Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Jan 3, 2006; 103(1): 171–176.
Published online Dec 22, 2005. doi:  10.1073/pnas.0507407103
PMCID: PMC1324988

Multiple sensors control reciprocal expression of Pseudomonas aeruginosa regulatory RNA and virulence genes


The opportunistic pathogen Pseudomonas aeruginosa is responsible for a wide range of acute and chronic infections. The transition to chronic infections is accompanied by physiological changes in the bacteria favoring formation of biofilm communities. Here we report the identification of LadS, a hybrid sensor kinase that controls the reciprocal expression of genes for type III secretion and biofilm-promoting polysaccharides. Domain organization of LadS and the range of LadS-controlled genes suggest that it counteracts the activities of another sensor kinase, RetS. These two pathways converge by controlling the transcription of a small regulatory RNA, RsmZ. This work identifies a previously undescribed signal transduction network in which the activities of signal-receiving sensor kinases LadS, RetS, and GacS regulate expression of virulence genes associated with acute or chronic infection by transcriptional and posttranscriptional mechanisms.

Keywords: biofilm, pel genes, small RNA, two-component system, type III secretion

Bacteria occupy a wide range of environmental and pathogenic niches by growing planktonicly or in sessile communities known as biofilms. The opportunistic pathogen Pseudomonas aeruginosa can form biofilms on a variety of surfaces, including the cystic fibrosis lung and abiotic surfaces such as contact lenses and catheters (13). Biofilm development is a coordinated series of events beginning with surface attachment by planktonic bacteria, followed by formation of microcolonies and subsequent development of differentiated structures in which individual bacteria as well as the entire community are surrounded by exopolysaccharides (4). This process is likely governed by the activities of regulatory networks that coordinate the temporal expression of various motility, adhesion, and exopolysaccharide genes in response to environmental cues.

Extracellular polysaccharide matrix production is a key attribute of P. aeruginosa growth in biofilm communities, and it contributes to the overall architecture and antibiotic resistance of bacteria in biofilms (5, 6). At least two distinct genetic loci, pel and psl, are required for the formation of these polysaccharides and for biofilm development in vitro (610). Expression of the pel and psl genes is coordinated by the global regulator RetS (11). RetS is an unusual sensor kinase that contains a periplasmic sensor domain linked by a seven-membrane-spanning region to a histidine kinase and two response regulator receiver domains. Deletion of retS results in overexpression of pel and psl genes and increased formation of surface-attached biofilms and matrix-enclosed bacterial aggregates at the air–liquid interface (pellicles). Simultaneously, retS mutants are unable to respond to host-cell contact or media-derived signals that normally activate the expression of genes encoding the type III secretion system (TTSS) (11, 12). The TTSS mediates the delivery of toxic effectors into the host-cell cytoplasm and is required for virulence in diverse animal models of acute infection (13, 14). This reciprocal relationship between biofilm formation and TTSS expression has implicated RetS as a regulator of bacterial behavior during infection. Through the RetS-signaling pathway, P. aeruginosa can sense environmental signals and respond by activating the expression of genes required for acute infection while repressing genes that promote biofilm formation. In addition to causing a variety of acute infections, in cystic fibrosis patients P. aeruginosa can also persist in a biofilm state for decades (15). The observation that chronic clinical isolates frequently accumulate mutations in the TTSS suggests that, although central to virulence in animal models of acute infection and very likely important in certain acute human infections (13), this secretion system is not necessary for maintaining chronic infection (16). P. aeruginosa could mediate this transition to chronic infection through the disappearance of the RetS-activating signals. Alternatively, it is also conceivable that novel signals that become dominant during conditions leading to chronic infections activate different regulatory networks, leading to large-scale reprogramming of the transcriptome to optimize bacterial survival in the new environment.

Here we describe the identification of LadS, a hybrid sensor kinase with domain architecture reminiscent of RetS but that acts in a manner opposite of RetS. We provide evidence that the two signaling pathways both function by influencing the levels of the small RNA RsmZ, an antagonist of the translational repressor RsmA (17, 18). Our results suggest that reciprocal regulation of virulence factors required for acute and chronic infection is achieved not just through the presence or absence of one signal but instead by distinct signals acting through different sensory components.

Materials and Methods

Bacterial Strains, Plasmids, and Growth Conditions. The P. aeruginosa ladS mutants, PAIH21, PAIC41 and PAIJ68, were recovered from a Tn5 insertion library constructed in the nonpiliated PAKΔpilA strain (19, 20). Other P. aeruginosa strains used were PAK, the ladS deletion mutants PAKΔladS and PAKΔpilAΔladS or their gfp-tagged derivatives, the pel mutant PAKΔpelF (10), the retS mutant PAKΔretS (11), the gacS mutant PAKΔgacS, and the rsmA mutant PAKΔrsmA.

Primer sequences are provided in Table 1, which is published as supporting information on the PNAS web site. To construct the ladS deletion mutant, a 3-kbp DNA fragment containing the ladS gene together with its flanking regions was amplified by PCR using primers IV1 and IV2. The PCR product was digested with BstBI to excise an internal fragment (2,379 bp) of the ladS gene. Purified fragments containing the 5′ and 3′ regions of ladS were ligated in the pCR2.1 vector. The truncated ladS gene was excised as a SpeI/EcoRV fragment and subcloned into the pKNG101 suicide vector (21). The resulting plasmid, pKΔS, contains the truncated ladS and its flanking regions. The truncated gene was introduced by allelic exchange into the homologous region of the PAK and PAKΔpilA chromosome as described in ref. 21.

The ladS/retS double mutant was constructed by making a deletion of the retS gene in PAKΔladS as described in ref. 11. Deletions in gacS and rsmA were constructed as described in ref. 22 by using primers PA0928 L5′, PA0928 L3′, PA0928 R5′, PA0928 R3′, and PA0905 L5′, PA0905 L3′, PA0905 R5′, and PA0905 R3′, respectively.

The ladS gene was amplified by PCR by using primers IV3 and IV4 and cloned into the broad host range vector pBBR1MCS4, yielding pBBRladS.

The exoS–lacZ reporter fusion was previously constructed (23). To construct the pel–lacZ and rsmZ–lacZ reporter fusions, upstream DNA fragments were PCR-amplified from P. aeruginosa PAK by using primers Ripel01 and Ripel03 (for pel–lacZ) and rsmZ-rep 5′ and rsmZ-rep 3′ (for rsmZ–lacZ). The pel fragment was cloned into the low copy vector, pMP220, yielding the transcriptional fusion pelA–lacZ. The rsmZ promoter fragment was cloned into pMiniCTX-lacZ and integrated into the P. aeruginosa att site as described in ref. 24.

Plasmids were introduced in P. aeruginosa by electroporation or by conjugation. The recombinant bacteria were selected on Pseudomonas isolation agar. Antibiotics were used at the following concentrations for Escherichia coli: 50 μg/ml ampicillin, 50 μg/ml streptomycin, and 15 μg/ml tetracycline. For P. aeruginosa, 500 μg/ml carbenicillin, 50 μg/ml gentamycin, 200 μg/ml tetracycline, and 2,000 μg/ml streptomycin were used. Bacteria were grown in LB or M63 minimal medium supplemented with 0.2% glucose, 1 mM MgSO4, and 0.5% casamino acids.

Adherence Assay and Pellicle Formation. The P. aeruginosa adherence assay was performed in 24-well polystyrene microtiter dishes as described in ref. 20 or by inoculating individual polystyrene tubes containing 1 ml of medium. Phenotypes are visualized after 6–12 h of incubation at 30°C. Attached bacteria were stained with 1% crystal violet.

Biofilm Formation in the Flow-Cell System. Cultivation of P. aeruginosa biofilms, microscopy, and image acquisition were performed as described in ref. 25. The strains were fluorescently tagged at an intergenic neutral chromosomal locus with gfp in miniTn7 constructs (25). Biofilms were grown at 30°C in flow chambers with individual channel. The images were recorded after 7 days of growth by using a confocal laser-scanning microscope (Zeiss LSM 510).

Measurements of β-Galactosidase Activity. The fusion-carrying strains were grown in minimal medium under agitation at 30°C. Cells (OD600 = 3.5) were collected by centrifugation, and the β-galactosidase activity was measured by using the method of Miller as described in ref. 26. The rsmZ–lacZ fusion experiments were carried out in LB under agitation at 37°C, and cultures were harvested at OD600 of 5.

CHO Cell Cytotoxicity Assay. Exponentially growing bacteria were added to near-confluent CHO cells at a multiplicity of bacteria to CHO cells of 10:1 as described in ref. 11. After a 4-h incubation, the extent of killing of CHO cells was determined by quantifying the release of lactate dehydrogenase (LDH) into culture supernatants by using the LDH cytotoxicity detection kit (Roche Applied Sciences). We selected conditions in which the wild-type (WT) strain causes ≈30% release of the total LDH as determined by treatment of CHO cells with Triton X-100.

Microarray Analysis. P. aeruginosa PAK and its isogenic ladS derivative were grown in minimal medium under agitation at 30°C. Under these conditions the growth kinetics of both of these strains are nearly identical. RNA was isolated, converted to cDNA, and labeled as described in ref. 22. Target hybridization to GeneChip P. aeruginosa Genome Arrays (Affymetrix), washing, and scanning were performed according to Affymetrix protocols. Intrachip normalizations were performed by using Affymetrix microarray suite 5.0 software. The data were subsequently analyzed, and experimental comparisons were made by using genespring (Version 4.2.1). Microarray analysis was performed in duplicate from independent cultures. Three criteria (>2-fold change, Student's t test P value of ≤0.05, and Affymetrix present call in at least one sample) were used to determine significant changes.


Identification of a Sensor Kinase That Regulates Biofilm Formation. In P. aeruginosa, initiation of biofilm formation is in part mediated by type IV pili. Bacteria lacking these appendages show a delay, but not a complete block, in biofilm initiation and maturation. We screened a Tn5 transposon mutant library generated in a nonpiliated P. aeruginosa PAKΔpilA strain for variants in biofilm formation by using an adherence assay (20, 27). Three insertion mutants unable to form a biofilm (PAIH21, PAIC41, and PAIJ68) (Fig. 1A) contained the transposon in the same gene, which mapped to PA3974 in the P. aeruginosa PAO1 genome (28). These strains showed no growth defect in liquid, suggesting that the bacteria were remaining planktonic in the conditions tested. The PA3974 ORF is predicted to encode a two-component sensor histidine kinase, which we have called ladS for “lost adherence sensor.” The ladS gene was required for biofilm formation in a P. aeruginosa PAK-expressing pili, because deletion of the gene in this strain, yielding PAKΔladS, resulted in a biofilm-deficient phenotype (Fig. 1B). Complementation of a ladS mutant with a plasmid containing the ladS gene (pBBRladS) suggests that this phenotype was not due to polar effects (data not shown).

Fig. 1.
Characterization of P. aeruginosa ladS mutants. (A) Microtiter plate assay for biofilm formation. P. aeruginosa PAK, PAKΔpilA nonpiliated strain, and PAKΔpilA ladS::Tn5 mutant derivatives PAIH21, PAIC41, and PAIJ68 are shown. (B) Plastic ...

As these mutants were isolated in a screen for the early steps of biofilm formation, we investigated whether LadS also plays a role in biofilm maturation. A gene coding for GFP was introduced into the WT and ladS deletion mutant of P. aeruginosa PAK, and biofilm formation was followed over several days in a flow cell system (25). Confocal images showed that WT PAK developed a typical mushroom-shaped structure (Fig. 1C). In contrast, the ladS mutant formed a poorly structured, flat biofilm, even after 7 days of growth. These observations suggest that LadS controls factors that contribute to early and late stages of biofilm development; however, it may also play a role in biofilm maturation.

LadS Regulates the pel Exopolysaccharide Operon. Overexpression of the ladS gene not only restored the ability of the bacteria (PAK/pBBRladS) to grow as a biofilm but also caused the formation of a thick bacterial aggregate at the air–media interface (Fig. 1B). Previous studies have associated pellicle formation with the pel exopolysaccharide biosynthesis operon (7, 10). When the ladS gene is overexpressed in a strain deleted for the pelF gene (PAKΔpelF/pBBRladS), no pellicle is formed (data not shown). Because deletion of pel genes suppressed the pellicle formation characteristic of PAK/pBBRladS, it appears that the ladS-induced pellicle is pel-dependent. A lacZ transcriptional fusion to pelA indicated that pel expression was reduced 4-fold in the P. aeruginosa ladS mutant (PAKΔladS) as compared with PAK (Fig. 2A). These observations are consistent with the hypothesis that LadS is a positive regulator of the pel operon.

Fig. 2.
Expression of pel and exoS genes in the ladS mutant. (A) Expression of the pelA-lacZ fusion in either PAK or PAKΔladS strains. (B) Expression of the exoS-lacZ fusion in either PAK or PAKΔladS strains. Filled bars correspond to the activity ...

Analysis of the LadS amino acid sequence and domain organization shows that it contains an N-terminal 7TMR-DISMED2 (7-transmembrane-receptor with diverse intracellular signaling modules extracellular domain 2), followed by a 7 TMR-DISM_7TM domain (seven transmembrane segments found adjacent to 7TMR-DISM domains) (29), which is followed by cytoplasmic, C-terminal histidine kinase and response regulator receiver domains (Fig. 3). Interestingly, this domain architecture is shared by RetS, a P. aeruginosa global regulator that negatively governs expression of the pel operon. The hybrid sensor RetS is larger than LadS, 946 and 795 aa, respectively, essentially because RetS possesses an additional response regulator domain at its C terminus (Fig. 3). Neither gene is linked to a typical response regulator, suggesting that they may function with one of the proteins encoded by the “orphan” response regulators in the P. aeruginosa genome (30, 31) or that the signals are integrated with those transmitted by another two-component system. The opposite biofilm phenotype observed between the retS and ladS mutants are likely the consequence of an antagonistic influence of the corresponding sensors on the expression of the pel genes, possibly in response to different input signals sensed by their unique 7TMR-DISMED2 domains.

Fig. 3.
Domain organization of LadS and RetS hybrid sensors. The transmitter domains are represented in black. This region contains the histidine kinase domain (HisKA) and the kinase domain (HATPase_C). The response regulator receiver domains (Response_reg) have ...

We further investigated the relationship between LadS and RetS by constructing a strain carrying deletions of both ladS and retS genes. As shown in Fig. 7A, which is published as supporting information on the PNAS web site, the biofilm phenotype of the double mutant was similar to that of the retS mutant. Therefore, LadS very likely acts upstream of RetS in the signal transduction pathway, which controls hyperbiofilm (pellicle) formation.

LadS Represses TTSS Gene Expression. In addition to repressing the expression of pel genes, the RetS sensor positively controls the expression of the TTSS regulon. To determine whether the shared regulatory targets of LadS and RetS extend beyond the pel operon, we analyzed the expression of a transcriptional lacZ fusion to exoS, a TTSS effector gene coordinately regulated with the rest of the TTSS regulon (22). We observed a significant increase (2.5-fold) in the activity of the exoS promoter in a ladS mutant as compared with the PAK parental strain (Fig. 2B). Upon growth of P. aeruginosa strains in TTSS-inducing conditions (i.e., in a medium depleted for calcium), a much larger amount of ExoS protein is found in the supernatant fraction of the strain lacking ladS (PAKΔladS) as determined by Western blotting with anti-ExoS antibody (data not shown). Because ExoS is not only overproduced, but also hypersecreted in the ladS mutant, this result suggests that the LadS sensor negatively controls the expression of the exoS gene and likely the entire TTSS regulon. In summary, whereas RetS has been shown to repress pel gene expression and biofilm formation while activating the TTSS, LadS activates pel gene expression and biofilm formation and represses TTSS expression.

We have compared the TTSS-dependent cytotoxicity of ladS, retS, and ladS/retS mutants by measuring the release of lactate dehydrogenase from infected CHO cells (Fig. 7A). As reported previously, the retS mutant was significantly impaired in its cytotoxic activity. In contrast, the ladS mutant was hypercyto-toxic as compared with infection with the WT strain. These results provide additional evidence that RetS and LadS act reciprocally in regulating contact-dependent TTSS activation. Moreover, we observed that the ladS/retS double mutant was noncytotoxic; as for biofilm formation, the position of LadS in the regulatory hierarchy leading to TTSS expression is upstream of RetS.

LadS and RetS Exert Opposite Effects on the Small Regulatory RNA RsmZ. A genetic screen for transposon insertions capable of suppressing both the biofilm and TTSS phenotypes of a retS deletion strain implicated the GacS/GacA two-component system and its downstream target, the small regulatory RNA RsmZ in the RetS signaling network (11). RsmZ is an antagonist of the mRNA-binding protein RsmA, which has been shown to shuttle between an inactive, RsmZ-bound state, and an active state in which it binds target mRNAs and regulates their translation (18). The phenotypes of the retS mutant were suppressed by second site mutations in the sensor kinase gacS, its unlinked but cognate response regulator gacA, and rsmZ. The consequence of each of these suppressor insertions is to eliminate the production of the small RNA rsmZ. These data are consistent with the hypothesis that in strains lacking retS, rsmZ levels are elevated. To assess this hypothesis and determine whether LadS is also a component of this signaling network, the rsmZ promoter was fused to lacZ and examined for activity in retS, ladS, gacS, and rsmA mutant backgrounds (Fig. 4). In agreement with published studies, expression of rsmZ required GacS and RsmA (17). Deletion of retS resulted in a substantial increase in rsmZ expression. In contrast, deletion of ladS abolished the expression of the rsmZ–lacZ fusion to the same extent as deletion of gacS, the gene encoding the known regulator of rsmZ. The activity of the rsmZ–lacZ fusion in a ladS/retS double mutant was also high, although somewhat lower than that measured in the retS background. These data provide strong evidence that the LadS and RetS signal transduction pathways are truly antagonistic, and their reciprocal effects on biofilm formation and TTSS expression are accomplished, at least in part, through their opposite effect on the expression of rsmZ.

Fig. 4.
The activity of the rsmZ-lacZ fusion in various P. aeruginosa mutant backgrounds. Levels of β-galactosidase were measured in a culture of P. aeruginosa PAK, and the isogenic ladS, retS, gacS, and rsmA mutants. On the y axis, a break in scale has ...

Transcriptional Profiling. To establish the extent of this opposite regulatory effect, we used P. aeruginosa microarrays (Affymetrix) to perform a transcriptome analysis of PAKΔladS compared with its WT parent. These data were then compared with a previously published RetS transcriptome analysis (11). In agreement with the reporter fusions, pel genes were repressed, and TTSS genes were activated in the ladS mutant (see Table 2, which is published as supporting information on the PNAS web site). Significantly changing (see Materials and Methods) genes are shown in Fig. 5. Globally, 79 genes were significantly affected by deletion of ladS. Of these, 1% were coregulated in the retS mutant, whereas 49% were oppositely regulated. Although only 27% of the RetS-dependent genes changed >2-fold in the LadS microarray data, 91% of those that did change were reciprocally regulated. These findings show that although a number of genes are regulated independently by LadS and RetS, a substantial fraction are reciprocally coordinated by these two sensors. Interestingly, several genes that were clearly reciprocally regulated by LadS and RetS (Figs. (Figs.22 and and44 and Table 2) did not show significant variation in the microarray analysis. This result is not entirely unexpected, given the involvement of RsmA/RsmZ in posttranscriptional control. Therefore, the effects seen could be due to controls of LadS/RetS at the level of transcription as well as mRNA stability. It is also likely that the growth conditions used may not fully activate these signal transduction pathways, reducing the potential for differential expression between WT and mutant strains.

Fig. 5.
Comparison between LadS-dependent and RetS-dependent gene regulation. The selected genes presented at least a 2-fold variation in the ladS mutant. (A) Genes that are up-regulated in a ladS mutant. (B) Genes that are down-regulated in a ladS mutant. The ...


The data presented here describe a regulatory network responsible for the control of central aspects of P. aeruginosa virulence: Type III secretion and biofilm formation. The observation that mutants in ladS, a gene encoding a hybrid sensor kinase, remained planktonic in conditions in which the parental strain transitions to biofilm growth suggests that this gene may mediate a switch between these bacterial lifestyles. Further analysis showed that LadS shares domain organization and downstream targets with RetS, a recently described regulator of biofilms and other virulence traits of P. aeruginosa (11). On the basis of the data presented here and of previously published work (11, 17), we propose an integrative model of this signaling network (Fig. 6). LadS and RetS contain 7TMR-DISMED2 periplasmic sensor modules, which show a modest 35% sequence identity, suggesting that they bind related but nonidentical ligands. This domain has been predicted to bind carbohydrates and is always associated with a transmembrane domain (7TM-DISM) (29). Activation of LadS or RetS, however, has reciprocal effects on a shared set of positively and negatively regulated virulence determinants. We show here that in a ladS mutant rsmZ transcription is repressed, resulting in up-regulation of TTSS genes and concomitant repression of transcription of the exopolysaccharide genes within the pel operon. A mutant with a deletion of retS behaves in an opposite manner, resulting in dramatic increase in the transcription of rsmZ. This mutant becomes insensitive to cues that normally trigger activation of the TTSS and derepresses multiple operons encoding the exopolysaccharide biosynthetic enzymes necessary for biofilm formation. These experiments suggest that the small RNA RsmZ at least partially mediates this global switch in adaptive behavior. The model presented in Fig. 6 suggests that both RetS and LadS signal through the GacS/GacA/RsmZ pathway. Hybrid sensors often mediate signal transduction through a histidine phosphotransfer (Hpt) domain or protein. It is possible that RetS and LadS control expression of rsmZ and other genes by regulating the phosphorylation state of an as-yet-unidentified Hpt, which in turn controls a downstream response regulator.

Fig. 6.
A model for the convergence of the signaling pathways during reciprocal regulation of virulence factors by LadS, RetS, and GacS through transcription of the small regulatory RNA RsmZ. The three sensors are anchored into the cytoplasmic membrane via their ...

Gene activation and repression by a single response regulator is not uncommon. For example, cooperative occupancy of multiple regulatory sites by different forms of the response regulator OmpR has been shown to control activation and repression of ompF and ompC genes in E. coli (32). Our discovery that some or perhaps most of the regulatory activities of LadS and RetS occur through the transcriptional control of a small regulatory RNA is reminiscent of the quorum-sensing circuitry in several Vibrio species (33). In Vibrio harveyi, the two-component sensor kinases LuxN and LuxQ each respond to a different quorum-sensing signaling molecule. At low cell density, LuxN and LuxQ phosphorylate the response regulator LuxO via the LuxU phosphorelay. LuxO consequently activates the expression of the small quorum-sensing (Qrr) RNAs 1–4. The Qrr RNAs anneal to the 5′ end of the mRNA encoding the transcriptional regulator LuxR and promote degradation of the luxR message. LuxR acts to activate transcription at some promoters and repress at others (including the TTSS), and small RNA-dependent degradation of the luxR mRNA affects the expression of a wide range of downstream genes. When the cell (and autoinducer) density increases, the sensors act as phosphatases, and consequently the luxR message is no longer degraded. Although the Vibrio quorum-sensing network and the P. aeruginosa LadS/RetS pathway share certain common features, there are significant differences in their mechanisms of signal assimilation and transmission. In the Vibrio systems, two sensor kinases appear to control small RNA transcription in a coordinate fashion, both acting as kinases or phosphatases depending on the presence of autoinducer molecules. Whereas LadS and GacS appear to coordinately positively regulate transcription of the small RNA RsmZ, the third sensor kinase RetS acts in an opposite manner. Such a system allows for adaptive transitions to be triggered by both the disappearance of a signal maintaining the current state and the appearance of a signal prompting the future state.

We have speculated that the RetS-controlled regulatory network functions in response to signals present in tissues during the acute phase of bacterial infection, maintaining the bacterium in a planktonic state competent for TTSS activation. Work presented here suggests that the transition to biofilm formation, TTSS repression, and chronic infection may require both the disappearance of the RetS signal and also the appearance of signals sensed by LadS and/or GacS. Interestingly, unlike the majority of two-component systems, genes coding for LadS and RetS are not linked to an adjacent response regulator. The gacS gene is also distant from gacA (31), and the conclusion that GacS and GacA form a cognate pair comes from the shared phenotypes of gacS and gacA mutants and homology to a similar system in E. coli. It is therefore conceivable that a significant level of crosstalk occurs to assimilate various signals which regulate expression of RsmZ. Our data support a model in which RetS, GacS, and LadS are key bacterial sensors that constantly sample the environmental conditions. In response to these inputs, P. aeruginosa can optimize expression of genes that favor acute interactions with the host or persistence as biofilm communities. The precise nature of the input signals for the RetS/LadS pathway is unknown at this time, as is the signal received by GacS. The 7TMR-DISMED2 domains found in RetS and LadS have been identified in a variety of carbohydrate binding proteins (29), suggesting that both hybrid sensors sense carbohydrates present in host tissue, which can serve to direct the course of infection. Alternatively, the carbohydrates sensed by RetS and LadS may be produced by P. aeruginosa and therefore can represent a novel form of quorum sensing. Identification of the chemical nature of the signal recognized by RetS and LadS would greatly enhance our understanding of these types of regulatory networks among various bacterial pathogens. Equally important is the elucidation of the intermediate steps in this signal transduction network, which allow for a global transcriptional response through a small RNA molecule.

Supplementary Material

Supporting Information:


We thank the P. aeruginosa Sequencing Consortium for providing updated annotations at www.pseudomonas.com. We thank Vincent Lee for help with the cytotoxicity assays. We thank Cystic Fibrosis Foundation Therapeutics Inc. for subsidizing the GeneChip P. aeruginosa genome arrays. I.V. was supported by a grant from the French cystic fibrosis foundation (VLM). Research in the A.F. laboratory was supported by grants from the VLM, European Union Grant QLK3CT-2002-02086, and the Bettencourt-Schueller foundation. A.L.G. is a Howard Hughes Medical Institute Predoctoral Fellow. This work also was supported by National Institutes of Health Grant R37-AI021451 (to S.L.).


Author contributions: S.L. and A.F. designed research; I.V., A.L.G., I.V.-G., P.V., and C.S. performed research; I.V., A.L.G., S.M., S.B., A.L., S.L., and A.F. analyzed data; and I.V., A.L.G., S.L., and A.F. wrote the paper.

Conflict of interest statement: No conflicts declared.

This paper was submitted directly (Track II) to the PNAS office.

Abbreviation: TTSS, type III secretion system.


1. Nickel, J. C., Downey, J. A. & Costerton, J. W. (1989) Urology 34, 284–291. [PubMed]
2. Fletcher, E. L., Weissman, B. A., Efron, N., Fleiszig, S. M., Curcio, A. J. & Brennan, N. A. (1993) Curr. Eye Res. 12, 1067–1071. [PubMed]
3. Govan, J. R. & Deretic, V. (1996) Microbiol. Rev. 60, 539–574. [PMC free article] [PubMed]
4. O'Toole, G., Kaplan, H. B. & Kolter, R. (2000) Annu. Rev. Microbiol. 54, 49–79. [PubMed]
5. Mah, T. F., Pitts, B., Pellock, B., Walker, G. C., Stewart, P. S. & O'Toole, G. A. (2003) Nature 426, 306–310. [PubMed]
6. Matsukawa, M. & Greenberg, E. P. (2004) J. Bacteriol. 186, 4449–4456. [PMC free article] [PubMed]
7. Friedman, L. & Kolter, R. (2004) Mol. Microbiol. 51, 675–690. [PubMed]
8. Friedman, L. & Kolter, R. (2004) J. Bacteriol. 186, 4457–4465. [PMC free article] [PubMed]
9. Jackson, K. D., Starkey, M., Kremer, S., Parsek, M. R. & Wozniak, D. J. (2004) J. Bacteriol. 186, 4466–4475. [PMC free article] [PubMed]
10. Vasseur, P., Vallet-Gely, I., Soscia, C., Genin, S. & Filloux, A. (2005) Microbiology 151, 985–997. [PubMed]
11. Goodman, A. L., Kulasekara, B., Rietsch, A., Boyd, D., Smith, R. S. & Lory, S. (2004) Dev. Cell 7, 745–754. [PubMed]
12. Laskowski, M. A., Osborn, E. & Kazmierczak, B. I. (2004) Mol. Microbiol. 54, 1090–1103. [PMC free article] [PubMed]
13. Hauser, A. R., Cobb, E., Bodi, M., Mariscal, D., Valles, J., Engel, J. N. & Rello, J. (2002) Crit. Care Med. 30, 521–528. [PubMed]
14. Roy-Burman, A., Savel, R. H., Racine, S., Swanson, B. L., Revadigar, N. S., Fujimoto, J., Sawa, T., Frank, D. W. & Wiener-Kronish, J. P. (2001) J. Infect. Dis. 183, 1767–1774. [PubMed]
15. Costerton, J. W. (2001) Trends Microbiol. 9, 50–52. [PubMed]
16. Jain, M., Ramirez, D., Seshadri, R., Cullina, J. F., Powers, C. A., Schulert, G. S., Bar-Meir, M., Sullivan, C. L., McColley, S. A. & Hauser, A. R. (2004) J. Clin. Microbiol. 42, 5229–5237. [PMC free article] [PubMed]
17. Heeb, S., Blumer, C. & Haas, D. (2002) J. Bacteriol. 184, 1046–1056. [PMC free article] [PubMed]
18. Heurlier, K., Williams, F., Heeb, S., Dormond, C., Pessi, G., Singer, D., Camara, M., Williams, P. & Haas, D. (2004) J. Bacteriol. 186, 2936–2945. [PMC free article] [PubMed]
19. Kagami, Y., Ratliff, M., Surber, M., Martinez, A. & Nunn, D. N. (1998) Mol. Microbiol. 27, 221–233. [PubMed]
20. Vallet, I., Olson, J. W., Lory, S., Lazdunski, A. & Filloux, A. (2001) Proc. Natl. Acad. Sci. USA 98, 6911–6916. [PMC free article] [PubMed]
21. Kaniga, K., Delor, I. & Cornelis, G. R. (1991) Gene 109, 137–141. [PubMed]
22. Wolfgang, M. C., Lee, V. T., Gilmore, M. E. & Lory, S. (2003) Dev. Cell 4, 253–263. [PubMed]
23. Bleves, S., Soscia, C., Nogueira-Orlandi, P., Lazdunski, A. & Filloux, A. (2005) J. Bacteriol. 187, 3898–3902. [PMC free article] [PubMed]
24. Hoang, T. T., Kutchma, A. J., Becher, A. & Schweizer, H. P. (2000) Plasmid 43, 59–72. [PubMed]
25. Klausen, M., Heydorn, A., Ragas, P., Lambertsen, L., Aaes-Jorgensen, A., Molin, S. & Tolker-Nielsen, T. (2003) Mol. Microbiol. 48, 1511–1524. [PubMed]
26. Sambrook, J., Fritsch, E. F. & Maniatis, T., eds. (1989) Molecular Cloning (Cold Spring Harbor Lab. Press, Woodbury, NY).
27. O'Toole, G. A. & Kolter, R. (1998) Mol. Microbiol. 30, 295–304. [PubMed]
28. Stover, C. K., Pham, X. Q., Erwin, A. L., Mizoguchi, S. D., Warrener, P., Hickey, M. J., Brinkman, F. S., Hufnagle, W. O., Kowalik, D. J., Lagrou, M., et al. (2000) Nature 406, 959–964. [PubMed]
29. Anantharaman, V. & Aravind, L. (2003) BMC Genomics 4, 34. [PMC free article] [PubMed]
30. Rodrigue, A., Quentin, Y., Lazdunski, A., Mejean, V. & Foglino, M. (2000) Trends Microbiol. 8, 498–504. [PubMed]
31. Ventre, I., Filloux, A. & Lazdunski, A. (2004) in Virulence Gene and Regulation, The Pseudomonads, ed. Ramos, J. L. (Kluwer, New York), Vol. 2, pp. 257–288.
32. Mattison, K., Oropeza, R., Byers, N. & Kenney, L. J. (2002) J. Mol. Biol. 315, 497–511. [PubMed]
33. Lenz, D. H., Mok, K. C., Lilley, B. N., Kulkarni, R. V., Wingreen, N. S. & Bassler, B. L. (2004) Cell 118, 69–82. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...