Structure/Function Analysis of the Pantoea stewartii Quorum-Sensing Regulator EsaR as an Activator of Transcription

Abstract

In Pantoea stewartii subsp. stewartii, two regulatory proteins are key to the process of cell-cell communication known as quorum sensing: the LuxI and LuxR homologues EsaI and EsaR. Most LuxR homologues function as activators of transcription in the presence of their cognate acylated homoserine lactone (AHL) signal. However, EsaR was initially found to function as a repressor in the absence of AHL. Previous studies demonstrated that, in the absence of AHL, EsaR retains the ability to function as a weak activator of the lux operon in recombinant Escherichia coli. Here it is shown that both the N-terminal and the C-terminal domains of EsaR are necessary for positive regulation. A site-directed mutagenesis study, guided by homology modeling to LuxR and TraR, has revealed three critical residues in EsaR that are involved in activation of RNA polymerase. In addition, a native EsaR-activated promoter has been identified, which controls expression of a putative regulatory sRNA in P. stewartii.

Quorum sensing is a population-wide behavioral response that has been identified in a variety of prokaryotes. The term quorum sensing is used to describe the ability of a microorganism to sense and initiate a response to a self-produced intercellular signaling molecule commonly known as an autoinducer (14). Several important bacterial processes are regulated through quorum sensing, including antibiotic production, the release of exoenzymes, the production of virulence factors, the induction of genetic competence, conjugative plasmid transfer, biofilm formation, and bioluminescence (13, 22, 38, 39).

Similar to the model system of Vibrio fischeri, the quorum-sensing systems in many gram-negative proteobacteria typically contain a LuxI homologue that functions as an autoinducer synthase producing an acylated homoserine lactone (AHL) signal and a LuxR homologue that responds to AHL and controls gene expression. Of the more than 50 LuxR homologues identified, the majority function as AHL-dependent activators similar to LuxR (13, 22, 38). However, a subset of them, including EsaR from Pantoea stewartii subsp. stewartii, has been found to regulate expression of specific target genes by repression and AHL-dependent derepression (2, 3, 7, 25, 35).

In P. stewartii, quorum sensing is involved in the control of virulence. This plant pathogen is the causative agent of Stewart's wilt disease and leaf blight in maize. Disease is initiated when the bacterium begins producing large amounts of an exo/capsular polysaccharide (EPS), which blocks the corn xylem vessels and induces necrotic lesions (9). EPS production is under the control of a multitiered regulatory cascade (4, 30). EsaI/R reside at the top of this regulatory hierarchy. The LuxI homologue, EsaI, synthesizes AHL, which at high enough concentrations can induce the production of EPS through the AHL-dependent inactivation of EsaR (4, 35). EsaR functions in the absence of AHL by binding to DNA as a dimer and blocking the transcription of genes involved in EPS production (23, 24). Like LuxR, which requires the lux box for DNA binding, EsaR also requires a DNA binding site known as the esaR box. Unlike the lux box, which is centered around the −42.5 region of the promoter in a class II promoter orientation (12, 20), the esaR box is centered around the −10 site and blocks transcription by RNA polymerase at the esaR and rcsA promoters (6, 24). The lux box and the P_esaR esaR box sequences differ by five bases. Interestingly, both LuxR and EsaR respond to the same AHL, 3-oxo-hexanoyl-l-homoserine lactone, albeit in an opposite manner.

A previous study demonstrated that in the absence of AHL EsaR retains the ability to function as a weak activator of the V. fischeri lux operon in recombinant Escherichia coli (34). The goal of the present study was to further characterize EsaR as an activator of transcription. Studies of LuxR and TraR from Agrobacterium tumefaciens identified critical residues necessary for interaction with RNA polymerase (RNAP) for activation to occur (11, 21, 26, 37). In the case of LuxR several interacting residues were found solely in the C-terminal domain, whereas critical residues were found in both the N-terminal and C-terminal domains of TraR (11, 21, 26, 37). Through the use of sequence alignment with LuxR and homology modeling of TraR to EsaR (A. Thode, D. Donham, and M. Churchill, unpublished data), several of these previously characterized positive control (PC) variants were mapped on EsaR. A combination of deletion and site-directed mutagenesis strategies was used to identify critical residues of EsaR, which are required for activation. The relevance of EsaR as an activator has been established through the identification of a native promoter of P. stewartii that controls expression of a sRNA.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Strains and plasmids used in the present study are described in Table 1. The E. coli strains were grown at 37°C in Luria-Bertani broth (LB) or RM minimal medium (RM) (2% Casamino Acids, M9 salts [12.8 g of Na₂HPO₄·7H₂O, 3 g of KH₂PO₄, 0.5 g of NaCl, and 1 g of NH₄Cl per liter], 0.4% glucose, and 1 mM MgCl₂), and supplemented, where indicated below, with 100 μg of ampicillin/ml, 50 μg of kanamycin/ml, 10 μM N-(β-ketocaproyl)-l-homoserine lactone (3-oxo-C6-HSL) (AHL; Sigma, St. Louis, MO), 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside), or 4 μg of X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside; Sigma)/ml. The P. stewartii strains were grown at 28°C in LB in the presence of 30 μg of nalidixic acid/ml.

TABLE 1.

Strains and plasmids

Strain or plasmid	Relevant informationa	Source or reference
Strains
E. coli
DH5α	F− φ80dlacZΔM15 Δ(lacZYA-argF)U169 decR recA1 endAI hsd17 phoA supE44 thi-l gyrA96 relA1	17
Top10	F−mcrA Δ(mrr-hsdRMS-mcrBC) φ80dlacZΔM15 ΔlacX74 deoR recA1 araD139 Δ(ara-leu)7697 galU galK rpsL (Strr) endA1 nupG	15
P. stewartii
DC283	Wild-type strain SS104; Nalr	9
ESN51	esaI::Tn5seqN51	4
ESN10	P. stewartii esaI::cat	23
Plasmids
pGEM-T	Cloning vector, f1 ori; Apr; used as an intermediate cloning vector	Promega
pBAD22	Arabinose-inducible vector, Apr	16
pEXT22	Low-copy-number (one to two copies/cell) plasmid; Kmr	10
pBAD-LuxR	luxR ligated into EcoRI and SmaI sites in pBAD22	34
pBAD-EsaR	esaR ligated into EcoRI sites in pBAD22, 15-bp carryover of pGEM vector	34
pluxI-lacZ	luxI-lacZYA fusion; Apr	34
pEXT-luxI	luxI-lacZYA fusion ligated into EcoRI and MfeI sites in pEXT22	This study
pEXT-esa1	esaR box in first orientation-lacZYA fusion ligated into EcoRI and MfeI sites in pEXT22	This study
pEXT-esa2	esaR box in second orientation-lacZYA fusion ligated into EcoRI and MfeI sites in pEXT22	This study
pRNP-lacZ	pEXT22 with natural promoter region of esaR	This study
pBAD-161EsaRΔN	pBAD-EsaR with deletion variant of N terminus (Δ2-160)	This study
pBAD-179EsaRΔN	pBAD-EsaR with deletion variant of N terminus (Δ2-178)	This study
pBAD-ΔABD	pBAD-EsaR with deletion of autoinducer-binding domain (Δ65-107)	This study
pBAD-EsaRΔ171-8	pBAD-EsaR with deletion of residues 171 to 178	This study
pBAD-EsaRΔCT	pBAD-EsaR with deletion of residues 237 to 249	This study
pBAD-EsaRΔΔ	pBAD-EsaR with deletion of residues 171 to 178 and residues 237 to 249	This study
pBAD-E7A	pBAD-EsaR with variant E7A	This study
pBAD-N119A	pBAD-EsaR with variant N119A	This study
pBAD-196EsaR	pBAD-EsaR with variant K196A	This study
pBAD-199EsaR	pBAD-EsaR with variant A199W	This study
pBAD-204EsaR	pBAD-EsaR with variant I204A	This study
pCR2.1	Intermediate cloning vector	Invitrogen
pCR-PesaR	pCR2.1 containing EsaR activated promoter	This study
pFPV25	Vector with promoterless gfp	32
pPesaR-AC::gfp	pFPV25 containing EsaR activated promoter	This study
pSUP102	Mobilizable plasmid; Tcr Cmr	28
pSUP-EsaR	araC PBAD promoter and esaR in pSUP102; Cmr	This study

^aCm^r, chloramphenicol resistance; Tc^r, tetracycline resistance; Ap^r, ampicillin resistance; Str^r, streptomycin resistance; Km^r, kanamycin resistance; Nal^r, nalidixic acid resistance.

Plasmid construction.

To construct variants of the luxI promoter containing the esaR box, overlap PCR (18, 19) was performed with the template pluxI-lacZ (34), which contained the lux operon promoter fused to a lacZ reporter. Two sets of overlapping mutagenic primers, EBUP/EBDOWN and EBUP2/EBDOWN2, were synthesized (Table 2) that contained the esaR box in two different orientations. They were used in two separate PCRs with an upstream external primer, ECORI, containing an EcoRI site and a downstream external primer, BAMHI, containing a BamHI site (Table 2). A second round of PCR was then performed to generate the desired 400-bp products, which were cloned into pGEM-T (Promega, Madison, WI) and sequenced to confirm their integrity (Virginia Bioinformatics Institute Core Laboratory, Virginia Tech, Blacksburg). The intermediate pGEM-T constructs and pluxI-lacZ were digested with EcoRI and BamHI, and the promoter regions from pGEM-T were ligated into pluxI-lacZ replacing the lux operon promoter with the newly constructed esaR box containing promoters. The constructs containing the promoters with the lux or esaR box fused to lacZ were digested with EcoRI and MfeI and subcloned into pEXT22 (pEXT-luxI, -esa1, or -esa2) (10), which replicates at 1 to 1.5 copies/cell.

TABLE 2.

Primers^a

Primer	Sequence (5′-3′)	Source or reference
EBUP	CATAAGCGCCTGTACTATAGTGCAGG	This study
EBDOWN	CCTGCACTATAGTACAGGCGCTTATG	This study
EBUP2	GCACCTGCACTATAGTACAGGCTTAC	This study
EBDOWN2	GTAAGCCTGTACTATAGTGCAGGTGC	This study
ECORI	AAGAATTCACAATGTACCATTTTAGTCATATCAG	This study
BAMHI	AAGGATCCTTATACTCCTCCGATGGAATTGCC	This study
NPesaF	GAATTCGGACGTTTTCCCTAGTGTTGGCTG	This study
NPesaR	GGATCCATGTAAGTCTGAAGCGTATCC	This study
ESAR161	GAATTCACCATGCTGGCCGGTACCGAAGG	This study
ESAR179	GAATTCACCATGACGATATTTTCCTCGCGTG	This study
ESARRC	AAGCTTTCACTACCTGGCCGCTGAC	This study
C-termfwd	ACGATATTTTCCTCGCGTGA	This study
T7rev	CTAGTTATTGCTCAGCGG	This study
T7	TAATACGACTCACTATAGGG	This study
N-termrev	TCACGCGAGGAAAATATCGTCGGGGCTCGCTCGCCTTCGG	This study
CTDELR	TCTGATAAGCTCGAGTTATACACC	This study
EE7F	CCTTGCGAATCAAACAATAACGG	This study
EE7R	CCGTTATTGTTTGATTCGCAAGG	This study
EN119F	GCATGACCACATGGCAAACCTTGC	This study
EN119R	GCAAGGTTTGCCATGTGGTCATGC	This study
ESF196	GTATGGGCGCAACATATGCTGAGATTG	This study
ESR196	CAATCTCAGCATATGTTGCGCCCATAC	This study
ESF199	GTATGGGCAAAACCTATTGGGAGATTGCC	This study
ESR199	CAATCTCCCAATAGGTTTTGCCCATACTCG	This study
ESF204	CAAAACATATGCTGAGATTGCCGCTGCTACGGGCATTTC	This study
ESR204	CCGTAGCAGCGGCAATCTCAGCATATGTTTTGCCCATAC	This study
BADR	CTTCTCTCATCCGCCAAAAC	34
ESARF2	GGAATTCACCATGTTTTCTTTTTTCCTTG	34
BADVF	TAACCTTTCATTCCCAGCGGTCG	This study
PesaRbamfwd	AAACAACTGGATGGATTGTAAC	This study
PesaRbamrev	CATTTGAAGGATCCTTTTTGCT	This study
SALARACF	GGTGTCGACTTATGACAACTTGACGGCTACAT	This study
ESSRF	GGTTACAATGGCTTCAGTTGTTTAGCG	This study
ESSRR	CACATTCATATGCTTATTATTTGCGCTCAGGC	This study
ESAS	TGGCCGGACGTTTTCCCTAGTGTTGGCTGTTTTAG	This study

^aUnderlined sequences correspond to unique restriction site utilized for cloning.

To measure the activity of EsaR as a repressor, a new low-copy-number reporter was constructed by using the primers NPesaF and NPesaR (Table 2) to amplify P_esaR via PCR with pSVB5-18 as a template (4). The promoter region was then fused to lacZ and cloned into pEXT22 (pRNP-lacZ) using EcoRI and MfeI sites as described above.

A series of deletion variants of EsaR was also constructed by using PCR to amplify the regions of interest, followed by subsequent subcloning into pBAD22. Two constructs, pBAD-161EsaRΔN and pBAD-179EsaRΔN, were generated with the primers ESAR161 and ESAR179 in combination with ESARRC (Table 2) to create N-terminal domain deletions of amino acids 2 to 160 and 2 to 178, respectively (Table 1 and Fig. 1). An in-frame deletion of a portion of the autoinducer-binding domain (residues 65 to 107) from esaR was created by performing a set of sequential digestions. First, pBAD-EsaR was digested with PvuII and HindIII. In the second reaction pBAD-EsaR was digested with HpaI and HindIII. A 450-bp fragment was recovered from the second reaction and was ligated to a 6-kb fragment from the first reaction to generate pBAD-ΔABD. A deletion of the linking loop between the N-terminal and C-terminal domains corresponding to residues 171 to 178 was accomplished by amplifying the C terminus with the primers C-termfwd and T7rev and the N terminus with the primers T7 and N-termrev (Table 2) from the template pET28b::H6esaR. The two products were then used in another round of PCR to obtain a 1-kb product that was subcloned by using NcoI and XhoI sites into pET28b (Novagen) (pET28b::H6-esaRD1). Other deletions were created where the C terminus of EsaR (residues 237 to 249) was removed or both the linking loop (residues 171 to 178) and the C terminus (residues 237 to 249) were removed. Primers T7 and CTDELR (Table 2) were used in these PCRs with either pET28b::H6esaR or pET28b::H6-esaRD1 as templates. The two products were cloned into pET28b (pET28b::H6-esaRD2 and pET28b::H6-esaRD3). The three His-tagged constructs along with pBAD-EsaR (contains non-His-tagged esaR) were then digested with KpnI and HindIII, and the truncated esaR genes were individually subcloned into pBAD-EsaR creating pBAD-EsaRΔ171-8, pBAD-EsaRΔCT, and pBAD-EsaRΔΔ. (Table 1 and Fig. 1).

FIG. 1.

Stability and activity of deletion variants of EsaR. Shading: ▧, AI binding; ░⃞, DNA binding; ▪, extended regions. Y, yes; N, no.

Overlap PCR was also used for site-directed mutagenesis of esaR in pBAD-EsaR to generate possible PC variants. Two overlapping mutagenic (internal) primers (Table 2) were designed to change one specific amino acid (E7A, N119A, K196A, A199W, and I204A) and to add a restriction endonuclease site (NdeI) to constructs K196A and I204A for screening purposes. In all cases, except for the E7A construct, external primers (BADR and ESARF2) (Table 2) were also used in amplifying the gene. In the case of the E7A construct the BADVF primer was used in the amplification of the N-terminal domain. The second round PCR products were ligated into pGEM-T. The pGEM-T constructs, along with pBAD-EsaR, were digested with HindIII and HpaI, and the region of esaR containing the corresponding residues changes were ligated into the pBAD vector.

pPesaR-AC::gfp (Table 1) was constructed by amplifying a 253-bp fragment from the chromosome of DC283 (wild-type P. stewartii) (9) by PCR using the primers PesaRbamfwd and PesaRbamrev (Table 2). The PCR product was ligated into TOPO vector pCR2.1 (Invitrogen, Carlsbad, CA) to generate pCR-PesaR. To generate a transcriptional fusion of the promoter region to gfp, pCR-PesaR was digested with BamHI and subcloned into the BamHI restriction site of pFPV25 (32). The orientation of the promoter fusion was verified by PCR and DNA sequencing.

So that EsaR-dependent control of pPesaR-AC::gfp could be examined, the compatible plasmid pSUP-EsaR was constructed. First, a fragment containing esaR downstream of the P_BAD promoter and araC was amplified from pBAD-EsaR by PCR using the primers SALARACF and BADR (Table 2). The PCR fragment was ligated into pGEM-T and sequenced. This intermediate plasmid was digested with SalI, and a 1.8-kb fragment was ligated into the tetracycline resistance gene of pSUP102 (28) that was digested with SalI. Proper fragment insertions were verified via phenotype screening for the loss of tetracycline resistance and restriction mapping with KpnI and NcoI to ensure that the insert was in an orientation as to not be under the control of the tetracycline resistance gene promoter.

EsaR activation of promoter constructs in vivo.

E. coli Top10 strains containing pBAD-EsaR, pBAD-LuxR, or pBAD constructs encoding deletion or PC variants were cotransformed with either pluxI-lacZ, pEXT-esa1, or pEXT-esa2. Cells were grown as previously described (34) except they were induced at an optical density at 600 nm (OD₆₀₀) of 0.25 and harvested at an OD₆₀₀ of 1.0. Then, 5-μl aliquots of the cells were stored at −70°C prior to analysis of LacZ expression via chemiluminescent β-galactosidase assays measuring relative light units (RLU) (Tropix, Bedford, MA).

P. stewartii ESN10 containing pPesaR-AC::gfp was grown overnight in LB containing ampicillin and nalidixic acid. Cells were subcultured to an OD₆₀₀ of 0.05 with or without 20 μM AHL and then grown to an OD₆₀₀ of 1.0. E. coli Top10 strains containing pSUP102 or pSUP-EsaR were cotransformed with pPesaR-AC::gfp. Cells were grown as previously described (35) except they were induced at an OD₆₀₀ of 0.25 and harvested at an OD₆₀₀ of 1.0. For both the P. stewartii and the E. coli assays, at the desired OD, 200 μl of each culture was placed in a 96-well optical bottom microtiter plate for the analysis of fluorescence output (excitation and emission wavelengths of 485 and 535 nm, respectively), and the OD₅₉₀ was determined on a Tecan SpectraFluor Plus plate reader (Tecan, Mannedorf/Zurich, Switzerland). The output was normalized by dividing the relative fluorescence units (RFU) by the OD. Assays were performed as two independent triplicate sets.

Assay for repression by EsaR variants in vivo.

pBAD constructs encoding wild-type EsaR, deletion variants, or PC variants were also cotransformed with pRNP-lacZ and assayed under the following conditions to determine the ability of a variant to repress transcription. Overnight cultures were subcultured to an OD₆₀₀ of 0.05 in 5 ml of RM with 0.2% arabinose. Cells were harvested at an OD₆₀₀ of 0.5, and 5-μl aliquots were stored in the −70°C freezer prior to analysis of LacZ expression via chemiluminescent β-galactosidase assays (Tropix).

Western immunoblot analysis.

Western immunoblots were performed in duplicate according to published protocols (5) with the primary antibody, polyclonal anti-EsaR, at a dilution of 1:500.

Northern blot analysis.

Two protocols were used for Northern blots, which were performed in duplicate. For initial analysis of the production of the sRNA, an esaI mutant strain (ESN51) was grown up overnight in LB at 30°C and subcultured into 25 ml of LB to an OD₆₀₀ of 0.025. Specific volumes of culture were harvested as follows: OD₆₀₀ of 0.25 (10 ml), 0.5 (5 ml), 1.0 (1 ml), and 2.0 (0.5 ml). The pellets were thawed at room temperature, and lysed by adding 1 ml of TRIzol reagent (Invitrogen). The RNA was extracted according to the manufacturer's protocol. Then, 15 to 20 μg of RNA was precipitated and electrophoresed on a 1% 3-(N-morpholino)propanesulfonic acid agarose gel. The RNA was then transferred to a nitrocellulose membrane through the use of Turboblotter (Whatman/GE Healthcare) according to the manufacturer's protocol. The primers ESSRF and ESSRR (Table 2) were used to amplify a 500-bp region, downstream of the hypothesized promoter region in P. stewartii, from chromosomal DNA to be used as a probe in the identification of a transcript from a Northern blot. Random primed ³²P-labeled probes were made from the PCR products using a Boehringer-Mannheim kit (Roche Applied Science, Indianapolis, IN). The ³²P-labeled probes were hybridized to the membrane with QuickHyb according to the manufacturer's protocol (Stratagene, La Jolla, CA). The membrane was then air dried and exposed to a Storm phosphorimager (General Electric Company).

A second set of Northern blots was completed to establish that production of the sRNA was under the control of quorum sensing. The esaI mutant strain (ESN51) was grown overnight in LB at 30°C, subcultured into 25 ml of LB containing no AHL or 10 nM, 100 nM, 1 μM, or 10 μM AHL to an OD₆₀₀ of 0.025, and 10-ml samples were harvested from each of the growth conditions at an OD₆₀₀ of 0.25. RNA was extracted as described above, and Northern blot analysis of EsaS was performed using the method described by De Lay and Gottesman (8). A 5′-biotinylated probe, ESAS (Table 2) was used for detection of the sRNA. Detection of the probe was achieved with a LAS 4000 miniseries luminescent image analyzer (FujiFilm, Stamford, CT).

RESULTS AND DISCUSSION

Ability of EsaR to activate promoters with the esaR box versus the lux box.

The quorum-sensing regulator EsaR from P. stewartii was initially identified and characterized as a repressor of transcription, versus the majority of its homologues from the LuxR family of proteins, which function as activators of transcription. Recent studies on EsaR in recombinant E. coli demonstrated that EsaR retains the ability to also activate transcription of the V. fischeri luxI promoter, which was fused to a lacZ reporter. EsaR activated the reporter at a level ∼4-fold lower than LuxR (34). Taken together, with electrophoretic mobility shift assay analysis, which suggested that EsaR has an ∼4-fold weaker affinity for the lux box (34), it appeared that this variation in levels of expression between the two homologues could be attributed to their differential affinity for the DNA. Nevertheless, it was demonstrated that EsaR can bind to the lux box and make appropriate interactions with E. coli RNAP to activate transcription from the luxI promoter.

The lux box is a 20-bp palindrome, which differs by 5 bp from the native DNA binding site of EsaR at P_esaR, the esaR box. Thus, two new reporters were constructed that contained the esaR box in two separate orientations, in place of the lux box in the luxI promoter, to establish whether or not this would enhance DNA binding by EsaR in vivo (Fig. 2A). In this assay system, EsaR can function as an activator only in the absence of AHL, whereas LuxR requires AHL to activate transcription. Incorporating the native DNA binding site of EsaR into the reporter did indeed allow for a stronger interaction between EsaR and its native target DNA. EsaR in the presence of the esa1 box but, in the absence of AHL, expressed LacZ at levels comparable to those generated when LuxR drove expression of the reporter in the presence of the lux box and 3-oxo-C6-HSL. Hence, it was a difference in protein-DNA interactions rather than an inability to recognize E. coli RNA polymerase that yielded the observed fourfold difference in activation between LuxR and EsaR in the previous study (34). The reporter under the expression of the promoter containing the esa2 box in the second orientation had lower levels of expression than the one containing the esa1 box (Fig. 2B). Hence, the esa1 box is in the preferred orientation for EsaR to activate transcription, suggesting that bases at positions 6 and 20 are more important for recognition of the DNA than bases 1 and 15.

FIG. 2.

Promoter constructs and activation assays. (A) Sequences are shown for constructs with the native binding sites for LuxR, the lux box, and for EsaR, the esaR box, which were cloned into reporter constructs at a position centered around −42.5. The esaR box was subcloned in two different orientations, and differences from the lux box are in grey and underlined. (B) β-Galactosidase assays were performed on E. coli strains with these pEXT22-based promoter constructs and either pBAD-LuxR or pBAD-EsaR. Constructs and growth conditions are as indicated. Raw data values in terms of LacZ activity in RLU are reported from samples obtained in two independent trials that were tested in triplicate. Error bars represent the standard deviation from the mean.

Effect of deletions in EsaR on its ability to regulate transcription.

A comparative analysis of EsaR to LuxR and TraR was initiated, with regard to domain stability and function. LuxR from V. fischeri is stable and functions in an AHL-independent manner in the absence of its N-terminal domain (29). In comparison, TraR from A. tumefaciens requires both the N- and C-terminal domains for the protein to remain stable and functional (21, 27). Several deletion variants of EsaR were analyzed with respect to their stability and ability to regulate transcription. Two truncated variants (Δ2-160 and Δ2-178) were constructed that lacked portions of the N-terminal domain (Fig. 1). Western immunoblots from recombinant E. coli cells overexpressing these constructs revealed that the variants were unstable (Fig. 3A). These results suggest that in regards to stability of the C-terminal domain, EsaR, more closely behaves like TraR.

FIG. 3.

Expression and activity of EsaR deletion variants. (A) Western immunoblot illustrating the expression levels of deletion variants of EsaR with wild-type EsaR (lanes 1 and 5) and the variants Δ171-178 (lane 2), Δ237-249 (lane 3), ΔΔ (lane 4), Δ2-160 (lane 6), Δ2-178 (lane 7), and Δ65-107 (ΔABD) (lane 8). (B) Repression assays of E. coli strains expressing EsaR deletion variants as indicated. The strain containing the pBAD22 vector was used as a control to determine maximum levels of expression from the constitutively expressed promoter. Raw data values in terms of LacZ activity in RLU are reported from samples obtained in two independent trials that were tested in triplicate. Error bars represent the standard deviation from the mean.

The subset of the LuxR protein family originally identified as repressors, like EsaR (3, 14), which functions in the absence of the AHL, contains two unique regions in comparison to the majority of the LuxR protein family. These two regions consist of (i) an extended linker region between the AHL-binding N-terminal domain and the DNA-binding C-terminal domain and (residues 171 to 178) and (ii) an extension at the C terminus (residues 237 to 249) (Fig. 1) (Thode et al., unpublished). Deletion variants in which these two regions were removed from EsaR singly and in combination were constructed (Δ171-178, Δ237-249, and ΔΔ) (Fig. 1) to reveal the possible roles these unique regions have in protein stability and/or function. Another deletion made at residues 65 to 107 (ΔABD) removed a large portion of the hypothesized 3-oxo-C6-HSL binding region in the N-terminal domain. Western blot analysis of these variants confirmed that all four variants were expressed and remained stable within the cells (Fig. 3A). However, none of the deletion variants were capable of binding to the esaR box and repressing transcription (Fig. 3B). Therefore, the deletions rendered EsaR nonfunctional as a transcriptional regulator.

Role of individual amino acid residues in PC by EsaR.

A further comparison of EsaR, LuxR and TraR was conducted to examine the specific protein-protein interactions that occur between EsaR and RNAP during transcriptional activation. A large number of residues in both the N- and C-terminal domains of TraR (D10, G123, W184, V187, K189, E193, V197, and D217) play critical roles in activation (21, 37). Previous work on LuxR revealed three specific residues (K200, W203, and I208) in the C-terminal domain that were required for activation of transcription at the luxI promoter (11). Interestingly, three residues from TraR (K189, E193, and V197) align closely with the PC variants of LuxR. Through sequence alignments and homology modeling of EsaR (based on the structure of TraR) (Thode et al., unpublished), five amino acids (E7 and N119 in the N-terminal domain, and K196, A199, and I204 in the C-terminal domain) were selected for examination as playing a role in PC control of transcription through analysis of EsaR variants.

In order to characterize the PC variants of EsaR, repression assays were used to differentiate between variants that lost their affinity for DNA compared to those that retained it. Several of the variants (E7A, N119A, K196A, and A199W) retained their ability to block transcription from the esaR promoter by binding the esaR box, which is centered around the −10 site. One variant, I204A, had higher levels of LacZ expression compared to wild type, suggesting that this variant lost some affinity for the DNA (Fig. 4B).

FIG. 4.

Examination of PC variants of EsaR. (A) Western immunoblot illustrating the expression levels of PC variants of EsaR as follows: wild-type EsaR (lane 1), E7A (lane 2), N119A (lane 3), K196A (lane 4), A199W (lane 5), and I204A (lane 6). (B) Repression assays of E. coli strains expressing EsaR point variants were performed as indicated. The strain containing the pBAD22 vector was used as a control to determine maximum levels of expression from the constitutively expressed promoter. Raw data values in terms of LacZ activity in RLU are reported from samples obtained in two independent trials that were tested in triplicate. Error bars represent the standard deviation from the mean. (C) Ability of EsaR point variants to activate transcription in recombinant E. coli. Constructs and growth conditions are as indicated. Raw data values in terms of LacZ activity in RLU are reported from samples obtained in two independent trials that were tested in triplicate. Error bars represent the standard deviation from the mean.

The newly developed promoter construct with the esa1 box centered at −42.5 of the luxI promoter was used in activation assays to test the putative PC variants. Activation assays performed on the C-terminal domain variants K196A and A199W yielded significant decreases in activation compared to the wild-type protein. The I204A variant also had decreased levels of activation, but this is attributed to its decreased affinity to the DNA (Fig. 4C). Activation assays were also performed on two N-terminal domain variants of EsaR. The E7A variant activated transcription at an ∼3-fold lower level than wild-type EsaR. Assays performed on the N119A variant showed no significant decrease in activation (Fig. 4C).

Overall, results from these assays revealed that the E7A, K196A, and A199W variants lost the ability to activate transcription but were still capable of binding to the DNA at levels comparable to wild-type EsaR based on standard repression assays. This result was surprising for the E7A variant since it is not as stable as wild-type EsaR based on Western immunoblot analysis (Fig. 4A). This mutant protein may be capable of binding to the DNA with higher affinity, but this should not have influenced the activation assay results. The I204A variant appears to have a decreased affinity for DNA, as is demonstrated by the increased levels of β-galactosidase activity in the repression assay. Taken together, these results demonstrate that like TraR and unlike LuxR, both the N-terminal domain (E7A) and the C-terminal domain (K196A and A199W) of EsaR appear to play key roles in transcriptional activation.

EsaR-dependent activation of a native promoter in P. stewartii.

The ability of EsaR to function as an activator was examined in the heterologous host, E. coli, using primarily artificial promoter constructs. However, a native promoter activated by EsaR in P. stewartii has now been identified. The EsaR-activated promoter is divergently transcribed from the esaR promoter (Fig. 5A). A gfp fusion to this class I-type promoter, which extends 60 bp downstream from the center of the esaR box, was downregulated 35-fold by the addition of AHL in the presence of EsaR in P. stewartii (41,189 ± 615 RFU in the absence of AHL and 1177 ± 55 RLU in the presence of AHL). This was a surprising result, since the construct had been made to serve as a negative control for EsaR-dependent transcription. To establish that EsaR was directly regulating this native promoter, its activity was also examined in recombinant E. coli. In the presence of EsaR a 45-fold increase in activation was observed in the absence of AHL. In comparison, when AHL was included in the assays no significant increase in activation was observed compared to background (Fig. 5B). These results confirm that biologically activate at EsaR is the only protein from P. stewartii necessary to active transcription of this native promoter in a heterologous host and that all of the essential promoter elements required for EsaR-dependent control were within the cloned region.

FIG. 5.

Analysis of the sRNA divergently transcribed from esaR. (A) Diagram model, not to scale, of the promoter region controlling the divergently transcribed genes esaR and esaS. The +1 site for esaS indicates the point of the fusion with the gfp reporter. (B) Ability of EsaR to activate transcription of P_esaS in E. coli Top10. Constructs and growth conditions are as indicated. Dividing the RFU by the OD normalized the output. Assays were performed as two independent triplicate sets. Error bars represent the standard deviation from the mean. (C) Northern blot analysis examining the expression of the EsaR-dependent sRNA in an esaI mutant strain of P. stewartii, ESN51, at different OD₆₀₀ values. Size standards in bases are indicated to the left. (D) Northern blot analysis examining the expression of the EsaR-dependent sRNA in an esaI mutant strain of P. stewartii, ESN51, in the presence of no AHL or 10 nM, 100 nM, 1 μM, or 10 μM AHL at an OD₆₀₀ of 0.25. Size standards in bases are indicated to the left.

Analysis of the sequence downstream of the EsaR-activated promoter suggested that the closest open reading frame was 564 bp away, with homology to a glutathionylspermidine synthase. Therefore, Northern blot analysis was performed to determine the size of the transcript produced from this EsaR-activated promoter. No transcript was observed from the wild-type strain DC283 (data not shown); apparently, it is produced at levels below the limits of detection. However, using a probe that encompassed a 500-bp region, a single band slightly larger than 100 bases in size was found to be constitutively produced from the ESN51 strain lacking EsaI (and thereby the ability to produce AHL) (Fig. 5C). When exogenous AHL was added to cultures of ESN51, it was demonstrated that increasing concentrations of AHL decreased transcript production (Fig. 5D). Thus, the transcript is synthesized in P. stewartii in an EsaR-dependent manner. There are no obvious rho-independent terminators at the end of the transcript region. Preliminary primer extension analysis (data not shown) suggested that the transcript might be processed at the 5′ end. However, additional analysis will be required to define the precise endpoints of the mature transcript. Because of the relatively small size of the transcript and the fact that no clear open reading frame is present, it is hypothesized that the transcript may function as a small regulatory RNA within P. stewartii, and it has been tentatively named EsaS.

Not unexpectedly, there are no obvious targets for the sRNA in the P. stewartii genome based on complementary nucleotide sequences. Efforts to establish its function within the cell are ongoing. Interestingly, in the homologous YenR/I quorum-sensing system from Yersinia enterocolitica, a YenR-activated promoter controls expression of a putative sRNA divergently transcribed from yenR (36). Multiple cases of coupled regulation of a divergently encoded small regulatory RNA, and its transcriptional regulators have also been characterized in E. coli. Examples of this type of organization include SgrR/S (33), OxyR/S (1), and GcvA/B (31).

Identification of a transcript activated by EsaR in P. stewartii establishes that this regulatory function has been retained in the native host. Therefore, EsaR can function as both an activator and a repressor at low cell density, with accumulation of AHL leading to inactivation and derepression of certain transcripts at high cell density. This dual-level control may afford an advantage to P. stewartii as it progresses from its initial colonization of the host to expression of tissue-destructive virulence factors.