The rcsA Promoter of Pantoea stewartii subsp. stewartii Features a Low-Level Constitutive Promoter and an EsaR Quorum-Sensing-Regulated Promoter

Abstract

The upstream region of the Pantoea stewartii rcsA gene features two promoters, one for constitutive basal-level expression and a second autoregulated promoter for induced expression. The EsaR quorum-sensing repressor binds to a site centered between the two promoters, blocking transcription elongation from the regulated promoter under noninducing conditions.

Pantoea stewartii subsp. stewartii is the etiological agent of Stewart's wilt disease in maize (2, 3). The organism grows in the plant xylem, where it produces large amounts of stewartan capsular polysaccharide/exopolysaccharide (EPS), thereby impeding normal vascular transport (16). A cluster of 12 genes, designated cps, encodes the functions required for stewartan biosynthesis and translocation (5, 8). This gene system is related to group 1 cps gene clusters based on the chromosomal linkage to the rfb and his locus and regulation by the RcsC/YojN/RcsA/B phosphorelay signal transduction system (17, 24, 32). The RcsAB heterodimeric transcription factor binds to an RcsAB box element generally located some distance (∼100 bp) from the regulated promoters (31).

In P. stewartii subsp. stewartii, the synthesis of stewartan EPS is cell density dependent and governed by the EsaI/EsaR quorum-sensing system (28). The EsaI protein is an acyl homoserine lactone (AHL) signal synthase (27, 30), and EsaR is the cognate AHL-responsive transcription factor (19, 28). This system differs from the LuxR paradigm in that EsaR dimerizes and binds DNA in the absence of the AHL coinducer, thereby repressing or activating target genes depending on the location of the esaR box DNA binding site (18, 21, 26). We recently reported that EsaR controls the expression of the cps gene system indirectly through transcriptional repression and AHL-mediated derepression of the rcsA gene (18). Quorum-sensing systems govern the synthesis of EPS in other phytopathogenic bacteria, but the underlying regulatory mechanisms differ (7, 9, 22). To our knowledge, P. stewartii subsp. stewartii represents the only described example of a species in which the rcsA gene is under direct quorum-sensing control (18).

In this study, we identify the transcriptional start site(s) and promoter site(s) within the upstream region of the rcsA gene to further define the mode of quorum-sensing regulation at this promoter. We observe that the rcsA gene expresses at significant basal levels from a previously unrecognized constitutive promoter and that the induced expression initiates from a separate, AHL-inducible, RcsA-autoregulated promoter. The previously established EsaR DNA binding site (18, 19) is centered between the two promoters, suggesting that EsaR represses transcription by interfering with transcription elongation rather than by steric exclusion of RNA polymerase from the inducible promoter.

Bacterial strains and growth conditions.

The P. stewartii subsp. stewartii strains used in this study were ESN10 and ESN51, which are mutant strains derived from the wild-type strain DC283 (8). These strains carry different mutated alleles of the esaI gene (18, 27). The esaI gene in ESN10 is disrupted with a chloramphenicol resistance cassette (18), while strain ESN51 carries a Tn5 (kanamycin resistance) transposon insertion (27). Strain PSS11 is an rcsA mutant derivative of ESN10 (18). Escherichia coli strain S17-1 (23) was used for the transfer of reporter plasmids into P. stewartii subsp. stewartii. Strains were grown in Luria-Bertani medium and AB minimal medium-0.2% glucose (4). P. stewartii subsp. stewartii was grown at 28°C and E. coli at 37°C.

Constitutive, RcsA-independent transcription of the rcsA gene.

Preliminary data suggested that the rcsA gene expresses at significant basal levels. To verify this observation, we created plasmid pAUC30, which carries the 600-bp upstream DNA region of the rcsA gene cloned into a gfpmut promoter probe vector, pFPV25 (18, 25). This plasmid was conjugally transferred into the esaI, AHL signal synthase-deficient P. stewartii subsp. stewartii strain ESN10 and separately into the esaI rcsA double mutant strain PSS11 for parallel evaluation. The levels of green fluorescent protein (GFP)-specific fluorescence expressed by ESN10(pAUC30) and PSS11(pAUC30) were measured on 200-μl culture aliquots using a FLUOstar OPTIMA microtiter plate reader (BMG Labtech, Durham, NC). The data summarized in Table 1 show that the rcsA promoter expresses at similar significant basal levels in both strains under noninducing conditions, indicating that basal-level expression is RcsA and AHL independent. These data also show that induction of the reporter fusion by AHL-mediated quorum sensing requires a functional copy of rcsA, which is present in strain ESN10 but lacking in PSS11. These genetic data corroborate our prediction that EsaR-mediated quorum sensing governs rcsA transcription by blocking the RcsA-positive feedback loop (18).

TABLE 1.

GFP expression from the rcsA promoter in the esaI strain ESN10 and esaI rcsA double mutant strain PSS11 under noninducing and inducing conditions

Induction	Mean GFP expression (relative units)± 95% CIa
	ESN10		PSS11
	pFPV25	pAUC30	PFPV25	pAUC30
No AHL	2.5 ± 0.3	29.6 ± 0.8	2.4 ± 0.2	29.6 ± 3.2
AHL	2.4 ± 0.3	100.9 ± 3.6	2.4 ± 0.3	31.2 ± 4.6

^aThe values (10³, average of four measurements) have been rounded off to the nearest decimal. Error values indicate 95% confidence intervals (CI).

The upstream region of the rcsA gene features two promoters.

To further define the mechanism for basal and induced rcsA expression, we created a set of pFPV25-based GFP reporter fusions carrying partial DNA fragments of the 600-bp rcsA upstream region. We cloned DNA segments (schematically illustrated in Fig. 1A) corresponding to bp −588 to −380, −588 to −316, and −385 to +13 from the translation initiation codon after PCR amplification by using the primer sets 5′-CCATAGGATCCAAATTCACAACTATCC and 5′-ACAAGAAGCTTCACACAATATTTTTTCT to create the reporter plasmid pAUC31, 5′-CCATAGGATCCAAATTCACAACTATCC and 5′-CATTTAAGCTTCGAAAGTGTAAGGCTGA to generate the reporter plasmid pAUC32, and 5′-CAAGAGGATCCATTGTGTGATTTTTCTT and 5′-ATAATAAGCTTCATGTTAGCGACCCTCA to create the reporter plasmid pAUC33, respectively. The individual reporter constructs were transferred into the esaI mutant strain ESN51 (28). Each strain was grown to an optical density at 600 nm of 0.5 in AB minimal medium-glucose with and without inducing levels of AHL extracts (1,000×). AHL from crude wild-type and esaI mutant (control) culture supernatants was extracted by a method adapted from Elasri et al. (10). GFP fluorescence of the induced and uninduced cultures was measured as described previously. The data summarized in Fig. 1B show that the reporter constructs pAUC30 (full-length promoter region), pAUC32 (partial fragment, bp −588 to −316), and pAUC33 (partial fragment, bp −385 to + 13) all exhibit similar, significant basal levels of GFP expression in the absence of exogenous AHL. In contrast, the reporter plasmid pAUC31 (partial promoter fragment, bp −588 to −380) lacks basal-level GFP expression. We conclude that the region between bp −380 and −326 from the ATG translation initiation codon contains a putative constitutive promoter. The promoter fragments carried on pAUC30, pAUC31, and pAUC32 are AHL responsive, while that borne on pAUC33 is not. The responsive promoter fragments each feature a conserved RcsAB box (31) and the previously established EsaR binding site (18), both of which are lacking in the promoter fragment of pAUC33. The fact that the promoter fusion of pAUC31 is AHL inducible establishes a likely location for the regulated promoter in the region between bp −588 and −380. Interestingly, the expression levels from the partial promoter fragment carried on plasmid pAUC32 (bp −588 to −316) are approximately twofold higher than those obtained from the full-length rcsA promoter fusion carried on plasmid pAUC30. This subtle difference may be due to undefined structural silencing effects associated with the untranslated 5′ region of the mRNA, as suggested by Majdalani and Gottesman (17).

FIG. 1.

Genetic analysis of rcsA promoter activity. (A) Schematic illustration of the promoter constructs used in this study. Numbers refer to the distance from the translational start site of rcsA, with the +1 position corresponding to the A residue of the ATG translation initiation codon. The EsaR binding site and conserved RcsAB box are indicated. (B) Relative GFP activities of the promoter fusions. Strain ESN51 harboring the reporter plasmids was grown in the presence (open bars) or absence (shaded bars) of AHL. Error bars indicate 95% confidence intervals. All values are based on multiple assays of three independent cultures.

Biochemical evidence for two promoters.

We performed primer extension analysis of mRNAs extracted from strain ESN51 grown in AB minimal medium-glucose with and without AHL supplementation. Total RNA was isolated using the Masterpure RNA kit (Epicenter, Madison, WI) by following the protocol recommended by the manufacturer. Contaminating genomic DNA was removed by RQ1 DNase (Promega, Madison, WI). The primer 5′-GTCGAATCCACATCGGGTATTGATGTAA was 5′-end phosphorylated in the presence of T4 kinase (Epicenter, Madison, WI) and [γ-³²P]ATP (Perkin-Elmer, Boston, MA). The primer extension reactions were carried out in the presence of Superscript III reverse transcriptase (Invitrogen) at 55°C for 1 h. The autoradiogram in Fig. 2A shows two major primer extension products. Uninduced cells yield significant levels of a shorter transcript with an apparent adenosine (A) transcriptional start site at residue −336 from the translation initiation codon. A conserved σ⁷⁰ promoter consensus sequence is properly positioned from the identified transcriptional start site and represents the likely constitutive promoter. The second, larger extension product corresponds to a transcript with an apparent guanosine (G) transcriptional start site located at bp −460 from the translation initiation codon. This transcript appears primarily under AHL-inducing conditions. A likely inducible promoter located between bp −497 and −470 (Fig. 2B) precedes the G transcriptional start site. Interestingly, this promoter is located upstream of the previously established EsaR binding site (bp −421 and −401) (18). In addition, a well-conserved RcsAB binding motif (78.6% identity to the consensus TaAGaatatTCctA sequence [31], where the uppercase letters indicate conserved residues) is located 25 bp upstream of the −35 promoter consensus sequence of this predicted inducible promoter. The assignment of these inducible promoter elements differs from those specified in a previous publication, which was based on bioinformatics predictions and the assumption that EsaR blocks transcription initiation (18).

FIG. 2.

Primer extension analysis of the rcsA promoter. (A) Radiolabeled primer extension products of RNA templates extracted from strain ESN51 grown under noninducing and inducing conditions were resolved by denaturing polyacrylamide gel electrophoresis (lanes 1 and 2, respectively) and detected by phosphorimaging (Bio-Rad molecular imager FX). The reference DNA sequence was generated with the same primer using plasmid pAUC30 as a template. The gray and black arrows mark the positions of the constitutive and inducible transcription start sites, respectively. (B) rcsA promoter sequence. Numbers in the margin correspond to the distance from the rcsA translation start codon. Underlined characters indicate σ⁷⁰ promoter consensus sequences. The esaR and rcsAB boxes are highlighted in black and gray, respectively. Arrows designate the corresponding transcription start sites.

Quorum-sensing regulation of the rcsA gene.

The genetic and biochemical data presented in this report allow us to develop a model for the expression of an RcsA coactivator. In this model, the rcsA promoter of P. stewartii subsp. stewartii features an upstream, quorum-sensing-responsive, RcsA-autoactivated promoter and a downstream constitutive promoter. The two promoters are separated by a 97-bp region that includes the cis-acting esaR box element (18). At a low cell density and low AHL concentration, ligand-free EsaR binds to the esaR box, which presumably blocks transcription elongation from the upstream promoter. Transcription from the constitutive promoter remains unaltered. At a high cell density and high AHL concentration, EsaR repression relaxes (19), allowing transcription from the upstream promoter and priming of a positive RcsA-dependent feedback loop for the induced transcription of the rcsA gene. This hierarchical regulatory scheme remains unique, to our knowledge.

An inherent feature of many quorum-sensing systems is the positive-feedback regulatory circuit at the level of autoinducer signal synthesis, which is thought to be a key factor in the switch-like control of quorum-sensing-regulated phenotypes (11-14, 29). However, in P. stewartii subsp. stewartii, AHL synthesis is constitutive, and the expression of esaR is autorepressed (19, 27). Thus, the positive-feedback regulatory loop at the RcsA level may be a key contributing factor to a rapid “off-on” switching mechanism for EPS synthesis. In this regard, we have shown that the onset of EPS synthesis in P. stewartii subsp. stewartii is strictly cell density linked, generally ensuing at ∼2 × 10⁸ cells/ml when grown in liquid culture (28).

The EsaR/RcsA regulatory hierarchy may represent a strategy for coupling the Rcs environmental sensory pathway with the EsaR population density sensory system, presumably to ensure that EPS synthesis occurs exclusively when both inputs coincide. The integration of the rcsA and quorum-sensing signaling networks, however, may play an even broader role in governing the transition between successive stages of biofilm development in P. stewartii subsp. stewartii. For example, in E. coli and Salmonella enterica serovar Typhi, the RcsAB heterodimer represses surface adhesion and motility, which are functions typical of early biofilm development (1, 6, 15, 17, 20). Conversely, RcsAB activates the synthesis of EPS, which contributes to biofilm maturation. In addition, RcsB plays a role in the activation of cell division independently of RcsA (17). Assuming that RcsA and RcsB serve similar roles in P. stewartii subsp. stewartii, the quorum-sensing control of the RcsA coactivator may represent a key mechanism for coordinating the differential expression of physiological processes that define population development.