The end of the reign of a “master regulator’’? A defect in function of the LasR quorum sensing regulator is a common feature of Pseudomonas aeruginosa isolates

ABSTRACT

Pseudomonas aeruginosa, a bacterium causing infections in immunocompromised individuals, regulates several of its virulence functions using three interlinked quorum sensing (QS) systems (las, rhl, and pqs). Despite its presumed importance in regulating virulence, dysfunction of the las system regulator LasR occurs frequently in strains isolated from various environments, including clinical infections. This newfound abundance of LasR-defective strains calls into question existing hypotheses regarding their selection. Indeed, current assumptions concerning factors driving the emergence of LasR-deficient isolates and the role of LasR in the QS hierarchy must be reconsidered. Here, we propose that LasR is not the primary master regulator of QS in all P. aeruginosa genetic backgrounds, even though it remains ecologically significant. We also revisit and complement current knowledge on the ecology of LasR-dependent QS in P. aeruginosa, discuss the hypotheses explaining the putative adaptive benefits of selecting against LasR function, and consider the implications of this renewed understanding.

PERSPECTIVE

Current understanding of Pseudomonas aeruginosa quorum sensing

The bacterium Pseudomonas aeruginosa is a versatile opportunistic pathogen found mostly in environments related to human activity such as soil, water, and hospital settings (1). P. aeruginosa causes infections in immunocompromised individuals and people living with cystic fibrosis (CF) (2–4). To adapt to diverse and dynamic environments, this bacterium uses quorum sensing (QS), an intercellular communication system. Through the production and detection of small diffusible autoinducers, QS coordinates the expression of numerous key cellular functions at the population level, including virulence, in a cell density-dependent manner (5).

P. aeruginosa has three interdependent QS systems, known as las, rhl, and pqs. Based on observations of prototypical strains such as PA14 and PAO1, the las system is generally considered to be at the top of the QS regulatory cascade (6, 7). This system comprises the LasI synthase, which produces the autoinducer N-(3-oxododecanoyl)-L-homoserine lactone (3-oxo-C₁₂-HSL). This signal molecule subsequently binds to its cognate LuxR-type transcriptional regulator, LasR. Once activated by its ligand, LasR regulates the transcription of several target genes, such as lasB encoding for the LasB elastase, as well as lasI, which codes for the LasI synthase, completing a positive feedback loop that is a characteristic of many QS systems (8–10). LasR activation also induces the transcription of the rhlI and rhlR genes, which constitute the rhl QS system (6, 7, 10–12). The RhlI synthase produces the N-butanoyl-L-homoserine lactone (C₄-HSL) autoinducer signal, which can bind to another LuxR-type transcriptional regulator, RhlR, activating the transcription of several genes, including those implicated in the production of virulence factors, including pyocyanin (phz1 and phz2 operons) and rhamnolipids (rhlAB and rhlC) (11, 13–15). The third QS system, pqs, is regulated by MvfR (also known as PqsR), a LysR-type transcriptional regulator. The pqs system relies on the production of 4-hydroxy-2-alkylquinolines (HAQs) by the enzymes encoded by the pqsABCDE operon (16, 17). Indeed, the PqsABCD enzymes synthesize 4-hydroxy-2-heptylquinoline (HHQ), which can be converted into the Pseudomonas quinolone signal (PQS; 3,4-dihydroxy-2-heptylquinoline) by the genetically unlinked PqsH monooxygenase (16, 18, 19). HHQ and PQS can both act as autoinducing ligands of MvfR, leading to the transcriptional activation of the pqs operon (18, 20).

The interconnection of all three systems is widely recognized. LasR activation positively regulates the expressions of rhlR, mvfR, and pqsH, while RhlR activation represses pqs operon activity (10, 15, 17, 21–24). Additionally, PqsE, a thioesterase enzyme encoded by the last gene of the pqs operon, plays an important role in the regulation of P. aeruginosa QS (25). Indeed, this protein interacts with RhlR through an incompletely defined chaperone-like function to activate the rhl system and heighten the expression of some target genes (12, 26–29). These examples highlight the complex regulation that characterizes P. aeruginosa QS. However, while most of the research that informed this interconnected but hierarchical model of QS was performed in a few, well-defined laboratory strains, the frequent detection of LasR-defective strains still able to produce QS-regulated factors has revealed certain inconsistencies regarding the central role of the LasR regulator within the QS hierarchy. Here, we will propose an alternative perspective and discuss this issue based on current literature, unveiling new hypotheses regarding the relative positions of LasR, RhlR, and MvfR (PqsR) within the P. aeruginosa QS hierarchy.

LasR-defective strains are generally prevalent in both chronic and acute clinical environments

Over time, numerous studies have highlighted the natural occurrence of P. aeruginosa strains with impaired LasR activity. Initial reports of LasR-defective variants reflected their detection within microbial populations that had evolved for years within the chronically infected lungs of people with CF, which was unexpected given the requirement of QS-controlled factors for full expression of bacterial virulence in vitro (30). However, multiple reports have identified LasR-defective strains in CF respiratory cultures, and the selection of these strains by the CF lung environment is now a well-accepted dogma. The emergence of lasR mutants has been associated with increased inflammatory markers, increased neutrophilic inflammation, and deteriorated pulmonary function (31–37). These variants have also been implicated in pathogenesis in corneal ulcers and in the airways of individuals with chronic obstructive pulmonary disease (COPD) and ventilator-associated pneumonia (38–40). Moreover, we recently identified a high prevalence of LasR-defective strains in non-clinical environments as well, such as sinks, hydrocarbon-contaminated soils, and animal products (4). While a defect in LasR activity has seldom been reported in strains isolated from acute infections, a recent study from O’Connor et al. (41) identified P. aeruginosa isolates carrying lasR mutations commonly present within many environments, although the impact of these mutations on LasR activity per se was not investigated (38, 41, 42).

Based on the emerging picture of a widespread prevalence of LasR-defective strains in various environmental niches, we hypothesized that defects in LasR activity could also be found in acute clinical contexts. To evaluate the prevalence of LasR-defective strains isolated from both chronic and acute infections, we assessed LasR activity in 92 P. aeruginosa strains from diverse sources, including burn wounds, keratitis, urinary tract infections, bronchitis, CF lungs, COPD, and others (Table S1) (43). LasR activity was evaluated using a previously published method relying on unbiased phenotypic profiling based on the quantification of QS-regulated metabolites such as HAQs and pyocyanin (4). Based on this model, we found that 41% of the 92 isolates from our panel of clinical strains exhibited impaired LasR activity (Fig. 1). These data are consistent with our previous findings, where 40% of isolates from a panel of 176 P. aeruginosa environmental strains were found to be LasR-defective (4). These results also support the conclusions of O’Connor et al., as well as our hypothesis that LasR-defective strains are common in all clinical contexts and are not restricted to CF clinical isolates, as previously assumed (41), due to a predominant focus of prior analyses on CF isolates. Here, we found the prevalence of defects in LasR function to be similar among isolates from various environments, including when comparing CF and non-CF sources.

Fig 1.

Clustering analysis performed on a panel of 92 strains of P. aeruginosa isolated from acute and chronic infections. Clustering analysis is based on selected variables from a previous study. Briefly, HAQ concentrations and pyocyanin production were measured in King’s A medium at two different time points (4). Strains PA14 (Functional LasR; top) and PA14 lasR::Gm (LasR-defective; bottom, “lasR”) were included in the analysis as references. Three robust clusters were identified. One of these clusters comprises LasR-defective strains, whereas another includes strains with a functional LasR protein. The third cluster comprises strains that produce negligible levels of HAQs. Further analyses are required to confirm the functionality of LasR in this subset of isolates, as described before (4). Raw data used to generate this analysis are presented in Table S2. Statistical analyses were made as previously described using R software (4, 44). Robustness (*) represents the proportion of clustering runs in which a pair of isolates appeared together in some clusters, given that they were clustered together in at least one run, averaged over all such pairs (***: 80%–89%).

In the literature, the prevalence of lasR mutants among P. aeruginosa isolates from CF patients has been estimated to be from 20% to 25%, as defined using genotypic techniques (35, 45, 46). However, different studies have used a variety of methods to estimate the proportion of LasR-defective strains, hampering direct comparison of results. Relying solely on the identification of mutations in the lasR coding sequence, as it is usually reported, could misidentify LasR-defective strains, given its complex regulation and the unpredictable relationship between coding sequence and protein function (4, 35, 47). Indeed, a great diversity of mutations in the lasR gene has been identified, not all of which completely abrogate function (35). Additionally, some mutations in regulatory elements could also modify LasR expression or activity (4, 35). Examining a single phenotypic trait could also bias LasR-defective strain identification given the complex regulation of many phenotypes. Our approach, based on phenotypic profiling rather than examining only one trait, allows for unbiased identification of all isolates with a defect in LasR activity (4). Therefore, we refer to LasR-defective strains rather than lasR mutants.

Taken together with previous studies, the data presented here further highlight that LasR-defective strains occur commonly, regardless of their clinical or environmental origin (4, 35). As a result, we must reconsider prior assumptions, including the notion that chronically colonized CF lungs distinctively provide a selective environment promoting the loss of LasR function (32, 33).

Factors driving the emergence of LasR-defective strains remain elusive

Despite decades of research, the precise drivers for the emergence of LasR-defective strains remain elusive. LasR function seems especially prone to be lost, and the lasR gene might even be considered a hotspot for various types of genetic variations (4, 35, 41). One particularly popular hypothesis suggests that LasR-defective cells arise as “cheaters,” taking advantage of neighboring cells with a functional LasR that provide “public goods,” such as exoproteases (e.g., LasB elastase), to the entire population. By exploiting this strategy, LasR-defective variants reduce the population metabolic cost associated with the expression of these proteins when they are essential for bacterial growth. Interestingly, this behavior has been mostly observed under specific in vitro conditions (39, 48–51), as well as suggested under in vivo conditions (52, 53). While such social cheating behavior could potentially extend to different contexts and environments beyond the CF lung, providing an interesting explanation for the emergence of these variants in multiple niches, evidence suggests that social conflict is not the sole source of LasR loss-of-function and that lasR mutants are not always “cheaters” (54). It is important to consider that LasR-defective strains can also arise in various conditions that do not implicate “public goods” (55). Therefore, other factors must contribute to the emergence of LasR variants.

As stated above, LasR-defective strains have been typically associated with the chronically infected lungs of people with CF (30, 32, 56). Extensive investigations have also explored the non-social benefits conferred by defective LasR function in this specific environment, revealing their relatively high fitness in the presence of specific amino acids, such as phenylalanine, which is especially abundant in CF secretions (32, 57). Relative to wild-type strains, these variants exhibit altered metabolism, including lower oxygen consumption and enhanced nitrogen utilization, providing them with a competitive edge over their wild-type counterpart (56, 58). Furthermore, LasR-defective strains exhibit relative resistance to specific antimicrobials and enhanced tolerance to alkaline stress, resulting in protection from cell lysis (32, 33, 59–61). These findings support the notion that diminished LasR activity confers substantial advantages in conditions known to be common in CF lungs. However, the growing evidence for the prevalence and abundance of LasR-defective variants in diverse infections and environmental contexts disproves any specificity for the CF lung and warrants expanding current models of their emergence. For instance, it was proposed that diversification of P. aeruginosa populations in the CF lung could favor long-term survival to multiple unpredictable stresses (62). This phenomenon could also apply outside of the CF lung. Accordingly, the effects of LasR impairment on growth on different carbon sources were suggested to explain the emergence of LasR-defective variants in the context of the CF lung, but such non-social mechanisms could likely arise in many other environmental settings, as growth conditions and carbon sources differ (63).

Alternatively, we might consider the emergence of LasR-defective variants as beneficial for a population that contains them (64). Supporting this model, controlled evolution experiments performed in vitro demonstrated the tendency of LasR-defective clones to rapidly emerge, with their proportion frequently stabilizing at about 50% of the total population (48, 50, 55, 59, 65). For instance, swarming colonies of P. aeruginosa with higher proportions of LasR-defective cells tend to have fitness advantages over those without LasR variants (55). The population dynamics of mixed populations of P. aeruginosa, comprising wild-type and LasR-defective strains, could partially explain the prevalence of the latter in diverse environments. In natural settings, P. aeruginosa forms biofilms, which are bacterial communities composed of diverse physicochemical niches. Jeske et al. showed that LasR function is especially prone to be lost in a biofilm context (42). Within these biofilms, factors produced by strains with a functional LasR could influence the behavior of surrounding LasR-defective strains within specific niches. Similarly, LasR-defective strains can modulate the behavior of LasR-functional strains by affecting, for instance, QS signaling and function (66, 67).

Hence, the concomitant mixed presence of LasR-functional and LasR-defective strains appears to be beneficial to the overall population. This observation suggests that LasR plays an important ecological role and remains essential for the viability of P. aeruginosa populations. There could be various reasons for the selection of such mixed populations, including both social and non-social drivers, considering the multitude of factors that influence the diversification of this protein.

LasR-defective clinical isolates could be acquired directly from surrounding environments

The same high prevalence of LasR-defective strains in both clinical and environmental contexts suggests potential adaptative benefits associated with modulating LasR activity in populations of P. aeruginosa, irrespective of the environment. Therefore, it is possible that LasR-defective strains isolated from infections originated from environmental sources, as previously proposed, rather than emerging in situ (4, 38). Two distinct studies provided evidence supporting this notion, as the lasR gene of some P. aeruginosa isolates already exhibited mutations at initial detection in CF respiratory samples, consistent with a direct acquisition from environmental sources (40, 68). While P. aeruginosa lineages have been observed to lose LasR function over time during an infection, the discussion above indicates that such events are not necessarily specifically selected by the CF lung environment (30, 69–72).

Presence of naturally occurring LasR-impaired function is not necessarily synonymous with loss of quorum sensing

A widely held notion is that LasR occupies the top position in the QS regulatory hierarchy (47, 73). An alternative hypothesis is that the presence of this hierarchy is not a universal feature of P. aeruginosa and reflects studies conducted primarily with one strain, PAO1, which could itself be considered as an outlier regarding its QS mechanisms (74). We now know that isolates with defective LasR activity can still exhibit robust QS-dependent regulation of virulence factors (4, 35, 75–77). Since LasR upregulates the rhl and the pqs QS systems in studied laboratory-adapted strains, it has been generally assumed that a defect in LasR activity should result in deficient QS regulation and thus reduced transcription of target genes and production of virulence factors. However, LasR-defective strains with a functional rhl QS system, referred to as “RAIL’’ (RhlR Active Independently of LasR) strains, have been identified. Such strains have been isolated from CF lungs, as well as from diverse environments (4, 35, 75–77). Using the same method as previously mentioned, we found that about half of the LasR-defective clinical strains from our panel (Fig. 1) possess LasR-independent RhlR activity (Fig. 2), in accordance with previously published findings (4). Hence, an impaired LasR protein does not necessarily lead to diminished production of virulence factors. Thus, the dogma based on the study of prototypical strains stating that LasR is the conserved “master regulator” of the QS regulatory cascade should be reconsidered. Besides regulating known RhlR-dependent virulence factors such as pyocyanin, RhlR can activate virulence factors typically regulated by LasR, such as various exoproteases (4, 35, 78, 79). Together, these observations and concepts highlight the crucial role of RhlR, accentuating its importance as a QS regulator in P. aeruginosa. However, the precise regulatory mechanisms involved in LasR-independent RhlR regulation remain elusive. One mechanism might involve PqsE, encoded by the last gene of the pqs operon. PqsE plays a major role in promoting RhlR activity, at least in prototypical strains (12, 26, 29, 80, 81), underscoring the importance of MvfR and the pqs system for maintaining the full suite of QS regulation. Since an important subset of naturally evolved P. aeruginosa strains activates QS and produces virulence factors without relying on LasR, we need to better understand the nature of the ecological pressure sustaining its regulatory activity in some lineages.

Fig 2.

Clustering analysis performed on a subset of LasR-defective isolates of Pseudomonas aeruginosa shown in Fig. 1. Clustering analysis was based on chosen variables from a previous study. Briefly, HAQ concentrations, pyocyanin production, and activity of a rhlA-gfp reporter were measured in King’s A medium at two different time points using the same methods as previously described (4, 82). RhlR Active Independently of LasR (RAIL) strain E90 was included in the analysis as a reference. Raw data used to generate this analysis are presented in Table S2. Statistical analyses were made as previously described using R software (4, 44). Robustness (*) represents the proportion of clustering runs in which a pair of isolates appeared together in some clusters, given that they were clustered together in at least one run, averaged over all such pairs (****: >90%).

Improving quorum sensing inhibition by targeting RhlR-dependant factors

P. aeruginosa is naturally tolerant of, and resistant to, a wide range of antibiotics, limiting the efficacy of treating infections by this bacterium (83). To address this issue, anti-virulence therapies have emerged as a promising approach for the development of new drugs, since they offer distinct advantages. Unlike antibiotics that directly target the survival of the bacteria, anti-virulence therapies aim to inhibit non-essential virulence factors without affecting viability. This approach theoretically reduces the development of resistance since no selective pressure for survival is applied. In P. aeruginosa, QS represents an interesting target, since it modulates the expression of multiple virulence factors unrelated to bacterial survival (84–87). Rationally, many anti-virulence therapies target P. aeruginosa LasR or the production of 3-oxo-C₁₂-HSL, since the las system has been considered to be on top of the QS regulation cascade (88–90). Unfortunately, the high prevalence of LasR-defective strains and the clear absence of QS hierarchy in some strains suggest that LasR is not an ideal target. Instead, we propose that researchers working on anti-QS therapies should consider focusing on other QS targets, such as the rhl or pqs system, which can still be active in the absence of LasR and seem better conserved. There have been a very few reports of strains lacking RhlR or MvfR/PqsR activity, perhaps because these regulators are indispensable for the proper functioning of QS (41, 91–93). Thus, targeting rhl-dependent QS, such as the production/function of C₄-HSL or PqsE, could be an interesting approach in the control of P. aeruginosa infections and their clinical consequences. Studying the QS ecology in strains from a broader range of origins (clinical and environmental) should allow for better target selection.

Outstanding questions and conclusion

Acknowledging that maybe 40% of all P. aeruginosa isolates are LasR-defective raises several important questions. One particularly intriguing observation is the prevalence of mutations in lasR compared to those in its cognate synthase gene, lasI (41). A possible explanation for this discrepancy could be the absence of social benefit for lasI mutations: the inactivation of LasI would not prevent an autologous response to neighboring strains producing the diffusible 3-oxo-C₁₂-HSL signal in a mixed population (63). Another potential explanation is that 3-oxo-C₁₂-HSL does not exclusively serve as a LasR autoinducer and contributes to alternative functions in P. aeruginosa. For example, 3-oxo-C₁₂-HSL increases pyocyanin production in the absence of LasR (67). In fact, the maintenance of 3-oxo-C₁₂-HSL production in LasR-defective strains is an intriguing question that warrants further investigation. Globally, there seems to be an advantage for P. aeruginosa to conserve the function of LasI, while it is often not the case for LasR, as previously discussed. Finally, LasR-defective strains seem to maintain most QS-regulated functions; how this is working represents an important question to address in the coming years.

Loss of LasR function is common among P. aeruginosa isolated from any source, including both environmental and clinical settings, regardless of the acute or chronic nature of the infections. Furthermore, despite the absence of LasR activity, the rhl system is still functional in a subset of LasR-defective strains, ensuring QS functionality and the expression of survival and virulence factors. Nevertheless, it appears to be beneficial to have both LasR-defective and LasR-functional strains present in a population, as this combination of genotypes may play a beneficial role in maintaining P. aeruginosa’s ecological balance.

Based on available knowledge, current hypotheses, and accumulating data, we propose that LasR-defective strains isolated from infection-related settings could be acquired directly from the environment, where genomic diversification occurs in P. aeruginosa populations based on differences in regulation in various niches. We also propose that LasR is not as essential for functional QS as it has been commonly believed and that there may be a larger variety of QS architectures in P. aeruginosa than previously thought. Based on the rarity of isolates with completely deficient RhlR activity, and the relatively high frequency of LasR deficiency, we suggest that RhlR plays a more central role in QS regulation than LasR. Additionally, RhlR activity depends on MvfR-dependent regulation through the expression of pqsE via the pqs operon. More investigation on the importance of this partnership, until now mainly investigated in prototypical strains, throughout a larger panel of strains is needed, given the possible importance of the pqs system in maintaining QS activity in P. aeruginosa. Accordingly, to better understand the diversity of QS, future studies should focus on P. aeruginosa obtained from a broader range of environments and consider the dynamics of cocultures with naturally co-isolated strains.

SUPPLEMENTAL MATERIAL

The following material is available online at https://doi.org/10.1128/mbio.02376-23.

Supplemental Tables. mbio.02376-23-s0001.pdf.

Tables S1 and S2.

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.