Albumin Inhibits Pseudomonas aeruginosa Quorum Sensing and Alters Polymicrobial Interactions

ABSTRACT

Polymicrobial interactions are complex and can influence the course of an infection, as is the case when two or more species exhibit a synergism that produces a disease state not seen with any of the individual species alone. Cell-to-cell signaling is key to many of these interactions, but little is understood about how the host environment influences polymicrobial interactions or signaling between bacteria. Chronic wounds are typically polymicrobial, with Staphylococcus aureus and Pseudomonas aeruginosa being the two most commonly isolated species. While P. aeruginosa readily kills S. aureus in vitro, the two species can coexist for long periods together in chronic wound infections. In this study, we investigated the ability of components of the wound environment to modulate interactions between P. aeruginosa and S. aureus. We demonstrate that P. aeruginosa quorum sensing is inhibited by physiological levels of serum albumin, which appears to bind and sequester some homoserine lactone quorum signals, resulting in the inability of P. aeruginosa to produce virulence factors that kill S. aureus. These data could provide important clues regarding the virulence of P. aeruginosa in albumin-depleted versus albumin-rich infection sites and an understanding of the nature of friendly versus antagonistic interactions between P. aeruginosa and S. aureus.

INTRODUCTION

Interactions between microbes are complex and range from fierce competition for nutrients and niches, manifested by antagonistic behavior, to highly evolved cooperative mechanisms between different species that support their mutual growth in specific environments (1). These interactions influence not only which species live or die but also the course of an infection. A growing appreciation of the importance of polymicrobial interactions during infection has resulted in increased efforts to develop representative in vitro and in vivo infection models with which to elucidate these mechanisms of interaction.

Our group has focused on understanding the interactions between microbial species in wound infections (2–5), which are typically polymicrobial, associated with the presence of biofilms, and exhibit increased tolerance to antimicrobials (6, 7). Staphylococcus aureus and Pseudomonas aeruginosa are the two most commonly isolated species from chronic wounds, many of which are coinfected with both species (8–10). There is also evidence that patients with P. aeruginosa-S. aureus coinfected wounds have worse outcomes than those with P. aeruginosa or S. aureus monospecies infections (11–13). Despite clinical evidence that S. aureus and P. aeruginosa coinfect chronic wounds, there have been few studies investigating their interspecies interactions, which may be partially explained by the technical difficulty of growing them together in vitro.

Under standard microbiological culturing conditions (planktonic growth in complex liquid medium), S. aureus is typically eradicated from the coculture within 8 h, presumably due to killing by P. aeruginosa (5). P. aeruginosa produces multiple exoproducts that inhibit the growth of, or kill, S. aureus, such as the LasA protease (14, 15), 4-hydroxy-2-heptylquinoline-N-oxide (HQNO) (16), pel and psl products (17), and phenazines such as pyocyanin (18). While we have a good understanding of the mechanisms controlling the regulation and expression of these P. aeruginosa virulence factors, little is known about how the composition of the bacterial environment influences their production.

We previously described an in vitro wound-like model that prevented P. aeruginosa from killing S. aureus and allowed the two species to persist in coculture for up to 7 days, similar to coinfections seen in mouse wounds (5). This model was used to demonstrate synergism in regard to antibiotic tolerance when S. aureus and P. aeruginosa were grown together in a wound-like environment (5). However, one intriguing facet of the study that was not elucidated was how the wound environment modulates the interactions between S. aureus and P. aeruginosa to promote coinfection. Therefore, in this study we sought to understand how components of the wound environment alter the ability of P. aeruginosa to kill S. aureus in coculture.

RESULTS AND DISCUSSION

Albumin prevents P. aeruginosa from killing S. aureus in the in vitro wound environment.

Our group (5), in addition to others (11, 19, 20), have noted the ability of Pseudomonas aeruginosa to rapidly kill Staphylococcus aureus in standard laboratory cocultures. This is an active killing process that occurs as P. aeruginosa approaches stationary phase and is dependent on quorum sensing (QS). Killing is conserved across a wide range of isolates and growth conditions (see Fig. S1 in the supplemental material); however, our group recently reported that P. aeruginosa and S. aureus can be stably cocultured in vitro using a “wound-like” medium (WLM) that contains many of the same components found in the wound environment (5). In order to determine the specific medium components required for stable coculture, we systematically tested the components of WLM.

WLM components were tested in the same concentrations used in WLM, which includes 45% Bolton broth, 50% adult bovine plasma, and 5% laked horse red blood cells (RBC). Bolton broth alone was not able to rescue S. aureus in coculture (Fig. 1A). This result is not surprising since Bolton broth is a rich, undefined complex medium, similar to Luria-Bertani (LB) broth, which also cannot rescue S. aureus. A marginal rescue effect was observed when Bolton broth was supplemented with RBC (Fig. 1A), but it was not nearly as robust as that seen with whole WLM. Coculturing S. aureus and P. aeruginosa in Bolton broth supplemented with adult bovine plasma dramatically increased the numbers of S. aureus bacteria that survived coculture with P. aeruginosa, similar to the results seen with whole WLM (Fig. 1A). This effect was also observed when LB broth was supplemented with plasma. These data demonstrate that the plasma fraction of the WLM is largely responsible for preventing P. aeruginosa from killing S. aureus in WLM coculture.

FIG 1.

The albumin component of WLM rescues S. aureus (SA) from eradication when it is in coculture with P. aeruginosa (PA). (A) PAO1-SA31 cocultures were grown overnight in Luria-Bertani (LB) broth, full WLM (wound-like medium), or specific components of WLM, including Bolton broth (BB), Bolton broth plus red blood cells (BB+RBC) or plasma, or Bolton broth plus heat-denatured plasma. (B) PAO1-SA31 cocultures were grown overnight in LB broth, full WLM, or LB broth supplemented with the indicated components. Overnight cultures were then serially diluted and plated on Pseudomonas and Staphylococcus isolation agar to determine the number of CFU per milliliter. Plasma was heat denatured by heating whole plasma to 80°C for 15 min. Albumin was depleted from plasma using DEAE Affi-Gel Blue gel columns (Bio-Rad) according to the manufacturer's recommendations. Experiments were repeated at least three times, with at least three technical replicates.

We next investigated which component of plasma allowed S. aureus to coexist with P. aeruginosa in coculture. When plasma was heat denatured (21), the rescue effect was lost (Fig. 1A), suggesting a role for plasma proteins. Serum is the component of plasma generated when clotting factors and immune cells are removed while proteins and other molecules present in whole blood are retained (22). At physiologically relevant concentrations of serum, there was a stepwise increase in the amount of S. aureus rescued in coculture with P. aeruginosa (Fig. 1B), indicating that the protective factor was present in serum. As bovine serum albumin (BSA) is the most abundant protein found in serum, at about 65%, and constitutes 5% of the total blood volume (23, 24), we reasoned that this protein may be responsible for the protective effect. Addition of BSA at physiological concentrations caused a stepwise increase in the amount of S. aureus that remained viable in cocultures (Fig. 1B). In addition, depleting albumin from plasma resulted in a loss of the rescue effect, and P. aeruginosa regained the ability to kill S. aureus (Fig. 1B). Importantly, significant inhibition of S. aureus killing was not seen when S. aureus and P. aeruginosa were cocultured in LB broth supplemented with other serum proteins, such as gamma globulins and fibrinogen (Fig. S2), suggesting that the inhibitory effect observed was not simply due to the addition of excess protein and implying that the mechanism of action is albumin specific. Human serum albumin (HSA) also rescued S. aureus from killing by P. aeruginosa at a level similar to that of BSA, as did Albuminar (CSL Behring), which contains 25% human albumin in normal saline and is a common commercial albumin replacement therapy administered to patients suffering from hypoalbuminemia (Fig. S3). Taken together, these results support a role for albumin in preventing loss of S. aureus from in vitro WLM cocultures.

The ability of P. aeruginosa to kill S. aureus is quorum sensing controlled and does not require cell-to-cell contact.

In order to elucidate how albumin prevents P. aeruginosa from killing S. aureus in the presence of albumin, we sought to determine the mechanism of S. aureus killing by P. aeruginosa in coculture. P. aeruginosa produces many virulent exoproducts, many of which have been shown to kill or inhibit the growth of S. aureus (14–18). P. aeruginosa exoproducts can be either secreted into the extracellular environment or delivered in a contact-dependent manner (25–27). Thus, we next sought to determine whether cell-to-cell contact was required for S. aureus eradication in S. aureus-P. aeruginosa cocultures.

To test whether P. aeruginosa requires cell-to-cell contact to kill S. aureus, cocultures were grown in transwells, with the two species separated by a 0.4-μm-pore-size filter. This experimental setup allows the bacterial species to grow separately, without the ability to physically interact, while still allowing secreted exoproducts to move freely through the membrane. We observed that S. aureus was similarly eradicated from coculture when the two species were grown mixed in both compartments or when they were separated in different compartments by the membrane (Fig. S4). These data indicate that P. aeruginosa is able to kill S. aureus in coculture independent of direct cell-to-cell contact, suggesting that some P. aeruginosa-secreted exoproduct(s) must be responsible for the eradication of S. aureus from coculture.

The P. aeruginosa quorum sensing systems affect expression of over 300 genes either directly or indirectly, controlling up to 6% of the total P. aeruginosa genome (28). P. aeruginosa possesses two prototypical acyl-homoserine lactone (AHL) QS systems, termed Las and Rhl. These systems function through production of diffusible signal molecules by the enzymes LasI and RhlI, which are subsequently sensed by the cognate transcriptional regulator proteins LasR and RhlR. These systems do not function independently but are arranged hierarchically, with Las controlling expression of Rhl (29). The production of many P. aeruginosa-secreted exoproducts known to inhibit or kill S. aureus is controlled by QS (14–18), including the Las system-controlled staphylolytic enzyme LasA and the Rhl system-controlled rhamnolipid biosurfactant, hydrogen cyanide, and pyocyanin (29). P. aeruginosa also possesses a third QS system termed PQS (Pseudomonas quinolone signal). The PQS system directs the synthesis of a quinolone-based autoinducer, 2-heptyl-3-hydroxy-4-quinolone (HHQ), which is sensed by the transcriptional regulator PqsR (MvfR) (30). PQS production and sensing require activation by Las and Rhl systems. As with the Las and Rhl systems, the PQS system controls production of several antimicrobials, including pyocyanin and 4-hydroxy-2-heptylquinoline N-oxide (HQNO) (31).

To evaluate the contribution of P. aeruginosa QS to S. aureus eradication, we cocultured S. aureus with a panel of P. aeruginosa strains that have mutations in different genes in the QS regulatory hierarchy. We observed that P. aeruginosa strains with mutations that control the Las and Rhl QS systems (PAO1 Δvfr), with mutations in the Las system (PAO1 ΔlasI rhlI, PAO1 ΔlasI, and PAO1 ΔlasR), or with mutations in the PQS system were unable to kill S. aureus in coculture (Fig. 2). However, mutations in the Rhl system (PAO1 ΔrhlI and PAO1Δ rhlR) did not affect the ability of P. aeruginosa to kill S. aureus in coculture. As the Las system is known to regulate the production of the staphylolytic LasA protease (14, 15), we tested whether a PAO1 strain unable to produce LasA could kill S. aureus. Surprisingly, we found that the ability of PAO1 ΔlasA to kill S. aureus was equivalent to that of PAO1. This indicates that in this coculture system LasA is not a significant contributor to S. aureus killing.

FIG 2.

P. aeruginosa mutants with defects high in the QS regulatory system and in the PQS system are unable to kill S. aureus. A total of 10⁴ CFU of the indicated strains was inoculated into LB medium and grown overnight with SA31. Cocultures were initiated as a 1:1 mix of P. aeruginosa and S. aureus. Samples were then taken for serial dilution and plating on Pseudomonas and Staphylococcus isolation agar to determine the number of CFU per milliliter. Experiments were repeated at least three times, with at least three technical replicates.

Based on these results, we hypothesized that albumin was inhibiting P. aeruginosa killing of S. aureus via inhibition of QS. To test this hypothesis, we first measured the impact of albumin addition on the expression of several QS-controlled genes. While albumin did not affect the growth of P. aeruginosa (Fig. S5), the expression levels of the Las-controlled genes lasI, lasB, rhlI, and rsaL were all significantly decreased by the addition of albumin (Fig. 3A). As expected, albumin also decreased production of the Las-controlled extracellular products LasA and LasB (Fig. 3B). Surprisingly, the production of pyocyanin, a redox-active small molecule primarily controlled by the PQS QS system, was increased (Fig. 3D). Taken together, these data demonstrated that albumin impacts P. aeruginosa QS; however, while the Las and Rhl systems are repressed by albumin, the PQS system was stimulated. Our results are consistent with those of Kruzcek et al. (32), who reported that pyocyanin and PQS were increased in the presence of serum at late stages of P. aeruginosa growth. Thus, while pqsA mutants are attenuated in their ability to kill S. aureus, this effect does not appear to be due to a reduction in pyocyanin or the expression of pqsA.

FIG 3.

Albumin inhibits P. aeruginosa QS. (A) P. aeruginosa GFP reporters were grown for 24 h in LB broth with and without 5% BSA, monitored for their level of GFP fluorescence, and normalized to growth (OD₆₀₀). In the presence of BSA, the expression of most P. aeruginosa QS-controlled genes was significantly reduced in comparison to growth in LB alone. Activities of the Las system-regulated exoproducts LasA (B) and LasB (C) and the PQS system-regulated pyocyanin (D) were measured from the supernatants of overnight cultures of PAO1 grown in LB medium with or without 5% BSA. The QS mutant JP2 (PAO1 ΔlasI rhlI) supernatant was utilized as a negative control. *, P < 0.01; **, P < 0.001; ***, P < 0.0001 (one-way analysis of variance, Tukey posttest). Experiments were repeated at least three times, with at least three technical replicates.

It should also be noted that we also tested P. aeruginosa strains with mutations in other genes, including lasB, pilA, pel and psl, narGH, phzA, phzE, pcrV, pchE, algD, and others, and they were not attenuated in their ability to kill S. aureus in planktonic coculture (data not shown). In addition we tested several purified exoproducts and signals for their ability to kill S. aureus (without P. aeruginosa), including 2-heptyl-4-quinolone (HHQ), HQNO, PQS, and the Las- and Rhl-associated QS signals, and we saw no effect on S. aureus growth at the concentrations tested. So, despite great effort, we were not able to pinpoint the specific exoproduct(s) responsible for S. aureus killing in our system. We think that it is likely to be a combination of exoproducts, possibly at specific concentrations or produced at specific times during growth. Importantly, while the specific S. aureus-killing products were not identified, it is clear from our data that they are QS controlled.

Albumin binds and sequesters P. aeruginosa QS molecules.

Albumin is a common serum transport protein that carries numerous metabolites, hormones, and fatty acids throughout the body (23). The protein consists of three homologous domains, with nine binding sites that can bind a variety of molecules with different affinities (33). Examples of known albumin ligands include cholesterol, ibuprofen, and fatty acids (34–36). When we compared the structure of the P. aeruginosa QS signaling molecules to that of known BSA ligands, we noticed a great deal of structural similarity. Since albumin is known to nonspecifically bind to an array of molecules present in the physiological environment, we hypothesized that serum albumin can inhibit P. aeruginosa QS by binding and sequestering QS signaling molecules. This would effectively remove them from the environment, and therefore P. aeruginosa QS gene expression would be decreased.

To test the hypothesis that albumin inhibits P. aeruginosa QS by binding and sequestering autoinducers, we incubated purified N-3-oxo-dodecanoyl homoserine lactone (3OC12-HSL) and N-butanoyl homoserine lactone (C4-HSL) (the cognate signaling molecules of the LasI/R and RhlI/R QS systems, respectively) with phosphate-buffered saline (PBS) alone or PBS supplemented with 5% BSA for 3 h. The solutions were then passed through a 30-kDa-molecular-mass-cutoff filter to remove the BSA and anything bound to it. Any QS molecules remaining in the eluate were then extracted with ethyl acetate and detected using luminescent bioreporters as described in Materials and Methods. We observed a significant, 6.25-fold, decrease in the amount of 3OC12-HSL, but not C4-HSL, in the eluates of solutions that were incubated with BSA prior to column separation (Fig. 4A).

FIG 4.

BSA binds 3OC12-HSL and its degradation product C12-TA-HSL. 3OC12-HSL, C12-TA-HSL, C4-HSL, PQS, HHQ, or HQNO was added to PBS or PBS–5% BSA at a concentration of 5 μM. After a 3-h incubation the solutions were passed through a 30-kDa column to remove the BSA and any of its ligands. Autoinducers were extracted from the eluates, and their levels were measured using LasR, RhlR, or PQS-based luciferase bioreporter strains as described in Materials and Methods. ***, P < 0.0001 (one-way analysis of variance, Tukey posttest). Experiments were repeated at least three times, with at least three technical replicates.

We also examined the ability of BSA to bind C12-TA-HSL (where TA is tetramic acid), which is a degradation product of 3OC12-HSL. C12-TA-HSL is generated by nonenzymatic deprotonation and has been detected in P. aeruginosa cultures and in sputum from patients infected with P. aeruginosa (37, 38). C12-TA-HSL has bactericidal activity and can also act as a siderophore (37). In our experiments, C12-TA-HSL activated our LasR-based bioreporter just as well, if not better than, 3OC12-HSL. This suggests that C12-TA-HSL could physiologically activate the Las QS system in addition to 3OC12-HSL. We detected a significant, 10.4-fold decrease in the amount of C12-TA-HSL, similar to level of 3OC12-HSL, in eluates from solutions incubated with BSA. We saw no evidence of albumin binding to the quinolone-type autoinducers PQS, HHQ, and HQNO (Fig. 4). Taken together, these results indicate that albumin decreases expression of Las QS genes through sequestration of 3OC12-HSL.

To determine the affinity of QS molecules to albumin, we used isothermal titration calorimetry (ITC), which has been used extensively to study albumin and fatty acid binding (39–42). With ITC, affinities are determined by directly measuring the binding of a ligand to a protein by recording the heat absorbed or released by the binding process. ITC with 3OC12-HSL and BSA resulted in an oscillating thermogram (Fig. 5), which is indicative of multiple complex binding events. This implies that one ligand binds more strongly, and as the particular binding site involved becomes saturated or the ligand becomes exhausted, a second, less strongly bound ligand begins to bind. While it is clear from our ITC measurements that 3OC12-HSL binds BSA, computing accurate binding affinity (K_d) values for multiple binding events of many different affinities would not be possible.

FIG 5.

ITC thermograph demonstrates the binding of 3OC12-HSL to BSA. 3OC12-HSL (2 mM) in 5% ethyl acetate was utilized as a ligand for measuring the binding affinity to 30 μM fatty-acid-free BSA in 5% ethyl acetate protein by ITC as described in Materials and Methods. Qi, heat effect.

Our observations correlate with the findings of Davis et al. (43), who also proposed that albumin may act as “a sink or perhaps a reservoir for AHLs.” This group demonstrated that, in the presence of HSA, the critical micelle concentration (the amount of AHL or HSL needed to self-aggregate into a micelle) for 3OC14-HSL increased from 17 μM to over 500 μM, suggesting that albumin effectively reduced the concentration of free AHLs in the medium. A similar mechanism has also been described in S. aureus, where human apolipoprotein B binds to the S. aureus autoinducing pheromone and prevents its attachment to bacteria and subsequent signaling through its receptor, AgrC (44). Mice deficient in plasma apolipoprotein B were significantly more susceptible to invasive agr⁺ strain infection but not to infection with an agr deletion mutant. Therefore, the authors proposed that apolipoprotein B at homeostatic levels in blood is an essential innate defense effector against invasive S. aureus infection.

Potential clinical significance of our findings and future studies.

If albumin is capable of binding and sequestering P. aeruginosa HSLs, this could have important clinical implications during P. aeruginosa infections. Albumin concentrations differ depending on anatomical location. For example, body sites rich in blood, such as wounds, are rich in albumin. Based on our data, it is possible that when P. aeruginosa is in an albumin-rich environment, such as in chronic wounds, QS will be downregulated, or even nonfunctional, due to the binding of QS molecules by serum albumin. Conversely, there are body sites that have reduced levels of albumin, such as the urinary tract, lung, and the peripheral blood during burn trauma (45, 46). In these types of infections, P. aeruginosa is more likely to become the predominant microbe (47, 48). We propose that the reduced levels of albumin in these environments allow the P. aeruginosa QS to function, thus enhancing P. aeruginosa fitness through production of Las-controlled antimicrobials that eliminate coinfecting bacteria. This may be particularly relevant to burn patients who have extremely low levels of albumin in the infection site (46, 49, 50), and P. aeruginosa is generally the predominant pathogen. P. aeruginosa is highly virulent in burns, quickly disseminating and causing sepsis and septic shock, which can be lethal (51), and the contribution of QS has been shown to be paramount for virulence and dissemination in animal burn models (52, 53).

While our study represents a first step in understanding both how specific components present in the infection site can modulate cell-to-cell signaling and impact polymicrobial interactions, there are still many questions to answer. One major limitation of this study was that our observations were largely based on planktonic growth, which may not correlate with the biofilm growth occurring in wounds. Thus, important future studies will be to investigate the effect of albumin on QS in P. aeruginosa biofilms, both in vitro and in vivo. We were also unable to determine a specific binding affinity (K_d) for HSL-albumin. This is critical to understanding the in vivo relevance of our findings as albumin binds many small molecules, such as free fatty acids or steroid-based ligands, that will likely compete with AHLs for binding. Finally, while P. aeruginosa QS has been demonstrated to be an important factor controlling virulence in burn wounds (52, 53), the same is not true for chronic wounds. Based on our in vitro findings, we would predict that P. aeruginosa QS gene expression is higher in the albumin-depleted burn wound than in the albumin-rich chronic wound. If this is true, it is possible that P. aeruginosa virulence can be attenuated by the administration of exogenous albumin given locally at the burn wound. This finding would also be extremely relevant to investigators who are developing P. aeruginosa QS inhibitors for the treatment of chronic wound infections, which may ultimately prove ineffective. Thus, our future studies will focus on experiments conducted in vivo to characterize the interaction of albumin (and other host proteins) with P. aeruginosa quorum signals.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Bacterial strains used in this study are listed in Table 1. P. aeruginosa and S. aureus clinical isolates were provided by Abdul Hamood (54) and obtained from patients at the Texas Tech University Health Sciences Center's affiliated hospital, University Medical Center, under an approved IRB protocol. S. aureus and P. aeruginosa planktonic cultures were grown in Luria-Bertani (LB) broth or LB broth supplemented as specified in the figure legends. The in vitro wound-like medium (WLM) used was previously described by DeLeon et al. (5). WLM consists of 45% Bolton broth (complex meat-based broth), 50% bovine plasma, and 5% laked horse red blood cells (RBC). A 5-ml volume of medium was inoculated with 10⁴ to 10⁵ cells from overnight cultures of S. aureus, P. aeruginosa, or a 1:1 mixture of both and grown, shaking at 220 rpm, at 37°C overnight. Samples were then serially diluted, and plated on Pseudomonas isolation agar to determine the number of CFU of P. aeruginosa per milliliter and/or on mannitol salt agar (MSA) or Staphylococcus isolation agar to determine the number of CFU of S. aureus per milliliter (Difco, Sparks, MD); (Remel, Thermo Fisher Scientific, Lenexa, KS).

TABLE 1.

Bacterial strains and plasmids used in this study

Strain	Description	Reference or source
P. aeruginosa PAO1	Wild type	58
P. aeruginosa PA14	Wild-type P. aeruginosa	59
S. aureus SA31	S. aureus sp. aureus strain ATCC 29213	Remel Microbiology Products
P. aeruginosa mutants
PAO Δvfr	vfr deletion of PAO-SW	60
PAO-JP2	PAO1 ΔlasI ΔrhlI mutant, Tcr Hgr	61
PAO-JM2	PAO1 ΔrhlI mutant, Tcr	62
PAO1 lasI::Gmr	PAO1 insertion mutant (Gmr) in the lasI gene	63
PAO1 lasR::Gmr	PAO1insertion mutant (Gmr) in the lasR gene	64
PA14 rhlI::Tcr	PA14 insertion mutant (Tcr) in the rhlI gene	65
PA14 rhlR::Tcr	PA14 insertion mutant (Tcr) in the rhlR gene	65
PAO1 lasA::Tcr	PAO1 insertion mutant (Tcr) in the lasA gene	66
PAO1 ΔpqsA	pqsA deletion mutant; PQS negative	67
PAO1 ΔpqsA pqsH	Double deletion mutant in PAO1	30
PAO1 ΔpqsE	pqsE in-frame deletion mutant in PAO1	68
Bioreporters
PAO1(prhlI-LVAgfp)	P. aeruginosa PAO1 carrying the rhlI-gfp transcriptional reporter on plasmid prhlI-LVAgfp	69
PAO1(plasI-LVAgfp)	P. aeruginosa PAO1 carrying the lasI-gfp transcriptional reporter on plasmid plasI-LVAgfp	69
PA14(pGJB5)	P. aeruginosa PA14 carrying the transcriptional rsaL-gfp reporter on plasmid pGJB5	70
PAO1 lasB::GFP	MH451 lac::lasR lasB::gfp(ASV), GFP reporter of the PAO1 lasB gene	71
PA14(pAW1)	P. aeruginosa PA14 carrying the pqsA-gfp transcriptional reporter on plasmid pAW1	This study
E. coli(pSB1075)a	E. coli 3OC12-HSL luminescent reporter strain	72
E. coli(p536)	E. coli C4-HSL luminescent reporter strain	73
PAO1 pqsA CTX-lux::pqsA	pqsA mutant containing a chromosomal pqsA-luxCDABE transcriptional reporter	74

^aE. coli, Escherichia coli.

DNA manipulations.

PCR was performed using an Expand Long Template PCR system (Roche). Qiagen kits were used for purification of plasmids (QIAprep spin miniprep kit), gel purifications (QIAquick gel extraction kit), and PCR purifications (MinElute PCR purification kit). Restriction endonucleases and buffers were purchased from New England BioLabs or Fermentas Life Sciences. DNA sequencing was performed at the DNA Core Facility at the University of Texas at Austin Institute for Cell and Molecular Biology.

Plasmid construction.

To construct the pqsA transcriptional reporter plasmid pAW1, the region upstream of pqsA was amplified from P. aeruginosa chromosomal DNA using primers CGGAATTCGTAGGTGTCCTCTTCGGCAG and GCTCTAGATCAATCAATCAGCGATATGCATCCGGATCAG. The resulting amplicon was digested with EcoRI/XbaI and ligated into EcoRI/XbaI-digested pGJB5 (Table 1). The resulting plasmid, pAW1, was confirmed by sequencing and transformed into P. aeruginosa by electroporation.

Staphylolytic assay.

A staphylolytic assay was performed as previously described (55). LasA protease activity was measured by determining the ability of P. aeruginosa culture supernatants to lyse S. aureus cells. Briefly, an overnight culture of S. aureus was boiled for 10 min, followed by centrifugation for 10 min at 10,000 × g. The resulting pellet was resuspended in 10 mM Na₂PO₄ (pH 7.5) to an optical density at 600 nm (OD₆₀₀) of approximately 0.8. An overnight culture of P. aeruginosa was centrifuged at 10,000 × g for 10 min to pellet the cells, and then the supernatant was passed through a 0.2-μm-pore-size filter for sterilization. A 100-μl aliquot of supernatant was then added to 1 ml of S. aureus suspension. LasA activity results in clearance of the suspension (reduced turbidity), indicative of S. aureus cell lysis, as measured by the OD₆₀₀ at 2, 6, 10, 14, 18, 22, 26, 30, 45, and 60 min.

Pyocyanin assay.

A pyocyanin assay was performed as previously described (56). Briefly, overnight cultures were centrifuged at 6,000 × g for 10 min at room temperature, and sterile supernatant was collected after filtration through a 0.2-μm-pore-size filter. A 5-ml portion of the supernatant was mixed with 3 ml of chloroform. Following centrifugation for 3 min at 1,200 × g at room temperature, 2.5 ml of the phase under the chloroform was collected and mixed with 1 ml of 0.2 M HCl. A sample of 200 μl was collected, and the absorbance was measured at 530 nm in a spectrophotometer.

Detection of QS gene expression.

Reporter strains carrying green fluorescent protein (GFP) transcriptional fusions to QS promoters (Table 1) were utilized to monitor QS gene expression in environments with and without 5% BSA supplementation. A total of 10⁴ to 10⁵ cells from overnight cultures were inoculated into LB broth with or without 5% BSA supplementation and grown at 37°C, with shaking at 220 rpm, for 24 h. Samples from these cultures were diluted 1:100 in 1× phosphate-buffered saline (PBS), and the level of GFP fluorescence (excitation wavelength, 480 nm; emission wavelength, 535 nm) was detected using a Synergy H1 hybrid microtiter plate reader (BioTek, Winooski, VT).

Detection of autoinducers.

Solutions (5 μM) of purified 3OC12-HSL, C4-HSL (Fluka Chemie, Sigma-Aldrich, Saint Louis, MO), and the quinolones (PQS, HHQ, and HQNO; obtained from Steve Diggle, Nottingham University) were prepared in 1× PBS and incubated for 3 h, with shaking at 130 rpm at 37°C, either with or without 5% BSA. Solutions were then passed through a 30-kDa-molecular-mass-cutoff Vivaspin column (Corning, Lowell, MA) via centrifugation at 8 × g for 10 min. QS molecules were separated from the eluate via three successive rounds of acidified ethyl acetate (high-performance liquid chromatography [HPLC]-grade) extraction as previously described (57). Final extracts were dried in an Eppendorf 5301 concentrator (Eppendorf, Hauppauge, NY) and resuspended in 30 μl of methanol (HPLC grade). Overnight cultures of the 3OC12-HSL, C4-HSL, or PQS luciferase bioreporter strains (Table 1) were diluted to an OD₆₀₀ of 0.5 in LB broth. Five-microliter aliquots of methanol-dissolved extracts were combined with 100-μl aliquots of diluted reporter strain cultures in the wells of a 96-well, clear, flat-bottom microtiter plate (Costar, Corning, NY) and incubated statically at 37°C for 3 h. Luminescence was detected using a luminometer (Modulus microplate reader; Turner Biosystems, Promega, Madison, WI), and values were normalized to the value for the reporter biomass (OD₆₀₀).

ITC.

Isothermal titration calorimetry (ITC) measurements were characterized using a TA Instruments small-volume ITC instrument at 20°C. Titrations were performed using a stir rate of 250 rpm and either 1.5- or 2-μl injection volume. Titrations of 2 mM 3OC12-HSL into 30 μM defatted BSA in 5% ethyl acetate and 1× PBS buffer showed drifting of the thermogram binding profile upon repeated measurements of the same sample over time due to the degradation of the 3OC12-HSL. Measurement of the C12-TA-HSL degradation product of 3OC12-HSL was measured by allowing a reconstituted 3OC12-HSL to degrade completely (for 72 h) before the titration was performed. All titrations were corrected for any heat of dilution of the ligand through separate control experiments in which the ligand was injected into sample buffer lacking BSA in the cell.

Statistics.

Statistical analyses were performed using GraphPad Prism, version 6, or InStat, version 3. Specific tests for significance are indicated in figure legends.