Antimicrobial resistance is based on the multifarious strategies that bacteria adopt to face antibiotic therapies, making it a key public health concern of our era. Among these strategies, efflux pumps (EPs) contribute significantly to increase the levels and profiles of resistance by expelling a broad range of unrelated compounds – buying time for the organisms to develop specific resistance. In Gram-negative bacteria, many of these chromosomally encoded transporters form multicomponent ‘pumps’ that span both inner and outer membranes and are driven energetically by a primary or secondary transporter component.

One of the strategies to reinvigorate the efficacy of antimicrobials is by joint administration with EP inhibitors (EPI), which either block the substrate binding and/or hinder any of the transport-dependent steps of the pump. In this review, we provide an overview of multidrug-resistance EPs, their inhibition strategies and the relevant findings from the various computational simulation studies reported to date with respect to deciphering the mechanism of action of inhibitors with the purpose of improving their rational design.

Introduction

Over the last two decades, there has been a dramatic surge in the number of multidrug-resistant bacteria, yet paradoxically the number of pharmaceutical companies developing new antimicrobial drugs has dwindled. These conflicting coincidences have contributed to making antimicrobial resistance one of the world’s most demanding health problems [1,2].

Continuous efforts to develop better antimicrobial agents against resistant microbes have yielded successful milestones in the case of infections associated with Gram-positive organisms [3], whereas the Gram-negative pathogens still remain a major challenge. This is due to their very high intrinsic drug resistance, largely attributed to the permeability barrier imposed by the outer membrane (OM), which is absent in Gram-positive organisms, and to the expression of chromosomally encoded drug efflux pumps (EPs) [4,5]. In particular, promiscuous multidrug EPs are part of the primary survival kit of microorganisms and remove most of the xenobiotics from the cell interior, enabling the organism to acquire resistance to agents through more specific adaptive mechanisms [610].

The current shortage of new antimicrobials in the development pipeline to replace the ineffective ones adds to the urgency to maintain the efficacy of existing drugs. One possible way of reinvigorating the previously effective drugs attenuated by bacterial efflux mechanism is by the combinatorial use of efflux pump inhibitors (EPIs). In the present review, we provide an updated view on the main achievements of computational studies on EPIs that have been reported to date.

Efflux-mediated resistance and multidrug resistance

Among the various mechanisms that bacteria exploit to counter the effectiveness of multiple drugs [11,12], efflux-mediated strategy is the predominant one in multidrug resistance (MDR) [13], working in synergy with low OM permeability in Gram-negative bacteria to keep a tight check on the entry of unwanted toxic compounds. In these bacteria, synergistic activity also involves EPs of different families: drug molecules that have gained access to the periplasmic space can further penetrate the inner membrane (IM) via diffusion, but they can be expelled to the periplasm by single-component pumps (e.g. MdfA of Escherichia coli [14]) and successively out of the cell by multicomponent pumps (e.g. AcrAB-TolC of E. coli and MexAB-OprM of Pseudomonas aeruginosa) [1517]. A detailed description of the different families of EPs can be found in several articles [1821].

We here just recall that the wide distribution and overlapping functions of MDR EPs in bacteria hint at their probable role in physiological functions (including virulence, stress response, bacterial cell communication, colonization, fitness and intracellular survival and transport of endogenous toxic compounds) in addition to mediating intrinsic and acquired MDR [1921].

Inhibition strategies for efflux pumps

Inactivation of EPs may open up a wide arena of possibilities for better adjuvant therapy. This strategy has several advantages [2228]: increase of the intracellular concentration of antimicrobials; reduction in the efflux-mediated intrinsic bacterial resistance; reversal of the acquired resistance associated with EP overexpression; reduction in the appearance of highly resistant mutant strains by making unavailable the development of additional mechanisms of resistance-like target-based mutations; and prevention of the export of endogenous microbial virulence factors, thus inhibiting their invasiveness.

To revive the activity of an efflux-susceptible drug, the following strategies have been investigated so far:

  • Targeting the regulatory network involving activators and repressors of EPs [29].

  • Modifying the existing susceptible antimicrobials to make them devoid of the chemophore recognized by EP (e.g. telithromycin bypasses MefA/E and AcrAB systems [30]). However, resistance against these new compounds was described shortly after their deployment [8].

  • Blocking the IM proteins with a high-affinity competitive substrate (an EPI) to trap the EP in an inactive conformation [27,31]. However, toxicity issues have hampered the clinical applications of these EPIs [27], although new compounds are being developed with minimal toxicity but strong inhibitory effects on AcrB [32].

  • Depleting the proton gradient blocks the motive force necessary for the IM proteins to work (e.g. reserpine inhibits the activity of Gram-positive EPs Bmr and NorA [13]). However, these inhibitors affect the entire energetics of bacterial and also of eukaryotic cells, which make them less attractive for clinical implication [33].

  • Specifically, in the case of resistance-nodulation-division (RND) EPs, inhibition may be achieved by preventing the functional tripartite assembly formation (e.g. designed ankyrin repeat proteins inhibited AcrAB-TolC formation [34]) or by blocking the exit duct (e.g. large cations targeting the negatively charged aspartate-rich entrance of TolC in E. coli [35]).

Among these strategies, inhibition of EPs appears a viable one [28] as a single potent inhibitor capable of competitively binding to a pump, and hindering expulsion of its substrate antimicrobials could in principle also bind and block other MDR pumps overlapping in their substrate profiles [32]. In addition to revitalizing antimicrobials, these EPIs also contribute to antibacterial action by hindering the transport of compounds needed for the normal growth and/or maintenance of the microorganism.

Lomovskaya and Bostian [27] proposed the following criteria to qualify a compound as an ideal clinically significant EPI: it must (i) potentiate the activity of antimicrobials in resistant strains expressing functional drug EP; (ii) not have a significant effect on susceptible strains lacking the specific drug EP and not have pharmacological activity on eukaryotic cells; (iii) not potentiate the activity of antimicrobials that are not effluxed; (iv) increase the level of accumulation and decrease the level of extrusion of substrates of the pump; (v) not permeabilize the OM; and (vi) not affect the proton gradient across the IM.

The rational design of an inhibitor fulfilling these requirements (as for that of more efficient drugs that can escape EPs) is challenging and must rely on a detailed knowledge of the molecular features behind the functioning of the EPs. For this purpose, useful insights come from several studies of the structural and functional aspects of drug transporters at a molecular level [10,13,2225,36].

Computational studies on drug efflux pumps and their inhibitors

Role of molecular modelling

Deepening the structural aspects of MDR pumps from crystallographic structures has been remarkably significant but not sufficient to fruitfully assist structure-based drug design. To address mechanistic knowledge gaps, computational techniques are a great resource as they can pinpoint functional dynamics of biological systems. In particular, molecular docking is a valuable in silico tool to identify drug-binding locations, and MD simulations are close to becoming routine computational tools both for rationalizing existing data as well as for various predictions, for instance, about drug recognition and binding, free energies of interactions, translocation mechanism and structural relations with the surrounding environment, including solvent and lipid bilayer [3740]. Additionally, they are often employed to validate homology models [41]. Moreover, normal mode analysis and functional mode analysis of the protein movements allow the comparison between simulations of the apo and holo structures [42], which could lead to the identification of factors intimately related with the first steps of the efflux mechanism [4345]. Recently, ‘coarse-grained’ simulations, where typically four atoms are combined into one particle [46,47], and biased MD simulations [48] have enabled to sample large conformational changes that would normally be inaccessible by standard MD techniques because of the large free energy barriers between such conformations and the correspondingly unaffordable computational time.

Various computational studies on the EPIs of MDR pumps, mainly RND and ATP-binding cassette (ABC) transporters, have been focused on identification of effective EPIs by high-throughput screening of compound databases and determination of the mechanism of inhibition by inspecting the molecular level inhibitor–pump interactions and the coupled conformational changes occurring in the transporters. In the following, we describe few examples from the relevant literature for several families of MDR EPs.

RND pumps

For a short description of structural features of RND transporters, we used AcrB, the RND transporter of E. coli, as prototype (Figure 1) [23]. There are two major substrate-binding pockets located along the substrate translocation pathway and are named with reference to their proximity to the cell membrane and the entrance. The proximal binding pocket (proximal pocket, PP) is located closer to the periplasmic bulk and is supposedly the preferred binding site for high molecular mass compounds (e.g. erythromycin and rifampicin or the doxorubicin dimer) in the access (A or L [23]) monomer of the transport cycle. The distal binding pocket (distal pocket, DP) is located much deeper within the substrate transport pathway, and is likely the recognition site for low molecular mass compounds (e.g. minocycline and doxorubicin). The DP is wide open in the binding (B or T) monomer of the transport cycle. A characteristic flexible G-loop (switch loop), rich in glycine residues, separates the two pockets and modulates the path from the PP in monomer L to the DP in monomer T. Finally, the exit gate (EG) modulates the entrance in the TolC-docking domain.

Side (A) and top (B) views of AcrB

Figure 1
Side (A) and top (B) views of AcrB

The putative affinity sites, PP and DP, are highlighted as red and blue surfaces respectively, the EG as a brown surface. The G-loop is coloured in green.

Figure 1
Side (A) and top (B) views of AcrB

The putative affinity sites, PP and DP, are highlighted as red and blue surfaces respectively, the EG as a brown surface. The G-loop is coloured in green.

Molecules that inhibit these pumps include both medicinal plant extracts [49,50] and synthetic compounds (see Figure 2 for their chemical structures) [32,5153]. Regarding the former compounds, two successful studies have identified potent EPIs by in silico screening of natural compound databases. In one case, by performing molecular-docking-based screening, Ohene-Agyei et al. [49] identified six molecules, of which plumbagin and nordihydroguaiaretic acid showed promising inhibitory activity. In the second study, Aparna et al. [50] singled out hits, which are non-substrates of AcrB and MexB, by using high-throughput virtual screening of an in-house database of phytochemicals. These hits were then subjected to extra-precision docking against AcrB and MexB proteins, and in vitro efflux inhibitory activity testing that selected lanatoside C and daidzein as promising EPIs effective for use in combination therapy against drug-resistant strains of P. aeruginosa and E. coli.

Chemical structures of natural compounds identified as potential RND pump inhibitors

Figure 2
Chemical structures of natural compounds identified as potential RND pump inhibitors
Figure 2
Chemical structures of natural compounds identified as potential RND pump inhibitors

On the basis of molecular docking of approximately 30 compounds (including substrates and inhibitors) to the DP of AcrB [18], Takatsuka et al. [51] showed the presence of two large binding subsites, constituted respectively by a narrow groove and a wide cave (Figure 3). The former corresponds to the upper part of the DP (analogous to the minocycline-binding site in the AcrB crystal structure 2DRD) whereas the lower part of this pocket becomes wider (resembling a cave) and binding mode of ligands at this site is referred to as cave binding.

(Left) Side view of a protomer of AcrB asymmetric trimer with the proximal portion clipped away to reveal the DP (surface with carbons in orange) and the co-crystallized minocycline (PDB code: 2DRD) shown as green sticks

Figure 3
(Left) Side view of a protomer of AcrB asymmetric trimer with the proximal portion clipped away to reveal the DP (surface with carbons in orange) and the co-crystallized minocycline (PDB code: 2DRD) shown as green sticks

(Inset) Enlargement of the binding pocket, (Right) predicted binding site of inhibitor 1-(1-naphthylmethyl)-piperazine (NMP) (cave binder) and substrate doxorubicin (groove binder). Reproduced from [51].

Figure 3
(Left) Side view of a protomer of AcrB asymmetric trimer with the proximal portion clipped away to reveal the DP (surface with carbons in orange) and the co-crystallized minocycline (PDB code: 2DRD) shown as green sticks

(Inset) Enlargement of the binding pocket, (Right) predicted binding site of inhibitor 1-(1-naphthylmethyl)-piperazine (NMP) (cave binder) and substrate doxorubicin (groove binder). Reproduced from [51].

The first MD simulation on RND transporters interacting with EPIs is due to Vargiu and Nikaido [52], who compared the binding of nine substrates, two inhibitors and two non-substrates to the DP of AcrB in the presence of explicit water. Both the inhibitors phenylalanine-arginine β-naphthylamide (PAβN) and NMP (see Figure 4 for their chemical structures) were found to bind to the lower part of the DP that is rich in phenylalanine residues, lately renamed “hydrophobic trap” [53,54]. During the MD simulations, the two compounds left their initial positions extracted from preliminary docking runs and straddled the G-loop during the MD simulations (Figure 5). This possibly explains the inhibition caused by PAβN and NMP by reducing the flexibility of G-loop, which was indeed shown by mutagenesis experiments to be crucial for the functioning of the pump [23,5457].

Chemical structures of the inhibitors MBX2319, NMP, PAβN and D13-9001

Figure 4
Chemical structures of the inhibitors MBX2319, NMP, PAβN and D13-9001
Figure 4
Chemical structures of the inhibitors MBX2319, NMP, PAβN and D13-9001

Comparison among different binding modes of PAβN and NMP to the distal and proximal (NMP’) binding pockets of AcrB displaying straddling of G-loop by these inhibitors

Figure 5
Comparison among different binding modes of PAβN and NMP to the distal and proximal (NMP’) binding pockets of AcrB displaying straddling of G-loop by these inhibitors

Ligands are shown with spheres coloured according to atom types (with non-polar hydrogens removed). The DP, PP and the PC1/PC2 subdomain cleft are shown with transparent red, green and orange surfaces respectively, whereas the G-loop is shown in grey cartoon. Residues within 3.5 Å (1 Å=0.1 nm) from the ligand are represented as coloured beads (red, green, orange and yellow for those of DP, PP, cleft and G-loop respectively). Blue and grey beads identify residues common to both the pockets and those defining the EG (far away from the ligand) respectively. Reproduced from [23,52].

Figure 5
Comparison among different binding modes of PAβN and NMP to the distal and proximal (NMP’) binding pockets of AcrB displaying straddling of G-loop by these inhibitors

Ligands are shown with spheres coloured according to atom types (with non-polar hydrogens removed). The DP, PP and the PC1/PC2 subdomain cleft are shown with transparent red, green and orange surfaces respectively, whereas the G-loop is shown in grey cartoon. Residues within 3.5 Å (1 Å=0.1 nm) from the ligand are represented as coloured beads (red, green, orange and yellow for those of DP, PP, cleft and G-loop respectively). Blue and grey beads identify residues common to both the pockets and those defining the EG (far away from the ligand) respectively. Reproduced from [23,52].

A recent work by Vargiu et al. [53] identified the underlying molecular mechanism of inhibition of MBX2319 (a novel pyranopyridine EPI potent against RND pumps of the Enterobacteriaceae species, see Figure 3 for the chemical structure) by comparing it with that of D13-9001, PAβN and NMP by molecular docking and MD simulations. They observed that D13-9001 and MBX2319 bound more tightly than the typical substrate minocycline to the DP. By binding to the lower part of the DP, similar to doxorubicin in the F610A variant of AcrB [58,59], MBX2319 interacted with AcrB in a manner analogous to that of the hydrophobic portion of D13-9001 (Figure 6) [54]. Investigation of the minocycline binding to such AcrB-inhibitor complexes supports the picture of a competitive binding of all these inhibitors (except D13-9001). As MBX2319 neither contains any charged groups nor can utilize common specific channels to penetrate across the OM of P. aeruginosa, it does not remarkably inhibit efflux in them [13].

Position of inhibitors D13-9001 (B) and MBX2319 (C) with respect to the hydrophobic trap in T protomer [23] of AcrB, as extracted from MD simulations

Figure 6
Position of inhibitors D13-9001 (B) and MBX2319 (C) with respect to the hydrophobic trap in T protomer [23] of AcrB, as extracted from MD simulations

The channel found in AcrB free of ligands (A) is also shown for reference. Ligands are depicted in thick sticks, protein is shown with the molecular surface coloured in orange, yellow and iceblue at the PC1/PC2 cleft, the G-loop tip and the EG respectively, and white elsewhere. Reproduced from [53].

Figure 6
Position of inhibitors D13-9001 (B) and MBX2319 (C) with respect to the hydrophobic trap in T protomer [23] of AcrB, as extracted from MD simulations

The channel found in AcrB free of ligands (A) is also shown for reference. Ligands are depicted in thick sticks, protein is shown with the molecular surface coloured in orange, yellow and iceblue at the PC1/PC2 cleft, the G-loop tip and the EG respectively, and white elsewhere. Reproduced from [53].

A success story in developing more efficient broad-spectrum EPIs concerns some derivatives of MBX2319, some of which are 30 times more potent than the original inhibitor [60]. By combining cellular, X-ray crystallographic and MD simulations data, the molecular basis for pyranopyridine-based inhibition of AcrB was disclosed [61]. All of the derivatives bind within the hydrophobic trap forming extensive hydrophobic interactions. Moreover, the increasing potency of improved inhibitors correlates with the formation of a delicate protein- and water-mediated hydrogen bond network.

Yilmaz et al. [62] reported another successful development of EPIs based on computer modelling. They identified and characterized the binding site of 2-substituted benzothiazoles (see Figure 7 for the chemical structure) as potential EPIs with the ability to restore the antibacterial activity of ciprofloxacin. Among the compounds experimentally tested, BSN-004, BSN-006 and BSN-023 topped the list with clinically significant EPI activity and were found to bind similar to the co-crystallized AcrB substrates minocycline and doxorubicin in the DP. Also, the higher calculated binding energies of BSN-006 and BSN-023 compared with that of ciprofloxacin (approximately −18.0 and −12.7 kcal/mol to be compared with −10.2 kcal/mol) indicated their possibile roles as competitive inhibitors in contrast with BSN-004, which might act as an uncompetitive inhibitor by steric hindrance.

Chemical structures of BSN-coded 2-substituted benzothiazoles

Figure 7
Chemical structures of BSN-coded 2-substituted benzothiazoles
Figure 7
Chemical structures of BSN-coded 2-substituted benzothiazoles

Recently, Nikaido and co-workers for the first time determined quantitatively the efflux kinetics of PAβN and of its homologues alanine, arginine and phenylalanine β-naphthylamides [63]. Experimental and computational data support the hypothesis that PAβN inhibits the efflux of AcrB substrates by both binding to the hydrophobic trap and interfering with the binding of other drug substrates to the upper part of the binding pocket.

ABC transporters

The mammalian P-glycoprotein of the ABC transporters has been a member of prime interest offering valuable insights on the drug recognition and transport [64], which can be translated to their bacterial counterparts MsbA and LmrA [65,66]. We, therefore, have included studies related to P-glycoprotein inhibitors that have increased our understanding of various structural and dynamic aspects underlying the multidrug binding property of this transporter.

Vandevuer et al. [67] published the first computational study of EPIs on ABC transporters where they performed molecular docking of several first- and second-generation P-glycoprotein inhibitors (dexniguldipine, quinidine, quinine, S9788, tamoxifen and verapamil, whose chemical structures are shown in Figure 8) and found different binding positions for the docked ligands corroborating the experimental evidence of multiple drug-binding sites in this transporter. They also identified several key interactions like H-bonds, π–π and cation–π made by the inhibitor in the docked complexes.

Chemical structures of the first- and second-generation P-glycoprotein inhibitors (dexniguldipine, quinine, quinidine, S9788, verapamil and tamoxifen)

Figure 8
Chemical structures of the first- and second-generation P-glycoprotein inhibitors (dexniguldipine, quinine, quinidine, S9788, verapamil and tamoxifen)
Figure 8
Chemical structures of the first- and second-generation P-glycoprotein inhibitors (dexniguldipine, quinine, quinidine, S9788, verapamil and tamoxifen)

Successively, many studies were aimed at identifying possible inhibitor-specific features in terms of their binding locations and types of interactions that could be exploited in inhibitor design. One such study was reported by Ferreira et al. [42], where using docking and MD simulations, they identified several residues (Met68, Phe332, Leu335 and Tyr946) and at least two regions (Figure 9 [Inset 1]) of the P-glycoprotein exclusively interacting with modulators (latilagascene E, QZ59–SSS, tariquidar and verapamil, depicted in Figure 10). The first region was located at the beginning of transmembrane domain (TMD) 6 whereas the other included residues from TMD7, TMD10 and TMD11. Interestingly, though tariquidar showed competitive binding with substrates vinblastine and colchicine, as deducible from the overlap among the networks of interactions established by the three compounds, the fact that the additional interactions with Met874, Leu875 and Phe934 in TMD10/11 are not observed for any of the substrates and could correlate with the increased modulatory effect of tariquidar. Indeed, tariquidar is nearly 67 times more active than verapamil [42].

Binding sites identified for various inhibitors within the TMD and NBD regions of ABC transporters

Figure 9
Binding sites identified for various inhibitors within the TMD and NBD regions of ABC transporters

The structure of P-glycoprotein (PDB code: 3G60) is on the left side (cartoon representation) and the TMD and NDB domains are coloured white and iceblue respectively. Surface representation is used to highlight the overall positions of binding sites (cyan and yellow surfaces respectively, for the TMD- and NBD-binding sites). In the insets on the right side, the magnified residue level details of the binding of a few inhibitors at that site reported from different studies are shown: (Inset 1) Binding site interactions of QZ59–RRR (A) and QZ59–SSS molecules each in the lower (red) and upper (blue) sites (B) in the P-glycoprotein internal cavity as seen in the co-crystallized structures (PDB codes: 3G60 and 3G61 [69]). The inhibitors are shown with CPK representation and coloured according to the atom type (C, N, O and S atoms are coloured white, blue, red and yellow respectively), whereas sticks depict the side chains of residues within 4 Å of the ligand. (Inset 2) (A). The docked structure of the low energy conformation of inhibitor XR9576 (C, N, O and H atoms are coloured cyan, blue, red and white respectively) superimposed on substrate rhodamine 123 (yellow sticks) and another inhibitor GP240 (pink sticks). Inhibitors GP240 (B) and XR9576 (C) are stabilized by formation of H-bond with specific residues of P-glycoprotein. Obtained with permission from Elsevier [70]. (Inset 3) (A, B and C) Different binding modes of the inhibitor QZ59–RRR (black sticks) to the TMD drug-binding pocket of P-glycoprotein as obtained by docking the compound on three different conformations of the protein extracted from MD simulations [68]. Each subfigure (A, B and C) shows all the three conformations of the binding site residues. The thick cyan sticks correspond to the conformation used in that specific docking run whereas the thin lines correspond to two additional conformations. The crystal structure of mouse P-glycoprotein with QZ59–RRR bound is shown in (D). Reproduced from [68].

Figure 9
Binding sites identified for various inhibitors within the TMD and NBD regions of ABC transporters

The structure of P-glycoprotein (PDB code: 3G60) is on the left side (cartoon representation) and the TMD and NDB domains are coloured white and iceblue respectively. Surface representation is used to highlight the overall positions of binding sites (cyan and yellow surfaces respectively, for the TMD- and NBD-binding sites). In the insets on the right side, the magnified residue level details of the binding of a few inhibitors at that site reported from different studies are shown: (Inset 1) Binding site interactions of QZ59–RRR (A) and QZ59–SSS molecules each in the lower (red) and upper (blue) sites (B) in the P-glycoprotein internal cavity as seen in the co-crystallized structures (PDB codes: 3G60 and 3G61 [69]). The inhibitors are shown with CPK representation and coloured according to the atom type (C, N, O and S atoms are coloured white, blue, red and yellow respectively), whereas sticks depict the side chains of residues within 4 Å of the ligand. (Inset 2) (A). The docked structure of the low energy conformation of inhibitor XR9576 (C, N, O and H atoms are coloured cyan, blue, red and white respectively) superimposed on substrate rhodamine 123 (yellow sticks) and another inhibitor GP240 (pink sticks). Inhibitors GP240 (B) and XR9576 (C) are stabilized by formation of H-bond with specific residues of P-glycoprotein. Obtained with permission from Elsevier [70]. (Inset 3) (A, B and C) Different binding modes of the inhibitor QZ59–RRR (black sticks) to the TMD drug-binding pocket of P-glycoprotein as obtained by docking the compound on three different conformations of the protein extracted from MD simulations [68]. Each subfigure (A, B and C) shows all the three conformations of the binding site residues. The thick cyan sticks correspond to the conformation used in that specific docking run whereas the thin lines correspond to two additional conformations. The crystal structure of mouse P-glycoprotein with QZ59–RRR bound is shown in (D). Reproduced from [68].

Modulators (latilagascene E, tariquidar, QZ59-SSS and QZ59–RRR) of P-glycoprotein reported in the studies of Ferreira et al. [42] and Wise [68]

Figure 10
Modulators (latilagascene E, tariquidar, QZ59-SSS and QZ59–RRR) of P-glycoprotein reported in the studies of Ferreira et al. [42] and Wise [68]
Figure 10
Modulators (latilagascene E, tariquidar, QZ59-SSS and QZ59–RRR) of P-glycoprotein reported in the studies of Ferreira et al. [42] and Wise [68]

In another study, Wise [68] performed ensemble-docking of 26 catalytically relevant non-redundant structures of P-glycoprotein against 21 known substrates or inhibitors, and advocated that no specific “inhibitor-binding site” is located within the drug-binding domain and a competitive inhibition is seen in this transporter. The author also found that in the fully opened outward conformation (ATP hydrolysis transition state), drug docking in the extracellular half of the TMDs seemed to be destabilized as transport ligand EGs opened to the extracellular space. Whereas, in conformations close to fully opened inward state, the side chain of Phe978 (top of the binding site) moved out of the way favouring access of QZ59–RRR analogue to binding pocket (Figure 9 [Inset 3]) [68,69]. This supports the postulation of putative aromatic gating structures in the drug-binding sites of P-glycoprotein.

Likewise, Jara et al. [70] also reported the presence of a common binding site for rhodamine 123 and modulators (derivatives of propafenone and tariquidar) in the mouse P-glycoprotein (Figure 9 [Inset 2A]), suggesting a competitive inhibition in which the higher binding affinity of the modulators (between −11.9 and −14.2 kcal/mol compared with −9.2 kcal/mol of rhodamine) is contributed largely by the aromatic residues in the region between TM 4, 5 and 6. The inhibitor binding to this site could reduce the mobility of TMs (especially TM6), which in turn could affect the subsequent ATP hydrolysis.

The role of flexibility in poly-specific drug binding in P-glycoprotein was addressed by Liu et al. [71] with comparative MD simulations of inward-facing structure with/without inhibitor ligands (QZ59–RRR or QZ59–SSS) in explicit lipid and water environment. Among the transmembrane segments forming the binding pocket in P-glycoprotein, TM4 and TM5 are rigid and stabilize the whole structure, whereas TM6 and TM12 are highly flexible with side chains of aromatic residues exploring different orientations critical for poly-specificity of the drug-binding cavity. The big cavity and flexible nature of binding pocket allow the inhibitor to interact with different residues of the transporter instead of having it locked.

In order to have a comprehensive understanding of EPI action of desmosdumotin, an anti-cancer agent, Gadhe et al. [72] explored its inhibition mechanism against P-glycoprotein (NBD2) by performing molecular docking and MD simulations [72]. They found van der Waals and electrostatic interactions to be playing a predominant role in stabilizing desmosdumotin binding to NBD2 especially with a π–π stacking interaction between the B-ring of desmosdumotin and side chain of Tyr1044.

Recently, Ma et al. [73] systematically characterized and compared substrate (daunorubicin) and inhibitor (QZ59–RRR and QZ59–SSS) effects on NBD and TMD conformational dynamics using apo murine P-glycoprotein [73]. They simulated the apo form, the co-crystals with inhibitor QZ59 bound [69] and docking-generated complexes with daunorubicin bound to each of the two sites where QZ59 was found. The presence of an inhibitor inside the drug-binding pocket kept NBD site 2 open (maintaining crystallographic distances) with ATP–protein interaction energies significantly higher than the ones reported for substrates. This suggests that inhibitors function by keeping the NBDs apart, thus preventing ATP hydrolysis. Moreover, the inhibitor QZ59–RRR exhibited higher affinities compared with that of the substrate owing to much more favourable van der Waals interactions. On the contrary, the apo protein and ligand-bound complexes (with daunorubicin bound at the upper site) (Figure 11) triggered a closure event with an asymmetrical association of the NBDs leading to the formation of nucleotide-binding sites competent to bind ATP with one of the two putative nucleotide-binding sites further dissociated than the other.

(A) The representative structure of ABC transporter Sav1866 with the subunits coloured yellow and turquoise and highlighting the important domains (reproduced with permission from Macmillan Publishers Ltd: Nature [73,77])

Figure 11
(A) The representative structure of ABC transporter Sav1866 with the subunits coloured yellow and turquoise and highlighting the important domains (reproduced with permission from Macmillan Publishers Ltd: Nature [73,77])

(B) Schematic illustration of the NBDs 1 and 2. Here, the N-terminal Walker A motif and the C-terminal Signature sequence form “Site 1” whereas the C-terminal Walker A motif and the N-terminal Signature sequence form “Site 2” ([reproduced from 73,77]).

Figure 11
(A) The representative structure of ABC transporter Sav1866 with the subunits coloured yellow and turquoise and highlighting the important domains (reproduced with permission from Macmillan Publishers Ltd: Nature [73,77])

(B) Schematic illustration of the NBDs 1 and 2. Here, the N-terminal Walker A motif and the C-terminal Signature sequence form “Site 1” whereas the C-terminal Walker A motif and the N-terminal Signature sequence form “Site 2” ([reproduced from 73,77]).

In another study, Prajapati et al. [74] modelled P-glycoprotein in three different catalytic states (inward open (IO) (NBDs far apart), intermediate open (IIO) and outward open (OO) (NBDs in close proximity)), and studied totally 17 systems including eight substrates, eight inhibitors and one without ligand by multi-targeted MD [74]. Similar to the studies discussed above, no distinct sites for substrate and inhibitor were identified, yet significant differences in their binding interactions and stability were observed during the simulation from IO to OO state. The loss of stable binding interactions destabilized the substrate binding in the active site and dislodged it during the IO to OO transformation. On the other hand, the inhibitors maintained stable interactions with drug-binding hydrophobic residues, possibly inhibiting the conformational change in this transporter (Figure 12).

Variations in molecular interactions of verapamil (inhibitor) detected during multi-targeted MD simulation; (A), (B), (C) and (D) show the P-glycoprotein transition states: initial IO, at starting of IIO, after IIO and OO respectively

Figure 12
Variations in molecular interactions of verapamil (inhibitor) detected during multi-targeted MD simulation; (A), (B), (C) and (D) show the P-glycoprotein transition states: initial IO, at starting of IIO, after IIO and OO respectively

The magnified images of corresponding encircled regions are shown as I, II, III and IV respectively. Obtained with permission from Elsevier [74].

Figure 12
Variations in molecular interactions of verapamil (inhibitor) detected during multi-targeted MD simulation; (A), (B), (C) and (D) show the P-glycoprotein transition states: initial IO, at starting of IIO, after IIO and OO respectively

The magnified images of corresponding encircled regions are shown as I, II, III and IV respectively. Obtained with permission from Elsevier [74].

In addition to studies on understanding the mechanism of action of existing ABC pump inhibitors, attempts have been made to potentiate their activity and to design new ones. For example, Tardia et al. [75] reported a new series of 21 polymethoxy benzamides with P-glycoprotein inhibitory activity reaching submicromolar IC50 level due to their modulated lipophilicity by establishment of an intramolecular hydrogen bond (IMHB). MD simulations and density functional theory calculations evidenced the presence of a unique conformation of the hit 4b (Figure 13), which was characterized by a very stable IMHB. They also identified the strength of such IMHB interaction as a sensitive parameter for soft modulation of the P-glycoprotein response as evident from 2,4,5-trimethoxybenzamide derivatives 3b, 4b and 5b, which displayed the highest activity and also the strongest IMHB.

2D structural representation of the regioisomers 4a and 4b pointing out the location of the IMHB

Figure 13
2D structural representation of the regioisomers 4a and 4b pointing out the location of the IMHB

Obtained with permision from the American Chemical Society [75].

Figure 13
2D structural representation of the regioisomers 4a and 4b pointing out the location of the IMHB

Obtained with permision from the American Chemical Society [75].

Singh et al. [76] designed peptides as inhibitors of the transporter P-glycoprotein in Leishmania, responsible for the extrusion of miltefosine, a drug to treat leishmaniasis. Molecular docking of the peptides confirmed the high affinity of inhibitor-9 (I9), −8.3 kcal/mol, (sequence ‘QFIYYSAYALCFWY’, Figure 14A) interacting with Asp1029, Ala1022 and His55 of the transporter (Figure 14B). The present study provided insights into the possibility of targeting P4-ATPase (important for the import of alkylphospholipid drugs into the parasite) and ABC transporters.

(A) Designed peptide inhibitors of the ABC transporters coloured in rainbow from the N-terminus (blue) to the C-terminus (red)

Figure 14
(A) Designed peptide inhibitors of the ABC transporters coloured in rainbow from the N-terminus (blue) to the C-terminus (red)

Their corresponding amino acid sequence is also shown. (B) Affinity region of ABC transporter for peptide inhibitor I9. Reproduced from [76] with permission of the Royal Society of Chemistry.

Figure 14
(A) Designed peptide inhibitors of the ABC transporters coloured in rainbow from the N-terminus (blue) to the C-terminus (red)

Their corresponding amino acid sequence is also shown. (B) Affinity region of ABC transporter for peptide inhibitor I9. Reproduced from [76] with permission of the Royal Society of Chemistry.

The results from the various computational studies on inhibitors of ABC pumps reported above reflect the need for copious non-bonded interactions to be formed by an inhibitor molecule to compete and establish itself strongly in the binding site of the pump, thenceforth hampering the requisite conformational changes for substrate transport.

MATE transporters

MD simulations were also employed in the study of EPIs for the multi-antimicrobial extrusion (MATE) transporter NorA [77]. The three isomeric hybrid compounds, SS14, SS14-M and SS14-P, contain berberine, an antibacterial alkaloid known to be a substrate of NorA, fused at different positions of INF55 (5-nitro-2-phenylindole), an inhibitor of NorA. The authors found that subtle repositioning of the pump-blocking INF55 moiety in berberine–INF55 hybrids minimally affects their antibacterial activitiy but has a significant effect on their inhibitory action. Based on the experimental results, they reported all the three hybrids to have a very similar activity against S. aureus and Caenorhabditis elegans though SS14 showed a slightly higher potency than its isomers against the wild-type and NorA-knockout strains. Also, the SS14 hybrid showed only a minor inhibitory effect on MDR pumps when compared with that of SS14-M and SS14-P. According to MD simulations, the hybrid SS14 prefers a more compact globular conformation with the INF55 moiety folded back over the berberine unit, whereas in SS14-M and SS14-P the INF55 moiety extends away from berberine (Figure 15). The unique conformation for SS14 identified here may explain why it shows different bacterial cell uptake kinetics and reduced inhibitory effects on MDR pumps relative to those of SS14-M and SS14-P.

Chemical structures of berberine and INF55 moieties as well as the isomeric hybrid compounds (SS14, SS14-M and SS14-P)

Figure 15
Chemical structures of berberine and INF55 moieties as well as the isomeric hybrid compounds (SS14, SS14-M and SS14-P)
Figure 15
Chemical structures of berberine and INF55 moieties as well as the isomeric hybrid compounds (SS14, SS14-M and SS14-P)

Concluding remarks

MDR is an unavoidable natural phenomenon and needs to be effectively countered with highest priority to prevent the advent of a post-antibiotic era with untreatable life-threatening infections. Efflux transporters like those of the MFS members in Gram-positive bacteria and RND members in Gram-negative bacteria are the primary saviours in these clinically important pathogens. These transporters, if inhibited, can hinder the normal physiology as well as the MDR, eventually reviving the era of antibiotic treatable infections. The recent reports on computational studies significantly contributing towards the development of several EPIs of such transporter systems provide a positive ray of hope towards development of better EPIs and novel antimicrobial agents that can bypass efflux. It would definitely be important to improve these molecules to widen their spectrum, even if attainment of a universal prokaryotic EPI might not be pragmatic. In addition to focusing solely on the competitive inhibitors of the MDR pumps, scientists are now considering inhibition of transcription of the genes coding for EPs or inhibition of other members of tripartite complexes as possible alternatives.

Summary

  • Promiscuous multidrug EPs are ubiquitous and are considered part of the primary survival kit of microorganisms.

  • Structural plasticity and flexibility (with multisite binding) are the identified basis of multidrug recognition and transport in most of the MDR efflux pumps (e.g. ABC, MFS and RND members).

  • Computational studies have been crucial towards addressing mechanistic knowledge gaps in structural and functional facets of drug transporters, thus laying a cornerstone for rational design of EPIs.

  • Effective strategies of EPIs: competitive binding with higher binding affinities and subsequently blocking and/or reducing flexibility of the binding sites to interfere with the binding of other drug substrates.

  • EPIs are characterized by copious non-bonded interactions (like π–π, cation–π and H-bonds).

  • Drug design attempts are in progress to reduce toxicity and increase inhibitory action of EPIs to achieve a wide spectrum, if not universal, EPI for clinical applications.

Funding

The research leading to the results discussed here was partly conducted as part of the Translocation Consortium (http://www.translocation.eu) and has received support form the Innovative Medicines Initiative Joint Undertaking under Grant Agreement no. 115525, resources that are composed of financial contribution from the European Union’s Seventh Framework Programme (FP7/2007-2013) and EFPIA companies’ in kind contribution; VKR is a Marie Skłodowska-Curie fellow within the “Translocation” Network, project no. 607694.

Abbreviations

     
  • ABC

    ATP-binding cassette

  •  
  • DP

    distal pocket

  •  
  • EG

    exit gate

  •  
  • EP

    efflux pump

  •  
  • EPI

    efflux pump inhibitor

  •  
  • IM

    inner membrane

  •  
  • IMHB

    intramolecular hydrogen bond

  •  
  • MATE

    multi-antimicrobial extrusion

  •  
  • MD

    molecular dynamics

  •  
  • MDR

    multidrug resistance

  •  
  • MFS

    major facilitator superfamily

  •  
  • MFP

    membrane fusion protein

  •  
  • NBD

    nucleotide-binding domain

  •  
  • NMP

    1-(1-naphthylmethyl)-piperazine

  •  
  • OM

    outer membrane

  •  
  • PAβN

    phenylalanine-arginine β-naphthylamide

  •  
  • PP

    proximal pocket

  •  
  • RND

    resistance-nodulation-division

  •  
  • SMR

    small multidrug resistance

  •  
  • TM

    transmembrane

  •  
  • TMD

    transmembrane domain

References

References
1
WHO
. (
2014
)
Antimicrobial resistance: global report on surveillance
,
Geneva, Switzerland
2
Howell
L.
(
2013
)
Global Risks 2013: an initiative of the risk response network
, Eighth edn,
Geneva, Switzerland
3
Butler
M.S.
and
Cooper
M.A.
(
2011
)
Antibiotics in the clinical pipeline in 2011
.
J. Antibiot. (Tokyo)
64
,
413
425
4
Nikaido
H.
(
1994
)
Prevention of drug access to bacterial targets: permeability barriers and active efflux
.
Science
264
,
382
388
5
Ceccarelli
M.
and
Ruggerone
P.
(
2008
)
Physical insights into permeation of and resistance to antibiotics in bacteria
.
Curr. Drug Targets
9
,
779
788
6
Poole
K.
(
2005
)
Efflux-mediated antimicrobial resistance
.
J. Antimicrob. Chemother.
56
,
20
51
7
Piddock
L.J.
(
2006
)
Clinically relevant chromosomally encoded multidrug resistance efflux pumps in bacteria
.
Clin. Microbiol. Rev.
19
,
382
402
8
Nikaido
H.
and
Pagès
J.-M.
(
2012
)
Broad-specificity efflux pumps and their role in multidrug resistance of Gram-negative bacteria
.
FEMS Microbiol. Rev.
36
,
340
363
9
Blair
J.M.
,
Richmond
G.E.
and
Piddock
L.J.
(
2014
)
Multidrug efflux pumps in Gram-negative bacteria and their role in antibiotic resistance
.
Future Microbiol.
9
,
1165
1177
10
Blair
J.M.
,
Webber
M.A.
,
Baylay
A.J.
,
Ogbolu
D.O.
and
Piddock
L.J.
(
2015
)
Molecular mechanisms of antibiotic resistance
.
Nat. Rev. Microbiol.
13
,
42
51
11
George
A.M.
(
1996
)
Multidrug resistance in enteric and other Gram-negative bacteria
.
FEMS Microbiol. Lett.
139
,
1
10
12
Pagès
J.-M.
and
Amaral
L.
(
2009
)
Mechanisms of drug efflux and strategies to combat them: challenging the efflux pump of Gram-negative bacteria
.
Biochim. Biophy. Acta
1794
,
826
833
13
Li
X.-Z.
,
Plésiat
P.
and
Nikaido
H.
(
2015
)
The challenge of efflux-mediated antibiotic resistance in Gram-negative bacteria
.
Clin. Microbiol. Rev.
28
,
337
418
14
Fluman
N.
,
Adler
J.
,
Rotenberg
S.A.
,
Brown
M.H.
and
Bibi
E.
(
2014
)
Export of a single drug molecule in two transport cycles by a multidrug efflux pump
.
Nat. Comm.
5
,
1
9
15
Lee
A.
,
Mao
W.
,
Warren
M.S.
,
Mistry
A.
,
Hoshino
K.
,
Okumura
R.
et al
(
2000
)
Interplay between efflux pumps may provide either additive or multiplicative effects on drug resistance
.
J. Bacteriol.
182
,
3142
3150
16
Tal
N.
and
Schuldiner
S.
(
2009
)
A coordinated network of transporters with overlapping specificities provides a robust survival strategy
.
Proc. Natl. Acad. Sci. U.S.A.
106
,
9051
9056
17
Lewinson
O.
,
Adler
J.
,
Sigal
N.
and
Bibi
E.
(
2006
)
Promiscuity in multidrug recognition and transport: the bacterial MFS Mdr transporters
.
Mol. Microbiol.
61
,
277
284
18
Higgins
C.F.
(
2007
)
Multiple molecular mechanisms for multidrug resistance transporters
.
Nature
,
446
,
749
757
19
Piddock
L.J.
(
2006
)
Multidrug-resistance efflux pumps – not just for resistance
.
Nat. Rev. Microbiol.
4
,
629
636
20
Wong
K.
,
Ma
J.
,
Rothnie
A.
,
Biggin
P.C.
and
Kerr
I.D.
(
2014
)
Towards understanding promiscuity in multidrug efflux pumps
.
Trends Biochem. Sci.
39
,
8
16
21
Matsumura
K.
,
Furukawa
S.
,
Ogihara
H.
and
Morinaga
Y.
(
2011
)
Roles of multidrug efflux pumps on the biofilm formation of Escherichia coli K-12
.
Biocontrol Sci.
16
,
69
72
22
Ramaswamy
V.K.
,
Cacciotto
P.
,
Malloci
G.
,
Ruggerone
P.
and
Vargiu
A.V.
(
2016
)
Multidrug efflux pumps and their inhibitors characterized by computational modeling
.
In
Efflux-Mediated Drug Resistance in Bacteria: Mechanisms, Regulation and Clinical Implications
(
Li
Xian-Zhi
Elkins
Christopher A.
Zgurskaya
Helen I.
, eds), pp.
797
831
,
Springer Verlag
,
Heidelberg, New York
23
Ruggerone
P.
,
Murakami
S.
,
Pos
K.M.
and
Vargiu
A.V.
(
2013
)
RND efflux pumps: structural information translated into function and inhibition mechanisms
.
Curr. Top. Med. Chem.
13
,
3079
3100
24
Venter
H.
,
Mowla
R.
,
Ohene-Agyei
T.
and
Ma
S.
(
2015
)
RND-type drug efflux pumps from Gram-negative bacteria: molecular mechanism and inhibition
.
Front. Microbiol.
6
,
377
25
Du
D.
,
van Veen
H.W.
,
Murakami
S.
,
Pos
K.M.
and
Luisi
B.F.
(
2015
)
Structure, mechanism and cooperation of bacterial multidrug transporters
.
Curr. Opin. Struct. Biol.
33
,
76
91
26
Dreier
J.
and
Ruggerone
P.
(
2015
)
Interaction of antibacterial compounds with RND efflux pumps in Pseudomonas aeruginosa
.
Front. Microbiol.
6
,
660
27
Lomovskaya
O.
and
Bostian
K.A.
(
2006
)
Practical applications and feasibility of efflux pump inhibitors in the clinic – a vision for applied use
.
Biochem. Pharmacol.
71
,
910
918
28
Zechini
B.
and
Versace
I.
(
2009
)
Inhibitors of multidrug resistant efflux systems in bacteria
.
Recent Pat. Antiinfect. Drug Discov.
4
,
37
50
29
Grkovic
S.
,
Brown
M.H.
and
Skurray
R.A.
(
2002
)
Regulation of bacterial drug export systems
.
Microbiol. Mol. Biol. Rev.
66
,
671
701
30
Chollet
R.
,
Chevalier
J.
,
Bryskier
A.
and
Pages
J.M.
(
2004
)
The AcrAB-TolC pump is involved in macrolide resistance but not in telithromycin efflux in Enterobacter aerogenes and Escherichia coli
.
Antimicrob. Agents Chemother.
48
,
3621
3624
31
Mahamoud
A.
,
Chevalier
J.
,
Alibert-Franco
S.
,
Kern
W.V.
and
Pages
J.-M.
(
2007
)
Antibiotic efflux pumps in Gram-negative bacteria: the inhibitor response strategy
.
J. Antimicrob. Chemother.
59
,
1223
1229
32
Opperman
T.J.
and
Nguyen
S.T.
(
2015
)
Recent advances toward a molecular mechanism of efflux pump inhibition
.
Front. Microbiol.
6
,
421
33
Kourtesi
C.
,
Ball
A.R.
,
Huang
Y.Y.
,
Jachak
S.M.
,
Vera
D.M.
,
Khondkar
P.
et al
(
2013
)
Microbial efflux systems and inhibitors: approaches to drug discovery and the challenge of clinical implementation
.
Open Microbiol. J.
7
,
34
52
34
Tikhonova
E.B.
,
Yamada
Y.
and
Zgurskaya
H.I.
(
2011
)
Sequential mechanism of assembly of multidrug efflux pump AcrAB-TolC
.
Chem. Biol.
18
,
454
463
35
Andersen
C.
,
Koronakis
E.
,
Hughes
C.
and
Koronakis
V.
(
2002
)
An aspartate ring at the TolC tunnel entrance determines ion selectivity and presents a target for blocking by large cations
.
Mol. Microbiol.
44
,
1131
1139
36
Yamaguchi
A.
,
Nakashima
R.
and
Sakurai
K.
(
2015
)
Structural basis of RND-type multidrug exporters
.
Front. Microbiol.
6
,
327
37
Galeazzi
R.
(
2009
)
Molecular dynamics as a tool in rational drug design: current status and some major applications
.
Curr. Comput. Aid. Drug Des.
5
,
225
240
38
Zhao
H.
and
Caflisch
A.
(
2015
)
Molecular dynamics in drug design
.
Eur. J. Med. Chem.
91
,
4
14
39
Mortier
J.
,
Rakers
C.
,
Bermudez
M.
,
Murguetiu
M.S.
,
Riniker
S.
and
Wolber
G.
(
2015
)
The impact of molecular dynamics on drug design: applications for the characterization of ligand–macromolecule complexes
.
Drug Disc. Today
20
,
686
702
40
Biarnes
X.
,
Bongarzone
S.
,
Vargiu
A.V.
,
Carloni
P.
and
Ruggerone
P.
(
2011
)
Molecular motions in drug design: the coming age of the metadynamics method
.
J. Comput. Aid. Mol. Des.
25
,
395
402
41
Nurisso
A.
,
Daina
A.
and
Walker
R.C.
(
2012
)
A practical introduction to molecular dynamics simulations: applications to homology modeling
.
Methods Mol. Biol.
857
,
137
173
42
Ferreira
R.J.
,
Ferreira
M.-J.U.
and
dos Santos
D.J.
(
2012
)
Insights on P-glycoprotein’s efflux mechanism obtained by molecular dynamics simulations
.
J. Chem. Theory Comput.
8
,
1853
1864
43
Ruggerone
P.
,
Vargiu
A.V.
,
Collu
F.
,
Fischer
N.
and
Kandt
C.
(
2013
)
Molecular dynamics computer simulations of multidrug RND efflux pumps
.
Comput. Struct. Biotechnol. J.
5
,
e201302008
44
Collu
F.
,
Vargiu
A.V.
,
Dreier
J.
,
Cascella
M.
and
Ruggerone
P.
(
2012
)
Recognition of imipenem and meropenem by the RND-transporter MexB studied by computer fimulations
.
J. Am. Chem. Soc.
134
,
19146
19158
45
Collu
F.
and
Cascella
M.
(
2013
)
Multidrug resistance and efflux pumps: insights from molecular dynamics simulations
.
Curr. Top. Med. Chem.
13
,
3165
3183
46
Takada
S.
(
2012
)
Coarse-grained molecular simulations of large biomolecules
.
Curr. Opin. Struct. Biol.
22
,
130
137
47
Parkin
J.
,
Chavent
M.
and
Khalid
S.
(
2015
)
Molecular simulations of Gram-negative bacterial membranes: a vignette of some recent successes
.
Biophys. J.
109
,
461
468
48
Bernardi
R.C.
,
Melo
M.C.R.
and
Schulten
K.
(
2015
)
Enhanced sampling techniques in molecular dynamics simulations of biological systems
.
Biochim. Biophys. Acta
1850
,
872
877
49
Ohene-Agyei
T.
,
Mowla
R.
,
Rahman
T.
and
Venter
H.
(
2014
)
Phytochemicals increase the antibacterial activity of antibiotics by acting on a drug efflux pump
.
Microbiol. Open
3
,
885
896
50
Aparna
V.
,
Dineshkumar
K.
,
Mohanalakshmi
N.
,
Velmurugan
D.
and
Hopper
W.
(
2014
)
Identification of natural compound inhibitors for multidrug efflux pumps of Escherichia coli and Pseudomonas aeruginosa using in silico high-throughput virtual screening and in vitro validation
.
PLoS ONE
9
,
e101840
51
Takatsuka
Y.
,
Chen
C.
and
Nikaido
H.
(
2010
)
Mechanism of recognition of compounds of diverse structures by the multidrug efflux pump AcrB of Escherichia coli
.
Proc. Natl. Acad. Sci. U.S.A.
107
,
6559
6565
52
Vargiu
A.V.
and
Nikaido
H.
(
2012
)
Multidrug binding properties of the AcrB efflux pump characterized by molecular dynamics simulations
.
Proc. Natl. Acad. Sci. U.S.A.
109
,
20637
20642
53
Vargiu
A.V.
,
Ruggerone
P.
,
Opperman
T.J.
,
Nguyen
S.T.
and
Nikaido
H.
(
2014
)
Molecular mechanism of MBX2319 inhibition of Escherichia coli AcrB multidrug efflux pump and comparison with other inhibitors
.
Antimicrob. Agents Chemother.
58
,
6224
6234
54
Nakashima
R.
et al
(
2013
)
Structural basis for the inhibition of bacterial multidrug exporters
.
Nature
500
,
102
106
55
Nakashima
R.
,
Sakurai
K.
,
Yamasaki
S.
,
Nishino
K.
and
Yamaguchi
A.
(
2011
)
Structures of the multidrug exporter AcrB reveal a proximal multisite drug-binding pocket
.
Nature
480
,
565
569
56
Eicher
T.
,
Cha
H.-J.
,
Seeger
M.A.
,
Brandstätter
L.
,
El-Delik
J.
,
Bohnert
J.A.
et al
(
2012
)
Transport of drugs by the multidrug transporter AcrB involves an access and a deep binding pocket that are separated by a switch-loop
.
Proc. Natl. Acad. Sci. U.S.A.
109
,
5687
5692
57
Feng
Z.
,
Hou
T.
and
Li
Y.
(
2012
)
Unidirectional peristaltic movement in multisite drug binding pockets of AcrB from molecular dynamics simulations
.
Mol. Biosyst.
8
,
2699
2709
58
Bohnert
J.A.
,
Schuster
S.
,
Seeger
M.A.
,
Fahnrich
E
,
Pos
K.M.
and
Kern
W.V.
(
2008
)
Site-directed mutagenesis reveals putative substrate binding residues in the Escherichia coli RND efflux pump AcrB
.
J. Bacteriol.
190
,
8225
8229
59
Vargiu
A.V.
,
Collu
F.
,
Schulz
R.
,
Pos
K.M.
,
Zacharias
M.
,
Kleinekathofer
U.
et al
(
2011
)
Effect of the F610A mutation on substrate extrusion in the AcrB transporter: explanation and rationale by molecular dynamics simulations
.
J. Am. Chem. Soc.
133
,
10704
10707
60
Nguyen
S.T.
,
Kwasny
S.M.
,
Ding
X.
,
Cardinale
S.C.
,
McCarthy
C.T.
,
Kim
H.-S.
et al
(
2015
)
Structure–activity relationships of a novel pyranopyridine series of Gram-negative bacterial efflux pump inhibitors
.
Bioorg. Med. Chem.
23
,
2024
2034
61
Sjuts
H.
,
Vargiu
A.V.
,
Kwasny
S.M.
,
Nguyen
S.T.
,
Kim
H.-S.
,
Ding
X.
et al
(
2016
)
Molecular basis for inhibition of AcrB multidrug efflux pump by novel and powerful pyranopyridine derivatives
.
Proc. Natl. Acad. Sci. U.S.A.
113
,
3509
3514
62
Yilmaz
S.
,
Altinkanat-Gelmez
G.
,
Bolelli
K.
,
Guneser-Merdan
D.
,
Ufuk Over-Hasdemir
M.
,
Aki-Yalcin
E.
et al
(
2015
)
Binding site feature description of 2-substituted benzothiazoles as potential AcrAB-TolC efflux pump inhibitors in E. coli
.
SAR QSAR Environ. Res.
26
,
853
871
63
Kinana
A.D.
,
Vargiu
A.V.
,
May
T.
and
Nikaido
H.
(
2016
)
Aminoacyl β-naphthylamides as substrates and modulators of AcrB multidrug efflux pump
.
Proc. Natl. Acad. Sci. U.S.A.
113
,
1405
1410
64
Davidson
A.L.
,
Dassa
E.
,
Orelle
C.
and
Chen
J.
(
2008
)
Structure, function, and evolution of bacterial ATP-binding cassette systems
.
Microbiol. Mol. Biol. Rev.
72
,
317
364
65
van Veen
H.W.
,
Venema
K.
,
Bolhuis
H.
,
Oussenko
I.
,
Kok
J.
,
Poolman
B.
et al
(
1996
)
Multidrug resistance mediated by a bacterial homolog of the human multidrug transporter MDR1
.
Proc. Natl. Acad. Sci. U.S.A.
93
,
10668
10672
66
Reuter
G.
,
Janvilisri
T.
,
Venter
H.
,
Shahi
S.
,
Balakrishnan
L.
and
van Veen
H.W.
(
2003
)
The ATP binding cassette multidrug transporter LmrA and lipid transporter MsbA have overlapping substrate specificities
.
J. Biol. Chem.
278
,
35193
35198
67
Vandevuer
S.
,
Van Bambeke
F.
,
Tulkens
P.M.
and
Prévost
M.
(
2006
)
Predicting the three‐dimensional structure of human P‐glycoprotein in absence of ATP by computational techniques embodying crosslinking data: insight into the mechanism of ligand migration and binding sites
.
Proteins
63
,
466
478
68
Wise
J.G.
(
2012
)
Catalytic transitions in the human MDR1 P-glycoprotein drug binding sites
.
Biochemistry
51
,
5125
5141
69
Aller
S.G.
,
Yu
J.
,
Ward
A.
,
Weng
Y.
,
Chittaboina
S.
,
Zhuo
R.
et al
(
2009
)
Structure of P-glycoprotein reveals a molecular basis for poly-specific drug binding
.
Science
323
,
1718
1722
70
Jara
G.E.
,
Vera
D.M.
and
Pierini
A.B.
(
2013
)
Binding of modulators to mouse and human multidrug resistance P-glycoprotein. A computational study
.
J. Mol. Graph. Model.
46
,
10
21
71
Liu
M.
,
Hou
T.
,
Feng
Z.
and
Li
Y.
(
2013
)
The flexibility of P-glycoprotein for its poly-specific drug binding from molecular dynamics simulations
.
J. Biomol. Struct. Dyn.
31
,
612
629
72
Gadhe
C.G.
,
Kothandan
G.
and
Joo Cho
S.
(
2013
)
In silico study of desmosdumotin as an anticancer agent: homology modeling, docking and molecular dynamics simulation approach
.
Anti-Cancer Agents Med. Chem.
13
,
1636
1644
73
Ma
J.
and
Biggin
P.C.
(
2013
)
Substrate versus inhibitor dynamics of P‐glycoprotein
.
Proteins
81
,
1653
1668
74
Prajapati
R.
and
Sangamwar
A.T.
(
2014
)
Translocation mechanism of P-glycoprotein and conformational changes occurring at drug-binding site: insights from multi-targeted molecular dynamics
.
Biochim. Biophys. Acta
1838
,
2882
2898
75
Tardia
P.
,
Stefanachi
A.
,
Niso
M.
,
Stolfa
D.A.
,
Mangiatordi
G.F.
,
Alberga
D.
et al
(
2014
)
Trimethoxybenzanilide-based P-glycoprotein modulators: an interesting case of lipophilicity tuning by intramolecular hydrogen bonding
.
J. Med. Chem.
57
,
6403
6418
76
Singh
S.
and
Mandlik
V.
(
2015
)
Structure based investigation on the binding interaction of transport proteins in leishmaniasis: insights from molecular simulation
.
Mol. BioSyst.
11
,
1251
1259
77
Tomkiewicz
D.
,
Casadei
G.
,
Larkins-Ford
J.
,
Moy
T.I.
,
Garner
J.
,
Bremner
J.B.
et al
(
2010
)
Berberine-INF55 (5-nitro-2-phenylindole) hybrid antimicrobials: effects of varying the relative orientation of the berberine and INF55 components
.
Antimicrob. Agents Chemother.
54
,
3219
3224