|Home | About | Journals | Submit | Contact Us | Français|
The opportunistic pathogen Pseudomonas aeruginosa uses intercellular signals to control the density-dependent expression of many virulence factors. The las and rhl quorum-sensing systems function, respectively, through the autoinducers N-(3-oxododecanoyl)-l-homoserine lactone and N-butyryl-l-homoserine lactone (C4-HSL), which are known to positively regulate the transcription of the elastase-encoding gene, lasB. Recently, we reported that a second type of intercellular signal is involved in lasB induction. This signal was identified as 2-heptyl-3-hydroxy-4-quinolone and designated the Pseudomonas quinolone signal (PQS). PQS was determined to be part of the quorum-sensing hierarchy since its production and bioactivity depended on the las and rhl quorum-sensing systems, respectively. In order to define the role of PQS in the P. aeruginosa quorum-sensing cascade, lacZ gene fusions were used to determine the effect of PQS on the transcription of the quorum-sensing system genes lasR, lasI, rhlR, and rhlI. We found that in P. aeruginosa, PQS caused a major induction of rhlI′-lacZ and had lesser effects on the transcription of lasR′-lacZ and rhlR′-lacZ. We also observed that the transcription of both rhlI′-lacZ and lasB′-lacZ was cooperatively effected by C4-HSL and PQS. Additionally, we present data indicating that PQS was not produced maximally until cultures reached the late stationary phase of growth. Taken together, our results imply that PQS acts as a link between the las and rhl quorum-sensing systems and that this signal is not involved in sensing cell density.
Pseudomonas aeruginosa is a ubiquitous environmental organism capable of infecting a wide variety of animals, plants, and insects. As a human pathogen, this bacterium is a major source of opportunistic infections in both immunocompromised individuals and cystic fibrosis patients. P. aeruginosa is now the leading source of gram-negative nosocomial infections (25) and causes chronic lung infections in approximately 90% of cystic fibrosis patients (7). The ability of this organism to cause devastating infections stems from the production of an arsenal of virulence factors, several of which are controlled according to cell density through an elegant mechanism known as quorum sensing. In gram-negative bacteria, most quorum-sensing systems consist of a LuxR-type transcriptional activator (R protein) and an acylated homoserine lactone signal molecule (autoinducer) (see reference 4 for a review). When a bacterial culture is at a low cell density, basal levels of autoinducer and R protein are produced. As a population grows, the concentration of autoinducer increases with cell density until it reaches a threshold concentration where it binds to and thereby activates an R protein. Activated R protein then activates specific genes, causing cell density-dependent gene expression.
In P. aeruginosa, there are at least two quorum-sensing systems, las and rhl, which control the expression of numerous genes (see reference 21 for a review). The las quorum-sensing system consists of the transcriptional activator LasR and the autoinducer N-(3-oxododecanoyl)-l-homoserine lactone (3-oxo-C12-HSL), the synthesis of which is directed by the LasI autoinducer synthase (5, 15, 16). Similarly, the rhl system consists of the transcriptional activator RhlR and the autoinducer N-butyryl-l-homoserine lactone (C4-HSL), the synthesis of which is directed by the RhlI autoinducer synthase (12, 13, 17). Together, the las and rhl systems have been shown to regulate between 1 and 4% of the genes carried by P. aeruginosa, demonstrating the global importance of these intercellular signaling systems (27).
While the las and rhl quorum-sensing systems consist of two separate regulons, their functions are apparently not independent. Both of these systems have been shown to control the transcription of lasB, which encodes the major virulence factor LasB elastase (2, 5, 8, 18). It has also been shown that the las quorum-sensing system controls the rhl quorum-sensing system at both the transcriptional and posttranslational levels (9, 20). The discovery of this link between the systems indicated that they were arranged in a hierarchy where the las quorum-sensing system is dominant over the rhl quorum-sensing system.
Recently, we reported that P. aeruginosa produced a third intercellular signal in addition to the two homoserine lactone-type autoinducers. The Pseudomonas quinolone signal (PQS), which is capable of inducing lasB in P. aeruginosa, was identified as 2-heptyl-3-hydroxy-4-quinolone (22). Our initial characterization of PQS indicated that the production of this novel intercellular signal occurred only in the presence of an active form of LasR. We also demonstrated that exogenous PQS exhibited bioactivity in the lasR mutant strain PAO-R1 but not in the lasR rhlR double mutant strain PAO-JP3. Taken together, these results suggested that LasR was required for PQS production and that RhlR was important for PQS bioactivity (22). This indicated that PQS was intertwined in the quorum-sensing hierarchy, but because of the induction cascade that occurs, its exact role was difficult to determine. Therefore, we set out to further characterize PQS bioactivity and production.
In this report, we present data indicating that PQS provides a link between the las and rhl quorum-sensing systems. We show that PQS strongly induced rhlI in P. aeruginosa and had lesser positive effects on the transcription of lasR and rhlR. Our data indicated that rhlI and lasB are both cooperatively regulated by PQS and C4-HSL. We also determined that PQS was produced maximally in late stationary phase, suggesting that this intercellular signal was not involved in sensing cell density. These exciting results imply an expanded role for P. aeruginosa cell-to-cell signaling and suggest a revised model of P. aeruginosa quorum sensing.
The bacterial strains and plasmids utilized in this study are listed in Table Table1.1. For each experiment, strains were plated from 10% milk stocks maintained at −70°C. All P. aeruginosa strains were grown in PTSB (5% peptone, 0.25% tryptic soy broth, pH 7.0) (14) supplemented with 200 μg of carbenicillin per ml when appropriate. All cultures were grown at 37°C, and liquid cultures were shaken at 270 rpm. Plasmids were transformed into P. aeruginosa by electroporation (24).
PQS bioassays were performed as described previously (22). Briefly, cells from overnight cultures were harvested by centrifugation and resuspended in PTSB. These washed cells were used to inoculate fresh medium to a starting A660 of 0.05. At mid-log phase, subcultures were adjusted (using washed cells) to an A660 of 0.02 in fresh medium and added to tubes containing dried signal molecules or extracts. Synthetic PQS was used in all bioassays except those performed with the extracts from the growth curve experiments. The cultures were grown for 18 h and assayed in duplicate for β-galactosidase (β-Gal) activity as described by Miller (10).
Freshly plated cells of P. aeruginosa strain PAO-JP2(pECP39) were inoculated into 10 ml of medium and grown overnight. Subcultures were started at an A660 of 0.05 and grown until they reached the log phase of growth. At that time, the cells were washed with PTSB and used to adjust 400 ml of medium to an A660 of 0.05. At specified time points over the next 48 h, a 10-ml aliquot was removed for A660 determinations and ethyl acetate extractions. Aliquots were centrifuged at 10,000 × g for 10 min at room temperature, and the supernatant was extracted twice with equal volumes of ethyl acetate (22). Water was removed from the extracts by the addition of sodium sulfate followed by evaporation of the solvent in a rotary evaporator. The dried extracts were stored at −20°C until being assayed for PQS activity as described previously, using the bioassay strain P. aeruginosa PAO-R1(pTS400) (22). This experiment was repeated four times, and results are presented as the percentage of the maximal activation seen during each separate experiment.
Washed cells from late-log-phase cultures of P. aeruginosa strains PAO-JP2(pECP39) and PAO1 were used to start a 35-ml subculture at an A660 of 0.05. At specified intervals, 1 ml of each culture was removed and the A660 was determined. Serial dilutions were spread on duplicate plates which were incubated overnight at 37°C. Colonies on plates with between 30 and 300 colonies were counted, and these data were used to produce viability curves. Viability curves were repeated at least twice, and results are presented as the mean CFU/ml ± ςn − 1.
Previous research has shown that PQS was capable of inducing lasB expression in P. aeruginosa (22). During the initial characterization of PQS, it was also suggested that active LasR protein is required for PQS production and that RhlR is required for PQS bioactivity (22). The latter conclusion was based on the fact that exogenous PQS induced a lasB′-lacZ fusion in a lasR mutant but not in a lasR rhlR double mutant. This inability to activate lasB in the absence of RhlR suggested that PQS may be acting directly or indirectly through rhl quorum sensing. While PQS appeared to be an integral part of the P. aeruginosa quorum-sensing hierarchy, its exact role in cell-to-cell signaling was not apparent. To determine if PQS was affecting the components of rhl or las quorum sensing, we monitored β-Gal activity in strain PAO-R1 (lasR) containing rhlI′-lacZ (pLPRI), rhlR′-lacZ (pPCS1002), lasI′-lacZ (pPCS223), or lasR′-lacZ (pPCS1001) reporter fusions. (The lasR mutant strain PAO-R1 was used because it does not produce PQS , which allows the effects of exogenously added PQS to be determined.) These experiments produced a very interesting result. We discovered that PQS caused a major induction of rhlI′-lacZ in strain PAO-R1(pLPRI) (Fig. (Fig.1).1). The addition of 50 μM PQS to this strain led to the production of 48,428 ± 7,610 Miller units of β-Gal activity, a notable increase over the 2,380 ± 611 Miller units of β-Gal activity produced in the absence of PQS. In contrast, the lasI′-lacZ fusion was not affected by exogenously added PQS, and as expected, β-Gal expression from strain PAO-R1 containing the vector control, pLP170, was unaffected by the addition of PQS (Fig. (Fig.1).1). This suggested that the production of the C4-HSL signal, but not the 3-oxo-C12-HSL signal, was positively regulated by PQS. It should be noted here that the induction of rhlI caused by the addition of PQS to strain PAO-R1(pLPRI) shown in Fig. Fig.11 (48,428 ± 7,610 Miller units) does not fully complement rhlI′-lacZ expression to the level seen in the wild-type strain containing pLPRI (139,797 ± 10,291 Miller units [see Table Table2]).2]). This indicates that either LasR or a LasR-controlled factor (in addition to PQS) is required for complete induction of rhlI in P. aeruginosa. We also found that the expression of lasR′-lacZ and rhlR′-lacZ increased approximately twofold in the presence of PQS (Fig. (Fig.1).1). This was interesting because the activation of lasR by a LasR-controlled factor such as PQS indicates that PQS may be part of a positive feedback loop within the quorum-sensing hierarchy. However, we must point out that both lasR and rhlR are partially expressed in the absence of PQS and LasR (Fig. (Fig.1),1), demonstrating that other factors are important for their expression.
The results presented in Fig. Fig.11 show that the addition of PQS had the greatest effect on the expression of rhlI compared to the other quorum-sensing system genes. It has previously been reported that rhlI is positively regulated by both C4-HSL–RhlR and 3-oxo-C12-HSL–LasR (9). To learn more about the factors affecting rhlI induction, we monitored the expression of rhlI′-lacZ from plasmid pLPRI contained in the P. aeruginosa wild-type strain PAO1 and the defined mutant strains PDO100 (rhlI) and PAO-R1 (lasR). In the presence or absence of C4-HSL and/or PQS, the expression of rhlI′-lacZ in strain PDO100(pLPRI) was very similar to that in the wild-type strain (Table (Table2).2). This indicated that both C4-HSL and PQS did not affect rhlI expression in an rhlI mutant, implying that elements capable of inducing full rhlI expression were present in strain PDO100 (rhlI). However, examination of the expression of rhlI′-lacZ in the lasR mutant strain PAO-R1 indicated that both C4-HSL and PQS are important regulators of rhlI. The addition of 5 μM C4-HSL or 50 μM PQS to strain PAO-R1(pLPRI) caused the expression of rhlI′-lacZ to be 20 or 35%, respectively, of that seen in strain PAO1(pLPRI) (Table (Table2).2). Most interestingly, the addition of both C4-HSL and PQS caused rhlI′-lacZ to be expressed at approximately the wild-type level, indicating that these signals cooperatively effect rhlI expression. Taken together, the data presented in Fig. Fig.11 and Table Table22 indicate that rhlI expression falls under the control of both PQS and C4-HSL. In addition, these data confirm the results of Latifi et al. (9), who reported that both the las and rhl quorum-sensing systems affect the expression of rhlI.
Previous data have shown that three intercellular signals, 3-oxo-C12-HSL, C4-HSL, and PQS, are each capable of activating lasB in P. aeruginosa (15, 17, 22). The fact that these signals are not produced independently of each other has made it difficult to determine the role of each individual signal with regard to lasB activation. Research focused on the role of PQS in lasB expression had indicated that lasB induction by PQS required a functional rhl quorum-sensing system (22). This led us to speculate that PQS and C4-HSL were functioning in a cooperative manner to activate lasB. In order to determine whether C4-HSL and PQS have cooperative effects on lasB expression, the activation of lasB′-lacZ in strains PDO100 and PAO-R1 was monitored in the presence of PQS and/or C4-HSL. In strain PAO-R1, PQS caused lasB′-lacZ to be induced to 28% of the wild-type level seen in strain PAO1(pTS400) (Table (Table3).3). The addition of C4-HSL alone had a minor effect on lasB′-lacZ in this strain, but the addition of C4-HSL and PQS together caused lasB′-lacZ to be induced to 58% of the wild-type level (Table (Table3).3). This indicated that these two signals had a cooperative effect on lasB induction. A similar effect was seen in strain PDO100(pTS400), where the lack of a functional rhl quorum-sensing system caused lasB′-lacZ to be expressed at 36% of the wild-type level seen in strain PAO1(pTS400) (Table (Table3).3). The addition of PQS alone caused an increase of lasB′-lacZ expression to 64% of the wild-type level, and the addition of C4-HSL alone restored lasB′-lacZ expression to the expected wild-type level (Table (Table3).3). Addition of PQS and C4-HSL together had an additive effect on the induction of lasB′-lacZ, which was expressed at approximately twice the wild-type level seen in strain PAO1(pTS400).
The production of the C4-HSL and 3-oxo-C12-HSL signals is initiated during the log phase of growth. Pearson et al. (16) purified 3-oxo-C12-HSL from late-log-phase cultures of strain PAO1, and C4-HSL has been isolated from extracts of early-stationary-phase cultures of strain PAO1 (17). It has also been shown that lasR and rhlR are induced during the last half of log-phase growth (20). These data combined with the fact that the production of PQS required active LasR led us to speculate that PQS would be produced during the late log phase of growth. In order to learn more about the production of PQS, we monitored its synthesis throughout the growth cycle. The PQS-producing strain PAO-JP2(pECP39) (22) was grown in PTSB for 48 h, and crude PQS was recovered from aliquots removed at specific intervals.
The results presented in Fig. Fig.22 indicate that PQS was not found in strain PAO-JP2(pECP39) culture medium until the end of the log phase of growth. The maximal amount of this signal was found in cultures that were in the late stationary phase of growth (30 to 42 h after inoculation), and the concentration decreased after 48 h of growth (Fig. (Fig.2A).2A). These results showed that PQS was produced maximally long after the las and rhl quorum-sensing systems had been activated, which suggested that its production required a factor, in addition to the las quorum-sensing system, that is not available until the stationary phase of growth. These results are especially interesting when one considers that the genes controlled by the las quorum-sensing system will be induced in strain PAO-JP2(pECP39) because the autoinducer-independent LasR encoded by pECP39 is expressed constitutively (data not shown). The determination of CFU and A660 for cultures of strain PAO-JP2(pECP39) indicated that bacteria were viable and in the stationary phase of growth from approximately 8 to 48 h after inoculation (Fig. (Fig.2B).2B). The CFU and optical density of strain PAO-JP2(pECP39) were very similar throughout the growth curve to those of the wild-type strain PAO1 (Fig. (Fig.2B),2B), indicating that strain PAO-JP2(pECP39) was growing normally.
We have discovered that the intercellular signal PQS controls the rhl quorum-sensing system at a transcriptional level. PQS was found to have a major positive effect on rhlI induction and lesser positive effects on the induction of rhlR and lasR (Fig. (Fig.1).1). These results, along with previous data which showed that PQS production required an active LasR protein (22), indicate that PQS may act as an additional connecting signal between the las and rhl quorum-sensing systems by transcriptionally regulating the autoinducer synthase gene, rhlI. This would be the third link between these two systems. LasR and 3-oxo-C12-HSL were shown to positively regulate the transcription of rhlR (9, 20) and rhlI (9) in P. aeruginosa and Escherichia coli, and 3-oxo-C12-HSL posttranslationally affected RhlR by blocking the interaction of C4-HSL and RhlR in E. coli (20). The purpose of PQS acting as an additional linking signal between the two quorum-sensing systems is not known. Speculatively, the upregulation of rhlI by PQS (which implies a subsequent increase in C4-HSL production) may be necessary to overcome the inhibitory effect of 3-oxo-C12-HSL on the rhl quorum-sensing system. It is also interesting that Pearson et al. (19) showed that 3-oxo-C12-HSL accumulated inside E. coli cells, while C4-HSL freely diffused to reach an internal-external equilibrium. This could lead to an imbalance of autoinducer concentration that would require the upregulation of rhlI in order to provide enough intracellular C4-HSL to compete with 3-oxo-C12-HSL for binding to RhlR.
Another interesting finding with regard to rhlI regulation is that it was induced by PQS or C4-HSL in our lasR mutant, but neither signal had an effect in our rhlI mutant (Table (Table2).2). Additionally, when added together PQS and C4-HSL had a cooperative effect on rhlI induction in strain PAO-R1 (lasR) (Table (Table2).2). The reason that rhlI expression was not affected in strain PDO100 (rhlI) is not clear, but there is a plausible explanation. Since both PQS and the las quorum-sensing system have been shown to regulate rhlI (Fig. (Fig.1)1) (9), it is most likely that these elements are able to complement the rhlI mutation with regard to rhlI′-lacZ expression. Whether this theory is correct remains to be proven, but it is apparent that rhlI regulation is quite complex and probably involves multiple layers of control, perhaps working at different stages of growth (see below).
We also learned that PQS and C4-HSL cooperatively induce lasB. Previous reports have shown that lasB transcription is positively regulated by all three P. aeruginosa intercellular signals (15, 17, 22). In addition, we knew that the induction of lasB by PQS required a functional rhl quorum-sensing system (22). Therefore, we examined the effect of adding both PQS and C4-HSL to strains PAO-R1 and PDO100 containing pTS400 (Table (Table3).3). In both strains, PQS and C4-HSL had a cooperative effect on the induction of lasB. In strain PAO-R1(pTS400) the two signals had a synergistic effect on lasB induction, and in strain PDO100(pTS400) the signals' effect was additive. The reason for this difference is not clear, but in either case, lasB′-lacZ was induced to a greater level in the presence of both signals than with either individual signal. This indicates that the induction of lasB requires a complex chain of events that we are continuing to learn about.
Finally, our analysis of PQS production indicated that this signal was produced much later in the growth cycle than a typical quorum-sensing signal. The concentration of PQS in culture medium was negligible in the late log phase of growth and was at a maximum late in the stationary phase of growth (Fig. (Fig.2).2). PQS was most abundant between 30 and 42 h of growth and had decreased after 48 h of growth (Fig. (Fig.2).2). This indicated that strain PAO-JP2(pECP39) either made less PQS after 48 h of growth or produced a factor capable of degrading PQS at that time. Nevertheless, it is apparent that PQS is not involved in sensing cell density, because the signal is produced at a time after cell density has become stable. This leads one to ponder an interesting question. Given that C4-HSL is produced during the log phase of growth and PQS is not produced until late in the stationary phase of growth, then what is the purpose of PQS inducing rhlI? We speculate that PQS may induce rhlI in order to further upregulate the rhl quorum-sensing system. This could be beneficial to the organism, since the rhl quorum-sensing system has been shown to control the production of elastase, alkaline protease, and the biosurfactant rhamnolipid (2, 8, 12, 17). The production of proteases at a later stage of growth could help increase the availability of nutrients during infections because of the tissue-destroying capabilites of these enzymes. At the same time, rhamnolipid helps cells to utilize long-chain fatty acids as sources of carbon (11), which would benefit bacteria that have depleted available nutrients. These theories lead us to conclude that P. aeruginosa may use PQS as a signal to induce genes controlled by the rhl quorum-sensing system in order to respond to stress encountered by late-stationary-phase cultures.
As more is learned about the quorum-sensing hierarchy of P. aeruginosa, it becomes clear that this is an extremely complex signal transduction pathway. To help clarify the circuitry of these interrelated systems, we have provided a schematic diagram that summarizes our present level of understanding (Fig. (Fig.3).3). While the effects of PQS are intriguing, we are only beginning to understand how this intercellular signal fits in the grand scheme of P. aeruginosa quorum sensing. The ability of PQS to induce an important virulence factor gene (lasB) and genes for quorum-sensing components (rhlI, rhlR, and lasR) indicates that understanding its role in quorum sensing will be a key to determining its effect on virulence.
This work was supported by research grants from the Cystic Fibrosis Foundation (grant PESCI99I0 to E. C. Pesci), the American Lung Association of North Carolina (to E. C. Pesci), the North Carolina Biotechnology Center (to E. C. Pesci), and the NIH (grant R01-AI33713 to B. H. Iglewski).
We thank C. J. Smith, T. deKievit, R. Smith, M. W. Calfee, E. Batten, and C. S. Pesci for help in manuscript preparation and thoughtful insight.