Functional Characterization of the Quorum Sensing Regulator RsaL in the Plant-Beneficial Strain Pseudomonas putida WCS358

Abstract

In many bacteria, quorum sensing (QS) systems rely on a signal receptor and a synthase producing N-acyl-homoserine lactone(s) as the signal molecule(s). In some species, the rsaL gene, located between the signal receptor and synthase genes, encodes a repressor limiting signal synthase expression and hence signal molecule production. Here we investigate the molecular mechanism of action of the RsaL protein in the plant growth-promoting rhizobacterium Pseudomonas putida WCS358 (RsaL_WCS). In P. putida WCS358, RsaL_WCS displayed a strong repressive effect on the promoter of the QS signal synthase gene, ppuI, while it did not repress the same promoter in Pseudomonas aeruginosa. DNase I protection assays showed that purified RsaL_WCS specifically binds to ppuI on a DNA region overlapping the predicted σ⁷⁰-binding site, but such protection was observed only at high protein concentrations. Accordingly, electrophoretic mobility shift assays showed that the RsaL_WCS protein was not able to form stable complexes efficiently with a probe encompassing the ppuI promoter, while it formed stable complexes with the promoter of lasI, the gene orthologous to ppuI in P. aeruginosa. This difference seems to be dictated by the lower dyad symmetry of the RsaL_WCS-binding sequence on the ppuI promoter relative to that on the lasI promoter. Comparison of the results obtained in vivo and in vitro suggests that RsaL_WCS needs a molecular interactor/cofactor specific for P. putida WCS358 to repress ppuI transcription. We also demonstrate that RsaL_WCS regulates siderophore-mediated growth limitation of plant pathogens and biofilm formation, two processes relevant for plant growth-promoting activity.

INTRODUCTION

The cell-cell communication system termed quorum sensing (QS) allows bacterial populations to coordinate gene expression in response to cell density. It is believed that QS constitutes a central element for the social life of bacteria, conferring to the members of a bacterial community the ability to behave as an organized multicellular organism (2, 4).

QS systems are based on the production and secretion of signal molecules that accumulate in the extracellular milieu. At a certain concentration, corresponding to the “quorum” cell density, the signal molecules are perceived by dedicated receptors that, once activated, trigger a physiological response concerted in the whole population. The majority of QS systems thus far described in Gram-negative proteobacteria rely on N-acyl homoserine lactones (N-acyl-HSL) as signal molecules; these QS systems have been found in more than 100 bacterial species (14).

Several members of the Pseudomonas genus regulate their social behavior through N-acyl-HSL-based QS systems. In particular, N-acyl-HSL-dependent regulation plays a major role in the pathogenesis of the opportunistic pathogen Pseudomonas aeruginosa and is involved in plant colonization by several Pseudomonas rhizosphere strains (5, 12, 18, 23, 24).

Pseudomonas putida WCS358 is a plant growth-promoting rhizobacterium (PGPR) originally isolated from the rhizosphere of potato (15). It can protect plants from pathogenic bacteria by competing for iron through the production of the fluorescent siderophore pyoverdine (formerly named pseudobactin) and can induce systemic resistance to different pathogens in Arabidopsis thaliana, bean, and tomato (6, 7, 25, 40).

The orthologous ppu and las QS loci of P. putida WCS358 and P. aeruginosa PAO1, respectively, share an elevated degree of homology at both the nucleotide and amino acid levels and are similarly organized (Fig. 1); the ppuI and lasI genes code for the enzymes that produce the signal molecule N-(3-oxododecanoyl)homoserine lactone (3OC₁₂-HSL), while the ppuR and lasR genes code for the receptors of this signal. The rsaL gene is located intergenically between the synthase and receptor genes and codes for the RsaL transcriptional regulator, which, in both species, strongly represses the transcription of the signal synthase gene and the production of 3OC₁₂-HSL (Fig. 1) (8, 30).

Fig 1.

The las and ppu genetic loci of P. aeruginosa PAO1 and P. putida WCS358. (A) Schematic model representing the interplay between LasR and RsaL in regulating lasI transcription in P. aeruginosa PAO1. The LasI synthase produces the QS signal molecule 3OC₁₂-HSL. Once 3OC₁₂-HSL reaches the threshold concentration, it binds to LasR. The LasR/3OC₁₂-HSL complex binds to the lasI-rsaL bidirectional promoter, activating the transcription of lasI and rsaL. Subsequently, RsaL binds close to LasR on the same bidirectional promoter, repressing the transcription of both the lasI and rsaL genes. (B) Schematic organization of the ppu QS locus in P. putida WCS358. The putative PpuR-binding site, the lux box, is indicated. The levels of sequence identity between the genes of the ppu and las loci are shown.

The regulatory role of RsaL in P. aeruginosa has been thoroughly investigated. In this bacterium, the rsaL mutation leads to enhancement of surface motility (twitching and swarming) and of the production of secreted virulence factors (i.e., elastase, hemolysins, and hydrogen cyanide) with respect to the wild-type strain. Accordingly, this mutant is hypervirulent in the Galleria mellonella wax moth virulence model. Moreover, the P. aeruginosa PAO1 rsaL mutant produces less biofilm and shows enhanced antibiotic susceptibility with respect to the wild type (31). These phenotypic alterations are partially due to the inability of a P. aeruginosa rsaL mutant to maintain the QS signal molecule 3OC₁₂-HSL at physiologically advantageous levels (32). Indeed, the repressive effect exerted by RsaL on lasI transcription counteracts the positive-feedback loop generated by the activator LasR, resulting in QS homeostasis (Fig. 1A).

The ppu QS locus identified in P. putida WCS358 is 99% identical to those described for the P. putida strains IsoF and PCL1445. In the latter strains, QS plays a key role in biofilm formation, a process important for rhizosphere colonization and plant growth-promoting activity (3, 11, 35). The roles of the ppu QS system and rsaL in the physiology of P. putida WCS358 have not been investigated thus far.

From a structural point of view, the RsaL protein of P. aeruginosa PAO1 (RsaL_PAO) belongs to a new subfamily within the tetrahelical superclass of helix-turn-helix (H-T-H) proteins. Although RsaL_PAO is a monomer in solution, it binds as a dimer to an inverted repeat on the lasI promoter. The palindromic structure of this cis-acting element is necessary for RsaL_PAO dimerization and DNA binding (29). The RsaL protein of P. putida WCS358 (RsaL_WCS) is 42% identical to RsaL_PAO. The predicted tridimensional structures of the two proteins revealed a common fold consisting of four α-helices, where helices 2 and 3 form an H-T-H motif for DNA binding. The residues that in RsaL_PAO are involved in protein-DNA interaction are also conserved in RsaL_WCS (29). Interestingly, the structure of RsaL_PAO includes a fifth α-helix, involved in protein dimerization, that is lacking in RsaL_WCS. In addition, obvious palindromic sequences could not be identified on the ppuI promoter (29). These observations suggest that, despite the overall similarity between the las and ppu QS systems, RsaL_WCS could exert its regulatory role through a mechanism of action different from that of RsaL_PAO.

In this study, the DNA-binding properties of the RsaL_WCS protein showed that RsaL_WCS behaves differently from RsaL_PAO and requires an additional factor for full activity. In addition, the effects of RsaL_WCS and QS on phenotypes related to the plant growth-promoting activity of strain WCS358 have been investigated.

MATERIALS AND METHODS

Bacteria, growth conditions, and plasmids.

The bacterial strains and plasmids used in this study are listed in Table 1. Bacterial strains were routinely grown in Luria-Bertani (LB) broth (33) at 37°C (Escherichia coli and Pseudomonas aeruginosa) or 30°C (Pseudomonas putida and Erwinia carotovora) in 10 ml of broth and 100-ml flasks with shaking at 200 rpm. When required, the following antibiotics were added (at the concentrations [in micrograms per milliliter] given in parentheses): for E. coli, ampicillin (Ap) (100), chloramphenicol (Cm) (30), kanamycin (Km) (25), or gentamicin (Gm) (10); for Pseudomonas spp., Gm (50), tetracycline (Tc) (100), or Km (100). To select Pseudomonas spp. from E. coli after mating experiments, LB agar plates were supplemented with nalidixic acid at 15 μg/ml.

Table 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant characteristic(s)a	Reference or source
Strains
P. putida WCS358	Wild type	15
P. putida WCS358 rsaL	rsaL1640::Tn5 mutant (formerly IBE3); Kmr	8
P. putida WCS358 ppuI	ppuI::Km mutant (formerly IBE5); Kmr	39
P. aeruginosa PAO1	Wild type	University of Washington Genome Center
P. aeruginosa PAO1 rsaL	rsal::ISlacZ/hah; Tcr	University of Washington Genome Center
E. coli DH5α	Cloning strain; Nalr	16
E. coli BL21(DE3, pLysS)	High-stringency expression host; Cmr	Merck KGaA, Darmstadt, Germany
E. coli DH5α(pRK2013)	Helper strain for triparental mating; Kmr	13
E. carotovora subsp. carotovora	Wild type, prototroph	1
Plasmids
pBluescript SK II(+)	Cloning vector; Apr	Stratagene, La Jolla, CA
pET-28b(+)	Expression vector; Kmr	Merck KGaA, Darmstadt, Germany
pBBR1MCS-5	Broad-host-range expression vector; Gmr	21
pDRIVE	Cloning vector; Apr	Qiagen, Venlo, Netherlands
pPUI220	pMP220 derivative plasmid containing the PppuI::lacZ transcriptional fusion; Tcr	8
pRsaLWCS6H	pET-28b(+) derivative plasmid for the overexpression in E. coli of a recombinant RsaLWCS protein carrying a 6-histidine tag at the C terminus; Kmr	This study
pPSRsaLWCS6H	pBBR1MCS-5 derivative plasmid for the expression in Pseudomonas of a recombinant RsaLWCS protein carrying a 6-histidine tag at the C terminus; Gmr	This study
pPppuI5′	pBluescript SK II(+) derivative plasmid for the generation of the PppuI 5′ probe used in the DNase I protection expt; Apr	This study
pPppuI3′	pBluescript SK II(+) derivative plasmid for the generation of the PppuI 3′ probe used in the DNase I protection expt; Apr	This study
pPppuIEMSA	pDRIVE derivative plasmid for the generation of the PppuI probe used in the EMSA expt; Apr	This study

^aAntibiotic resistance is indicated by a superscript “r.” Abbreviations for antibiotics are defined in Materials and Methods.

DNA manipulation.

The oligonucleotides used in this study are listed in Table 2. Plasmid DNA preparation, enzyme digestion, agarose gel electrophoresis, and ligation were performed using standard methods (33). Triparental matings between Pseudomonas strains and E. coli were carried out with the helper strain E. coli DH5α(pRK2013) (13). Automated sequencing was performed to check the DNA inserts of all the plasmids generated in this study (MWG Biotech, Ebersberg, Germany).

Table 2.

Oligonucleotides used in this study

Primer	Sequencea	Restriction site
FW119	5′-CATGCCATGGAGCTACTCAATACCTC-3′	NcoI
RV120	5′-CCCAAGCTTCTGCGGGCGTTGGGCC-3′	HindIII
FW140	5′-CGGAATTCGGCTGATGGCTTGATGGG-3′	EcoRI
RV141	5′-CGGGATCCCAGCGGGATGTCCCACAT-3′	BamHI
FW142	5′-GCTCTAGAGGCTGATGGCTTGATGGG-3′	XbaI
FW195	5′-ACAAGTCTGGTTAATGTCGG-3′
RV196	5′-GCAATTACCGGCAGGGGT-3′
FW0	5′-GGGTACCAATAATTTTGTTTAACTTTA-3′	KpnI
RV0	5′-GGGATCCATTGCTCAGCGGTGGCAGC-3′	BamHI

^aRestriction sites are underlined.

Protein purification.

The rsaL gene from P. putida WCS358 was amplified with primers FW119 and RV120 (Table 2) and was cloned by NcoI-HindIII digestion into pET-28b(+) (Merck KGaA, Darmstadt, Germany). The resulting plasmid, pRsaL_WCS6H, was introduced into E. coli BL21(DE3, pLysS) (Merck KGaA, Darmstadt, Germany) for protein overexpression. Six-histidine (6×His)-tagged RsaL_WCS (RsaL_WCS6H) was purified by Ni²⁺ affinity chromatography as described previously for RsaL_PAO6H (30). The protein was further purified by gel permeation chromatography using a Superdex 200 HR column (Amersham Pharmacia Biotech Inc., Piscataway, NJ) according to the manufacturer's instructions.

SDS-PAGE and immunoblotting.

E. coli BL21(DE3, pLysS, pRsaL_WCS6H) cells were lysed and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) prior to electrophoretic transfer to nitrocellulose membranes. Western immunoblotting (33) was carried out with a mouse anti-6×His antibody (1:1,000; Sigma-Aldrich, St. Louis, MO) and with horseradish peroxidase-conjugated anti-mouse IgG as the secondary antibody (1:7,500; Promega, Madison, WI). The blot was then processed with the Amersham ECL chemiluminescent reagents (Amersham Pharmacia Biotech Inc., Piscataway, NJ), as recommended by the manufacturer.

DNase I protection assay.

DNase I protection assays were performed as described previously (30). Briefly, a 305-bp DNA fragment encompassing the PppuI promoter region was amplified either with primers FW140 and RV141 or with primers FW142 and RV141 (Table 2) and was cloned into the pBluescript SK II(+) vector by using the EcoRI-BamHI and XbaI-BamHI restriction sites, respectively. The resulting pPppuI3′ and pPppuI5′ plasmids (Table 1) were digested by EcoRI-SacI, and the DNA fragments containing the PppuI promoter were labeled by fill-in with [α-³²P]dATP and Klenow enzyme (22). The labeled probes (concentration, 0.5 nM) were mixed with different amounts of RsaL_WCS6H in DNase buffer [43 mM Tris-acetate, 30 mM potassium acetate, 8 mM MgCl₂, 27 mM ammonium acetate, 1 mM dithiothreitol, 80 mM KCl, 2 mM CaCl₂, 10% (vol/vol) glycerol, 4% (wt/vol) polyethylene glycol, 100 μg/ml bovine serum albumin, 100 μg/ml poly(dI-dC) (pH 8.0)]. DNA-protein complexes were formed in a total volume of 50 μl per reaction for 15 min at 30°C, followed by 1 min at 25°C. Afterwards, DNase I (0.4 U; Roche, Basel, Switzerland) was added to the reaction mixtures. The reaction mixtures were incubated for 1 min at 25°C; then the reactions were stopped by the addition of 150 μl of Stop solution (0.2 M sodium acetate, 0.1 M EDTA, 0.15% [wt/vol] sodium dodecyl sulfate, 100 μg/ml tRNA [pH 8.0]). DNA from the samples was extracted with phenol-chloroform, precipitated with ethanol, and dissolved in 5 μl of sequence loading buffer (33). After 3 min of denaturation at 95°C, DNA was loaded onto a 7% DNA sequencing gel (33). The A+G Maxam-Gilbert reaction was carried out with the same probe, and the reaction product was loaded onto the gel along with the samples (33).

EMSA.

Electrophoretic mobility shift assays (EMSA) were performed as described previously (30). In brief, a 103-bp DNA fragment encompassing the PppuI promoter was PCR amplified with primers FW195 and RV196 (Table 2) and was TA cloned into the pDRIVE vector (Qiagen, Venlo, Netherlands). The resulting pPppuIEMSA plasmid (Table 1) was digested with EcoRI, and the DNA fragment corresponding to the PppuI promoter was labeled by fill-in with Klenow enzyme and [α-³²P]dATP (22). The radiolabeled DNA probe was purified by using a SigmaSpin postreaction purification column (Sigma-Aldrich), followed by phenol-chloroform (1:1) extraction and ethanol precipitation. In a typical assay, the labeled DNA probe (final concentration, 2 nM) was incubated with different amounts of RsaL_WCS6H in binding buffer [20 mM Tris-HCl, 2 mM EDTA, 5 mM MgCl₂, 30 mM KCl, 5% (vol/vol) glycerol, 0.025% (vol/vol) NP-40, 30 μg/ml poly(dI-dC) (pH 8.0)]. The reactions were carried out in a final volume of 10 μl. After 15 min of incubation at 30°C, the reaction mixtures were loaded onto a 30-min-prerun 8% (wt/vol) polyacrylamide gel under nondenaturing conditions. Electrophoresis was carried out at 4°C in 0.5× Tris-borate-EDTA buffer at 5 V/cm for 2.5 h. The gels were then dried and autoradiographed.

RsaL_PAO6H purification, EMSA with RsaL_PAO6H on the PlasI promoter, and the generation of the PlasI probe were performed as described previously (30).

Measurement of β-galactosidase activity.

The pPUI220 plasmid (Table 1), containing the PppuI::lacZ transcriptional fusion, was introduced independently into the wild-type and rsaL genetic backgrounds of P. putida WCS358 and P. aeruginosa PAO1. In the resulting strains, the pPSRsaL_WCS6H plasmid (for RsaL_WCS complementation) or the empty plasmid pBBR1MCS-5 was introduced (Table 1).

β-Galactosidase activity was determined in the resulting strains as described previously (26). The experiments were performed in triplicate at different points of the growth curve on bacterial cultures grown in LB broth supplemented with 100 μg of Tc/ml and 50 μg of Gm/ml at 30°C (P. putida WCS358) or 37°C (P. aeruginosa PAO1) with shaking at 200 rpm. The statistical significance of the differences observed in mean invasion frequencies was determined by calculating the P values using the two-tailed Student t test for unpaired data sets.

pPSRsaL_WCS6H was obtained by cloning into the KpnI-BamHI-digested vector pBBR1MCS-5 a PCR fragment corresponding to RsaL_WCS6H amplified with primers FW0 and RV0 (Table 2), with plasmid pRsaL_WCS6H as the template (Table 1).

Phenotypic assays.

Siderophore-dependent inhibition of the growth of E. carotovora subsp. carotovora was carried out as described previously (1) with a few modifications. M9 agar plates (33) supplemented with 0.4% (wt/vol) glucose were inoculated in the center with 5-μl suspensions of the different P. putida strains at an absorbance at 600 nm (A₆₀₀) of 0.1. The plates were first incubated for 24 h at 30°C, then sprayed with a suspension of E. carotovora subsp. carotovora at an A₆₀₀ of 0.5, and finally incubated for an additional 24 h at 30°C. To assess the dependency of growth inhibition on siderophore production, the experiment was also performed on M9 agar plates supplemented with 0.4% (wt/vol) glucose and 100 μM FeCl₃.

Pyoverdine production was measured in cultures of P. putida strains grown in M9 liquid medium (33) supplemented with 0.4% (wt/vol) glucose as described previously (19). Briefly, pyoverdine was measured as the A₄₀₅ of culture supernatants appropriately diluted in 100 mM Tris-HCl (pH 8.0), and this value was normalized to the cell density of bacterial cultures (A₆₀₀).

Biofilm formation was determined as described previously (27) with a few modifications. Briefly, P. putida strains were inoculated into polystyrene microtiter plates in AB medium (9) supplemented with 0.2% (wt/vol) glucose at an A₆₀₀ of 0.1. Cultures were incubated for 24 h at 30°C, and cells were washed out of the polystyrene microtiter plates prior to crystal violet staining and measurement of A₅₉₅.

For all the phenotypic assays, the statistical significance of the observed differences in mean invasion frequencies was determined by calculating the P values using the two-tailed Student t test for unpaired data sets.

RESULTS

Identification of the RsaL_WCS binding site on PppuI.

To investigate the molecular mechanism of action of RsaL_WCS, a chimeric variant of this protein fused with a six-histidine (6×His) tag at the C-terminal domain was purified by Ni-nitrilotriacetic acid (NTA) affinity chromatography under nondenaturing conditions. As shown in the SDS-PAGE analysis reported in Fig. 2A, a protein with a molecular size consistent with RsaL_WCS6H (∼10 kDa) was eluted from the Ni-NTA resin with a good degree of purity (>90%). The identity of this purified protein as RsaL_WCS6H was verified by Western blot analysis with an anti-6×His antibody (Fig. 2B).

Fig 2.

Purification of RsaL_WCS6H. (A) SDS-PAGE analysis of protein samples from key steps of RsaL_WCS6H purification. Lane 1, PageRuler protein ladder (Fermentas) (only molecular size markers from 10 kDa to 50 kDa are indicated); lane 2, crude extract of uninduced protein; lane 3, crude extract of induced protein; lane 4, crude extract of the insoluble fraction of the induced protein; lane 5, crude extract of the soluble fraction of the induced protein; lane 6, purified protein. (B) Western blot analysis performed with an anti-6×His antibody on an SDS-PAGE gel identical to that shown in panel A.

DNase I protection assays showed that RsaL_WCS6H protects a 22-bp DNA region spanning nucleotide −40 to nucleotide −61 with respect to the ppuI translational start site (Fig. 3A). This region partially overlaps with the putative −10 consensus sequence for σ⁷⁰ binding to the promoter, in accordance with the role of RsaL_WCS as a transcriptional repressor of ppuI (Fig. 3B) (8).

Fig 3.

Identification of the RsaL_WCS6H binding site on PppuI. (A) DNase I protection assay performed on the PppuI promoter with increasing concentrations of purified RsaL_WCS6H. Both strands of the tested sequence were independently labeled in order to investigate the protection pattern on the bottom strand (left) and top strand (right). Thick lines indicate the regions showing specific protection by RsaL_WCS6H. The putative transcriptional start site (+1) and −10 sequence are indicated. Lanes 1, Maxam and Gilbert sequencing reaction (A+G); lanes 2, no RsaL_WCS6H added; lanes 3 and 4, RsaL_WCS6H added to final concentrations of 30 and 60 μM, respectively. (B) Sequence of the PppuI promoter region. The putative transcriptional start site (+1) and −10 consensus sequence for σ⁷⁰ binding are indicated. The ATG start codon of PpuI is shown in boldface. The RsaL_WCS6H-protected region is shaded. The putative PpuR-binding site, the lux box, is boxed. (C) Alignment of the RsaL_PAO-binding site on PlasI with the RsaL_WCS-binding site on PppuI. Arrows indicate the residues organized as inverted repeats. Nucleotides whose replacement affects the binding of RsaL_PAO to PlasI are shown in boldface (29). Conserved nucleotides are shaded.

In a previous study, a mutagenesis analysis allowed the identification of single nucleotides necessary for the binding of RsaL_PAO to a palindromic sequence on the PlasI promoter, defining the half-site of the palindrome as TATGNAA (29). Since the amino acids of RsaL_PAO involved in protein-DNA interaction are conserved in RsaL_WCS (29), we reasoned that the two proteins could bind to a similar DNA sequence. In the PppuI region protected by RsaL_WCS, a 15-bp sequence showing 60% identity with the RsaL_PAO-binding site was identified. As shown in Fig. 3C, of the first seven nucleotides at the 5′ end of this sequence, TATTTAA, five are conserved with respect to the half-site for RsaL_PAO binding on PlasI, while the rest of this sequence lacks clear dyad symmetry, with only two residues organized as an inverted repeat (Fig. 3C). Therefore, the RsaL_WCS-binding site on PppuI is not palindromic, but it contains a motif resembling the half-site site for RsaL_PAO binding to PlasI. This suggests that RsaL_WCS might bind its target promoter as a monomer. The monomeric status of purified RsaL_WCS6H in solution was verified by gel filtration (data not shown).

The differences in the DNA sequences of the two promoters might explain why the affinity of RsaL_WCS for PppuI was lower than that of RsaL_PAO for PlasI. Indeed, RsaL_WCS was able to protect its target promoter from DNase I digestion at a concentration of 30 to 60 μM, while RsaL_PAO had the same effect on PlasI at concentrations ranging from 0.5 to 5 μM (30). We also tested the ability of RsaL_WCS to protect the lasI promoter from DNase I digestion and found that RsaL_WCS and RsaL_PAO protected the same DNA sequence on this promoter (data not shown).

The RsaL_WCS/PppuI complex is not stable in vitro.

To clarify the DNA-binding properties of RsaL_WCS for the PppuI promoter, purified RsaL_WCS6H was tested in electrophoretic mobility shift assays (EMSA) on a DNA probe encompassing the PppuI promoter. As shown in Fig. 4A, RsaL_WCS6H was not able to form stable complexes with the tested probe, and only a faint band corresponding to an unstable RsaL_WCS/probe complex appeared after a long exposure of the film. In these assays, RsaL_WCS6H was not able to shift more than 10% of the probe even at high concentrations, and further addition of the protein resulted in aspecific binding (data not shown). This result is in accordance with the high concentration of RsaL_WCS6H required to obtain protection from DNase I digestion and indicates that in vitro, RsaL_WCS6H has a low binding affinity overall for the ppuI promoter.

Fig 4.

Interactions of the RsaL proteins with PppuI and PlasI. EMSA show the interactions of RsaL_WCS6H with a probe encompassing the PppuI promoter (A), RsaL_WCS6H with a probe encompassing the PlasI promoter (B), and RsaL_PAO6H with a probe encompassing the PppuI promoter (C). The final protein concentration (μM) in each lane is indicated. The free DNA probes are indicated by arrows.

Interestingly, purified RsaL_WCS6H was able to form stable protein-DNA complexes with a probe encompassing the PlasI promoter region (Fig. 4B) (30). This result demonstrates that the RsaL_WCS6H protein is functional in DNA binding and that the in vitro binding affinity of RsaL_WCS6H for its cognate promoter is lower than that for the heterologous lasI promoter. In contrast, the purified RsaL_PAO6H protein, previously shown to bind the PlasI promoter (30), was not able to form stable complexes with the PppuI probe (Fig. 4C), a behavior very similar to that of RsaL_WCS6H with the same probe in the EMSA reported in Fig. 4A. These data are in line with our previous results showing that RsaL_PAO6H requires a concrete palindromic structure for dimerization and in vitro DNA binding (29).

In summary, these data demonstrated that in vitro, the purified RsaL proteins are not effective in binding the PppuI promoter region, while both proteins can efficiently form stable complexes with the PlasI promoter. This difference is likely driven by the palindromic structure of the RsaL-binding DNA sequence present on PlasI (29) but lacking on the PppuI promoter.

RsaL_WCS represses PppuI activity in P. putida WCS358 but not in P. aeruginosa PAO1.

The poor binding of RsaL_WCS to its target promoter was not expected, since RsaL_WCS strongly represses ppuI transcription and 3OC₁₂-HSL production in vivo (8). To explain this discrepancy, we postulated that the lack of a palindromic structure in the RsaL_WCS-binding site on PppuI, resulting in low affinity in vitro, could be compensated for by a cofactor/interactor required for full RsaL_WCS activity in P. putida WCS358 in vivo. To test this hypothesis, PppuI activity was measured in wild-type and rsaL mutant strains of P. putida WCS358 and P. aeruginosa PAO1 by means of a PppuI::lacZ transcriptional fusion.

As shown in Fig. 5A, PppuI activity was about 10-fold higher in an rsaL mutant of P. putida WCS358 than in the parental strain, and this effect was complemented by expressing RsaL_WCS with the pBBR1MCS-5-derived vector pPSRsaL_WCS6H. The wild-type and rsaL mutant strains carrying the empty pBBR1MCS-5 vector were used as controls. The growth curves of the mutant strains carrying the pBBR1MCS-5 or pPSRsaL_WCS6H plasmid were comparable to the growth curves obtained for the respective parental strains carrying the pBBR1MCS-5 vector (data not shown).

Fig 5.

Effects of rsaL mutation and complementation on PppuI activity. Histograms represent the activity of the ppuI promoter at different points of the growth curve (expressed as A₆₀₀) in P. putida WCS358 (A) or P. aeruginosa PAO1 (B) carrying a PppuI::lacZ transcriptional fusion in the pPUI220 plasmid. Open bars, promoter activity in wild-type strains carrying the pBBR1MCS-5 empty vector; light and dark shaded bars, promoter activity in rsaL mutant strains carrying the empty vector pBBR1MCS-5 or the pPSRsaL_WCS6H plasmid, respectively. Averages for three independent experiments with standard deviation are shown. Statistical significance with respect to the wild-type strains is indicated by one asterisk (P < 0.05) or two asterisks (P < 0.01). M.u., Miller units.

In P. aeruginosa PAO1, the PppuI promoter was active even in the absence of the cognate PpuR activator, as indicated by the high promoter activity measured in this strain (maximal activity, >14,000 Miller units, about 3-fold higher than that in P. putida WCS358 [Fig. 5B]). This effect is likely due to a cross-complementation of PpuR with the orthologous LuxR family regulator LasR. As shown in Fig. 5B, PppuI activity was less than 1.5-fold higher in the P. aeruginosa PAO1 rsaL mutant than in the wild-type strain, indicating that RsaL_PAO has only a minor repressive effect on PppuI expression. Most importantly, since introduction of the pPSRsaL_WCS6H plasmid, expressing RsaL_WCS, into the rsaL mutant of P. aeruginosa had no effect on PppuI activity, it is clear that RsaL_WCS cannot repress its target promoter in P. aeruginosa (Fig. 5B). The empty plasmid pBBR1MCS-5 was introduced into the wild-type and rsaL P. aeruginosa PAO1 strains as a control.

In conclusion, RsaL_WCS can fully repress ppuI transcription in P. putida WCS358 but not in the heterologous host P. aeruginosa PAO1. These data suggest that an additional cofactor/interactor required for RsaL_WCS activity is present in P. putida WCS358.

RsaL is involved in processes relevant for plant growth promotion in P. putida WCS358.

We performed experiments aimed at determining whether RsaL regulates pyoverdine-mediated biocontrol of plant pathogens and biofilm formation, two pro-cesses relevant for plant growth-promoting activity in P. putida WCS358.

An antibiosis plate assay based on inhibition of the growth of the plant pathogen Erwinia carotovora (1) was used to compare the inhibitory activity of wild-type P. putida WCS358 with those of its isogenic rsaL and ppuI mutants. The results showed that all strains were able to inhibit E. carotovora under conditions of iron limitation (Fig. 6A). UV inspection of iron-poor plates disclosed the fluorescent halo typical of pyoverdine, corresponding to the E. carotovora growth inhibition zone (data not shown). Conversely, E. carotovora growth was not inhibited when iron-rich plates were used to suppress pyoverdine production (data not shown). Interestingly, the rsaL mutant strain caused the formation of an inhibition halo larger than those with the other two strains (Fig. 6A). In accordance with this observation, measurements of pyoverdine production in liquid cultures showed that the levels of this fluorescent siderophore were about 5-fold higher in the P. putida WCS358 rsaL mutant than in the wild type and the ppuI mutant (Fig. 6B).

Fig 6.

Phenotypic effects of rsaL and ppuI mutations in P. putida WCS358. (A) Histogram representing the diameters of the E. carotovora subsp. carotovora inhibition halos due to pyoverdine production in the indicated P. putida WCS358 strains grown on M9 agar plates. (B) Quantification of pyoverdine in the indicated P. putida WCS358 strains. Strains were grown in M9 liquid medium. (C) Biofilm formation by the indicated P. putida WCS358 strains, grown in AB medium, was measured by adhesion to polystyrene microtiter plates. Each bar represents the average for three independent experiments; error bars, standard deviations. Statistical significance with respect to the wild-type strain is indicated by two asterisks (P < 0.01).

Overall, these experiments confirm that pyoverdine production could play an important role in the biocontrol activity of P. putida WCS358, and they demonstrate that RsaL is a negative regulator of this process. Conversely, the abolition of 3OC₁₂-HSL production in the ppuI mutant had no relevant effect on this phenotype.

We also tested the effects of the ppuI and rsaL mutations on biofilm formation by using a standard adhesion assay in microtiter plates (27). The wild-type strain and the ppuI mutant exhibited similar levels of biofilm production, while the rsaL mutant produced less biofilm than the wild type (Fig. 6C). This indicates that RsaL has a positive effect on biofilm formation, while, for this phenotype also, the abolition of 3OC₁₂-HSL production in the ppuI mutant showed no significant effect.

DISCUSSION

To date, the rsaL gene has been found linked to acyl-HSL-dependent QS systems in P. aeruginosa (10, 36), in P. putida strains (8, 11, 35), in Pseudomonas fuscovaginae (24), and in a cluster of 30 Burkholderia spp. (37, 38). In all these bacteria, RsaL represses the expression of the signal synthase gene and (with the sole exception of P. fuscovaginae) causes a decrease in signal molecule production. However, the RsaL protein has been investigated at the physiological and molecular levels only in P. aeruginosa PAO1 (29–32).

In this study, we carried out a comparative analysis of RsaL proteins from P. putida WCS358 and P. aeruginosa, disclosing new properties of this regulator. Moreover, the role played by RsaL and QS in regulating phenotypes relevant for the phytostimulatory activity of P. putida WCS358 has been investigated for the first time.

The molecular mechanism by which RsaL represses the transcription of the synthase genes ppuI and lasI seems to be conserved in P. putida WCS358 with respect to P. aeruginosa PAO1, since in both strains, RsaL binds the respective promoter (PppuI or PlasI) in vitro on a region overlapping the −10 consensus sequence for σ⁷⁰ (30). However, the DNA sequences protected by the RsaL proteins on their respective target promoters disclose interesting differences. In P. aeruginosa PAO1, the RsaL_PAO-binding site on PlasI is a palindromic motif that allows RsaL_PAO dimerization on the target promoter (29, 32). Conversely, the presence of a site with poor dyad symmetry on PppuI most probably affects the abilities of both RsaL_WCS and RsaL_PAO to form stable protein-DNA complexes with a PppuI probe in vitro. However, RsaL_WCS has a strong repressing effect on ppuI transcription in P. putida WCS358 in vivo, while it does not repress the same promoter in the heterologous host P. aeruginosa PAO1. These observations suggest that RsaL_WCS needs a molecular interactor/cofactor specific for P. putida WCS358 in order to repress ppuI transcription.

Previous in silico modeling of the structures of RsaL_PAO and RsaL_WCS supports this hypothesis. This analysis revealed that the overall structure of the RsaL proteins is very similar to those of the N-terminal domain of the lambda cI repressor (λcI-NTD) and the POU-specific domain of the mammalian transcription factor Oct-1 (Oct-1 POUs), and it clustered the RsaL proteins in the tetrahelical superclass of helix-turn-helix proteins (29). λcI-NTD binds to a palindromic sequence of DNA as a homodimer, and dimer formation in this protein is driven by α-helix 5 (20). This α-helix is absent in Oct-1 POUs, a protein domain that binds to DNA in association with the POU homeodomain (28). Therefore, the observation that α-helix 5 is also present in RsaL_PAO but absent in RsaL_WCS (29) supports the idea that RsaL_PAO, like λcI-NTD, binds to a palindromic DNA sequence as a homodimer, while the binding of RsaL_WCS to DNA, like that of Oct-1 POUs, requires an association with another DNA-binding protein.

In this context, it is noteworthy that a motif corresponding to the half-binding site of RsaL_PAO was also identified in the promoter regions of other genes controlled by RsaL_PAO in P. aeruginosa, such as cupA1, hcp1, cmpX, and PA3174 (Fig. 7) (31, 32). Interestingly, purified RsaL_PAO was not able to interact with the promoter regions of genes such as cupA1 and PA3174 in vitro (data not shown). Therefore, it is possible that RsaL_PAO acts as a homodimer in controlling the activity of promoters containing a palindromic sequence, while it could act as a heterodimer on other promoters. This appealing hypothesis is under investigation in our laboratory.

Fig 7.

Alignment of the palindromic RsaL_PAO-binding site on the PlasI promoter with DNA motifs containing a half-site site for RsaL_PAO binding on other promoters regulated by RsaL in P. aeruginosa PAO1. Arrows indicate the residues organized as inverted repeats on PlasI. Nucleotides whose replacement affected the binding of RsaL_PAO to PlasI are shown in boldface (29). The conserved RsaL_PAO half-binding site is shaded.

This putative molecular mechanism resembles the tissue specificity described for Oct-1 in mammals. Indeed, small polypeptides with a tissue-specific pattern of expression can alter the DNA-binding ability of Oct-1 by interacting with this transcriptional regulator, resulting in distinct regulons controlled by Oct-1 in different tissues (28).

The notion that QS does not rely solely on cell density is now well established. In P. aeruginosa, many regulators have been shown to affect the expression of QS signal synthase and receptor genes, likely integrating cell density sensing with the response to other environmental signals. However, the biological relevance and the mechanisms underlying the regulation of QS genes/proteins by many of these regulators remain far from clear (34, 41, 43).

The evidence that QS in P. putida WCS358 is repressed by RsaL_WCS, most probably in association with another factor, raises the possibility that the cell density-dependent regulation of QS in this bacterium is integrated with other physiological and/or environmental cues at the level of RsaL. In this context, it is important that in WCS358, the transcriptional regulator PpuR does not affect RsaL_WCS expression (8). This implies that in P. putida WCS358, RsaL_WCS and its putative interactor might be expressed or active under defined conditions not necessarily related to cell density. Therefore, RsaL-mediated repression could constitute a key element connecting the expression of ppuI, and hence the production of the QS signal molecule, with other metabolic or external stimuli in P. putida WCS358. Conversely, in P. aeruginosa PAO1, the expression of RsaL_PAO is dependent on LasR (10, 32); therefore, RsaL_PAO alone is sufficient to repress lasI and signal molecule production in the absence of an interactor, independently of stimuli other than cell density. This does not exclude the possibility that environmentally dependent modulation of lasI expression could still occur in P. aeruginosa PAO1, for instance, at the level of lasR expression. Of note, for the RsaL_PAO-dependent promoters containing only a half-binding site, the response to cell density and to other environmental cues could be integrated by modulating the availability of an RsaL_PAO interactor. The identification of these possible RsaL interactors in P. aeruginosa PAO1 and P. putida WCS358 merits further investigation.

In this study, we also provide evidence that in P. putida WCS358, RsaL plays a positive role in biofilm formation, while it acts as a repressor of pyoverdine production. These effects could be due to the strong increase in the level of 3OC₁₂-HSL production in the rsaL mutant, even if RsaL could also exert its regulative role independently of QS. The observation that the abolition of 3OC₁₂-HSL production in the ppuI mutant had no significant effect on biofilm formation and pyoverdine production supports the latter hypothesis. The physiological role played by RsaL in P. putida WCS358 is similar to that observed in P. aeruginosa PAO1, since the RsaL protein of P. aeruginosa was also shown to have a positive effect on biofilm formation and to repress the production of secreted secondary metabolites (31). In contrast, the lack of an effect of the ppuI mutation on the biofilm phenotype was surprising, since previous studies with P. putida strains IsoF and PCL1445 showed that this mutation causes a significant alteration in biofilm production (3, 11, 35). However, this difference could be strain specific or could depend on the specific experimental conditions and growth media used in this study.

The phytostimulatory activity of PGPR is due to a combination of different mechanisms, including biocontrol of plant pathogens, production of plant growth-promoting hormones, and induction of systemic resistance (17). In P. putida WCS358, biofilm formation could be important for root colonization, while it is well established that iron sequestration by siderophores plays an important role in limiting the growth of plant pathogens, as we have shown in the case of E. carotovora. Since RsaL has opposite effects on these two phenotypes, it is not possible to conclude whether this regulator has a beneficial or a detrimental effect on phytostimulatory activity overall. RsaL-like proteins present in different bacteria seem to play a key role in the control of energy-consuming processes involved in adaptation to complex environmental niches, such as the tissues of higher organisms (42). It is tempting to speculate that the opposite regulation of two energetically expensive processes, such as biofilm formation and siderophore production, mediated by RsaL could help to optimally manage energy utilization with respect to the availability of nutritional resources, ultimately enhancing the fitness of root-colonizing bacteria. Overall, the results discussed above indicate that RsaL-dependent regulation could play a leading role in P. putida WCS358 root colonization and phytostimulatory activity, and they should inspire further studies in planta.