PqsE has a conserved sequence, yet a variable impact in Pseudomonas aeruginosa

ABSTRACT

Pseudomonas aeruginosa is an opportunistic pathogen responsible for several acute and chronic infections. The production of many of its virulence factors is tightly regulated by three interlinked quorum sensing (QS) systems named las, rhl, and pqs. The pqs system relies on 4-hydroxy-2-alkylquinolines (HAQs) as signaling molecules to activate the transcriptional regulator MvfR (PqsR), which drives HAQ biosynthesis via the pqsABCDE operon. The final gene in this operon encodes PqsE, a multifunctional protein unique to P. aeruginosa. Beyond its thioesterase activity in HAQ biosynthesis, PqsE stabilizes RhlR, the transcriptional regulator of the rhl system, facilitating the regulation of virulence-related genes. Due to its pathogenic relevance, PqsE is considered a potential therapeutic target against P. aeruginosa infections. While the role of PqsE toward the RhlR regulon is increasingly understood in reference P. aeruginosa strains such as PA14 and PAO1, its broader relevance remains underexplored. In this study, pqsE and rhlR were found to be genetically conserved across a diverse panel of 12 P. aeruginosa strains. Phenotypic assays and metabolite quantification revealed that PqsE broadly influences virulence factors and multicellular behaviors, including pyocyanin production and biofilm formation. While the magnitude of PqsE-dependent phenotypes varied between strains, key functions were consistently maintained, underscoring both conservation and strain-specific modulation. Notably, PqsE proved essential for HAQ biosynthesis in most strains, challenging prior assumptions of its dispensability. This study contributes to a deeper understanding of QS regulation, highlighting that while PqsE contributes to conserved functions across P. aeruginosa strains, its impact is strain dependent.

IMPORTANCE

Pseudomonas aeruginosa is a versatile opportunistic pathogen, naturally tolerant and readily acquiring resistance to multiple antibiotics. Consequently, the World Health Organization identified this bacterium as a high-priority pathogen for researching and developing new antimicrobial strategies. P. aeruginosa utilizes quorum sensing, a cell-to-cell communication system, to regulate the expression of several of its virulence factors. Here, we confirm that the PqsE protein is conserved, and its function in quorum sensing, especially toward the RhlR regulator, is variable across a panel of 12 P. aeruginosa strains. Since PqsE is conserved and unique to this bacterium, it has been proposed as an ideal target for antivirulence therapies, offering new alternatives to combat antimicrobial resistance. However, our results question the relevance of PqsE as an appropriate target.

INTRODUCTION

The bacterium Pseudomonas aeruginosa is an opportunistic pathogen closely associated with human activity and is a leading cause of infections, both acute and chronic, in immunocompromised individuals (1, 2). Its capacity to adapt to various environments and produce virulence factors is mainly regulated by a process named quorum sensing (QS), a cell-to-cell communication system that modulates gene expression through the production and detection of small autoinducer molecules in response to population density (3, 4).

P. aeruginosa has three interdependent QS systems: las, rhl, and pqs. The las system relies on the LasI synthase, which produces the signal molecule 3-oxo-dodecanoyl-homoserine lactone (3-oxo-C12-HSL). This molecule binds and activates LasR, its cognate LuxR-type transcriptional regulator, which in turn promotes the expression of numerous genes, including lasI, thereby sustaining autoinducer production through a positive feedback loop (5, 6). In prototypical strains, LasR activates the rhl system by initiating the transcription of rhlI and rhlR (7, 8). The RhlI synthase produces the butanoyl-homoserine lactone (C4-HSL) autoinducer, while RhlR functions as a transcriptional regulator. Upon binding to C4-HSL, RhlR induces the expression of genes responsible for the production of virulence factors, including those involved in pyocyanin synthesis (two phzABCDEFG operons) and the production of rhamnolipids (rhlAB and rhlC) (6, 8–11). While the las system is generally considered atop the QS hierarchy in prototypical strains, loss of LasR activity is frequent in strains isolated from both clinical and environmental settings (12–15). Nevertheless, some LasR-defective strains still retain RhlR function and fully express virulence factors, suggesting that RhlR could assume a central role in a malleable QS hierarchy (13–15). The third QS system in P. aeruginosa, the pqs system, is driven by signaling molecules called 4-hydroxy-2-alkylquinolines (HAQs). The transcriptional regulator of this system, MvfR (also known as PqsR), is activated upon binding either 4-hydroxy-2-heptylquinoline (HHQ) or 3,4-dihydroxy-2-heptylquinoline (Pseudomonas quinolone signal [PQS]) (16). Once bound to one of its autoinducing ligands, MvfR regulates the transcription of the pqsABCDE operon, which encodes the enzymes responsible for HAQ biosynthesis (17–20). Furthermore, the pqs system is thoroughly regulated by the two other QS systems, as LasR positively regulates the transcription of mvfR, whereas RhlR negatively regulates the expression of pqsABCDE (20–22).

The final gene in the pqsABCDE operon, pqsE, encodes the multifunctional protein PqsE. It was characterized as a thioesterase implicated in HAQ biosynthesis (23). However, PqsE seems dispensable for this process, as pqsE mutants show no defects in HAQ production in reference strains, presumably because redundant enzymes can take over its function (18, 19, 23). In addition to its enzymatic activity, PqsE modulates the transcription of multiple target genes, most of which belong to the RhlR regulon, including the phzABCDEFG operons, and to some extent, the rhlAB operon (19, 24–29). Indeed, PqsE has a minimal impact on the P. aeruginosa transcriptome in the absence of RhlR (24). Recent studies have revealed that PqsE interacts directly with RhlR through a protein-protein interaction, enhancing RhlR stability and increasing its affinity for target promoters (26, 30–32). The chaperone-like activity of PqsE is particularly noteworthy, as it is unique to P. aeruginosa (30). Notably, the PqsE-RhlR complex functions independently of the catalytic thioesterase activity of PqsE (30, 31, 33). Furthermore, an implication for PqsE in biofilm formation and virulence has also been reported in strains PA14 and PAO1 (25, 28, 34, 35).

Due to its specificity and role in virulence, PqsE is often considered an attractive target for the development of antivirulence therapies (26, 30, 31, 36, 37). However, although the functional role of PqsE toward the RhlR regulon and HAQ production has been extensively studied in well-known P. aeruginosa strains PA14 and PAO1, its role in other isolates has not been explored. To establish its potential as a target for antivirulence strategies, it is essential to first confirm the genetic and functional conservation of PqsE across diverse strains.

In this study, we conducted a genetic and functional analysis of PqsE in diverse isolates to gain knowledge on its ecological role. We focused on virulence determinants reported as being regulated by PqsE in reference strains. Our results show that the pqsE gene is conserved, but the functional role of PqsE generally varies across isolates, although its importance in regulating some virulence determinants and social behaviors is widespread. Importantly, our results reveal that the enzymatic activity of PqsE is consistently required for HAQ biosynthesis across strains, unlike in the PAO1 and PA14 reference strains, where thioesterase activity appears complemented by alternative enzymes.

RESULTS AND DISCUSSION

The pqsE and rhlR genes are conserved in different P. aeruginosa backgrounds

To be considered a key target for developing antivirulence strategies, PqsE would need to be genetically and functionally conserved. Thus, a panel of 12 P. aeruginosa strains, including PA14 and PAO1, was selected to study the functionality and conservation of PqsE (Table 1; Table S1). These strains were chosen to capture ecological diversity and were obtained from various clinical contexts and isolation sites. Some of them also display diverse QS functional patterns. Notably, our panel includes LasR-defective strains, some of which possess an independently active RhlR. In certain cases, strains maintain full functionality of the rhl system, despite a defective LasR, which is typically considered to be the master regulator of the QS hierarchy. We refer to those strains as LasR defective and RhlR active (13–15, 38–40). Additionally, our panel also includes one strain unable to produce HAQs, as determined previously (13, 14). Notably, whole-genome sequences are available for all strains (41).

TABLE 1.

Characteristics of the P. aeruginosa isolates used in this study

Strains	Origin	Pathology or isolation site	LasR classificationa	Production of HAQsa	Source
39016	United Kingdom	Keratitis	Functional LasR	Yes	(41)
A22	France	Wound	Functional LasR	No	(42)
E90	USA	Cystic fibrosis	LasR defective, RhlR active (RAIL)	Yes	(15)
JJ692	USA	Urinary tract infection	LasR defective	Yes	(43)
PA14	USA	Burn wound	Functional LasR	Yes	(44)
PAO1	Australia	Wound infection	Functional LasR	Yes	Nottingham collection
PA-CL512	Canada	Hospital sink	LasR defective, RhlR active (RAIL)	Yes	(45)
PA-CL513	Canada	Hospital sink	Functional LasR	Yes	(45)
PA-CL521b	Canada	Hospital sink	LasR defective, RhlR active (RAIL)	Yes	(45)
PA-W9	United Kingdom	Leg ulcer	Functional LasR	Yes	(41)
PG201 (Rsan-ver)	Switzerland	Soil	LasR defective	Yes	(46)
SMC1596	Canada	Cystic fibrosis	Functional LasR	Yes	(47)

^a Classification of LasR function and HAQ production is based on Groleau et al. (13) and Trottier et al. (14). For the PAO1 strain, this classification is based on Pearson et al. (48) and Pesci et al. (49).

To gain deeper insight into the evolutionary dynamics of this panel of strains, we conducted a phylogenetic analysis based on the conservation of the core genome. The generated distance tree provides information about the core genome diversity of the analyzed P. aeruginosa strains (Fig. 1). Many strains exhibit a high core genome similarity, suggesting limited divergence in conserved functional pathways. However, strains located on distinct branches of the tree (e.g., the PA14 group vs the PAO1 group) may possess slight differences in their set of genes, explained by evolutionary divergences.

Fig 1.

Core genome phylogeny and average nucleotide identity (ANI) heatmap of selected P. aeruginosa strains. (A) Core genome phylogenetic tree of P. aeruginosa strains from this study (red), along with reference strains (blue). Node labels indicate statistical support (in %) among 1,000 bootstrap replicates. Core genes included in this phylogeny were inferred from a core gene alignment generated by PyMLST v.2.1.6 using the cgMLST.org public scheme for P. aeruginosa, which includes 3,687 core loci from 6,730 genomes. The GTR + F + I + G4 model was selected according to the Bayesian Information Criterion by IQ-TREE v.2.3.6 as the best-fit model. Branch lengths indicate the number of substitutions per site. (B) ANI matrix aligned with the phylogenomic tree in panel A. ANI calculations were done with pyani v.2.1.6 using the “ANIm” all-versus-all comparison mode, with default parameters.

Within the main branch of the P. aeruginosa species, the strains form two distinct clusters: one that includes 10 strains along with PAO1 and another where 2 strains cluster with PA14, thus confirming that all strains used in this study accurately belong to the P. aeruginosa species. Our collection reflects the general distribution of phylogroups, as the PAO1-containing clade is more prevalent than the two other clades (41). None of our strains cluster with the PA7 outlier group, recently reclassified as the Pseudomonas paraeruginosa species (50, 51). Although the interaction between PqsE and RhlR seems to be conserved in strains belonging to this newly described species, we chose to exclude them from our analysis to focus specifically on strains belonging to the P. aeruginosa species (52, 53). These findings further support that there are some variations within the core genomes of our strains, which could further explain phenotypic differences among them. These potential differences are also reflected in the broad average nucleotide identity (ANI) range both within and between each P. aeruginosa core phylogenomic cluster (99.1% to 99.7%) (Fig. 1B).

Since previous studies have shown that pqsE is highly conserved across diverse P. aeruginosa isolates (54, 55), we aimed to confirm whether conserved mutations or polymorphisms were present. To do so, we analyzed the genomic sequence of pqsE and the predicted amino acid sequence of the PqsE protein across our panel of P. aeruginosa strains. Our genomic analysis confirms that pqsE is constitutively present and highly conserved, revealing a 99.5% (902 out of 906 nucleotides) identity across strains. Three nucleotide positions with polymorphisms among the strains and one unique mutation were identified (Table S2). Notably, these variations do not affect the predicted amino acid sequence (Fig. S1), further supporting the genetic conservation of pqsE across isolates (Fig. S1; Table S2).

Since PqsE and RhlR function as a pair to regulate the transcription of various genes and variations in rhlR could thus impact PqsE-dependent phenotypes, we also verified the conservation of the rhlR gene. Indeed, this gene has a low mutation rate according to the literature (56, 57). Like pqsE, our analysis revealed only a few polymorphisms in the rhlR sequence (Table S3), which have no impact on the predicted amino acid sequence (Fig. S2).

Overall, our results reveal that the predicted PqsE and RhlR proteins are conserved in our panel of independent strains. Importantly, previously identified codons encoding key amino acids involved in PqsE-RhlR interactions are preserved (Fig. S1 and S2) (30). From a strictly genetic perspective, pqsE remains a promising target for antivirulence therapies due to its high conservation, as shown here and in prior studies, and its specificity to P. aeruginosa (54, 55).

PqsE impacts the production of virulence determinants in P. aeruginosa isolates

While pqsE is genetically conserved, the functional conservation of its corresponding protein, PqsE, remains unclear. We first examined the production of virulence determinants regulated by PqsE in well-studied strains, such as pyocyanin and rhamnolipids. This approach allowed us to determine whether phenotypes observed in a couple of strains could be reproduced in independent isolates. We included PA14 and PAO1 in our panel of strains since their PqsE-dependent phenotypes are well described (19, 25, 27–29). To investigate the role of PqsE, we deleted pqsE in all strains from our panel. Importantly, we confirmed that ∆pqsE mutants exhibited no differences in growth compared to their respective wild-type (WT) counterparts (Fig. S3). We also included the previously characterized ∆pqsE mutants of PA14 and PAO1 (18, 25). The next step was to phenotypically characterize all the generated ∆pqsE mutants and verify if the impact on regulated traits was conserved between isolates.

The most studied PqsE-dependent phenotype is pyocyanin production. Pyocyanin is a redox-active phenazine responsible for the characteristic blue pigmentation of P. aeruginosa cultures. Its biosynthesis involves two redundant operons, phzA1-G1 (phz1) and phzA2-G2 (phz2), which code for enzymes to produce phenazine-1-carboxylic acid (PCA), a precursor that is subsequently converted to pyocyanin by PhzM and PhzS (58). The phz1 operon possesses a lux-box recognized by RhlR and is the predominant phz operon expressed in planktonic cultures of the PA14 strain (59, 60). Regulation of the expression of phz1 in PA14 or PAO1 depends on the functionality of PqsE on RhlR activity (29, 61, 62).

To investigate the role of PqsE in this process across our panel of strains, we quantified PCA, the direct product of the phz1 operon, along with its final metabolite, pyocyanin, using liquid chromatography coupled to tandem mass spectrometry (LC/MS/MS) in WT strains and their isogenic ∆pqsE mutants. In 11 out of the 12 strains, we measured a significant reduction or complete loss of PCA and pyocyanin in the ∆pqsE mutant, at least for one of the two time points (Fig. 2A). This widespread reduction confirms that the role of PqsE in promoting pyocyanin biosynthesis is broadly conserved across diverse P. aeruginosa isolates, consistent with previous findings (29).

Fig 2.

Production of virulence determinants in the ∆pqsE mutants relative to their respective WT strain. (A) Heatmap of the production of PCA and pyocyanin (PYO) in the ∆pqsE mutants relative to their respective WT strain. Concentrations of PCA and PYO were measured by LC/MS/MS at 6 and 24 h. (B) Heatmap of the production of rhamnolipids in the ∆pqsE mutant relative to their respective WT strain. Concentrations of mono-rhamnolipids (Rha-C10-C10) and di-rhamnolipids (Rha-Rha-C10-C10) were measured with LC/MS/MS after 24 h of growth in King’s A medium. For both panels, values are log₂-transformed ratios (log₂ fold change [Log₂FC]) of production in ∆pqsE mutants relative to their WT strain (∆pqsE/WT production). Red shades indicate reduced production of metabolites in the mutant. The intensity of the color reflects the magnitude of the change: darker shades indicate greater differences in the logarithmic scale. In panel A, the color scale ranges from −7 (strong reduction in the mutant) to 7 (strong increase). For visual purposes, the range in panel B is from −2 to 2. To enable log transformation and visual consistency, zero values (absence of production) in the mutant were adjusted using a pseudo-count of 1. Ratios were then capped at the lower end of the scale (minimum log₂ ratio of −7 for panel A and −2 for panel B). Raw data for these measurements are provided in Tables S4 to S6.

While the role of PqsE in regulating pyocyanin production is well established in reference strains, the influence on other virulence factors remains unclear (29, 61, 62). Given that PqsE modulates the activity of RhlR, we chose to look at another RhlR-dependent determinant: rhamnolipid production (11, 63). Rhamnolipids are biosurfactants that play key roles in social motility, biofilm development, and virulence (64). In P. aeruginosa, the two principal rhamnolipid congeners are the mono-rhamnolipid Rha-C10-C10 and the di-rhamnolipid Rha-Rha-C10-C10 (65). Their biosynthesis is primarily driven by the rhlAB operon, which encodes key enzymes involved in the production of these surface-active molecules (63, 66). Given prior evidence suggesting that PqsE plays a role in rhamnolipid production (27, 37, 67) and rhlAB transcription through RhlR (24, 34, 68), the production of the two major rhamnolipid congeners in our panel of pqsE mutants was investigated using LC/MS/MS quantification.

Rhamnolipid production was reduced in only 4 of the 12 P. aeruginosa panel strains (Fig. 2B). Additionally, under our culture conditions, loss of pqsE in both PAO1 and PA14 did not affect rhamnolipid production, suggesting that any effect of PqsE on the RhlR-mediated rhlAB transcription is, at most, very limited. Indeed, previous studies in PA14 have shown only a minimal influence of PqsE on the transcription of the rhlA gene, in contrast with an important impact of C4-HSL (61). The fact that many pqsE mutants do not exhibit reduced rhamnolipid production suggests that the RhlR transcriptional regulator may rely primarily on C4-HSL rather than on PqsE for rhlAB transcription, as previously suggested (52). In addition, it is important to note that the expression of rhlAB is highly dependent on environmental conditions, and various regulatory elements can affect this process (69–71). Therefore, PqsE may be just one of the many factors influencing RhlR-mediated transcription of rhlAB.

Curiously, three of the four strains affected by PqsE for rhamnolipid production have a defective LasR protein, and two have a RhlR that functions independently of LasR (Table 1) (13, 14). In these strains, we hypothesize that the absence of LasR increases the dependence of RhlR on PqsE for the transcription of the rhlAB operon, as RhlR becomes central to the QS hierarchy (29). However, more strains should be studied to verify this hypothesis.

Globally, the role of PqsE in rhamnolipid production varies among strains. While PqsE can influence rhamnolipid production, its effect seems to be strain specific, indicating that rhlAB is not a conserved target for PqsE-mediated RhlR activity. Interestingly, no rhamnolipid production defect is observed in the ∆pqsE mutants of the PA14 and PAO1 reference strains. While previous reports showed limited differences between the WT strain and ∆pqsE mutants, these were largely based on the assumption that RhlR-regulated genes are generally also regulated by PqsE. However, more recent data suggest that this is not the case (24, 61). There is also a slight possibility that rhlAB does not depend on RhlR in some strains, though this possibility is unlikely, as rhlA is considered part of the RhlR core regulon (38). Additionally, while the culture conditions we used are suitable for rhamnolipid production in PA14 and PAO1, we cannot rule out that results could be condition dependent.

Overall, even if data regarding rhamnolipid production are variable, our results regarding pyocyanin production confirm that PqsE is globally relevant for the regulation of the rhl system and the production of virulence determinants, even in LasR-deficient strains (Table 1, Fig. 2) (29). However, the extent of the variation of PqsE-dependent impact on the production of virulence factors complicates its candidacy as a relevant therapeutic target.

PqsE plays a role in multicellular behaviors of P. aeruginosa strains

RhlR is associated with biofilm formation through the production of various factors, including rhamnolipids and lectins (72, 73). Furthermore, RhlR influences colony biofilm formation (74). Again, since PqsE is linked to RhlR activity, the impact of PqsE on biofilm formation was investigated.

Biofilm formation is a complex process influenced by multiple factors. In P. aeruginosa, the biofilm matrix comprises various components, including exopolysaccharides, proteins, and extracellular DNA (72, 75–77). Previous studies based on in vitro assays and transcriptomic analyses, primarily conducted with the PAO1 strain, have shown that PqsE contributes to biofilm formation. Indeed, pqsE mutants typically exhibit reduced biofilm production (25, 68). PqsE also influences the structure of colony biofilms in P. aeruginosa PA14 (34, 36). Specifically, a pqsE mutant exhibits a hyper-rugose phenotype when grown on Congo red agar, primarily due to the loss of phenazine production (34, 78). To further investigate the role of PqsE in this process, we examined biofilm formation in polystyrene plates and on Congo red agar in the strains from our panel and their isogenic ∆pqsE mutants.

Our results show variability among the ∆pqsE mutants, with some producing more biofilms than their respective WT strains and others producing less on polystyrene (Fig. 3). Interestingly, under our experimental conditions, the ∆pqsE mutant in the PAO1 strain demonstrates increased biofilm formation compared to the WT strain, whereas the opposite effect is seen for the PA14 strain. Overall, PqsE typically impacts in vitro biofilm formation in P. aeruginosa, with 11 out of 12 isolates showing a significant increase or decrease in biofilm production when pqsE is disrupted (Fig. 3). However, its role does not appear to be conserved, as it can have drastically different effects.

Fig 3.

Heatmap of the production of biofilms in polystyrene plates in the ∆pqsE mutant relative to their respective P. aeruginosa WT strain. Biofilm formation (OD₅₅₀) was measured after 24 and 48 h of incubation at 37°C (79). Values are log₂-transformed ratios (log₂ fold change [Log₂FC]) of production in ∆pqsE mutants relative to their WT strain (∆pqsE/WT production). Red shades indicate reduced biofilm formation in the mutant, whereas blue indicates increased formation. The intensity of the color reflects the magnitude of the change: darker shades indicate greater differences in the logarithmic scale. For visual purposes, the color scale ranges from −2.5 (strong reduction in the mutant) to 2.5 (strong increase). Raw data corresponding to these measurements are provided in Table S6.

We also examined colony biofilm formation using Congo red agar. Congo red dye is known to bind to the glucose-rich exopolysaccharide Pel, which is an essential component of biofilm architecture in P. aeruginosa PA14 and plays a role in virulence and survival (78, 80, 81). Our results show that PqsE visually affects the global appearance and binding to Congo red of colony biofilms in 10 out of 12 strains, even though the resulting morphologies vary rather than consistently displaying the hyper-rugose colony phenotype observed for PA14 (Fig. 4). The hyper-rugose colony phenotype observed for the PA14∆pqsE mutant has been linked to the absence of phenazine production (78). Since most pqsE mutants have reduced PCA and pyocyanin production (Fig. 2), strain-specific differences in phenazine concentrations may influence colony architecture in distinct ways. Strain A22, which does not produce HAQs, shows no difference in biofilm formation on either polystyrene plates or Congo red agar upon pqsE deletion, aligning with observations for phenazines and rhamnolipids, in which no differences were noted between the mutant and the WT strain (Fig. 2 and 3).

Fig 4.

Colony biofilm formation of P. aeruginosa strains and their isogenic ∆pqsE mutants. Overnight cultures of each strain and their respective ∆pqsE mutants were diluted to an initial OD₆₀₀ of 0.05, in tryptic soy broth medium, and incubated at 37°C with agitation until they reached an OD₆₀₀ of 0.6. Five microliters of each culture was spotted on 1% agar plates containing 1% tryptone, 40 µg/mL Congo red, and 20 µg/mL Coomassie brilliant blue, as described before (82). Pictures were taken with a binocular microscope (Olympus Life Sciences) after 6 days of incubation at room temperature. The scale bar is 2 mm.

The effects of PqsE on biofilm formation vary between strains, likely due to multiple interacting factors. For instance, rhamnolipids play a crucial role in biofilm dispersion and architecture (72, 83–86). In strains where pqsE disruption leads to reduced rhamnolipid production, biofilm formation may be enhanced (72, 84). Additionally, recent studies with PAO1 show that PqsE modulates levels of c-di-GMP through its interaction with the ProE phosphodiesterase (87). A pqsE mutant exhibits higher c-di-GMP levels, which could correlate with enhanced biofilm production (87–89). Conversely, in reference strains, PqsE positively regulates genes involved in biofilm formation, including cupA1, lecA, and lecB (68). Overall, the interplay of these opposing factors likely contributes to the variations in biofilm formation across different strains, highlighting the complexity of this process. While PqsE has a consistent role in biofilm formation across P. aeruginosa strains, its effect is variable across strains, suggesting that strain-specific contexts might modulate its activity.

We also examined swarming motility, another multicellular behavior depending on RhlR. Swarming motility is a collective behavior characterized by the rapid and coordinated movement of groups of cells on a semisolid surface (90–93). To swarm, P. aeruginosa requires both rhamnolipid production and a functional flagellum (94, 95). PqsE is thought to be involved in this social behavior, as a pqsE mutant in the PAO1 strain shows reduced swarming motility under some conditions (25). Since swarming motility requires rhamnolipids, we assessed swarming motility in the four strains where the ∆pqsE mutation led to reduced rhamnolipid production (E90, JJ692, PA-CL512, and PA-W9) (Fig. 2B). We compared the swarming patterns of the WT strains and their ∆pqsE mutants to that of the PA14 strain, for which rhamnolipid production remains unaffected by pqsE depletion. Among these strains, two (E90 and JJ692) exhibited a modest reduction in swarming motility upon ∆pqsE deletion (Fig. 5), consistent with the rhamnolipid quantification results (Fig. 2B). Notably, swarming motility in strain PA-CL512∆pqsE was severely impaired, aligning with its pronounced decrease in rhamnolipid production. The remaining strain (PA-W9) displayed a minor alteration in the swarming pattern but no significant change in swarm coverage (Fig. 5). Indeed, this strain only showed a very slight decrease in mono-rhamnolipid (Rha-C10-C10) production in the ∆pqsE mutant, which may not be sufficient to have a significant impact on the swarming behavior (Fig. 2B). Our results confirm that PqsE can indeed be involved in swarming motility, most likely through its impact on rhamnolipid production.

Fig 5.

Swarming motility of four P. aeruginosa isolates exhibiting reduced rhamnolipid production in their ∆pqsE mutant. PA14 and its isogenic ∆pqsE mutant were used as negative controls. Images were acquired after 24 h of incubation at 37°C on semisolid MD9CAA medium.

Overall, PqsE generally affects RhlR-dependent multicellular behaviors, including swarming motility and biofilm development. However, the extent of this influence varies across isolates. Given the complexity of biofilm regulation, PqsE likely influences multiple regulatory pathways, leading to strain-specific outcomes. Despite the variability of phenotypes between strains, PqsE appears to play a consistent and functionally relevant role in shaping multicellular structures in P. aeruginosa isolates.

The function of PqsE in HAQ-defective strains remains unknown

After confirming the genetic conservation of pqsE, we assumed it was transcribed in all our strains. Among the 12 P. aeruginosa isolates from our panel, 11 produce HAQs (13, 14) (Table 1). In those strains, there is likely transcription of the pqs operon and, consequently, pqsE (18). Supporting this, all pqsE mutants of HAQ-producing strains exhibit differences in at least one of the tested phenotypes compared to their respective WT strains (Fig. 2 to 4). Strain A22 does not produce HAQs but still produces pyocyanin (14) (Table S5). Notably, previous studies have shown that pqsE can still be expressed in some strains that lack HAQ production yet still produce pyocyanin (52, 53). However, A22 is the only isolate from our panel with no significant differences with the ∆pqsE mutant in any tested virulence determinants, including the production of phenazines and rhamnolipids (Fig. 2). The same goes for multicellular behaviors (Fig. 3 and 4). This raised the possibility that pqsE may not be transcribed in this strain under our experimental conditions, and these factors would be produced without involvement for a PqsE homolog. To investigate this further, we performed reverse transcription polymerase chain reaction (RT-PCR) on pqsE in the A22 strain. The PA14 reference strain, as well as PA14 carrying a non-polar mutation in pqsA were used as control. This mutant can still express pqsE and produce detectable pyocyanin while lacking HAQ production, which represents an ideal control for the A22 HAQ-defective strain (Fig. 6, data not shown). Our results confirm that A22 does transcribe pqsE, reinforcing that it can occur independently of HAQ production, as previously reported for some P. paraeruginosa strains (Fig. 6) (52, 53). Interestingly, despite pqsE transcription, deletion of pqsE in A22 does not affect pyocyanin production, suggesting that PqsE-dependent phenotypes may be independent of PqsE in this strain. However, this interpretation warrants further investigation. Previous studies have reported only partial dependence of RhlR towards PqsE, supporting the possibility of a subset of strains having PqsE-independent RhlR activity (52). Globally, the role of PqsE, as well as the whole QS regulatory circuitry, remains largely unknown in HAQ-defective strains. The existence of such strains questions the development of PqsE-targeted anti-virulence therapies. Given that HAQ-negative strains represent approximately 20% of all P. aeruginosa isolates (13, 14), elucidating the mechanisms underlying this adaptation is an important avenue for future research.

Fig 6.

RT-PCR of pqsE transcription in the A22 strain. A 2% agarose gel displaying RT-PCR amplification of pqsE mRNA levels from strain A22 alongside the reference strain PA14 and a PA14pqsA non-polar mutant (PA14pqsA_np). Cultures were grown in King’s A medium at 37°C with agitation for 6 h. The nadB gene was used as a housekeeping control. Genomic DNA (gDNA) served as a positive control (+), while a no-reverse transcriptase (no-RT) sample was included as a negative control (−).

HAQ production can also be a PqsE-dependent determinant

In addition to regulating the production of virulence factors through RhlR, PqsE has a thioesterase activity. PqsE catalyzes the conversion of 2-aminobenzoylacetyl-coenzyme A to 2-aminobenzoylacetate, which is ultimately used to produce HHQ and PQS, the key ligands of the transcriptional regulator of the PQS system, MvfR (16, 18, 23). However, some studies have shown that PqsE is not strictly essential for this function, as a pqsE mutant in PA14 and PAO1 is reported to produce WT levels of HHQ and PQS (18, 19, 23). This was explained by the fact that other thioesterase enzymes, such as TesB, can compensate for PqsE’s enzymatic activity (23). To verify whether what was observed in well-studied strains remains true in other strains, we assessed HAQ levels in our panel (Fig. 7). Specifically, we quantified HHQ and PQS production at two distinct time points in HAQ-producing strains (i.e., A22 was not included). These time points were selected based on prior studies analyzing HAQs in large collections of independent strains (13, 14).

Fig 7.

Heatmap of the production of HHQ and PQS in the ∆pqsE mutants compared to the P. aeruginosa WT strains. Concentrations of HAQs (HHQ and PQS) were measured by LC/MS/MS at 6 and 24 h. Values are log₂-transformed ratios (log₂ fold change [Log₂FC]) of production in ∆pqsE mutants relative to their WT strain (∆pqsE/WT production). Red shades indicate reduced production of metabolites in the mutant, whereas blue indicates increased production. Gray indicates no detectable production of PQS. The intensity of the color reflects the magnitude of the change: darker shades indicate greater differences in the logarithmic scale. The color scale ranges from −5 (strong reduction in the mutant) to 5 (strong increase). To enable log transformation and visual consistency, zero values (absence of production) in the mutants were adjusted using a pseudo-count of 1. Ratios were then capped at the lower end of the scale (minimum log₂ ratio of −5). Raw data for these measurements are provided in Tables S8 and S9.

Surprisingly, most strains have significantly lower HHQ and/or PQS production in ∆pqsE mutants compared to their respective WT strains. In some cases, HAQ production was completely abolished in the absence of pqsE (Fig. 7). This suggests that other thioesterase enzymes may not be as effective in most strains to fully compensate for the loss of PqsE activity, in contrast with the previous suggestion (23), and that this corresponds to an expected involvement for the presence of pqsE in the HAQ biosynthetic gene cluster.

Also unexpectedly, a subset of strains (39016, PA-CL513, PAO1, and PA14) behaved similarly: the PqsE enzymatic activity did not appear to be essential for HAQ production at either time point. In these strains, HAQ levels were even higher in the mutants than in the parental strain, with an accumulation especially observed in the later stages of growth (24 h) (Fig. 7).

Furthermore, PqsE exerts a negative autoregulatory feedback loop on the transcription of the pqsABCDE operon in PAO1 again through its impact on RhlR (25). However, whether this is also the case in other P. aeruginosa strains remains to be determined.

These findings highlight the crucial role of PqsE thioesterase activity in the synthesis of HAQs across many P. aeruginosa strains, contradicting previous conclusions that were based on limited assessment in reference strains. Notably, this challenges the widely accepted notion that PqsE is not involved in HAQ production.

Conclusion

The thioesterase PqsE moonlighting as a quorum sensing effector protein has been extensively studied in recent years, particularly concerning its role in promoting the RhlR regulon through a protein-protein interaction (24, 26, 29–31). Because of its impact on virulence and its presence restricted to very few species such as P. aeruginosa, PqsE could be an interesting target for the development of specific antivirulence therapies. To reinforce its appeal, it is important to show that the presence and functionality of PqsE is a conserved trait among P. aeruginosa strains outside of the few prototypical strains described in the literature. In this study, we explored the conservation and function of PqsE in a panel of 12 independent strains of P. aeruginosa, including PA14 and PAO1. To achieve this, we looked at the conservation of the pqsE gene as well as the impact on phenotypes considered to be PqsE dependent.

Our findings confirm the genetic conservation of pqsE and rhlR, consistent with previous reports (54, 55). Indeed, the few identified single-nucleotide polymorphisms do not affect the predicted amino acid sequence. Furthermore, our phenotypical survey of the pqsE mutants from our panel reveals that while the impacts might vary, phenotypes affected by PqsE remain consistent across strains. Furthermore, our study highlights that PqsE is essential for complete HAQ biosynthesis in most strains, challenging previous assumptions (18, 19, 23).

Since our strains are not closely related and contain variations within QS systems (14, 29), our results likely reflect the diversity of the PqsE function, emphasizing the importance of studying multiple isolates before reaching general conclusions. The reasons why strain A22 does not require PqsE to activate the RhlR regulon and other known PqsE targets remain unclear and warrant further investigation. The possibility that some strains retain full RhlR activity independently of PqsE remains of interest and questions the relevance of PqsE as a therapeutic target.

The interaction between PqsE and RhlR has been proposed as an interesting target for antivirulence therapies due to its conservation and functional importance in producing some virulence factors in well-studied strains (26, 30, 31). Further studies on additional PqsE targets are needed to gain a better understanding of the role of this protein in QS regulation. Moreover, testing independent P. aeruginosa strains and their isogenic ∆pqsE mutants in animal models is essential to deepen our understanding of PqsE’s role in virulence. Such in vivo studies are critical to determine whether the functional relevance of PqsE translates into therapeutic potential.

MATERIALS AND METHODS

Strains and growth conditions

The isolates selected for this study were collected from the environment and various chronic and acute clinical infections. Clinical strains were obtained from the International Pseudomonas Consortium Database (41), while other strains were previously isolated from hospital sinks and soils. Reference strains UBCPP-PA14 (96) and PAO1-N (University of Nottingham collection) were also included (Table S1). A detailed list of the selected isolates and some of their genetic and phenotypic characteristics is presented in Table 1. Unless indicated otherwise, tryptic soy broth (TSB) medium (BD Difco) was used for the routine growth of bacteria, and cultures were incubated at 37°C in a TC-7 roller drum (New Brunswick) at 150 r.p.m.

Genome annotation and sequence analyses

Genomes were annotated with Prokka v.1.14.6 (97) with default parameters except for the use of a Prokka genus-specific database (--usegenus --genus Pseudomonas). Subsequently, MAFFT v.7.511 was used to align the pqsE and rhlR sequences and identify mutations (98). Alignments were then visualized with Jalview v.2.11 (99).

Generation of phylogenetic tree and ANI heatmap

A core gene alignment was done with PyMLST v.2.1.6 (100) and the cgMLST.org core genome MLST scheme for P. aeruginosa (https://www.cgmlst.org/ncs/schema/Paeruginosa85/). Monomorphic and underrepresented sites were filtered from the nucleotide matrix with BMGE v.1.12 (101). Then, a maximum likelihood phylogenetic tree was constructed using IQ-TREE v.2.3.6 (102), with statistical support from 1,000 bootstrap replicates. The optimal substitution model was automatically inferred by IQ-TREE (GTR + F + I + G4). The resulting tree was visualized with FigTree v.1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/). ANI calculations were done with pyani v.2.1.6 (https://github.com/widdowquinn/pyani) using the “ANIm” all-versus-all comparison mode, with default parameters. The resulting percentage matrix was aligned with the nodes of the phylogenetic tree described above.

Plasmid construction

The pMT01 plasmid (pEX18Gm-∆pqsE) was constructed by modifying the existing sacB-containing suicide plasmid pEX18Ap-∆pqsE, which contains a 570 bp deletion allele of the pqsE gene (18). The pEX18Ap-∆pqsE plasmid was digested with EcoRI and BamHI, releasing a 1,905 bp ∆pqsE fragment. This fragment was purified and ligated into the pEX18Gm backbone with the same restriction enzymes (103), using the T4 DNA ligase (NEB). A comprehensive list of the plasmids used in this study is provided in Table S10.

Construction of in-frame deletion mutants

An allelic replacement method adapted from Hmelo et al. was used to create deletion mutants in the pqsE gene (104). Suicide vector pMT01 was introduced into recipient P. aeruginosa strains by conjugation with donor auxotrophic Escherichia coli strain χ7213 on plates containing 50 µg/mL diaminopimelic acid (DAP). Merodiploid cells were selected on media containing gentamicin at predetermined concentrations for each strain. Double crossover mutants were isolated through sucrose counterselection, and the mutation in pqsE was confirmed by PCR, using primers listed in Table S11.

Quantification of secondary metabolites and quorum sensing molecules

For the quantification of PCA, pyocyanin, and HAQ molecules (HHQ and PQS), overnight cultures of WT strains and their isogenic ∆pqsE mutants were diluted to an OD₆₀₀ of 0.05 in King’s A broth supplemented with 100 µM of FeCl₃ (105). Cultures were prepared in triplicate and incubated at 37°C under agitation for 6 and 24 h, with sampling time points selected based on previous studies (13). To extract metabolites, 375 µL of acetonitrile containing tetradeuterated 4-hydroxy-2-heptylquinoline as an internal standard was added to 1.5 mL of culture sample. The suspension was vortexed and centrifuged for 10 minutes at 17,000 × g to pellet the bacteria. Supernatants were transferred into vials and analyzed using a LC/MS/MS method, as described previously (106).

For rhamnolipid quantification, a similar procedure was followed for the preparation of samples with some modifications. Only the 24 h time point was used to measure rhamnolipid accumulation, and supernatants were diluted 20-fold before analysis. The quantification of rhamnolipids in the supernatant was performed by LC/MS/MS, as described before (64).

Due to the tendency of many strains to form clumps, precluding the use of absorbance to assess growth, relative concentrations of secondary metabolites and HAQ molecules were normalized to the total protein content of the cell pellet collected from the whole culture at the time of sampling. The pellets were resuspended in 0.1 N NaOH and incubated at 70°C for 1 h. Total protein concentrations were measured using the Bradford protein assay (Bio-Rad Laboratories, Montreal, Canada), with bovine serum albumin serving as a standard.

RT-PCR

Overnight cultures were diluted in triplicate to an OD₆₀₀ of 0.05 in King’s A medium and incubated at 37°C with agitation. Cells were harvested after 6 h of growth. Total RNA was extracted using Aurum Total RNA Mini Kit (Bio-Rad Laboratories). To eliminate any residual DNA, the extracted RNA was treated with TURBO DNA-Free Kit (Ambion, Life Technologies). Reverse transcription was performed using iScript gDNA Clear cDNA Synthesis Kit (Bio-Rad Laboratories). A portion of the pqsE gene was amplified using specific primers (Table S11), and the resulting PCR products were analyzed by electrophoresis on a 2% agarose gel. The nadB gene was used as a housekeeping control (107), and no-RT and gDNA controls were also included.

Swarming motility

Swarming motility assays were performed as previously described (108). Briefly, 20 mL of M9DCAA medium with 0.5% Bacto-agar (Difco) was poured into 100 mm Petri dishes and allowed to dry in a laminar flow cabinet. Overnight cultures were adjusted to an OD₆₀₀ of 3.0, and 5 µL was inoculated in the center of an agar plate. Plates were then incubated at 37°C for 16 h. Each strain was tested in triplicate.

Biofilm formation

Biofilm formation was quantified using crystal violet staining, as described before (79). Briefly, overnight cultures of all strains and their isogenic ∆pqsE mutants were diluted to an OD₆₀₀ of 0.05 in M63 minimal medium supplemented with 1 mM magnesium sulfate, 0.2% dextrose, and 0.5% casamino acids. For each strain, a 100 µL aliquot was added to 5 wells of polystyrene 96-well plates, which were incubated at 37°C for 24 and 48 h. After incubation, the plates were rinsed thoroughly with water, and 125 µL of 0.1% crystal violet was added to each well. Following a 15 min incubation at room temperature, the plates were rinsed again, and the bound dye was solubilized in 125 µL of 30% acetic acid. Absorbance was measured at 550 nm using a Cytation microplate reader (Biotek).

Colony biofilm formation

Overnight cultures of WT strains and ∆pqsE mutants were diluted to an OD₆₀₀ of 0.05 in TSB medium and incubated at 37°C under agitation until they reached an OD₆₀₀ of 0.6. As described before, 5 μL of each dilution was inoculated onto 1% agar plates containing 1% tryptone, 40 µg/mL Congo red, and 20 µg/mL Coomassie brilliant blue (82). Plates were incubated at room temperature for 6 days. Pictures of colony biofilms were taken with a binocular microscope (Olympus Life Science). This experiment was repeated at least twice for each strain.

Statistical analyses

Statistical significance in metabolite production between the WT strains and their respective ∆pqsE mutants was determined using the Holm-Sidak method, with a significance threshold (P-value) of 0.05 (GraphPad Prism). When a significant difference was detected, the production or expression levels in the ∆pqsE mutants were divided by those of the WT strains to calculate the proportion. In cases where no significant difference was observed, the production of the ∆pqsE mutant relative to the WT strain was assumed to be 1. Proportions were subsequently log₂-transformed for analysis. A small pseudo-count of 1 was added for instances where the mutant’s production or expression was zero to enable log transformation. Log₂-transformed proportions were capped at a subjective minimum threshold (indicating the absence of production) to prevent extreme values. Heatmaps were generated with the R software (109) using the package “pheatmap’’ (110).

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/jb.00402-25.

Supplemental figures and tables. jb.00402-25-s0001.pdf.

Figures S1 to S3 and Tables S1 to S10.

ASM does not own the copyrights to Supplemental Material that may be linked to, or accessed through, an article. The authors have granted ASM a non-exclusive, world-wide license to publish the Supplemental Material files. Please contact the corresponding author directly for reuse.