Multidrug efflux inhibition in Acinetobacter baumannii: comparison between 1-(1-naphthylmethyl)-piperazine and phenyl-arginine-β-naphthylamide

Abstract

Objectives: 1-(1-Naphthylmethyl)-piperazine (NMP) has been shown to reverse multidrug resistance (MDR) in Escherichia coli overexpressing RND-type efflux pumps but there are no data on its activity in non-fermenters like Acinetobacter.

Methods: Antimicrobial susceptibility in the absence and presence of NMP and, for comparison, phenyl-arginine-β-naphthylamide (PAβN), another putative efflux pump inhibitor (EPI), was tested in laboratory and mutant strains with differing intracellular dye accumulation and expression of adeB, and in clinical isolates of Acinetobacter baumannii.

Results: Based on a 4-fold or greater MIC reduction, the effects of both EPIs at low concentrations (25 mg/L) were limited. PAβN had a highly selective action on the reduction in the MIC of rifampicin and clarithromycin. At a higher concentration of the putative EPIs (100 mg/L), NMP was more active than PAβN. This effect was not limited to strains with adeB overexpression, but affected the susceptibility to linezolid, chloramphenicol and tetracycline most, and was enhanced in clinical isolates with reduced fluoroquinolone susceptibility.

Conclusion: NMP can partially reverse MDR in A. baumannii and differs substantially in its activity from that of PAβN.

Introduction

Acinetobacter baumannii has become an important cause of nosocomial infections with multiple outbreaks recently reported.^1–5 Treatment of infections caused by A. baumannii can be difficult because of the intrinsic resistance of the organism to several antimicrobial agents and acquired resistance to many others. Such multidrug-resistant A. baumannii often are susceptible only to carbapenems, amikacin and to polymyxins, and some may be susceptible only to polymyxins. Recent studies have identified a proton motive force-dependent resistance-nodulation-cell division (RND) type tripartite efflux pump in A. baumannii, named AdeABC.⁶ Inactivation of this pump revealed that it was responsible for resistance towards aminoglycosides and for decreased susceptibility to fluoroquinolones, tetracyclines, chloramphenicol, macrolides and ethidium bromide. Isolates with overexpression of adeABC can be selected in vivo under the selective pressure of fluoroquinolones,⁷ and it is likely that AdeABC expression contributes to the multidrug-resistant phenotypes commonly found in A. baumannii clinical isolates.

Pharmacological inhibition of MDR efflux pumps might be an attractive goal to reverse drug resistance in Acinetobacter species and to improve therapy options. A few putative bacterial efflux pump inhibitors (EPIs) have been described.⁸ An example is phenyl-arginine-β-naphthylamide (PAβN). This compound was reported to be a broad-spectrum EPI capable of reversing the MDR phenotype of Pseudomonas aeruginosa and several other Gram-negative bacteria.⁹ It has also been tested in A. baumannii clinical isolates.¹⁰ A 4-fold or greater reduction in the MIC of nalidixic acid after PAβN addition was observed in approximately half of the tested isolates, though there was no significant effect of PAβN addition on the susceptibility to ciprofloxacin. In the present paper, we analyse the effect of another EPI, 1-(1-naphthylmethyl)-piperazine (NMP),^11–13 on drug susceptibility in A. baumannii and show that NMP can partially reverse MDR and differs in its activity from that of PAβN in this species.

Materials and methods

Bacterial strains

A. baumannii strains U10247 and U11177 were clonal outbreak isolates and have previously been described.⁷ The two isolates differed from each other in dye accumulation and adeB gene expression (see below, and Table 1). Strain SB13 was a clinical isolate that was used in selection experiments with increasing concentrations of moxifloxacin (Table 1). Two resulting mutants, first-step mutant SBMox1 and second-step mutant SBMox2, differed in dye accumulation and adeB gene expression but had no new mutations in the quinolone resistance determining regions (QRDR) of topoisomerase genes gyrA and parC (Table 1). Sequencing of the QRDR and gene expression studies by quantitative RT–PCR (qRT–PCR) were done as described.⁷ Clinical isolates of A. baumannii were from collections of non-replicate isolates from intensive care patients. The standard A. baumannii ATCC 19606 was used as a control strain.

Chemicals and media

PAβN, pyronin Y, polymyxin B nonapeptide (PMBN) and CCCP (carbonylcyanide-3-chlorophenyl hydrazone) were purchased from Sigma-Aldrich (Steinheim, Germany), and NMP was obtained from Chess (Mannheim, Germany). Nitrocefin, Luria–Bertani (LB) broth and agar were obtained from Oxoid (Basingstoke, England) and ethidium bromide from Merck (Darmstadt, Germany).

Susceptibility testing

Susceptibilities to a panel of different antibiotics were studied by microbroth dilution in the presence or absence of NMP or PAβN, in accordance with NCCLS performance and interpretive guidelines. Custom microtitre plates containing selected antimicrobials at increasing concentrations were purchased from Merlin Diagnostics (Bornheim, Germany). A 4-fold or greater reduction in the MIC values after addition of NMP or PAβN was considered significant. Microdilution tests were also performed to determine the MIC of ethidium bromide (EtBr) and pyronin Y.

Fluorescent dye whole cell accumulation assays and nitrocefin uptake

Cells were grown overnight on LB agar plates and diluted in 1 mL of PBS + 0.4% glucose (pH 7.4) until an OD at 600 nm of ∼1 was reached. The cells were then transferred to a 96-well plate, and NMP was added. Thereafter, EtBr was added to a final concentration of 1 mg/L, and the relative fluorescence intensity was measured over time in a Safire (Tecan, Crailsheim, Germany) fluorescence plate-reader (excitation 518 nm, emission 605 nm). A similar assay was performed with pyronin Y (final concentration, 5 mg/L, excitation 545 nm, emission 570 nm).

In order to test potential effects on the outer membrane permeability, rates of hydrolysis of a chromogenic β-lactam, nitrocefin, by intact cells of A. baumannii ATCC 19606, were measured in the absence or presence of NMP, PAβN and PMBN. This highly reproducible assay (coefficient of variation, <10%) was essentially done as described by Lomovskaya et al.⁹ Briefly, hydrolysis of nitrocefin was monitored spectrophotometrically by measurement of the increase in absorbance at 490 nm after addition of nitrocefin (final concentration, 32 mg/L) to a cell suspension (OD at 600 nm of ∼0.5, in Mg²⁺-free PBS + 0.4% glucose).

Results and discussion

The reference strain A. baumannii ATCC 19606 showed a susceptibility pattern after addition of NMP which differed from that seen with PAβN addition (Table 1). Both EPIs had little or no effect when added at a concentration of 25 mg/L (data not shown). NMP at 100 mg/L (approximately one-fourth of its intrinsic MIC) reduced by 4-fold, or more, the MICs of many test drugs except aminoglycosides, while the effect of PAβN at 100 mg/L was limited to clarithromycin, rifampicin and linezolid. It was interesting to see whether similar susceptibility patterns were obtained with strains SB13 and U10247 that appeared to have very little adeB expression in qRT–PCR evaluations. As shown in Table 1, PAβN consistently reduced the MICs of clarithromycin, rifampicin and linezolid in these two strains while effects by PAβN on the susceptibility of other drugs were only seen in adeB-overexpressing mutants. Together with the observation that the rifampicin MIC did not change with increasing adeB expression, this indicated the presence of two different mechanisms of action of PAβN, one of which was independent of the inhibition of the AdeABC pump. The results of the nitrocefin uptake assay suggested that this second effect of PAβN was an increased outer membrane permeability of A. baumannii, similar to that found in P. aeruginosa.⁹ The mean absorbance values 30 min after addition of PAβN were 1.15 (at 25 mg/L) and 1.22 (at 100 mg/L), which were similar to those observed with PMBN (1.21 at 10 mg/L), an agent with little intrinsic antimicrobial activity but significant effects on membrane permeability in Gram-negative bacterial organisms. In comparison, values for NMP were not (0.89 at 25 mg/L), or only slightly, (1.0 at 100 mg/L) increased compared with control values in the absence of EPIs (0.93). Earlier experiments with PMBN demonstrated that PMBN addition through its permeabilizing effects leads to large reductions in the MIC of macrolides and rifampicin and rather small reductions in the MIC of other agents.¹⁴ Thus, the differential effect of PAβN on macrolide and rifampicin (and presumably linezolid) MICs is in fact probably due to increased membrane permeability rather than efflux pump inhibition.

Magnet and co-workers have shown that aminoglycosides, fluoroquinolones, cefotaxime, erythromycin, tetracycline, chloramphenicol and EtBr were substrates for the AdeB pump while ceftazidime and rifampicin were not.⁶ Testing the SB13 series of mutants that showed increasing adeB expression and did not differ in QRDR nucleotide sequences confirmed and extended this observation. The MICs of rifampicin (Table 1) and ceftazidime, carbapenems and piperacillin (data not shown) did not change in mutants SBMox1 and SBMox2 compared with the parental strain SB13. The increases in the MIC of tetracycline, chloramphenicol and EtBr were small but reproducible. Dye accumulation experiments were consistent with decreasing EtBr intracellular concentrations as the reason for the increasing resistance. Intermediate MIC increases were observed for clarithromycin, linezolid, aminoglycosides and cefepime while larger MIC increases were seen for fluoroquinolones (Table 1). Interestingly, NMP was active in reducing the MIC of several drugs in SB13 and U10247 which indicated a resistance reversal activity, at least in part, independent of adeB expression but broader than and different from that observed with PAβN. On the one hand, a significant effect on the MIC of ciprofloxacin, for example, was only found for NMP but not for PAβN. On the other hand, a significant effect by NMP on the aminoglycoside MIC was only observed in mutant SBMox2 which showed a large reduction in the MIC of streptomycin after addition of NMP, while in the parental strain SB13 that showed very little adeB expression NMP was paradoxically increasing the resistance to streptomycin. Surprisingly, there were no significant effects of NMP or PAβN on the susceptibility of amikacin or tobramycin even in strains with relatively high adeB expression such as SBMox2.

The testing of clinical isolates essentially confirmed the different patterns of resistance reversal of the two putative EPIs, NMP and PAβN. The latter had marked effects on the susceptibility to clarithromycin and rifampicin, which were already observed at the lower concentration of 25 mg/L (Table 2) and possibly reflected membrane permeability changes. The effects of PAβN addition on the susceptibility to other drugs including EtBr were much more limited. As expected from the results in strains ATCC 19606, SB13, its mutants, U10247 and U11117, and as reported previously for clinical isolates,¹⁰ PAβN was virtually ineffective in reducing the MIC of fluoroquinolones. NMP, in contrast, affected more agents than PAβN, in particular linezolid, chloramphenicol and tetracycline (Table 2).

The effects of NMP were clearly enhanced in clinical isolates with reduced fluoroquinolone susceptibility (Table 2). Significant effects of NMP addition on the MICs of fluoroquinolones were observed in roughly half of the fluoroquinolone-resistant isolates compared with <25% of fluoroquinolone-susceptible isolates (Table 2). There were few clinical isolates that showed significant effects with either EPI on enhanced susceptibility to aminoglycosides. Possibly, the known multitude of aminoglycoside resistance genes among clinical A. baumannii precludes the observation of clear effects of pump inhibition on aminoglycoside susceptibility even in strains with increased efflux pump activity.

In conclusion, the findings of the present study demonstrate the capacity of the two putative EPIs NMP and PAβN to partly reverse drug resistance in A. baumannii but strongly indicate different mechanisms of action of the two compounds and effects that are in part independent of expression of adeB, the gene for an important RND-type multidrug efflux pump. Whether part of the effects can be explained by the effects on another pump such as the newly described AbeM¹⁵ remains to be determined.

Transparency declarations

None to declare.

This study was supported in part by the Landesstiftung Baden-Württemberg.

References

Author notes

1Center for Infectious Diseases and Travel Medicine, University Hospital, Freiburg, Germany; 2Institute of Medical Microbiology, Immunology and Hygiene, University Hospital, Köln, Germany; 3Institute of Environmental Medicine and Hospital Epidemiology, University Hospital, Freiburg, Germany; 4Section of Infectious Diseases, Hacettepe University, Ankara, Turkey