Biochimica et Biophysica Acta 1794 (2009) 782793

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / b b a p a p

Review

Drug transport mechanism of the AcrB efux pump

Klaas M. Pos
Cluster of Excellence Frankfurt Macromolecular Complexes and Institute of Biochemistry, Goethe-University Frankfurt, Max-von-Laue-Str. 9, D-60438 Frankfurt am Main, Germany

a r t i c l e i n f o a b s t r a c t

Article history: In Gram-negative bacteria such as Escherichia coli and Pseudomonas aeruginosa, tripartite multidrug efux
Received 19 October 2008 systems extrude cytotoxic substances from the cell directly into the medium bypassing periplasm and the
Received in revised form 18 December 2008 outer membrane. In E. coli, the tripartite efux system AcrA/AcrB/TolC is the pump that extrudes multiple
Accepted 19 December 2008
antibiotics, dyes, bile salts and detergents. The inner membrane component AcrB, a member of the Resistance
Available online 3 January 2009
Nodulation cell Division (RND) family, is the major site for substrate recognition and energy transduction of
Keywords:
the entire tripartite system. The drug/proton antiport processes in this secondary transporter are suggested
Macromolecular complex to be spatially separated, a feature frequently observed for primary transporters like membrane-bound
Tripartite drug pump system ATPases. The recently elucidated asymmetric structure of the AcrB trimer reveals three different monomer
AcrB conformations proposed to represent consecutive states in a directional transport cycle. Each monomer
RND shows a distinct tunnel system with entrances located at the boundary of the outer leaet of the inner
Membrane transporter membrane and the periplasm through the periplasmic porter (pore) domain towards the funnel of the trimer
Drug efux and TolC. In one monomer a hydrophobic pocket is present which has been shown to bind the AcrB
Peristaltic mechanism
substrates minocyclin and doxorubicin. The energy conversion from the proton motive force into drug efux
Alternate access
includes proton binding in (and release from) the transmembrane part. The conformational changes
Drug transport
Antiport observed within a triad of essential, titratable residues (D407/D408/K940) residing in the hydrophobic
Binding change mechanism transmembrane domain appear to be transduced by transmembrane helix 8 and associated with the
conformational changes seen in the periplasmic domain. From the asymmetric structure a possible peristaltic
pump transport mechanism based on a functional rotation of the AcrB trimer has been postulated. The novel
drug transport model combines the alternate access pump mechanism with the rotating site catalysis of F1Fo
ATPase as originally postulated by Jardetzky and Boyer, respectively, and suggests a working hypothesis for
the transport mechanism of RND transporters in general.
2009 Elsevier B.V. All rights reserved.

1. Resistance nodulation division (RND) pump AcrB
Acriavine resistance protein B (AcrB) is a member of the
Resistance Nodulation cell Division (RND) superfamily. Members of
the RND superfamily (TC#2.A.6) are found ubiquitously in bacteria,
archaea and eukaryotes. They mainly function as proton/substrate
antiporters and belong to eight phylogenetic families [1]. Clustering
patterns in these families of the RND superfamily correlate with
substrate specicity. For example, the heavy metal efux (HME) family
(TC#2.A.6.1) catalyzes export of heavy metal ions (Zn2+, Cu2+) and the
hydrophobe/amphiphile efux-1 (HAE1) family (TC#2.A.6.2) cata-
lyzes the export of multiple drugs (Fig. 1). The best characterized
example of the HME family is CzcA from Ralstonia eutropha, acting as a
heavy metal-ion/proton antiporter [2]. AcrB of Escherichia coli and
Abbreviations: Acr, acriavine resistance; AcrB, acriavine resistance protein B; L
monomer, loose monomer (access monomer); T monomer, tight monomer (binding
monomer); O monomer, open monomer (extrusion monomer); MDR, multidrug
resistance; OM, Outer Membrane; IM, Inner Membrane; MIC, Minimal Inhibitory
Concentration; RMSD, root mean square distance; PDB, Protein Data Bank
* Tel.: +49 69 798 29251.
E-mail address: pos@em.uni-frankfurt.de.
MexB of Pseudomonas aeruginosa are the best studied members of the
HAE1 family [3,4]. Human NPC1 being involved in lipid transport
through the membrane and associated with the NiemannPick C1
disease belongs to the eukaryotic sterol transporter (EST) family
(TC#2.A.6.6) [5].
The RND members of the HME and HAE1 families function in
conjunction with a membrane fusion protein (MFP; TC #8.A.1) and an
outer membrane factor (OMF; TC #1.B.17) [1] (Fig. 2). The RND
proteins are the actual pumps of the tripartite RND/MFP/OMF efux
system, since both substrate specicity and energy transduction are
attributed to this inner-membrane component. Nevertheless, it
should be emphasized that every single component of the entire
tripartite RND/MFP/OMF efux machinery appears to be essential,
indicating that communication between the three subunits is central
to a concerted drug transport mechanism.
2. Structure of AcrB
Twenty-ve AcrB structures have been submitted to the Protein
Data Bank (www.pdb.org) to date. The rst AcrB structure (year 2002,
PDB entry: 1IWG) [6] was based on 3.5 X-ray diffraction data from

1570-9639/$ see front matter 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.bbapap.2008.12.015
 K.M. Pos / Biochimica et Biophysica Acta 1794 (2009) 782793 783

Fig. 1. Substrates and inhibitors of the AcrABTolC efux system. The system confers resistance to a wide variety of noxious substances like dyes, different classes of antibiotics,
detergents, bile salts and small organic molecules. Phe-Arg--naphthylamide and 1-(1-naphthylmethyl)-piperazine (NMP) inhibit RND/MFP/OMF efux systems. From Seeger et al.
[11] with permission.

crystals containing one unliganded AcrB monomer in the asymmetric
unit and for which the trimer is generated by exact crystallographic
symmetry so that the structure is enforced to be symmetric (Fig. 3).
In 2006 and 2007, there were independent reports of structures
describing an asymmetric AcrB trimer (PDB entries: 2DHH, 2DR6,
2DRD, 2GIF, 2HRT, 2J8S) [79]. The best resolution structure (2.5 ,
PDB entry 2J8S) was obtained with AcrB/designed ankyrin repeat
protein (DARPin) co-crystals [9]. Two of the deposited asymmetric
structures were derived from 3.1 and 3.3 data of AcrB/minocyclin
and AcrB/doxorubicin co-crystals, respectively [7]. From the asym-
metric structures, a model for drug transport based on conformational
cycling of the monomers by the RND pump AcrB has been proposed
(Fig. 4) [7,8,10,11].
2.1. Symmetric AcrB
The AcrB monomer consists of 1049 amino acids and exhibits
sequence homology and similar structural architecture between its
N- and C-terminal half, indicating an early gene duplication event
[12]. Full length AcrB contains twelve transmembrane -helices
constituting the transmembrane domain (Fig. 3A, D). TM4 and TM10
are surrounded by the other transmembrane helices of the
monomer and harbour the residues D407, D408 (TM4) and K940
(TM10) (Fig. 3D), which appear to play an essential role in proton
translocation [13,14]. The connecting loops between TM1 and TM2
as well as TM7 and TM8 fold into two large periplasmic domains
(Fig. 3A), which account for approximately half of the amino acids of
the full length AcrB molecule. The AcrB monomers form a trimer
which appears to be stabilized by the intermonomer connecting
loops protruding into the adjacent monomer (Fig. 3B). The
periplasmic part of AcrB consists of the TolC docking domain
(divided into the DN and DC subdomains, not shown), which is
located most distant from the membrane plane (Fig. 3A), and the
porter (pore) domain. The TolC docking domain exhibits a funnel-
like structure narrowing to a central pore located in the porter
domain, the latter composed of subdomains PN1, PN2, PC1 and PC2
784 K.M. Pos / Biochimica et Biophysica Acta 1794 (2009) 782793

Fig. 2. Schematic drawing based on the X-ray structures of the tripartite multidrug efux system AcrABTolC of Gram-negative E. coli. AcrB (RND component, in blue color)
resides in the inner membrane and is responsible for substrate recognition/selection and energy transduction. Drugs are captured at the outer leaet of the inner membrane
and extruded in a coupled exchange with protons. TolC (OMF component, yellow) forms a pore in the outer membrane which is extended by a long periplasmic conduit.
AcrA (MFP component, red) mediates contact between AcrB and TolC. The presence of all three components is essential for the MDR phenotype. From Seeger et al. [11] with
permission.

(Fig. 3A, C). The characteristic central pore is formed by three -
helices (designated pore helices), donated by the PN1 subdomains
of each AcrB monomer (Fig. 3B, C). The pore has a small diameter
and therefore does not allow drug passage in this conformation. The
PC1 and PC2 subdomains constitute a cleft at the periplasmic
periphery of the porter domain which is suggested to accommodate
AcrA (Fig. 3C) [6]. Chimeric analysis of AcrA (397 residues) pointed
towards a main role of C-terminal domain residues 290357 in this
interaction with AcrB [15]. At the interior of the latter protein near
the membrane plane, the central pore leads to a central cavity and
further to a 3035 wide, presumably lipid-lled transmembrane
hole dened by the ring-like arrangement of the TM helices of the
trimer (Fig. 3A, D). Three vestibules at the monomer interface
located just above the membrane plane lead towards the central
cavity (Fig. 3A, D).
2.2. Asymmetric AcrB
Recently, we [8] and others [7,9] solved a structure of AcrB which
does not entail a three-fold symmetry of the trimer. In contrast to the
symmetric structure, this asymmetric structure reveals three different
monomer conformations representing consecutive states in a trans-
port cycle, loose (L), tight (T) and open (O). The structural changes in
the T monomer create a hydrophobic pocket, which is not present in
the other monomers. We [8] assumed that this pocket is a substrate
binding pocket inside the porter domain and Murakami et al. [7]
showed binding of minocyclin, 9-bromo-minocyclin and doxorubicin
to this pocket. The impact of systematic alanine scanning mutations in
the binding pocket was evaluated by determination of minimal
inhibitory concentration of drug and the results pointed towards a
special role for F610 in the transport process [16].
From the asymmetric structures, a model for drug transport based
on conformational cycling of the monomers has been proposed (Fig.
4) [79].
3. Conformational cycling and binding change mechanism
Comparison of the periplasmic domain of the asymmetric AcrB
structure (PDB entry: 2GIF) viewed from the periplasmic side
perpendicular to the membrane plane with the structure of the /
subunits and subunits of bovine F1Fo ATP synthase (PDB entry: 1BMF
[17]) viewed from the cytoplasm leads to the overwhelming
impression that the structures are analogous (Fig. 5). The rod of the
ATPase is tilted towards one of the three / subunits. Likewise, in
one monomer of the AcrB trimer the pore -helix is tilted towards the
neighbouring monomer's PN2 subdomain. In the ATPase and AcrB
structures, the monomers are in distinct different conformations,
designated loose (L), tight (T) and open (O). However, there is a
marked difference between the designation of the / subunits of
bovine F1Fo ATP synthase and that of the AcrB monomers. The rod of
the ATPase is tilted towards the bopenQ / subunits (despite
appearances in Fig. 5), whereas the PN1 subdomain of the bopenQ
(O) monomer of AcrB is tilted towards the btightQ (T) monomer. The
functional relevance of the rod of the ATPase as energy transducer
and for catalytic cooperativity [18] is more likely comparable with
TM8 of AcrB. The coil to helix transition of TM8 in the AcrB bopenQ
monomer appears to emphasize its role as an energy transducer.
Before structural details of the F1Fo ATP synthase were known,
Paul Boyer postulated the binding change mechanism for this
enzyme, describing a conformational cycling of the ( and)
subunits through the states loose, tight and open, leading to the
synthesis of ATP [1921].
One of the three key points of his binding change mechanism is
that the energy input (i.e. proton motive force) is not used to form ATP,
but to release the ATP molecule from the subunit. The second key
point is the catalytic cooperativity: ATP can only be released when
ADP and Pi are bound to another monomer within the hexameric F1
part of the ATP synthase (so called bi-site activation). The third key
point is the rotational catalysis. It is important to note that physical
 K.M. Pos / Biochimica et Biophysica Acta 1794 (2009) 782793 785

Fig. 3. AcrB structural features. (A) Side view. Trimeric AcrB is composed of the TolC docking domain exhibiting a funnel-like structure narrowing to a central pore located in the
porter (pore) domain and the transmembrane domain conning a central cavity. (B) Top view. The trimeric state of AcrB is mainly stabilized by the intermonomer connecting loops.
A central pore is formed by three -helices, donated by the PN1 subdomains of each AcrB monomer. (C) Top view onto the porter domain with its subdomains PN1, PN2, PC1 and PC2.
The lateral cleft between subdomains PC1 and PC2 is suggested to accommodate AcrA [6]. (D) Top view onto the transmembrane domain. TM4 and TM10 are surrounded by other 10
transmembrane helices. The transmembrane helices of the monomers conne a large central cavity. From Seeger et al. [11] with permission.

rotation (of the rod) and functional rotation (/ subunits) are both
features of the F1Fo ATPase, whereas the rotational catalysis postulated
below for AcrB only suggests functional rotation, not physical rotation.
In analogy to the binding change mechanism (also called the
alternating site mechanism) our group [8] and Murakami et al. [7]
independently postulated an analogue alternating site functional
rotation mechanism for the transport of drugs by AcrB (Fig. 4). The
general concept describes a functional rotation starting with the
(loose) binding of substrate to a low afnity site on the L monomer,
followed by conformational change to the T conformer and tight
binding of the substrate in the designated binding pocket and nally
conversion to the O monomer resulting in the release of the substrate
towards TolC. The O monomer nally converts to the L conformation
restarting the cyclic event (Fig. 4). It has to be pointed out that the
direction of events e.g. conversion from L to T, T to O, and back to L is
based on the interpretation of structural data only and that further
experimental proof is needed to establish the consecutive conforma-
tional states. To date, three considerably different conformational
states of the AcrB trimer have been crystallized: the ball-looseQ or LLL
conformation derived from diffraction data of crystals grown in the
R32 space group [6], the asymmetric or LTO conformation obtained
from crystals lacking the three-fold symmetry [7,8] and the ball-tightQ
or TTT conformation (Eicher, Pos, unpublished). Moreover, cross-
linking data supports the conformational exibility within the AcrB
trimer to form the LLT, LTT and TTO conformations [22,23]. As is
indicated in Fig. 6, these conformational states have been incorporated
into a more detailed scheme for drug transport, taking into account
the cooperativity or bi-site activation, in analogy to the binding
786 K.M. Pos / Biochimica et Biophysica Acta 1794 (2009) 782793

Fig. 4. Schematic representation of the AcrB alternating site functional rotation transport mechanism. The conformational states loose (L), tight (T), and open (O) are colored blue,
yellow and red, respectively. (A) Side-view schematic representation of two of the three monomers of the AcrB trimer. AcrA and TolC are indicated in light green and light purple
colors, respectively. (B) The lateral grooves in the L and T monomer indicate the substrate binding sites. The different geometric forms reect low (triangle), high (rectangle), or no
(circle) binding afnity for the transported substrates. In the rst state of the cycle, a monomer binds a substrate (acridine) in its transmembrane domain (L conformation),
subsequently transports the substrate from the transmembrane domain to the hydrophobic binding pocket (conversion to T conformation) and nally releases the substrate in the
funnel toward TolC (O conformation). AcrA is postulated to participate in the transduction of the conformational changes from AcrB to TolC, which results in the opening of the TolC
channel and the facilitation of drug extrusion to the outside of the cell. Adapted from Seeger et al. [8] and modied. From Seeger et al. [11] with permission.

Fig. 5. Structural analogy between the / subunits and subunits of bovine F1Fo ATP synthase (PDB entry: 1BMF [17]) (left, viewed from the cytoplasm) and the periplasmic domain
of the asymmetric AcrB structure (PDB entry: 2GIF [8]) (right, viewed from the periplasmic side perpendicular to the membrane plane). The structures are presented as ribbon
diagrams and the designation of the individual monomers is indicated (Loose (L), Tight (T) and Open (O)). See text for details. From Seeger et al. [11] with permission.
 K.M. Pos / Biochimica et Biophysica Acta 1794 (2009) 782793 787

Fig. 6. Schematic representation of the AcrB alternating site functional rotation transport mechanism extended by postulated intermediate steps. The conformational states loose (L),
tight (T), and open (O) are colored blue, yellow and red, respectively. The lateral grooves in the L and T monomer indicate the substrate binding sites. The different geometric forms
reect low (triangle), high (rectangle), or no (circle) binding afnity for the transported substrates. Both states LTO at the far left and far right are identical to the states shown in Fig.
4. State TTT is postulated to occur at high substrate concentrations. The states LLL and LLT are postulated to occur in the absence or at low substrate concentrations. See text for details.
From Seeger et al. [11] with permission.

change mechanism by the F1Fo ATP synthase. In the proposed scheme
(Fig. 6), the LLL conformation of the AcrB trimer describes a state
when the protein is depleted from substrate (bresting stateQ [24]). It is
notable that a structure of the substrate-free 33 F1-ATPase
subcomplex (without the subunit) of Bacillus PS3 has been
crystallized [25] and was shown to contain only low-afnity
nucleotide binding sites (analogous to the LLL conformation). The
presence of the rod introduces the structural and functional
asymmetry between the 33 subunits and is responsible for
cooperative catalysis [18]. Even in the axle-less (i.e. short ) F1-
ATPase mutant [26], interaction of the short and the / subunits is
necessary for directed rotation and torque generation.
Binding of substrate to the L monomer results in the conversion
from L to T and, according to the binding change mechanism, a second
substrate has to bind to another monomer as a prerequisite for the
release of the substrate to the outside i.e. substrate has to bind to
another monomer before conversion of the substrate-occupied T
monomer to the O monomer can occur (Fig. 6). The conformational
change from the T to the O monomer is proposed to be an energy
dependent step, in analogy with the energy dependence of the release
788 K.M. Pos / Biochimica et Biophysica Acta 1794 (2009) 782793

of ATP from the -subunit. Proton uptake from the periplasm is
anticipated in this step (Fig. 6).
Structurally, the conversion of the T monomer to the O monomer
might be sterically facilitated by the formation of an adjacent T
monomer (Figs. 6 and 7). Conversion of the L monomer to the T
monomer comprises substantial movement of the PN2 subdomain
(LTO to TTO). This causes the loss of restraints for the adjacent PN1
subdomain of the already existent T monomer from the LTO trimer
and facilitates the pronounced PN1 subdomain inclination towards
the other T monomer (Figs. 5 and 7) [7,8]. The next step in the
proposed cycle would be the conversion of the TTO trimer conforma-
tion into a TOO conformation. However, two neighbouring O
monomers would cause an intermonomeric steric clash like has
been shown for neighbouring O and L monomers (Fig. 7). Hence, TTO
to TTL (i.e. LTT) conversion precedes the LTT to LTO conversion (Fig. 6).
The combined effect of binding of a second substrate to the AcrB
trimer (leading to the formation of a second T monomer) with the
binding of (a) proton(s) to the transmembrane domain (driven by the
proton motive force), results in the formation of the O monomer.
Binding of a third substrate to the L conformer of an LTT AcrB
trimer (Fig. 6) is considered to be possible (e.g. in the case of high
substrate concentration) and will trigger the creation of the third T
monomer resulting in the ball-tightQ conformation (TTT), a conforma-
tion found in AcrB crystals (Eicher, Pos et al., unpublished).
Interestingly, cross-link data supports the formation of more than
one of the same conformer (TTL and LLT) in E. coli membranes [23],
whereas in crystallization experiments the formation of symmetric
forms of the AcrB trimer (LLL, TTT) appears to be predominant to the
formation of asymmetric AcrB trimers in crystals. For the crystalline
state, this may be explained by the reduced free energy of symmetric
crystal contacts as compared to asymmetric ones.
4. Tunnels guiding the substrate
The asymmetric AcrB structure [8,9] (PDB entries: 2GIF, 2J8S)
accommodates tunnels leading to and away from the hydrophobic
substrate binding pocket (Fig. 8). The L monomer (Fig. 8A, blue)
contains a tunnel (designated tunnel 2) starting at the lateral cleft
(subdomain PC1/PC2 interface) about 15 above the putative
membrane plane and protruding into the porter domain. In the T
monomer (Fig. 8A, yellow), the formation of an additional tunnel
(tunnel 1) in the porter domain is apparent with its entrance located
at height of the TM8 and TM9 groove, which has been postulated to
accommodate substrates originating from the membrane [69]. Good
resolution (2.5 ) data [9] show the presence of a dodecylmaltoside
molecule (which is both detergent and AcrB substrate) located in the
TM8/TM9 groove of the L monomer and conversion of L to T might
open the entrance of tunnel 1 towards the substrate binding pocket.
Tunnel 1 merges with tunnel 2 close to the hydrophobic substrate
binding pocket (Fig. 8, inset with bound minocyclin). In the O
monomer (Fig. 8B, red), the lateral opening and tunnels 1 and 2 are
absent due to the closure of the lateral cleft caused by the large
conformational change of the PC2 subdomain and the coil to helix
transition of TM8 during the T to O monomer transition (Fig. 8B) [7,8].
In the O monomer, tilting of the central helix due to the movement of
the PN1 subdomain creates an exit pathway (tunnel 3, Fig. 8B) which
leads to the funnel of the AcrB trimer. In a conformational cycling
scenario, the tunnels form an alternate access pathway between the
TM8/TM9 groove (tunnel 1) or/and the lateral side (tunnel 2) of the
AcrB trimer and tunnel 3 towards the funnel and TolC (Fig. 8) [8,9]. In
addition, we postulated a peristaltic mode of drug transport through
the tunnels, actively guiding the substrates from the periplasmic
space/membrane towards TolC and the outside of the cell [8,11].
5. Substrate transport through the tunnel system
The lateral access of tunnel 2 is located 15 above the putative
membrane plane (Fig. 8A) and at rst sight appears to be suitable for
the collection of substrates residing in the periplasmic space, i.e. the -
lactams. However, resistance against -lactams due to the action of the
AcrABTolC system of Salmonella typhimurium (AcrB of S. typhimurium
shares 94k identical residues with AcrB of E. coli) was positively

Fig. 7. Main structural differences of the PN1 and PN2 subdomains in the AcrB trimer. Right: The three AcrB PN1 and PN2 subdomains of the L (in blue), T (yellow) and O (red)
monomers in top view as ribbon presentation are superimposed onto the PN1 and PN2 subdomains of the symmetric AcrB trimer model depicted in grey. Inset (on the left): The PN1
subdomain of the O monomer would sterically clash with the PN2 subdomain of a neighbouring L monomer (in grey). Conversion of the L monomer (grey) into the T monomer
(yellow) would allow movement of the PN1 subdomain of the O monomer without steric hindrance. From Seeger et al. [11] with permission.
 K.M. Pos / Biochimica et Biophysica Acta 1794 (2009) 782793 789

Fig. 8. Visualization of tunnels in the porter (pore) domain of the trimeric AcrB peristaltic drug efux pump. The AcrB monomers are presented in (A) blue (loose, L), (B) yellow
(tight, T) and (c) red (open, O). The tunnels are highlighted as green surfaces in a ribbon model of the AcrB trimer. Inset: In the T monomer (yellow), a hydrophobic pocket is dened
by phenylalanines 136, 178, 610, 615, 617, and 628; valines 139 and 612; isoleucines 277 and 626; and tyrosine 327 at the PN2/PC1 interface. Bound minocyclin is depicted with the
observed electron density in a 2Fo-Fc electron density map contoured 1 (Eicher, Seeger, Pos et al., unpublished). Panels A and B represent in each case a one third conversion of a full
LYTOYL cycle. From Seeger et al. [11] with permission.

correlated with the increment in the penicillin side chain octanol-
water partition coefcient, the preferred substrates for the pump
being nafcillin, cloxacillin and benzyl-penicillin [27]. The hydrophobic
side chains of these amphiphilic substrates are likely to submerge into
the outer leaet of the inner membrane, whereas the -lactam ring
would remain in the periplasm. The outer leaet localization makes
these compounds suitable candidates for entering the AcrB tunnel 1
via the TM8/TM9 groove. It remains elusive whether other substrates
like e.g. uoroquinolones and macrolides are also garnered from the
membrane or whether tunnel 1 can be considered the general
entrance tunnel for all substrates. The role of tunnel 2 remains elusive
as well. The lateral opening of tunnel 2 is located at the PC1/PC2 cleft,
which was postulated to be an interaction site for AcrA [6,28]. The
latter might function as a exible lid and might alternately open and
shield the entrance towards the periplasm. An alternative hypothesis
would be that tunnel 2 rather functions as an exit tunnel for non-
substrates to leave the tunnel system (Fig. 8). The porter domain of
AcrB and its homologues is responsible for substrate selectivity [29
31], which is in complete accordance to the localization of the
substrates minocyclin and doxorubicin in the hydrophobic pocket of
the T monomer (Fig. 8 inset) (Eicher, Pos et al., unpublished) and [7].
Therefore, substrate binding occurs in the porter domain, where the
hydrophobic pocket acts as a selectivity lter [7, 8]. The results from
the periplasmic domain swapping experiments [2931] imply that
selection takes place not until in the porter domain and suggest that
non-substrates are also transported from the transmembrane domain
towards the binding pocket. Deselected compounds have to leave the
porter domain either via back-diffusion through tunnel 1, but might
also be bypassed through tunnel 2 (Fig. 8). Residues lining tunnel 1
and 2 are charged and/or polar rather than hydrophobic. To allow
transport through the rather hydrophilic tunnel(s), a peristaltic
transport mechanism would account for the unidirectionality of
substrate transport and the movement of hydrophobic compounds
through the rather hydrophilic tunnels.
Once substrate is recognized and bound to the binding pocket in
the T monomer, conformational change is triggered and leads to
conversion from the T monomer to the O monomer (Figs. 4, 6 and 8).
However, as stated above, binding of substrate is most likely not the
exclusive trigger for conformational change from T to O. In accordance
to the binding change mechanism theory for the ATP synthase, bound
substrate might not be released before a second substrate binds to
another monomer i.e. the adjacent L monomer (Fig. 6). In this context,
it is postulated for AcrB that only if the trimer is in an LTT (or possibly
TTT) conformation, release of the substrate and conformational
change to LTO (or possibly via TTO and LTT) occur. It appears that
detergent solubilized AcrB is preferably present in its asymmetric
conformation, as is implied from cross-linking [22,23,32] and DARPin
binding [9]. Notwithstanding, AcrB readily crystallizes at the ball-
looseQ symmetric form [6,24,3337] and also in its ball-tightQ state
(Eicher, Pos et al., unpublished) in a wide range of crystallization
conditions, including the presence of substrates, whereas the
asymmetric LTO conformation was only retrieved at a more dened
crystallization condition and in combination with DARPins. The
formation of crystals containing AcrB in different conformational
states shows the exibility of this membrane protein to adopt
intermediate cycling states.
As has been shown very recently, exibility and the formation of
intermediate states appear also to be common for the symmetric AcrB
trimer [24,35,37]. The extent of the conformational differences
between the published symmetric AcrB forms is small, whereas the
monomers within the asymmetric trimer show substantial conforma-
tional differences. The AcrB structure derived from AcrB/DARPin co-
crystals diffracting to 2.5 (PDB entry: 2J8S) showed very low RMSD
values compared to the 2.93.0 structures derived from our group
(PDB entries: 2GIF, 2HRT), indicating near identity of the structures [9].
The suggested path of the substrate does not yet consider a role for
prominent features such as the vestibules [6] and the central cavity
(Fig. 3A, D). Recent structural studies [24,3437] are, however,
indicative for an implication of the central cavity on drug binding
and transport. Careful analysis of this structural data derived from R32
crystals e.g. on twinning and more distinct substrate assignment
would be helpful to substantiate the role of the cavity on substrate
binding. Substrate binding studies with isolated detergent soluble
AcrB from E. coli indicated drug binding afnities between 5.5 and 74
M [38] comparable to drug binding afnities of MdfA (MF Super-
family) [39] and EmrE (SMR Superfamily) [40]. The results suggested
a drug to AcrB monomer stoichiometry of 1, which is in agreement
with the proposed binding of substrates in the cavity [34], but stands
790 K.M. Pos / Biochimica et Biophysica Acta 1794 (2009) 782793

in contrast to the drug to AcrB trimer stoichiometry of 1 found in the
asymmetric AcrB structure [7].
6. Transport of protons and energy transduction
Transport of drugs from the cell to the outside by the AcrA/AcrB/
TolC system is coupled to proton transport from the periplasm to the
cytoplasm. Proton binding and release takes place in the transmem-
brane domain, where key residues involved in proton transport have
been identied by mutagenesis, functional assays and studying the
asymmetric AcrB structure. The drug/proton antiport catalysis i.e.
energy transduction and substrate transport in this secondary
transporter appears to be spatially separated, a feature frequently
observed for primary transporters like membrane-bound ATPases.
Single AcrB mutants where D407, D408, K940, T978 or R971 were
substituted by alanine caused complete loss of function of the
tripartite efux system [13,14]. In the asymmetric structure [79],
K940 and R971 (and to some extent also D407) have distinct different
side chain conformations when comparing the L and T monomers
with the O monomer (Fig. 9A, B). The subtle side chain reorientations
observed in the O monomer appear to be leveraged via TM8 to the
periplasmic subdomains PC2 and PN1 where large rigid body
conformational changes appear to have taken place (Fig. 9A, B). The
PC2 and PN1 movement leads to closure of tunnels 1 and 2 (and the
substrate binding pocket) and opening of tunnel 3 towards the funnel
(Fig. 9C). Moreover, the periplasmic loop connecting TM3 and TM4 is
downshifted approx. 3 towards the membrane during the T to O
monomer transition and appears to relieve constraints for the PN1
domain to adopt its inclined conformation as described above. The
exact chain of events from proton binding towards coil-to-helix
transition of TM8 (Fig. 9B) and the apparent downshift of the
transmembrane domain during T to O monomer conformational
change are not completely clear from the structure per se [79] and
need further investigation.
From the local environment of the D407, D408, K940 and R971 side
chains in the different monomers, the protonation state of these side
chains can be assessed. As AcrB is energized by the proton-motive
force, transient protonation of titratable groups within the transmem-
brane domain of the protein can be expected to be the mechanism
which delivers the energy required for the conformational changes
described above. A prominent K940 (TM10) side chain reorientation
away from D408 and towards D407 (both on TM4) in the O
conformation (Fig. 9C, lower panel), and a bulging of TM5 towards
TM4 and TM10 [8] are indicative of protonation and deprotonation
events. The side chain arrangement and especially the putative role of
R971 in pKa determination of Asp407 and/or Asp408 are reminiscent
to that of bacteriorhodopsin [41].
As stated above, the conversion of the T monomer to the O
monomer is postulated to be the energy-dependent step in accor-
dance with the F1Fo ATP synthase binding change mechanism. Hence,

Fig. 9. (A) Side view superimposition of the AcrB L monomer (grey) and the PN1, PC2 subdomains as well as TM8 of the O monomer (red). (B) In a close-up view of the boxed region,
(A) the N-terminal parts of TM8 (residues 859 to 880) and the PC2 subdomain (residues 679721 and 822858) are superimposed. The structures in blue, yellow and red represent
the conformations of the TM8 and the C-terminal -sheets of the PC2 subdomains of the L, T and O monomers, respectively. The rest of the PC2 subdomain is depicted in transparent
grey (L and T monomers) or red (O monomer). (C) Side view representation of the AcrB trimer as shown in Fig. 8. The boxed regions of the transmembrane domains comprising the
essential residues Asp407, Asp408, Lys940, Arg971, and Thr978 of the L, T and O monomers are shown in a close-up view. Proton uptake is anticipated in the T monomer (middle box)
and is postulated to lead to the side chain reorientation of Asp407, Lys940 and Arg971. These side chain conformational changes might be coupled to the coil-to-helix transition
shown in (B) and the PN1/PC2 subdomain movement as shown in (A). From Seeger et al. [11] with permission.
 K.M. Pos / Biochimica et Biophysica Acta 1794 (2009) 782793 791

proton uptake from the periplasm is expected in the T monomer.
Further uptake of protons from the periplasm and subsequent release
to the cytoplasm in any of the monomers is anticipated but cannot be
elucidated from the structural information up to date. High resolution
data of AcrB in its asymmetric conformation to locate structural water
molecules in the transmembrane domain, as well as further
biochemical data are necessary to obtain more certainty about the
chain of protonation/deprotonation events.
7. Stoichiometry of drug/proton antiport
To date there are only a few reports on in vitro assays using
reconstitution of the inner membrane RND component (AcrB/AcrD)
into liposomes [42,43]. Reconstitution of activity was dependent on
addition of the soluble form of the MFP component (AcrA) to AcrB or
AcrD containing proteoliposomes. Despite the success of retrieving
AcrB and AcrD activity outside the native tripartite setup, readout is
still not suitable for quantication of the drug/proton antiport activity.
Transport has been shown to be dependent on pH and assumed to be
dependent on as well. Considering the wide variety of substrates
(anionic, cationic neutral) stoichiometric prediction becomes highly
complicated and suggests a coupling exibility of the drug pump
system as has been shown for MdfA [44]. As has been strongly
advocated by Nikaido et al. [27,43,45] and is supported by the latest
asymmetric AcrB structures [69], substrates might be in general
collected from the outer leaet of the inner membrane, where they
most probably are at a high concentration compared to the periplasm/
cytoplasm. The question on electrogenicity of the system would then
become obsolete, since substrates would not be transported across a
membrane. Hence, the coupling proton(s) transported from the
periplasm to the inside of the cell by AcrB would dene the overall
electrogenic mechanism.
8. AcrA/AcrB/TolC drug transport hypothesis
The tripartite AcrA/AcrB/TolC setup has been suggested to occur
independent of the presence of substrates or proton motive force [46
48], with one report describing a stimulation in the formation of the
complex in the presence of substrates [47], as has been shown for the
E. coli type I extrusion system HlyB/HlyD/TolC in the presence of its
substrate HlyA (haemolysin) [49]. A hypothesis on drug transport by
the AcrA/AcrB/TolC three component system is depicted in Fig. 10.
Aside from the conformational signal transduction from AcrB to TolC
via the AcrA linker [7,8,28,5054], there is a line up of drug molecules
from the AcrB funnel continuing into the TolC channel. The latter
postulate results from two observations: 1) residues lining the
observed AcrB tunnels and the TolC channel are polar rather than
hydrophobic and 2) the TolC channel is recruited by other systems to
extrude a wide variety of compounds including polypeptides as in the
E. coli type I extrusion system HlyB/HlyD/TolC transporting haemo-
lysin. The ABC transporter HlyB is energized by the free energy of
hydrolysis of ATP to drive HlyA transport to the medium. The export
process is dependent on HlyD, a member of the MFP family, and TolC.
It seems plausible that HlyA extrusion is reliant on a mechanical
(maybe peristaltic) movement through the transport system rather
than diffusion. Despite the lack of covalent bonds between the

Fig. 10. Schematic representation of the AcrB alternating site functional rotation transport mechanism. The conformational states loose (L), tight (T), and open (O) are colored blue,
yellow and red, respectively. Only two of the three monomers of the AcrB trimer are shown in side-view. AcrA and TolC are indicated in light green and grey, respectively. The
proposed proton translocation site (D407, D408, and K940) is indicated in the membrane part of each monomer. In the rst state of the cycle (from left to right), a monomer binds a
substrate (acridine) in its transmembrane domain (L conformation), subsequently transports the substrate from the transmembrane domain to the hydrophobic binding pocket
(conversion to T conformation) and nally releases the substrate in the funnel toward TolC (O conformation). Peristaltic transport of drugs through the AcrB tunnels (indicated by the
red arrow) and through TolC in combination to the line up of drug molecules inside the AcrB funnel and the TolC channel would account for a strict unidirectional movement towards
the outside of the cell. The conversion from the T monomer to the O monomer conformation is suggested to be the major energy-requiring (proton motive force-dependent) step in
this functional rotation cycle and requires the binding of a proton to the proton translocation site (D407, D408, and K940) from the periplasm. The release of a proton from the proton
translocation site to the cytoplasm might occur during conversion from the O monomer to the L monomer (as depicted) or from the latter to the T monomer. AcrA is expected to
participate in the transduction of the conformational changes from AcrB to TolC (indicated by black arrows), which results in the movement of the proximal part of TolC and the
facilitation of drug extrusion to the outside of the cell. From Seeger et al. [11] with permission.
792 K.M. Pos / Biochimica et Biophysica Acta 1794 (2009) 782793
separate AcrB substrate molecules, the mode of transport by a
mechanical, peristaltic movement [55] of a chain of substrates
would account for strict unidirectional movement towards the outside
of the cell.
The last decade has brought us an enormous leap in understanding
on how the AcrA/AcrB/TolC and homologue systems transport drugs
and other compounds across the inner and outer membrane of the
Gram-negative cell. Despite the wealth of the recent structural and
functional data on all components of the tripartite system, we still lack
the basic understanding on drug/proton coupling and stoichiometry,
energy transduction, drug binding and transport kinetics.
Acknowledgements
I like to thank the members of our research group, Thomas Eicher
and Lorenz Brandsttter, the past member, Markus Seeger, and my
collaborators at the University of Konstanz, Kay Diederichs and Andr
Schiefner. This review includes many ne ideas of scientist whom I
had the pleasure to discuss transport mechanisms in general and of
AcrB in particular. I like to thank C. von Ballmoos, S. Murakami, E. Bibi,
S. Schuldiner, D. Friedman, K. Locher, D. Picot, D. Mueller, C. Tate, B.
Luisi, H. van Veen, G. Sennhauser, P. Amstutz, M. Grtter, J. Bohnert, O.
Lomovskaya, W. Kern and many others for their open discussion
philosophy.
References
[1] M.H. Saier Jr., I.T. Paulsen, Phylogeny of multidrug transporters, Semin. Cell Dev.
Biol. 12 (2001) 205213.
[2] M. Goldberg, T. Pribyl, S. Juhnke, D.H. Nies, Energetics and topology of CzcA, a
cation/proton antiporter of the resistance-nodulation-cell division protein family,
J. Biol. Chem. . 274 (1999) 2606526070.
[3] S. Murakami, A. Yamaguchi, Multidrug-exporting secondary transporters, Curr.
Opin. Struct. Biol. 13 (2003) 443452.
[4] K. Poole, Multidrug efux pumps and antimicrobial resistance in Pseudomonas
aeruginosa and related organisms, J. Mol. Microbiol. Biotechnol. 3 (2001)
255264.
[5] N. Ohgami, D.C. Ko, M. Thomas, M.P. Scott, C.C. Chang, T.Y. Chang, Binding between
the NiemannPick C1 protein and a photoactivatable cholesterol analog requires a
functional sterol-sensing domain, Proc. Natl. Acad. Sci. U. S. A. 101 (2004)
1247312478.
[6] S. Murakami, R. Nakashima, E. Yamashita, A. Yamaguchi, Crystal structure of
bacterial multidrug efux transporter AcrB, Nature 419 (2002) 587593.
[7] S. Murakami, R. Nakashima, E. Yamashita, T. Matsumoto, A. Yamaguchi, Crystal
structures of a multidrug transporter reveal a functionally rotating mechanism,
Nature 443 (2006) 173179.
[8] M.A. Seeger, A. Schiefner, T. Eicher, F. Verrey, K. Diederichs, K.M. Pos, Structural
asymmetry of AcrB trimer suggests a peristaltic pump mechanism, Science 313
(2006) 12951298.
[9] G. Sennhauser, P. Amstutz, C. Briand, O. Storchenegger, M.G. Grutter, Drug export
pathway of multidrug exporter AcrB revealed by DARPin inhibitors, PLoS Biol. 5
(2007) e7.
[10] S. Murakami, Multidrug efux transporter, AcrBthe pumping mechanism, Curr.
Opin. Struct. Biol. 18 (2008) 459465.
[11] M.A. Seeger, K. Diederichs, T. Eicher, L. Brandstatter, A. Schiefner, F. Verrey, K.M.
Pos, The AcrB efux pump: conformational cycling and peristalsis lead to
multidrug resistance, Curr. Drug Targets 9 (2008) 729749.
[12] T. Dinh, I.T. Paulsen, M.H. Saier Jr., A family of extracytoplasmic proteins that allow
transport of large molecules across the outer membranes of gram-negative
bacteria, J. Bacteriol. 176 (1994) 38253831.
[13] L. Guan, T. Nakae, Identication of essential charged residues in transmembrane
segments of the multidrug transporter MexB of Pseudomonas aeruginosa, J.
Bacteriol. 183 (2001) 17341739.
[14] Y. Takatsuka, H. Nikaido, Threonine-978 in the transmembrane segment of the
multidrug efux pump AcrB of Escherichia coli is crucial for drug transport as a
probable component of the proton relay network, J. Bacteriol. 188 (2006)
72847289.
[15] C.A. Elkins, H. Nikaido, Chimeric analysis of AcrA function reveals the importance
of its C-terminal domain in its interaction with the AcrB multidrug efux pump, J.
Bacteriol. 185 (2003) 53495356.
[16] J.A. Bohnert, S. Schuster, M.A. Seeger, E. Fahnrich, K.M. Pos, W.V. Kern, Site-
directed mutagenesis reveals putative substrate binding residues in the Escheri-
chia coli RND efux pump AcrB, J. Bacteriol. 190 (2008) 82258229.
[17] J.P. Abrahams, A.G. Leslie, R. Lutter, J.E. Walker, Structure at 2.8 A resolution of F1-
ATPase from bovine heart mitochondria, Nature 370 (1994) 621628.
[18] C. Kaibara, T. Matsui, T. Hisabori, M. Yoshida, Structural asymmetry of F1-ATPase
caused by the gamma subunit generates a high afnity nucleotide binding site, J.
Biol. Chem. 271 (1996) 24332438.
[19] D.D. Hackney, G. Rosen, P.D. Boyer, Subunit interaction during catalysis:
alternating site cooperativity in photophosphorylation shown by substrate
modulation of [18O]ATP species formation, Proc. Natl. Acad. Sci. U. S. A. 76
(1979) 36463650.
[20] R.L. Hutton, P.D. Boyer, Subunit interaction during catalysis. Alternating site
cooperativity of mitochondrial adenosine triphosphatase, J. Biol. Chem. 254
(1979) 99909993.
[21] P.D. Boyer, The ATP synthasea splendid molecular machine, Annu. Rev. Biochem.
66 (1997) 717749.
[22] Y. Takatsuka, H. Nikaido, Site-directed disulde cross-linking shows that cleft
exibility in the periplasmic domain is needed for the multidrug efux pump AcrB
of Escherichia coli, J. Bacteriol. 189 (2007) 86778684.
[23] M.A. Seeger, C. von Ballmoos, T. Eicher, L. Brandstatter, F. Verrey, K. Diederichs, K.
M. Pos, Engineered disulde bonds support the functional rotation mechanism of
multidrug efux pump AcrB, Nat. Struct. Mol. Biol. 15 (2008) 199205.
[24] C.C. Su, M. Li, R. Gu, Y. Takatsuka, G. McDermott, H. Nikaido, E.W. Yu, Conformation
of the AcrB multidrug efux pump in mutants of the putative proton relay
pathway, J. Bacteriol. 188 (2006) 72907296.
[25] Y. Shirakihara, A.G. Leslie, J.P. Abrahams, J.E. Walker, T. Ueda, Y. Sekimoto, M.
Kambara, K. Saika, Y. Kagawa, M. Yoshida, The crystal structure of the nucleotide-
free alpha 3 beta 3 subcomplex of F1-ATPase from the thermophilic Bacillus PS3 is
a symmetric trimer, Structure 5 (1997) 825836.
[26] S. Furuike, M.D. Hossain, Y. Maki, K. Adachi, T. Suzuki, A. Kohori, H. Itoh, M.
Yoshida, K. Kinosita Jr., Axle-less F1-ATPase rotates in the correct direction,
Science 319 (2008) 955958.
[27] H. Nikaido, M. Basina, V. Nguyen, E.Y. Rosenberg, Multidrug efux pump AcrAB of
Salmonella typhimurium excretes only those beta-lactam antibiotics containing
lipophilic side chains, J. Bacteriol. 180 (1998) 46864692.
[28] D. Nehme, K. Poole, Interaction of the MexA and MexB components of the MexAB
OprM multidrug efux system of Pseudomonas aeruginosa: identication of MexA
extragenic suppressors of a T578I mutation in MexB, Antimicrob. Agents
Chemother. 49 (2005) 43754378.
[29] S. Eda, H. Maseda, T. Nakae, An elegant means of self-protection in gram-negative
bacteria by recognizing and extruding xenobiotics from the periplasmic space, J.
Biol. Chem. 278 (2003) 20852088.
[30] C.A. Elkins, H. Nikaido, Substrate specicity of the RND-type multidrug efux
pumps AcrB and AcrD of Escherichia coli is determined predominantly by two
large periplasmic loops, J. Bacteriol. 184 (2002) 64906498.
[31] E.B. Tikhonova, Q. Wang, H.I. Zgurskaya, Chimeric analysis of the multicomponent
multidrug efux transporters from gram-negative bacteria, J. Bacteriol. 184 (2002)
64996507.
[32] S. Murakami, N. Tamura, A. Saito, T. Hirata, A. Yamaguchi, Extramembrane central
pore of multidrug exporter AcrB in Escherichia coli plays an important role in drug
transport, J. Biol. Chem. 279 (2004) 37433748.
[33] K.M. Pos, A. Schiefner, M.A. Seeger, K. Diederichs, Crystallographic analysis of AcrB,
FEBS Lett. 564 (2004) 333339.
[34] E.W. Yu, G. McDermott, H.I. Zgurskaya, H. Nikaido, D.E. Koshland Jr., Structural
basis of multiple drug-binding capacity of the AcrB multidrug efux pump,
Science 300 (2003) 976980.
[35] E.W. Yu, J.R. Aires, G. McDermott, H. Nikaido, A periplasmic drug-binding site of
the AcrB multidrug efux pump: a crystallographic and site-directed mutagenesis
study, J. Bacteriol. 187 (2005) 68046815.
[36] D. Das, Q.S. Xu, J.Y. Lee, I. Ankoudinova, C. Huang, Y. Lou, A. DeGiovanni, R. Kim, S.H.
Kim, Crystal structure of the multidrug efux transporter AcrB at 3.1A resolution
reveals the N-terminal region with conserved amino acids, J. Struct. Biol. 158
(2007) 494502.
[37] S. Tornroth-Horseeld, P. Gourdon, R. Horseeld, L. Brive, N. Yamamoto, H. Mori, A.
Snijder, R. Neutze, Crystal structure of AcrB in complex with a single
transmembrane subunit reveals another twist, Structure 15 (2007) 16631673.
[38] C.C. Su, E.W. Yu, Ligand-transporter interaction in the AcrB multidrug efux
pump determined by uorescence polarization assay, FEBS lett. 581 (2007)
49724976.
[39] O. Lewinson, E. Bibi, Evidence for simultaneous binding of dissimilar substrates by
the Escherichia coli multidrug transporter MdfA, Biochemistry 40 (2001)
1261212618.
[40] T.R. Muth, S. Schuldiner, A membrane-embedded glutamate is required for ligand
binding to the multidrug transporter EmrE, EMBO J. 19 (2000) 234240.
[41] R. Govindjee, S. Misra, S.P. Balashov, T.G. Ebrey, R.K. Crouch, D.R. Menick, Arginine-
82 regulates the pKa of the group responsible for the light-driven proton release in
bacteriorhodopsin, Biophys. J. 71 (1996) 10111023.
[42] H.I. Zgurskaya, H. Nikaido, Bypassing the periplasm: reconstitution of the AcrAB
multidrug efux pump of Escherichia coli, Proc. Natl. Acad. Sci. U. S. A. 96 (1999)
71907195.
[43] J.R. Aires, H. Nikaido, Aminoglycosides are captured from both periplasm and
cytoplasm by the AcrD multidrug efux transporter of Escherichia coli, J. Bacteriol.
187 (2005) 19231929.
[44] O. Lewinson, J. Adler, G.J. Poelarends, P. Mazurkiewicz, A.J. Driessen, E. Bibi, The
Escherichia coli multidrug transporter MdfA catalyzes both electrogenic and
electroneutral transport reactions, Proc. Natl. Acad. Sci. U. S. A. 100 (2003)
16671672.
[45] X.Z. Li, D. Ma, D.M. Livermore, H. Nikaido, Role of efux pump(s) in intrinsic
resistance of Pseudomonas aeruginosa: active efux as a contributing factor to
beta-lactam resistance, Antimicrob. Agents Chemother. 38 (1994) 17421752.
[46] T. Touze, J. Eswaran, E. Bokma, E. Koronakis, C. Hughes, V. Koronakis, Interactions
underlying assembly of the Escherichia coli AcrABTolC multidrug efux system,
Mol. Microbiol. 53 (2004) 697706.
 K.M. Pos / Biochimica et Biophysica Acta 1794 (2009) 782793
[47] E.B. Tikhonova, H.I. Zgurskaya, AcrA, AcrB, and TolC of Escherichia coli form a
stable intermembrane multidrug efux complex, J. Biol. Chem. 279 (2004)
3211632124.
[48] F. Husain, M. Humbard, R. Misra, Interaction between the TolC and AcrA proteins
of a multidrug efux system of Escherichia coli, J. Bacteriol. 186 (2004) 85338536.
[49] L. Balakrishnan, C. Hughes, V. Koronakis, Substrate-triggered recruitment of the
TolC channel-tunnel during type I export of hemolysin by Escherichia coli, J. Mol.
Biol. 313 (2001) 501510.
[50] S. Lobedanz, E. Bokma, M.F. Symmons, E. Koronakis, C. Hughes, V. Koronakis, A
periplasmic coiled-coil interface underlying TolC recruitment and the assembly of
bacterial drug efux pumps, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 46124617.
[51] D. Nehme, X.Z. Li, R. Elliot, K. Poole, Assembly of the MexAB-OprM multidrug
efux system of Pseudomonas aeruginosa: identication and characterization of
793
mutations in MexA compromising MexA multimerization and interaction with
MexB, J. Bacteriol. 186 (2004) 29732983.
[52] J.F. Stegmeier, G. Polleichtner, N. Brandes, C. Hotz, C. Andersen, Importance of the
adaptor (membrane fusion) protein hairpin domain for the functionality of
multidrug efux pumps, Biochemistry 45 (2006) 1030310312.
[53] V.N. Bavro, Z. Pietras, N. Furnham, L. Perez-Cano, J. Fernandez-Recio, X.Y. Pei, R.
Misra, B. Luisi, Assembly and channel opening in a bacterial drug efux machine,
Mol. Cell 30 (2008) 114121.
[54] Z. Pietras, V.N. Bavro, N. Furnham, M. Pellegrini-Calace, E.J. Milner-White, B.F.
Luisi, Structure and mechanism of drug efux machinery in Gram negative
bacteria, Curr. Drug Targets 9 (2008) 719728.
[55] L. Vaccaro, K.A. Scott, M.S. Sansom, Gating at both ends and breathing in the
middle: conformational dynamics of TolC, Biophys. J. 95 (2008) 56815691.