SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Inner membrane transporters
  5. Periplasmic membrane fusion proteins
  6. Outer membrane components of tri-partite efflux systems
  7. Acknowledgements
  8. References

A set of multidrug efflux systems enables Gram-negative bacteria to survive in a hostile environment. This review focuses on the structural features and the mechanism of major efflux pumps of Gram-negative bacteria, which expel from the cells a remarkably broad range of antimicrobial compounds and produce the characteristic intrinsic resistance of these bacteria to antibiotics, detergents, dyes and organic solvents. Each efflux pump consists of three components: the inner membrane transporter, the outer membrane channel and the periplasmic lipoprotein. Similar to the multidrug transporters from eukaryotic cells and Gram-positive bacteria, the inner membrane transporters from Gram-negative bacteria recognize and expel their substrates often from within the phospholipid bilayer. This efflux occurs without drug accumulation in the periplasm, implying that substrates are pumped out across the two membranes directly into the medium. Recent data suggest that the molecular mechanism of the drug extrusion across a two-membrane envelope of Gram-negative bacteria may involve the formation of the membrane adhesion sites between the inner and the outer membranes. The periplasmic components of these pumps are proposed to cause a close membrane apposition as the complexes are assembled for the transport.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Inner membrane transporters
  5. Periplasmic membrane fusion proteins
  6. Outer membrane components of tri-partite efflux systems
  7. Acknowledgements
  8. References

Multidrug resistance presents a serious problem in the treatment of bacterial infections. The importance of this mechanism of resistance in clinical settings is reflected in a large number of recent review articles which address various aspects of this problem (Paulsen et al., 1996a; Nikaido, 1998a, b; Saier et al., 1998; Lomovskaya et al., 2000).

The multidrug resistance phenomenon is often associated with the overexpression of the transporters that recognize and efficiently expel from the cells a broad range of structurally unrelated compounds. Analysis of the available genome sequences of various bacteria revealed that known and putative drug efflux transporters constitute from 6% to 18% of all transporters (Paulsen et al., 1998). A growing number of studies demonstrate that resistance to almost any antibiotic could be achieved through the activity of multidrug efflux pumps. Thus, evolution has tailored bacteria with enormous capabilities to survive in a toxic environment (Saier et al., 1998).

In Gram-negative bacteria, the majority of multidrug transporters share a common three-component organization: a transporter located in the inner membrane (IM) functions with an outer membrane (OM) channel and a periplasmic accessory protein (Fig. 1A). In this arrangement, efflux complexes traverse both, the inner and the outer, membranes and thus facilitate direct passage of the substrate from the cytoplasm or the cytoplasmic membrane into the external medium. Direct efflux of drugs into the medium is advantageous for Gram-negative bacteria because, in order to come in again, the expelled drug molecules must cross the low permeability OM. Hence, drug efflux works synergistically with the low permeability of OM. The synergy explains the observation that Gram-negative bacteria become hypersusceptible to various drugs either by the inactivation of efflux pumps or by the permeabilization of the OM. The synergy is also apparent as some carbapenems such as meropenem select for a loss of specific basic amino acid/carbapenem channel, OprD, in the OM, as well as for the overproduction of MexAB-OprM efflux complex (Köhler et al., 1999).

image

Figure 1. The hypothetical structure and mechanism of tri-partite efflux pumps in Gram-negative bacteria, using as an example the AcrAB-TolC efflux pump from E. coli.The top picture shows the domains within AcrA, with its N-terminal lipid, an interrupted coiled–coil domain and a pair of lipoyl arm domains (shown as arrows flanking the coiled coil domain).

A. In the non-functioning state, the inner-membrane-associated transporter AcrB simply exists as a complex with the periplasmic component AcrA, which is anchored to the outer surface of IM through its N-terminal lipid moiety, and is presumably loosely associated with OM through its C-terminal domain. The OM channel TolC is also shown as a part of the tri-partite complex, but currently there is no evidence for this association (see text).

B. AcrA might bring TolC and AcrB together by folding back on itself, using the interrupted coiled–coil domain and the two lipoyl arm domains. This will allow the drug efflux across the two membranes directly into the medium. The model is based on results obtained by Zgurskaya and Nikaido (1999a, b) and the theoretical analysis of Johnson and Church (1999). For details, see text.

Download figure to PowerPoint

All three components of the efflux pumps are often coded for in the same gene cluster, e.g. MexAB-OprM efflux complex from Pseudomonas aeruginosa (Poole et al., 1993) and MtrCDE from Neisseria gonorrhoeae (Hagman et al., 1995). Some pumps, however, such as MexXY from P. aeruginosa (Aires et al., 1999; Mine et al., 1999) and AcrAB from Escherichia coli (Ma et al., 1993), lack the gene for the OM component in the encoding gene cluster. Yet AcrAB extrudes the substrates from cells without the accumulation of drugs in the periplasm, suggesting that AcrAB recruits an OM channel for its functioning. Genetic studies suggested that TolC, a multifunctional OM channel, may associate with AcrAB to form a functional tri-partite complex (Fralick, 1996). MexXY appears to share the OM component with MexAB efflux complex because both these systems require the presence of a functional OprM channel (Aires et al., 1999). Interestingly, when expressed in E. coli cells, both MexAB and MexXY can also function with TolC protein (Mine et al., 1999).

The TolC channel was proposed to associate with multidrug transporters and also with several other transport systems including a haemolysin export machinery (Holland et al., 1990) and the system for uptake of colicin (Gilson et al., 1990). Thus, it is not surprising that the architecture of the tri-partite multidrug resistance complexes is thought to be analogous to the assemblies that secrete large proteins, such as haemolysin, directly into the medium (Wandersman, 1992). Many mechanistic details of the substrate translocation across two membranes have emerged from the studies of haemolysin translocator HlyBD-TolC of E. coli and this assembly may be thought as a model of the transport across two membranes.

It will be useful to consider the mechanism of tri-partite multidrug efflux systems in two parts. First, structurally diverse substrates must be recognized, most probably by the IM transporters. Second, the transport process, bypassing the periplasm and the OM, presumably occurs by a mechanism involving the two accessory proteins. This review discusses the structural and mechanistic features of these two likely phases of such efflux transport processes.

Inner membrane transporters

  1. Top of page
  2. Abstract
  3. Introduction
  4. Inner membrane transporters
  5. Periplasmic membrane fusion proteins
  6. Outer membrane components of tri-partite efflux systems
  7. Acknowledgements
  8. References

Genetic and biochemical studies indicated that IM efflux transporters of the three-component multidrug resistance (MDR) complexes are responsible for drug recognition (Zgurskaya and Nikaido, 1999a) and that they efficiently expel from cells drugs which are targeted to both the cytoplasm and the periplasm of the Gram-negative bacteria (Li et al., 1994; 1995).

MDR pumps from Gram-negative bacteria so far known belong to several families of transporters (for reviews see Nikaido, 1998a, b). However, most of the efflux transporters that mediate resistance to clinically relevant antibiotics belong to the Resistance-Nodulation-cell Division (RND) superfamily of transporters (Saier et al., 1994). The RND superfamily had previously been known to include only Gram-negative bacterial efflux transporters of heavy metals, hydrophobic drugs and lipooligosaccharides. Recently, this superfamily has undergone an extensive revision, and as a result, new members were identified in all major kingdoms (Tseng et al., 1999).

All members of RND superfamily have similar but unusual topological features. Their proposed structure consist of 12 transmembrane segments (TMS) or α-helices with two large hydrophilic extracytoplasmic domains between TMS 1 and 2 and TMS 7 and 8 (Saier et al., 1994; 1998). This topology was experimentally confirmed for the MexB and MexD drug transporters from P. aeruginosa (Gotoh et al., 1999; Guan et al., 1999) and CzcA heavy metal transporter from Ralstonia sp. (Goldberg et al., 1999). Comparative sequence analyses have indicated that the N- and C-terminal halves of RND proteins share sequence similarity, suggesting that they may have evolved via tandem duplication (Saier et al., 1994; 1998). Reconstitution studies of two RND-type transporters, multidrug transporter AcrB and CzcA, have confirmed the previous predictions that the RND family members are proton–substrate antiporters (Goldberg et al., 1999; Zgurskaya and Nikaido, 1999a).

The best studied RND drug transporters such as AcrB of E. coli and MexB of P. aeruginosa pump out mostly lipophilic and amphiphilic compounds (Nikaido, 1996). The substrates of AcrB/MexB have very diverse structures, many of them carrying either negative or positive charges. Extensive studies of substrate specificity of these transporters suggested that the only requirement for the drug to be a substrate of these efflux pumps is the presence of a hydrophobic domain capable of insertion into the phospholipid bilayer (Nikaido, 1996). This notion is strengthened by a good correlation between the hydrophobicity of the substituent at position C-6 of penicillin or at position C-7 of cephalosporin and the efficacy of these drugs to be pumped out by AcrB (Nikaido et al., 1998). In view of such correlations it was suggested that the substrate initially interacts with the bilayer and access to the core of the membrane-associated transporter is gained directly from the lipid phase, an idea similar to that independently proposed for other types of efflux pumps in Gram-positive bacteria and eukaryotes (Bolhuis et al., 1996a, b). More direct evidence in support of such a mechanism was obtained by the reconstitution of purified AcrB transporter into proteoliposomes (Zgurskaya and Nikaido, 1999a). This pump was shown to extrude, in an energy-dependent manner, fluorescent-labelled derivatives of phosphatidylethanolamine and this activity was inhibited by the well-established AcrB substrates such as bile salts and erythromycin. Capture of pump substrates from within the lipid bilayer is also consistent with the observation that human and animal homologues in the currently enlarged RND superfamily include several proteins that contain a ‘cholesterol-sensing’ (and sometimes putative cholesterol transport) domain, such as HMG-CoA reductase and Niemann-Pick Type C protein (NPC1) (Tseng et al., 1999).

Recently a ‘lead compound’ that inhibits all three P. aeruginosa pumps, MexAB-OprM, MexCD-OprJ and MexEF-OprN, was identified, and its various modifications have been evaluated (Renau et al., 1999). Interestingly, the presence of two amino (or equivalent) groups appears to be essential in addition to a lipophilic core. This is reminiscent of the finding that lipophilic monoamines inhibit the export of cholesterol from the lysosomes in animal cells, presumably by interfering with the function of another RND transporter, NPC1 (Lange and Steck, 1998).

Even if the substrate is recognized by a multidrug efflux pump on the basis of its hydrophobicity, the pump still must discriminate between its substrates and the natural components of cell membrane. The mechanism of multidrug recognition by RND-type transporters remains unknown. However, the recently solved structure of BmrR regulator protein of Bacillus subtilis in complex with the tetraphenylphophonium (TPP+) suggested how the substrate might be selected by multidrug pumps (Zheleznova et al., 1999). BmrR appears to have a cone-like drug-binding pocket which is formed by hydrophobic amino acids and contains a glutamate residue at the bottom, which carries the negative charge inside the binding pocket necessary to attract and bind a positively charged drug. This study also showed how the conformation of this protein can be altered through the interaction with cationic ligands so that it can accommodate molecules that are quite different in chemical structure. Nevertheless, a cone-like structure of the binding pocket imposes certain restrictions on the size and shape of ligands: planar molecules, such as ethidium bromide, do not bind to BmrR. Thus, in this case at least, three ligand properties affect its interaction with the binding protein: hydrophobicity, charge and shape of the molecule.

Efflux transporters with preference to cationic hydrophobic molecules, such as members of SMR (Small Multidrug Resistance) and MFS (Major Facilitator Superfamily) families have a conserved, membrane-embedded, glutamate–aspartate residue in TMS1 (Paulsen et al., 1996b; Edgar and Bibi, 1999). Site-directed mutagenesis studies in MFS transporter MdfA from E. coli have shown that this residue is important for the transport of cationic substrates but not for that of uncharged substrates (Edgar and Bibi, 1999). Similarly, the negative charge in TMS10 of the QacA, an MFS transporter from Staphylococcus aureus, was shown to play a critical role in resistance to divalent cations (Paulsen et al., 1996c). Hence, the electrostatic interactions between lipophilic cations and the membrane-embedded negatively charged residues of transporters may be analogous to those occurring in the binding site of BmrR (Zheleznova et al., 1999).

Membrane-embedded charged residues are also found in the RND-type transporters, which contain a conservative (D)DE motif in TMS 4. Heavy metal efflux pump CzcA with defects in any of these aspartate residues was not able to catalyse the proton–zinc antiport in vitro(Goldberg et al., 1999). However, mutated proteins were still able to support the facilitated diffusion of cations, suggesting the presence of a two-channel system. The DDE residues were suggested to comprise the proton transport channel, whereas the second channel is required for the transport of metal cations. These two channels might form the charge-relay system, where an electrical field generated by the transport of protons drives the metal cations through the cation channel to the periplasm.

AcrB and MexB multidrug transporters also have the (D)DE motif with the aspartate residues embedded into TMS 4 and the glutamate residue presumably exposed to the cytoplasmic side of the membrane, consistent with the experimental data that these transporters also function by proton antiport. However, the mechanism of substrate export apparently differs from that of CzcA. First, the substrates of AcrB/MexB are highly hydrophobic molecules. Some of them are uncharged, such as organic solvents or chloramphenicol, and others carry either positive or negative charges. Hence, the charge-relay system (Goldberg et al., 1999) is very unlikely to drive the export of such substrates. Second, AcrB/MexB can efflux drugs, such as many β-lactams, that do not cross the IM (Nikaido et al., 1998), suggesting that the substrate binding occurs either in the outer leaflet of the lipid bilayer or in the lipid–water interface in the periplasmic domains of these transporters. Thus, it is unlikely that drug-conducting channel traversing the phospholipid bilayer plays a major or indispensable role in the AcrB transporter and most of its homologues.

If the binding of the drug usually occurs on the periplasmic face of transporter, the large periplasmic domains of transporters may play a critical role in the drug efflux. One could imagine that the binding of the substrate triggers the proton transport, which, in turn, affects the charge distribution in the periplasmic domains of the transporter (Goldberg et al., 1999). The hydrophobic substrate bound to the site may dissociate then into the periplasm when the site becomes more hydrophilic or due to the change in the site conformation. However, recently identified RND transporters AmrB of Burkholderia pseudomallei (Moore et al., 1999) MexY of P. aeruginosa (Aires et al., 1999), and AcrD of E. coli(Rosenberg et al., 2000) export aminoglycosides, which are positively charged, highly hydrophilic molecules, along with hydrophobic substrates such as erythromycin and tetracycline. Thus, it is possible that the same transporter may contain multiple substrate-binding sites to expel drugs of different types, as was suggested recently in a study of the LmrP transporter of Lactococcus lactis, an MDR pump of the MFS superfamily (Putman et al., 1999).

Periplasmic membrane fusion proteins

  1. Top of page
  2. Abstract
  3. Introduction
  4. Inner membrane transporters
  5. Periplasmic membrane fusion proteins
  6. Outer membrane components of tri-partite efflux systems
  7. Acknowledgements
  8. References

The periplasmic components of multidrug efflux complexes of Gram-negative bacteria belong to a loose family of proteins named Membrane Fusion Proteins (MFPs) (Dinh et al., 1994). MFPs are common constituents of Gram-negative transport systems that export a variety of proteins, oligosaccharides, small molecules and divalent metal cations. The family owes its name to the C-terminal homology of its members with paramyxoviral membrane fusion proteins. Two hydrophobic domains were identified near the N- and C-termini of MFPs, which were proposed to interact with the inner and outer membrane components of complexes and by this means provide the transfer of substrates across the periplasm. Some MFPs, such as AcrA, a periplasmic component of the AcrAB-TolC multidrug efflux system, also carry covalently linked lipids at their N-terminus, although the lipid-deficient variant of AcrA is functional in drug efflux (Zgurskaya and Nikaido, 1999b). Consistent with the notion that MFPs connect the two membranes, the AcrA protein was found to be a highly asymmetric molecule with a length of about 20 nm, which is sufficient to span the periplasm (Zgurskaya and Nikaido, 1999b). The large α-helical region centrally located in the primary sequence of the AcrA protein was suggested to be responsible for its asymmetric shape (Fig. 1B). This α-helical region was also identified in almost all MFPs and was predicted with high probability to form a coiled coil (Pimenta et al., 1996; Johnson and Church, 1999).

Phylogenetic analysis indicates that MFPs cluster into several groups according to their substrate specificity and to the type of transporter with which they interact (Paulsen et al., 1997). However, regardless of the substrate, MFPs most probably function in the tri-partite complexes by preventing the substrate release into the periplasm. Two general models of MFP action were proposed: either through oligomerization with the OM component MFP may form a closed channel, which could become a route for the expelled substrates to pass through the periplasm, or the inner and outer membranes are simply brought together into close apposition by MFP for substrate transfer (Holland et al., 1990). Recent experimental evidence appears to support the latter model: the purified AcrA protein was shown to promote the close association of two membranes, and in addition possibly a hemifusion event under conditions similar to those found in periplasmic space (Zgurskaya and Nikaido, 1999a). Structural features of AcrA, such as its highly asymmetric shape, the presence of coiled–coil domains and two hydrophobic regions, are reminiscent of viral membrane fusion proteins. AcrA could thus act as a true MFP under conditions found in intact cells, perhaps creating the adhesion sites between the inner and outer membranes (Fig. 1B).

Structure prediction analysis of MFPs (Johnson and Church, 1999) suggested one possible mechanism for closing the distance between membranes using the AcrA protein. The α-helical region of MFPs appears to have a twofold symmetry (Fig. 1B): two regions of high coiled–coil probability of approximately equal length (four to five heptad repeats) are separated by a gap of five to 10 residues. This region is further bracketed by two copies of the lipoyl motif, which was previously described in biotin carboxyl carrier proteins and in the lipoyl domains of 2-oxo acid dehydrogenases (Neuwald et al., 1997; Johnson and Church, 1999). Such twofold symmetry of coiled–coil and lipoyl motifs suggests that a MFP could simply fold back on itself at the gap between helical regions, forming an α-helical hairpin (Johnson and Church, 1999).

However, helical regions differ substantially by size among MFPs and some of them, such as the haemolysin translocator accessory protein HlyD of E. coli, do not have clearly symmetrical coiled coils and could therefore have different structures. Two MFPs, AcrA and HlyD, appear to be trimeric in cross-linking experiments (Thanabalu et al., 1998; H. I. Zgurskaya and H. Nikaido, submitted) and possibly their coiled coils constitute part of a larger helical assembly. The structural and biochemical similarity between AcrA and viral MPFs is provocative and suggests that the central part of all these proteins may under certain conditions exist as a long, triple-stranded coiled coil (Weissenhorn et al., 1999). Furthermore, if MFPs and viral MFPs have a similar fold then the N- and C-termini of MFPs might be at the same end of a rod-shaped molecule, and such a model may explain the observation that a mutation in the HlyB IM pump can be suppressed by a C-terminal mutation of its periplasmic accessory protein HlyD (Schlor et al., 1997).

Even highly homologous MFPs, such as MexA and MexC that are the periplasmic components of multidrug resistance systems from P. aeruginosa, cannot be interchanged among the tri-partite complexes, suggesting the presence of specific interactions between MFPs and the corresponding IM transporters (Yoneyama et al., 1998). At least two MFPs, HlyD and AcrA, were shown to interact specifically with the corresponding IM transporter in cross-linking experiments (Thanabalu et al., 1998; H. I. Zgurskaya and H. Nikaido, submitted). The availability of substrate did not affect this interaction, suggesting that the complex between MFP and the IM transporter is stable. In contrast, in protein secretion systems such as the PrtDEF complex from Erwinia chrysanthemi, which secretes a metalloprotease (Letoffe et al., 1996), and the haemolysin transporter HlyBD-TolC (Thanabalu et al., 1998), the tri-partite complex, including the OM channel, appears to be transient and was shown to be formed only in the presence of the protein substrate. It was also suggested that the MFP component in such complexes plays an active role in the recruitment of the OM channel. However, presently it is not known how the specificity of assembly of the bi- and tri-partite complexes is determined.

Outer membrane components of tri-partite efflux systems

  1. Top of page
  2. Abstract
  3. Introduction
  4. Inner membrane transporters
  5. Periplasmic membrane fusion proteins
  6. Outer membrane components of tri-partite efflux systems
  7. Acknowledgements
  8. References

Unlike MFPs, the OM components of Gram-negative efflux systems (OMFs) (reviewed by Paulsen et al., 1997) are interchangeable between different multidrug efflux systems (Yoneyama et al., 1998). Furthermore, a single OMF protein can function with different pumps. For example, the most widely studied OM protein TolC from E. coli is essential for the activity of the drug transporter AcrB (Fralick, 1996) and the protein transporters HlyB (Holland et al., 1990) and CvaB (Gilson et al., 1990). Interestingly this protein also supports the activity of the RND drug transporters MexD and MexY from P. aeruginosa when these transporters are expressed in E. coli cells in the absence of their native OMFs (Srikumar et al., 1998). Another example is OprM from P. aeruginosa, which is a component of at least two tri-partite RND pumps, MexB and MexY.

OMF sequences are rather divergent. However, a recently derived structure of TolC from two-dimensional crystals (Koronakis et al., 1997) and the extensive sequence analysis (Johnson and Church, 1999) of OMFs led to the identification of several unusual structural features of these proteins. OMFs appear to comprise the transmembrane and periplasmic domains and form trimers in the OM. The primary sequences of these proteins have two tandem repeats suggesting that they have arisen from a duplication event. Hence, the two halves of OMF molecules might have similar three-dimensional structures. β-Strands were predicted in both halves of OMF sequences and probably constitute the transmembrane domains of these proteins. Consistent with this analysis is the finding that trimeric TolC forms ion-permeable pores in reconstituted bilayers (Benz et al., 1993). Besides, members of the OMF family seem to contain a large proportion of helical structure with high coiled–coil probability, which may correspond to the periplasmic domains of OMFs (Johnson and Church, 1999). As judged from the two-dimensional structure of the TolC channel, these periplasmic domains are sufficient to span most of the way to the IM (Koronakis et al., 1997).

The mechanistic role of OMFs in the export processes is unclear and it is not known how these proteins might receive translocating substrates from the IM transporters. In export assemblies that secrete large proteins directly into the medium, such as HlyBD-TolC and PrtDEF, substrate itself appears to play a critical role in the translocation process (Letoffe et al., 1996; Thanabalu et al., 1998). Although the ABC translocator HlyB and an MFP, HlyD, form a stable precomplex in the IM independently of the HlyA substrate, the formation of a functional tri-partite unit requires the substrate, HlyA (Thanabalu et al., 1998). Furthermore, translocating HlyA induces conformational changes in each of the three exporter proteins HlyD, HlyB and TolC, concomitant with assembly of a tri-partite complex.

The most plausible mechanism for the translocation of proteins seems to be ‘threading through’ in their unfolded state so that the translocating substrate interacts simultaneously with all three components of the export assembly. The interaction with unfolded HlyA presumably induces the conformational changes in an MFP, HlyD, that make possible the recruitment of the OM channel TolC and arrangement of the tri-partite complex (Thanabalu et al., 1998). The simultaneous interaction of the substrate with all three components appears to stabilize the complex because after translocation the complex disintegrates.

However, such a scenario seems unlikely for the exporters of small molecules, such as RND multidrug transporters, because of the completely different nature of their substrates and the structural differences between these systems. Although the overall structures of OM proteins and periplasmic MFPs that support translocation of proteins and small molecules are similar (see above), the topologies of IM transporters differ dramatically. As described above, RND multidrug transporters have two unique periplasmic domains protruding deep into the periplasm. One could imagine that these periplasmic domains may play a role similar to HlyA, that is to induce conformational changes in MFP and OMF and create the functional tri-partite complex. Thus, despite the analogous composition of the tri-partite complexes, the mechanism of transport across the two-membrane envelope of Gram-negative bacteria may be different between those that extrude large polypeptides and those that pump out diverse drug molecules.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Inner membrane transporters
  5. Periplasmic membrane fusion proteins
  6. Outer membrane components of tri-partite efflux systems
  7. Acknowledgements
  8. References

Studies in the authors' laboratory were supported by National Institute of Health (Grant AI-09644) and a grant from Microcide Pharmaceuticals, Inc.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Inner membrane transporters
  5. Periplasmic membrane fusion proteins
  6. Outer membrane components of tri-partite efflux systems
  7. Acknowledgements
  8. References
  • Aires, J.R., Köhler, T., Nikaido, H., Plésiat, P. (1999) Involvement of an active efflux system in the natural resistance of Pseudomonas aeruginosa to aminoglycosides. Antimicrob Agents Chemother 43: 26242628.
  • Benz, R., Maier, E., Gentschev, I. (1993) TolC of Escherichia coli functions as an outer membrane channel. Int J Med Microbiol Virol Parasitol Infect Dis 278: 183196.
  • Bolhuis, H., Van Veen, H.W., Brands, J.R., Putman, M., Poolman, B., Driessen, A.J.M., et al. (1996a) Energetics and mechanism of drug transport mediated by the lactococcal multidrug transporter LmrP. J Biol Chem 271: 2412324128.
  • Bolhuis, H., Van Veen, H.W., Molenaar, D., Poolman, B., Driessen, A.J.M., Konings, W.N. (1996b) Multidrug resistance in Lactococcus lactis: evidence for ATP-dependent drug extrusion from the inner leaflet of the cytoplasmic membrane. EMBO J 15: 42394245.
  • Dinh, T., Paulsen, I.T., Saier, M.H. Jr (1994) A family of extracytoplasmic proteins that allow transport of large molecules across the outer membranes of Gram-negative bacteria. J Bacteriol 176: 38253831.
  • Edgar, R. & Bibi, E. (1999) A single membrane-embedded negative charge is critical for recognizing positively charged drugs by the Escherichia coli multidrug resistance protein MdfA. EMBO J 18: 822832.
  • Fralick, J.A. (1996) Evidence that TolC is required for functioning of the Mar/AcrAB efflux pump of Escherichia coli. J Bacteriol 178: 58035805.
  • Gilson, L., Mahantym, H.K., Kolter, R. (1990) Genetic analysis of an MDR-like export system: the secretion of colicin V. EMBO J 9: 38753894.
  • Goldberg, M., Pribyl, T., Juhnke, S., Nies, D. (1999) Energetics and topology of CzcA, a cation/proton antiporter of the resistance-nodulation-cell division protein family. J Biol Chem 274: 2606526070.
  • Gotoh, N., Kusumi, T., Tsujimoto, H., Wada, T., Nishino, T. (1999) Topological analysis of an RND family transporter, MexD of Pseudomonas aeruginosa. FEBS Microbiol Lett 458: 3236.
  • Guan, L., Ehrmann, M., Yoneyama, H., Nakae, T. (1999) Membrane topology of the xenobiotic-exporting subunit, MexB, of the MexA,B-OprM extrusion pump in Pseudomonas aeruginosa. J Biol Chem 274: 1051710522.
  • Hagman, K.E., Pan, W., Spratt, B.G., Balthazar, J.T., Judd, R.C., Shafer, W.M. (1995) Resistance of Neisseria gonorrhoeae to antimicrobial hydrophobic agents is modulated by the mtrRCDE efflux system. Microbiology 141: 611622.
  • Holland, I.B., Kenny, B., Blight, M. (1990) Haemolysin secretion from E. coli. Biochimie 72: 131141.
  • Johnson, J.M. & Church, G.M. (1999) Alignment and structure prediction of divergent protein families: periplasmic and outer membrane proteins of bacterial efflux pumps. J Mol Biol 287: 695715.
  • Köhler, T., Michéa-Hamzehpour, M., Epp, S.F., Péchère, J.C. (1999) Carbapenem activities against Pseudomonas aeruginosa: Respective contributions of OprD and efflux systems. Antimicrob Agents Chemother 43: 424427.
  • Koronakis, V., Li, J., Koronakis, E., Stauffer, K. (1997) Structure of TolC, the outer membrane component of the bacterial type I efflux system, derived from two-dimentional crystals. Mol Microbiol 23: 617626.
  • Lange, Y. & Steck, T.L. (1998 Four cholesterol-sensing proteins. Curr Op Struct Biol8: 435439.
  • Letoffe, S., Delepelaire, P., Wandersman, C. (1996) Protein secretion in gram-negative bacteria: assembly of the three components of ABC protein-mediated exporters is ordered and promoted by substrate binding. EMBO J 15: 58045811.
  • Li, X.-Z., Livermore, D.M., Nikaido, H. (1994) Role of efflux pump(s) in intrinsic resistance of Pseudomonas aeruginosa: resistance to tetracycline, chloramphenicol, and norfloxacin. Antimicrob Agents Chemother 38: 17321741.
  • Li, X.Z., Nikaido, H., Poole, K. (1995) Role of mexA-mexB-oprM in antibiotic efflux in Pseudomonas aeruginosa. Antimicrob Agents Chemother 39: 19481953.
  • Lomovskaya, O., Warren, M.S., Lee, V. (2000)Efflux mechanisms: molecular and clinical aspects. In Antibiotic Resistance and Antibiotic Development. Hudges, D., and Andersson, D.I. (eds). Amsterdam: Harwood Academic Publishers, (in press).
  • Ma, D., Cook, D.N., Albert, M., Pon, N.G., Nikaido, H., Hearst, J.E. 1993 Molecular cloning and characterization of acrA and acrE genes of Escherichia coli. J Bacteriol 175: 62996313.
  • Mine, T., Morita, Y., Kataoka, A., Mizushima, T., Tsuchiya, T. (1999) Expression in Escherichia coli of a new multidrug efflux pump, MexXY, from Pseudomonas aeruginosa. Antimicrob Agents Chemother 43: 415417.
  • Moore, R.A., DeShazer, D., Reckseidler, S., Weissman, A., Woods, D.E. (1999) Efflux-mediated aminoglycoside and macrolide resistance in Burkholderia pseudomallei. Antimicrob Agents Chemother 43: 465470.
  • Neuwald, A.F., Liu, J.S., Lipman, D.J., Lawrence, C.E. (1997) Extracting protein alignment models from the sequence database. Nucleic Acid Res 25: 16651677.
  • Nikaido, H. (1996) Multidrug efflux pumps of gram-negative bacteria. J Bacteriol 178: 58535859.
  • Nikaido, H. (1998a) Antibiotic resistance caused by Gram-negative multidrug efflux pumps. Clin Infect Dis 27 (Suppl. 1): S32S41.
  • Nikaido, H. (1998b) Multiple antibiotic resistance and efflux. Curr Opin Microbiol 1: 516523.
  • Nikaido, H., Basina, M., Nguyen, V., Rosenberg, E.Y. (1998) Multidrug efflux pump AcrAB of Salmonella typhimurium excretes only those β-lactam antibiotics containing lipophilic side chains. J Bacteriol 180: 46864692.
  • Paulsen, I.T., Brown, M.H., Skurray, R.A. (1996a) Proton-dependent multidrug efflux systems. Microbiol Rev 60: 575608.
  • Paulsen, I.T., Skurray, R.A., Tam, R., Saier, M.H. Jr, Turner, R.J., Weiner, J.H., et al. (1996b) The SMR family: a novel family of multidrug efflux proteins involved with the efflux of lipophilic drugs. Mol Microbiol 19: 11671175.
  • Paulsen, I.T., Brown, M.H., Littlejohn, T.G., Mitchell, B.A., Skurray, R.A. (1996c) Multidrug resistance proteins QacA and QacB from Staphylococcus aureus: membrane topology and identification of residues involved in substrate specificity. Proc Natl Acad Sci USA 93: 36303635.
  • Paulsen, I.T., Park, J.H., Choi, P.S., Saier, M.H., Jr (1997) A family of gram-negative bacterial outer membrane factors that function in the export of proteins, carbohydrates, drugs and heavy metals from Gram-negative bacteria. FEMS Microbiol Lett 156: 18.
  • Paulsen, I.T., Sliwinski, M.K., Saier, M.H., Jr (1998) Microbial genome analyses: global comparisons of transport capabilities based on phylogenies, bioenergetics and substrate specificities. J Mol Biol 277: 573592.
  • Pimenta, A., Blight, M., Clarke, D., Holland, I.B. (1996) The gram-negative cell envelope ‘springs’ to life: coiled-coil trans-envelope proteins. Mol Microbiol 19: 643645.
  • Poole, K., Krebes, K., McNally, C., Neshat, S. (1993) Multiple antibiotic resistance in Pseudomonas aeruginosa: evidence for involvement of an efflux operon. J Bacteriol 175: 73637372.
  • Putman, M., Koole, L.A., Van Veen, H.W., Konings, W.N. (1999) The secondary multidrug transporter LmrP contains multiple drug interaction sites. Biochemistry 38: 1390013905.
  • Renau, T.E., Leer, R., Flamme, E.M., Sangalang, J., She, M.W. et al. (1999 Inhibitors of efflux pumps in Pseudomonas aeruginosa potentiates the activity of the fluoroquinolone antibacterial levofloxacin. J Med Chem 42: 49284931.
  • Rosenberg, E.Y., Ma, D., Nikaido, H. 2000 AcrD of Escherichia coli in an aminoglycoside efflux pump. J Bacteriol 182: 17541756.
  • Saier, M.H., Jr, Tam, R., Reizer, A., Reizer, J. (1994) Two novel families of bacterial membrane proteins concerned with nodulation, cell division and transport. Mol Microbiol 11: 841847.
  • Saier, M.H., Jr, Paulsen, I.T., Sliwinski, M.K., Pao, S.S., Skurray, R.A., Nikaido, H. (1998) Evolutionary origins of multidrug and drug-specific efflux pumps in bacteria. FASEB J 12: 265274.
  • Schlor, S., Schmidt, A., Maier, E., Benz, R., Goebel, W., Gentschev, I. (1997) In vivo and in vitro studies on interactions between the components of the haemolysin translocator protein HlyD. Mol Gen Genet 256: 306319.
  • Srikumar, R., Kon, T., Gotoh, N., Poole, K. (1998) Expression of Pseudomonas aeruginosa multidrug efflux pumps MexA-MexB-OprM and MexC-MexD-OprJ in a multidrug-sensitive Escherichia coli strain. Antimicrob Agents Chemother 42: 6571.
  • Thanabalu, T., Koronakis, E., Hudges, C., Koronakis, V. (1998) Substrate-induced assembly of a contiguous channel for protein export from E. coli: reversible bridging of an inner-membrane translocase to an outer membrane exit pore. EMBO J 17: 64876496.
  • Tseng, T., Gratwick, K.S., Kollman, J., Park, D., Nies, D.H., Goffeau, A., et al. (1999) The RND permease superfamily: an ancient, ubiquitous and diverse family that includes human disease and development proteins. J Mol Microbiol Biotechnol 1: 107125.
  • Wandersman, C. (1992) Secretion across the bacterial outer membrane. Trends Genet 8: 317322.
  • Weissenhorn, W., Dessen, A., Calder, L.J., Harrison, S.C., Skehel, J.J., Wiley. D.C. (1999) Structural basis for membrane fusion by enveloped viruses. Mol Membr Biol 16: 39.
  • Yoneyama, H., Ocaktan, A., Gotoh, N., Nishino, T., Nakae, T. (1998) Subunit swapping in the Mex-extrusion pumps in Pseudomonas aeruginosa. Biochem Biophys Res Commun 244: 898902.
  • Zgurskaya, H.I. & Nikaido, H. (1999a) Bypassing the periplasm: reconstitution of the AcrAB multidrug efflux pump of Escherichia coli. Proc Natl Acad Sci USA 96: 71907195.
  • Zgurskaya, H.I. & Nikaido, H. (1999b) AcrA from Escherichia coli is a highly asymmetric protein capable of spanning the periplasm. J Mol Biol 285: 409420.
  • Zheleznova, E.E., Markham, P.N., Neyfakh, A.A., Brennan, R.G. (1999) Structural basis of multidrug recognition by BmrR, a transcriptional activator of a multidrug transporter. Cell 96: 353362.