1008
Biochemical Society Transactions (2005) Volume 33, part 5
Similarities between ATP-dependent and
ion-coupled multidrug transporters
H. Venter, S. Shahi, L. Balakrishnan, S. Velamakanni, A. Bapna, B. Woebking and H.W. van Veen1
Department of Pharmacology, University of Cambridge, Tennis Court Road, Cambridge CB2 1PD, U.K.
Abstract
The movement of drugs across biological membranes is mediated by two major classes of membrane
transporters. Primary-active, ABC (ATP-binding cassette) multidrug transporters are dependent on ATPbinding/hydrolysis, whereas secondary-active multidrug transporters are coupled to the proton (or sodium)motive force that exists across the plasma membrane. Recent work on LmrA, an ABC multidrug transporter
in Lactococcus lactis, suggests that primary- and secondary-active multidrug transporters share functional
and structural features. Some of these similarities and their implications for the mechanism of transport by
ABC multidrug transporters will be discussed.
Introduction
The ever-increasing incidence of MDR (multidrug resistance)
due to the overexpression of multidrug transporters is an
important cause of failure of the drug-based treatment of
patients with cancers or infections by pathogenic microorganisms [1,2]. Even in non-overexpressing cells, drug efflux
by multidrug transporters may interfere with the effective
dosing of potentially efficacious drugs by altering the
pharmacokinetic properties of these drugs [2,3].
Multidrug transporters exhibit a broad substrate specificity
and are able to lower the intracellular drug concentration
to subtoxic levels by mediating the active extrusion of
drugs from the cell. They are integral membrane proteins,
which can be broadly classified according to their bioenergetic requirements into primary- and secondary-active transporters. In primary-active transport, as is the case for ABC
(ATP-binding cassette) multidrug transporters, the translocation of substrate is coupled directly to the hydrolysis
of ATP [4]. In contrast, secondary-active transporters utilize
the chemiosmotic energy derived from the electrochemical
gradient of proton/sodium ions, which is present across
the cytoplasmic membrane (also referred to as the proton/
sodium motive force). These systems mediate drug efflux
in a drug/ion antiport reaction (Figure 1) [5]. However,
recent studies on the ABC multidrug transporter LmrA in
Lactococcus lactis suggest that the boundaries between these
two major classes of transporters might not be as distinct as
was originally thought [6]. In this review, these studies will
be discussed in the context of functional and structural similarities that can be observed between primary- and secondary-
Key words: ATP-binding cassette transporter (ABC transporter), carboxylate, LmrA, multidrug
resistance, secondary-active transporter, transmembrane segment (TMS).
Abbreviations used: ABC, ATP-binding cassette; BCRP, breast cancer resistance protein;
MD, membrane domain; MDR, multidrug resistance; MFS, major facilitator superfamily; MRP,
multidrug resistance protein; NBD, nucleotide-binding domain; RND, resistance-nodulation-cell
division; SMR, small multidrug resistance; TMS, transmembrane segment.
1
To whom correspondence should be addressed (email hwv20@cam.ac.uk).
C 2005
Biochemical Society
active multidrug transporters in prokaryotic and eukaryotic
cells.
ABC transporter with a secondary-active
multidrug translocator domain
Bacterial LmrA is a homologue of the human MDR P-glycoprotein MDR1 (also termed ABCB1) and can functionally
substitute for P-glycoprotein in human lung fibroblast
cells [7]. LmrA consists of an N-terminal MD (membrane
domain), containing six TMS (transmembrane segments),
followed by the NBD (nucleotide-binding domain). The
protein homodimerizes to form the minimal functional unit
containing two MDs and two NBDs [8]. In P-glycoprotein,
these four domains are fused into a single polypeptide. The
MDs contain the substrate translocation pathways, whereas
the NBDs couple the transport reaction at the MD to ATPbinding/hydrolysis [2,4]. When a truncated LmrA protein
was generated containing the MD in the absence of the
NBD, this protein (termed LmrA-MD) retained the substrate
specificity of the full-length protein, but mediated drug influx
instead of efflux [6]. Detailed biochemical analysis of purified
LmrA-MD, functionally reconstituted into proteoliposomes,
revealed that the protein transports drugs in a proton motive
force-dependent fashion by catalysing a drug/H+ symport
reaction analogous to that of secondary-active solute uptake
systems. Similar to LmrA-MD, full-length LmrA can mediate
drug/H+ co-transport during ATP-dependent efflux [6].
In another study, it was observed that in the presence of
an inwardly directed drug concentration gradient in ATPdepleted cells, LmrA can mediate the reverse transport (or
uptake) of drug [9]. This uptake reaction was competitively
inhibited by LmrA substrates and was also affected by
mutations in the NBD or MD, which inactivate these domains
or reduce their activity. As LmrA can mediate drug influx and
efflux, these observations on LmrA and LmrA-MD pointed
to the reversibility of drug transport. This conclusion is striking, as reversibility of substrate transport is usually linked
Mechanistic and Functional Studies of Proteins
Figure 1 Two classes of multidrug transporters
ABC multidrug transporters are ATP-dependent. Secondary-active multidrug transporters are driven by the transmembrane electrochemical
gradient of protons (proton motive force) or sodium ions (sodium motive
force). ‘In’ and ‘Out’ refers to the inside and outside of the cytoplasmic
membrane respectively.
to secondary-active transporters; ABC transporters are often
thought to be unidirectional [9]. Interestingly, a significant
sequence similarity has recently been observed between
LmrA and MexB, a secondary-active multidrug transporter
belonging to the RND (resistance-nodulation-cell division)
family. The shared sequence corresponds to a hairpin region
of two membrane-spanning helices and the connecting
loop between these helices [10]. Our findings generate the
interesting question whether substrate translocation mechanisms are conserved between primary- and secondary-active
multidrug transporters and also whether the functional similarity is supported by further structural similarities between
these two classes of proteins.
Conservation of global structure of MDs
As the MDs of transporters form the pathways through which
substrates are translocated across the phospholipid bilayer
and hence, contain substrate-binding sites, it is interesting to
compare MDs of primary- and secondary-active transporters
at the structural level. Even though, in general there is little
overall sequence identity between MD regions of members
of these two classes of transporters, or even between members within each class, interesting similarities exist between
the secondary and tertiary structures of MDs. Recent crystal
structures for members of the MFS (major facilitator superfamily) and for ABC transporters reveal that MDs are
mostly α-helical [11–16] and that these helices often span
the membrane in a two-times-six helices motif. Although
some members of the MFS consist of 14 TMS, the two
additional segments can easily be removed without complete
loss of transport activity [17]. Likewise, certain mammalian
MRPs (multidrug resistance proteins) contain an additional
N-terminal MD containing five TMS, in addition to the
two MDs found in P-glycoprotein and dimeric LmrA. This
extra domain can be removed without compromising drug
transport activity [18]. Even transporters of the SMR (small
multidrug resistance) family, containing four TMS, seem
to abide to the two-times-six helices rule. A recent X-ray
crystal structure of the Escherichia coli SMR member EmrE
revealed a tetramer consisting of two heterodimers with
inverted orientation with respect to each other. Each heterodimer forms a six-helical bundle, consisting of helices 1, 2
and 3 of each monomer, while helix 4 does not form part
of this six-helix arrangement [19]. However, some caution
is needed here, as a different tertiary structure was obtained
for dimeric EmrE by electron crystallography [20]. The
overall similarity between structures of MDs of ABC transporters and secondary-active transporters may suggest that
the mechanical movements of helices induced by substrate
binding and transport might be similar, e.g. based on rigidbody movement of the two six-helix domains against one
another. This movement would then allow alternating access
of substrate-binding sites to the inner and outer membrane
surface. Indeed, this notion is consistent with structural
data for both secondary-active transporters (LacY, GlpT
and OxltT) and ABC transporters (MsbA and BtuCD)
[11–16,21].
Similarities in drug-binding sites
Given that the substrate specificities of primary- and
secondary-active multidrug transporters overlap, it follows
that there may be similarities in the drug-binding pockets
of these transporters. Much insight into multidrug binding
has been gained from X-ray crystallography on soluble
multidrug-binding transcriptional regulators such as BmrR in
Bacillus subtilis and QacR in Staphylococcus aureus [22–24].
BmrR activates the expression of the secondary-multidrug
transporter Bmr in response to the binding of amphiphilic
compounds [24]. Cationic tetraphenylphosphonium is able to
penetrate into the hydrophobic core of BmrR, where it forms
hydrogen bonds and stacking interactions with hydrophobic
and aromatic residues, and makes an ion-pair interaction
with a buried glutamate residue (Glu134 ) [23,24]. Similar to
BmrR, drug binding to QacR creates an expansive multisite
drug-binding pocket in which four buried glutamate residues,
Glu57 , Glu58 , Glu90 and Glu120 , are available to neutralize
positively charged moieties in stereochemically dissimilar
drugs [22].
Since the amphiphilic compounds that bind to BmrR
and QacR are transported by Bmr and QacA/B and many
C 2005
Biochemical Society
1009
1010
Biochemical Society Transactions (2005) Volume 33, part 5
other multidrug transporters, similar drug–protein interactions may occur in all these proteins [25–27]. Indeed, the
importance of carboxylates in a hydrophobic environment,
in the translocation of cationic drugs by secondary-active
multidrug transporters is well documented. A carboxylate in
the first TMS is highly conserved in multidrug transporters
of the MFS and the SMR families [28]. The MFS transporter,
MdfA from E. coli, for example, contains a single membrane
embedded glutamate (Glu26 ) in the middle of TMS 1.
Mutation of this glutamate residue has a profound effect on
the substrate recognition profile of MdfA. The carboxylate
at position 26 is essential for the transport of cationic drugs,
while the transport of neutral chloramphenicol is unaffected.
Therefore Glu26 is likely to be part of the substrate-binding
site [29].
QacA and QacB are 14 TMS secondary-active MFS
multidrug efflux proteins from the pathogen S. aureus [30].
These transporters share substantial sequence identity, but
differ in that QacA confers resistance on uni- as well as
bi-valent cations, while QacB only confers resistance on
univalent cations. This difference is due to the presence of
an aspartate (Asp323 ) in TMS 10 of QacA. QacB mutants
containing an acidic residue at the equivalent position displayed resistance to bivalent cations similar to QacA [30].
Carboxylates important for substrate specificity have also
been identified in the secondary-active multidrug transporter
LmrP from L. lactis [31].
To date, much less is known about the role of carboxylates
in drug–protein interactions in ABC transporters. The human
BCRP (breast cancer resistance protein; also referred to
as ABCG2 or mitoxanthrone resistance protein) contains
only one membrane-embedded acidic residue, Glu446 in
TMS 2. Even a conserved mutation of this residue to an
aspartate resulted in a marked decrease in mitoxanthrone and
SN-38 efflux in BCRP-expressing cells [32]. A single membrane-embedded carboxylate in TMS 2 is also a feature of the
‘topological homologues’ of BCRP such as the ABC multidrug transporter Pdr5p in Saccharomyces cerevisiae and
Cdr1p in Candida albicans. Although as yet untested,
these acidic residues are probably important for substrate
recognition too. Carboxylates with a possible role in
drug binding have also been identified in the second MD
(TMS 6–11) of mammalian MRP1 (ABCC1) [33]. Interestingly, certain ABC transporters, including LmrA and
P-glycoprotein, do not contain Glu/Asp residues in the TMS.
Interactions between these proteins and cationic drugs will
be based on similar principles as observed in mammalian ion
channels and receptor proteins, where aromatic residues can
stabilize the binding of ligands with cationic and aromatic
moieties through cation–π and π –π (stacking) interactions
respectively [34–36].
Many of the biochemical observations listed above were
confirmed in crystallographic studies on AcrB, an MFS member in E. coli [37,38]. The structure shows the presence of an
expanded drug-binding pocket, as observed for BmrR and
QacR, lined with aromatic amino acids and containing buried
carboxylates. Drugs bind to this surface through stacking of
C 2005
Biochemical Society
their aromatic rings to aromatic residues, while the positively
charged moieties form electrostatic interactions with the
dissociated carboxylates [37,38].
Role of carboxylates in proton transport
Carboxylates are also known to play an essential role in
proton translocation by many secondary-active transporters,
including drug transporters such as EmrE and the tetracycline transporter TetA(B). EmrE contains only one membrane-embedded carboxylate (Glu14 ). An array of biochemical data implicates Glu14 in the binding of cationic drugs as
well as protons [39–41]. A mechanism was proposed, where
drugs and protons bind to Glu14 in a mutually exclusive
fashion: (i) a drug interacts with Glu14 in a substrate-binding
site at the inside surface of the membrane, (ii) the liganded
substrate-binding site reorients to the outside surface of
the membrane, and (iii) reprotonation of Glu14 promotes
dissociation of the substrate [42].
Recently, evidence was obtained for a role of carboxylates
in proton transport by ABC transporters. In a homology
model of LmrA based on the Vibrio cholera MsbA structure,
a glutamate residue (Glu314 ) was identified that is present
in a hydrophobic environment at the membrane–cytoplasm
interface of TMS 6 [43]. Interestingly, Glu314 is essential for
proton-coupled drug transport by LmrA [6,43]. Glu14 in
EmrE and Glu314 in LmrA both displayed an elevated pKa
value, and a conserved Glu to Asp mutation changed the
pH dependence of drug transport for both proteins. These
observations suggest that, similar to Glu14 in EmrE, Glu314 in
LmrA may have a mechanistic role in drug transport, e.g. by
coupling conformational changes in the NBD (induced
by ATP-binding/hydrolysis) to conformational changes in
the MD associated with transport. In this context, it is
noteworthy that replacement of Glu1204 at the membrane–
cytoplasm interface of TM17 of MRP1 uncoupled its catalytic
and transport activities [44]. Alternatively, proton transport
by LmrA might result from drug/proton–protein interactions
reminiscent of those described for EmrE. The ability of LmrA
to mediate proton extrusion may also relate to an unidentified
role of this transporter in the physiology of L. lactis.
Conclusion
It is clear that ABC-type and secondary-active multidrug
transporters share many structural and functional features.
These features might reflect a similarity in conformational
changes during transport, which could be based on common
biochemical mechanisms for drug binding and translocation.
The research described in this review was funded by Cancer
Research UK, Association of International Cancer Research (AICR),
Biotechnology and Biological Sciences Research Council (BBSRC),
Medical Research Council (M.R.C.) and Royal Society. S.V. is a
recipient of a Cambridge Nehru Scholarship.
Mechanistic and Functional Studies of Proteins
References
1 Van Bambeke, F., Balzi, E. and Tulkens, P.M. (2000) Biochem. Pharmacol.
60, 457–470
2 Borst, P. and Elferink, R.O. (2002) Annu. Rev. Biochem. 71, 537–592
3 Polgar, O. and Bates, S.E. (2005) Biochem. Soc. Trans. 33, 241–245
4 Davidson, A.L. and Chen, J. (2004) Annu. Rev. Biochem. 73, 241–268
5 Murakami, S. and Yamaguchi, A. (2003) Curr. Opin. Struct. Biol. 13,
433–452
6 Venter, H., Shilling, R.A., Velamakanni, S., Balakrishnan, L. and van Veen,
H.W. (2003) Nature (London) 426, 866–870
7 van Veen, H.W., Callaghan, R., Soceneantu, L., Sardini, A., Konings, W.N.
and Higgins, C.F. (1998) Nature (London) 391, 291–295
8 van Veen, H.W., Margolles, A., Muller, M., Higgins, C.F. and Konings, W.N.
(2000) EMBO J. 19, 2503–2514
9 Balakrishnan, L., Venter, H., Shilling, R.A. and van Veen, H.W. (2004)
J. Biol. Chem. 279, 11273–11280
10 Kim, S.H., Chang, A.B. and Saier, Jr, M.H. (2004) Microbiology 150,
2493–2495
11 Abramson, J., Smirnova, I., Kasho, V., Verner, G., Kaback, H.R. and
Iwata, S. (2003) Science 301, 610–615
12 Chang, G. and Roth, C.B. (2001) Science 293, 1793–1800
13 Chang, G. (2003) J. Mol. Biol. 330, 419–430
14 Huang, Y., Lemieux, M.J., Song, J., Auer, M. and Wang, D.N. (2003)
Science 301, 616–620
15 Locher, K.P., Lee, A.T. and Rees, D.C. (2002) Science 296, 1091–1098
16 Reyes, C.L. and Chang, G. (2005) Science 308, 1028–1031
17 Jin, J., Guffanti, A.A., Beck, C. and Krulwich, T.A. (2001) J. Bacteriol. 183,
2667–2671
18 Bakos, E., Evers, R., Szakacs, G., Tusnady, G.E., Welker, E., Szabo, K.,
de Haas, M., van Deemter, L., Borst, P., Varadi, A. et al. (1998)
J. Biol. Chem. 273, 32167–32175
19 Ma, C. and Chang, G. (2004) Proc. Natl. Acad. Sci. U.S.A. 101, 2852–2857
20 Tate, C.G. (2005) in Structural Biology of Membrane Proteins
(Grisshammer, R. and Buchanan, S., eds.), Royal Society of Chemistry,
Cambridge, U.K., in the press
21 Hirai, T., Heymann, J.A., Shi, D., Sarker, R., Maloney, P.C. and
Subramaniam, S. (2002) Nat. Struct. Biol. 9, 597–600
22 Schumacher, M.A. and Brennan, R.G. (2003) Res. Microbiol. 154, 69–77
23 Vazquez-Laslop, N., Markham, P.N. and Neyfakh, A.A. (1999)
Biochemistry 38, 16925–16931
24 Zheleznova, E.E., Markham, P.N., Neyfakh, A.A. and Brennan, R.G. (1999)
Cell (Cambridge, Mass.) 96, 353–362
25 Grkovic, S., Hardie, K.M., Brown, M.H. and Skurray, R.A. (2003)
Biochemistry 42, 15226–15236
26 Schumacher, M.A., Miller, M.C. and Brennan, R.G. (2004) EMBO J. 23,
2923–2930
27 Zheleznova, E.E., Markham, P., Edgar, R., Bibi, E., Neyfakh, A.A. and
Brennan, R.G. (2000) Trends Biochem. Sci. 25, 39–43
28 Paulsen, I.T., Brown, M.H. and Skurray, R.A. (1996) Microbiol. Rev. 60,
575–608
29 Edgar, R. and Bibi, E. (1999) EMBO J. 18, 822–832
30 Paulsen, I.T., Brown, M.H., Littlejohn, T.G., Mitchell, B.A. and Skurray, R.A.
(1996) Proc. Natl. Acad. Sci. U.S.A. 93, 3630–3635
31 Mazurkiewicz, P., Konings, W.N. and Poelarends, G.J. (2002)
J. Biol. Chem. 277, 26081–26088
32 Miwa, M., Tsukahara, S., Ishikawa, E., Asada, S., Imai, Y. and Sugimoto, Y.
(2003) Int. J. Cancer 107, 757–763
33 Haimeur, A., Conseil, G., Deeley, R.G. and Cole, S.P. (2004)
Mol. Pharmacol. 65, 1375–1385
34 Beene, D.L., Price, K.L., Lester, H.A., Dougherty, D.A. and Lummis, S.C.
(2004) J. Neurosci. 24, 9097–9104
35 Dougherty, D.A. (1996) Science 271, 163–168
36 Pawagi, A.B., Wang, J., Silverman, M., Reithmeier, R.A. and Deber, C.M.
(1994) J. Mol. Biol. 235, 554–564
37 Murakami, S., Nakashima, R., Yamashita, E. and Yamaguchi, A. (2002)
Nature (London) 419, 587–589
38 Yu, E.W., McDermott, G., Zgurskaya, H.I., Nikaido, H. and Koshland, Jr, D.E.
(2003) Science 300, 976–980
39 Muth, T.R. and Schuldiner, S. (2000) EMBO J. 19, 234–240
40 Yerushalmi, H. and Schuldiner, S. (2000) J. Biol. Chem. 275, 5264–5269
41 Yerushalmi, H. and Schuldiner, S. (2000) FEBS Lett. 476, 93–97
42 Yerushalmi, H. and Schuldiner, S. (2000) Biochemistry 39, 14711–14719
43 Shilling, R.A., Federici, F., Walas, F., Venter, H., Velamakanni, S.,
Woebking, B., Balakrishnan, L., Luisi, B. and van Veen, H.W. (2005)
FASEB J., in the press
44 Situ, D., Haimeur, A., Conseil, G., Sparks, K.E., Zhang, D., Deeley, R.G. and
Cole, S.P. (2004) J. Biol. Chem. 279, 38871–38880
Received 28 June 2005
C 2005
Biochemical Society
1011