Research Article | Spotlight

Substrate Specificities and Efflux Efficiencies of RND Efflux Pumps of Acinetobacter baumannii

Inga V. Leus, Jon W. Weeks, Vincent Bonifay, Lauren Smith, Sophie Richardson, Helen I. Zgurskaya
Thomas J. Silhavy, Editor
DOI: 10.1128/JB.00049-18

This article has a correction. Please see:

ABSTRACT

Antibiotic-resistant Acinetobacter baumannii causes infections that are extremely difficult to treat. A significant role in these resistance profiles is attributed to multidrug efflux pumps, especially those belonging to the resistance-nodulation-cell division (RND) superfamily of transporters. In this study, we analyzed functions and properties of RND efflux pumps in A. baumannii ATCC 17978. This strain is susceptible to antibiotics and does not contain mutations that are commonly selected upon exposure to high concentrations of antibiotics. We constructed derivatives of ATCC 17978 lacking chromosomally encoded RND pumps and complemented these strains by the plasmid-borne genes. We analyzed the substrate selectivities and efficiencies of the individual pumps in the context of native outer membranes and their hyperporinated variants. Our results show that inactivation of AdeIJK provides the strongest potentiation of antibiotic activities, whereas inactivation of AdeFGH triggers the overexpression of AdeAB. The plasmid-borne overproduction complements the hypersusceptible phenotypes of the efflux deletion mutants to the levels of the parental ATCC 17978. Only a few antibiotics strongly benefitted from the overproduction of efflux pumps and antibacterial activities of some of those depended on the synergistic interaction with the low permeability barrier of the outer membrane. Either overproduction or inactivation of efflux pumps change dramatically the lipidome of ATCC 17978. We conclude that efflux pumps of A. baumannii are tightly integrated into physiology of this bacterium and that clinical levels of antibiotic resistance in A. baumannii isolates are unlikely to be reached solely due to the overproduction of RND efflux pumps.

IMPORTANCE RND-type efflux pumps are important contributors in development of clinical antibiotic resistance in A. baumannii. However, their specific roles and the extent of contribution to antibiotic resistance remain unclear. We analyzed antibacterial activities of antibiotics in strains with different permeability barriers and found that the role of active efflux in antibiotic resistance of A. baumannii is limited to a few select antibiotics. Our results further show that the impact of efflux pump overproduction on antibiotic susceptibility is significantly lower than the previously reported for clinical isolates. Additional mechanisms of resistance, in particular those that improve the permeability barriers of bacterial cells and act synergistically with active efflux pumps are likely involved in antibiotic resistance of clinical A. baumannii isolates.

INTRODUCTION

Acinetobacter baumannii exhibits high levels of intrinsic resistance to several β-lactams, chloramphenicol and trimethoprim. Over the past 40 years susceptibilities of A. baumannii clinical isolates continuously decreased and comprised an increasing number of antibiotics (1–3). Recent surveillance data from the United States suggest that ca. 50 to 60% of A. baumannii isolates were multidrug resistant, including resistance to such classes of antibiotics as fluoroquinolones, aminoglycosides, and tetracyclines (4–6). Infections caused by such strains are extremely difficult to manage by chemotherapies and are associated with high mortality.

As in other Gram-negative bacteria, the development of multidrug-resistant A. baumannii involves overexpression of the chromosomally encoded efflux pumps, especially those belonging to the resistance-nodulation-cell division (RND) superfamily of transporters (7). The RND transporters are present in genomes of all Gram-negative bacteria, and their numbers and polyspecificities correlate with high intrinsic and clinical resistance (8, 9). Overproduction of one of the three major RND pumps was reported in A. baumannii clinical isolates and associated with antibiotic resistance: AdeABC, AdeFGH, and AdeIJK (10, 11). The three-component composition of these pumps empowers them with the ability to transport their substrates across both the inner and the outer membrane and make use of the synergistic interaction with the low permeability barrier of the outer membrane (OM) (12–14). Typically, the first gene in the operon encodes a membrane fusion protein (MFP), e.g., AdeA, AdeF, and AdeI. The second gene encodes the RND transporters AdeB, AdeG, and AdeJ, respectively. Finally, the third component and the last gene in the operon is the outer membrane factor (OMF), exemplified by AdeC, AdeH, and AdeK in A. baumannii. The substrate recognition and transport are predominantly the function of the RND component of efflux complexes (15–17). The OMF provides a channel for the substrate to cross the outer membrane, and the MFP enables functional interactions between transporters and channels and are responsible for the coupling of reactions separated in two different membranes (18–22).

The polyspecificity is a mechanistic signature of RND transporters, although some of these transporters are selective for only one or two compounds. In different A. baumannii strains and clinical isolates, AdeIJK was identified as the major constitutively expressed efflux pump with broad substrate specificity (7, 23). Among AdeIJK substrates are β-lactams, chloramphenicol, fluoroquinolones, and tetracyclines.

AdeABC is also polyspecific, but the spectrum of its substrates is narrower and different from that of AdeIJK. Inactivation of adeB in the clinical strain BM4587 did not alter susceptibility to tested antibiotics (24). However, overproduction of AdeABC is the major marker in the acquired multidrug resistance in A. baumannii clinical isolates from different geographical regions (25, 26). Surprisingly, in addition to β-lactams, tetracyclines, and macrolides, the typical substrates of RND transporters in different Gram-negative bacteria, mutants with the overexpression of adeABC are also resistant to aminoglycosides, antibiotics that act on the outer membrane and are actively transported into the cytoplasm (11). Overexpression of either AdeABC or AdeIJK in clinical strains affects the expression of various proteins involved in adhesion and biofilm formation, diminishing biofilm formation and natural transformation and pointing to a significant physiological impact of these pumps (27). Interestingly, in the ATCC 17978 strain, the adeAB operon lacks the adeC gene encoding the OMF (28). Furthermore, there are two adeA genes (A1S_1751 and A1S_1752) and one adeB gene (A1S_1750) in the operon. This modified AdeA1A2B pump appears to be functional, and its expression correlates with changes in tigecycline resistance (28).

Mutants overproducing AdeFGH were selected in vitro upon exposure of hypersusceptible A. baumannii strains lacking adeABC and adeIJK to chloramphenicol or norfloxacin (29, 30). This pump appears to have the narrowest substrate specificity among these three characterized RND pumps. Surprisingly, neither deletion nor overproduction of AdeFGH changes the bacterial susceptibility in the presence of AdeABC or AdeIJK, but AdeFGH overproduction appears to complement the hypersusceptibility of efflux-deficient strains (24, 31). Despite the low effect on antibiotic susceptibilities, AdeFGH was reported as one of the prevalent overproduced pump in clinical isolates in different geographical regions (30, 32, 33).

We recently showed that the RND efflux pumps of A. baumannii function synergistically with the low permeability barrier of the outer membrane (12). The AbΔ3 mutant lacking all three adeAB, adeIJK, and adeFGH operons was more susceptible to a variety of antibiotics than its parental ATCC 17978 strain. However, for some antibiotics these changes in susceptibilities were synergistic with hyperporination of the outer membrane, suggesting that some of the pump activities could be dependent on the permeability properties of the outer membrane. In this study, we constructed A. baumannii ATCC 17978 derivatives either lacking or overproducing individual efflux pumps in the genetic background lacking antibiotic exposure changes characteristic for clinical isolates and analyzed their properties in the context of the outer membrane with different permeability properties. We report that all three pumps are functional in ATCC 17978. Only a few antibiotics benefit from overproduction of either one of these pumps and activities of these antibiotics are strongly dependent on synergistic interactions between efflux pumps and the outer membrane permeability barrier. The overexpression of efflux pumps has a dramatic pump-specific effect on A. baumannii physiology and metabolism, implicating each pump in specific physiological functions.

RESULTS

Deletions of efflux pumps lead to strain-specific changes in antibiotic susceptibility profiles.The ATCC 17978 derivative, strain AbWT, was used to construct a series of efflux pump knockout mutants (Fig. 1; see also Table S2 in the supplemental material). Inactivation of either one of the three RND pumps did not notably affect the growth rate of A. baumannii (see Fig. S1 in the supplemental material). Antibacterial activities of antibiotics were unaffected in ΔAB strain with a chromosomal deletion of adeA1A2B operon (see Table S2 in the supplemental material), suggesting that the AdeA1A2B pump, even if expressed, does not contribute to intrinsic levels of antibiotic resistance. Surprisingly, a chromosomal deletion of adeFGH (ΔFGH) decreased the susceptibility of AbWT to specific antibiotics such as azithromycin (16-fold increase), ciprofloxacin (4-fold), gentamicin (4-fold), and zeocin (>8-fold) (Fig. 1; see also Table S2 in the supplemental material). These results suggest that on one hand, AdeFGH does not contribute to the intrinsic levels of antibiotic resistance and on the other, its deletion activates the expression of another pump, which protects ΔFGH cells from specific antibiotics. Interestingly, the double-knockout mutant ΔΑΒΔFGH has an antibiotic susceptibility phenotype similar to that of ΔAB, suggesting that the latter could be responsible for the changes in antibiotic susceptibility of ΔFGH (see Table S2 in the supplemental material). The protein alignment analyses showed that the previously described AdeA1 and AdeA2 are in fact the N-terminal 167 amino acid residues and the C-terminal 143 amino acid residues that are 98 and 99% identical to the corresponding domains of AdeA from A. baumannii clinical strain AYE. Cloning of the adeAB operon into an expression vector confirmed that ATCC 17978 carries an intact adeA gene coexpressed with adeB (see below).

FIG 1
FIG 1

Fold changes in MICs in A. baumannii efflux-deficient strains compared to the WT (MICmutant/MICWT, n = 3). Abbreviations: AZI, azithromycin; ERY, erythromycin; NAL, nalidixic acid; LEV, levofloxacin; CIP, ciprofloxacin; KAN, kanamycin; GM, gentamicin; MER, meropenem; CAR, carboxycillin; CLO, cloxacillin; AMP, ampicillin; TET, tetracycline; TIG, tigecycline; MIN, minocycline; NOV, novobiocin; RIF, rifampin; ZEO, zeocin; VAN, vancomycin; CF, chloramphenicol; TRI, triclosan; HT, Hoechst 33342; TMP, trimethoprim; SDS, sodium dodecyl sulfate.

As expected inactivation of AdeIJK (ΔIJK) led to the hypersusceptibility to almost all tested antibiotics (Fig. 1; see also Table S2 in the supplemental material). The exceptions are aminoglycosides kanamycin and gentamicin, azithromycin, vancomycin, and trimethoprim. The double-knockout mutants ΔΑΒΔIJK and ΔFGHΔIJK had antibiotic susceptibility profiles identical to that ΔIJK (see Table S2 in the supplemental material). Likewise, the antibiotic susceptibility profile of AbΔ3 mutant with all three efflux pumps knocked out is dominated by the lack of AdeIJK (see Table S2 in the supplemental material). Thus, the activity of AdeIJK defines the intrinsic levels of antibiotic resistance in AbWT.

Overproduced efflux pumps have complementary substrate specificities.We next constructed plasmids carrying individual efflux pumps under the control of an arabinose-inducible promoter. These plasmids integrate the expression cassette onto the A. baumannii chromosome and also self-replicate in the constructed strains (12). The plasmids were introduced into AbΔ3 and in some cases into double-knockout ΔABΔFGH and ΔABΔIJK variants, and their integration onto chromosome and self-replication was confirmed by a set of genetic, biochemical, and functional assays. The adeFGH and adeIJK operons were cloned from ATCC 17978 strain. To analyze the functionality of AdeAB, in addition to the native adeAB operon from ATCC 17978 (AdeAB17978), we also cloned the adeABC operon from the A. baumannii AYE strain (AdeABCAYE) (34). As in the case of adeA (see above), adeB genes from ATCC 17978 and AYE strains are 98% identical to each other.

Protein profile analyses showed the presence of additional proteins with a molecular mass of ∼45 kDa in membrane fractions isolated from the cells overproducing AdeABCAYE, AdeAB17978, AdeIJK, or AdeFGH (Fig. 2A). Based on the apparent size, the detected proteins could correspond to MFPs AdeA, AdeI, and AdeF, respectively. Indeed, liquid chromatography-tandem mass spectrometry (LC-MS/MS) analyses of the intact protein, as well as the trypsin digestion, confirmed that the major band in membranes isolated from the cells with AdeABCAYE corresponds to AdeA. Therefore, both AdeABCAYE and AdeAB17978 are significantly overproduced in cells, whereas the expressions of AdeIJK and AdeFGH are lower, especially the latter, but still detectable in membrane fractions (Fig. 2A). No other significant changes could be seen in the protein profiles.

FIG 2
FIG 2

Protein expression profiles and morphology of A. baumannii cells overproducing different efflux pumps. (A) SDS-PAGE analyses of membrane fractions isolated from indicated cells. WT, Ab-ARA(pTJ1); Δ3, AbΔ3-ARA(pTJ1); AB, AbΔ3-ARA(pAdeAB17978); ABC, AbΔ3-ARA(pAdeABCAYE); FGH, AbΔ3-ARA(pAdeFGH); IJK, AbΔ3-ARA(pAdeIJK). Cells either contained an empty expression cassette (No Pore) or the cassette with a gene encoding the Pore. Protein bands corresponding to MFPs are indicated by asterisks. (B) Growth rates of A. baumannii strains overproducing the indicated efflux pumps in the presence or absence of the Pore. (C) Phase-contrast microscopy of the exponentially growing A. baumannii cells. Strains are abbreviated as in panel A.

Overproduction of AdeABCAYE but not of the other efflux pumps in AbΔ3 slightly reduced the growth rate of these cells (Fig. 2B). In addition, only the induced AbΔ3(pAdeABCAYE) cells were morphologically different from other efflux pump overproducers and deletion mutants. Unlike typical coccobacilli of A. baumannii ATCC 17978, cells overproducing AdeABCAYE were more spheroidal and varied in size (Fig. 2C). This growth defects and morphology were specific to AbΔ3(pAdeABCAYE), whereas AbΔ3(pAdeAB17978) cells were morphologically similar to other A. baumannii strains. This result suggests that overproduction of AdeC could be the reason for morphological and growth changes.

Overproduction of AdeABCAYE complemented the hypersusceptibility phenotype of AbΔ3 to several classes of antibiotics (Fig. 3 and Table 1; see also Table S3 in the supplemental material). Among its substrates are macrolides, both azithromycin and erythromycin, fluoroquinolones, tetracyclines, and zeocin. For the majority of these antibiotics, overproduction of AdeABCAYE increased the MICs to the same levels or below those of the WT cells. The exceptions are azithromycin and zeocin, the MICs of which in AbΔ3(pAdeABCAYE) cells were significantly greater than the MICs in the wild type (WT). Azithromycin stands apart because its MIC is 32-fold higher in cells overproducing AdeABCAYE than in AbWT, which contrasts with only a 2- to 4-fold MIC change for zeocin. In contrast to previous reports, overproduction of AdeABC does not provide resistance to chloramphenicol and aminoglycosides, as seen from the lack of significant changes in the MICs of these antibiotics.

FIG 3
FIG 3

Fold changes in MICs for A. baumannii strains overproducing efflux pumps compared to AbWT (A) and efflux-deficient strain AbΔ3 (B). MICover/MICparent, n = 3. Antibiotics are abbreviated as defined in the legend to Fig. 1, with the following additions: AMI, amikacin; TOB, tobramycin; DOX, doxycycline.

View this table:
TABLE 1

Fold changes in MICs for A. baumannii efflux-deficient and efflux-overproducing strainsa

Interestingly, the overproduction of AdeAB17978 failed to complement the hypersusceptibility phenotype of AbΔ3 (Table 1), indicating that in this genetic background, the OMF required for the activity of this efflux pump is missing. Furthermore, the construct failed to complement the hypersusceptibility phenotypes of the double ΔABΔIJK mutant, but it was functional in ΔABΔFGH cells. This result suggests that the OMF AdeK can be engaged by AdeAB into a functional complex. A comparison of MICs in strains with functional AdeABCAYE and AdeAB17978 shows that both constructs enable high levels of resistance to azithromycin, erythromycin, and zeocin (Table 1). This substrate specificity matches that of the efflux pump overproduced in ΔFGH, providing further evidence that the inactivation of AdeFGH in AbWT triggers overproduction of AdeAB17978.

AbΔ3(pAdeFGH) cells carrying a plasmid-borne AdeFGH were less susceptible to ciprofloxacin (8- to 16-fold increase in MIC), nalidixic acid (2- to 4-fold), novobiocin (4-fold), chloramphenicol (2- to 4-fold), and triclosan (4-fold) (Fig. 3; see also Table S2 in the supplemental material). However, only the MICs of fluoroquinolones and chloramphenicol matched the respective MICs in WT cells. For all other antibiotics, MICs were below the WT levels. Hence, AdeFGH is expressed, functional and differs from AdeAB(C/K) in its substrate specificity.

The AdeIJK pump again was the most robust, and its overproduction reduced susceptibility not only to antibiotics that are affected by overproduction of AdeABC and AdeFGH but also to meropenem, Hoechst 33342 (HT), and sodium dodecyl sulfate (SDS) (Fig. 3; see also Table S3 in the supplemental material). Surprisingly, overproduction of AdeIJK increases the MIC of erythromycin at and above the levels of the WT cells but fails to increase the MIC of azithromycin. This pump is also the most efficient in efflux of fluoroquinolones, nalidixic acid, meropenem, and triclosan.

Taken together, these results show that the constructed strains overproduce the specific efflux pumps. These pumps provide complementary protection against antibiotics belonging to different classes, but only for select antibiotics does efflux pump overproduction increase the MICs above WT levels.

Contribution of the outer membrane barrier in antibacterial activities correlates with the expression of efflux pumps.In Gram-negative bacteria, the intracellular accumulation of antibiotics is defined by two kinetic constants: the barrier constant (B), which is the ratio of total passive flux and active efflux across the cell envelope, and the efflux constant (KE), which is the measure of efflux efficiency (35). To characterize relationships between efflux pumps and the OM barrier in A. baumannii, we constructed efflux-deficient and efflux-overproducing strains with the controlled permeability of the OM. In these strains, the B constant is reduced due to the expression of a large pore (Pore) that increases diffusion of antibiotics across the OM (12). On the other hand, the deletion and overexpression of efflux pumps change both B and KE. To enable the co-overproduction of the Pore and efflux pumps, the Pore was integrated onto the chromosomes of A. baumannii strains, whereas efflux pumps were overproduced from plasmids. The quantitative immunoblotting analyses showed that upon induction, the hyperporinated cells contain ca. 25 to 30 copies of the Pore per cell, which is ∼5 times lower than the previously reported pTJ1-mediated expression (12). The expression of the Pore did not affect significantly the expression of MFPs, as judged from the protein profiles of isolated membrane fractions (Fig. 1A) and had no significant effect on the growth rate of the cells (see Fig. S1 in the supplemental material).

In agreement with the previous studies, hyperporination of the AbWT strain increased susceptibility only to large antibiotics such as erythromycin, bacitracin, and vancomycin and to β-lactams such as meropenem and carbenicillin (see Table 3; see also Table S4 in the supplemental material). Activities of the same antibiotics were also potentiated in the hyperporinated efflux-deficient strains lacking different combinations of efflux pumps, suggesting that the OM barrier significantly contributes to activities of these antibiotics independently of active efflux (see Table 3). Surprisingly, inactivation of any one of the three efflux pumps resulted in the loss of contribution of the OM in the activity of meropenem, suggesting changes in the cell permeability for this antibiotic. It is possible that the efflux of meropenem is increased in single and double mutants of adeAB and adeFGH due to the increased activity of AdeIJK in these mutants.

Overproduction of single efflux pumps in AbΔ3 cells led to a stronger potentiation of antibiotic activities by hyperporination and affected larger number of antibiotics (Tables 2 and 3). Furthermore, the effect of hyperporination was efflux pump specific. Hyperporination of AdeIJK-overproducing cells strongly potentiated antibiotics, the activities of which are affected by either AdeIJK inactivation or overproduction, e.g., ciprofloxacin and tetracycline. In addition, the hyperporination-mediated potentiation of the activities of azithromycin and rifampin was stronger in AdeIJK overproducers than in AbWT or efflux deletion mutants (Table 3). Apparently, the weak efflux of these antibiotics by AdeIJK is amplified when the pump is overproduced, therefore increasing the B constant for azithromycin and rifampin. The antibacterial activities of novobiocin, doxycycline, and levofloxacin were specifically potentiated by hyperporination of cells overproducing AdeFGH. Hyperporination of AdeABC-overproducing cells affected a smaller number of antibiotics, among which were macrolides, tetracycline, and rifampin.

View this table:
TABLE 2

MICs for A. baumannii strains overproducing efflux pumps in cells with native and hyperporinated outer membranesa

View this table:
TABLE 3

Fold changes in MICs for A. baumannii wild-type and efflux-deficient and efflux-overproducing strains and their hyperporinated variants

Taken together, these results suggest that overproduction of efflux pumps increases efflux efficiency for even weak substrates of the pumps. The synergistic interactions with the OM membrane amplify the contribution of active efflux and create a more efficient permeability barrier for a wider range of structurally diverse antibiotics.

Intracellular accumulation of NPN is affected by both active efflux and outer membrane permeability.We previously found that the fluorescent probe N-phenyl-naphtylamide (NPN) is a sensitive reporter of changes in bacterial permeability barriers induced either by hyperporination or efflux inactivation (12, 35). The fluorescence of NPN is enhanced when the probe intercalates into the hydrophobic cores of membrane bilayers and is highly sensitive to changes in the permeability properties of OM. We next used NPN to analyze changes in permeability barriers of efflux-deficient and -overproducing cells with intact and hyperporinated membranes.

The intracellular accumulation of NPN in ΔAB cells in the presence or absence of the Pore was similar to that in the parent AbWT, suggesting that inactivation of AdeAB17978 does not lead to changes in the permeability properties of the cell envelope (Fig. 4A and B; see also Fig. S2 in the supplemental material). Unlike results for cells with high expression of the Pore (12), a moderate hyperporination of AbWT and ΔAB did not increase the uptake rate or steady-state levels of NPN. Hence, in both strains efflux efficiency is high enough to maintain the permeability barrier against NPN even with the increased influx of NPN across the outer membrane. The deletion of AdeFGH only slightly increased uptake of NPN (Fig. 4C). However, even such small reduction in efflux was sufficient for hyperporination to further reduce the barrier and to increase intracellular accumulation of NPN in ΔFGH cells. The NPN accumulation in ΔABΔFGH was similar to that in ΔAB cells (Fig. 4D). In all of these mutants the initial rates of NPN accumulation were too fast to quantify them (see Fig. S2 in the supplemental material), but the differences in the steady-state levels show the differences in efflux efficiencies between the strains.

FIG 4
FIG 4

Intracellular accumulation of the membrane fluorescent probe NPN in A. baumannii strains lacking different efflux pumps. (A to D) Concentration dependencies of the intracellular steady-state accumulation levels of NPN in the indicated strains. (E and F) Steady states and initial rates of NPN accumulation in ΔIJK cells. (G and H) Steady states and initial rates of NPN accumulation in AbΔ3 cells.

The inactivation of AdeIJK resulted in a dramatic >10-fold increase in both rates and steady-state accumulation levels of NPN. Thus, AdeIJK is the major pump responsible for efflux of NPN. In agreement, the intracellular accumulation of NPN in AbΔ3 cells was similar to that in ΔIJK. In both mutants, we were able to determine both the initial rates of NPN uptake and its intracellular steady-state levels (Fig. 4E to H). Hyperporination increased the initial rates of NPN uptake by 3 to 4 times, but the steady-state levels were the same in both ΔIJK and AbΔ3 cells with and without the Pore. Thus, inactivation of AdeIJK diminishes the active efflux of NPN, but the OM barrier still limits the rate of NPN accumulation.

In cells overproducing either AdeFGH or AdeIJK, the intracellular levels of NPN were reduced 10-fold and were significantly lower than those in AbΔ3 and AbΔ3(pAdeABCAYE) (Fig. 5A; see also Fig. S3 in the supplemental material). However, only hyperporinated AbΔ3-Pore(pAdeIJK) cells were able to maintain the same steady-state levels of NPN, whereas hyperporination of AbΔ3-Pore(pAdeFGH) was sufficient to reduce the barrier by at least 2-fold (Fig. 5B). This difference in efflux efficiency could be explained by finding that the amounts of overproduced AdeFGH are lower than those of AdeIJK (Fig. 2A). It is also possible that the differences in NPN affinities between the two pumps also contribute to different efflux efficiencies.

FIG 5
FIG 5

Intracellular accumulation of the membrane fluorescent probe NPN in A. baumannii strains overproducing different efflux pumps with (Pore) and without (No Pore) hyperporination of the outer membrane.

The NPN steady-state levels in AbΔ3(pAdeABCAYE) were the same as in AbΔ3 cells (Fig. 5C and D). Surprisingly, the initial rates of NPN accumulation were at least 3 times faster in cells overproducing AdeABCAYE than in efflux-deficient AbΔ3 cells. Hyperporination increased the initial rates of NPN accumulation in AbΔ3 but not in AbΔ3-Pore(pAdeABCAYE), suggesting that the OM barrier does not limit the diffusion of NPN in cells overproducing AdeABCAYE. This result further suggests that overproduction of AdeABCAYE makes the outer membrane of AbΔ3 cells more permeable to the hydrophobic NPN but at the same time enables high levels of resistance to azithromycin, zeocin, and several other antibiotics (Table 1; see also Table S3 in the supplemental material).

Lipid profiles of A. baumannii follow the changes in efflux efficiencies and permeability barriers.The results described above showed that changes in the expression of efflux pumps and hyperporination affect the permeability barriers in A. baumannii strains. We next investigated lipid composition of isogenic wild-type AbW-ARA(pTJ1), its efflux deficient AbΔ3-ARA(pTJ1) derivative, and AbΔ3 overproducing the three efflux pumps using ultrahigh-performance liquid chromatography-quadrupole time of flight mass spectrometry (UPLC/Q-TOF-MS). All of these strains contain an insertion in the chromosomal attTn7 site and pTJ1-based plasmids, which carry either an empty expression cassette with an arabinose-controllable promoter or the cassette with an efflux pump operon (see Table S5 in the supplemental material). In addition, Ab-ARA-EcPore(pPore) and AbΔ3-ARA-EcPore(pPore) with arabinose-controllable hyperporination (12) were included to investigate the effect of hyperporination. The whole cells were extracted with chloroform–methanol–phosphate-buffered saline (PBS), followed by a second chloroform extraction to enrich the cell extracts with lipids. On average, ∼2,560 peaks were present on chromatograms of different strains, ∼1,400 of which were identified in the KEGG database, including ∼470 molecules classified as lipids, ∼400 peptides, and ∼530 metabolites (see Tables S5 and S6 in the supplemental material).

The ∼1,400 metabolites clearly separate the parent AbWT with and without Pore strains from efflux-deficient and -overproducing derivatives into two well-defined clusters, as seen both on the clustered heatmap and principal component analysis plots (Fig. 6). The first two principal components PC1 and PC2 account for 54% variability in the data sets. The separation of the wild-type variants from other strains is mainly defined by PC1 (40.4%), whereas PC2 (13.9%) separates AbΔ3(+/−Pore) from its AdeABC, AdeIJK, and AdeFGH overproducers. The AbΔ3(pAdeIJK) metabolome is the closest to that of AbΔ3(+/−Pore). However, metabolomes of all overproducers are more similar to each other and AbΔ3(+/−Pore) and equally distant from AbWT(+/−Pore). Overall, hyperporination of AbWT and AbΔ3 cells resulted only in relatively small changes with 10% and 8% metabolites, respectively, affected by 1.5-fold or more. Both strains grouped with the respective nonhyperporinated parental strains. In contrast, inactivation of efflux in AbΔ3(+/−Pore) leads to the largest changes and affects ∼20% total metabolites. Overproduction of either one of the three efflux pumps results in the intermediate level of changes, with the overproduction of AdeABCAYE affecting ∼15% of metabolites. Hence, either inactivation or overproduction of efflux pumps, but not hyperporination of the OM, generates strong metabolic responses.

FIG 6
FIG 6

Metabolic changes in different A. baumannii strains. (A) Principal component analysis. Technical and biological repeats are shown as smaller circles, and averaged data are shown as larger circles. (B) Heatmap of the abundances of total identified metabolites. (C) Percentages of metabolites by classes, the abundances of which changed by 50% or more in the indicated strains. Strains: WT, Ab-ARA(pTJ1); Δ3, AbΔ3-ARA(pTJ1); ABC, AbΔ3-ARA(pAdeABCAYE); FGH, AbΔ3-ARA(pAdeFGH); IJK, AbΔ3-ARA(pAdeIJK); WT-P, Ab-ARA-EcPore(pPore); Δ3-P, AbΔ3-ARA-EcPore(pPore).

Lipids constitute ∼30% of the identified peaks with significant representation of fatty acyls and glycerophospholipids (274 of 407 lipid molecules [see Table S5 in the supplemental material]). Glycerophospholipids phosphatidylethanolamine (PE) and phosphatidylglycerol (PG) are the major lipids of A. baumannii membranes (36). The amounts of PE-containing lipids were only modestly, if at all, affected by hyperporination or changes in active efflux. In contrast, abundances of negatively charged lipids such as PG, phosphatidylserine (PS), and phosphatidic acid (PA) varied significantly between the strains (Fig. 7). In general, PGs with longer-chain fatty acids (>32 carbons) were depleted in efflux-deficient and -overproducing strains with some shorter-chain PGs accumulated in cells overproducing AdeABC, AdeIJK, and AdeFGH (Fig. 7). Majority of the identified PAs contained fatty acids with >34 carbons. Inactivation of efflux and hyperporination of AbWT cells, as well as the overproduction of either one of the three efflux pumps in AbΔ3, reduced the amounts of PA in cells. These results suggest that changes in the abundance of PGs and PAs reflect alterations in the protein content or permeability of cellular membranes and are not related to specific roles of the efflux pumps. In agreement with this conclusion, the overproduction of any of the three efflux pumps resulted in similar trends in the abundance ratios of lipids (Fig. 6B and 7).

FIG 7
FIG 7

Fold change in amounts of identified glycerophospholipids in different strains. (A) Changes in lipid compositions due to hyperporination and efflux inactivation. The classes of lipids are indicated by longer to shorter fatty acid chains from top to bottom. PS, phosphatidylserine; PG, phosphatidylglycerol; PE, phosphatidylethanolamine; PC, phosphatidylcholine; PA, phosphatidic acid; Lyso, single-chain glycerophospholipids. (B) Same as for panel A but showing changes in lipid compositions due to the overproduction of indicated efflux pumps. Strains are abbreviated as in Fig. 6.

Along with the major lipids, PS lipids were identified in the metabolomes of A. baumannii before (37). Both hyperporination and efflux inactivation in AbWT cells affected the abundance of these lipids. In contrast, neither hyperporination nor efflux overproduction led to a significant changes of PS lipids in AbΔ3 cells. In addition, we identified the presence of phosphatidylcholine (PC) lipids. The identities of these and several other lipids that differentiate hyperporinated and efflux variants were confirmed by MS/MS analyses (see Fig. S4 in the supplemental material). The levels of PCs were significantly perturbed in the strain-specific manner (Fig. 6 and 7). In particular, efflux-deficient AbΔ3(+/−Pore) cells were depleted of these lipids, whereas overproduction of AdeABC or AdeIJK restored the abundance of these lipids to the AbWT(+/−Pore) levels (Fig. 7). A similar specificity could be seen with some fatty acids, such as palmitoleamide (Fig. 7), the depletion of which in AbΔ3 and AbΔ3-Pore cells is reversed by the overproduction of AdeIJK and, to a lesser extent, AdeABC.

Thus, these results show that hyperporination, inactivation, and overproduction of efflux pumps lead to perturbations in the lipid composition of A. baumannii cells. Although changes in major phospholipids PGs and PAs contents are nonspecific to efflux pumps or hyperporination, other phospholipids such as PS and PC are strain specific. In particular, PCs are specific markers of AdeABC and AdeIJK functions, whereas PSs are reflective of the compromised permeability of cell envelope either by hyperporination of the outer membrane or by efflux inactivation.

DISCUSSION

RND-type efflux pumps undoubtedly play an important role in the development of clinical antibiotic resistance in A. baumannii. Here, we analyzed functional interactions of the three known RND pumps of A. baumannii with the permeability barrier of the outer membrane, identified their antibiotic substrates and separated the roles of these efflux pumps in bacterial physiology and antibiotic resistance. We used the antibiotic-susceptible ATCC 17978 derivatives that do not contain resistance islands identified in clinical isolates and were not exposed to high concentrations of antibiotics (34, 38). Our results show that the role of efflux pumps in antibiotic resistance is more complex and depends on a specific physiological context.

AdeABC is believed not to be expressed in natural isolates, and the multidrug resistance phenotypes in clinical isolates are due to its overexpression (11, 39). The ATCC 17978 strain contains only two of the genes of the operon adeA and adeB, but both genes are complete and produce a functional pump. The overexpression of the chromosomal AdeAB17978 in ΔFGH cells, as well as its plasmid-borne copy in ΔABΔFGH(pAdeAB17978), increased the MICs of several specific antibiotics (Table 1). Apparently, AdeAB17978 engages the outer membrane channel AdeK for its activity, as seen from the loss of activity in cells lacking adeIJK (Table 1) and as was previously reported for a recombinant AdeABK pump expressed in Escherichia coli (40).

The substrate preferences of AdeAB17978 and its homolog AdeABCAYE are similar and distinct from other pumps. Azithromycin and zeocin are the most strongly affected by the overproduction of either one of the two constructs. Erythromycin is also affected by both variants, albeit to a lesser extent (Table 1). Importantly, the overproduction of AdeABCAYE or AdeAB17978 not only complemented the hypersusceptible phenotype of the efflux-deficient A. baumannii strains but also increased the MICs of zeocin and azithromycin above the wild-type levels, suggesting the high efficiency of efflux of these antibiotics and a potential major contribution to clinical resistance against these antibiotics. In addition, the overproduction of AdeABCAYE, but not AdeAB17978, partially complemented the hypersusceptibility of AbΔ3 cells to SDS, fluoroquinolones, and tetracyclines. We did not find significant changes in susceptibilities to aminoglycosides and chloramphenicol, the antibiotics suggested to be the substrates of AdeABC based on studies of antibiotic-resistant clinical isolates or strains overproducing the pump as a result of selection on drug concentration gradients (11). Thus, AdeAB(C/K) is essential but not sufficient for resistance against these antibiotics, and clinical levels of resistance likely involve additional mechanisms. The overexpression of AdeABCAYE also makes cells leaky for the hydrophobic probe NPN (Fig. 5), affects cell growth and morphology (Fig. 2), and has the strongest impact on the metabolic makeup of AbΔ3 cells (Fig. 6). Some of these responses could be due to the high levels of expression and the presence of AdeC, which is not encoded on the ATCC 17978 chromosome. Others are likely to be specific to the physiological function of the AdeAB(C/K) pump. Interestingly, AdeABC seems to be always expressed at levels higher than other pumps (29). Our results show that such relatively higher levels are also achieved when pumps are expressed under the control of the same promoter in an inducer-controlled manner (Fig. 2A). Hence, the differences in efflux pump expression based on evaluation of transcription alone may not reflect the full extent of changes in the expression of this pump.

Our results show that A. baumannii cells need either AdeAB(C/K) or AdeFGH to be present, since inactivation of AdeFGH leads to the overexpression of AdeAB17978 (see Table S2 in the supplemental material). This result implies cross-regulation of the two operons and their distinct from AdeIJK roles in bacterial physiology. Indeed, despite significant differences in the levels of expression (Fig. 2), the metabolic changes induced by overproduction of either AdeABC or AdeFGH are quite similar and distinct from those in AdeIJK-overproducing cells (Fig. 6). The plasmid-borne expression of AdeFGH although low, complements the susceptibility of AbΔ3 cells to a broader range of antibiotics than AdeAB(C/K) (see Table S3 in the supplemental material and Fig. 3B). However, the only antibiotic exclusively affected by AdeFGH but not the other two pumps is chloramphenicol. Unlike AdeAB(C/K), AdeFGH reduces the accumulation of NPN and does not have significant growth phenotype, but its metabolome profile is unique and further distinguishes it from other pumps.

In agreement with previous studies, we found that AdeIJK is constitutively expressed in ATCC 17978 and provides protection against various antibiotics. It demonstrates the lowest specificity toward substrates among the three pumps and yet is quite specific to antibiotics within a given class. For example, the overproduction of AdeIJK significantly increases the MICs of erythromycin, at least 2-fold above the MIC in AbWT cells. However, this pump completely fails with azithromycin (see Table S2 in the supplemental material). Its inactivation strongly affects the activities of carbenicillin and cloxacillin, but the effect on ampicillin is modest. AdeIJK is a powerful protector against the accumulation of NPN. Importantly, AdeIJK expression alone fails to restore or bring the metabolome of AbΔ3 cells closer to the wild-type cells, suggesting that the impact of inactivation of all three pumps on cell physiology is much deeper than the loss of AdeIJK function. Further studies are needed to establish regulatory links between changes in the expression of efflux pumps and changes in the abundance of specific lipids and peptides.

One of the surprising findings of this study is that hyperporination alone has a limited impact on A. baumannii antibiotic susceptibilities and metabolites. This result shows that in ATCC 17978 cells the permeability barrier is relatively low compared to other Gram-negative species. More powerful efflux in strains overproducing efflux pumps increases the permeability barrier for the specific substrates of efflux pumps by altering both B and KE constants. Importantly, even if antibiotics are only poorly expelled by efflux pumps, such as rifampin or azithromycin by AdeIJK, but diffuse across the outer membrane slowly, the overexpression of the efflux pump effectively reduces their intracellular concentrations, as seen from the amplified effects of hyperporination on antibacterial activities (see Table S4 in the supplemental material). Comparison of MICs of antibiotics in strains with different permeability barriers (Fig. 1 and 3) shows that the role of active efflux in antibiotic resistance in A. baumannii is limited to a few select antibiotics and that the impact of efflux pump overproduction on antibiotic susceptibility is significantly lower than the previously reported for clinical isolates (11, 23, 24, 30, 41). These results point to the involvement of additional mechanisms of resistance, in particular those that improve the permeability barriers of bacterial cells and act synergistically with active efflux pumps.

MATERIALS AND METHODS

Strains and growth conditions.The strains and plasmids used in this study are listed in Table S1 in the supplemental material. Luria-Bertani (LB) broth (10 g of Bacto tryptone, 5 g of yeast extract, and 5 g of NaCl per liter [pH 7.0]) or LB agar (LB broth supplemented with 15 g/liter of agar) was used for bacterial growth. When indicated, cultures were induced with 1% l-arabinose to induce the expression of “Pore” or efflux pump proteins. For selection, gentamicin (30 μg/ml), kanamycin (30 μg/ml), carbenicillin (200 μg/ml), trimethoprim (100 μg/ml), and streptomycin (100 μg/ml) were used.

A. baumannii ΔadeIJK, ΔadeFGH, and ΔadeAB deletion strains were constructed as described previously (12). To construct hyperporinated strains, the suicide delivery vector carrying the FhuAΔC/Δ4L gene and the helper plasmid pTNS3 were electroporated into A. baumannii AbWT and all deletion strains and grown for 1 h in LB medium containing 1 mM glucose as described previously (12). The cells were then plated on LB agar containing kanamycin (30 μg/ml), followed by incubation for 16 h at 37°C. The resulting colonies were selected and confirmed for the insertion of the gene encoding the Pore by PCR using Ab(17978) glmSFWD (5′-TTCGCTGATGAAAATAGTGG-3′) and Ab(17978) glmSREV (5′-ATTCACCTCAAACCGTACAACG-3′) primers.

Construction of plasmids for overproduction of efflux pumps.To amplify genes encoding AdeABCAYE, AdeAB17978, AdeFGH, and AdeIJK efflux pumps, three sets of primers were designed with the following restriction sites for cloning: PciI and KpnI were used for adeAB and adeABC, NcoI and KpnI were used for adeFGH, and NcoI and SmaI were used for adeIJK. The adeAB, adeFGH, and adeIJK DNA fragments were amplified by PCR using the genomic DNA of A. baumannii ATCC 17978 as a template. The adeABC fragment was amplified using A. baumannii AYE (42) as a template. We previously found that the cloning vector pTJ1 replicates in A. baumannii ATCC 17978 and can be used for significant overexpression of genes from both the chromosome and the plasmid (12). The corresponding PCR products were cloned into pTJ1 to generate pTJ1-AdeABCAYE and pTJ1-AdeAB17978, pTJ1-AdeIJK, and pTJ1-AdeFGH expressing the efflux pumps under the control of the arabinose-inducible PBAD promoter. After transformation, clones containing plasmids were selected on trimethoprim (100 μg/μl) plates, and successful insertions were confirmed by restriction analyses.

Triparental mating was used to construct AbΔ3 cells overproducing the individual efflux pumps. For this purpose, plasmids were transformed into SM10 cells and mated with SM10(pTNS3) and AbΔ3 as described earlier (12). The chromosomal insertions of adeAB, adeABC, adeFGH, and adeIJK were confirmed by PCR. Biparental mating was used to introduce pTJ1-adeABC, pTJ1-adeIJK, and pTJ1-adeFGH into AbΔ3-ARA and AbΔ3-ARA-Pore strains.

Drug susceptibility assay (MIC determination).Susceptibility to different classes of antibiotics was determined by 2-fold broth dilution method as described previously (43). Briefly, the cells were grown in LB broth with appropriate antibiotics wherever necessary at 37°C with shaking at 200 rpm. The MICs of various antimicrobial agents were measured in 96-well microtiter plates. For this purpose, exponentially growing cells were inoculated at a density of 105 cells per ml into wells containing LB medium in the presence of 2-fold-increasing concentrations of drugs under investigation. Cell growth was determined visually after incubation of the microtiter plates at 37°C for 16 h.

NPN uptake.Cells from frozen stocks were inoculated into LB medium containing appropriate antibiotics and incubated for 16 h at 37°C. Cells were then subcultured into 30 ml of a fresh LB medium and grown at 37°C to an optical density at 600 nm (OD600) of 0.3. The cells were then induced with 1% arabinose and grown until the OD600 was 1.0. Cells were collected by centrifugation at 3,266 × g for 30 min at room temperature, and the pellet was resuspended in 25 ml of 50 mM HEPES-KOH buffer (pH 7.0) containing 1 mM MgSO4 and 0.4 mM glucose (HMG buffer) and pelleted at 3,266 × g at room temperature. The pellet was resuspended in HMG buffer, adjusted to an OD600 of ∼1.0, and kept at room temperature during the experiment.

An uptake assay was performed in a temperature-controlled multimode microplate reader (Tecan Spark 10M) equipped with a sample injector in fluorescence mode. The fluorescence of NPN was monitored at λex-350 nm and λem-405 nm at a gain of 75 for 10 s, as described earlier (12). All measurements were done in duplicates and repeated at least three times.

Protein expression and analyses.SDS-polyacrylamide gels and quantitative immunoblotting analyses were used to analyze protein profiles and the amounts of the pore proteins in the outer membrane of A. baumannii and its derivatives. Membrane fractions were isolated from A. baumannii cells by ultracentrifugation as described before (43). Outer membrane fractions were enriched by solubilization of inner membrane proteins in buffer containing 50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, and 0.2% Triton X-100, followed by separation of the insoluble outer membrane fractions by ultracentrifugation. Proteins were normalized and analyzed by SDS–12% PAGE and staining with Coomassie brilliant blue or transferred to polyvinylidene difluoride membranes for immunoblotting with primary monoclonal anti-histidine tag antibodies (Fisher Scientific) and a secondary alkaline phosphatase-conjugated anti-mouse immunoglobulin antibody (Sigma). BCIP (5-bromo-4-chloro-3-indolylphosphate) and nitroblue tetrazolium substrates were used to visualize the bands.

Lipid extraction.Overnight cultures of A. baumannii cells were diluted in a fresh LB medium (1:100) and grown at 37°C until achieving an OD600 of 0.3. The expression of AdeABC, AdeFGH, and AdeIJK proteins was induced by the addition of arabinose to a final concentration of 1% and further incubation until cultures reached an OD600 of 0.7. Cells were pelleted by centrifugation at 4°C and three times washed with PBS. Equal amounts of cells were taken (wet cell weight) for lipid extraction. A. baumannii cells were extracted by the Bligh and Dyer method, as described elsewhere (44). Briefly, equal amounts of cells were resuspended into a monophasic mixture with chloroform, methanol, and PBS in a of 1:2:0.8 (vol/vol/vol) ratio, followed by incubation for 1 h at room temperature on a rotator. The resulting cell suspension was centrifuged at 2,600 × g for 15 min at room temperature to remove cell debris. The supernatant was transferred to a fresh tube and made into a biphasic mixture with equal volumes of chloroform and PBS. The resulting suspension was vortexed and centrifuged as described above at 2,600 × g for 15 min at room temperature. The lower organic phase enriched with lipids was collected carefully after removal of the upper aqueous phase. The organic phase was dried under N2, and the lipid pellet was resuspended in chloroform.

Metabolites and lipid identification.An Agilent 1290 Infinity UPLC system coupled to an ultrahigh mass resolution quadrupole-time of flight mass spectrometer (6545 Q-TOF-MS; Agilent, Santa Clara, CA) was used for metabolomic analysis. Samples (5 μl) were processed in positive ion mode. Analysis of each sample was performed in duplicate. A SeQuant ZIC-HILIC column (5 μm, 150 by 4.6 mm; The Nest Group, Inc., Southborough, MA) was used with a flow rate of 0.3 ml/min. A linear gradient from 80 to 20% acetonitrile was used for the first 30 min, followed by 5% acetonitrile for an additional 8 min. Solvents contained 0.1% formic acid to promote positive ion formation in the electrospray. The MS parameters were as follows: ion source gas temperature, 350°C; capillary voltage, 3,500 V; fragmentor voltage, 160 V; m/z range, 50 to 1,000; data acquisition rate, 4 GHz; and 1 spectrum recorded per second.

As described previously (45, 46), raw MS data were processed using IDEOM version 19 workflow (47). IDEOM processing utilizes XCMS Centwave (48) for peak detection and mzMatch.R (49) for peak alignment between triplicates and between samples, for filtering, and for storage of the data in peakML formatted files. Feature alignment was performed with a retention time window of 0.5 min and a mass error window of 5 ppm. Scripts for XCMS (50) and mzMatch are presented in the R environment.

Detected features were matched against the IDEOM's version of the Kyoto Encyclopedia of Genes and Genomes (KEGG) metabolite database (51), as well as the Metacyc (52), Lipidmaps (53), and HMDB (54) using a mass tolerance of 2 ppm. The identification was also based on a comparison of experimental and calculated retention times. The retention time calculator is based on a quantitative structure-retention relationship model that incorporates six physicochemical variables in a multiple-linear regression based on 124 authentic standard metabolites. For the present work, 28 metabolite standards were used to calibrate the retention time predictor. Putative metabolites, whose experimentally observed retention time was more than 45% off from that calculated by the model, were rejected. With the use of an RT calculator, an identification level of 2 can be claimed (55, 56), but given the level of uncertainty associated with identification, metabolites detected here are referred to as “putative.” The IDEOM software assigns a level of confidence to each identification using a numerical value that ranges from 1 to 10, where 10 reflects the highest level of confidence. The confidence value assigned is based on the magnitude of the mass error, on the difference between experimental and expected retention times, and on a database priority order internal to IDEOM. For any feature, the metabolite with the highest confidence score is selected. The following parameters were used for the positive ESI-MS/MS experiments, precursors were collected every 2 s with a collisions energy of 10, 15, 20, and 30 eV over multiple injections within of a window of ±1 min of the targeted retention time.

Statistical method analyses were performed on the Web-based application Metaboanalyst 3.0 (57, 58). Clustered heatmaps were constructed to visualize clustering based on the Euclidean distance measurement with the Ward clustering algorithm.

Protein identification.After the protein sample has been separated by SDS-PAGE, the protein band of interest was localized for excision and elution. We used a Bio-Rad model 422 Electro-Eluter to recover proteins separated by SDS-PAGE, as described previously (59). The electroeluted protein after washing with ice-cold acetone was redissolved in 25 mM ammonium bicarbonate and trypsin digested.

Peptides were analyzed by LC-MS/MS with Agilent 1290 HPLC system and an Agilent 6545 Q-TOF-MS. The peptides were separated on an Acquity UPLC HSS C18 SB column (1.8 mm, 2.1 mm by 100 mm; Waters, Inc., Milford, MA); mobile phase A contains 95:5 (vol/vol) water-acetonitrile with 0.1% formic acid, and mobile phase B contains 100% acetonitrile with 0.1% formic acid. A flow of 0.5 ml/min was used with a gradient of 0 to 60% of mobile phase B over 60 min, with data treated using Agilent MassHunter qualitative analysis vB.06.00 and Agilent BioConfirm vB.08.00.

ACKNOWLEDGMENTS

This study was sponsored by the Department of the Defense, Defense Threat Reduction Agency (HDTRA1-14-1-0019), and by NIH/NIAID grant RO1AI132836.

The content of the information does not necessarily reflect the position or the policy of the federal government, and no official endorsement should be inferred.

FOOTNOTES

    • Received 26 January 2018.
    • Accepted 9 April 2018.
    • Accepted manuscript posted online 16 April 2018.
  • Address correspondence to Helen I. Zgurskaya, elenaz{at}ou.edu.

  • Citation Leus IV, Weeks JW, Bonifay V, Smith L, Richardson S, Zgurskaya HI. 2018. Substrate specificities and efflux efficiencies of RND efflux pumps of Acinetobacter baumannii. J Bacteriol 200:e00049-18. https://doi.org/10.1128/JB.00049-18.

  • Supplemental material for this article may be found at https://doi.org/10.1128/JB.00049-18.

REFERENCES

  1. 1.↵
    1. Fournier PE,
    2. Richet H
    . 2006. The epidemiology and control of Acinetobacter baumannii in health care facilities. Clin Infect Dis 42:692–699.
    OpenUrlCrossRefPubMedWeb of Science
  2. 2.↵
    1. Gonzalez-Villoria AM,
    2. Valverde-Garduno V
    . 2016. Antibiotic-resistant Acinetobacter baumannii increasing success remains a challenge as a nosocomial pathogen. J Pathog 2016:7318075. doi:10.1155/2016/7318075.
    OpenUrlCrossRef
  3. 3.↵
    1. Peleg AY,
    2. Seifert H,
    3. Paterson DL
    . 2008. Acinetobacter baumannii: emergence of a successful pathogen. Clin Microbiol Rev 21:538–582. doi:10.1128/CMR.00058-07.
    OpenUrlAbstract/FREE Full Text
  4. 4.↵
    1. Perez F,
    2. Hujer AM,
    3. Hujer KM,
    4. Decker BK,
    5. Rather PN,
    6. Bonomo RA
    . 2007. Global challenge of multidrug-resistant Acinetobacter baumannii. Antimicrob Agents Chemother 51:3471–3484. doi:10.1128/AAC.01464-06.
    OpenUrlFREE Full Text
  5. 5.↵
    1. Blanco N,
    2. Harris AD,
    3. Rock C,
    4. Johnson JK,
    5. Pineles L,
    6. Bonomo RA,
    7. Srinivasan A,
    8. Pettigrew MM,
    9. Thom KA
    , the CDC Epicenters Program. 2018. Risk factors and outcomes associated with multidrug-resistant Acinetobacter baumannii upon intensive care unit admission. Antimicrob Agents Chemother 62:e01631-17. doi:10.1128/AAC.01631-17.
    OpenUrlAbstract/FREE Full Text
  6. 6.↵
    1. Lob SH,
    2. Hoban DJ,
    3. Sahm DF,
    4. Badal RE
    . 2016. Regional differences and trends in antimicrobial susceptibility of Acinetobacter baumannii. Int J Antimicrobial Agents 47:317–323. doi:10.1016/j.ijantimicag.2016.01.015.
    OpenUrlCrossRef
  7. 7.↵
    1. Pagdepanichkit S,
    2. Tribuddharat C,
    3. Chuanchuen R
    . 2016. Distribution and expression of the Ade multidrug efflux systems in Acinetobacter baumannii clinical isolates. Can J Microbiol 62:794–801. doi:10.1139/cjm-2015-0730.
    OpenUrlCrossRef
  8. 8.↵
    1. Li XZ,
    2. Plesiat P,
    3. Nikaido H
    . 2015. The challenge of efflux-mediated antibiotic resistance in Gram-negative bacteria. Clin Microbiol Rev 28:337–418. doi:10.1128/CMR.00117-14.
    OpenUrlAbstract/FREE Full Text
  9. 9.↵
    1. Blair JMA,
    2. Webber MA,
    3. Baylay AJ,
    4. Ogbolu DO,
    5. Piddock LJV
    . 2015. Molecular mechanisms of antibiotic resistance. Nat Rev Microbiol 13:42–51. doi:10.1038/nrmicro3380.
    OpenUrlCrossRefPubMed
  10. 10.↵
    1. Hou PF,
    2. Chen XY,
    3. Yan GF,
    4. Wang YP,
    5. Ying CM
    . 2012. Study of the correlation of imipenem resistance with efflux pumps AdeABC, AdeIJK, AdeDE, and AbeM in clinical isolates of Acinetobacter baumannii. Chemotherapy 58:152–158. doi:10.1159/000335599.
    OpenUrlCrossRefPubMed
  11. 11.↵
    1. Coyne S,
    2. Courvalin P,
    3. Perichon B
    . 2011. Efflux-mediated antibiotic resistance in Acinetobacter spp. Antimicrob Agents Chemother 55:947–953. doi:10.1128/AAC.01388-10.
    OpenUrlAbstract/FREE Full Text
  12. 12.↵
    1. Krishnamoorthy G,
    2. Leus IV,
    3. Weeks JW,
    4. Wolloscheck D,
    5. Rybenkov VV,
    6. Zgurskaya HI
    . 2017. Synergy between active efflux and outer membrane diffusion defines rules of antibiotic permeation into gram-negative bacteria. mBio 8:e01172-17. doi:10.1128/mBio.01172-17.
    OpenUrlAbstract/FREE Full Text
  13. 13.↵
    1. Zgurskaya HI,
    2. Lopez CA,
    3. Gnanakaran S
    . 2015. Permeability barrier of gram-negative cell envelopes and approaches to bypass it. ACS Infect Dis 1:512–522. doi:10.1021/acsinfecdis.5b00097.
    OpenUrlCrossRef
  14. 14.↵
    1. Masi M,
    2. Refregiers M,
    3. Pos KM,
    4. Pages JM
    . 2017. Mechanisms of envelope permeability and antibiotic influx and efflux in Gram-negative bacteria. Nat Microbiol 2:17001. doi:10.1038/nmicrobiol.2017.1.
    OpenUrlCrossRef
  15. 15.↵
    1. Eicher T,
    2. Cha HJ,
    3. Seeger MA,
    4. Brandstatter L,
    5. El-Delik J,
    6. Bohnert JA,
    7. Kern WV,
    8. Verrey F,
    9. Grutter MG,
    10. Diederichs K,
    11. Pos KM
    . 2012. Transport of drugs by the multidrug transporter AcrB involves an access and a deep binding pocket that are separated by a switch-loop. Proc Natl Acad Sci U S A 109:5687–5692. doi:10.1073/pnas.1114944109.
    OpenUrlAbstract/FREE Full Text
  16. 16.↵
    1. Nakashima R,
    2. Sakurai K,
    3. Yamasaki S,
    4. Hayashi K,
    5. Nagata C,
    6. Hoshino K,
    7. Onodera Y,
    8. Nishino K,
    9. Yamaguchi A
    . 2013. Structural basis for the inhibition of bacterial multidrug exporters. Nature 500:102–106. doi:10.1038/nature12300.
    OpenUrlCrossRefPubMedWeb of Science
  17. 17.↵
    1. Nakashima R,
    2. Sakurai K,
    3. Yamasaki S,
    4. Nishino K,
    5. Yamaguchi A
    . 2011. Structures of the multidrug exporter AcrB reveal a proximal multisite drug-binding pocket. Nature 480:565–569. doi:10.1038/nature10641.
    OpenUrlCrossRefPubMedWeb of Science
  18. 18.↵
    1. Bavro VN,
    2. Pietras Z,
    3. Furnham N,
    4. Perez-Cano L,
    5. Fernandez-Recio J,
    6. Pei XY,
    7. Misra R,
    8. Luisi B
    . 2008. Assembly and channel opening in a bacterial drug efflux machine. Mol Cell 30:114–121. doi:10.1016/j.molcel.2008.02.015.
    OpenUrlCrossRefPubMedWeb of Science
  19. 19.↵
    1. Lobedanz S,
    2. Bokma E,
    3. Symmons MF,
    4. Koronakis E,
    5. Hughes C,
    6. Koronakis V
    . 2007. A periplasmic coiled-coil interface underlying TolC recruitment and the assembly of bacterial drug efflux pumps. Proc Natl Acad Sci U S A 104:4612–4617. doi:10.1073/pnas.0610160104.
    OpenUrlAbstract/FREE Full Text
  20. 20.↵
    1. Pei XY,
    2. Hinchliffe P,
    3. Symmons MF,
    4. Koronakis E,
    5. Benz R,
    6. Hughes C,
    7. Koronakis V
    . 2011. Structures of sequential open states in a symmetrical opening transition of the TolC exit duct. Proc Natl Acad Sci U S A 108:2112–2117. doi:10.1073/pnas.1012588108.
    OpenUrlAbstract/FREE Full Text
  21. 21.↵
    1. Mikolosko J,
    2. Bobyk K,
    3. Zgurskaya HI,
    4. Ghosh P
    . 2006. Conformational flexibility in the multidrug efflux system protein AcrA. Structure 14:577–587. doi:10.1016/j.str.2005.11.015.
    OpenUrlCrossRefPubMed
  22. 22.↵
    1. Zgurskaya HI,
    2. Weeks JW,
    3. Ntreh AT,
    4. Nickels LM,
    5. Wolloscheck D
    . 2015. Mechanism of coupling drug transport reactions located in two different membranes. Front Microbiol 6:100. doi:10.3389/fmicb.2015.00100.
    OpenUrlCrossRefPubMed
  23. 23.↵
    1. Damier-Piolle L,
    2. Magnet S,
    3. Brémont S,
    4. Lambert T,
    5. Courvalin P
    . 2008. AdeIJK, a resistance-nodulation-cell division pump effluxing multiple antibiotics in Acinetobacter baumannii. Antimicrob Agents Chemother 52:557–562. doi:10.1128/AAC.00732-07.
    OpenUrlAbstract/FREE Full Text
  24. 24.↵
    1. Yoon E-J,
    2. Nait Chabane Y,
    3. Goussard S,
    4. Snesrud E,
    5. Courvalin P,
    6. Dé E,
    7. Grillot-Courvalin C
    . 2015. Contribution of resistance-nodulation-cell division efflux systems to antibiotic resistance and biofilm formation in Acinetobacter baumannii. mBio 6:e00309-15. doi:10.1128/mBio.00309-15.
    OpenUrlAbstract/FREE Full Text
  25. 25.↵
    1. Marchand I,
    2. Damier-Piolle L,
    3. Courvalin P,
    4. Lambert T
    . 2004. Expression of the RND-type efflux pump AdeABC in Acinetobacter baumannii is regulated by the AdeRS two-component system. Antimicrob Agents Chemother 48:3298–3304. doi:10.1128/AAC.48.9.3298-3304.2004.
    OpenUrlAbstract/FREE Full Text
  26. 26.↵
    1. Ruzin A,
    2. Keeney D,
    3. Bradford PA
    . 2007. AdeABC multidrug efflux pump is associated with decreased susceptibility to tigecycline in Acinetobacter calcoaceticus-Acinetobacter baumannii complex. J Antimicrob Chemother 59:1001–1004. doi:10.1093/jac/dkm058.
    OpenUrlCrossRefPubMedWeb of Science
  27. 27.↵
    1. Lee CR,
    2. Lee JH,
    3. Park M,
    4. Park KS,
    5. Bae IK,
    6. Kim YB,
    7. Cha CJ,
    8. Jeong BC,
    9. Lee SH
    . 2017. Biology of Acinetobacter baumannii: pathogenesis, antibiotic resistance mechanisms, and prospective treatment options. Front Cell Infect Microbiol 7:55. doi:10.3389/fcimb.2017.00483.
    OpenUrlCrossRef
  28. 28.↵
    1. Lin M-F,
    2. Lin Y-Y,
    3. Yeh H-W,
    4. Lan C-Y
    . 2014. Role of the BaeSR two-component system in the regulation of Acinetobacter baumannii adeAB genes and its correlation with tigecycline susceptibility. BMC Microbiol 14:119. doi:10.1186/1471-2180-14-119.
    OpenUrlCrossRefPubMed
  29. 29.↵
    1. Coyne S,
    2. Guigon G,
    3. Courvalin P,
    4. Perichon B
    . 2010. Screening and quantification of the expression of antibiotic resistance genes in Acinetobacter baumannii with a microarray. Antimicrob Agents Chemother 54:333–340. doi:10.1128/AAC.01037-09.
    OpenUrlAbstract/FREE Full Text
  30. 30.↵
    1. Coyne S,
    2. Rosenfeld N,
    3. Lambert T,
    4. Courvalin P,
    5. Perichon B
    . 2010. Overexpression of resistance-nodulation-cell division pump AdeFGH confers multidrug resistance in Acinetobacter baumannii. Antimicrob Agents Chemother 54:4389–4393. doi:10.1128/AAC.00155-10.
    OpenUrlAbstract/FREE Full Text
  31. 31.↵
    1. Amin IM,
    2. Richmond GE,
    3. Sen P,
    4. Koh TH,
    5. Piddock LJ,
    6. Chua KL
    . 2013. A method for generating marker-less gene deletions in multidrug-resistant Acinetobacter baumannii. BMC Microbiol 13:158. doi:10.1186/1471-2180-13-158.
    OpenUrlCrossRefPubMed
  32. 32.↵
    1. He X,
    2. Lu F,
    3. Yuan F,
    4. Jiang D,
    5. Zhao P,
    6. Zhu J,
    7. Cheng H,
    8. Cao J,
    9. Lu G
    . 2015. Biofilm formation caused by clinical Acinetobacter baumannii isolates is associated with overexpression of the AdeFGH efflux pump. Antimicrob Agents Chemother 59:4817–4825. doi:10.1128/AAC.00877-15.
    OpenUrlAbstract/FREE Full Text
  33. 33.↵
    1. Fernando D,
    2. Zhanel G,
    3. Kumar A
    . 2013. Antibiotic resistance and expression of resistance-nodulation-division pump- and outer membrane porin-encoding genes in Acinetobacter species isolated from Canadian hospitals. Can J Infect Dis Med Microbiol 24:17–21.
    OpenUrlPubMed
  34. 34.↵
    1. Adams MD,
    2. Goglin K,
    3. Molyneaux N,
    4. Hujer KM,
    5. Lavender H,
    6. Jamison JJ,
    7. MacDonald IJ,
    8. Martin KM,
    9. Russo T,
    10. Campagnari AA,
    11. Hujer AM,
    12. Bonomo RA,
    13. Gill SR
    . 2008. Comparative genome sequence analysis of multidrug-resistant Acinetobacter baumannii. J Bacteriol 190:8053–8064. doi:10.1128/JB.00834-08.
    OpenUrlAbstract/FREE Full Text
  35. 35.↵
    1. Westfall DA,
    2. Krishnamoorthy G,
    3. Wolloscheck D,
    4. Sarkar R,
    5. Zgurskaya HI,
    6. Rybenkov VV
    . 2017. Bifurcation kinetics of drug uptake by Gram-negative bacteria. PLoS One 12:e0184671. doi:10.1371/journal.pone.0184671.
    OpenUrlCrossRef
  36. 36.↵
    1. Lopalco P,
    2. Stahl J,
    3. Annese C,
    4. Averhoff B,
    5. Corcelli A
    . 2017. Identification of unique cardiolipin and monolysocardiolipin species in Acinetobacter baumannii. Sci Rep 7:2972. doi:10.1038/s41598-017-03214-w.
    OpenUrlCrossRef
  37. 37.↵
    1. Maifiah MH,
    2. Creek DJ,
    3. Nation RL,
    4. Forrest A,
    5. Tsuji BT,
    6. Velkov T,
    7. Li J
    . 2017. Untargeted metabolomics analysis reveals key pathways responsible for the synergistic killing of colistin and doripenem combination against Acinetobacter baumannii. Sci Rep 7:45527. doi:10.1038/srep45527.
    OpenUrlCrossRef
  38. 38.↵
    1. Gallagher LA,
    2. Ramage E,
    3. Weiss EJ,
    4. Radey M,
    5. Hayden HS,
    6. Held KG,
    7. Huse HK,
    8. Zurawski DV,
    9. Brittnacher MJ,
    10. Manoil C
    . 2015. Resources for genetic and genomic analysis of emerging pathogen Acinetobacter baumannii. J Bacteriol 197:2027–2035. doi:10.1128/JB.00131-15.
    OpenUrlAbstract/FREE Full Text
  39. 39.↵
    1. Magnet S,
    2. Courvalin P,
    3. Lambert T
    . 2001. Resistance-nodulation-cell division-type efflux pump involved in aminoglycoside resistance in Acinetobacter baumannii strain BM4454. Antimicrob Agents Chemother 45:3375–3380. doi:10.1128/AAC.45.12.3375-3380.2001.
    OpenUrlAbstract/FREE Full Text
  40. 40.↵
    1. Sugawara E,
    2. Nikaido H
    . 2014. Properties of AdeABC and AdeIJK efflux systems of Acinetobacter baumannii compared with those of the AcrAB-TolC system of Escherichia coli. Antimicrob Agents Chemother 58:7250–7257. doi:10.1128/AAC.03728-14.
    OpenUrlAbstract/FREE Full Text
  41. 41.↵
    1. Richmond GE,
    2. Evans LP,
    3. Anderson MJ,
    4. Wand ME,
    5. Bonney LC,
    6. Ivens A,
    7. Chua KL,
    8. Webber MA,
    9. Sutton JM,
    10. Peterson ML,
    11. Piddock LJ
    . 2016. The Acinetobacter baumannii two-component system AdeRS regulates genes required for multidrug efflux, biofilm formation, and virulence in a strain-specific manner. mBio 7:e00430-16. doi:10.1128/mBio.00430-16.
    OpenUrlAbstract/FREE Full Text
  42. 42.↵
    1. Fournier P-E,
    2. Vallenet D,
    3. Barbe V,
    4. Audic S,
    5. Ogata H,
    6. Poirel L,
    7. Richet H,
    8. Robert C,
    9. Mangenot S,
    10. Abergel C,
    11. Nordmann P,
    12. Weissenbach J,
    13. Raoult D,
    14. Claverie J-M
    . 2006. Comparative genomics of multidrug resistance in Acinetobacter baumannii. PLoS Genet 2:e7. doi:10.1371/journal.pgen.0020007.
    OpenUrlCrossRefPubMed
  43. 43.↵
    1. Krishnamoorthy G,
    2. Wolloscheck D,
    3. Weeks JW,
    4. Croft C,
    5. Rybenkov VV,
    6. Zgurskaya HI
    . 2016. Breaking the permeability barrier of Escherichia coli by controlled hyperporination of the outer membrane. Antimicrob Agents Chemother 60:7372–7381. doi:10.1128/AAC.01045-15.
    OpenUrlAbstract/FREE Full Text
  44. 44.↵
    1. Garrett TA,
    2. Raetz CR,
    3. Richardson T,
    4. Kordestani R,
    5. Son JD,
    6. Rose RL
    . 2009. Identification of phosphatidylserylglutamate: a novel minor lipid in Escherichia coli. J Lipid Res 50:1589–1599. doi:10.1194/jlr.M800549-JLR200.
    OpenUrlAbstract/FREE Full Text
  45. 45.↵
    1. Bonifay V,
    2. Aydin E,
    3. Aktas DF,
    4. Sunner J,
    5. Suflita JM
    . 2017. Metabolic profiling and metabolomic procedures for investigating the biodegradation of hydrocarbons, p 111–161. In McGenity TJ, Timmis KN, Nogales B (ed), Hydrocarbon and lipid microbiology protocols: genetic, genomic, and system analyses of communities. Springer, Berlin, Germany.
  46. 46.↵
    1. Bonifay V,
    2. Wawrik B,
    3. Sunner J,
    4. Snodgrass EC,
    5. Aydin E,
    6. Duncan KE,
    7. Callaghan AV,
    8. Oldham A,
    9. Liengen T,
    10. Beech I
    . 2017. Metabolomic and metagenomic analysis of two crude oil production pipelines experiencing differential rates of corrosion. Front Microbiol 8:99. doi:10.3389/fmicb.2017.00099.
    OpenUrlCrossRef
  47. 47.↵
    1. Creek DJ,
    2. Jankevics A,
    3. Burgess KEV,
    4. Breitling R,
    5. Barrett MP
    . 2012. IDEOM: an Excel interface for analysis of LC–MS-based metabolomics data. Bioinformatics 28:1048–1049. doi:10.1093/bioinformatics/bts069.
    OpenUrlCrossRefPubMedWeb of Science
  48. 48.↵
    1. Tautenhahn R,
    2. Böttcher C,
    3. Neumann S
    . 2008. Highly sensitive feature detection for high-resolution LC/MS. BMC Bioinformatics 9:504. doi:10.1186/1471-2105-9-504.
    OpenUrlCrossRefPubMed
  49. 49.↵
    1. Scheltema RA,
    2. Jankevics A,
    3. Jansen RC,
    4. Swertz MA,
    5. Breitling R
    . 2011. PeakML/mzMatch: a File Format, Java Library, R Library, and Tool-Chain for mass spectrometry data analysis. Anal Chem 83:2786–2793. doi:10.1021/ac2000994.
    OpenUrlCrossRefPubMed
  50. 50.↵
    1. Smith CA,
    2. Want EJ,
    3. O'Maille G,
    4. Abagyan R,
    5. Siuzdak G
    . 2006. XCMS: processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching, and identification. Anal Chem 78:779–787. doi:10.1021/ac051437y.
    OpenUrlCrossRefPubMed
  51. 51.↵
    1. Kanehisa M,
    2. Goto S,
    3. Sato Y,
    4. Kawashima M,
    5. Furumichi M,
    6. Tanabe M
    . 2014. Data, information, knowledge and principle: back to metabolism in KEGG. Nucleic Acids Res 42:D199–D205. doi:10.1093/nar/gkt1076.
    OpenUrlCrossRefPubMedWeb of Science
  52. 52.↵
    1. Caspi R,
    2. Billington R,
    3. Fulcher CA,
    4. Keseler IM,
    5. Kothari A,
    6. Krummenacker M,
    7. Latendresse M,
    8. Midford PE,
    9. Ong Q,
    10. Ong WK,
    11. Paley S,
    12. Subhraveti P,
    13. Karp PD
    . 2017. The MetaCyc database of metabolic pathways and enzymes. Nucleic Acids Res 46:D633–D639. doi:10.1093/nar/gkx935.
    OpenUrlCrossRef
  53. 53.↵
    1. Fahy E,
    2. Sud M,
    3. Cotter D,
    4. Subramaniam S
    . 2007. LIPID MAPS online tools for lipid research. Nucleic Acids Res 35:W606–W612. doi:10.1093/nar/gkm324.
    OpenUrlCrossRefPubMedWeb of Science
  54. 54.↵
    1. Wishart DS,
    2. Jewison T,
    3. Guo AC,
    4. Wilson M,
    5. Knox C,
    6. Liu Y,
    7. Djoumbou Y,
    8. Mandal R,
    9. Aziat F,
    10. Dong E,
    11. Bouatra S,
    12. Sinelnikov I,
    13. Arndt D,
    14. Xia J,
    15. Liu P,
    16. Yallou F,
    17. Bjorndahl T,
    18. Perez-Pineiro R,
    19. Eisner R,
    20. Allen F,
    21. Neveu V,
    22. Greiner R,
    23. Scalbert A
    . 2013. HMDB 3.0—the human metabolome database in 2013. Nucleic Acids Res 41:D801–D807. doi:10.1093/nar/gks1065.
    OpenUrlCrossRefPubMedWeb of Science
  55. 55.↵
    1. Sumner LW,
    2. Amberg A,
    3. Barrett D,
    4. Beale MH,
    5. Beger R,
    6. Daykin CA,
    7. Fan TW,
    8. Fiehn O,
    9. Goodacre R,
    10. Griffin JL,
    11. Hankemeier T,
    12. Hardy N,
    13. Harnly J,
    14. Higashi R,
    15. Kopka J,
    16. Lane AN,
    17. Lindon JC,
    18. Marriott P,
    19. Nicholls AW,
    20. Reily MD,
    21. Thaden JJ,
    22. Viant MR
    . 2007. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI). Metabolomics 3:211–221. doi:10.1007/s11306-007-0082-2.
    OpenUrlCrossRefPubMedWeb of Science
  56. 56.↵
    1. Salek RM,
    2. Steinbeck C,
    3. Viant MR,
    4. Goodacre R,
    5. Dunn WB
    . 2013. The role of reporting standards for metabolite annotation and identification in metabolomic studies. GigaScience 2:1–3. doi:10.1186/2047-217X-2-1.
    OpenUrlCrossRefPubMed
  57. 57.↵
    1. Xia J,
    2. Wishart DS
    . 2002. Using MetaboAnalyst 3.0 for comprehensive metabolomics data analysis, current protocols in bioinformatics. John Wiley & Sons, Inc, New York, NY.
  58. 58.↵
    1. Xia J,
    2. Sinelnikov IV,
    3. Han B,
    4. Wishart DS
    . 2015. MetaboAnalyst 3.0—making metabolomics more meaningful. Nucleic Acids Res 43:W251–W257. doi:10.1093/nar/gkv380.
    OpenUrlCrossRefPubMed
  59. 59.↵
    1. Görg A,
    2. Weiss W,
    3. Dunn MJ
    . 2004. Current two-dimensional electrophoresis technology for proteomics. Proteomics 4:3665–3685. doi:10.1002/pmic.200401031.
    OpenUrlCrossRefPubMedWeb of Science
Back to top
Download PDF
Citation Tools
Print

Alerts
Email

KEYWORDS

Acinetobacter baumannii
antibiotic resistance
multidrug efflux
outer membrane
permeability barrier

Related Articles

Cited By...