At a hospital, a breast cancer patient is fighting a battle against cancer because the tenaciously dividing cancer cells in her body are spitting out anticancer drugs back into her blood stream, preventing them from attacking her tumour. In another setting, a pathogenic bacterium in our gut is collecting nutrients required for maintenance and growth of the cell. But within the cocktail trapped in the gut lies a mixture of chemical toxins, some natural (such bile acids which can permeabilize and dissolve the microbial cells membrane) and some man-made (such as antibiotics). Yet as soon as the poisons find their way into the interior of the bug, they are jettisoned out of the cell by the same molecular gadgetry as those present in the tumour cells of the cancer patient. Nature’s common currency is a protein called a ‘multidrug’ transporter. Embedded within a cell’s plasma membrane, this protein protects a cell by ejecting a variety of molecules - in many cases, toxins - on contact. With cancer cancer cells, the toxins are anticancer agents. The cell might also be a bacterium, in which case the toxins are antibiotics.
We live in an era when antibiotic resistance among pathogenic microorganisms has spread at an alarming rate. The World Health Organization states that antibiotic resistance is one the three greatest threats to human health. There are so many interesting questions about multidrug transporters: How do they recognize so many different drugs? Can they be inhibited so that existing antibiotics and anticancer drugs can still be used in treatment of diseases? Can we develop new antibiotics and anticancer agents that bypass recognition by multidrug transporters? In this laboratory, we work on the answers to these questions. Our methods are based on biochemistry, cell biology and biophysics, and include techniques such as DNA cloning, gene expression, molecular modelling, site-directed mutagenesis, protein purification and reconstitution, and drug binding and transport using radiolabelled and fluorescent substrates. We study multidrug transporters that are members of different families of membrane proteins, that originate from various pathogenic microorganisms and human tumour cells, and that are structurally different but have a conserved function. It is an exciting research group to be part of.
The following sections give an overview of some of our projects:
ABC (ATP-binding cassette) transporters comprise one of the largest families of structurally related membrane proteins, and are thought to mediate substrate transport in a reaction driven by ATP-binding and hydrolysis. One of the human transporter that we study is the Breast Cancer Resistance Protein (ABCG2). A variety of techniques is used in this laboratory to investigate the basic mechanisms by which the Breast Cancer Resistance Protein is able to recognize and transport multiple drugs. Our recent work on ABCG2 revealed for the first time the presence of a steroid hormone-binding element at the external face of the transporter (Fig. 1). As this element modulates ABCG2 activity, its further characterization might provide an opportunity for the development of new therapeutic ligands for ABCG2 in a clinical setting.
Certain human ABC transporters are well conserved in bacteria. The ABC multidrug transporter LmrA from Lactococcus lactis is a homologue of the human multidrug resistance P-glycoprotein (ABCB1), another major drug pump involved in drug resistance of human cancer cells. LmrA and ABCB1 have very similar specificities for chemotherapeutic drugs and modulators. LmrA can even functionally substitute for ABCB1 in human lung fibroblast cells. Although ABC transporters are generally thought to be unidirectional extrusion systems, we discovered that LmrA contains a reversible pathway for drugs. Moreover, we demonstrated that if the membrane-embedded, drug-binding part of the transporter is decoupled from the cytosolic part that breaks down ATP to provide metabolic energy (the 'engine' of the transporter) then LmrA acts to import drugs rather than export them. Interestingly, the membrane-embedded part of LmrA was shown to mediate a protonmotive force-dependent transport reaction, which might suggest a functional link between this ABC transporter and ion-coupled multidrug transporters (see below). Our findings open up the possibility to use LmrA-like transporters for drug delivery in cells.
This laboratory studies the molecular properties of the LmrA and ABCB1 homologue Sav1866 from the Gram-positive pathogen Staphylococcus aureus, which is responsible for outbreaks of MRSA (methicilline-resistant staphylococcus aureus) in hospitals and clinics. As the structure of the staphylococcal ABC transporter has been solved by X-ray crystallography (Fig. 2), we can examine detailed structure-function relationships underlying drug binding and translocation using structural biology as a basis.
- Figure 2. Ribbon representation of the crystallized homodimeric ABC transporter Sav1866 from Staphylococcus aureus (side view) showing 2 x 6 transmembrane helices and 2 intracellular nucleotide-binding domains. Out and In refer to the outside and inside of the plasma membrane, respectively. Panel on left-hand side: S. aureus.
One of the interesting issues regarding multidrug transporters is their role in cell physiology. Multidrug transporters might have a purely protective function, but they could also be involved in the transport of substrates (e.g. lipids and lipid soluble metabolites) that share physico-chemical properties with drugs. The ABC transporter MsbA is expressed in the plasma membrane of many Gram-negative bacteria including pathogens such as Salmonella typhimurium, Vibrio cholerae and Pseudomonas aeruginosa. MsbA is thought to interact with lipid A, which is the lipid anchor of lipopolysaccharides in the outer membrane, and which is a potent activator of innate immunity in mammals via Toll-like receptors. In a study on the kinetics of drug and lipid-A binding/transport by MsbA from Escherichia coli, we found that both types of substrates can interact with MsbA with a comparable affinity, and that lipid-A binding can inhibit drug binding. Hence, the lipid environment of a multidrug transporter can be an important factor its drug selectivity. In another study, we investigated the substrate-binding chamber (Fig. 3) that is enclosed by the two membrane domains in the MsbA dimer. Two centrally located residues Ser289 and Ser290 were identified that affect the antibiotic and toxic compound selectivity of this bacterial transporter.
Figure 3. Stereoview of the substrate-binding chamber in the membrane domain of an MsbA monomer. (A) View of membrane domain from the dimer interface showing a hotspot of residues that are implicated in drug binding near the apex of the chamber. Similar observations can be made for the other membrane domain in the dimer. (B) Surface representation of the binding chamber formed by the transmembrane helices in Panel A.
Recently, we used genetic engineering to tweak a critical part of MsbA – a group of four helices embedded in the transporter (referred to as the tetrahelix bundle) (Fig. 4) that acts as a ‘spring-lock’ for the transition from the inward-facing to outward-facing shape. By manipulating the msbA gene, we were able to produce bacteria that made MsbA mutants. We then built the modified proteins into ‘inside-out’ membrane vesicles in which the transporters import the drug from the outside, thus making it easier to follow the transport process. We were able to show that both transport and the associated changes in protein conformation were greatly impaired by the structural alterations in the tetrahelix bundle, while other features, such as the transporter’s ability to bind ATP and substrate, were not significantly affected. Tetrahelix bundles are shared by all ABC exporters in eubacteria and eukaryotes, and are essential for their activity. As their primary amino acid sequences are different, the tetrahelix bundle might provide an attractive drug target for modulation of ABC transporter activity. See Research at Cambridge for further information.
Figure 4. Tetrahelix bundle in the MsbA dimer. (A,B) Proposed conformational changes of the MsbA transporter in the transition from the inward-facing shape (A) to the outward-facing shape (B). Current mechanistic models for substrate transport by the MsbA dimer suggest substrate (Lipid A) binding to the inward-facing conformation followed by a transition to the ATP-bound, outward-facing structure with dimerized nucleotide-binding domains (NBDs), from which Lipid A is released (arrows in red). (C) Zoomed-in snapshot of the tetrahelix bundle in (B) (in red and blue) viewed from the cytoplasmic side showing examples of stabilizing side-chain interactions.
ABC exporters transport substrates by an alternating access mechanism that is driven by ATP binding and hydrolysis. The general mechanism is a motion from an inward to an outward state, with a different intertwining of the half-transporters in both states (Fig. 5). In collaboration with colleagues at Imperial College, London, we recently determined the function and crystal structure of the antimicrobial peptide ABC exporter McjD that exports the antibacterial peptide microcin MccJ25 in Enterobacteriaceae, and ensures self-immunity of the producing strain through efficient export of the toxic mature peptide from the cell. Our findings in detergent solution, proteoliposomes and intact cells demonstrate that McjD mediates MccJ25 transport in an ATP-dependent fashion, and that in the absence of Mccj25, the protein catalyses multidrug transport. The McjD structure represents a novel nucleotide-bound, outward-occluded state. It does not possess subunit intertwining and shows a well-defined binding cavity that is closed to all sides, consistent with it being an intermediate between the inward and outward-facing state. Our structure provides valuable insights in the role of a transition state in the mechanism of an ABC exporter (Fig. 5), and is consistent with our earlier biochemical observations on the existence of this conformation in the transport cycle of MsbA.
Figure 5. Proposed role of the outward-occluded state in the mechanism of ABC exporters. Ligand binding to the inward-open (apo) conformation (state i) facilitates transition of the transporter into the inward-closed (apo) conformation (state ii). ATP binding is associated with a transition into a hypothetical occluded conformation (state iii), and subsequent formation of a nucleotide-bound outward-open conformation (state iv). Upon release of the ligand, the ABC exporter adopts an outward-occluded conformation (state v) (current McjD structure). ATP hydrolysis resets the transporter back to the inward-facing conformation (state i). The TMs 1 and 2 from each monomer are highlighted in dark blue and red. The substrate is shown as a yellow diamond.
Our research group also studies the molecular properties of ion-coupled multidrug transporters, which mediate drug efflux in an antiport reaction that is driven by the transmembrane electrochemical gradient of protons and/or sodium ions. Ion-coupled multidrug transporters can be classified into various families including the Major Facilitator Super (MFS) family, the Resistance-Nodulation-cell Division (RND) family, and the multidrug and toxic compound extrusion (MATE) family. Even though members of these secondary-active transporter families are structurally different from each other and different from ABC exporters, they all show an ability to transport structurally dissimilar drugs. This phenomenon raises interesting questions.
LmrP is a MFS multidrug transporter from L. lactis that represents a common antibiotic efflux pump in bacteria. Previous work has indicated that ethidium, a monovalent cationic substrate, is exported by LmrP by electrogenic antiport with two (or more) protons. This observation raised the question whether these protons are translocated sequentially along the same pathway, or through different routes. To address this question we constructed a 3-D homology model of LmrP based on the high-resolution structure of the glycerol-3P/Pi antiporter GlpT from E. coli. Interestingly, LmrP is predicted to contain an internal cavity formed at the interface between the two halves of the transporter (each containing 6 transmembrane helices). On the surface of this cavity lie two clusters of polar, aromatic and carboxylate residues with potentially important function in proton shuttling and interactions with cationic drugs (Fig. 6). We tested by mutagenesis the possible proton conduction points suggested by this model, and found that there are two proton conduction pathways at play in the drug transport mechanism, of which one might be dedicated to proton transport and the other to drug transport. Proton translocation pathways in LmrP are subject of further research.
- Figure 6. Internal cavity in the proton motive force-dependent MFS member LmrP. (A) The predicted LmrP structure represents an inward facing conformation, exposing an internal cavity to the inner membrane leaflet. Out and In refer to the outside and inside of the plasma membrane, respectively. (B) Residues lining the surface of the internal cavity. Carboxylates (in red) exposed at this surface are organized into two distinct clusters containing additional neutral-polar (in green) and aromatic residues (in blue). The role of these clusters in proton and drug transport by LmrP are being investigated.
RND multidrug transporters are important for antibiotic resistance in Gram-negative pathogenic bacteria, where they interact with a membrane fusion protein and outer membrane porin to enable transport from the periplasm across the outer membrane. Even though X-ray crystallography has provided many insights in the structures of the components of RND transporters, much less information is available about the question how these components work together at the functional level.
In a collaborative work with Prof. Satoshi Murakami at the Tokyo Institute of Technology, we investigated how β-lactams are recognized by the AcrA-AcrB-TolC system in E. coli. β-Lactam antibiotics are mainstream antibiotics that are indicated for the prophylaxis and treatment of bacterial infections. The AcrA-AcrD-TolC multidrug efflux system confers much stronger resistance on E. coli to clinically relevant anionic β-lactam antibiotics (such as carbenicillin and sulbenicillin) than the homologous AcrA-AcrB-TolC system. Using an extensive combination of chimeric analysis and site-directed mutagenesis, we searched for residues that determine the difference in β-lactam specificity between AcrB and AcrD. We identified three crucial residues at the “proximal” (or access) substrate binding pocket. The simultaneous replacement of these residues in AcrB by those in AcrD (Q569R, I626R, and E673G) transferred the β-lactam specificity of AcrD to AcrB (Fig. 7). Our findings indicate for the first time that the difference in β-lactam specificity between AcrB and AcrD relates to interactions of the antibiotic with residues in the proximal binding pocket.
Figure 7. RND efflux pumps are crucial for the intrinsic resistance of Gram-negative bacteria to antibiotics, detergents and toxic ions. Stereo view of the regions in AcrB important for recognition of negatively charged β-lactam antibiotics. Substrate translocation pathway (in yellow) including proximal and distal substrate binding pockets, and direction of substrate transport (dashed arrow) are indicated. Green-colored β-sheet sequences indicate regions that determine the difference in substrate specificity between AcrB and AcrD. The important residues (Gln569, Ile626, and Glu673) for recognition of negatively charged β-lactams are shown in CPK representation. The space colored in yellow is the solvent accessible inner cavity of the AcrB molecule. To put the size of the substrate translocation pathway in AcrB in perspective, a stick model of the β-lactam antibiotic carbenicillin is included at the same scale.
Membrane transporters belonging to MATE family mediate the efflux of unrelated pharmaceuticals from the interior of the cell in organisms ranging from bacteria to human. These proteins are thought to fall into two classes that couple substrate efflux to the influx of either Na+ or H+. Recently, we studied the energetics of drug extrusion by NorM from Vibrio cholerae (Fig. 8) in proteoliposomes in which purified NorM protein was functionally reconstituted in an inside-out orientation. We established that NorM simultaneously couples to the sodium-motive force and proton-motive force, and biochemically identified protein regions and residues that play important roles in Na+ or H+ binding. As the positions of protons are not available in current medium and high-resolution crystal structures of MATE transporters, our findings add a previously unrecognized parameter to mechanistic models based on these structures.
Figure 8. Structure model of outward-facing NorM from Vibrio cholerae. Stereoview (PDB 3MKT) showing TM1–6 in gray and TM7–12 in green. Asp36 (red) and neighboring Asn174 and Asn178 residues (light blue) are shown in a blue circle. Glu255 (purple) and Asp371 (orange) are shown in a purple circle with neighboring aromatic residues in yellow. Top, extracellular side; bottom, intracellular side. Panel on right-hand side: V. cholerae.