2-AA promotes accumulation of antibiotic tolerant cells in P. aeruginosa

The synthesis of 2-AA is under the control of the QS regulator MvfR via the transcriptional regulation of the pqsABCD operon genes and requires pqsA and pqsD genes but not pqsB or pqsC [29]. To assess the 2-AA role on persisters formation first we used the PA14 isogenic mvfR^- and pqsBC^- mutants and compare them to the parental strain (Figure 1A) following exposure to high concentrations of a bactericidal antibiotic (i.e. meropenem). mvfR^- cells, which do not produce 2-AA and HAQs, exhibited 10 times fewer persisters than the wild-type (WT) cells (Figure 1A), while pqsBC^- that produces 2-AA but not HAQs [29], exhibited higher levels of persisters than did mvfR^- mutant (Figure 1A). As also shown in Figure S1A the majority of PA14, mvfR^- and pqsBC^-exponential-phase cells died quickly, showing a sharp but different drop-off in survival kinetics within 24 h and 48 h. Their persister fractions are comparable to previously reported persister cells rate [32-34]. Consistently, a smaller fraction of mvfR^- cells compared to those of PA14 and pqsBC^- survived the treatment even following 48 h exposure to high concentrations of antibiotic (Figure S1). Exogenous addition of 2-AA increased significantly (p < 0.05) in a dose dependent manner (Figure S1B) the surviving fraction of PA14 cells (Figure 1A), by up to 16 fold. Similarly, exogenous addition of 2-AA increased the antibiotic tolerant cell fraction in mvfR^-cells (Figures 1A and S1B), in which it restored persister formation in a manner toward that of WT cells, confirming that the mvfR^- cell’s deficit in persister formation is likely due to 2-AA. The surviving cells were confirmed to be truly antibiotic tolerant and not resistant mutants; this tolerance characteristic was demonstrated by their unchanged minimal inhibitory concentrations (MICs) (Table S1) and by repetition of the killing curve for a culture inoculated from a single surviving colony (Figure 1B). These effects were observed without any significant alterations on either the growth rate (Figures S2A-C) or sensitivity to meropenem (Figure S2D and Table S1). It should be noted that 2-AA effect is not limited to bacteria that tolerate meropenem only but also tolerant to other antibiotic (i.e. tetracycline, Figure S3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 1. 2-AA promotes antibiotic tolerance in P. aeruginosa.

A. Survival fraction of controls PA14 (black), PA14 + 2-AA (red), and mvfR^- (white), mvfR^- + 2-AA (pink) and pqsBC^- (green) cells grown to OD_600nm= 2.0 followed by the addition of 10 μg/ml meropenem. All experiments were performed in triplicates, and results expressed as mean ± SD. Differences in persister fractions between PA14, mvfR^- and pqsBC^- (p-value = 0.0016), between PA14 and PA14 + 2-AA (p-value = 0.0078) and between mvfR^- and mvfR^- + 2-AA (p-value = 0.0062), are statistically significant (one-way ANOVA, Tukey’s HSD test).

B. Persister cells are antibiotic tolerant, not antibiotic resistant. After meropenem killing (1^st cycle), a single PA14 surviving colony was re-inoculated in fresh medium in the presence (red) or absence (black) of 2-AA (p-value = 0.0078, Student’s t-test), and antibiotic was applied again for 24 h (2^nd cycle) (p-value = 0.0003, Student’s t-test). Additional single colonies tested gave similar results.

C, D. Chemical inhibition/stimulation of 2-AA production modulates persisters formation. Inhibition of 2-AA production (C) by compound M62 (orange) added to PA14 (black) culture decreased persisters formation (D) to a level similar to mvfR^- mutant (white), and stimulation of 2-AA production (C) by M29 (blue) increased PA14 persisters formation (D). The experiment was performed in triplicates, and results are expressed as mean ± SD. Differences in the persisters fractions between PA14, PA14 + M62 and PA14 + M29 are statistically significant (p-value=0.0001, one-way ANOVA, Tukey HSD).

E. The accumulation of persister cells is 2-AA specific, as neither the precursor/analog AA (grey) nor the analog 3-AA (brown) increased the persister cell fraction. The experiment was performed in triplicates, and results are expressed as mean ± SD. Differences in the persisters fractions are only statistically significant between PA14 (black) and PA14 + 2-AA (red) (p-value=0.0004, one-way ANOVA, Tukey HSD).

https://doi.org/10.1371/journal.pone.0080140.g001

To further confirm the correlation between 2-AA production and persisters formation we used two novel MvfR regulon inhibitors we have identified recently (Starkey et. al submitted). The first, M62 significantly reduced the production of HHQ, PQS and 2-AA, while M29 although inhibited HHQ and PQS production, increased 2-AA levels (Figure 1C). Accordingly, while M62 significantly reduced PA14 persisters to a similar level as in mvfR^-, M29 increased the levels of PA14 persisters (Figure 1D). Furthermore, Figure 1E shows that promoting persister formation is specific to 2-AA, as it does not occur with 2-AA analogs such as 3-AA and anthranilic acid (AA). Taken together, these results demonstrate that the small volatile QS molecule 2-AA promotes persisters formation in P. aeruginosa.

2-AA decreases the transcription of many genes involved in the translational capacity of the cell, contributing to persister formation

Previous studies in E. coli have shown that translation arrest is involved in dormancy and persister formation [11,35-37]. A series of findings, as presented in Figure 2, support our deduction that the 2-AA impacts the translational capacity of the cell. First, the transcription of almost all ribosomal protein genes, most tRNA synthesis genes, and numerous translational factors involved in translation initiation, elongation, and release—including EF-Tu, which has been hypothesized to be involved in persister formation [11] is significantly down-regulated in 2-AA treated mvfR^- and pqsBC^-compared to mvfR^- (Table S2 and Figure 2A). Second, expression of the aforementioned translation-related genes, including that of 16S and 23S rRNA (Figure 2B), is significantly up regulated in mvfR^- transcriptome profiles compared to PA14, (Figures 2A-B and Table S2). Third, gene expression (Table S2) and protein levels (Figure 2C) of the ribosomal modulation factor (RMF), which is known to promote ribosomal inactivity [38,39], was increased by 12.65-fold following addition of 2-AA and by 15.9 fold in pqsBC^- cells (Table S2). Moreover, as shown in Figure 2C, RMF protein levels in PA14 cells were prematurely increased at OD_600nm 1.5 following exogenous addition of 2-AA, and the low RMF protein levels in mvfR^- cells increased at the late exponential phase (OD_600nm 2.8) when 2-AA was added. Accordingly, loss of RMF function leads to a decreased persister fraction of about 5 fold compared to the parental strain (Figure 2D). Moreover, the increased rmf expression observed in pqsBC^- cells (Figure 2C) is in agreement with their increased persister fractions (Figures 1A and S1). Fourth, relative to non-2-AA-treated PA14 cells, the overall translational capacity of 2-AA treated PA14 cells was found to be decreased, as reflected by the low ratio between the translating polysomes and the 70S subunit (Figure 2E). Lastly, as shown in Figure 2F modulation of the cell translation capacity that does not affect growth impacted persisters formation. Addition of sub-MIC concentration of the translation inhibitor chloramphenicol increased persisters levels by 16-64-fold in both PA14 and mvfR^- mutant cells (Figure 2F).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 2. 2-AA down-regulates the transcription of many genes involved in the translational capacity of the cell.

A. Graphical analysis of mvfR^-, mvfR^-+ 2-AA and pqsBC^- transcriptomes to examine expression of all genes involved in translation (Table S2). Down-regulated genes are shown as red (2-fold or more reduction), or pink (1.5- to 2-fold reduction), and up regulated genes as green (>2-fold increase) or light green (1.5- to 2-fold increase). No change in gene expression is shown as grey. B. Expression of 16 rRNA (black) and 23S rRNA (red) in mvfR^- and pqsBC^-cells was assessed by qRT-PCR in the absence or presence (+ 2-AA) of exogenously added 2-AA. The PA14 group +/- 2-AA was used as the calibration reference. C. Western blot showing Rmf levels in PA14, mvfR^-, and pqsBC^- cells. Baseline Rmf levels was markedly lower in mvfR^- cells than in PA14 cells; Rmf levels were increased in pqsBC^-, and in both PA14 and mvfR^- cell types in the presence of exogenously added 2-AA. D. rmf^-- cells (purple) exhibit a reduced persister fraction compared to PA14 cells (black). Differences between both strains are statistically significant (p-value = 0.0113, Student’s t-test). Experiment was performed in triplicate. E. Exogenously added 2-AA decreases polysomes levels. Ribosomal extracts from PA14 cells grown in the absence (left panel) or presence (right panel) of exogenously added 2-AA were separated in a 25% to 5% sucrose density gradient. The ratio of translating polysomes to 70S was higher in the untreated cultures (left) than in the cultures grown in presence of exogenously added 2-AA (right) reflecting a decrease in translational activity. F. Inhibition of PA14 translational capacity increase persisters formation. PA14 and mvfR^- cells were grown in absence (black) or presence (orange) of sub-MIC concentration of translational inhibiting antibiotic chloramphenicol (15 mg/L), which did not affect PA14 growth (data not shown). Persisters fraction was determined as described in Figure 1. The results are presented as log₂ fold of change of the untreated PA14 persisters fraction as mean ± SD.

https://doi.org/10.1371/journal.pone.0080140.g002

These results provide strong support for the notion that the 2-AA mediated transcriptional modulation alters the translation capacity of the cell, and thus promotes persister cell accumulation. Taking into consideration the conserved nature of the translational machinery, we hypothesized that 2-AA may also promote persisters formation in other bacteria. Indeed, Figure 3 shows that 2-AA increased persisters accumulation in other two pathogens, Burkholderia thailandensis and Acinetobacter baumannii that are frequently isolated together with P. aeruginosa [40,41]. These data suggest that 2-AA may promote antibiotic tolerance in polymicrobial communities across bacterial genera.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Figure 3. 2-AA increase persister cell formation in other bacteria.

Survival fraction of Acinetobacter baumannii (left panel), and Burkholderia thailandensis (right panel) cells grown to OD_600nm= 2.0 in the absence (black) or presence (red) of 2-AA followed by the addition of 10 μg/ml meropenem. All experiments were performed in triplicates, and results expressed as mean ± SD. Differences in persister fractions between absence and presence of 2-AA are statistically significant for both A. baumannii (p-value = 0.0068, Student’s t-test) and B. thailandensis (p-value = 0.0203, Student’s t-test).

https://doi.org/10.1371/journal.pone.0080140.g003

Strains, antibiotics, and chemicals

Unless otherwise stated, bacteria were grown overnight at 37 °C in LB (Difco, Detroit) with aeration, or on LB agar plates containing the called for antibiotic or compounds. The pqsBC^-mutant construction procedure is described elsewhere [29]. Acinetobacter baumannii is a clinical isolate (Shriners Hospitals for Children Boston). Burkholderia thailandensis is closely related to Burkholderia pseudomallei [49]. AA, 2-AA, 3-AA and tetracycline were obtained from Sigma-Aldrich (St. Louis, MO), Meropenem from Astra-Zeneca (Newark, DE). Compounds were added at concentration of 0.75mM except for Figure 3 where 1.5mM was used for Burkholderia. M29 and M62 compounds are described in (Starkey et al. submitted).

Persisters assessment

Persisters were assessed by using the standard plating method [50] and the new Start of Growth Time (SGT) method recently described by our group [51], where over-night cultures were diluted 1:100 in triplicate in fresh LB medium in the presence or absence of 2-AA, and re-grown to an OD_600nm of 2.0 while shaking under aeration at 37°C. This took approximately 4 hours. Before the antibiotic treatment, aliquots were plated for reference colony counts. The cultures were then treated with 10 mg/L of meropenem or tetracycline and further incubated at 37°C while shaking under aeration. At 24 h, the number of surviving bacteria was assessed by plating. The survival fraction (persisters) was calculated by dividing the number of surviving bacteria by the number of live bacteria that were present before the antibiotic treatment and expressed as a mean ± SD of at least three replicate experiments.

Statistical analysis

Growth curves were compared by permutation tests using the Thompson and Smyth method (http://bioinf.wehi.edu.au/software/compareCurves/) [52,53] with 10,000 permutations. Killing curves were compared by applying a one step decay exponential model (GraphPad Prism 5.04, Graphpad Software, Inc.). Persister fractions were compared using Student’s t-test or Analysis of Variance (ANOVA) and Tukey’s Honest Significant Difference (HSD) as indicated. Statistical significance was set at a p value of 0.05.

RNA isolation and transcriptome data generation and analysis

Transcriptome data of mvfR^-, mvfR^- supplemented with 2-AA (mvfR^- + 2-AA) and pqsBC^- were generated as previously described [13]. These experiments were performed independently in triplicate. Normalized expression values were analyzed with the Bioconductor limma package and R functions (http://www.bioconductor.org/) [54]. The fold change relative to WT sample was log₂ transformed before the sample-sample correlation coefficients were calculated. Pearson correlation coefficients of log₂ transformed fold changes between samples were calculated with the R function “cor()” in the base “stats” package (http://www.bioconductor.org/) [54]. The transcriptome results of genes known to be involved in persister cells formation were validated by RT-PCR and results are expressed as log₂ fold of change of 2-AA treated versus untreated cells: i) induced gene rmf (PA3049) (3.8 +/- 0.2); ii) repressed genes lexA (PA3007) (-0.8 +/- 0.2), recA (PA3617) (-2.1 +/- 0.4) and sulA (PA3008) (-1.2 +/-0.1). The microarray data have been deposited in the NCBI Gene Expression Omnibus under GEO Series accession number GSE24036 and can be accessed using the following link:

http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?token=vzcrpuuwkwcymxu&acc=GSE24036

Western Blotting

Cells grown to the desired OD_600nm at 37°C while shaking, in presence or absence of 2-AA, were harvested by centrifugation, re-suspended in 500 µl of Bug Buster protein extraction reagent (Novagen) and lysed in presence of 10 mg/L lysozyme according to the manufacturer’s instructions at room temperature by gentle inversion for 30 min. After removal of cellular debris by centrifugation, the Bradford method was used to calculate protein concentrations according to the manufacturer’s protocol (Thermo Scientific Product #23238). Equivalent amounts of total proteins (5 µg) were heated for 5 min at 95°C in Laemmli sample buffer (Biorad), separated by SDS-PAGE on 12% pre-cast polyacrylamide gels (Biorad), and transferred to PVDF membranes (Millipore). The membranes were blocked with an excess of non-fat milk and then probed with RMF rabbit antiserum [1:2,000] in TBS-Tween. RMF bands were revealed by incubating the membranes with horseradish peroxidase-labeled goat anti-rabbit secondary antibodies [1:20,000] and TMB One Solution (Promega), according to the manufacturer’s instructions. Intensities of RMF bands within a pre-defined area were quantified using ImageJ 1.42q software (NIH, http://rsb.info.nih.gov/ij).

Antibiotic sensitivity profile

The possible effect of 2-AA addition on antibiotic susceptibility of PA14 cells was evaluated using the following two methods

a): Broth macrodilution method

MICs of antibiotics were determined using a previously described broth macrodilution method [55] and 10⁵ CFU/mL as the starting inoculum. The MIC was defined as the lowest antibiotic concentration able to inhibit visible bacterial growth after 24 h of incubation at 37 °C.

b): Population analysis profiles

PA14 cells sensitivity to meropenem in the presence or absence of exogenously added 2-AA in various concentrations was determined by plating high number of cells (up to 10⁹ CFU/ml) onto LB plates containing 2-fold serial dilutions of antibiotics as described [56]. Population analysis profile curves were generated by plotting the numbers of colonies growing on the plates against the concentrations of antibiotic present on each plate.

Ribosomal profiles

PA14 cells were harvested from logarithmic (OD_600nm 2) cultures grown in the presence or absence of 2-AA exogenously added to 1 ml of a 20 mM Tris-HCL (pH 7.4) buffer supplemented with 15 mM MgCl₂, 30 mM NH₄Cl₂, and 125 mg/L chloramphenicol. The cells were lysed in the presence of 10 g/L lysozyme by three freeze/thaw cycles in liquid nitrogen. The lysed solution was subsequently treated with 10% sodium deoxycholate and 4 µl of a 1 mg/L DNaseI solution. Cellular debris was spun down and the supernatants solutions were collected. Ribosomes were subjected to centrifugation on 5–20% linear sucrose density gradients. After ultra-centrifugation in a SW40 Ti rotor (Beckman) at 285,000 ×g for 80 min at 4°C, gradients were fractionated by upward displacement with 70% (wt/vol) sucrose, and absorbance at 260 nm was monitored continuously by using an ISCO UA-6 UV monitor.