Probing the Impact of Ligand Binding on the Acyl-Homoserine Lactone-Hindered Transcription Factor EsaR of Pantoea stewartii subsp. stewartii

Abstract

The quorum-sensing regulator EsaR from Pantoea stewartii subsp. stewartii is a LuxR homologue that is inactivated by acyl-homoserine lactone (AHL). In the corn pathogen P. stewartii, production of exopolysaccharide (EPS) is repressed by EsaR at low cell densities. However, at high cell densities when high concentrations of its cognate AHL signal are present, EsaR is inactivated and derepression of EPS production occurs. Thus, EsaR responds to AHL in a manner opposite to that of most LuxR family members. Depending on the position of its binding site within target promoters, EsaR serves as either a repressor or activator in the absence rather than in the presence of its AHL ligand. The effect of AHL on LuxR homologues has been difficult to study in vitro because AHL is required for purification and stability. EsaR, however, can be purified without AHL enabling an in vitro analysis of the response of the protein to ligand. Western immunoblots and pulse-chase experiments demonstrated that EsaR is stable in vivo in the absence or presence of AHL. Limited in vitro proteolytic digestions of a biologically active His-MBP tagged version of EsaR highlighted intradomain and interdomain conformational changes that occur in the protein in response to AHL. Gel filtration chromatography of the full-length fusion protein and cross-linking of the N-terminal domain both suggest that this conformational change does not impact the multimeric state of the protein. These findings provide greater insight into the diverse mechanisms for AHL responsiveness found within the LuxR family.

INTRODUCTION

The ability of a bacterium to thrive in a particular environment is dependent upon its ability to sense and respond to various environmental factors, including osmolarity, pH, temperature, nutrient availability, or even the presence of other bacteria. Cell-cell communication, or quorum sensing, has been characterized in a number of bacteria that produce and respond to signal molecules commonly termed autoinducers. The well-studied Vibrio fischeri quorum-sensing system has served as a model for quorum sensing in Gram-negative proteobacteria. In this system, the LuxI autoinducer synthase protein produces an acyl-homoserine lactone (AHL) autoinducer, N-3-oxo-hexanoyl-l-homoserine lactone (10). At a threshold concentration of AHL, the LuxR receptor protein forms complexes with this ligand. Binding of the autoinducer molecule permits LuxR to take on a functional conformation capable of binding DNA and activating genes involved in the production of bioluminescence (reviewed in references 11, 30, and 32).

Many other Gram-negative proteobacteria have been shown to carry out a similar type of signaling and recognition. For the majority of the LuxR homologues identified in different bacterial species, a model has been proposed whereby AHL binding induces multimerization of the protein, leading to stabilization and DNA binding (18, 19, 26, 36). In the case of TraR from Agrobacterium tumefaciens, in vivo experiments demonstrated that the half-life of TraR is significantly shortened in the absence of its cognate AHL. More specifically, the Clp and Lon proteases play a direct role in regulating the levels of TraR within the cell (37). Whether or not this mechanism of posttranslational regulation is conserved across the entire LuxR family is unknown.

It has been proposed that the LuxR family of proteins can be divided into at least five classes (23). The majority of the LuxR family members are activators that become functional after interacting with AHL. TraR (class I) is stable only when AHL is present during translation and AHL binding is not readily reversible (37). LuxR (class II) is stabilized by the presence of AHL, but AHL binding is reversible (26), while MrtR (class III) from Mesorhizobium tianshanense (33) is stable without AHL but biologically active only when it is present. Class V regulators, such as SdiA from Escherichia coli, have not been shown to dimerize yet are capable of recognizing multiple noncognate AHLs (34). The class IV subset of LuxR homologues, represented by EsaR from Pantoea stewartii subsp. stewartii, was initially described as repressors that function using a reverse mechanism in comparison to the other classes of LuxR homologues (29). The class IV proteins maintain a functional conformation and bind DNA in the absence of their cognate AHL; biological activity is actually hindered by the presence of the AHL N-3-oxo-hexanoyl-l-homoserine lactone, in the majority of cases (reviewed in reference 24). Examples of these homologues, besides EsaR from P. stewartii, include ExpR from Erwinia chrysanthemi and Pectobacterium carotovorum (7, 21), YenR from Yersinia enterocolitica (1, 25), and EanR from Pantoea ananatis (16). Structurally, the members of the EsaR subfamily also share the characteristics of having an extended linker region between the AHL-binding N-terminal domain (NTD) and an extended DNA-binding C-terminal domain (CTD) (23).

At low cell densities, EsaR negatively regulates its own expression and the expression of RcsA, an activator of exopolysaccharide (EPS) biosynthesis genes, which in turn positively regulates EPS production (6, 15, 29). EsaR also positively regulates expression of a small RNA (sRNA) native to P. stewartii (20). In P. stewartii the LuxI homologue EsaI synthesizes N-3-oxo-hexanoyl-l-homoserine lactone (2), and at high cell densities this AHL interferes with the ability of EsaR to regulate gene expression (15, 28). Little biochemical information is available regarding the mechanism whereby AHL regulates the EsaR protein subfamily. However, in comparison to other members of the LuxR protein family, the relative stability and solubility of EsaR in the absence of its cognate AHL make it an attractive model for analyzing the impact of ligand binding.

A simple explanation for the AHL-dependent inactivation of EsaR would be a conformational shift, which causes the protein to become more susceptible to proteolytic processing. Hence, the susceptibility of EsaR to proteases in the absence and presence of its cognate AHL was examined both in vivo and in vitro. The impact of the ligand on the multimeric state of EsaR was also examined in vitro, as the conversion to and maintenance of a dimeric state are critical to the activity of many LuxR homologues (reviewed in reference 23). It is clear that the mode of AHL responsiveness by EsaR has distinctive features in comparison to the majority of quorum-sensing regulators.

MATERIALS AND METHODS

Measurement of EsaR stability in P. stewartii.

Luria-Bertani (LB) broth was inoculated with log-phase cells of P. stewartii DC283 (9) to an optical density at 600 nm (OD₆₀₀) of 0.025 and grown at 30°C. Aliquots were taken at time points during growth, and cells were pelleted by centrifugation and stored at −70°C. LB supplemented with 10 μM N-(β-ketocaproyl)-l-homoserine lactone (AHL) (Sigma-Aldrich) was similarly inoculated with P. stewartii DC283, grown at 30°C to an OD₆₀₀ of 1.5, and the cell pellet was harvested. After resuspension, 50 μg of each protein sample, as determined through a Bio-Rad Bradford protein assay, was analyzed via 12% SDS-PAGE to confirm equal protein loads and subjected to Western immunoblotting with a 1:500 dilution of polyclonal antiserum generated against EsaR (28). Quantitation of bands of interest was performed using Bio-Rad Quantity One 1D-analysis software.

Measurements of EsaR stability in recombinant E. coli.

To measure the half-life of EsaR in E. coli SG22163 (malP::lacI^q) (12), two plasmids, pT7-esaR and pJZ410 (37), were introduced into the strain. Plasmid pT7-esaR was constructed by amplifying the esaR-coding region from pBAD-EsaR (28) template. The reverse primer BAMESR1 (IDT) (5′GGATCCTTACTACCTGGCCGCTGACGCTG3′) included a BamHI site; the upstream primer PCIES1 (IDT) (5′ACATGTTTTCTTTTTTCCTTGAAAATCAAACAATAACGG3′) contained a PciI site overlapping the esaR ATG start codon. The PCR product was cloned into pGEM-T (Promega) and sequenced. A 760-bp PciI-BamHI fragment was digested from the pGEM-T construct and ligated into NcoI-BamHI-digested pMLU115 (26). In the resulting plasmid, pT7-esaR, esaR expression is controlled by the phage T7 promoter. Plasmid pJZ410 (37) has a gene encoding T7 RNA polymerase under the control of a heat-inducible promoter. Cells were cultured in LB containing 100 μg/ml ampicillin (Ap) and 20 μg/ml gentamicin at 28°C to an OD₆₀₀ of 0.4. The cells were shocked at 45°C for 20 min to induce expression of T7 RNA polymerase and were treated with 200 μg/ml rifampin to inhibit the host RNA polymerase. After 20 min at 45°C, the culture was shifted to 30°C for 30 min. At this point, half of the culture volume was transferred to a tube containing 10 μM AHL, and then [³⁵S]methionine was immediately added to both cultures (± AHL) to a final concentration of 5 μCi/ml. Radiolabeling was terminated after 3 min by adding excess nonlabeled methionine (5 mM). Aliquots were taken at time points up to 60 min, and cell pellets were placed at −20°C after centrifugation. Samples were analyzed using 12% SDS-PAGE and a Storm phosphorimager (GE Healthcare).

Generation and purification of HMGE fusion protein.

A His₆-MBP-Gly₅ linker-EsaR (HMGE) fusion protein was constructed through two rounds of PCR. Primers in the first round were used to amplify esaR from pKK-EsaR (EcoRI fragment encoding EsaR from pBAD-EsaR [28] ligated into pKK223-3 [5] with the P_tac promoter upstream of esaR). Forward primer TEVESAR2 (IDT) (5′GAGAACCTGTACTTCCAGGGTGGTGGTGGTGGTATGTTTTCTTTTTTCCTTGAAAATC3′) contained a TEV cleavage site and sequences encoding part of the N-terminal end of EsaR separated by 15 nucleotides coding for 5 glycine residues. Reverse primer ATTBR (IDT) (5′GGGGACAACTTTGTACAAGAAAGTTGCATTACTACCTGGCCGCTGACGCTC3′) contained an attB site downstream of sequences encoding the C-terminal end of EsaR. An 800-bp PCR product was used as template in a second round of PCR. The forward primer ATTBTEV (IDT) (5′GGGGACAACTTTGTACAAAAAAGTTGTGGAGAACCTGTACTTCCAG3′), containing the attB site upstream of the TEV proteolytic site, and the reverse primer from the first round of PCR were used in the second round of PCR. An 850-bp DNA fragment containing attB-TEV-Gly₅-esaR-attB was recovered. This DNA fragment was then used in a BP reaction following the vendor's protocol with the plasmid pDONR201 (Invitrogen). pDONR201 confers kanamycin (Kn) resistance at 50 μg/ml and is an entry vector in the Gateway system (Invitrogen). pDONR201 containing the fragment with esaR was recovered and screened by PCR. Using the vendor's protocol, an LR reaction was performed with the vector pDEST-HISMBP (17) in which the vector contains a P_lac promoter controlling the expression of HMGE (pHMGE). Recovered plasmid DNA was sequenced to confirm integrity.

E. coli Top 10 (13) harboring pHMGE was cultured at 30°C in LB broth containing 100 μg/ml Ap, in the presence or absence of 10 μM AHL, to an OD₆₀₀ of 0.5 to 0.6, at which point 0.5 mM IPTG (isopropyl-β-d-thiogalactopyranoside) was added, and the culture was then incubated overnight at 19°C. Cells were concentrated by centrifugation at 4°C for 10 min at 7,000 rpm in a Beckman JA-10 fixed angle rotor and resuspended in wash buffer (20 mM HEPES, 500 mM NaCl, 20 mM imidazole, 10% glycerol [pH 7.4]). The resuspension was lysed using a French pressure cell at 18,000 lb/in², and the lysates were cleared by ultracentrifugation at 40,000 rpm in a Beckman 70Ti ultra rotor at 4°C for 1 h. The clarified cell lysate was filtered using a 0.2-μm-pore-size syringe filter prior to loading it on to a Ni-nitrilotriacetic acid (NTA) column. HMGE was purified using fast-performance liquid chromatography at 4°C (AKTAprime plus; GE Healthcare) with a 5-ml HisTrap HP column containing Ni-NTA resin (GE Healthcare). A linear gradient elution with wash buffer containing 20 to 500 mM imidazole was used to elute the HMGE from the column. All buffers utilized in purifying HMGE in the presence of AHL were supplemented with 10 μM AHL.

Generation and purification of EsaR NTD178.

Using pHMGE as the DNA template, the forward primer ATTBTEV (IDT) and the reverse primer ATTBR178 (IDT) (5′GGGGACAACTTTGTACAAGAAAGTTGCATTATCATTTGTCCGCGCTCTG3′) were used to amplify a PCR product that contained attB-TEV-Gly₅-esaR (amino acids 1 to 178)-attB. The PCR product was ligated into pGEM-T easy (Promega) and sequenced. After confirmation of sequence integrity, the plasmid was used for the BP and LR reactions of Gateway cloning (Invitrogen) using the entry vector pDONR201 (Invitrogen) and the final destination vector pDEST-HISMBP (17), as described above. The resulting plasmid pHMGNTD178 was transformed and propagated in E. coli BL21(DE3) (Stratagene).

The HMGNTD178 fusion protein (His₆-MBP-Gly₅-EsaR NTD) was purified from E. coli BL21(DE3) cells in a manner very similar to that described above for HMGE, with the following changes. IPTG (1 mM) was used to induce protein expression overnight at 19°C. In addition to Ni-NTA resin, amylose resin (New England Biolabs) was used per the manufacturer's directions followed by gel filtration over a HiPrep 26/60 Sephacryl S-200 column (GE Healthcare) to further improve protein purify. After purification, fractions containing HMGNTD178, as indicated by 12% SDS-PAGE visualization, were pooled and dialyzed against an excess of wash buffer in the presence of TEV protease overnight at 4°C. The sample was then passed over a Ni-NTA column. The flowthrough (correlating to cleaved NTD178) was collected and passed over the Sephacryl S-200 column equilibrated with wash buffer without imidazole. Eluted protein was visualized on a 12% SDS-PAGE gel, and protein fractions correlating with the predicted molecular size of a monomer of NTD178 (∼21 kDa) were pooled, concentrated, and stored at −70°C.

EMSA analysis of HMGE activity.

The ability of purified HMGE to bind to DNA was analyzed through an electrophoretic mobility shift assay (EMSA) as previously described (15) using the ³²P-end-labeled 28-bp double-stranded oligonucleotide containing the 20-bp esa box sequence, PesaR28. HMGE (0 to 100 nM), additionally purified using a HiTrap Heparin HP column (GE Healthcare) per the manufacturer's directions and the Sephacryl S-200 column, was incubated with 2 nM probe DNA for 20 min in EMSA binding buffer [20 mM HEPES, 1 mM EDTA, 30 mM KCl, 0.2% Tween 20, 10 mM (NH₄)₂SO₄, 50 ng/μl poly(dI-dC), 150 μg/ml acetylated bovine serum albumin (BSA), and 10% glycerol] in a total reaction volume of 20 μl. Samples were analyzed using 6% Tris-glycine-EDTA native PAGE with 1× Tris-glycine buffer at 80 V for 2 h in an apparatus packed in ice. After electrophoresis, gels were dried and visualized using a Typhoon Trio variable mode imager (GE Healthcare).

Partial in vitro proteolysis.

Reaction mixtures of a 50-μl volume containing 13.5 μM purified HMGE were incubated with 72 nM thermolysin in the absence or presence of 67.5 μM AHL (5× concentration in relation to HMGE) for 1 h at 37°C under the following conditions: 25 μl of HMGE in wash buffer without imidazole was added to 25 μl of 2× thermolysin buffer (4 mM CaCl₂, 10% glycerol, 300 mM NaCl, 20 mM Tris-HCl [pH 8.0]). Reactions with thermolysin were stopped by the addition of 2 μl of 0.5 M EDTA (pH 8.0). Similarly, the fusion protein was also separately exposed to 54 nM trypsin for 1 h at room temperature with and without 67.5 μM AHL under the following conditions: 25 μl of HMGE in wash buffer without imidazole was added to 25 μl of 2× trypsin buffer (40 mM MgSO₄, 40 mM Tris-HCl [pH 7.5], 20 mM CaCl₂). Reactions were stopped by adding 12.5 μl of sample buffer (0.624 ml 1 M Tris [pH 6.8], 0.2 g SDS, 1.04 ml glycerol, 0.5 ml β-mercaptoethanol, trace bromophenol blue) and boiling prior to analysis via 12% SDS-PAGE.

For the limited proteolytic analysis of EsaR NTD178, 13.5 μM protein was exposed to decreasing concentrations of thermolysin (starting with 1.7 μM and serially diluted 2-fold to 110 nM). At each concentration of thermolysin, digestions were performed in the absence or presence of AHL (67.5 μM), in a final concentration of 1× thermolysin buffer and a final volume of 25 μl. After 1 h of incubation at 37°C, reactions were quenched by boiling in sample buffer prior to analysis on a 12% SDS-PAGE gel.

Similar control assays in the presence of thermolysin or trypsin were performed with 13.5 μM BSA (New England BioLabs) in the presence or absence of AHL. Matrix-assisted laser desorption ionization-tandem time of flight (MALDI-TOF/TOF) mass spectrometry analysis was performed by the Virginia Tech-Mass Spectrometry Incubator on protein fragments of interest.

Gel filtration of HMGE.

E. coli Top 10 cells expressing HMGE were grown separately in LB medium without and with 10 μM AHL. HMGE samples (5 ml) purified as described above, in the absence or presence of 10 μM AHL throughout the purification scheme, were passed over a 320-ml HiPrep 26/60 Sephacryl 200-S (GE Healthcare) gel filtration column using the AKTAPrime system. The concentration of the HMGE sample without AHL was 5.7 μM, while the sample with AHL was 7.8 μM. Each sample was separately loaded onto the column, following equilibration with conjugation buffer (20 mM HEPES, 300 mM NaCl, 10% glycerol [pH 7.4]) at 4°C, using a flow rate of 0.7 ml/min, and 5.0-ml fractions were collected. Peak fractions were assayed on a 12% SDS-PAGE gel to detect the 73-kDa HMGE monomer. Gel filtration molecular size markers (Sigma-Aldrich) were also passed over the same column under identical conditions at the following concentrations: 3 mg/ml carbonic anhydrase (33 kDa), 10 mg/ml albumin (66 kDa), 5 mg/ml alcohol dehydrogenase (150 kDa), and 4 mg/ml β-amylase (223 kDa). The elution volume of Blue Dextran (2,000 kDa) was used to determine the void volume (V_o) of 95 ml. The elution volume-to-void volume ratio (V_e/V_o) of each of the protein molecular size standards was used to generate a calibration curve for the Sephacryl 200-S column.

BS³ cross-linking of EsaR NTD178.

EsaR NTD178 (10 μM) was cross-linked with 100 μM BS³ (Thermo Fisher Scientific) in the absence or presence of 50 μM AHL. The 15-μl reactions were carried out in 20 mM HEPES, 500 mM NaCl, 10% glycerol (pH 7.4) at room temperature and quenched at time points up to 30 min by boiling in sample buffer. The protein was then visualized by 12% SDS-PAGE.

RESULTS

EsaR is stable in vivo in the absence and presence of AHL.

The majority of LuxR family members are known to be unstable/less stable in the absence of AHL (reviewed in reference 23), but very little is known about the actual mechanism of AHL responsiveness that stabilizes them. Comparative experiments in protease-proficient and-deficient E. coli strains revealed that the Clp and Lon proteases degraded TraR at a much higher rate when AHL was not present than when AHL was permitted to complex with the protein during translation. Thus, TraR does not accumulate to appreciable levels in the absence of AHL in protease-proficient E. coli (37). Using similar logic, it was hypothesized that EsaR might be less stable in its inactive state when associated with AHL. An in vivo analysis of cellular levels of EsaR in the native host, P. stewartii, from an OD₆₀₀ of 0.05 to an OD₆₀₀ of 1.6 via quantitative Western immunoblotting revealed that EsaR levels slowly increased from early to late exponential growth (Fig. 1A). A sample at low AHL/OD₆₀₀ clearly has less EsaR per mg total protein than a sample to which exogenous AHL was added (Fig. 1A). At low AHL/OD₆₀₀, EsaR is capable of autorepressing its own expression. However, at a high AHL/OD₆₀₀, EsaR is inactivated and derepression of EsaR expression occurs. If AHL-triggered proteolytic degradation of EsaR were rapidly occurring, its levels would be anticipated to decrease at high AHL/OD₆₀₀. However, our experiments showed no net reduction in EsaR levels, even when excess AHL was supplemented in the growth medium (Fig. 1A). This suggests that extensive degradation by proteases is not the means by which EsaR is inactivated in the presence of AHL.

Fig. 1.

EsaR accumulation and stability in vivo. (A) Results from in vivo Western immunoblot experiments, in which EsaR accumulation was measured in P. stewartii DC283 at the indicated OD₆₀₀ values with natural AHL levels, except for one sample where exogenous AHL was added as indicated. (B) Pulse-chase experiments, in which EsaR was labeled in E. coli SG22163 in the absence or presence of 10 μM AHL. Stability was examined at time points up to 60 min. The images are representative of experiments performed in duplicate.

Pulse-chase experiments were performed in recombinant E. coli to compare the turnover rates of EsaR in the presence or absence of AHL in a more precise manner. EsaR has previously been shown to respond to AHL in E. coli in the same manner as in its native host (20, 28). EsaR was equally stable in the absence or presence of AHL for up to 60 min after synthesis of the protein (Fig. 1B). Therefore, both the Western immunoblot and pulse-chase experiments demonstrate that EsaR is not targeted for rapid proteolysis within the cell upon exposure to AHL.

A His-maltose binding protein (MBP)-glycine linker-tagged EsaR construct (HMGE) is biologically active.

The native form of EsaR has been purified (15), but the method is time-consuming and has not proven to be easily reproducible (S. B. von Bodman, personal communication). An N-terminal MBP fusion to EsaR was successfully used to study the dimeric structure of EsaR when it binds to DNA (15). To further analyze the activity of EsaR in vitro, we have improved our ability to rapidly purify EsaR using a His₆-MBP-glycine linker (Gly₅)-EsaR fusion protein (HMGE) expressed from an IPTG-inducible promoter. TEV protease was unable to completely cleave the His-MBP tag from EsaR, resulting in the production of mixed heterodimers (data not shown). To determine if full-length HMGE was biologically active, its capacity to complement a chromosomal deletion of esaR was examined. Wild-type strain P. stewartii DC283 (9) expressed some EPS at high cell density on CPG medium (Fig. 2A) while a strain with a deletion of esaR and esaI, ESΔIR (29), without or with the vector control pDEST HisMBP (17), exhibited a hyper-mucoid phenotype since EsaR is absent and rcsA is derepressed (Fig. 2A). When the ESΔIR strain is complemented with pSVB60 (15) (encoding the native EsaR protein) or with pHMGE, EPS production is noticeably reduced due to constitutive repression of rcsA (Fig. 2A). These strains regained a hyper-mucoid phenotype in the presence of AHL due to inactivation of EsaR/HMGE and the subsequent derepression of rcsA (Fig. 2A). Thus, HMGE is biologically active in vivo, and it is capable of repressing the rcsA promoter in the absence of AHL and derepressing EPS production in the presence of the AHL ligand. Purified HMGE binds to the esa box target, the native EsaR binding site within the esaR promoter, in a manner similar to EsaR (15) as demonstrated by in vitro EMSAs (Fig. 2B). It is also able to associate with AHL in vitro enabling for an analysis of the responses of HMGE to the presence of ligand (see below).

Fig. 2.

HMGE activity assays. (A) Qualitative analysis of EPS production on CPG agar medium (4) (0.1% Casamino Acids, 1% peptone, 1% glucose, and 1.5% agar), with Ap (50 μg/ml) or tetracycline (Tc) (10 μg/ml) as appropriate, following incubation at 30°C for 48 h: 1, wild-type P. stewartii DC283 (9); 2, esaR esaI double-deletion strain P. stewartii ESΔIR (29), and 3, P. stewartii ESΔIR with the vector control pDEST HisMBP (Ap^r) (17); 4 and 5, P. stewartii ESΔIR pSVB60 (Tc^r) grown in the absence or presence of 10 μM AHL, respectively; 6 and 7, P. stewartii ESΔIR pHMGE (Ap^r) grown in the absence or presence of 10 μM AHL, respectively. (B) EMSA of HMGE with PesaR28 DNA probe. Radiolabeled PesaR28 DNA probe (2 nM) (lane 1) was incubated with various concentrations of HMGE as indicated (lanes 2, 3, and 4) and shown to produce a higher-molecular-size complex. Specificity of binding was verified by adding excess unlabeled PesaR28, which resulted in a downshift of the labeled probe (lanes 5, 6).

Partial in vitro proteolysis reveals AHL-dependent protection and conformational changes in EsaR.

Limited in vitro proteolytic digestions with HMGE were performed to further assess the effects of AHL binding on the conformation of EsaR. HMGE was purified in the absence or presence of 10 μM AHL, with the AHL added cotranslationally during growth (and present through protein purification) or posttranslationally after purification. Two different proteases (thermolysin and trypsin) were used to produce cleavage patterns of HMGE that were analyzed via SDS-PAGE. The activity of the two proteases, in the absence of EsaR, was not detectably altered by the presence of AHL as determined through control digestions of purified bovine serum albumin (data not shown).

The highly stable His-MBP tag (45 kDa) was clearly seen in all of the HMGE samples digested with thermolysin (Fig. 3A) in the presence or absence of AHL, but no band of the predicted size for full-length EsaR was observed. However, two extra bands (∼18 and 5 kDa) were observed when HMGE was incubated in the presence of AHL (either co- or posttranslationally) but not when HMGE was incubated without AHL (Fig. 3A). The bands were extracted, peptides were generated, and their minimal amino acid sequence was determined via MALDI-TOF/TOF (Virginia Tech-Mass Spectrometry Incubator). The bands produced from the thermolysin cleavage reactions were minimally comprised of EsaR amino acids 1 to 160 (the NTD of EsaR) and 23 to 74 (a region close to the putative AHL binding pocket) (Fig. 3C); there may be additional residues in the N or C termini of these protein fragments that were not detected. Both region 1 to 160 and region 23 to 74 of EsaR contain more than one thermolysin cleavage site, as predicted by PeptideCutter (ExPASy), suggesting that the presence of AHL produced a conformational shift that interfered with thermolysin cleavage. Subsequent purification of the EsaR NTD (amino acid residues 1 to 160) indicated that the polypeptide existed as a dimer (data not shown), which led to further analysis of the NTD.

Fig. 3.

Limited in vitro digestion of EsaR by thermolysin and trypsin. SDS-PAGE analysis of resistance of EsaR to (A) thermolysin and (B) trypsin ± AHL. Purified HMGE was used at a final concentration of 13.5 μM for all conditions (lanes 1 to 4) as follows: 1, no protease; 2, no AHL plus protease; 3, AHL present during growth and purification plus protease; 4, AHL added after purification plus protease. The dominant band in lanes 2 to 4 is the His-MBP tag (45 kDa). Arrows indicate EsaR-associated bands of interest, with the numbers providing the minimal number of residues present in the protein fragment. The images are representative of experiments performed in duplicate. (C) Cartoon model of the domain structure of EsaR. The AHL binding NTD is white, with the AHL binding pocket region in light gray and the DNA binding CTD in dark gray. Residues of EsaR found to be protected from thermolysin or trypsin digestion are labeled above or below the diagram, respectively.

Similar limited proteolytic digestion assays performed with trypsin produced the highly stable His-MBP tag (45 kDa) (Fig. 3B) and weak residual bands corresponding to the predicted size of EsaR (28 kDa) in all three samples ±AHL. However, there was an extra band (∼9 kDa) present in the samples with AHL. Through MALDI-TOF/TOF (Virginia Tech-Mass Spectrometry Incubator) it was determined that this extra band corresponded minimally to amino acids 130 to 213 of EsaR (Fig. 3B). This region includes the predicted linker region between the NTD and CTD (Fig. 3C) and it contains more than one trypsin cleavage site, as predicted by PeptideCutter (ExPASy). The fact that cleavage did not occur in this region as efficiently in the presence of AHL suggests that interdomain conformational changes are occurring in response to AHL binding, a phenomenon that has long been speculated to occur among all LuxR family members (22) but has never before been biochemically demonstrated with a native ligand. In sum, it appears that HMGE, whose activity is inhibited by AHL, alters its conformation such that some proteolytic cleavage sites are less accessible in the presence than in the absence of AHL. Interestingly, the finding that AHL affords protection from proteases has also been observed with other members of the LuxR protein family that are stimulated by AHL (31, 37).

EsaR dimerizes in the absence and presence of AHL.

Because EsaR was equally resistant to proteolytic digestion in vivo both in the presence and in the absence of AHL and more resistant to proteolytic digestion in vitro, another mechanism of AHL inactivation was considered. It is known that most LuxR homologues function as dimers when AHL is present (3, 26, 33, 37). Based on amino acid alignments, the corresponding dimerization interface for EsaR would be predicted to encompass residues 132 to 161. EsaR is known to bind to DNA as a dimer in the absence of AHL (15), but its quaternary structure in the presence of AHL has never been well defined. The effect of AHL on the multimeric state of EsaR was therefore determined through a combination of gel filtration chromatography and cross-linking experiments. During gel filtration experiments, HMGE exposed to 10 μM AHL during growth and purification eluted with a single peak at 122 ml from a 320-ml volume column (Fig. 4A). Through the use of four known molecular size markers, an apparent molecular size of 136 kDa was calculated for HMGE (Fig. 4B). The predicted monomeric molecular size of HMGE is 73 kDa, so a dimer would be ∼146 kDa; these experiments suggest that HMGE is dimeric in the presence of AHL. Gel filtration chromatography was also performed on HMGE in the absence of AHL. The protein eluted with a single peak at 118 ml (Fig. 4A), just slightly earlier than the protein in the presence of AHL. An apparent molecular size of 155 kDa was calculated for HMGE (Fig. 4B). AHL had no effect on the protein standard calibration curve (data not shown). These in vitro results suggest that at the concentrations examined, HMGE exists as a dimer in both the absence and presence of AHL. The observation that HMGE has a slightly smaller calculated molecular size in the presence of AHL suggests that the ligand promotes a tighter conformation of the protein.

Fig. 4.

Multimeric state of HMGE fusion protein in presence and absence of AHL. (A) UV absorption profiles (280 nm) of protein fractions containing HMGE eluted in the presence and absence of 10 μM AHL over the 320-ml 26/60 Sephacryl 200-S gel filtration column (solid line, HMGE without AHL in buffer; dashed line, HMGE with AHL in buffer). HMGE without AHL eluted at 118 ml and HMGE with AHL at 122 ml. This corresponds to a molecular size of 155 kDa in the absence of AHL and 136 kDa in the presence of AHL, as determined by the calibration plot generated using the elution volumes of molecular size markers in panel B. The two additional minor peaks seen on the left and right of the major peak are aggregates and truncated HMGE, respectively, as determined through SDS-PAGE analysis. (B) The data points on the calibration plot represent the following: 1, β-amylase (200 kDa); 2, alcohol dehydrogenase (150 kDa); 3, albumin (66 kDa); and 4, carbonic anhydrase (29 kDa). Plots are representative of experiments performed in duplicate.

The NTD of EsaR binds AHL and is capable of dimerizing in the absence and presence of AHL.

To further verify the finding that EsaR is capable of dimerizing in the presence of AHL, cross-linking experiments were performed in the absence or presence of AHL using only the NTD of EsaR from residues 1 to 178. The His-MBP tag was efficiently cleaved from the NTD using TEV protease; apparently the presence of the CTD hinders complete cleavage of HMGE by TEV. Limited proteolytic thermolysin digestion assays produced a similar banding pattern for the EsaR NTD in the presence of AHL (Fig. 5A) as had been observed with HMGE (Fig. 3A). MALDI-TOF/TOF analysis of the two EsaR NTD bands of highest molecular size determined that they were minimally comprised of amino acids 1 to 160 and 1 to 93. These regions encompass most of the NTD and the region around the AHL binding site; they appear to be protected from complete thermolysin digestion when AHL is present. Therefore, the EsaR NTD appears to be capable of binding AHL, independent of the CTD (Fig. 5A), consistent with what has been observed in other LuxR homologues (3, 8, 34).

Fig. 5.

Ability of EsaR NTD178 to resist proteolytic digestion and to form dimers. (A) Limited in vitro digestion of NTD178 by thermolysin. Purified NTD178 was used at a final concentration of 13.5 μM. Samples were as follows: 1, size standards; 2, thermolysin; 3 and 4, NTD178 with and without AHL (+ and −, respectively); 5 to 14, NTD178 with and without AHL and digested with the indicated concentration of thermolysin. Arrows indicate EsaR-associated bands of interest, with the numbers providing the minimal number of residues present in the protein fragment, and the mobility of thermolysin (T). (B) Time course of NTD178 cross-linking by BS³. NTD178 (10 μM) was exposed to 100 μM BS³ cross-linker in the presence (+) and absence (−) of AHL. Reaction pairs, left to right, correlate to 0, 1, 5, 10, 15, 20, and 30 min. The two arrows highlight the bands associated with the 42-kDa dimer (D) and 21-kDa monomer (M) of NTD178. The images are representative of experiments performed in duplicate.

The EsaR NTD protein was then exposed to the cross-linking agent BS³. Over a 30-min period, the EsaR NTD was cross-linked to a level similar to that observed in cross-linking experiments with full-length CarR, the LuxR-type activator from Erwinia carotovora (Fig. 5B) (31). The apparent molecular size of the cross-linked product suggests that the NTD of EsaR has a propensity to form dimers and that the CTD is not essential for this to occur. In addition, the EsaR NTD maintains a similar quaternary structure independent of the presence of ligand, further indicating that AHL binding does not significantly alter the multimeric state of the protein.

DISCUSSION

EsaR is the best-understood member of a distinctive subclass of the LuxR family of proteins (23, 24). These proteins have been described as sharing several common attributes: they are produced by Enterobacteriales, they preferentially bind N-3-oxo-hexanoyl-l-homoserine lactone, they do not directly regulate their cognate AHL synthase genes, they can function as repressors and activators depending on the position of the recognition site in the DNA, and their genes are convergently transcribed with the genes for their respective AHL synthases. In addition, all of these proteins have an extended interdomain linker region and an extended CTD in comparison to the majority of LuxR homologues (24). They regulate gene expression using a reverse mechanism whereby in the absence of AHL they are capable of binding DNA but take on an inactive conformation in the presence of AHL. Structurally, EsaR and its closest relatives deviate from most LuxR homologues because they are stable in the absence of their cognate AHL, making them amenable to in vitro analysis.

It has previously been demonstrated that EsaR is fully active as an apoprotein. In the absence of AHL, EsaR is capable of repressing transcription from the esaR and rcsA promoters (6, 14, 15) and activating transcription of an sRNA (20) at low cell density. AHL hinders the ability of EsaR to regulate transcription at high cell density. However, the molecular basis for inhibition of EsaR by AHL remains elusive. In this body of work, two possible ways that binding of AHL might modulate the level of active EsaR, an enhanced rate of proteolysis of the protein and an inability to dimerize, were examined.

This study established that EsaR does not appear to be extensively posttranslationally degraded by proteases in vivo in response to AHL binding. In P. stewartii, EsaR levels gradually rose from low cell density/low AHL to high cell density/high AHL, suggesting that rapid removal of the protein via proteases is not the manner in which the protein is inactivated in the presence of AHL. This is in stark contrast to the results of a similar analysis of TraR from A. tumefaciens in E. coli, which showed that apo-TraR is quickly degraded by proteases in vivo. Upon binding of AHL by TraR in vitro, the protein takes on an active conformation and in the process becomes more resistant to these proteases, as the half-life increases 20-fold (36, 37). Similarly, CarR from E. carotovora also exhibits enhanced protease resistance in the presence of AHL (31).

TraR is one of just a few members of the LuxR protein family for which detailed structural information is available. The structure of AHL-associated full-length TraR and the NTD of LasR from Pseudomonas aeruginosa have been solved with their native ligands by X-ray crystallography (3, 27, 35, 38), and the structure of CviR from Chromobacterium violaceum bound to an antagonist is also known (8). The structure of the NTD of SdiA from E. coli was examined using nuclear magnetic resonance (NMR) analysis (34). No structural information exists for EsaR or any member of its subfamily. However, previous studies of EsaR using fluorescence spectroscopy revealed that AHL altered the intrinsic fluorescence emissions from tryptophan residues in the binding pocket (15). In addition, AHL enhanced the thermal stability of EsaR (15). The in vitro limited proteolytic digestion assay performed in this study reinforced the previous findings and further probed the structural differences between apo-EsaR and ligand-associated EsaR. AHL binding to EsaR resulted in altered patterns of protease susceptibility using two different proteases, thermolysin and trypsin.

In the presence of AHL, the ligand binding NTD of EsaR was less susceptible to thermolysin digestion (Fig. 3A). Thermolysin attacks hydrophobic regions, and binding of AHL most likely directly interfered with the ability of the protein to cleave target sites in the AHL binding pocket. In the case of assays with trypsin, a region spanning from the NTD through the linker region into the CTD of EsaR was found to be more resistant to the protease when AHL was present (Fig. 3B). This result is very intriguing because it suggests that AHL binding in the NTD may elicit a conformational change across both domains of the protein, resulting in a tighter conformation that protects the linker from degradation. For LuxR family proteins, it has long been proposed that AHL binding at the NTD leads to subsequent changes in the CTD that alter the DNA binding abilities of the protein (22). However, an inability to study LuxR proteins in vitro in the presence and absence of their native ligand has limited a structural analysis of this mechanism. X-ray crystallography of CviR bound to an antagonist provides evidence in support of NTD-CTD interactions (8). It has been speculated that the antagonism of EsaR by its native AHL might work in a similar manner, resulting in a decreased affinity for the promoters it regulates.

In addition to intradomain interactions playing a role in the responsiveness of EsaR to AHL, changes in the interface between the two polypeptides in a homodimer may also be important. Binding of AHL may inhibit dimerization of the NTDs, leading to dissociation of EsaR into monomers and subsequent dissociation from DNA. Both gel filtration analysis of HMGE and in vitro cross-linking of the NTD demonstrate that EsaR has a propensity to form dimers at the concentrations examined and that the presence of AHL does not decrease the capacity for dimerization to occur. Further, only the EsaR NTD is required for both AHL binding and dimerization to occur. Thus, AHL does not appear to alter the multimeric state of EsaR.

A complete understanding of the mechanism by which AHL binds to the NTD of LuxR homologues and modulates the activity of the CTD remains an unresolved issue in the quorum-sensing field. Clearly, diverse mechanisms for AHL responsiveness exist within the LuxR family. Our report offers an explanation of the interaction between the AHL signal and its cognate EsaR receptor, a system that is the model for a unique subset of quorum-sensing regulators. In the case of EsaR, binding of AHL causes a conformational shift in the protein, in which the protein appears to take on a tighter, more stable and likely less flexible conformation. Under these conditions, EsaR is no longer able to make appropriate interactions with DNA, leading to its dissociation. Future structural studies of EsaR would provide valuable insights into the precise nature of this mechanism.