OUP user menu

The alternative sigma factor RpoN regulates the quorum sensing gene rhlI in Pseudomonas aeruginosa

Lyndal S. Thompson, Jeremy S. Webb, Scott A. Rice, Staffan Kjelleberg
DOI: http://dx.doi.org/10.1016/S0378-1097(03)00097-1 187-195 First published online: 1 March 2003

Abstract

The rhl quorum sensing (QS) circuit of Pseudomonas aeruginosa is known to regulate the expression of a number of virulence factors. This study investigates the regulation of rhlI, encoding the auto-inducer synthase RhlI responsible for the synthesis of N-butryl-l-homoserine lactone (BHL). A putative RpoN binding site was located upstream, in the promoter region of rhlI. Utilising a rhlI-lacZ transcriptional reporter, we demonstrate that under certain media conditions RpoN is a positive regulator of rhlI transcription. Measurements of BHL in extracted supernatant showed that the transcriptional patterns were reflected in the BHL levels, which were reduced in the rpoN mutant. Elastase and pyocyanin, known to be regulated by the rhl QS circuit, were shown to be reduced in a RpoN deficient strain. However, exogenous addition of BHL to the rpoN mutant did not restore these phenotypes suggesting that other regulatory factors apart from BHL are involved. Consistent with other rhl regulated phenotypes, we found that a rpoN mutant strain forms a biofilm that is different from that of the wild-type but similar to that displayed by a rhlI mutant.

Keywords
  • rhlI regulation
  • RpoN
  • Biofilm
  • Quorum sensing
  • Pseudomonas aeruginosa

1 Introduction

Many virulence factors of the opportunistic human pathogen, Pseudomonas aeruginosa, are known to be regulated by quorum sensing circuits [1]. These population monitoring circuits are based on the accumulation of N-acylated homoserine lactone (AHL) molecules. Such quorum sensing circuits have been identified in a number of proteobacteria and have been shown to regulate a range of phenotypes, from biofilm formation in Serratia liquefaciens (Labbate, M., unpublished data), P. aeruginosa [2] and Aeromonas hydrophila [3] to stationary phase survival in Rhizobium leguminosarum [4]. P. aeruginosa contains two AHL regulated circuits that are arranged in a hierarchical cascade with the las circuit regulating expression from the rhl circuit [5]. The AHL circuits in P. aeruginosa have been shown to contribute to the transcriptional regulation of a range of genes [6], however they have mostly been studied for their participation in the regulation of virulence factors [1].

Factors contributing to the pathogenicity of P. aeruginosa have been well studied. Global regulators such as Vfr [7,8], GacA [9], RpoS [10] and the AHL signalling circuits [1,11] have been shown to regulate the expression of virulence factors such as extracellular proteases, lipase, surface attachment and biofilm development, rhamnolipid production, alginate production, chitinases, and the siderophores pyocyanin and pyoverdine. The alternate sigma factor, RpoN, also appears to regulate virulence in this organism [12]. To date, it has been demonstrated that alginate, rhamnolipid, lipase, and type IV fimbrae are all regulated by RpoN (for review see [13]). Furthermore, it has been shown in vivo that a rpoN mutant of P. aeruginosa was 100-fold less virulent in a mouse thermal injury model and significantly impaired in its ability to kill the nematode Caenorhabditis elegans. However, the rpoN mutant did not display any differential virulence in Galleria mellonella (greater wax moth) or in the later stages on an Arabidopsis thaliana leaf infection assay [12].

A number of virulence factors regulated by RpoN are also under the control of the rhl quorum sensing. Due to this coordinate regulation, we hypothesised that expression of the rhl circuit genes and rpoN are linked. Specifically, we were interested in whether a hierarchical regulatory network exists between RpoN and AHL quorum sensing circuits.

This study describes the identification of a RpoN recognition sequence, using sequence analysis, in the promoter region of the rhlI gene, which encodes an N-butryl-l-homoserine lactone (BHL) synthase. RpoN (σ54) directs promoter specific transcription by binding to a relatively conserved sequence of 16 bp in the promoter region of targeted genes [14] making it possible to identify potential RpoN regulated genes from sequence analysis [13]. Using traditional analytical and genetic approaches to investigate the extent of RpoN regulation on the accumulation of BHL, we report that under specific nutrient conditions RpoN acts as a positive effector of BHL accumulation. To discern the significance of this regulation on phenotype expression in P. aeruginosa we compared known rhl regulated phenotypes, elastase and pyocyanin production, in both the wild-type PA01 and a rpoN knockout strain. Although we observed a decrease in expression of these virulence factors in the rpoN mutant, addition of exogenous BHL did not restore these phenotypes to wild-type levels suggesting a more complex regulatory sequence. In addition, we investigated biofilm formation and describe the development and morphology of a rpoN mutant biofilm and show that it is morphologically similar to the phenotype displayed by a rhlI minus strain.

2 Materials and methods

2.1 Bacterial strains, plasmids and growth conditions

Bacterial strains, plasmids and primer sequences used in this study are listed in Table 1. Strains were routinely cultured on Luria–Bertani (LB) medium or on LB plates containing 1.5% agar. Assays were performed as indicated in M9 minimal medium (48 mM Na2HPO4, 22 mM KH2PO4, 8.5 mM NaCl, 18.7 mM NH4Cl, 2 mM MgSO4, 0.1 mM CaCl2) supplemented with 0.2% glucose, low phosphate succinate minimal medium (LPSM) [15] containing 20 mM sodium succinate, 40 mM NH4Cl, 2 mM K2SO4, 0.4 mM MgCl2, 1 µM MnCl2, 1 µM CaSO4, 1 µM ZnCl2, 1 µM FeCl3 and 10 mM morpholino propanesulfonic acid (MOPS) or in Kings complex medium (peptone 1%, glycero1 1%, NaCl 1%). All media were supplemented with 0.2% glutamine. Cultures were grown at 37°C and aerated at 200 rpm. Antibiotics were used at the following concentrations: kanamycin (Km) 50 µg ml−1, gentamicin (Gm) 100 µg ml−1, tetracycline (Tc) 100 µg ml−1 for P. aeruginosa and Gm 35 µg ml−1, Tc 20 µg ml−1 and ampicillin (Amp) 50 µg ml−1 for Escherichia coli strains. 5′-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) was used at a final concentration of 40 µg ml−1.

View this table:
Table 1

Bacterial strains and plasmids used in this study

Strain or plasmidGenotypeSource or reference
Strains
P. aeruginosa
    PA01UNSWWild-type P. aeruginosaUNSW culture collection
    PA01ΔrpoNPA01UNSW ΔrpoN::aacC1, Gmrthis study
    PA01ΔrpoN/NPA01UNSW ΔrpoN::aacC1 with rpoN in attP site, Gmr, Tcrthis study
    PA01ATCCWild-type P. aeruginosa ATCC 15692American type culture collection
    PA01rhlI::TcPA01ATCC ΔrhlI:: Tc[8]
Eschericia. coli
    E. coli DH5αφ80dLacZΔM15 Δ (lacZYA-argF)U169 deoR recA1 hsdR17 (rk, mk+) supE44 λgyrA96 thi-1 relA1 FGibco BRL
    E. coli XL1-BlueΔ(mcrA) 183 Δ (mcrCB-hsdSMR-mrr)173 endA1 supE44 thi-1 recA1 gyrA96 relA1 lac F′proAB lacIqM15 Tn10 (Tetr)Stratagene
Chromobacterium. violaceum
    C. violaceum CV026mini Tn 5 mutant of ATCC 31532[32]
Plasmids
    pKS-(Bluescript)Apr; extended polylinker pUC derivativeStratagene
    pBSRPONpKS-(Bluescript) with 2.1 kb rpoNthis study
    pBSRPONGMGmr pBSRPON with 0.51 kb HindII fragment removed and aacC1 from pUCGM insertedthis study
    pBSRHLpKS-(Bluescript) with 0.3 kb promoter of rhlIthis study
    pBSRHLACpBSRHL with 7.8 kb lacZYAthis study
    pEX100Tgene replacement vector, AproriT+sacB+ (5.8 kb)[16]
    pEXRPONGMΔrpoNGmr insert in pEX100Tthis study
    mini CTX-1Tcr; self-proficient integration vector with tet, Ω-FRT-attP-MCS, ori, int, and oriT[17]
    mini CTX-rpoNmini CTX-1 with 2.1 kb rpoNthis study
    mini CTX-rhlImini CTX-1 with 0.82 kb rhlIthis study
    pME6031PVS1-p15A shuttle vector, Tcr[33]
    pMERPONpME6031 with 2.1 kb rpoNthis study
    pUCGMcloning vector, Gmr, Cbr[34]
    pRK600helper plasmid in triparental matings Cmr ColE1 oriV RP4oriT[35]
Primers
    RpoNF5′-GCTGCAGGAATTCCACATCC-3′this study
    RpoNR5′-CGTGTGTAAGTGGCAGGTTAGC-3′this study
    SacIF5′-ATCCAGAAGGAGCTCGACGC-3′this study
    rhl320R5′-ATCATGGCGGCGGAAAGCCC-3′this study

2.2 Construction of a rpoN knockout mutant

A rpoN deficient mutant was constructed by replacement of the chromosomal rpoN gene using allelic exchange. Briefly, polymerase chain reaction (PCR) amplification was used to obtain a 2.1 kb fragment including 650 bases upstream and 60 bases downstream of rpoN, using the primers RpoNF and RpoNR (Table 1). This fragment was blunt-end ligated into the HindII site of the pKS-(Bluescript) MCS creating pBSRPON. Next a 0.51 kb HindII fragment was deleted from rpoN and replaced with a 0.85 kb HindII fragment containing the aacC1 gene encoding Gm resistance from pUCGM forming pBSRPONGM. The disrupted rpoN::Gm construct was PCR amplified using primers RpoNF and RpoNR and ligated directly into the SmaI site of pEX100T [16], a conjugatable counter-selective suicide plasmid, yielding pEXRPONGM. Triparental conjugal transfer was employed, using pRK600 as the helper, to transfer pEXRPONGM into PA01. Clones in which a double crossover had occurred were selected on plates containing Gm 100 µg ml−1, Km 50 µg ml−1 and 5% sucrose. The resultant rpoN mutants were confirmed by PCR and Southern hybridisation. Further, the RpoN deficiency was confirmed by loss of twitching motility and a requirement for glutamine. One clone was designated PA01ΔrpoN and was subsequently used for further work.

For complementation studies, rpoN was replaced in the chromosome in single copy number at the attP site using the mini CTX-1 system [17]. Mini CTX-rpoN was constructed by ligating a BamHI/XhoI fragment from pBSRPON into the MCS of mini CTX-1 and conjugated into PA01ΔrpoN as previously described.

2.3 rhlI reporter strains

prhlI-lacZ reporters were constructed and inserted into the chromosome of PA01 in single copy at the attP site using the mini CTX-1 [17]. The promoter of the rhlI gene was PCR amplified using primers RhlIF and Rhl320R (Table 1) and ligated into EcoRV of pKS-(Bluescript). Vectors were selected according to the orientation of the prhlI fragment forming pBSRHL. Next a 7.8 kb BamHI fragment from pVK32 containing lacZYA was ligated into BamHI of PBSRHL. Again a vector containing the correctly orientated lacZYA was chosen and designated pBSRHLAC. The reporter construct was obtained on a NotI/XhoI digested fragment and ligated into mini CTX-1. Conjugal transfer was used to move the reporter into both PA01ΔrpoN and the PA01UNSW wild-type.

β-Galactosidase activity was assayed essentially as described [18]. Samples were taken during the growth of PA01UNSW, PA01ΔrpoN and a complemented PA01ΔrpoN containing plasmid pMERPON, grown in M9 and Kings media. Activity was measured using O-nitrophenyl β-d-galactopyranoside (ONPG) and expressed as Miller units, which had been normalised to OD610nm. Each assay was performed in triplicate.

2.4 BHL quantitation assays

BHL was quantitated employing a thin layer chromatography (TLC) bioassay as previously described [19,20]. Five ml samples were taken from different growth phases of the wild-type PA01UNSW, PA01ΔrpoN and the complemented strain PA01ΔrpoN/N. Cell-free supernatant was collected by 0.22 µm filtration and extracted three times with equal volumes of 0.1% acidified ethyl acetate to partition the rhlI native signal, BHL. The solvent was evaporated and the residue re-dissolved in 30 µl of dichloromethane (DCM). Samples were spotted onto C18 reverse-phase TLC plates using capillary tubes and the chromatograms developed with methanol/water (60:40, vol/vol). Once developed, the dried plates were overlaid with 50 ml warm LB agar seeded with 10 ml of a 24 h culture of Chromobacterium violaceum CV026. Plates were then incubated for 12–18 h at 30°C. TLC plates were digitally photographed and the size and intensity of the induced chromatin spot determined using Bio-Rad Multi-Analyst version 1.0.1. BHL levels are reported as pixel density mm−2, or as a percentage of levels measured in the wild-type. Samples were collected from duplicate cultures and separate experiments were conducted twice.

2.5 Bioassays

All bioassays were performed in triplicate. Assessment of virulence factor expression was performed at four stages of the growth cycle: mid exponential phase (OD610nm 0.1), late exponential phase (OD610nm 0.3), transition from exponential to stationary phase and in early (1 h) stationary phase. Cultures used for elastase assays were grown in M9 medium. Elastase was measured as previously described [21]. Cultures used for pyocyanin assays were grown in LPSM medium as the low phosphate increases pyocyanin expression [15]. Pyocyanin was measured as described [22]. Activities from biochemical assays were adjusted to OD610nm of the sampled culture. To complement phenotype expression in the rpoN mutant, BHL was added at the time of inoculation to 10 µM.

2.6 Biofilm experiments

Biofilms were grown on glass coverslips attached to polycarbonate flow cells, 1 mm, 4 mm, 40 mm as previously described [23]. M9 medium supplemented with 0.2% glucose was pumped through the flow cell reactors at approximately 4 ml h−1. Cells inoculated into the flow chamber reactors were left to attach for 1 h before the flow was started. Development of the biofilms was monitored by end point staining using 4,6-diamidino-2-phenylindole (DAPI) and subsequent photographing of the biofilms at 2 days, 4 days, and 6 days using an Olympus, Japan, LSM-HN GIU, confocal scanning laser microscope. PA01UNSW was compared to PA01ΔrpoN, whilst the rhlI mutant, PA01rhlI::Tc, was compared to its parent strain PA01ATCC.

3 Results

3.1 Regulation of rhlI transcription

We identified a recently published RpoN recognition sequence [13] in the promoter of rhlI (Table 2) 144 bp upstream of the start codon and overlapping a previously identified las box and σ70 binding site (Fig. 1) [6,24].

View this table:
Table 2

RpoN consensus sequence found in the promoter region of a specific subset of genes in P. aeruginosa known to be regulated by RpoN

GeneRpoN box
ConsensusYTGGCACGrNNNTTGCW
algDCCGGAACTTCCCTCGCA
glnATGGCACGCAACTTGCT
oprECGGGCCCACGCGTTCCA
lipATGGCACGGTTCCTGCG
fleSRAGGCACGGGTATTGCT
ureD1TGGCAAGGCCCTTGCT
rhlABCCGGGATACCCGTGGCA
−144 rhlICTGGCAGGTTGCCTGCC
  • The consensus sequence identified −144 bp from the rhlI ATG start codon. Bold type indicates the highly conserved GG-N10-GC motif. For review see [13].

Figure 1

The P. aeruginosa rhlI promoter sequence. Bold type indicates the highly conserved GG-N10-GC motif. Underlining indicates a conserved las box sequence. The ATG start of the open reading frame is italicised and the −35 and −10 regions homologous to a σ70 type promoter are shown in lower case (recently published [24]).

To determine whether RpoN regulates rhlI transcription we monitored the rhlI-lacZ reporter in the rpoN mutant. When this experiment was performed in M9 minimal medium, rhlI gene expression was reduced by 70% in the rpoN mutant compared to the wild-type (Fig. 2A). However, when grown in complex medium, we observed no significant difference in rhlI expression in the rpoN mutant compared to the wild-type strain (Fig. 2B). This reduction in expression of rhlI in the rpoN mutant compared to the wild-type was evident from mid exponential growth to early stationary phase. At 24 h (data not shown) there was an approximate 40% reduction in rhlI transcription in the rpoN mutant compared to the PA01 wild-type. This effect could be complemented by trans addition of rpoN on plasmid pMERPON.

Figure 2

Transcription of rhlI in PA01UNSW, the rpoN mutant; PA01ΔrpoN and the complemented rpoN mutant containing plasmid pMERPON, during growth into stationary phase in (A) M9 minimal medium and (B) Kings complex medium. Expression of a single copy rhlI-lacZ transcriptional fusion (open symbols) and growth (closed symbols) was determined in the wild-type PA01 (boxes), PA01ΔrpoN (triangles) and PA01ΔrpoN/N (diamonds). Transcription is expressed in Miller units and is adjusted to culture OD610nm.

3.2 BHL quantitation

It is well established that levels of transcription in the bacterial cell do not always correlate with accumulation patterns of the translational product [25]. Therefore, we were interested in whether changes observed in the transcription of rhlI resulted in changes in the accumulation patterns of BHL. Since P. aeruginosa produces at least two AHLs, we separated the different AHL molecules using thin layer chromatography before quantifying levels of BHL. We were able to visualise and semi-quantitate the amount of BHL by overlaying the TLC with C. violaceum CV026, a signal (luxI homologue) deficient strain that produces a purple pigment in response to short chain AHLs. BHL levels in the wild-type increased from mid exponential (1.3050e3 pixels mm−2) to late exponential phase (9.6350e3 pixels mm−2) where BHL levels peaked. Levels of BHL were lower in stationary phase cultures at 6.5579e3 pixels mm−2 and 1.3534e3 pixels mm−2 at the transition phase and 1 h into stationary phase respectively. We demonstrated that in M9 medium BHL levels did not accumulate in the rpoN mutant at the same rate or to the same level as seen in the wild-type PA01 (Fig. 3). This phenomenon, however, was observed only in the mid exponential phase of growth when BHL levels in the rpoN mutant were only 14% of that seen in the wild-type. This difference gradually diminished as the culture moved into stationary phase with the rpoN mutant accumulating BHL to 82% of that seen in the wild-type in late exponential phase and 100% in early stationary phase. Wild-type levels of BHL accumulation could be restored to the rpoN mutant by complementing the strain with a single copy of rpoN inserted in the chromosome at the attP site.

Figure 3

Density of violacein, produced by the AHL reporter strain C. violaceum, indicating BHL levels expressed as a percentage of the wild-type. Open bars, PA01UNSW; filled bars, PA01ΔrpoN; and hatched bars, the complemented strain PA01ΔrpoN/N.

3.3 Influence of RpoN regulation on AHL regulated phenotypes

The rhl quorum sensing circuit in P. aeruginosa consists of a response regulator, RhlR and a BHL signal synthase, RhlI. It is believed that RhlR can bind to DNA, up-regulating transcription, when associated with the effector molecule BHL [26]. The concentration of BHL in the cell is an important factor in gene regulation by the rhl circuit. Hence, we were interested in whether changes in the observed levels of BHL in the rpoN mutant strain corresponded to changes in the expression of AHL regulated phenotypes.

Several of the known AHL regulated phenotypes, such as rhamnolipid [26] and lipase [27] are known to be co-regulated by both RpoN and the AHL circuits. Elastase and pyocyanin to date have not been reported to be under the control of RpoN, nor did we identify a RpoN binding recognition sequence in the promoter regions of the genes involved in expression of these extracellular factors.

Although elastase is regulated by both the las and rhl circuits in P. aeruginosa we were able to show reduced extracellular accumulation of elastase of up to 70% in the rpoN mutant in late exponential phase (Table 3). This change in elastase expression could be restored by complemention with the rpoN gene, however it could not be restored by exogenous addition of 10 µM BHL.

View this table:
Table 3

Effect of rpoN mutation on the expression of virulence factors in P. aeruginosa in different growth phases

Mid exponentialLate exponentialTransitionEarly stationary
ΔrpoNrpoN/NΔrpoNrpoN/NΔrpoNrpoN/NΔrpoNrpoN/N
Elastase46.9105.030.775.064.097.473.5121
PyocyaninNANA23.7118.642.6131.671.7150.0
  • Elastase assays were performed in M9 medium and pyocyanin assays in low phosphate succinate medium (LPSM) as described in Section 2. NA, data not available due to immeasurable quantities at low cell densities. The complemented strain is designated rpoN/N (Table 1).

Pyocyanin was also used to assess regulation of rhl regulated phenotypes in the rpoN mutant. Glutamine, known to be regulated by RpoN, is a precursor for pyocyanin synthesis. However, by supplementing all media with excess glutamine (0.2%), we were able to complement the growth of the rpoN mutant to wild-type levels and demonstrate that pyocyanin production in the RpoN mutant was significantly reduced (Table 3). In late exponential phase, pyocyanin levels in the rpoN mutant were only 24% of that measured in the wild-type, however by early stationary phase this difference diminished, with the rpoN mutant producing 72% of that observed in the wild-type. Pyocyanin synthesis could be restored to wild-type levels in the rpoN mutant by complementation with a single copy of the rpoN gene introduced onto the chromosome, but as with elastase, could not be complemented by exogenous addition of 10 µM BHL. A statistical significance of P<0.05 was shown using a Student's t-test for all results of the bioassays.

3.4 Role of RpoN in biofilm development

Since there is an overlap in the regulation of virulence genes by RpoN and the rhl circuit, as well as in the regulation of rhlI by RpoN we predicted there would also be similarity in the structure of the rpoN and the rhlI mutant biofilms. Initially we compared the wild-type biofilm to the rpoN mutant. After 4 days the wild-type had formed a thick mat of cells with little micro-colony definition (Fig. 4A). After 6 days the wild-type phenotype was a complex arrangement of micro-colonies densely covering most of the surface. In contrast after 4 days the rpoN mutant had formed defined micro-colonies (Fig. 4B). The rpoN mutant biofilm changed little between 4 and 6 days, however the micro-colonies appeared more tightly packed. The rpoN mutant formed a much poorer biofilm than the wild-type with few cells attached to the surface between the micro-colonies. Wild-type biofilm morphology was restored to the rpoN mutant when complemented with a single copy of rpoN placed in the chromosome (data not shown). Once a phenotype had been established for the rpoN mutant compared to its isogenic parent we could use it in a comparison with the rhlI mutant biofilm morphology with its isogenic parent. The rhlI mutant formed a biofilm in comparison to its parent similar to that of the rpoN mutant (Fig. 4C). Micro-colonies were formed more quickly than observed in the parent wild-type. In contrast to the rpoN mutant however, the rhlI mutant had a thicker coverage of cells on the surface between the micro-colonies.

Figure 4

Biofilm structure of (A) PA01UNSW, (B) ΔrpoN UNSW, and (C) ΔrhlI ATCC at 2 days, 4 days, and 6 days.

4 Discussion

Unlike the binding sites of other sigma factors, RpoN recognises and binds to a highly conserved 16 bp sequence in the promoter region of controlled genes [28]. Since the genome of P. aeruginosa has been sequenced, this allows identification of genes that are putatively regulated by RpoN to be identified. Our analysis of the rhlI promoter sequence showed that a previously identified las box [6] and σ70 binding site [24] overlaps with a RpoN binding sequence, suggesting a dual regulatory mechanism for rhlI transcription.

Our data demonstrate that RpoN is involved in the nutrient dependent transcriptional regulation of rhlI. Using a rhlI-lacZ transcriptional reporter fusion we observed a decrease in the levels of rhlI transcription in the rpoN mutant compared to the wild-type. Under minimal medium conditions there was an approximate 70% reduction in rhlI transcription levels. The difference, observed in rhlI transcription in the rpoN mutant, was not apparent when a complex medium was used for growth. Explanations for this observation include the possibility that RpoN is only able to bind to or initiate transcription of rhlI under specific nutrient conditions. One of the functionally diverse genes that RpoN has been shown to regulate is glnA, encoding glutamine synthase. GlnA is required for synthesis of glutamine from glutamate and ammonia [29]. Hence RpoN activity in the cell may be regulated by the composition, in particular the ammonia concentration, of the growth medium. A recent publication [24] has shown that when P. aeruginosa is grown in tryptic soy broth, a complex medium, rhlI is transcribed predominantly as a single extension product. They suggest that rhlI transcription is from a σ70 type promoter and the las box is the primary operator controlling rhlI expression. The RpoN and σ70 type promoters overlap (Fig. 1), suggesting competition between the two sigma factors. Under conditions where σ70 levels are low, regulation from RpoN may be more consequential, providing an alternative account of the differences we observed between media.

Should RpoN have a significant regulatory influence over the quorum sensing circuits, its transcriptional regulation must impact on the synthesis of BHL. By measuring BHL levels using thin layer chromatography we were able to show that disruption of rpoN resulted in decreased levels of BHL accumulation and hence the involvement of RpoN in the regulation of the rhl quorum sensing circuit. Interestingly, it was only during early logarithmic growth that this phenomenon was apparent, suggesting that at later phases in growth BHL levels are independent of RpoN control. This is in contrast to the transcriptional data, where a difference in rhlI transcription levels is apparent throughout exponential growth and into the stationary phase of the growth cycle. Possible explanations for this discrepancy include changes in the stability of the rhlI mRNA, stability of the RhlI protein in the cytoplasm or the increased turnover of BHL in stationary phase cells [30].

We have demonstrated that RpoN is also controlling the expression of elastase and pyocyanin; both known to be regulated by the rhl circuit. We have, however, not been able to show that the decrease in BHL levels accounts for these changes in phenotype expression. While the phenotypes could not be restored to wild-type levels by the exogenous addition of 10 µM BHL to the rpoN mutant, both elastase and pyocyanin levels were restored by complementation with rpoN. These differences may reflect a number of influencing factors including a possible reduction in the functioning of the type II general secretory pathway in the rpoN mutant, resulting in a decrease in extracellular products. There is also the possibility that RpoN may be indirectly regulating these phenotypes by controlling other factors that are required for maximum phenotype expression.

Our observations show that the biofilm structure of the rpoN mutant was significantly altered compared to that displayed by the wild-type PA01. The rpoN mutant is deficient in a number of parameters that have to date been indicated as important for biofilm formation, including the production of functional fimbrae and flagellar [31], as well as rhamnolipid synthesis (Davey, M.E. and O'Toole, G.A., unpublished data). The rhlI mutant strain was also compared against its parent strain PA01ATTC and was found to have an altered biofilm structure similar to that of the rpoN mutant. It was beyond the scope of this study to identify specific phenotypes of the rpoN and rhlI mutants that contributed to their changed biofilm morphologies.

Many factors have been recognised to contribute to the regulation of the AHL circuits in P. aeruginosa. In this study, we have added RpoN to the list of regulators responsible for directing the rhl circuit. The regulation of rhlI by RpoN is affected at the transcriptional level with a rhlI-lacZ transcriptional fusion demonstrating differential expression in a rpoN mutant strain compared to the wild-type PA01. We have shown that this change in the transcription level is reflected in the accumulation patterns of the signalling molecule C4-HSL (BHL). Furthermore, expression of the rhl regulated phenotypes elastase and pyocyanin was shown to be regulated by RpoN, although it is not clear if this is through changes in BHL accumulation. Under specific conditions a rpoN minus genotype forms a biofilm with isolated micro-colonies, a morphology that is also displayed by the rhlI mutant strain. In conclusion this study has demonstrated that expression of rhlI is under the control of RpoN and that this regulation is nutrient dependent.

Acknowledgements

This study was supported by the Australian Research Council and the Centre for Marine Biofouling and Bio-Innovation (University of New South Wales, Sydney, NSW, Australia). We are grateful to S. Beatson for providing the wild-type PA01ATCC and PA01ΔrhlI::Tc strains.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
View Abstract