gfp fusion; Gmr	This study

^aThe positions with respect to the pcaD transcriptional start site are indicated.

Overexpression and purification of PcaQ.

Using S. meliloti Rm1021 genomic DNA as a template, pcaQ was PCR amplified using primers 5′-GTGATACATATGATCGACGCTCGCGTTAAG-3′ and 5′-ACTCGAGGGCCGTCCTCTTTGCTTCC-3′ and was cloned into the expression vector pET-21a (Novagen) via NdeI and XhoI restriction sites to create a C-terminal fusion with a His tag. The location of the hexahistidine tag at the carboxyl terminus of the protein was chosen to minimize any influence that the foreign tag might have upon the ability of PcaQ to bind DNA, as has been demonstrated with other LTTRs (5, 30). DNA sequencing of the cloned regions confirmed the absence of mutations within the 937-bp pcaQ gene fragment, and the designated plasmid, pTH1979, was then transformed into E. coli BL21(DE3)(pLysS) (Stratagene).

E. coli strain M924 [BL21(DE3)(pLysS)(pTH1979)] was subcultured into 100 to 200 ml of prewarmed LB broth (with appropriate antibiotics) to obtain an optical density at 600 nm (OD₆₀₀) of approximately 0.2 and was grown to an OD₆₀₀ of ∼0.6 to 0.8. Expression of pcaQ was induced with 0.1 mM isopropyl-β-d-thiogalactoside (IPTG) for 3 h at 37°C. Cells were then centrifuged for 20 min at 9,300 × g, and the pellets were stored at −20°C prior to lysis.

Pellets were thawed on ice, resuspended in 4 ml of buffer (50 mM NaH₂PO₄, 300 mM NaCl, 10 mM imidazole; pH 8.0), and sonicated with a model 350 Sonifier (Branson Sonic Power Co.) using five 10-s bursts on ice. The lysate was incubated with 2 ml Ni-nitrilotriacetic acid (NTA) agarose beads (Qiagen) for 1 h at 4°C before the chromatography column (Bio-Rad) was loaded. The column was washed with 20 to 100 mM imidazole in buffer (50 mM NaH₂PO₄, 300 mM NaCl; pH 8.0), and the protein was eluted with 250 mM imidazole.

For size exclusion fast protein liquid chromatography (FPLC), eluant from the Ni-NTA column was dialyzed into a column buffer (100 mM Tris-HCl, 300 mM NaCl, 1 mM dithiothreitol, 1 mM EDTA; pH 8.0). Six hundred microliters of sample was loaded onto a Sephadex G-200 size exclusion column and eluted at a rate of 0.2 ml/min. Fractions (1 ml) were collected, and the purified protein was visualized by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Pooled fractions were dialyzed into a buffer (50% [vol/vol] glycerol, 20 mM Tris-HCl, 50 mM KCl; pH 8.0) and stored at −20°C.

The molecular mass of PcaQ was determined using a Sephadex G-200 size exclusion column calibrated using the following molecular mass standards: catalase (232 kDa), aldolase (158 kDa), albumin (67 kDa), ovalbumin (43 kDa), chymotrypsin A (25 kDa), and RNase A (13 kDa). Blue Dextran 2000 was used to calculate the void volume of the column. Molecular mass standards were run in sets of three using a final volume of 600 μl.

Electrophoretic mobility shift assays.

PCR-amplified probes and annealed synthetic oligonucleotides were purified using polyacrylamide or agarose gels prior to labeling. For equilibrium dissociation constant (K_D) determination and experiments comparing PcaQ binding activities of wild-type and mutant binding sites, assays were performed with 194-bp PCR amplified products spanning the intergenic region, using primers 102 (5′-CGCTCTAGACAAATGTCTGCAGATGG-3′) and 103 (5′-ATTCTAGAGATAGTGAATCGTGACGTCG-3′) with either pTH2209 or a derivative as the template. Probes were 5′ end labeled using [γ-³²P]ATP (Perkin Elmer) and T₄ polynucleotide kinase (New England Biolabs). Binding reaction mixtures (total volume, 15 μl) contained 10,000 cpm of one of the labeled DNA probes and 500 ng of herring sperm nonspecific competitor DNA in a buffer containing 20 mM Tris-HCl (pH 8.0), 50 mM KCl, 5 mM MgCl₂, 1 mM dithiothreitol, 200 ng/ml bovine serum albumin, and 4% (vol/vol) glycerol. Purified PcaQ was added as specified below, and the binding reaction mixture was incubated on ice for 15 min and then incubated at room temperature for 25 min. Reaction mixtures were loaded onto a 6% nondenaturing polyacrylamide gel (200-bp probes) or 8% nondenaturing polyacrylamide gel (45- to 78-bp oligomers). The gels were run at 150 V for 5 min and at 80 V at room temperature for the remaining time. The gels were dried onto filter paper, exposed on a phosphorimager screen, and visualized. To determine the K_D, mobility assays were performed using eight concentrations of purified PcaQ (range, 0 to approximately 25 nM) in triplicate for a total of 24 assays; this experiment was performed twice.

DNase I footprinting reactions.

Primer 103 was 5′ end labeled using [γ-³²P]ATP (Perkin Elmer) and T₄ polynucleotide kinase (New England Biolabs) and was purified using a QIAquick nucleotide removal kit (Qiagen). Labeled primer was incorporated into the probe via PCR amplification using pFL2211 as a template and unlabeled primer 102. The PCR product was resolved on an 8% polyacrylamide gel, and purified product was electroeluted from the gel and purified further using a QIAquick PCR purification kit (Qiagen). Binding reactions were performed using ∼45,000 cpm of labeled DNA probe, 100 ng poly(dI-dC) (as a nonspecific competitor DNA) and binding buffer as described above for the electrophoretic mobility shift assays. DNase I (0.01 U; Invitrogen) was incubated in the presence or absence of 100 to 500 ng purified PcaQ (approximately 15 to 70 nM) for 2 min at room temperature (22°C). Reactions were stopped by addition of 10 volumes (total volume, 500 μl) of buffer PN (Qiagen), and DNase I-digested DNA was immediately purified with a QIAquick nucleotide removal kit used according to the manufacturer's directions. DNA was eluted using nuclease-free water (50 μl), and samples were concentrated with an Eppendorf Vacufuge at 30°C for 20 min. Pellets were resuspended in loading buffer, and samples were resolved on an 8% polyacrylamide (7 M urea) sequencing gel. Sequencing reactions were performed using primer 103 and a Sequenase version 2 DNA sequencing kit (USB) with plasmid pFL2211 as the template.

Site-directed mutagenesis.

Platinum Pfx DNA polymerase (Invitrogen) was used according to the supplier's protocol. Briefly, the reaction mixtures contained each deoxynucleoside triphosphate at a concentration of 0.3 mM (10 mM stock solution), 1 mM MgSO₄, 0.3 μM primer, 2× Pfx amplification buffer, 50 ng template DNA (pTH2209), and 2 U Platinum Pfx polymerase. To obtain a template for site-directed mutagenesis experiments, plasmid pTH2209 was constructed as follows. Primers 102 and 103 were used to PCR amplify a 194-bp product spanning the intergenic region (positions −122 to 55), which was cloned into pUCP30T using XbaI. Following amplification, 20 U DpnI (New England Biolabs) was added directly to the amplification mixture, which was incubated for 2 h at 37°C. Product DNA was purified using QIAquick spin columns (Qiagen) and was transformed into competent E. coli DH5α; transformants were selected for Gm^r, and all plasmids were sequenced to verify the presence of site-directed mutations (MOBIXlab, McMaster University, Hamilton, Ontario, Canada).

To construct the gfpuv reporter fusions, pUCP30T derivatives (each bearing a site-directed mutation) were used as template DNA for PCR amplification using primers 5′-TGTCTGCAGATAGTGAATCGTGACGTCG-3′ and 5′-GTATCTAGACAGATGGCGAAACTTAACGC-3′. PCR products were cloned into pOT1 (2) using XbaI and PstI sites and standard molecular biology techniques to obtain a pcaD::gfp fusion. All plasmids were sequenced to confirm the presence of mutations and the orientation of transcriptional fusions with gfpuv.

Construction of an S. meliloti pcaQ::Ω strain.

An S. meliloti pcaQ mutant was required for reporter enzyme assays (GfpUV); however, the only strain available was gentamicin resistant, and thus it was necessary to construct an additional strain which lacked resistance to this antibiotic. Accordingly, an Ω cassette encoding streptomycin/spectinomycin resistance from pHP45Ω (42) was introduced into a PstI site located 35 bp downstream of the predicted translational start site of pcaQ as follows. The Ω cassette was PCR amplified and cloned into plasmid pTH1577 using PstI, yielding pTH1958. pTH1577 has been described previously (23) and consists of an 893-bp fragment centered on the PstI site located within pcaQ and cloned into the suicide vector pJQ200 uc1 (43). pTH1958 was transferred via conjugation into rifampin-resistant S. meliloti strain Rm5000, and recombinants were selected by plating preparations onto LB agar supplemented with rifampin and gentamicin. A single Gm^r colony was grown overnight in LBmc in the absence of antibiotic selection, and the resulting culture was plated onto LB agar supplemented with 5% sucrose and spectinomycin. Sp^r sucrose-resistant colonies were patched onto LB agar containing gentamicin to confirm excision of the Gm^r suicide plasmid. The pcaQ::Ω allele was transferred from Rm5000 to RmP110 via transduction by selecting for Sp^r transductants. The resulting RmP110 pcaQ::Ω strain was designated RmP1676. As expected, RmP1676 exhibited a Pca⁻ phenotype when it was plated on M9 minimal medium with protocatechuate as a sole carbon source.

GfpUV assays.

Plasmids were transferred by conjugation from donor E. coli DH5α into S. meliloti recipient strains RmP110 and RmP1676 using triparental mating with E. coli helper strain MT616. Streak-purified transconjugants were used in GfpUV assays as follows. Overnight S. meliloti LBmc cultures were washed and subcultured into M9 minimal salts with a carbon source and gentamicin as specified below to obtain an OD₆₀₀ of ∼0.2 to 0.5. Cultures were incubated at 30°C for 4 to 6 h before they were harvested for enzyme assays. GfpUV fluorescence was quantified by dividing each emission output by its OD₆₀₀; all assays were performed in triplicate for each experiment, and experiments were performed a minimum of three times.

β-Galactosidase assays.

Plasmids were transferred by conjugation into S. meliloti recipient strains RmG212 (Rm1021 lac) and RmP134 (Rm1021 lac pcaQ::Ω) (23), and transconjugant colonies were streak purified prior to use in enzyme assays. β-Galactosidase enzyme assays were performed as previously described (23), and enzyme activities were calculated as described by Miller (27).

RESULTS

Overexpression and purification of PcaQ.

The LTTR PcaQ is known through genetic analyses to participate in the regulation of expression of the pcaDCHGB operons in S. meliloti and Agrobacterium tumefaciens (23, 34). The transcription of the pcaQ gene is divergent from the transcription of the pcaDCHGB operon (Fig. 1A), and we recently mapped the pcaD transcriptional start sites to G and C residues located 14 and 15 nucleotides upstream of the pcaD start codon (23). For simplicity, however, in this paper nucleotide positions are reported with respect to the upstream transcriptional start site identified (e.g., a C residue located 15 nucleotides upstream of pcaD). To facilitate the analysis of the pcaD promoter, we wished to purify PcaQ through use of a hexahistidine tag located at the C terminus of the protein (Fig. 1B). Accordingly, pcaQ was overexpressed in E. coli, and the purified protein (PcaQ·His) eluted from a size exclusion chromatography column as a 149-kDa protein peak, indicating that this regulator exists in solution as a tetramer (Fig. 1C). We determined that pcaQ·His complements the regulatory phenotype of a pcaQ strain of S. meliloti in trans, thus indicating that PcaQ·His is capable of activating transcription in vivo (data not shown).

FIG. 1.

(A) Schematic diagram of the 94-bp intergenic region located between pcaQ and pcaDCHGB on the pSymB megaplasmid of S. meliloti. The predicted translational start codons of pcaQ and pcaD are italicized, and the directions of translation are indicated by arrows. The two pcaD transcriptional start sites are identified by bold type, and the direction of transcriptions is indicated by bent arrows; inferred −10 and −35 hexanucleotide regions associated with pcaD are underlined. (B) Overexpression and purification of S. meliloti PcaQ carrying a C-terminal His tag. Protein samples were visualized by sodium dodecyl sulfate-polyacrylamide gel electrophoresis with a 10% polyacrylamide gel, followed by staining with Coomassie brilliant blue. Lane 1, BenchMark protein ladder (Invitrogen); lane 2, crude cell lysate obtained from uninduced E. coli strain M924; lane 3, crude cell lysate obtained from E. coli strain M924 induced with 0.1 mM IPTG; lane 4, flowthrough collected from Ni-NTA column; lane 5, eluate collected from 20 mM imidazole wash; lane 6, eluate collected from 50 mM imidazole wash; lane 7, eluate collected from 100 mM imidazole wash; lane 8, eluate collected from 250 mM imidazole wash; lane 9; pooled eluate obtained following size exclusion FPLC. (C) Standard curve used to estimate the molecular mass of purified PcaQ, established using globular proteins as described in Materials and Methods. The elution coefficient (K_av) obtained for PcaQ is 0.229, corresponding to an estimated PcaQ molecular mass of 149 kDa.

Purified PcaQ binds upstream of the pcaD promoter.

PcaQ regulates expression of the pcaDCHGB promoter in S. meliloti and A. tumefaciens, inducing expression of the operon in cells grown in the presence of protocatechuate (23, 34). To determine whether PcaQ recognizes and binds the pcaD promoter region, electrophoretic mobility shift assays were performed using purified PcaQ. A 194-bp probe spanning the pcaDQ intergenic region (positions −122 to 55) was shown to bind PcaQ in the absence of a coinducing metabolite (Fig. 2). The apparent K_D of PcaQ was determined to be 0.54 nM for the pcaD promoter region, indicating that there was a high-affinity interaction even in the absence of a coeffector molecule. As a negative control, a probe consisting of the pcaIJF upstream region was also tested for PcaQ binding; expression of these genes is not regulated by PcaQ, and thus the upstream sequence should not be recognized by the regulator (23). As expected, addition of up to 50 ng of PcaQ (24 nM) did not result in a shift of the pcaIJF promoter probe (data not shown), confirming that the interaction observed between PcaQ and the pcaD promoter region is specific.

FIG. 2.

Electrophoretic mobility shift assay for PcaQ binding to the S. meliloti intergenic region. A 194-bp PCR amplified probe was end labeled using [γ-³²P]ATP, and the resulting probe was incubated in the presence of increasing amounts of purified PcaQ prior to resolution on a 6% nondenaturing polyacrylamide gel. Each assay mixture contained 500 ng herring sperm DNA as a nonspecific competitor. Lanes 1 to 8 contained 0, 0.2, 0.5, 1.0, 2.5, 4.9, 12.3, and 24.7 nM PcaQ, respectively.

DNase I footprinting analysis of PcaQ binding.

LTTRs are known to recognize sequences with a TN₁₁A core motif (15, 47). Seven such motifs are present within the 94-bp pcaDQ intergenic region, and it was therefore necessary to perform DNase I footprinting experiments to identify specific sequences involved in PcaQ binding (Fig. 3A). PcaQ protected nucleotides located upstream of the pcaD promoter, with a footprint located at approximately positions −78 to −45 bp with respect to the pcaD transcriptional start site (Fig. 3C). These results are in agreement with previously published descriptions of other LysR-type regulators, which often bind upstream of the promoters subject to their regulation (14, 16, 17, 19, 21, 22, 47, 49, 51). To confirm that the area protected by PcaQ was sufficient for PcaQ recognition and binding, mobility shift assays were performed using oligonucleotide probes including or excluding the protected region (Fig. 3B). These assays confirmed that the region at positions −87 to −45 (Fig. 3B, lanes 1 to 4) is necessary and sufficient for PcaQ binding in vitro, consistent with results obtained in the DNase I footprinting experiments. In addition, a probe extending from position −83 to position −6 (lanes 9 to 12) was shifted by addition of purified regulator. In contrast, binding was not detected using a probe extending from position −51 to position 6 (lanes 5 to 8) in the presence of up to 48 nM PcaQ, suggesting that the primary (or only) binding site is located at positions −83 to −51 upstream of the pcaD transcriptional start site.

FIG. 3.

Identification of the sequence involved in PcaQ binding within the S. meliloti pcaDQ intergenic region. (A) A 194-bp labeled probe was subjected to DNase I digestion in the presence and absence of purified PcaQ. Lanes G, A, T, and C contained corresponding nucleotides as determined with sequencing reactions performed using pFL2211 as a template. Lane 1, probe digested with 0.01 U DNase I at room temperature for 2 min; lane 2, probe in the presence of 100 ng PcaQ (14.4 nM); lane 3, probe in the presence of 500 ng PcaQ (72.2 nM). Brackets 1, 2, and 3 correspond to probes used in subsequent mobility shift assays (panel B). The arrows on the right indicate regions of hypersensitivity. (B) Electrophoretic mobility shift assays performed to confirm the location of a putative PcaQ binding site as determined by a DNase I footprinting experiment. Group 1 (lanes 1 to 4), group 2 (lanes 5 to 8), and group 3 (lanes 9 to 12) assays were performed using probes corresponding to the regions at positions −87 to −45, −51 to 6, and −83 to −6, as indicated by brackets in panel A. The following amounts of purified PcaQ were added to each group of assays: 0 ng, 25 ng (12 nM), 50 ng (24 nM), and 100 ng (48 nM). (C) Sequence analysis of a region protected by PcaQ from DNase I digestion. The sequence upstream of pcaD included −35 and −10 hexanucleotide promoter regions (underlined); pcaD transcriptional start sites (G and C) are indicated by arrows bent in the direction of transcription; and the predicted translational start codons of pcaD and pcaQ are indicated by italics. The 34-nucleotide region protected by PcaQ from digestion is enclosed in a box, whereas bold type indicates the nucleotides identified by alignment as conserved residues (Fig. 4) upstream of pcaD in A. tumefaciens, R. etli, M. loti, and R. leguminosarum. The bracket indicates a “TN₁₁A” motif within the conserved sequence, and the arrows above the sequence indicate DNase I hypersensitive sites identified during DNase I footprinting experiments. The arrows below the sequence indicate inverted repeats within the PcaQ binding site.

Identification of a motif that is required for PcaQ binding in vivo.

It was previously demonstrated that an A. tumefaciens pcaD::lacZ transcriptional fusion exhibits protocatechuate-inducible expression in several related species, including S. meliloti (37). Accordingly, it is likely that PcaQ binding sites are conserved between these species, and alignment of the intergenic regions of S. meliloti, Rhizobium leguminosarum, A. tumefaciens, Mesorhizobium loti, and Rhizobium etli was performed to facilitate identification of possible conserved binding sites in the footprint region (Fig. 4). Examination of the alignment revealed two distinct areas that were conserved in all species; one of these areas encompasses and surrounds the −35 hexanucleotide promoter region associated with the S. meliloti pcaD gene and is located outside the region identified as being necessary and sufficient for PcaQ binding, as determined by our in vitro analyses. The second conserved region is comprised of two halves of an AT-rich sequence with partial dyad symmetry (5′-ATAACC-N₄-GGTTAA-3′) that includes a single “TN₁₁A” motif. This region is located upstream of the predicted −35 hexanucleotide region (positions −72 to −57) and falls within the region protected from DNase I digestion.

FIG. 4.

Alignment of pcaDQ intergenic regions of S. meliloti, R. leguminosarum, A. tumefaciens, M. loti, and R. etli. Residues conserved in all species are indicated by asterisks, and the bracket indicates the region protected from DNase I digestion by PcaQ. Positions targeted for site-directed mutagenesis in S. meliloti are enclosed in a box. Within the S. meliloti sequence, −35 and −10 hexanucleotide regions are underlined, and the pcaD transcriptional start sites are indicated by arrows bent in the direction of transcription. A consensus sequence (5′-ATAACYCCNNGGTTAAW-3′) was established based on the alignment; uppercase letters indicate nucleotides conserved in all species, and lowercase letters indicate nucleotides conserved in all but one species. The alignment was constructed using ClustalW (52; http://www.ebi.ac.uk/clustalw/).

The upstream region was targeted for site-directed mutagenesis experiments to confirm that the identified motif is required for PcaQ regulation in vivo. Highly conserved positions within the putative PcaQ binding site were systematically mutated by introduction of single point mutations at each position (Fig. 4); A and T residues were replaced with G, whereas G residues were replaced with C. The wild-type and mutant binding sites were cloned as 184-bp fragments (positions −112 to 54) into the broad-host-range reporter plasmid pOT1 to generate pcaD::gfpuv transcriptional fusions. These plasmids were conjugated into wild-type and PcaQ⁻ S. meliloti strains RmP110 and RmP1676, respectively, and GfpUV assays were performed to determine whether expression of pcaD::gfpuv was induced by growth in the presence of protocatechuate (Table 2).

TABLE 2.

Analysis of expression of pcaD::gfpuv in S. meliloti

Plasmid	Position(s) of mutation(s)a	GfpUV sp act (SD)b		Induction (fold)
		Uninduced	Induced
pTH2276	None	1,107 (80)	15,748 (316)	14.2
pTH2396	A(−72)G	317 (36)	691 (53)	2.2
pTH2397	T(−71)G	202 (16)	598 (30)	3.0
pTH2398	A(−70)G	198 (26)	1,821 (83)	9.2
pTH2399	A(−69)G	214 (10)	1,507 (29)	7.1
pTH2282	T(−63)G, A(−51)G	981 (48)	14,126 (665)	14.4
pTH2401	G(−62)C	326 (27)	3,596 (19)	11.0
pTH2400	G(−61)C	223 (30)	1,028 (18)	4.6
pTH2337	T(−60)G	264 (14)	720 (32)	2.7
pTH2395	T(−59)G	718 (41)	4,195 (193)	5.8
pTH2298	A(−58)G	343 (20)	419 (21)	1.2
pTH2299	A(−57)G	551 (48)	4,112 (32)	7.4
pOT1	None	208 (90)	158 (10)	0.8

^aPosition with respect to the pcaD transcriptional start site.

^bThe values are the averages for three independent cultures of S. meliloti strains subcultured into M9 minimal medium with 0.5% glycerol and with or without 5 mM protocatechuate.

Expression of the pcaD promoter (as determined by GfpUV specific activity) was induced in cells of wild-type S. meliloti strain RmP110 grown in the presence of protocatechuate (pTH2276) (Table 2). Inducible expression of the pcaD promoter was reduced upon mutation of any one of the highly conserved nucleotides targeted for mutagenesis (underlined) within the putative PcaQ binding site (5′-ATAACC-N₄-GGTTAA-3′) (Fig. 4). Single point mutations introduced at positions A(−72)G (pTH2396) and A(−58)G (pTH2298) yielded particularly severe reductions in inducible expression, resulting in 87 and 90% decreases in activation, respectively, compared to the wild-type regulatory region. In contrast, mutations introduced into other positions [such as G(−62)C and A(−57)G] did not have such a severe impact on the regulated expression of pcaD::gfpuv; in these instances, expression was induced by growth with protocatechuate at levels approaching that exhibited in the wild-type control. As expected, pcaD::gfpuv expression in all instances was not induced in a PcaQ⁻ background, despite the addition of protocatechuate to the growth medium (data not shown).

An additional “TN₁₁A” motif was identified at positions −63 to −51 from the transcriptional start site of the S. meliloti pcaD promoter. Simultaneous introduction of two mutations in the motif [T(−63)G and A(−51)G; pTH2282] yielded protocatechuate-inducible expression comparable to that observed with the wild-type sequence, indicating that these nucleotides are not important for PcaQ-mediated regulation. The lack of conservation of this region between species is consistent with these data (Fig. 4).

Identification of a motif that is required for PcaQ binding in vitro.

We have shown in vivo that introduction of mutations within the putative PcaQ binding site (5′-ATAACC-N₄-GGTTAA-3′) negatively affects the regulation of pcaD expression in S. meliloti. One mechanism by which this may occur is that mutations within this region impede PcaQ recognition and/or binding of the regulatory site; it is also possible that these mutations permit binding of the regulator but inhibit a conformational change in PcaQ that is necessary to elicit a regulated response (i.e., interaction with RNA polymerase). To examine whether mutations within this site also affect PcaQ binding, mobility shift assays were performed to compare the binding of PcaQ to wild-type and mutant regulatory sequences. Representative data are shown in Fig. 5. Densitometry analyses indicated that in all cases PcaQ binding was reduced at least twofold by introduction of mutations within the putative binding sequence compared to the data obtained with the wild-type sequence (Table 3). This was most evident with mutations at positions A(−72)G and A(−58)G, where PcaQ binding to the mutant binding site was decreased more than fivefold compared to the wild type.

FIG. 5.

Effect of site-directed mutations within a putative PcaQ binding site on the ability of PcaQ to bind upstream of pcaD in S. meliloti. Electrophoretic mobility shift assays were performed using probes containing the wild-type pcaD promoter region (lanes 1 and 2) or probes that included single point mutations within a putative PcaQ binding site (lanes 3 to 12), as indicated below the gel. Most assays were performed using 2.5 ng purified PcaQ (1.2 nM); the only exception was a control to which PcaQ was not added (lane 1). WT, wild type.

TABLE 3.

Quantification of PcaQ binding to the pcaD regulatory region in vitro

Position of mutationa	% of bound probe (SD)b
None (wild type)	84 (1)
A(−72)G	13 (<1)
T(−71)G	24 (6)
A(−70)G	57 (4)
A(−69)G	41 (8)
G(−62)C	60 (5)
G(−61)C	53 (7)
T(−60)G	33 (<1)
T(−59)G	53 (7)
A(−58)G	16 (4)
A(−57)G	64 (2)

^aThe position with respect to the pcaD transcriptional start site is indicated.

^bThe values are the averages obtained from two independent mobility shift assays.

PcaQ participates in negative autoregulation.

To determine whether PcaQ autoregulates its own expression in S. meliloti, the pcaDQ intergenic region was cloned into the replicating plasmid pMP220 (50), with the pcaQ promoter in the same orientation as the promoterless reporter gene lacZ. The resulting plasmid, pTH467, was then transferred via conjugation into RmG212 (Rm1021 lac mutant) and RmP134 (pcaQ::Ω derivative of RmG212) to examine expression of pcaQ (as measured by β-galactosidase activity). Reporter enzyme assays were performed using cells grown in 0.5% glycerol with and without 5 mM protocatechuate (Table 4). Expression of pcaQ was increased approximately fivefold in the PcaQ⁻ strain compared to wild-type S. meliloti, indicating that pcaQ expression was repressed by its encoded product under the conditions tested.

TABLE 4.

Expression of pcaQ::lacZ in S. meliloti

Strain	Relevant genotype	Growth conditions	β-Galactosidase activity (Miller units) (SD)a
RmP137	Rm1021 lac	Glycerol	272 (8.4)
	(pTH467)	Glycerol + protocatechuate	359 (12.2)
RmP138	Rm1021 lac	Glycerol	1,391 (35.2)
	pcaQ::Ω (pTH467)	Glycerol + protocatechuate	1,635 (18.4)

^aThe values are the averages for three independent cultures of S. meliloti strains subcultured in M9 minimal medium with 0.5% glycerol and with or without 5 mM protocatechuate.

DISCUSSION

Despite the widespread occurrence of members of the PcaQ family in the Gammaproteobacteria, this work describes the first purification of a member of this family and characterization of a PcaQ binding site. As a member of the LTTR superfamily, PcaQ exhibits many traits common to this group of proteins. Size exclusion FPLC performed with partially purified PcaQ (Fig. 1) indicated that this protein likely exists as a tetramer in solution, as has been reported for other LTTRs (26, 29, 47, 48). In particular, the crystal structure of the full-length LysR-type regulator CbnR indicates that this protein exists as a tetramer formed by association of a dimer of dimers, and it is believed that this corresponds to the biologically active form of this protein (29).

Electrophoretic mobility shift assays demonstrated that PcaQ binds to DNA fragments carrying the region from nucleotide −87 to nucleotide −51 upstream of the pcaDCHGB operon (Fig. 2 and 3) in the absence of any coinducing metabolite. Other LTTRs have also been shown to bind DNA in the absence of any coinducing molecule(s) (14, 41, 47, 49). The apparent K_D for PcaQ under these conditions was determined to be 0.54 × 10⁻⁹ M, indicating that there is a high-affinity interaction. This dissociation constant is comparable to that obtained for other LTTRs, whose reported values range from at least 7.0 × 10⁻¹¹ M (41) to 0.9 × 10⁻⁶ M (53).

DNase I footprinting experiments revealed that PcaQ strongly protected a region at positions −78 to −45 relative to the pcaD transcriptional start site. This location upstream of the −35 promoter region is consistent with that observed for several LysR-type proteins (14, 16, 19, 21, 22, 47, 49, 51). Figure 3A also shows a protected sequence spanning from approximately position −40 to position −21. This footprint was more difficult to discern; however, its presence was observed in multiple independent assays. In A. tumefaciens and R. leguminosarum, OccR and NodD have been demonstrated to bind DNA as tetramers, respectively (1, 12). The length of the protected sequence observed in our assays suggests that PcaQ may bind upstream of pcaD as a tetramer; however, we have not confirmed directly that this is the case. A probe encompassing positions −51 to 6 was not shifted by PcaQ in mobility shift assays (Fig. 3B, lanes 5 to 8), suggesting that the region from position −40 to position −21 is not sufficient to permit PcaQ binding. In contrast, the upstream region alone (positions −87 to −45) permitted PcaQ binding in the absence of inducer (Fig. 3B, lanes 1 to 4).

In order to document that conserved nucleotides within the region protected by PcaQ are essential for the regulation of pcaD expression, site-directed mutagenesis experiments were performed. Expression of the pcaD promoter was then monitored in S. meliloti in vivo through use of the reporter protein GfpUV. It is important to note that although the absolute levels of expression of each site-directed mutant were reduced compared to the wild-type sequence, most mutant sequences retained the ability of protocatechuate to up-regulate expression of the pcaD::gfpuv fusion, albeit to a lesser extent than that observed for wild-type sequence (Table 2). That this regulated response was nonetheless dependent upon PcaQ was evident when expression of each fusion in pcaQ mutant strain RmP1676 was examined; without exception, expression of each fusion in this genetic background remained at a comparable low level that was not altered by the presence of protocatechuate (data not shown). At present, we cannot explain the observation that the basal level of pcaD expression (i.e., uninduced level) in many mutants is lower than the wild-type levels.

Similar studies involving mutagenesis of LTTR binding sites have been performed, and often the mutations result in reduced transcription (6, 17, 22, 49). For example, introduction of point mutations within a putative MetR binding site reduced expression of a metH::lacZ fusion (6). However, gel mobility shift assays revealed that only a subset of the mutations affected MetR binding to the regulatory site. In the case of PcaQ, our assays revealed that although all mutations affecting pcaD::gfpuv expression in vivo reduced the PcaQ binding ability (as determined by densitometry analysis [Table 3]), only two mutations, (A(−72)G and A(−58)G), reduced the binding more than fivefold. It is worth noting that the same mutations had the strongest negative impact on pcaD::gfp expression, correlating the ability of PcaQ to bind in vitro with transcription activation. As observed by Byerly et al. (6), the A(−72)G and A(−58)G mutations correspond to positions (underlined) located at the outer edge of the palindromic binding site (5′-ATAACC-N₄-GGTTAA-3′), possibly reflecting the relative importance of outer positions for LTTR binding.

Systematic mutagenesis of a binding site recognized by the LysR-type protein AphB in Vibrio cholerae revealed that the promoter proximal dyad arm may be more important than the distal arm for activation of gene expression (22). Based upon our analyses, positions in both dyad arms of the PcaQ binding site contribute similarly to the transcriptional activation of pcaD expression (Table 2). It is possible that this inconsistency may reflect subtle variations in the manner in which these LysR-type proteins regulate gene expression; however, differences in experimental design (such as the type of nucleotide substitutions used in each study) may also account for this discrepancy.

In A. tumefaciens, the β-ketoadipate pathway intermediates β-carboxy-cis,cis-muconate and γ-carboxymuconolactone act as inducing agents in the presence of PcaQ, resulting in induction of pcaDCHGB expression (34, 36). The identity of comparable coinducing metabolites in S. meliloti is unknown; however, it is likely that the same two pathway intermediates have a similar function in S. meliloti (23). β-Carboxy-cis,cis-muconate and γ-carboxymuconolactone are unstable compounds (31, 32), and synthesis of these metabolites requires the enzymatic activities of protocatechuate 3,4-dioxygenase (PcaHG) and 3-carboxymuconate cycloisomerase (PcaB), respectively. We attempted but were unable to obtain a DNase I footprint with purified PcaQ in the presence of β-carboxy-cis,cis-muconate (data not shown). It is intriguing that multiple coeffector molecules have been found to act in concert with PcaQ, as a recent report has described the synergistic effect on BenM-mediated transcriptional activation exerted by the metabolites benzoate and muconate (5).

In the absence of data that include data from DNA binding assays performed with a coeffector, it is difficult to propose a comprehensive model of PcaQ regulation. Our results indicate that the region from position −78 to position −45 upstream of pcaD in S. meliloti encompasses a high-affinity binding site (or RBS), and a second region (positions −40 to −21) may include a secondary interaction site (or ABS). The ability of PcaQ to bind a probe containing only the putative RBS (spanning positions −87 to −45) but not a probe containing only the putative ABS (positions −51 to 6) is not inconsistent with this hypothesis. For example, OccR interacts with sequence flanking the −35 regulatory region associated with occQ; these binding sites are required for modulation of DNA bending in response to the ligand octopine and do not contribute to the high-affinity binding of this regulator (56). The presence of highly conserved sequences on both sides of the −35 region of pcaD (Fig. 4) is worth noting; these conserved residues may in fact be part of the ABS.

The gene encoding pcaQ is adjacent to, and transcribed divergently from, the pcaDCHGB operon. This spatial organization is a common feature in LTTRs (20, 21, 47, 53), and it has been proposed that this permits transcriptional coupling of the divergent promoters via DNA supercoiling introduced by the transcribing RNA polymerase, as modeled by the ilvYC operon in E. coli (44). In addition, the organization of pcaQ and pcaDCHGB as divergent transcriptional units raises the possibility that PcaQ influences the expression of both sets of genes by occupying the same binding site(s). While many LTTRs repress self-expression (9, 22, 33, 36, 47, 49), this is not always the case, as these proteins may also either act as activators (17, 19) or fail to influence the expression of their own genes (45, 47). In S. meliloti, PcaQ represses its own expression, as reflected by a fivefold increase in pcaQ expression in a PcaQ⁻ strain (Table 4), which is within the range reported for other LTTRs in comparable studies (22, 47). To further examine the regulation of pcaQ expression, we attempted to identify the pcaQ transcriptional start site via primer extension using mRNA isolated from wild-type and PcaQ⁻ strains of S. meliloti; however, we were unable to obtain an extension product. Similar problems have been reported in mapping the start sites of other LTTRs (5, 11, 22), and this is likely due to the relatively low expression of the encoding genes. In the absence of an identified transcriptional start site, we nonetheless noted that the close proximity of the PcaQ binding site and the predicted pcaQ translational start codon (which are separated by 6 nucleotides) suggests that binding of the LysR protein to this site may prevent efficient transcription of pcaQ. While effecting autorepression, LTTRs often bind to sites located within their own coding sequences (22, 33, 54). We examined the nucleotide sequence of pcaQ for candidate binding sites located within the gene, but we were unable to identify any additional sites with strong similarity to the consensus PcaQ binding site. Intriguingly, a scan of the entire S. meliloti genome with a consensus PcaQ binding site resulted in two positive hits. One of the sites is located upstream of pcaD and corresponds to the binding site characterized in this study. The second site is located upstream of an ABC-type transport system whose expression is induced by protocatechuate (24). It seems likely that this transport system is dedicated to the uptake of protocatechuate, and we are pursuing studies to determine whether this system in fact transports protocatechuate and whether expression of this system is regulated by PcaQ.