PA5470 Counteracts Antimicrobial Effect of Azithromycin by Releasing Stalled Ribosome in Pseudomonas aeruginosa

ABSTRACT

Pseudomonas aeruginosa causes various acute and chronic infections in humans. Treatment with azithromycin (AZM) has been shown to benefit patients with chronic P. aeruginosa infections. By binding to the exit tunnel of the 50S ribosome, AZM causes ribosome stalling and depletion of the intracellular tRNA pool. It has been shown that AZM is able to kill stationary-phase P. aeruginosa cells and repress quorum sensing-regulated virulence factors as well as swarming motility. In P. aeruginosa, the PA5470 gene encodes a putative peptide chain release factor whose expression is highly induced by macrolide antibiotics. However, its function remains unknown. Here, we found that overexpression of PA5470 increased bacterial tolerance against AZM and alleviated the repression of swarming motility. Ribosome pulldown assays revealed that PA5470 contributes to the release of ribosome stalled by AZM. We further demonstrate that overexpression of PA5470 counteracts AZM-mediated repression of the translation of the quorum sensing regulator RhlR. Overall, our results revealed a novel role of PA5470 in the bacterial response to AZM.

INTRODUCTION

Pseudomonas aeruginosa is a versatile Gram-negative pathogenic bacterium which can cause a wide range of infections in humans, especially in cystic fibrosis patients, burn victims, and immunocompromised patients (1). P. aeruginosa is intrinsically resistant to a variety of antibiotics, which enables its persistence after long-term antibiotic treatment. Its resistance mechanisms include low outer membrane permeability, chromosomally encoded inducible β-lactamase, aminoglycoside-modifying enzymes, multidrug efflux systems, and the ability to form biofilm (2–4).

Aminoglycoside and macrolide antibiotics have been widely used in treating pseudomonal infections, particularly pulmonary infections in cystic fibrosis (CF) patients (5, 6). Aminoglycoside antibiotics bind to the 30S subunit of the ribosome, causing misreading of mRNA. Accumulation of the truncated or aberrant proteins eventually leads to cell death (7). Although P. aeruginosa is highly resistant to macrolide antibiotics such as azithromycin (AZM), several clinical studies have demonstrated that CF patients who are chronically infected with P. aeruginosa benefit from AZM treatment (2, 8–12). The anti-inflammatory effect of AZM might alleviate tissue damage (13, 14). AZM also exhibits bactericidal activity on stationary-growth-phase P. aeruginosa cells (15–17). A subinhibitory concentration of AZM suppresses the expression of quorum sensing system-regulated virulence factors (18–22). The effects of AZM on P. aeruginosa can be counteracted by overexpression of peptidyl-tRNA hydrolase (pth), which increases the intracellular tRNA pool by hydrolyzing accumulated peptidyl-tRNA caused by AZM treatment (23).

The MexXY-OprM efflux pump plays a significant role in resistance to aminoglycoside and macrolide antibiotics in P. aeruginosa (24). Expression of the mexXY operon is repressed by MexZ (25), while MexZ is inhibited by PA5471 via direct binding, leading to the derepression of mexXY expression (26, 27). The expression of PA5471 is induced by protein synthesis inhibiting antibiotics through a transcriptional attenuation mechanism (28). Transcription of PA5471 and its upstream leader peptide (PA5471.1) is driven by a constitutive promoter (28). Complete translation of the PA5471.1 mRNA results in the formation of a transcription terminator upstream of the PA5471 coding region, thus abrogating its transcription (28). In the presence of ribosome-targeting antibiotics, such as aminoglycosides and macrolides, the ribosome is stalled at the PA5471.1 mRNA, which abolishes the formation of the transcription terminator structure, allowing the transcription of PA5471 to proceed (27–29).

PA5470 is a gene located downstream of PA5471 and forms an operon with PA5471 (27). Similar to expression of PA5471, the expression of PA5470 gene is induced by ribosome-targeting antibiotics such as chloramphenicol, erythromycin, tetracycline, and kanamycin but not by norfloxacin or cefotaxime (27). These observations indicate that PA5470 may be involved in the bacterial response to protein synthesis-inhibiting antibiotics in P. aeruginosa. Here, we demonstrated that PA5470 contributes to bacterial tolerance to AZM and counters AZM-mediated inhibition of the quorum sensing system. Our results suggest that PA5470 might function as a release factor that rescues ribosomes stalled by AZM.

RESULTS AND DISCUSSION

Deletion of PA5470 increased the susceptibility of P. aeruginosa to AMK and AZM.

Previous reports demonstrated that PA5470 and PA5471 were upregulated in the presence of translation-inhibitory antibiotics through a transcription attenuation mechanism (27, 28). To confirm the expression pattern of PA5470 gene, we treated the wild-type strain PA14 with four different types of antibiotics and performed real-time PCR. Consistent with a previous report (27), expression of PA5470 was induced by translation-inhibitory antibiotics, including amikacin (AMK) and AZM but not carbenicillin (CAR) or ciprofloxacin (CIP), which inhibits peptidoglycan cross-linking or DNA replication (Fig. 1). These results suggest that PA5470 is likely involved in the bacterial response to ribosome-targeting antibiotics. Accordingly, we examined the MICs of these antibiotics in a PA5470 deletion (ΔPA5470) mutant. However, the ΔPA5470 mutant and wild-type PA14 displayed similar MICs for AMK (2.5 μg/ml) and AZM (400 μg/ml). We then performed an antibiotic killing assay to determine the role of PA5470 in bacterial tolerance. Wild-type PA14 and the ΔPA5470 mutant were grown to exponential phase and treated with AMK. Compared to exponential-phase cells, the P. aeruginosa cells at stationary phase are susceptible to AZM (15, 23). Therefore, the bacterial susceptibility to AZM was examined at stationary growth phase as previously described (15–17, 23, 30, 31). Deletion of PA5470 slightly decreased the bacterial survival rate with AMK and AZM treatment, which was restored by complementation with a wild-type PA5470 gene (Fig. 2A and B). We suspected that the minor differences might be due to multiple tolerance mechanisms encoded by P. aeruginosa, such as multidrug efflux systems and low membrane permeability.

FIG 1.

Expression levels of PA5470 under treatment with various antibiotics. Wild-type PA14 was treated with carbenicillin (37.5 μg/ml; CAR), ciprofloxacin (0.78 μg/ml; CIP), amikacin (1.25 μg/ml; AMK), and azithromycin (200 μg/ml; AZM) for 3 h. Total RNA was isolated, and the mRNA levels of PA5470 were determined by real-time PCR. Error bars represent standard deviations. **, P < 0.01; ***, P < 0.001, compared to results with untreated PA14 (by a Student's t test).

FIG 2.

Increased susceptibility of the ΔPA5470 mutant to AMK and AZM. Exponential-phase bacteria were treated with 0, 2.5, 5, 10, and 20 μg/ml AMK for 1.5 h at 37°C with shaking (A and C). Stationary-phase bacteria were treated with 0, 2, 5, and 10 μg/ml AZM for 20 h at 37°C with shaking (B and D). The bacterial viabilities were determined by serial dilution and plating. Error bars represent standard deviations. *, P < 0.05; **, P < 0.01, compared to results with PA14 containing empty vector (by Student's t test).

In P. aeruginosa, the multidrug efflux system MexXY-OprM plays a major role in bacterial resistance against aminoglycoside and macrolide antibiotics (24). To verify the role of PA5470 in antibiotic resistance, we deleted the mexXY operon in PA14 or the ΔPA5470 mutant. The ΔmexXY and ΔmexXY ΔPA5470 mutants displayed similar MICs for AMK (1.25 μg/ml) and AZM (50 μg/ml). The survival rate of the ΔmexXY ΔPA5470 mutant was approximately 10-fold lower than that of the ΔmexXY mutant upon treatment with 20 μg/ml AMK or 10 μg/ml AZM (Fig. 2C and D). These results suggest that PA5470 is involved in bacterial tolerance to AMK and AZM.

Overexpression of PA5470 increases bacterial tolerance to AMK and AZM.

A previous study demonstrated that overexpression of pth protected P. aeruginosa from killing by AZM (23). To further elucidate the role of PA5470 in bacterial tolerance to antibiotics, we overexpressed the PA5470 gene by cloning it downstream of an arabinose-inducible promoter, P_l-ara (P_l-ara-PA5470). PA14 containing empty vector or P_l-ara-PA5470 was grown to exponential phase and treated with AMK or grown to stationary phase and treated with AZM. Overexpression of PA5470 slightly increased bacterial survival rates with AMK treatment. However, the bacterial survival rate was increased by approximately 100-fold in the presence of 10 μg/ml AZM (Fig. 3A and B). Of note, overexpression of the PA5470 gene conferred a level of protection against AZM similar to that of pth (Fig. 3B). Although overexpression of PA5471 increased the expression of mexX or mexY, overexpression of PA5470 did not affect the expression of these genes (Fig. 3C). In combination, these results suggest a role of PA5470 in bacterial tolerance against AMK and AZM.

FIG 3.

Overexpression of PA5470 increases bacterial tolerance to AMK and AZM without affecting the expression of mexX and mexY. (A) Exponential-phase bacteria were treated with 0, 2.5, 5, 10, and 20 μg/ml AMK for 1.5 h at 37°C with shaking. (B) Stationary-phase bacteria were treated with 0, 2, 5, and 10 μg/ml AZM for 20 h at 37°C with shaking. (C) Bacteria of the indicated strains were grown for 3 h at 37°C with shaking. Total RNA was isolated, and the mRNA levels of mexX and mexY were determined by real-time PCR. The bacterial viabilities were determined by serial dilution and plating. Error bars represent standard deviations. *, P < 0.05; **, P < 0.01, (compared to results for PA14 containing empty vector or overexpressing PA5470); ns, not significant (compared to results with PA14 containing empty vector) (Student's t test).

Thus far, our results revealed that the expression of PA5470 was induced the most by AZM, and overexpression of PA5470 increased the bacterial tolerance to AZM at a much higher degree than that of AMK. Since AMK causes mistranslation and since AZM mainly causes ribosome stalling (16, 20, 23, 32), PA5470 might mainly play a role in relieving the detrimental effect of ribosome stalling. Therefore, we next focused on determining the role of PA5470 in the bacterial response to AZM.

Overexpression of PA5470 counters the AZM-mediated inhibition on swarming motility.

Subinhibitory concentrations of AZM have been shown to inhibit the swarming motility of P. aeruginosa, which can be countered by overexpression of the pth gene (16, 20, 23). Since overexpression of PA5470 conferred similar levels of protection against AZM, we suspected that PA5470 might also be able to rescue the swarming motility. As shown in Fig. 4A and B, overexpression of PA5470 indeed rescued the swarming motility of PA14 in the presence of 10 μg/ml AZM.

FIG 4.

Effect of PA5470 overexpression on swarming motility and expression of rhlA in the presence of AZM. PA14 (I), the PA5470 overexpression strain (II), and the pth overexpression strain (III) were inoculated on swarming plates containing 0.2% l-arabinose without antibiotic (A) or with 10 μg/ml AZM (B). The plates were incubated overnight at 37°C. (C) Quantification of the rhlA mRNA levels. Bacteria were grown in the absence or presence of 5 μg/ml AZM with shaking for 10 h. Total RNA was isolated, and the mRNA levels of rhlA were determined by real-time PCR. The error bars represent standard deviations. **, P < 0.01, compared to results with PA14 containing empty vector (by Student's t test).

It has been demonstrated that AZM represses the production of rhamnolipid, which is required for swarming motility (16, 20, 23). So we tested the expression of rhlA, whose product is a key enzyme in the synthesis of rhamnolipid. Consistent with previous reports, AZM repressed the expression of rhlA; however, overexpression of PA5470 partially restored the expression of rhlA (Fig. 4C).

PA5470 interacts with the ribosome.

To understand the function of PA5470, we analyzed its structure using the Phyre2 Web server (33). A total of 199 residues (98% of PA5470 sequence) were modeled with 100% confidence by peptidyl chain release factor 1 (RCSB/PBD accession no. 4V63). The result was visualized using UCSF Chimera (Fig. 5A) (34). Release factor 1 enters the A-site of a ribosome when it recognizes a stop codon and hydrolyzes the peptidyl-tRNA to release the nascent polypeptide chain and ribosome (35–37). AZM blocks the peptide exit channel of the 50S subunit of the ribosome, resulting in stalling of the ribosome on mRNA and depletion of the intracellular tRNA pool (23, 38). The structural similarity between PA5470 and release factor 1 indicates that PA5470 might play a role in rescuing ribosomes stalled by AZM. So we first examined whether PA5470 can bind to the ribosome. A C-terminally His-tagged RplL (50S ribosomal protein L7/L12) was overexpressed in PA14 containing a PA5470-Flag fusion protein. The His-tagged 70S ribosome was isolated by Ni-nitrilotriacetic acid (NTA) chromatography as previously described (39). Western blot analysis revealed copurification of PA5470-Flag with RplL-His, indicating an interaction between PA5470 and the ribosome (Fig. 5B).

FIG 5.

Interaction between PA5470 and ribosome. (A) Structure modeling of PA5470. Molecular modeling of PA5470 was completed using the Phyre2 Web server with default parameters. The result was visualized using the UCSF Chimera program. (B) Interaction between PA5470 and the ribosome. PA14 carrying pMMB67EH-PA5470-Flag and pUCP24-rplL-His or pUCP24 was grown to an OD₆₀₀ of 0.8 and treated with 0.1 mM IPTG for 3 h. Bacteria were then lysed and subjected to chromatographic purification with Ni-NTA beads. His-tagged RplL and Flag-tagged PA5470 were detected by Western blotting. IB, immunoblotting.

PA5470 relieves AZM-mediated ribosome stalling.

To examine whether PA5470 plays a role in ribosome rescue, we constructed a reporter system based on the translational regulation of an exogenous gene, ermC (40). The ermC gene encodes a 23S rRNA-specific methylase, which confers bacterial resistance to macrolide-lincosamide-streptogramin B antibiotics (41). In the presence of macrolide antibiotics, ribosomes are stalled at the mRNA of the ermC leader peptide (ermC_leader), which alters the secondary structure of the ermC mRNA, resulting in exposure of the ribosome binding site of the ermC mRNA and subsequent translation of the ErmC protein (42, 43). This regulatory mechanism of ermC_leader makes it a perfect model for assessing ribosome stalling in vivo. For a reporter, we fused a Flag-tagged mApple gene downstream of the ribosome binding site of ermC. The DNA fragment containing the ermC_leader coding region and the fused mApple-Flag gene was cloned into pMMB67EH, resulting in pMMB67EH-ermC_leader-mApple-Flag (Fig. 6A). The plasmid was introduced into PA14 containing either an empty vector or P_L-ara-PA5470. Expression levels of the mApple-Flag were determined after treatment with 0, 2, 5, and 10 μg/ml AZM. In PA14/vector, the expression of mApple-Flag was increased upon treatment with 2 μg/ml and 5 μg/ml AZM, suggesting increased ribosome stalling at the ermC_leader mRNA. When the concentration of AZM was increased to 10 μg/ml, the expression of mApple-Flag was decreased, presumably due to the translation-inhibitory effect of AZM. Meanwhile, expression of mApple-Flag was lower in the PA5470 overexpression strain in the presence of 2 μg/ml and 5 μg/ml AZM (Fig. 6B), indicating reduced ribosome stalling at ermC_leader mRNA.

FIG 6.

Assessment of ribosome stalling at the ermC_leader mRNA in the PA5470 overexpression strain. (A) Structure of P_tac-ermC_leader-mApple-Flag and P_tac-ermC_leader-mApple-T0T1. The DNA sequence between the ermC_leader transcriptional start site and the ATG codon of the ermC gene was cloned under the control of P_tac, followed by the coding region of mApple-Flag or two consecutive transcription terminators, T0T1. (B) The protein level of mApple-Flag. Bacteria were treated with 0.1 mM IPTG and the indicated concentration of AZM. The protein levels of mApple-Flag were determined by Western blotting. (C) Quantification of ribosome-associated ermC_leader mRNA. Bacteria treated with 5 μg/ml AZM were lysed and subjected to Ni-NTA chromatography, followed by RNA purification. The relative levels of ermC_leader mRNA were determined by real-time PCR with the 16S ribosome RNA (PA0668.1) as an internal control. **, P < 0.01, compared with the results for PA14 containing empty vector (by a Student's t test).

We then utilized a modified RNA-binding protein immunoprecipitation assay coupled with real-time PCR to directly assess ribosome stalling at ermC_leader as previously described (44–46). First, we cloned ermC_leader in front of two consecutive transcription terminators, T0T1, into pMMB67EH, resulting in pMMB67EH-ermC_leader-T0T1 (Fig. 6A) (47). Then stationary-phase cells of PA14/vector and PA14/P_l-ara-PA5470 containing pUCP24-rplL-His and pMMB67EH-ermC_leader-T0T1 were treated with 5 μg/ml AZM for 1 h. Ribosomes from these cells were purified by Ni-NTA chromatography, and the associated RNA was isolated and subjected to real-time PCR assay. The total mRNA levels of ermC_leader were similar in both input samples (Fig. 6C); however, fewer ermC_leader mRNAs were associated with the ribosome isolated from the PA5470 overexpression strain (Fig. 6C). In combination, these results indicated that overexpression of PA5470 reduces the AZM-mediated ribosome stalling at ermC_leader mRNA.

Overexpression of PA5470 counteracts AZM-mediated repression of RhlR translation.

Previously, Gödeke et al. demonstrated that AZM-mediated ribosome stalling results in depletion of the tRNA pool, which might cause stronger repression of the translation of proteins with rare codons than of those with frequent codons (23). The second codon of rhlR, which is recognized by a rarely used arginine tRNA isoacceptor, renders the translation of RhlR more susceptible to AZM (23). The unavailability of the rarely used tRNA might lead to ribosome stalling at the cognate codon. Based on the potential function of PA5470 as a release factor, we suspected that overexpression of PA5470 might replenish the tRNA pool by rescuing ribosomes stalled by AZM and relieve the translational repression. To test our hypothesis, a C-terminal Flag-tagged rhlR (rhlR-Flag) gene was inserted into the chromosome of a PA14 or PA5470 overexpression strain according to a previously described protocol (47). Consistent with the previous report, treatment with AZM reduced the protein level of the RhlR-Flag. Of note, overexpression of PA5470 increased the translation of RhlR-Flag (Fig. 7A), which might be due to replenishment of the tRNA pool. In addition, replacing the rare codon (AGG) with a frequently used codon (CGC) (23) reduced AZM-mediated inhibition of RhlR translation, and overexpression of PA5470 did not affect the expression level of the mutated RhlR (Fig. 7A).

FIG 7.

Expression of RhlR and quantification of ribosome-associated rhlR mRNAs. (A) The protein level of RhlR-Flag and the mutated RhlR (RhlR-M-Flag) in PA14/vector or PA14/P_l-ara-PA5470 with a P_rhlR-rhlR-Flag or P_rhlR-rhlR-M-Flag integrated into the chromosome. Bacteria were treated with 2 μg/ml of AZM and grown to an OD₆₀₀ of 2.0, and the protein levels of RhlR-Flag and RhlR-M-Flag were determined by Western blotting. (B) Quantification of ribosome-associated rhlR mRNA. Stationary-phase bacteria treated with 5 μg/ml AZM were lysed and subjected to Ni-NTA chromatography, followed by RNA purification. The relative levels of rhlR mRNA were determined by real-time PCR with the 16S ribosome RNA (PA0668.1) as an internal control. **, P < 0.01, compared to results with PA14 containing empty vector (by Student's t test).

To examine whether overexpression of PA5470 indeed reduced ribosome stalling at the rhlR mRNA, we measured ribosome stalling on rhlR mRNA by the aforementioned methods. The mRNA levels of rhlR were similar in both input samples (Fig. 7B), whereas less rhlR mRNA was associated with the ribosome isolated from the PA5470 overexpression strain (Fig. 7B). These results suggest that overexpression of PA5470 decreased ribosome stalling on the rhlR mRNA.

Conclusion.

Here, we demonstrate that overexpression of PA5470 counteracts AZM-mediated killing and repression of RhlR expression. Our results indicate that PA5470 might function as a release factor which rescues ribosomes stalled by AZM. After release from mRNA, the peptidyl-tRNA complex might be hydrolyzed by Pth, which facilitates the recycling of ribosomes and tRNAs (48, 49). On the P. aeruginosa chromosome, PA5470 and PA5471 genes are in the same operon, whose expression is induced by translation-inhibiting antibiotics (27). We thus suspect that in response to these antibiotics, PA5471 increases the expression of the multidrug efflux system MexX-MexY, which mainly pumps out aminoglycoside and macrolide antibiotics (24). Meanwhile, PA5470 contributes to the release of ribosome stalled by the antibiotics. Therefore, PA5470 and PA5471 reduce the detrimental effect of the translation-inhibiting antibiotics through distinct mechanisms, which may contribute to bacterial resistance synergistically.

MATERIALS AND METHODS

Strains and plasmids.

The bacterial strains and plasmids used in this study are listed in Table 1. The Escherichia coli strain DH5α and P. aeruginosa strains were routinely cultured in Luria-Bertani (LB) broth at 37°C. Antibiotics were used at the following concentrations: for E. coli, 100 μg/ml ampicillin, 10 μg/ml tetracycline, and 10 μg/ml gentamicin; for P. aeruginosa, 150 μg/ml carbenicillin, 50 μg/ml tetracycline, and 50 μg/ml gentamicin.

TABLE 1.

Bacteria strains and plasmids used in this study

Strain or plasmid	Relevant characteristics or function	Reference or source
E. coli strain
DH5α	λ− ϕ80dlacZΔM15 Δ(lacZYA-argF)U169 recA1 endA1 hsdR17(rK− mK−) supE44 thi-1 gyrA relA1	TransGen
P. aeruginosa strains
PA14	Wild-type strain PA14	47
PA14 ΔPA5470	PA14 with an in-frame deletion of the PA5470 gene	This study
PA14 ΔmexXY	PA14 with deletion of the mexXY operon	This study
PA14 ΔmexXY ΔPA5470	PA14 with deletion of the mexXY operon and an in-frame deletion of the PA5470 gene	This study
Plasmids
pEX18Tc	Broad-host-range gene replacement vector; sacB+ Tcr	50
pEX18Tc-PA5470	For deletion of the PA5470 gene	This study
pUCP19-PA5470	For gene complementation of the PA5470 gene	This study
pRkaraRed	Broad-host-range Red recombinase vector, Tcr	10
Pl-ara-pth	For overexpression of pth under the arabinose inducible promoter Pl-ara	This study
Pl-ara-PA5470	For overexpression of the PA5470 gene under the arabinose inducible promoter Pl-ara	This study
pMMB67EH	Shuttle vector between E. coli and P. aeruginosa, Ampr	44
pMMB67EH-PA5470-Flag	Plasmid expressing Flag-tagged PA5470	This study
pUCP24-rplL-His	Plasmid expressing His-tagged ribosomal protein RplL	44
pUC18T-mini-Tn7T-Gm	For gene insertion into chromosome; Tcr	47
pUC18T-mini-Tn7T-Gm-rhlR-Flag	For chromosomal integration of Flag-tagged RhlR.	This study
pMMB67EH-ermCleader-mApple-Flag	Plasmid expressing Flag-tagged mApple under the control of ermCleader	This study
pMMB67EH-ermCleader-T0T1	Plasmid with ermCleader followed by two transcriptional terminators	This study

DNA methods.

DNA manipulations were performed according to standard protocols or the manufacturer's instructions. To construct a PA5470 deletion mutant in the wild-type strain PA14, 1,029-bp upstream and 1,069-bp downstream fragments of the PA5470 coding region were cloned into the SacI-HindIII sites of plasmid pEX18Tc, resulting in pEX18Tc-PA5470. Deletion of the chromosomal PA5470 was performed as previously described (50), and the deletion was confirmed by PCR with the primers PA5470-1Up-SacI and PA5470-2Dn-HindIII. For complementation, the PA5470 gene was amplified from the PA14 chromosome and cloned into the HindIII-EcoRI sites of pUCP19. For overexpression of pth and PA5470, the pth coding region and the PA5470 gene coding region, respectively, were amplified from the PA14 chromosome and cloned into the XhoI-XbaI sites of pRkareRed (51). For construction of pMMB67EH-PA5470-Flag, a 639-bp PA5470-Flag fragment was amplified from the chromosome of PA14 and cloned into the EcoRI-BamHI sites of pMMB67EH.

Antibiotic susceptibility assay.

MICs of antibiotics for P. aeruginosa were determined by serial 2-fold dilution in LB medium, as described previously (52–55). MICs were recorded as the lowest concentration of antibiotic inhibiting visible growth after 24 h of incubation at 37°C.

Exponential-phase bacterium-killing assay.

Overnight cultures of P. aeruginosa strains were diluted 100-fold into fresh LB medium and grown at 37°C with shaking (200 rpm). When the optical density at 600 nm (OD₆₀₀) reached 0.3 to 0.4, AMK was added at the indicated concentrations as shown in the figure legends. Then the bacteria were further cultured for 1.5 h at 37°C. Bacterial viabilities were determined by serial dilution and plating on LB agar plates. The survival rate of each strain was calculated as the number of live bacteria in the antibiotic-treated sample divided by the number of bacteria in the corresponding untreated sample.

Stationary-phase bacterial killing assay.

The stationary-phase bacterium-killing assay was performed as described previously (16). Briefly, bacteria were inoculated in LB medium and grown overnight at 37°C. The culture of each strain was diluted to an OD₆₀₀ of 0.05 in LB medium containing 0.2% l-arabinose and cultured at 37°C. After cultures reached stationary phase (OD₆₀₀ of 4.0), AZM at the indicated concentrations as shown in the figure legends was added to the medium. Then the bacteria were cultured for 22 h at 37°C. Bacterial viabilities were determined by serial dilution and plating. The survival rate of each strain was calculated as the number of live bacteria in the antibiotic-treated sample divided by the number of bacteria in the corresponding untreated sample.

Total RNA isolation and quantitative real-time PCR.

An overnight culture of PA14 was diluted 100-fold into fresh LB medium and grown at 37°C with shaking (200 rpm). When the OD₆₀₀ reached 0.3 to 0.4, antibiotic was added into the medium at the indicated concentrations as shown in the figure legends. After 3 h, bacterial cells were harvested by centrifugation. Total RNA was isolated with an RNAprep Pure Cell/Bacteria kit (Tiangen Biotech). Then cDNA was synthesized with reverse transcriptase and random primers (TaKaRa). Real-time PCR was performed with SYBR Premix Ex Taq (TaKaRa). The conserved hypothetical protein coding genes PA1769 and PA1805 were used as internal controls (56, 57). The primers used in quantitative real-time PCR are listed in Table 2.

TABLE 2.

Primers used in this study

Primer	Sequence (5′→3′)
PA1769-F	GAACATCAGCTTCGTCAA
PA1769-R	TCGGCAGCATTATTGATT
PA1805-F	ATATCAGTCTCAATGAAGTC
PA1805-R	CATGGATGGATCGAAATC
PA0668.1-F	AAGGTCTTCGGATTGTAA
PA0668.1-R	GTGCTTATTCTGTTGGTAA
PA5470-F	GAGGTAATCGAGGAGGTG
PA5470-R	CCGATGAACCAGTTCTTG
ermCleader-F	TTGTAATCAGCACAGTTCA
ermCleader-R	CCCTCTTTAATTTGGTTATAATGA
rhlR-F	GAAATCGCCATCATCCTG
rhlR-R	CGTCGAACTTCTTCTGGA

Motility assay.

Swarming motility was tested on swarming plates containing 1% tryptone, 0.5% NaCl, 0.6% glucose, and 0.35% agar. Two microliters of exponential-growth-phase bacteria was deposited on the plates, followed by incubation overnight at 37°C (16).

Ribosome isolation.

Isolation of ribosome by affinity chromatography was performed as previously described with minor modifications (39). Wild-type PA14 containing pMMB67EH-PA5470-Flag and pUCP24-rplL-His or the empty vector pUCP24 was grown to an OD₆₀₀ of 0.8, and the expression of PA5470-Flag was induced with 0.1 mM isopropyl-β-d-thiogalactopyranoside (IPTG) for 3 h. Bacteria were harvested by centrifugation, resuspended in lysis buffer (20 mM Tris-HCl, 150 mM NaCl, 3 mM β-mercaptoethanol, 10 mM imidazole, 0.5% NP-40, pH 8.0) containing 1 unit/μl recombinant RNase inhibitor (TaKaRa) and lysed by sonication. Cell debris was removed by centrifugation, and the supernatants were incubated with Ni-NTA agarose beads for 1 h at 4°C. The beads were then washed five times with the lysis buffer. The ribosomes were eluted with 100 μl of lysis buffer containing 250 mM imidazole. Protein samples were separated by 15% SDS-PAGE and probed with monoclonal anti-Flag (Sigma) or anti-His (Cell Signaling Technology) antibody.

Detection of ribosome-associated RNA.

The amount of ribosome-associated RNA was determined by a modified RNA-binding protein immunoprecipitation assay followed by real-time PCR as previously described (44–46). Stationary-phase bacteria were treated with 5 μg/ml AZM for 1 h. Ribosome was isolated as described above, and ribosome-associated RNA was purified with an RNAprep Pure Cell/Bacteria kit (Tiangen Biotech). The amounts of ermC_leader RNA or rhlR RNA in the samples were determined by real-time PCR using 16S ribosome RNA (PA0668.1) as the internal control for normalization.