JB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Aires, J. R.
Right arrow Articles by Nikaido, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Aires, J. R.
Right arrow Articles by Nikaido, H.
Journal of Bacteriology, March 2005, p. 1923-1929, Vol. 187, No. 6
0021-9193/05/$08.00+0     doi:10.1128/JB.187.6.1923-1929.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.

Aminoglycosides Are Captured from both Periplasm and Cytoplasm by the AcrD Multidrug Efflux Transporter of Escherichia coli

Julio Ramos Aires and Hiroshi Nikaido*

Department of Molecular and Cell Biology, University of California, Berkeley, California

Received 13 July 2004/ Accepted 10 September 2004


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
To understand better the mechanisms of resistance-nodulation-division (RND)-type multidrug efflux pumps, we examined the Escherichia coli AcrD pump, whose typical substrates, aminoglycosides, are not expected to diffuse spontaneously across the lipid bilayer. The hexahistidine-tagged AcrD protein was purified and reconstituted into unilamellar proteoliposomes. Its activity was measured by the proton flux accompanying substrate transport. When the interior of the proteoliposomes was acidified, the addition of aminoglycosides to the external medium stimulated proton efflux and the intravesicular accumulation of radiolabeled gentamicin, suggesting that aminoglycosides can be captured and transported from the external medium in this system (corresponding to cytosol). This activity required the presence of AcrA within the proteoliposomes. Interestingly, the increase in proton efflux also occurred when aminoglycosides were present only in the intravesicular space. This result suggested that AcrD can also capture aminoglycosides from the periplasm to extrude them into the medium in intact cells, acting as a "periplasmic vacuum cleaner."


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Multidrug efflux pumps play an important role in the intrinsic resistance of gram-negative bacteria to a number of antimicrobial agents, and their overproduction causes increased resistance to a broad range of agents, a major concern in view of the occurrence of multidrug-resistant bacterial pathogens in the clinical setting (17). Among the five known families of multidrug transporters, the resistance-nodulation-division (RND) family (26) tends to play major roles in the intrinsic resistance of gram-negative bacteria (17). Some of them also can extrude a very large range of compounds (29). The Escherichia coli genome contains several genes coding for RND transporters (20, 25). Among them, the best-studied transporter is AcrB, which provides the major mechanism for multiple drug resistance in E. coli (16, 17, 29) and whose crystal structure was elucidated by Murakami and coworkers (15). AcrB forms a multiprotein complex (31, 32) with AcrA, a periplasmic protein of the membrane fusion protein family (5), and the complex is likely to include an outer membrane channel protein, TolC (8). This tripartite construction allows the transporter to pump out drugs directly into the medium, rather than into the periplasm, making the complex much more efficient in producing resistance (16, 17, 29).

Earlier studies using intact cells suggested that the AcrAB-TolC complex and its homolog MexAB-OprM of Pseudomonas aeruginosa may capture their substrates from the periplasm or the outer leaflet of the cytoplasmic membrane (which is in equilibrium with the periplasm), because carbenicillin, which does not diffuse across the cytoplasmic membrane, is pumped out by this complex (13, 18). Furthermore, the crystallographic structure of the AcrB trimer (15) contains "vestibules," which connect the outer leaflet of the cytoplasmic membrane outside with the central cavity within the trimer, which was recently shown to be the binding site of the substrates (28). Thus, it seems likely that the substrates associated with the outer surface of the cytoplasmic membrane may diffuse laterally through these vestibules to reach the binding site inside AcrB. However, it proved difficult to test the hypothesis of periplasmic substrate capture in a more rigorous manner in the reconstituted AcrB-containing proteoliposomes (29) because most of the substrates of AcrB are lipophilic (16, 17, 29) and could spontaneously cross the vesicle membrane bilayer. In this study, we overcame this difficulty by using the AcrD transporter of E. coli, which is a close homolog of the AcrB RND pump, yet exports aminoglycosides, a polycationic, hydrophilic class of antibiotics (22) which are very unlikely to cross the membranes by simple diffusion. We purified hexahistidine-tagged AcrD, reconstituted it into proteoliposomes, and tested the function of this pump by adding aminoglycoside substrates on either side of the membrane. As the uptake of the substrates from the periplasmic side does not move them to another compartment, we utilized an indirect assay that examines the proton flux accompanying the transport of substrates; such an assay was successfully used with the proteoliposomes reconstituted with AcrB (29). Our results show that aminoglycosides can be captured and transported not only in a classical manner from the more alkaline, external medium (corresponding to cytosol) but also from the acidified intravesicular space (corresponding to the periplasm).


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Bacteria, plasmids, and growth conditions. E. coli KAM3 (a {Delta}acrB mutant of K-12 [20]) transformed with plasmid pTrcHacrD, obtained from A. Yamaguchi, was used for the expression of the RND protein AcrD with a C-terminal hexahistidine tag (20) (this tagged AcrD is called "AcrD" in this publication). Cells were grown in Luria-Bertani broth supplemented with ampicillin (100 µg/ml) with rotary shaking at 37°C.

Purification of AcrA and AcrD His-tagged proteins. AcrA, modified by the deletion of the N-terminal lipidation site and by the addition of a hexahistidine tag at the C terminus, was purified as described previously (29). For AcrD purification, an overnight culture of KAM3 containing pTrcHacrD was diluted into 100 vol of LB and incubated until the A600 reached 0.4. Expression of AcrD was induced by adding isopropyl-ß-D-thiogalactoside to a final concentration of 1 mM. Cells were harvested by centrifugation after 4 h of induction and resuspended in buffer A (20 mM Tris-HCl [pH 7.0], 500 mM NaCl, 5 mM imidazole). The cell suspension was lysed by two passes through a French pressure cell (American Instrument Co.) operating at 10,000 lb/in2 in the presence of DNase (100 µg/ml) and EDTA-free protease inhibitor cocktail (Roche). Once unbroken cells were removed by centrifugation, the supernatant was centrifuged at 100,000 x g for 1 h at 4°C. The pellet was resuspended in buffer A containing 1.5% dodecyl maltoside (DM) to solubilize membrane proteins. After the centrifugation step was repeated, the supernatant was used immediately for purification of AcrD, using Talon His-Bind cobalt chelating resin (Clontech) in a hybrid batch-gravity flow procedure as described by the manufacturer with some modifications. The binding and initial washing steps were carried out in buffer A. The last wash of the resin suspension was performed with buffer B (25 mM Tris-HCl [pH 7.0], 200 mM KCl, 10% glycerol, 5 mM imidazole, 0.2% DM). Elution of AcrD was obtained using buffer B containing 100 mM imidazole. Fractions containing AcrD were pooled and concentrated by centrifugation in a Centricon-30 (Mr 30,000-cutoff; Amicon). This solution was used in the reconstitution (see below) without removal of imidazole or DM because their carryover was minimal as the protein concentration was usually high (1.5 to 2 mg/ml).

Production of polyclonal anti-AcrD serum. Purified AcrD was used to generate polyclonal antibodies in rabbit. The antibodies were purified from serum by ammonium sulfate precipitation (70% saturation), and antibodies against non-AcrD proteins were removed by affinity chromatography using the crude extract from JZM320 (acrD::tet) (22) coupled to CNBr-activated Sepharose as described previously (29).

Protein assay and analysis. Protein concentration was determined with the BCA reagent (Pierce) with bovine serum albumin as the standard. Proteins were resolved in sodium dodecyl sulfate (SDS)-7.5% polyacrylamide gel electrophoresis. In some cases, the resolved proteins were transferred electrophoretically to a nitrocellulose membrane (Bio-Rad) in Tris base (20 mM), glycine (150 mM), and methanol (20%) for Western blot analysis. Binding of primary polyclonal antibodies (anti-AcrA [29] or anti-AcrD) was detected with an alkaline phosphate-conjugated anti-rabbit secondary antibody (Sigma). Protein visualization was performed with nitroblue tetrazolium and 5-bromo-4-chloro-3-indolylphosphate (4).

AcrD reconstitution into proteoliposomes. Proteoliposomes were prepared with E. coli lipids (Avanti Polar Lipids) by the octyl-ß-D-glucopyranoside (OG) dilution method, described previously (4, 29). Briefly, AcrD (50 µg, usually 25 to 30 µl) was added to 4.5 mg of lipids dispersed in the reconstitution buffer (25 mM HEPES-KOH [pH 7.0], 200 mM KCl, 1 mM dithiothreitol), and the total volume was adjusted to 0.5 ml with the reconstitution buffer and 10% OG so that the final concentration of OG was 1.1%. This solution was diluted with 5 ml of chilled reconstitution buffer, and the proteoliposomes were recovered by centrifugation (4) and were usually resuspended in 100 µl of the reconstitution buffer. As a control, vesicles without AcrD were prepared in the same way. In most cases, pyranine (1 mM) was added to the mixture before dilution so that it would become entrapped in the intravesicular space. In some cases, AcrA (30 µg), aminoglycosides (70 µM), and/or MgCl2 (1.5 mM) was similarly added to the 0.5-ml mixture before dilution.

Transmembrane H+ flux assay. Determination of transmembrane H+ flux was performed as described by Zgurskaya and Nikaido (29). Thus, pyranine, a hydrophilic, fluorescent, pH indicator dye, was incorporated into the intravesicular space. These proteoliposomes (up to 20 µl) were added to 2 ml of assay buffer (25 mM HEPES-NaOH [pH 7.0], 200 mM NaCl); valinomycin (10 µM) was added to produce {Delta}{Psi}, which becomes converted rapidly to {Delta}pH in medium containing a high concentration of chloride ion. Fluorescence was followed with a Shimadzu RF-5301 spectrofluorimeter to measure changes in intravesicular pH. When aminoglycosides were added to the external medium, they were present in the assay buffer before the addition of proteoliposomes. Excitation and emission wavelengths were, respectively, 455 and 509 nm.

Uptake experiments. Proteoliposomes (30 µl) were preequilibrated with 370 µl of K+-free buffer (25 mM HEPES-NaOH [pH 7.0], 200 mM NaCl) containing [3H]gentamicin (specific radioactivity, 200 mCi/g; American Radiolabeled Chemical) at a final concentration of 10 µM. After the addition of valinomycin, 50-µl samples were removed at different times and diluted into an ice-cold solution containing 0.1 M LiCl and 0.4 mg of poly-L-lysine (Sigma)/ml, and the mixture was filtered on a 0.22 µm-pore-size GSWP Millipore filter. The filter was washed with ice-cold 0.1 M LiCl, dried, and used for determination of radioactivity in a Beckman LS6500 liquid scintillation counter.

Aminoglycoside susceptibility assays. The susceptibility of E. coli to aminoglycosides (obtained from Sigma) was determined in Mueller-Hinton medium (Difco) by the broth microdilution method. The inoculum was 104 cells per ml, and the results were read after overnight incubation at 37°C.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Purification of AcrD and AcrA. The purified AcrD contained a hexahistidine tag at its C terminus (20) and was 95% pure as judged by Coomassie blue staining of SDS-polyacrylamide gels (Fig. 1A, lane 3). This hexahistidine-tagged AcrD was earlier shown to be fully functional in intact E. coli cells (20). Mass spectrometry analysis of its trypsin digest matched the expected pattern for the AcrD protein (data not shown). The purified AcrD preparation was used to produce polyclonal rabbit antibodies.



View larger version (39K):
[in this window]
[in a new window]
  		FIG. 1. Overexpression and purification of AcrD (A) and AcrA (B). Protein samples were loaded onto an SDS-7.5% polyacrylamide gel and stained with Coomassie brilliant blue. Lanes 1, crude extract; lanes 2, column flowthrough fractions; lanes 3, AcrD-6His after cobalt resin batch purification (A) and AcrA-6His after purification on an Ni2+ column (B). The rightmost and leftmost lanes in panels A and B, respectively, show molecular mass markers, with molecular masses shown in kilodaltons. (C) Western blot analysis of AcrD reconstituted into proteoliposomes. Here AcrD proteoliposomes were loaded onto an SDS-7.5% polyacrylamide gel and analyzed by immunoblotting with polyclonal anti-AcrD antibodies (left lane). The right lane shows molecular mass markers as for panels A and B.

 
The purified AcrA (Fig. 1B, lane 3) used in this work corresponds to the modified protein containing a hexahistidine tag at its C terminus and devoid of lipid modification at its N terminus (30). This lipid-deficient version of AcrA (starting at D26 of the native proprotein) was shown to be functionally active (30) and was used in order to avoid possible interactions with the lipid vesicles during AcrD reconstitution assays.

Reconstituted AcrD-containing proteoliposomes were inactive in aminoglycoside transport in the absence of AcrA. An uptake assay performed with intact cells of E. coli showed that aminoglycosides were substrates of AcrD (22). Since AcrD is a member of the RND family of transporters energized by proton motive force, movement of the substrates by the transporter should produce coupled transmembrane movement of H+, as was experimentally demonstrated for AcrB (29). We followed the changes in intravesicular pH with the water-soluble fluorescent pH probe pyranine (29), which does not diffuse across the vesicle membrane (3, 12).

When protein-free vesicles containing K+ and pyranine were diluted in the assay buffer containing Na+ but no K+, the addition of valinomycin produced a rapid efflux of K+ and generated membrane potential, which was replaced by interior-acid {Delta}pH in the presence of Cl–, as seen by the rapid drop of pyranine fluorescence. Pyranine fluorescence (therefore intravesicular pH) then increased slowly, reflecting the spontaneous leakage of H+ (data not shown). In preliminary experiments we tested whether proton leakage was nonspecifically enhanced by the presence of known substrates of AcrD. Addition of aminoglycosides (gentamicin, tobramycin, amikacin, and kanamycin) at 70 µM to the external solution did not alter the rate of proton leakage. In contrast, addition of SDS or deoxycholate at the same concentration caused the instantaneous collapse of the pH gradient, presumably through their detergent activity; thus, these substrates of AcrD (20) were not used in this study.

A transmembrane H+ flux assay was then conducted by first adding the aminoglycosides to the external medium bathing AcrD-containing proteoliposomes and then by starting the reaction by generating {Delta}pH (interior acid) by the addition of valinomycin as described above. In intact cells, the periplasm is more acidic than the cytoplasm, and thus the intravesicular and extravesicular spaces in our experimental system correspond to the periplasm and cytoplasm, respectively. AcrD is expected to be an H+/drug antiporter like its close homolog, AcrB (29). If aminoglycosides are captured and transported by the AcrD efflux pump from the cytoplasm, the pump action is thus expected to produce a countermovement of H+ (efflux of H+ from the vesicles) so that the intravesicular pH will return to the initial, more alkaline value faster than through the spontaneous leakage. Even though orientation of reconstituted AcrD in proteoliposomes is probably random, we expected that imposition of an interior-acid {Delta}pH would cause the AcrD proteins inserted in the correct orientation (with their periplasmic domain facing inside) to function, pumping in substrate in exchange for a proton(s). However, no enhanced efflux of H+ was observed in the presence of 70 µM aminoglycosides in the reaction buffer. We show the case for gentamicin in Fig. 2A, curve 2, which shows unaltered slope from the efflux without any substrate (Fig. 2A, curve 1). These results initially suggested to us that AcrD may not pump out aminoglycosides from the extravesicular space (or cytosol). We therefore carried out the H+ flux assay using reconstituted AcrD proteoliposomes containing gentamicin (70 µM) entrapped in the intravesicular space (corresponding to the periplasm). However, again, no stimulation of proton flux was observed (data not shown).



View larger version (16K):
[in this window]
[in a new window]
  		FIG. 2. Effects of aminoglycoside addition to the extravesicular space on proton efflux from AcrD-containing proteoliposomes. AcrD proteoliposomes were reconstituted in a buffer containing 0.2 M KCl and then were diluted in the same buffer containing 0.2 M NaCl instead. {Delta}pH was generated by the addition of 10 µM valinomycin. Decreased fluorescence of the pyranine corresponds to the interior-acid {Delta}pH, which dissipated slowly in the absence of drugs. Aminoglycosides were added to the extravesicular space at a concentration of 70 µM. (A) Curve 1, AcrD proteoliposomes alone; curve 2, gentamicin was added to AcrD proteoliposomes; curve 3, gentamicin was added to AcrD proteoliposomes containing AcrA plus Mg2+. (B) A single batch of AcrD proteoliposomes with entrapped AcrA and Mg2+ was used. Curve 1, without added substrates, as a control; curve 2, anti-AcrD antibodies and gentamicin added to the external medium; curve 3, gentamicin alone added to the external medium. (C) A single batch of AcrD proteoliposomes containing AcrA and Mg2+ was used. Curve 1, no-substrate control; curve 2, amikacin (70 µM) added externally; curve 3, gentamicin (70 µM) added externally; curve 4, tobramycin (70 µM) added externally. AU, arbitrary units. Insets in this and subsequent figures show schematically the experimental setup. S, substrate. Black squares, AcrA; grey structures, AcrD.

 
AcrA is necessary for the transport of aminoglycosides by the reconstituted AcrD proteoliposomes. Recently, Elkins and Nikaido (6) showed that the efficient efflux of amphiphilic substrates by AcrD in intact cells requires AcrA. Because of the apparent absence of activity of reconstituted AcrD in the presence of aminoglycosides, we examined whether the transport activity of AcrD for aminoglycosides required AcrA. When AcrA was added to the reaction buffer surrounding the reconstituted AcrD vesicles and gentamicin was also added to this medium, no enhancement of H+ efflux was observed (data not shown). This result was expected since under such conditions AcrA, a periplasmic protein, was present in the extravesicular fluid, corresponding to the cytosol based on the direction of {Delta}pH. However, when AcrA was entrapped in the intravesicular space of the AcrD-containing vesicles together with 1.5 mM MgCl2 and {Delta}pH was created, then a drug-induced proton efflux was observed, for example in the presence of 70 µM gentamicin (Fig. 2A, curve 3; compare with curve 1). Control experiments with AcrA entrapped in vesicles without AcrD unfortunately gave erratic results often showing rapid leakage of protons, possibly due to the presence of the hexahistidine tag on AcrA. Results were reproducible, however, when AcrD-containing proteoliposomes were used to entrap AcrA, and in later experiments vesicles of this type, without the addition of aminoglycosides, were always used as the control. (Each H+ efflux experiment reported in this study was successfully reproduced at least five times, although small quantitative variations were occasionally seen.)

Because we could not use control vesicles containing AcrA but without AcrD as mentioned above, it was important to confirm that the enhanced H+ efflux (as seen in Fig. 2A, curve 3) was caused by the pumping activity of AcrD. For this purpose we performed the H+ efflux assay with the AcrA-containing AcrD proteoliposomes in the presence of purified anti-AcrD polyclonal antibodies (added to the extravesicular space). As seen in Fig. 2B, curve 2, no visible drug-induced proton efflux was observed in the presence of gentamicin as a substrate (compare with the no-substrate control of curve 1). In the positive control without the antibody, gentamicin again produced significant acceleration of H+ efflux (curve 3). (The rate of H+ movement, especially that of H+ leakage, was strongly influenced by the particular batch of lipids used for reconstitution and other variables difficult to define. Thus, the slopes should be compared ideally within the same experiments done by using one batch of vesicles, or at least by using the vesicles made side by side on the same day. We cannot compare slopes of curves from different experiments, for example, Fig. 2A, curve 3, with Fig. 2B, curve 3.)

While the addition of 100 µM gentamicin to the external medium resulted sometimes in aggregation of vesicles, concentrations of 80, 70, 50, and 30 µM were able to produce drug-induced proton efflux (data not shown). Similar enhancement of H+ efflux was observed with 70 µM amikacin (Fig. 2C, curve 2, compare with curve 1 [no substrate] and with curve 3 [gentamicin control]), tobramycin (Fig. 2C, curve 4, compare with curve 1), or kanamycin (data not shown). The results shown in Fig. 2C indicate that AcrA in the intravesicular space activated AcrD and that the AcrA-AcrD complex acted as an antiporter, catalyzing the efflux of H+ coupled presumably to the influx of aminoglycosides into vesicles.

The presumed intravesicular accumulation of aminoglycosides under these conditions was tested by uptake experiments using radiolabeled gentamicin. AcrD proteoliposomes with entrapped AcrA and Mg2+ indeed accumulated more gentamicin than the control liposomes (Fig. 3). Using the intravesicular volume of 0.04 µl per 1 µg of AcrD (see Discussion), the accumulation of 3 to 6 pmol of gentamicin accumulated per µg of AcrD translates into an internal concentration 75 to 150 µM, about an order of magnitude higher than the concentration of gentamicin (10 µM) in the assay medium.



View larger version (12K):
[in this window]
[in a new window]
  		FIG. 3. Accumulation of [3H]gentamicin in AcrD-containing proteoliposomes. AcrD proteoliposomes were made in KCl buffer so that AcrA and Mg2+ were entrapped within. These were diluted into NaCl buffer containing 10 µM [3H]gentamicin as described in Materials and Methods ({blacksquare}). The results shown are averages of five independent experiments. Liposomes without AcrD and AcrA served as a control ({blacktriangleup}). At time zero 10 µM valinomycin was added to produce a proton motive force.

 
We also tested if AcrA was required for AcrD-mediated aminoglycoside efflux in intact cells, because an earlier study from our laboratory (6) examined the AcrA requirement only for amphiphilic substrates. Indeed, a {Delta}acrA mutant of E. coli showed levels of susceptibility to aminoglycosides in Mueller-Hinton medium equivalent to those of the isogenic {Delta}acrD strain (Table 1). (HNCE1b, lacking AcrA in addition to AcrB and AcrD, appeared to be slightly more susceptible than HNCE1a, lacking AcrB and AcrD alone. This could be because in an AcrBD-negative background the promiscuity of AcrA, which was shown to work together with other pumps such as YhiV [7], becomes detectable.)


View this table:
[in this window]
[in a new window]
  		TABLE 1. Aminoglycoside susceptibility of E coli strains with the acrA, acrAB, and/or acrD gene deleted

 
The AcrA-AcrD complex can capture substrates from within the vesicles. Because the very hydrophilic aminoglycosides are not expected to cross the lipid bilayer readily, we can entrap these compounds inside the vesicles and test if the AcrA-AcrD complex is capable of capturing them from this space, which corresponds to the periplasm according to the direction of {Delta}pH. When 70 µM gentamicin or amikacin was entrapped in the intravesicular space of the reconstituted AcrD proteoliposomes in the presence of AcrA and 1.5 mM MgCl2, the presence of the drug enhanced proton efflux strongly (Fig. 4, curve 2) in comparison with that of the control AcrD proteoliposomes containing only AcrA and Mg2+ (curve 1). Entrapment of amikacin showed similar stimulation (data not shown). AcrD proteoliposomes containing drugs and Mg2+, but not AcrA, did not show enhanced H+ efflux, indicating again that AcrA is needed for the activation of AcrD transporter activity. Thus, aminoglycosides present in the intravesicular space, corresponding to the periplasmic space, enhanced proton movement by the AcrD transporter if AcrA was present. Although we cannot create the barrier equivalent to the outer membrane in our reconstituted system, we believe that in intact cells the proton flux catalyzed by AcrD is energizing the efflux of aminoglycosides from the periplasmic space in intact cells with the help of AcrA and TolC.



View larger version (19K):
[in this window]
[in a new window]
  		FIG. 4. Effect of the presence of aminoglycosides in the intravesicular space on proton efflux from AcrD proteoliposomes. Experiments were carried out as for Fig. 2, using proteoliposome preparations made side by side on the same day, except that the substrate was entrapped within one batch of vesicles. Curve 1, AcrD vesicles with entrapped AcrA and Mg2+, as the control; curve 2, AcrD vesicles with entrapped AcrA, Mg2+, and gentamicin (70 µM).

 
Streptomycin as a substrate for the AcrD pump. Surprisingly, when streptomycin was used in our system as a substrate for the reconstituted AcrD, it induced a proton efflux response only when the antibiotic was entrapped in the intravesicular space together with AcrA and Mg2+ (Fig. 5, curve 3). Addition of streptomycin to the external medium surrounding the AcrD proteoliposomes containing AcrA and Mg2+did not alter the rate of H+ efflux in a detectable manner (Fig. 5, compare curve 2 with curve 1). It is likely that AcrD pumps out streptomycin preferentially from the periplasm.



View larger version (25K):
[in this window]
[in a new window]
  		FIG. 5. Effect of streptomycin on proton efflux from AcrD proteoliposomes. Experiments were carried out as for Fig. 2 and 4. Two kinds of proteoliposomes used were made side by side on the same day. Curve 1, AcrD proteoliposomes containing AcrA and Mg2+ (but no substrate) as the control; curve 2, streptomycin (70 µM) was added to the extravesicular space of AcrD proteoliposomes containing AcrA and Mg2+; curve 3, AcrD proteoliposomes containing AcrA, Mg2+, and 70 µM streptomycin. AU, arbitrary units.

 

   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Some RND family multidrug efflux pump complexes, such as AcrAB-TolC and MexAB-OprM, pump out mostly lipophilic and amphiphilic substrates, which include uncharged, anionic, cationic, and zwitterionic compounds (16). This observation, as well as the efficient efflux of carbenicillin, which does not cross the cytoplasmic membrane barrier, suggested already in 1994 that a major mechanism for substrate capture for these pumps may originate from the outer leaflet of the cytoplasmic membrane or from the periplasm that is in equilibrium with the outer leaflet (13). As described above, this concept was supported by further intact-cell studies (18) as well as the X-ray crystallographic structure of AcrB (15). Nevertheless, it was not possible to test the periplasmic substrate capture hypothesis by in vitro reconstitution of AcrB because most of its substrates are lipophilic or amphiphilic and are able to diffuse spontaneously across the lipid bilayer of proteoliposomes (29).

In this study, we took advantage of the fact that an RND transporter, AcrD, of E. coli extrudes aminoglycosides (22), which are totally hydrophilic compounds that are unlikely to diffuse spontaneously across the proteoliposome membrane. Hexahistidine-tagged AcrD was reconstituted into proteoliposomes by the OG dilution method. After the acidification of the intravesicular space, intravesicular pH was followed by using a fluorescent pH indicator. We first discovered that it was necessary to entrap AcrA in the intravesicular space in order to see the efflux-coupled movement of H+ (Fig. 2A), and thereafter we always used proteoliposomes containing AcrA in the intravesicular space. These experiments (Fig. 2A and C) showed that there was an increased H+ efflux when aminoglycosides were present in the extravesicular space, suggesting that AcrD can transport substrates inward, coupled to the outward movement of H+ from the proteoliposomes. In fact, intravesicular accumulation of gentamicin was experimentally observed (Fig. 3). Thus, AcrD appears to function as a proton-substrate antiporter, as was previously reported for multidrug transporter AcrB (29) and another RND efflux pump, the toxic cation transporter CzcA (9).

The more alkaline, extravesicular space corresponds to the cytosol, and these results imply that AcrD is capable of capturing aminoglycosides from the cytosol. This interpretation, however, requires a note of caution. {Delta}{Psi} (interior negative) was first generated by the valinomycin-mediated efflux of K+ from the vesicles, and there is no guarantee that this {Delta}{Psi} was completely dissipated through its conversion into {Delta}pH. If some of this {Delta}{Psi} still remained during the experiment, it might have caused the passive, uncatalyzed influx of polycationic aminoglycosides, possibly causing the secondary efflux of H+. We do not, however, believe that this is a plausible interpretation of our data, because streptomycin, in spite of the presence of multiple positive charges, did not cause H+ efflux when added to the extravesicular medium (Fig. 5).

An important finding of this study was that the presence of aminoglycosides in the intravesicular space also strongly enhanced H+ efflux (Fig. 4). The simplest interpretation of this observation is that AcrD captures its substrates from the periplasmic space as well in the living cells and the capture is followed by the active efflux of the substrate concomitant with the H+ influx. A close homolog of AcrD, AcrB, was earlier shown to function as an H+/substrate antiporter in a reconstitution assay (29). A recent X-ray crystallographic study (28) showed that the substrate binding to AcrB occurs in the central cavity approximately at the level of the outer surface of the cytoplasmic membrane. It is thus likely that most lipophilic and amphiphilic substrates reach this destination by lateral diffusion within the external leaflet of the membrane. We suggest that aminoglycosides, although hydrophilic, will become adsorbed onto the head groups of acidic phospholipids at the external surface of the membrane bilayer and diffuse laterally into the central cavity (27).

Interestingly, streptomycin, which has an unusual structure unrelated to other aminoglycosides used here, apparently is captured in the intravesicular space (periplasm) but not efficiently in the extravesicular space (cytoplasm) (Fig. 5). This results confirms our assumption that aminoglycosides do not cross the proteoliposome membrane spontaneously. The result also suggests that the cytosolic substrates are captured by a pathway quite different from that used for the periplasmic substrates.

We have been able to detect, reproducibly, stimulation of H+ efflux by as low a concentration as 30 µM gentamicin (about 15 µg/ml). It was not possible to estimate the Km value, in part because the higher concentrations of aminoglycosides caused aggregation of vesicles. However, these data suggest that the pump is active with clinically significant concentrations of aminoglycosides, because the interior-negative Donnan potential across the outer membrane (23) is expected to produce a huge accumulation of these polycationic compounds in the periplasm. It would have been desirable to confirm these results of in vitro reconstitution studies by the quantitation of aminoglycosides in the periplasm of intact cells. However, the strong adsorption of polycationic aminoglycosides to acidic constituents of the cell makes such a study practically impossible.

Proteoliposomes were made from 0.45 nmol (50 µg) of AcrD protein and 6.4 µmol of E. coli phospholipids. The molar ratio between AcrD and phospholipids was therefore 1:14,200. Each vesicle has an average diameter of about 100 nm and therefore a surface of 31,400 nm2 (24). Considering that each phospholipid occupies an area of about 0.6 nm2, each vesicle should contain about 105 lipid molecules (24). Thus, the ratio used corresponds to two or three AcrD trimers per vesicle.

Since we determined both the accumulation of radiolabeled gentamicin and the H+ flux, we can compare these values. The accumulation assay showed the uptake of 6 pmol (or 3 pmol if we subtract the control) of gentamicin per µg of AcrD protein in about 60 s. During the same period in the H+ efflux assay, we saw the internal pH change by about 0.1 U (Fig. 2 and 4). Using the Henderson-Hasselbach equation, we can calculate that this corresponds to deprotonation of 1.5 mM HEPES, which has a pKa of 7.5. If we assume that all of the lipids added to the reconstitution mixture became incorporated into proteoliposomes, we can calculate their total internal volume as about 2 µl from the parameters given above. Thus, the proteoliposome volume per microgram of AcrD is 0.04 µl, and the total amount of proton moved is 1.5 x 10–3 x 0.04 x 10–6 mol, or 60 pmol per µg of AcrD. This is not too different from the amount of the substrate translocated, 6 pmol per µg of AcrD, and validates the use of the proton efflux assay for AcrD function. It is not possible to determine the exact number of H+ molecules needed for the translocation of one substrate molecule, because "slippage" is known to occur frequently in transporters reconstituted into proteoliposomes (19).

The periplasmic protein AcrA is absolutely required for AcrB-catalyzed efflux of drugs by intact cells (14). Because AcrA is a very elongated protein (2, 30) with a length that could span the depth of the periplasm and because its N terminus is associated with the inner membrane through its lipid extension, it has often been assumed that its major function is to connect the outer and inner membranes. The earlier AcrB reconstitution assay (29) relied on the capture of the extruded fluorescent phospholipid molecules by neighboring acceptor vesicles, and thus the strong stimulation of this reaction by the added AcrA was interpreted as a consequence of connection, by AcrA, of donor and acceptor vesicles. However, in our current assay for AcrD activity, juxtaposition of two vesicles is not needed and the stimulation by AcrA presumably occurs because the AcrD activity requires direct activation by AcrA. Interestingly, AcrA was recently shown to go through extensive pH-induced conformational changes (11). (This dynamic nature of AcrA makes the interpretation of the recent crystallographic structure of MexA [1, 10], an AcrA homolog, somewhat difficult.)


   ACKNOWLEDGMENTS
 
We thank K. Nishino and A. Yamaguchi for providing us with plasmid pTrcHacrD and C. Elkins for providing us with E. coli strain HNCE3.

This study was supported by Public Health Service grant AI-09644 from the National Institute of Allergy and Infectious Diseases.


   FOOTNOTES
 
* Corresponding author. Mailing address: Department of Molecular and Cell Biology, 426 Barker Hall, University of California, Berkeley, CA 94720-3202. Phone: (510) 642-2027. Fax: (510) 643-9290. E-mail: nhiroshi{at}uclink4.berkeley.edu. Back


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Akama, H., T. Matsuura, S. Kashiwagi, H. Yoneyama, T. Tsukihara, A. Nakagawa, and T. Nakae. 2004. Crystal structure of the membrane fusion protein, MexA of the multidrug transporter in Pseudomonas aeruginosa. J. Biol. Chem. 279:25939-25942.[Abstract/Free Full Text]
  2. Avila-Sakar, A. J., S. Misaghi, E. M. Wilson-Kubalek, K. H. Downing, H. Zgurskaya, H. Nikaido, and E. Nogales. 2001. Lipid-layer crystallization and preliminary three-dimensional structural analysis of AcrA, the periplasmic component of a bacterial multidrug efflux pump. J. Struct. Biol. 136:81-88.[CrossRef][Medline]
  3. Clement, N. R., and J. M. Gould. 1981. Pyranine (8-hydroxy-1,3,6-pyrenetrisulfonate) as a probe of internal aqueous hydrogen ion concentration in phospholipid vesicles. Biochemistry 20:1534-1544.[CrossRef][Medline]
  4. Davidson, A. L., and H. Nikaido. 1991. Purification and characterization of the membrane-associated components of the maltose transport system from Escherichia coli. J. Biol. Chem. 266:8946-8951.[Abstract/Free Full Text]
  5. Dinh, T., I. T. Paulsen, and M. H. Saier, Jr. 1994. A family of extracytoplasmic proteins that allow transport of large molecules across the outer membranes of gram-negative bacteria. J. Bacteriol. 176:3825-3831.[Abstract/Free Full Text]
  6. Elkins, C. A., and H. Nikaido. 2002. Substrate specificity of the RND-type multidrug efflux pumps AcrB and AcrD of Escherichia coli is determined predominantly by two large periplasmic loops. J. Bacteriol. 184:6490-6498.[Abstract/Free Full Text]
  7. Elkins, C. A., and H. Nikaido. 2003. Chimeric analysis of AcrA function reveals the importance of its C-terminal domain in its interaction with the AcrB multidrug efflux pump. J. Bacteriol. 185:5349-5356.[Abstract/Free Full Text]
  8. Fralick, J. A. 1996. Evidence that TolC is required for functioning of the Mar/AcrAB efflux pump of Escherichia coli. J. Bacteriol. 178:5803-5805.[Abstract/Free Full Text]
  9. Goldberg, M., T. Pribyl, S. Juhnke, and D. H. Nies. 1999. Energetics and topology of CzcA, a cation/proton antiporter of the resistance-nodulation-cell division protein family. J. Biol. Chem. 274:26065-26070.[Abstract/Free Full Text]
  10. Higgins, M. K., E. Bokma, E. Koronakis, and V. Koronakis. 2004. Structure of the periplasmic component of a bacterial drug efflux pump. Proc. Natl. Acad. Sci. USA 101:9994-9999.[Abstract/Free Full Text]
  11. Ip, H., K. Stratton, H. Zgurskaya, and J. Liu. 2003. pH-induced changes in the conformation of AcrA, the membrane fusion protein of Escherichia coli multidrug efflux system. J. Biol. Chem. 278:50474-50482.[Abstract/Free Full Text]
  12. Kamp, F., and J. A. Hamilton. 1992. pH gradients across phospholipid membranes caused by fast flip-flop of un-ionized fatty acids. Proc. Natl. Acad. Sci. USA 89:11367-11370.[Abstract/Free Full Text]
  13. Li, X.-Z., D. Ma, D. M. Livermore, and H. Nikaido. 1994. Role of efflux pump(s) in intrinsic resistance of Pseudomonas aeruginosa: active efflux as a contributing factor to ß-lactam resistance. Antimicrob. Agents Chemother. 38:1742-1752.[Abstract/Free Full Text]
  14. Ma, D., D. N. Cook, M. Alberti, N. G. Pon, H. Nikaido, and J. E. Hearst. 1995. Genes acrA and acrB encode a stress-induced efflux system of Escherichia coli. Mol. Microbiol. 16:45-55.[CrossRef][Medline]
  15. Murakami, S., R. Nakashima, E. Yamashita, and A. Yamaguchi. 2002. Crystal structure of bacterial multidrug efflux transporter AcrB. Nature 419:587-593.[CrossRef][Medline]
  16. Nikaido, H. 1996. Multidrug efflux pumps of gram-negative bacteria. J. Bacteriol. 178:5853-5859.[Free Full Text]
  17. Nikaido, H. 1998. Antibiotic resistance caused by Gram-negative multidrug efflux pumps. Clin. Infect. Dis. 27(Suppl. 1):S32-S41.
  18. Nikaido, H., M. Basina, V. Nguyen, and E. Y. Rosenberg. 1998. Multidrug efflux pump AcrAB of Salmonella typhimurium excretes only those ß-lactam antibiotics containing lipophilic side chains. J. Bacteriol. 180:4686-4692.[Abstract/Free Full Text]
  19. Nikaido, H., and J. A. Hall. 1998. Overview of bacterial ABC transporters. Methods Enzymol. 292:3-20.[Medline]
  20. Nishino, K., and A. Yamaguchi. 2001. Analysis of a complete library of putative drug transporter genes in Escherichia coli. J. Bacteriol. 183:5803-5812.[Abstract/Free Full Text]
  21. Okusu, H., D. Ma, and H. Nikaido. 1996. AcrAB efflux pump plays a major role in the antibiotic resistance phenotype of Escherichia coli multiple-antibiotic-resistance (Mar) mutants. J. Bacteriol. 178:306-308.[Abstract/Free Full Text]
  22. Rosenberg, E. Y., D. Ma, and H. Nikaido. 2000. AcrD of Escherichia coli is an aminoglycoside efflux pump. J. Bacteriol. 182:1754-1756.[Abstract/Free Full Text]
  23. Sen, K., J. Hellman, and H. Nikaido. 1988. Porin channels in intact cells of Escherichia coli are not affected by Donnan potential across the outer membrane. J. Biol. Chem. 263:1182-1187.[Abstract/Free Full Text]
  24. Sugawara, E., and H. Nikaido. 1994. OmpA protein of Escherichia coli outer membrane occurs in open and closed channel forms. J. Biol. Chem. 269:17981-17987.[Abstract/Free Full Text]
  25. Sulavik, M. C., C. Houseweart, C. Cramer, N. Jiwani, N. Murgolo, J. Greeene, B. DiDomenico, K. J. Shaw, G. H. Miller, R. Hare, and G. Shimer. 2001. Antibiotic susceptibility profiles of Escherichia coli strains lacking multidrug efflux pump genes. Antimicrob. Agents Chemother. 45:1126-1136.[Abstract/Free Full Text]
  26. Tseng, T. T., K. S. Gratwick, J. Kollman, D. Park, D. H. Nies, A. Goffeau, and M. H. Saier, Jr. 1999. The RND permease superfamily: an ancient, ubiquitous and diverse family that includes human disease and development proteins. J. Mol. Microbiol. Biotechnol. 1:107-125.[Medline]
  27. Yu, E. W., J. R. Aires, and H. Nikaido. 2003. AcrB multidrug efflux pump of Escherichia coli: composite substrate-binding cavity of exceptional flexibility generates its extremely wide substrate specificity. J. Bacteriol. 185:5657-5664.[Free Full Text]
  28. Yu, E. W., G. McDermott, H. I. Zgurskaya, H. Nikaido, and D. E. Koshland, Jr. 2003. Structural basis of multiple drug-binding capacity of AcrB multidrug efflux pump. Science 300:976-980.[Abstract/Free Full Text]
  29. Zgurskaya, H. I., and H. Nikaido. 1999. Bypassing the periplasm: reconstitution of the AcrAB multidrug efflux pump of Escherichia coli. Proc. Natl. Acad. Sci. USA 96:7190-7195.[Abstract/Free Full Text]
  30. Zgurskaya, H. I., and H. Nikaido. 1999. AcrA is a highly asymmetric protein capable of spanning the periplasm. J. Mol. Biol. 285:409-420.[CrossRef][Medline]
  31. Zgurskaya, H. I., and H. Nikaido. 2000. Multidrug resistance mechanisms: drug efflux across two membranes. Mol. Microbiol. 37:219-235.[CrossRef][Medline]
  32. Zgurskaya, H. I., and H. Nikaido. 2000. Cross-linked complex between oligomeric periplasmic lipoprotein AcrA and the inner-membrane-associated multidrug efflux pump AcrB from Escherichia coli. J. Bacteriol. 182:4264-4267.[Abstract/Free Full Text]

Journal of Bacteriology, March 2005, p. 1923-1929, Vol. 187, No. 6
0021-9193/05/$08.00+0     doi:10.1128/JB.187.6.1923-1929.2005
Copyright © 2005, American Society for Microbiology. All Rights Reserved.



This article has been cited by other articles:


This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Aires, J. R.
Right arrow Articles by Nikaido, H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Aires, J. R.
Right arrow Articles by Nikaido, H.

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
Appl. Environ. Microbiol. Infect. Immun. Eukaryot. Cell
Mol. Cell. Biol. J. Virol. Microbiol. Mol. Biol. Rev.
ALL ASM JOURNALS