Aminoglycoside-inducible expression of the mexAB-oprM multidrug efflux operon in Pseudomonas aeruginosa: Involvement of the envelope stress-responsive AmgRS two-component system

Abstract

Exposure of P. aeruginosa to the aminoglycoside (AG) paromomycin (PAR) induced expression of the PA3720-armR locus and the mexAB-oprM multidrug efflux operon that AmgR controls, although PAR induction of mexAB-oprM was independent of armR. Multiple AGs promoted mexAB-oprM expression and this was lost in the absence of the amgRS locus encoding an aminoglycoside-activated envelope stress-responsive 2-component system (TCS). Purified AmgR bound to the mexAB-oprM promoter region consistent with this response regulator directly regulating expression of the efflux operon. The thiol-active reagent, diamide, which, like AGs, promotes protein aggregation and cytoplasmic membrane damage also promoted AmgRS-dependent mexAB-oprM expression, a clear indication that the MexAB-OprM efflux system is recruited in response to membrane perturbation and/or circumstances that lead to this. Despite the AG and diamide induction of mexAB-oprM, however, MexAB-OprM does not appear to contribute to resistance to these agents.

Fig 1. PAR induction of PA3720-armR and mexAB-oprM expression.

Expression of PA3720, armR and mexA was assessed in log-phase WT P. aeruginosa strain K767 following a 30-minute exposure to the MIC of the AG, paromomycin (PAR; 256 μg/ml) using qRT-PCR. mexA expression was also assessed in a ΔarmR derivative of K767, K3415, following exposure (30 min) to the MIC of PAR (256 μg/ml) using qRT-PCR. In all cases, expression was normalized to rpoD and is reported relative to the untreated WT P. aeruginosa strain K767 (fold-change). Values shown are means ± standard errors of the means (SEMs) from at least three independent determinations, each performed in triplicate. t-test: *, P < 0.05; ***, P < 0.001.

Fig 2. AG induction of mexAB-oprM expression.

mexA expression was assessed in log-phase WT P. aeruginosa strain K767 following a 30-minute exposure to the MIC of the AGs, paromomycin (PAR; 256 μg/ml), neomycin (NEO; 64 μg/ml), gentamicin (GEN; 2 μg/ml), tobramycin (TOB; 1 μg/ml), and kanamycin (KAN; 128 μg/ml) using qRT-PCR. Expression was normalized to rpoD and is reported relative to the untreated K767 (fold change). Values are means ± SEMs from at least three independent determinations, each performed in triplicate. t-test: *, P < 0.05; ***, P < 0.001.

Fig 3. MexR- and NalD-independent AG induction of mexAB-oprM expression.

mexA expression was assessed in P. aeruginosa strains K1491 (ΔmexR), K3793 (ΔnalD) and K3794 (ΔmexR ΔnalD) following a 30-minute exposure to the MIC of neomycin (NEO; 64 μg/ml for all strains) using qRT-PCR. Expression was normalized to rpoD and is reported relative to the untreated K767 strain (fold change). Values are means ± SEMs (error bars) from at least three independent determinations, each performed in triplicate. t-test: *, P < 0.05; **, P < 0.01.

Fig 4. AmgRS-dependent AG-inducible mexAB-oprM expression.

(A) mexA expression was assessed in log-phase cultures of WT P. aeruginosa strain K767 and its ΔamgR (strain K3159) and ΔamgS (strain K3583) derivatives following exposure to the MIC of PAR (K767, 256 μg/ml; K3159, 16 μg/ml; K3583, 16 μg/ml) for 30 min using qRT-PCR. (B) mexA expression was assessed in log-phase cultures of P. aeruginosa strains K767 (WT; AmgR⁺) and K3159 (ΔamgR; AmgR^-) exposed to the MIC of the indicated antimicrobials (-, no antimicrobial; NEO, neomycin; GEN, gentamicin; KAN; kanamycin) using qRT-PCR. Antimicrobials were used at: NEO, 64 (K767) and 4 (K3159) μg/ml; GEN, 2 (K767) and 0.5 (K3159) μg/ml; KAN, 128 (K767) and 16 (K3159) μg/ml. (C) mexA expression was assessed in strains harbouring WT (K767) and mutant (K3288, AmgS_V121G; K3249, AmgS_R182C) amgS genes using qRT-PCR. (D) armR and PA3720 expression was assessed in log-phase cultures of P. aeruginosa strains K767 (WT), K3159 (ΔamgR), and K3583 (ΔamgS) exposed to the MIC of PAR for 30 min using qRT-PCR. Expression in all cases was normalized to rpoD and is reported relative to the untreated K767 strain (fold change). Values are means ± SEMs from at least three independent determinations, each performed in triplicate. t-test: *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Fig 5. AmgR binds to the mexAB-oprM promoter region.

(A) Electromobility shift assay in which 40 ng of a 351-bp DNA fragment encompassing the intergenic region upstream of mexAB-oprM was incubated with increasing concentrations of purified AmgR as indicated. (B) Electromobility shift assay in which the above-mentioned 351-bp DNA fragment was incubated with the indicated concentration of AmgR in the absence (-) and presence (+) of 200 ng of sheared salmon sperm DNA (sssDNA).

Fig 6. Impact of chloramphenicol on AG induction of mexAB-oprM expression.

mexA expression was assessed in log phase cells of P. aeruginosa strain K767 exposed to NEO (at MIC; 64 μg/ml) for 30 min without (-) and with (+) a 15-min pre-treatment with chloramphenicol (CAM; 128 μg/ml) using qRT-PCR. In a control experiment, K767 was also exposed to 128 μg/ml CAM in the absence of NEO, to assess its impact on mexA expression on its own. Expression was normalized to rpoD and is reported relative to the untreated K767 strain (fold change). Values are means ± SEMs from at least three independent determinations, each performed in triplicate. t-test: **, P < 0.01.

Fig 7. AmgRS-independent macrolide induction of mexAB-oprM expression.

(A) mexA expression was assessed in log-phase cultures of WT P. aeruginosa strain K767 exposed to the MIC of tetracycline (TET; 16 μg/ml), erythromycin (ERY; 512 μg/ml) and azithromycin (AZI; 512 μg/ml) for 30 min using qRT-PCR. Expression was normalized to rpoD and is reported to that of the untreated (-) strain K767. (B) mexA expression was assessed in log-phase cultures of P. aeruginosa strains K767 (AmgR⁺) and K3159 (AmgR^-) exposed to the MIC of ERY or AZI (512 μg/ml for both drugs and both strains) for 30 min using qRT-PCR. In all cases, expression was normalized to rpoD and is reported relative to that of the untreated strain K767 (fold change). Values are means ± SEMs from at least three independent determinations, each performed in triplicate. t-test: **, P < 0.01; ***, P < 0.001.

Fig 8. AmgRS-dependent diamide induction of mexAB-oprM expression.

mexA expression was assessed in log-phase cultures of P. aeruginosa strains K767 (AmgR⁺) and K3159 (AmgR^-) following a 30-min exposure to the MIC of diamide (4 mM for both strains) using qRT-PCR. Expression was normalized to rpoD and is reported to that of the untreated strain K767 (fold change). Values are means ± SEMs from at least three independent determinations, each performed in triplicate. t-test: **, P < 0.01.

Fig 9. Diamide-promoted cytoplasmic membrane depolarization.

Cytoplasmic membrane depolarization, as assessed by DiBAC4(3) fluorescence (arbitrary units; A.U.), was measured over time following exposure of WT strain P. aeruginosa K767 (circles) and its AmgR^- derivative, K3159 (squares), to 4 mM diamide (filled symbols) at T = 0 hr. Unexposed cells are represented by unfilled symbols. The data are means ± SEMs of three independent experiments. Note that the symbols are larger than the error bars, which are thus not visible in the figure. t-test: ***, P < 0.001.