PitA Controls the H2- and H3-T6SSs through PhoB in Pseudomonas aeruginosa

ABSTRACT

Pseudomonas aeruginosa possesses three type VI secretion systems (T6SSs) that are involved in interspecies competition, internalization into epithelial cells, and virulence. Host-derived mucin glycans regulate the T6SSs through RetS, and attacks from other species activate the H1-T6SS. However, other environmental signals that control the T6SSs remain to be explored. Previously, we determined PitA to be a constitutive phosphate transporter, whose mutation reduces the intracellular phosphate concentration. Here, we demonstrate that mutation in the pitA gene increases the expression of the H2- and H3-T6SS genes and enhances bacterial uptake by A549 cells. We further found that mutation of pitA results in activation of the quorum sensing (QS) systems, which contributes to the upregulation of the H2- and H3-T6SS genes. Overexpression of the phosphate transporter complex genes pstSCAB or knockdown of the phosphate starvation response regulator gene phoB in the ΔpitA mutant reduces the expression of the QS genes and subsequently the H2- and H3-T6SS genes and bacterial internalization. Furthermore, growth of wild-type PA14 in a low-phosphate medium results in upregulation of the QS and H2- and H3-T6SS genes and bacterial internalization compared to those in cells grown in a high-phosphate medium. Deletion of the phoB gene abolished the differences in the expression of the QS and T6SS genes as well as bacterial internalization in the low- and high- phosphate media. Overall, our results elucidate the mechanism of PitA-mediated regulation on the QS system and H2- and H3-T6SSs and reveal a novel pathway that regulates the T6SSs in response to phosphate starvation.

IMPORTANCE Pseudomonas aeruginosa is an opportunistic pathogenic bacterium that causes acute and chronic infections in humans. The type VI secretion systems (T6SSs) have been shown to associate with chronic infections. Understanding the mechanism used by the bacteria to sense environmental signals and regulate virulence factors will provide clues for developing novel effective treatment strategies. Here, we demonstrate a relationship between a phosphate transporter and the T6SSs and reveal a novel regulatory pathway that senses phosphate limitation and controls bacterial virulence factors in P. aeruginosa.

INTRODUCTION

Pseudomonas aeruginosa is an opportunistic Gram-negative pathogen that causes various diseases, such as nosocomial infections and chronic lung colonization in cystic fibrosis (CF) patients (1, 2). The abilities to evade host immune clearance and outcompete the commensal microbes or other pathogens are essential to the bacterial pathogenesis.

The type VI secretion system (T6SS) is a widespread virulence factor in Gram-negative bacteria that can target prokaryotic and eukaryotic cells (3–5). It is a syringe-like machinery that consists of a puncturing device and a sheath (6). Contraction of the sheath propels the puncturing device to inject effector proteins into target cells (7, 8). P. aeruginosa encodes three T6SSs, namely, the H1-, H2- and H3-T6SSs. The H1-T6SS mainly targets and kills competing bacteria in a multimicrobial environment, which confers a growth advantage on the producer cells (9). The H2-T6SS plays a major role in the internalization of the bacterium by eukaryotic cells (10, 11). Two effector proteins, including VgrG2b and phospholipase D enzyme PldA, have been shown to be injected into epithelial cells through the H2-T6SS, targeting microtubule components and the Ser/Thr kinase Akt, respectively, both of which promote bacterial internalization (4, 10, 11). The H3-T6SS also contributes to bacterial internalization by injecting another Akt-targeting phospholipase D enzyme, PldB, into epithelial cells (4). Wettstadt et al. demonstrated a role for the H2-T6SS in the delivery of PldB (12). Recently, Li et al. revealed that mutation in the H3-T6SS gene clpV3 reduced biofilm formation, pyocyanin production, expression of the T3SS genes, and bacterial virulence in a Galleria mellonella infection model (13).

In P. aeruginosa reference strain PA14, all the three T6SSs are regulated by the RetS-GacS/GacA-RsmY/RsmZ-RsmA pathway. RsmA is an RNA binding protein that directly represses translation of T6SS mRNAs, such as tssA1, tssA2, and tssB3 (14, 15). The small RNAs RsmY and RsmZ derepress the T6SS genes by sequestering RsmA (16). Transcription of RsmY/RsmZ is directly regulated by the two-component regulatory system GacS-GacA (17). RetS is a hybrid sensor kinase that directly binds to the sensor GasS and suppresses its autophosphorylation (18). Recently, Wang et al. demonstrated that mucin glycans promote the inhibition of GacS by RetS (19). Two other hybrid sensor kinases, PA1611 and LadS, promote the expression of RsmY/RsmZ by antagonizing RetS and activating GacS, respectively (20–22).

Meanwhile, the expression of the three T6SSs has been shown to be differently regulated by a transcriptional regulator, AmrZ, and the quorum sensing (QS) systems. AmrZ represses the expression of the H2-T6SS genes and activates the expression of the H1-T6SS and H3-T6SS genes (14). P. aeruginosa contains three QS systems, namely, Las, Rhl, and PQS (23–26). The Las and Rhl QS systems activate the H2- and H3-T6SSs while repressing the H1-T6SS (10, 27, 28).

Previously, we found that mutation in a phosphate transporter gene, pitA (PA4292), resulted in increased resistance to β-lactam antibiotics and decreased resistance to aminoglycoside antibiotics by enhancing pyocyanin production and proton motive force, respectively (29, 30). In this study, we demonstrated that PitA regulates the expression of H2- and H3-T6SS genes through the phosphate stress response regulator PhoB and the QS systems, revealing a novel regulatory signaling pathway of the T6SSs in P. aeruginosa.

RESULTS

Mutation of pitA enhances the expression of the H2- and H3-T6SS genes and bacterial internalization.

Our previous transcriptomic profile analysis revealed upregulation of genes in the three T6SSs in a PA14 ΔpitA mutant (30). The H2- and H3-T6SS genes were upregulated to a greater extent than the H1-T6SS genes (30). To verify the expression levels of the genes, we utilized quantitative real-time PCR to examine the mRNA levels of the T6SS genes. Expression of the H2- and H3-T6SS effector genes hcp2 and hcp3 was upregulated more than 4-fold in the ΔpitA mutant, which was restored by the complementation of the pitA gene (Fig. 1A). Since the H2-T6SS and H3-T6SS both contribute to bacterial internalization by nonphagocytic cells (11, 31), we examined the intracellular bacterial number after incubating the bacteria with cells of the human lung carcinoma line A549. Mutation of pitA increased the bacterial internalization, which was restored by complementation of the pitA gene (Fig. 1B). Meanwhile, the mRNA level of the H1-T6SS effector gene hcp1 in the ΔpitA mutant was similar to that in the wild-type strain (Fig. 1A). Since the H1-T6SS displays antibacterial activity and confers on P. aeruginosa an advantage in competition over other bacteria inhabiting the same niche, bacterial competition assays were performed between P. aeruginosa and Escherichia coli. Previous reports demonstrated that mutation in the retS gene results in upregulation of the H1-T6SS, which enhances the bacterial killing ability (7, 8, 14, 32). We thus used a ΔretS mutant as a positive control in the competition assay. Wild-type PA14 and the ΔpitA mutant showed similar killing abilities against E. coli (Fig. 1C). As mutation of pitA did not affect expression of H1-T6SS genes, which is expected to control antibacterial activity of P. aeruginosa, the similar antibacterial activities between PA14 and the ΔpitA mutant reinforce the notion that PitA mainly controls the H2- and H3-T6SSs.

FIG 1.

PitA affects the H2- and H3-T6SSs. (A) Wild-type PA14, the ΔpitA mutant, and the complemented strain were grown to an OD₆₀₀ of 1. The mRNA levels of hcp1, hcp2, and hcp3 were determined by real-time PCR. (B) Bacterial internalization by A549 cells after incubation for 1.5 h at an MOI of 10. (C) Competition between P. aeruginosa and E. coli. The competitive index represents the ratio of CFU of P. aeruginosa stains to E. coli. Data represent the means ± standard deviations of assays performed in triplicate and are representative of three independent experiments with similar results. ns, not significant. **, P < 0.005; ***, P < 0.001 by one-way analysis of variance (ANOVA).

PitA influences the expression of the H2- and H3-T6SS genes through quorum sensing systems.

The H2- and H3-T6SSs have been shown to be regulated by the QS systems at the stationary phase (10, 28). We previously found that mutation of pitA results in upregulation of pyocyanin synthesis genes (29). Since the pyocyanin synthesis genes are positively regulated by the QS systems (33, 34), we hypothesized that the QS systems might play a role in the upregulation of the H2- and H3-T6SS genes in the ΔpitA mutant. At the exponential growth phase (optical density at 600 nm [OD₆₀₀] = 1.0), the lasI, rhlI, and pqsA mRNA levels were higher in the ΔpitA mutant (Fig. 2A). However, at the stationary growth phase, the lasI, rhlI, and pqsA mRNA levels were similar in wild-type PA14 and the ΔpitA mutant (Fig. 2A). These results indicate that mutation of pitA leads to early activation of the QS systems.

FIG 2.

Roles of the QS systems in the regulation of the H2- and H3-T6SS genes in the ΔpitA mutant. (A) Wild-type PA14, the ΔpitA mutant, and the complemented strain were grown to an OD₆₀₀ of 1 or 3. The mRNA levels of lasI, rhlI, and pqsA were determined by real-time PCR. (B) The mRNA levels of hcp2 and hcp3 in indicated strains at an OD₆₀₀ of 1 or 3 were determined by real-time PCR. Data represent the means ± standard deviations of assays performed in triplicate and are representative of three independent experiments with similar results. *, P < 0.05; **, P < 0.01; ***, P < 0.001 by ANOVA.

To investigate whether the QS systems are involved in the upregulation of the H2- and H3-T6SSs in the ΔpitA mutant, we deleted lasR, rhlR, or pqsA. Consistent with previous reports, the expression of hcp2 and hcp3 in wild-type PA14 was upregulated at the stationary phase, which was reduced by the deletion of lasR, rhlR, or pqsA (Fig. 2B). In the ΔpitA mutant, deletion of lasR, rhlR, or pqsA reduced the expression levels of hcp2 and hcp3 at both the exponential and stationary phases (Fig. 2B). In addition, deletion of lasR, rhlR, or pqsA in the ΔpitA mutant reduced the expression of the phzA gene, as well as pyocyanin production (see Fig. S1 in the supplemental material). These results indicate that the early activation of the QS systems contributes to the upregulation of the H2- and H3-T6SS genes and overproduction of pyocyanin in the ΔpitA mutant.

PitA regulates the QS systems through PhoB.

We previously demonstrated that mutation of pitA reduces intracellular phosphate level, which is restored by overexpression of the phosphate transporter complex genes pstSCAB (30). Overexpression of pstSCAB in the ΔpitA mutant reduced the expression levels of hcp2, hcp3, lasI, rhlI, pqsA, and phzA, pyocyanin production, and bacterial internalization (Fig. 3A to C). However, the overexpression of pstSCAB in wild-type PA14 did not affect the expression of these genes, pyocyanin production, or bacterial internalization. These results indicate a role for phosphate deficiency in the induction of the QS systems and H2- and H3-T6SSs.

FIG 3.

Effects of pstSCAB overexpression on the expression of the T6SSs and QS systems. (A) PA14/pMMB, ΔpitA/pMMB, PA14/pMMB-pstSCBA and ΔpitA/pMMB-pstSCBA strains were grown to the exponential phase. The mRNA levels of hcp2, hcp3, lasI, rhlI, pqsA, and phzA were determined by real-time PCR. (B) Internalization of PA14/pMMB, ΔpitA/pMMB, PA14/pMMB-pstSCBA, and ΔpitA/pMMB-pstSCBA strains by A549 cells. (C) Pyocyanin concentrations of overnight cultures of PA14/pMMB, ΔpitA/pMMB, PA14/pMMB-pstSCBA, and ΔpitA/pMMB-pstSCBA strains. The final values were normalized by the OD₆₀₀ of each bacterial culture. Data represent the means ± standard deviations of assays performed in triplicate and are representative of three independent experiments with similar results. *, P < 0.05; **, P < 0.01; ***, P < 0.001 by ANOVA. (D) Bacteria containing the phoB-FLAG driven by its native promoter were grown to an OD₆₀₀ of 1.0 in LB medium. The levels of PhoB-FLAG and RpoA were determined by Western blotting. Data are representative of three independent experiments with similar results. (E) Quantification of the relative densities of the PhoB-FLAG bands in the Western blot data in Fig. 3D. The densities of the bands were measured by ImageJ software. Data are the means ± standard deviations of three independent experiments with RpoA as the internal control.

In P. aeruginosa, phosphate deficiency results in activation of the PhoB-PhoR two-component regulatory system (35–38). The expression of phoB was upregulated in the ΔpitA mutant (Fig. 3A and D), which was reduced by overexpression of pstSCAB (Fig. 3A and D). We thus examined the role of PhoB in the upregulation of the QS systems. However, we were unable to delete the phoB gene in the ΔpitA mutant, although we could easily construct a ΔphoB mutant in wild-type PA14. Then we knocked down the expression of phoB by overexpressing an antisense small RNA (sRNA) (Fig. 4A), which reduced the expression of hcp2, hcp3, lasI, rhlI, pqsA, and phzA (Fig. 4B). In combination, these results demonstrate that mutation of pitA leads to phosphate deficiency, which activates the QS systems through PhoB, resulting in the upregulation of the H2- and H3-T6SS genes.

FIG 4.

Knockdown of phoB reduced the expression of T6SS genes and QS system genes. (A) Sequence and location of the antisense RNA. Bacteria were grown to an OD₆₀₀ of 1. The levels of PhoB-FLAG and RpoA were determined by Western blotting. Data are representative of three independent experiments with similar results. (B) Bacteria were grown to an OD₆₀₀ of 1.0. The mRNA levels of hcp2, hcp3, lasI, rhlI, pqsA, and phzA were determined by real-time PCR. Data represent the means ± standard deviations of assays performed in triplicate and are representative of three independent experiments with similar results. *, P < 0.05; ***, P < 0.001 by ANOVA.

Both the QS systems and T6SSs have been reported to contribute to bacterial biofilm formation and virulence (39–41). In LB medium, the ΔpitA mutant and wild-type PA14 formed similar levels of biofilm (Fig. S2). In an acute pneumonia model, mutation of pitA attenuated the bacterial virulence, which might have been due to the defective phosphate uptake (Fig. S3).

Phosphate depletion activates the H2- and H3-T6SSs in wild-type PA14.

We next examined whether phosphate limitation affects the expression of the H2- and H3-T6SSs through PhoB in wild-type P. aeruginosa. In the high- and low-phosphate media, the ΔphoB mutant and wild-type PA14 reached an OD₆₀₀ of 1.5 at similar times (Fig. S4). However, the bacterial growth was slower in the low-phosphate medium afterwards (Fig. S4). The ΔphoB mutant displayed a more severe growth defect than did wild-type PA14 in the low-phosphate medium after the OD₆₀₀ reached 1.5 (Fig. S4). Thus, to examine the gene expression levels, the bacteria were harvested at an OD₆₀₀ of 1.0. Growth of wild-type PA14 in the low-phosphate medium increased the mRNA levels of hcp-2, hcp-3, lasI, rhlI, and pqsA (Fig. 5A), which was reduced by the deletion of phoB (Fig. 5A). At an OD₆₀₀ of 1.5, the wild-type PA14 grown in the low-phosphate medium produced more pyocyanin than that grown in the high-phosphate medium (Fig. 5B), whereas deletion of phoB abolished the induction (Fig. 5B). We then investigated bacterial internalization in a low-phosphate buffer (0.9% NaCl plus 10% fetal bovine serum [FBS]) with or without 10 mM phosphate. The bacteria were grown in LB medium and incubated with the cells at a multiplicity of infection (MOI) of 10 for 1.5 h. However, we could barely detect internalized bacteria, presumably due to the unsuitable physiological condition. We thus cultured the bacteria in the low- or high-phosphate medium to an OD₆₀₀ of 1.0 and infected the cells in a regular tissue culture medium (RPMI medium plus 10% FBS) at a high MOI (1:100) for a shorter time (40 min). Growth in the low-phosphate medium led to increased bacterial internalization (Fig. 5C). These results demonstrate that phosphate depletion might increase bacterial internalization.

FIG 5.

Phosphate depletion activates the H2- and H3-T6SSs and QS systems through phoB. (A) Wild-type PA14 and the ΔphoB mutant were grown in high-phosphate or low-phosphate medium to an OD₆₀₀ of 1.0. The mRNA levels of hcp2, hcp3, lasI, rhlI, pqsA, and phzA were determined by real-time PCR. (B) Pyocyanin yields of PA14 and the ΔphoB mutant grown in high-phosphate or low-phosphate medium at an OD₆₀₀ of 1.5. (C) Internalization of PA14 or the ΔphoB mutant grown in high-phosphate or low-phosphate medium by A549 cells after incubation for 40 min at an MOI of 100. Data represent the means ± standard deviations of assays performed in triplicate and are representative of three independent experiments with similar results. *, P < 0.05; **, P < 0.01; ***, P < 0.001 by ANOVA.

DISCUSSION

In this study, we found that defect of the constitutive phosphate transporter PitA leads to activation of the QS systems and subsequent upregulation of the H2- and H3-T6SS genes and enhancement of bacterial internalization by epithelial cells. Previously, we demonstrated that mutation of pitA reduces the intracellular phosphate concentration (30). By overexpressing the phosphate transporter PstSCAB or deleting phoB, we demonstrated that phosphate deficiency contributes to the upregulation of the QS systems and H2- and H3-T6SSs. The proposed function of PitA and regulatory pathways were summarized in Fig. 6.

FIG 6.

The schematic diagram shows the proposed roles of PitA in the regulation of the H2- and H3- T6SSs. A defect of the constitutive phosphate transporter PitA or environmental phosphate depletion may reduce the bacterial intracellular phosphate concentration. The PhoBR two-component system sensor PhoR senses the low phosphate concentration and then phosphorylates the regulator PhoB, which activates the expression of the phoBR operon and the high-affinity phosphate transporter genes pstSCBA. Meanwhile, the phosphorylated PhoB activates the QS systems, which subsequently upregulates the expression of the H2- and H3-T6SS genes. The enhanced H2- and H3-T6SSs promote bacterial internalization into epithelial cells. The solid black arrows represent direct binding to the gene promoters, while the dashed arrows represent unclear regulatory mechanisms.

Phosphate is essential for bacterial growth and metabolism. During infection by P. aeruginosa, the available inorganic phosphate is limited in the host (36, 42). In response to phosphate limitation, the PhoBR system is activated, which subsequently upregulates the phosphate transporters and other genes to promote bacterial survival (35, 43). We found that phosphate depletion enhances bacterial internalization by epithelial cells. It has been shown that the concentration of serum phosphate in septic patients may drop below 0.3 to 0.4 mM, which is the same in vitro concentration as activates the PhoBR system of P. aeruginosa (44). In addition, hypophosphatemia was observed after major liver surgery, leading to induction of the pstS gene (45). These phosphate depletion conditions induce expression of multiple virulence genes (44, 45). Our results imply that bacterial internalization might be increased under these conditions, which might promote antibiotic tolerance and chronic infection (46–48). The internalization may also allow transit into deeper tissues through tight epithelial cell barriers. In this study, we lysed the A549 cells at about 1.5 h postinfection to determine the intracellular bacterial numbers. During the experiment, we did not observe change of viability in A549 cells infected by wild-type PA14 or the pitA mutant. Further studies are warranted to monitor the survival and response of cells with intracellular P. aeruginosa as well as the physiological status of the bacteria. In our experiment, the numbers of internalized bacteria were determined by plating. Besides the internalization rates, the intracellular survival abilities can also affect the results. Additional assays are needed to distinguish between internalization and intracellular survival. In addition, it has been reported that P. aeruginosa can disseminate by invading endothelial cells (49). Thus, further studies to examine the bacterial internalization using endothelial cells are warranted.

Besides the roles in interspecies competition and bacterium-host interactions, T6SSs have been found to play roles in transporting metal ions such as Zn²⁺, Mn²⁺, Cu²⁺, Fe²⁺, and Mo²⁺ (50–59). Lin et al. found that the effector TseF of the H3-T6SS participates in the acquisition of iron by interacting with iron-chelated PQS in the outer membrane vesicles (OMVs), the siderophore receptor FptA, or the outer membrane porin OprF (58). Transcription of the H2- and H3-T6SS genes is repressed by the ferric uptake regulator Fur under iron-replete conditions (10, 58).

PhoB/PhoR is a two-component system that generally exists in Gram-negative bacteria and plays an important role in bacterial response to environmental phosphate concentrations (60). In phosphate-limiting environments, PhoR self-phosphorylates and then transfers the phosphate group to the aspartate residue of the cognate response regulator PhoB. The phosphorylated PhoB activates the transcription of phoBR, pstSCAB, phnCD, etc., thereby improving the bacterium’s ability to obtain extracellular phosphate. In P. aeruginosa, PhoB was reported to activate QS systems under phosphate starvation conditions through binding to the promoter regions of the lasI gene (38). Sequence analysis on the lasI promoter region revealed overlap of the PhoB and LasR binding sites. Electrophoretic mobility shift assays (EMSA) demonstrated that PhoB outcompetes LasR for binding to the lasI promoter region (38). Global sequence analysis might reveal additional overlapping PhoB and LasR boxes, thereby deepening the understanding of the regulatory relationship between phosphate stress response and the QS systems. It has been reported that the H1-T6SS is repressed by the Las QS system (28). In this study, although mutation of pitA activates the Las QS system, the expression of the H1-T6SS genes was not significantly affected. We suspect that besides the Las QS system PitA might affect other regulatory pathways that control the H1-T6SS. Further studies are needed to elucidate the regulatory network.

We previously demonstrated that mutation of pitA reduces the intracellular phosphate level in a complete medium (LB) (30), indicating a role for PitA in transporting phosphate under phosphate-replete conditions. The ΔpitA mutant is more susceptible to aminoglycoside antibiotics due to enhanced drug uptake, which is caused by increased proton motive force (30). However, the intracellular ATP level is lower in the ΔpitA mutant, presumably owing to a lack of phosphate for ATP synthesis (61). Overexpression of the phosphate transporter genes pstSCAB in the ΔpitA mutant was able to restore the intracellular phosphate concentration, proton motive force, and ATP synthesis, as well as aminoglycoside resistance (30). In addition, the pitA mutation enhances pyocyanin production, which increases bacterial resistance to β-lactam antibiotics (29). Here, we demonstrate that mutation of pitA activates the QS and subsequent pyocyanin production and the H2- and H3-T6SSs through PhoB. Overexpression of pstSCAB in the ΔpitA mutant reduced the expression levels of the QS and T6SS genes and pyocyanin production. In combination, these results indicate that the phenotypes we observed in the ΔpitA mutant are mainly attributable to the defective phosphate uptake.

Although mutation of pitA enhanced the QS systems in LB medium, biofilm formation was not increased (Fig. S2). We suspected that the defective phosphate uptake might have hindered the growth of the biofilm mass. In the acute pneumonia model, the ΔpitA mutant displayed reduced virulence (Fig. S3). It has been reported that P. aeruginosa might encounter phosphate stress during infection (36, 42). Thus, we suspected that defective phosphate uptake might have hindered the growth of the ΔpitA mutant in the host.

To examine the occurrence of PitA mutations in P. aeruginosa isolates, we analyzed the PitA protein sequences of 7,906 isolates in the Pseudomonas database (https://www.pseudomonas.com) by using BLASTP. Compared to the PA14 PitA sequence, PitA variants were identified in 1,302 isolates (Table S1). Further studies are warranted to examine the consequences of the mutations in terms of protein functions and bacterial phenotypes.

Overall, our results reveal the relationship between PitA, QS systems, and T6SSs and uncover the influences of phosphate stress on expression of the virulence factors and antibiotic resistance.

MATERIALS AND METHODS

Bacterial strains, plasmids, and culture media.

The bacterial strains and plasmids used in this study are listed in Table S2. Bacterial cells were grown in Luria-Bertani (LB) broth (10 g/L of tryptone, 5 g/L of yeast extract, 5 g/L of NaCl [pH 7.4]). The low-phosphate medium was composed of 20 g of pancreatic digest of gelatin (Difco), 0.14 g of MgCl₂ (Sigma), and 10 g of K₂SO₄ (Sigma) per L of H₂O. For the high-phosphate medium, 4 mM KH₂PO₄ was added to the low-phosphate medium.

RNA isolation and quantitative real-time PCR.

Bacteria were cultured in LB medium overnight to an OD₆₀₀ around 3.0. The bacterial cultures were diluted 100-fold into indicated media. The bacterial growth in the high- and low-phosphate media was monitored by measuring the OD₆₀₀ every 2 h for 12 h (Fig. S4). The bacteria reached an OD₆₀₀ of 1.0 at similar times (Fig. S4). Then total bacterial RNA was extracted using the RNAprep pure cell/bacterium kit (Tiangen Biotec, Beijing, China). cDNA was synthetized with random primers and a reverse transcriptase (TaKaRa, Dalian, China). Quantitative real-time PCR was performed by using the CFX Connect real-time system (Bio-Rad, USA). The indicated specific primers, cDNA, and a SYBR II green supermix (Bio-Rad, Beijing, China) were mixed in a total volume of 20 μL. The 30S ribosomal protein gene rpsL was used as the internal control. Data represent the means ± standard deviations of the results from three biological replicates.

Construction of gene deletion mutants.

The whole open reading frames of the lasR, rhlR, pqsA, and phoB genes were deleted by homologous recombination as previously described (62). Fragments of around 1,000 bp upstream and downstream of the target genes were amplified by PCR using the PA14 genome DNA as the template with primers listed in Table S2. The PCR products were cloned into the plasmid pEX18Tc (62). The resultant plasmids were transferred into E. coli strain S17 by electroporation and then transferred into P. aeruginosa strains by conjugation. The single-crossover mutants were selected on plates containing 50 μg/mL of tetracycline and 25 μg/mL of kanamycin (to kill the E. coli S17 donor strain). Single colonies of the single-crossover mutants were grown overnight in LB medium and then plated on LB plates containing 7.5% sucrose to select double crossover mutants (62). Deletion of the target genes was verified by PCR (Fig. S5).

Downregulation of phoB by the AsRNA.

Construction of the antisense RNA (AsRNA) was performed as previously described (63). A fragment containing 35 bp upstream of the start codon and the first 65 bp of the phoB coding region was amplified by PCR using the PA14 chromosome as the template with primers listed in Table S2. The PCR product was cloned into the plasmid pMMB67EH, resulting in pMMB-AsR phoB, in which an RNA complementary to the phoB mRNA is driven by an inducible tac promoter. To examine the inhibitory effect of the AsRNA, the phoB promoter and gene fragment were fused with a FLAG tag at the 3′ terminus and cloned into the plasmid pUC18TminiTn7T to insert the genome. The levels of PhoB-FLAG were determined by Western blotting (Fig. 4A).

Bacterial internalization assay.

The internalization assay was performed by using A549 cells as described previously (11, 31). A549 cells were cultured to 90% confluence on the day of the experiment in 6-well plates with RPMI medium and 10% fetal bovine serum (FBS). Overnight bacterial cultures were diluted 100-fold into the indicated medium. The bacteria at an OD₆₀₀ of 1.0 were collected by centrifugation at 10,000 × g for 1 min and resuspended in RPMI medium. A549 cells were washed twice with sterile phosphate-buffered saline (PBS) and infected with P. aeruginosa at a multiplicity of infection (MOI) of 10. At 1.5 h postinfection, the cells were washed twice with PBS and resuspended in PBS containing 250 μg/mL of gentamicin, followed by incubation for 15 min. The cells were washed with PBS three times. Then 1 mL of cold deionized water was added to each well to lyse the cells, and the bacterial number was determined by plating. The internalization rates (CFU per cell) were calculated by dividing the numbers of intracellular bacteria by the numbers of A549 cells of the uninfected plates.

Competition assay.

The P. aeruginosa strains and E. coli DH5a/pDN19 were grown to an OD₆₀₀ of 1.0. The bacteria were washed twice and resuspended in fresh LB medium. The indicated strains of P. aeruginosa were mixed with the E. coli at a ratio of 1:1. One milliliter of the mixture was concentrated into 20 μL and spotted onto a nitrocellulose membrane with 0.22-μm pores (Solarbio, Beijing, China) on an LB agar plate. After incubation for 6 h at 37°C, bacterial cells were resuspended from the nitrocellulose membrane and plated onto LB agar plates with or without 50 μg/mL of tetracycline. The P. aeruginosa strains were unable to form colonies on the tetracycline plate. The total colony number (on LB plates) and the E. coli colony number (on tetracycline plates) were counted after incubation at 37°C for 20 h. The competitive index was calculated as the ratio of CFU of P. aeruginosa to E. coli, determined as follows: competitive index = CFU ratio of P. aeruginosa (CFU after coincubation/initial CFU)/CFU ratio of E. coli (CFU after coincubation/initial CFU). The ΔretS mutant was used as a positive control with highly expressed T6SSs (7, 8, 14, 32).

Pyocyanin quantification.

The pyocyanin concentration was determined as described previously (29). The bacteria were grown overnight in LB medium to an OD₆₀₀ around 3.0. A total of 1.5 mL of the bacterial culture was collected. After centrifugation at 10,000 × g for 1 min, 1 mL of supernatant was mixed with 0.6 mL of chloroform. For the samples grown in the high- and low-phosphate media, 30 μL of bacterial overnight culture in LB medium was inoculated into 3 mL of the indicated medium. When the OD₆₀₀ reached 1.5 (Fig. S4), the bacterial supernatant was collected by centrifugation. One milliliter of supernatant of each culture was mixed with 0.6 mL of chloroform. The organic layer was collected after centrifugation at 13,000 × g for 5 min and mixed with 0.2 mL of 0.2 N HCl. After centrifugation at 13,000 × g for 1 min, the upper liquid was collected. The concentration of pyocyanin (in micrograms per milliliter) was calculated as the absorbance at 520 nm of the upper liquid multiplied by 17.072.

Western blotting.

Protein samples were prepared by boiling equivalent numbers of bacterial cells in the loading buffer for 15 min and then loaded onto and separated with a 10% sodium dodecyl sulfate-polyacrylamide (SDS-PAGE) gel. Proteins were transferred onto a polyvinylidene difluoride (PVDF) membrane (Millipore, USA) and incubated with a rabbit monoclonal anti-FLAG antibody (Sigma, USA) or a mouse monoclonal anti-RpoA antibody (CST, USA) for 1 h at room temperature. After four washings with phosphate-buffered saline with Tween 20 (PBST), the membrane was incubated with the respective secondary antibodies for 1 h at room temperature. The membrane was washed with PBST four times. Signals were detected by an ECL Plus kit (Millipore) and visualized by a Bio-Rad molecular imager (ChemiDocXRS). Each experiment was performed at least three times.

Biofilm formation assays.

The bacteria were grown to an OD₆₀₀ of 1 at 37°C and then diluted 1:50 into fresh LB medium. A total of 150 μL of the bacterial suspension was aliquoted into each well of a 96-well plate and cultured at 37°C for 20 h. After removal of the liquid cultures, the wells were washed three times with double-distilled water (ddH₂O) and dried at 65°C for 15 min. The wells were stained with 1% crystal violet for 20 min. Then the wells were washed three times with ddH₂O and dried at 65°C for 15 min. A total of 200 μL of ethanol was added into each well and incubated for 15 min at room temperature. The absorbance of the crystal violet solution was measured at a wavelength of 595 nm with a Varioskan Flash reader (Thermo Fisher Scientific, Vantaa, Finland).

Animal assays.

Animal assays were carried out in accordance with national guidelines for laboratory animal care and usage in research. The protocols were approved by the animal care and use committee of Nankai University, College of Life Sciences, with permit number NK-04-2012. Overnight bacterial cultures were grown in LB medium to an OD₆₀₀ of 1.0 at 37°C. The bacteria were collected by centrifugation at 12,000 × g for 1 min, followed by washing with sterile saline twice. The pellet was resuspended in sterile saline to a concentration of 2 × 10⁸ CFU/mL. Female BALB/c mice 6 weeks old were anesthetized through intraperitoneal injection of 80 μL of 7.5% chloral hydrate and then intranasally inoculated with 20 μL of the bacterial suspension. At 12 h postinfection (hpi), the mice were sacrificed with CO₂. The lungs were removed and homogenized in 1% proteose peptone (Solarbio, Beijing, China), and bacterial loads were determined by plating.

Data availability.

The experimental data are shown in Table S3.