A single acidic residue can guide binding site selection but does not govern QacR cationic-drug affinity

Structures of the multidrug-binding repressor protein QacR with monovalent and bivalent cationic drugs revealed that the carboxylate side-chains of E90 and E120 were proximal to the positively charged nitrogens of the ligands ethidium, malachite green and rhodamine 6G, and therefore may contribute to drug neutralization and binding affinity. Here, we report structural, biochemical and in vivo effects of substituting these glutamate residues. Unexpectedly, substitutions had little impact on ligand affinity or in vivo induction capabilities. Structures of QacR(E90Q) and QacR(E120Q) with ethidium or malachite green took similar global conformations that differed significantly from all previously described QacR-drug complexes but still prohibited binding to cognate DNA. Strikingly, the QacR(E90Q)-rhodamine 6G complex revealed two mutually exclusive rhodamine 6G binding sites. Despite multiple structural changes, all drug binding was essentially isoenergetic. Thus, these data strongly suggest that rather than contributing significantly to ligand binding affinity, the role of acidic residues lining the QacR multidrug-binding pocket is primarily to attract and guide cationic drugs to the "best available" positions within the pocket that elicit QacR induction.     

Electron crystallography reveals plasticity within the drug binding site of the small multidrug transporter EmrE

EmrE is a Small Multidrug Resistance transporter (SMR) family member that mediates counter transport of protons and hydrophobic cationic drugs such as tetraphenylphosphonium (TPP⁺), ethidium, propidium and dequalinium. It is thought that the selectivity of the drug binding site in EmrE is defined by two negatively charged glutamate residues within a hydrophobic pocket formed from six of the α-helices, three from each monomer of the asymmetric EmrE homodimer. It is not apparent how such a binding pocket accommodates drugs of various sizes and shapes or whether the conformational changes that occur upon drug binding are identical for drugs of diverse chemical nature. Here, using electron cryomicroscopy of EmrE two-dimensional crystals we have determined projection structures of EmrE bound to three structurally different planar drugs, ethidium, propidium and dequalinium. Using image analysis and rigorous comparisons between these density maps and the density maps of the ligand-free and TPP⁺-bound forms of EmrE, we identify regions within the transporter that adapt differentially depending on the type of ligand bound. We show that all three planar drugs bind at the same pocket within the protein as TPP⁺. Furthermore, our analysis indicates that, while retaining the overall fold of the protein, binding of the planar drugs is accompanied by small rearrangements of the transmembrane domains that are different to those that occur when TPP⁺ binds. The regions in the EmrE dimer that are remodelled surround the drug binding site and include transmembrane domains from both monomers.     

Role of a conserved membrane-embedded acidic residue in the multidrug transporter MdfA

According to the current topology model of the Escherichia coli multidrug transporter MdfA, it contains a membrane-embedded negatively charged residue, Glu26, which was shown to play an important role in substrate recognition. To further elucidate the role of this substrate recognition determinant, various Glu26 replacements were characterized. Surprisingly, studies with neutral MdfA substrates showed that, unlike many enzymatic systems where the size and chemical properties of binding site residues are relatively defined, MdfA tolerates a variety of changes at position 26, including size, hydrophobicity, and charge. Moreover, although efficient transport of positively charged substrates requires a negative charge at position 26 (Glu or Asp), neutralization of this charge does not always abrogate the interaction of MdfA with cationic drugs, thus demonstrating that the negative charge does not play an essential role in the multidrug transport mechanism. Collectively, these results suggest a link between the broad substrate specificity profile of multidrug transporters and the structural and chemical promiscuity at their substrate recognition pockets.     

Molecular properties of bacterial multidrug transporters.

One of the mechanisms that bacteria utilize to evade the toxic effects of antibiotics is the active extrusion of structurally unrelated drugs from the cell. Both intrinsic and acquired multidrug transporters play an important role in antibiotic resistance of several pathogens, including Neisseria gonorrhoeae, Mycobacterium tuberculosis, Staphylococcus aureus, Streptococcus pneumoniae, Pseudomonas aeruginosa, and Vibrio cholerae. Detailed knowledge of the molecular basis of drug recognition and transport by multidrug transport systems is required for the development of new antibiotics that are not extruded or of inhibitors which block the multidrug transporter and allow traditional antibiotics to be effective. This review gives an extensive overview of the currently known multidrug transporters in bacteria. Based on energetics and structural characteristics, the bacterial multidrug transporters can be classified into five distinct families. Functional reconstitution in liposomes of purified multidrug transport proteins from four families revealed that these proteins are capable of mediating the export of structurally unrelated drugs independent of accessory proteins or cytoplasmic components. On the basis of (i) mutations that affect the activity or the substrate specificity of multidrug transporters and (ii) the three-dimensional structure of the drug-binding domain of the regulatory protein BmrR, the substrate-binding site for cationic drugs is predicted to consist of a hydrophobic pocket with a buried negatively charged residue that interacts electrostatically with the positively charged substrate. The aromatic and hydrophobic amino acid residues which form the drug-binding pocket impose restrictions on the shape and size of the substrates. Kinetic analysis of drug transport by multidrug transporters provided evidence that these proteins may contain multiple substrate-binding sites.     

Promiscuity in the geometry of electrostatic interactions between the Escherichia coli multidrug resistance transporter MdfA and cationic substrates

The Escherichia coli multidrug transporter MdfA contains a single membrane-embedded charged residue (Glu-26) that plays a critical role in the recognition of cationic substrates (Edgar, R., and Bibi, E. (1999) EMBO J. 18, 822-832). Using an inactive mutant (MdfA-E26T), we isolated a spontaneous second-site mutation (MdfA-E26T/V335E) that re-established the recognition of cationic drugs by the transporter. Only a negative charge at position 335 was able to restore the functioning of the inactive mutant MdfA-E26T. Intriguingly, the two genetically interacting residues are located at remote and distinct regions along the secondary structure of MdfA. Glu-26 is located in the periplasmic half of transmembrane helix 1, and as shown here, the complementing charge at position 335 resides within the cytoplasmic loop connecting transmembrane helices 10 and 11. The spatial relation between the two residues was investigated by cross-linking. A functional split version of MdfA devoid of cysteines was constructed and introduced with a cysteine pair at positions 26 and 335. Strikingly, the results indicate that residues 26 and 335 are spatially adjacent, suggesting that they both constitute parts of the multidrug recognition pocket of MdfA. The fact that electrostatic interactions are preserved with cationic substrates even if the critical acidic residue is placed on another face of the pocket reveals an additional dimension of promiscuity in multidrug recognition and transport.     

The key residue for substrate transport (Glu14) in the EmrE dimer is asymmetric

Transport proteins exhibiting broad substrate specificities are major determinants for the phenomenon of multidrug resistance. The Escherichia coli multidrug transporter EmrE, a 4-transmembrane, helical 12-kDa membrane protein, forms a functional dimer to transport a diverse array of aromatic, positively charged substrates in a proton/drug antiport fashion. Here, we report (13)C chemical shifts of the essential residue Glu(14) within the binding pocket. To ensure a native environment, EmrE was reconstituted into E. coli lipids. Experiments were carried out using one- and two-dimensional double quantum filtered (13)C solid state NMR. For an unambiguous assignment of Glu(14), an E25A mutation was introduced to create a single glutamate mutant. Glu(14) was (13)C-labeled using cell-free expression. Purity, labeling, homogeneity, and functionality were probed by mass spectrometry, NMR spectroscopy, freeze fracture electron microscopy, and transport assays. For Glu(14), two distinct sets of chemical shifts were observed that indicates structural asymmetry in the binding pocket of homodimeric EmrE. Upon addition of ethidium bromide, chemical shift changes and altered line shapes were observed, demonstrating substrate coordination by both Glu(14) in the dimer.     

Negatively charged residues adjacent to IFM motif in the DIII-DIV linker of hNa(V)1.4 differentially affect slow inactivation

The effects on slow inactivation (SI) of charge substitutions, neutralizations, and reversals were studied for the negatively charged residues D1309 and EE1314,15 surrounding the IFM motif in the DIII-DIV cytoplasmic linker - the putative fast inactivation particle - of human skeletal muscle voltage-gated sodium channel (hNa(V)1.4). Changing aspartate (D) at position 1309 to glutamate (E) (substitution) did not strongly affect SI, whereas charge neutralization to glutamine (Q) and charge reversal to arginine (R) right-shifted the midpoint of the steady-state SI curve. Charge neutralization (D-->Q) at position 1309 also reduced the apparent valence associated with SI. Glutamates (E) at positions 1314 and 1315 were similarly mutated. Charge reversal (EE-->RR) right-shifted the steady-state SI curve and both reversal and substitution (EE-->DD) reduced its apparent valence. Charge neutralization (EE-->QQ) and reversal decreased the maximum probability of SI. These mutations also had differential effects on the rate of SI onset and recovery. These results suggest that charged residues in the DIII-DIV linker may interact with structures that control SI.     

Glutamate substitution in repeat IV alters divalent and monovalent cation permeation in the heart Ca2+ channel.

In voltage-gated ion channels, residues responsible for ion selectivity were identified in the pore-lining SS1-SS2 segments. Negatively charged glutamate residues (E393, E736, E1145, and E1446) found in each of the four repeats of the alpha 1C subunit were identified as the major determinant of selectivity in Ca2+ channels. Neutralization of glutamate residues by glutamine in repeat I (E393Q), repeat III (E1145Q), and repeat IV (E1446Q) decreased the channel affinity for calcium ions 10-fold from the wild-type channel. In contrast, neutralization of glutamate residues in repeat II failed to significantly alter Ca2+ affinity. Likewise, mutation of neighboring residues in E1149K and D1450N did not affect the channel affinity, further supporting the unique role of glutamate residues E1145 in repeat III and E1446 in repeat IV in determining Ca2+ selectivity. Conservative mutations E1145D and E1446D preserved high-affinity Ca2+ binding, which suggests that the interaction between Ca2+ and the pore ligand sites is predominantly electrostatic and involves charge neutralization. Mutational analysis of E1446 showed additionally that polar residues could achieve higher Ca2+ affinity than small hydrophobic residues could. The role of high-affinity calcium binding sites in channel permeation was investigated at the single-channel level. Neutralization of glutamate residue in repeats I, II, and III did not affect single-channel properties measured with 115 mM BaCl2. However, mutation of the high-affinity binding site E1446 was found to significantly affect the single-channel conductance for Ba2+ and Li+, providing strong evidence that E1446 is located in the narrow region of the channel outer mouth. Side-chain substitutions at 1446 in repeat IV were used to probe the nature of divalent cation-ligand interaction and monovalent cation-ligand interaction in the calcium channel pore. Monovalent permeation was found to be inversely proportional to the volume of the side chain at position 1446, with small neutral residues such as alanine and glycine producing higher Li+ currents than the wild-type channel. This suggests that steric hindrance is a major determinant for monovalent cation conductance. Divalent permeation was more complex. Ba2+ single-channel conductance decreased when small neutral residues such as glycine were replaced by bulkier ones such as glutamine. However, negatively charged amino acids produced single-channel conductance higher than predicted from the size of their side chain. Hence, negatively charged residues at position 1446 in repeat IV are required for divalent cation permeation.     

Conserved aromatic residues in the transmembrane region VI of the V1a vasopressin receptor differentiate agonist vs. antagonist ligand binding.

Despite their opposite effects on signal transduction, the nonapeptide hormone arginine-vasopressin (AVP) and its V1a receptor-selective cyclic peptide antagonist d(CH2)5[Tyr(Me)2]AVP display homologous primary structures, differing only at residues 1 and 2. These structural similarities led us to hypothesize that both ligands could interact with the same binding pocket in the V1a receptor. To determine receptor residues responsible for discriminating binding of agonist and antagonist ligands, we performed site-directed mutagenesis of conserved aromatic and hydrophilic residues as well as nonconserved residues, all located in the transmembrane binding pocket of the V1a receptor. Mutation of aromatic residues of transmembrane region VI (W304, F307, F308) reduced affinity for the d(CH2)5[Tyr(Me)2]AVP and markedly decreased affinity for the unrelated strongly hydrophobic V1a-selective nonpeptide antagonist SR 49059. Replacement of these aromatic residues had no effect on AVP binding, but increased AVP-induced coupling efficacy of the receptor for its G protein. Mutating hydrophilic residues Q108, K128 and Q185 in transmembrane regions II, III and IV, respectively, led to a decrease in affinity for both agonists and antagonists. Finally, the nonconserved residues T333 and A334 in transmembrane region VII, controlled the V1a/V2 binding selectivity for both nonpeptide and cyclic peptide antagonists. Thus, because conserved aromatic residues of the V1a receptor binding pocket seem essential for antagonists and do not contribute at all to the binding of agonists, we propose that these residues differentiate agonist vs. antagonist ligand binding.     

A single membrane-embedded negative charge is critical for recognizing positively charged drugs by the Escherichia coli multidrug resistance protein MdfA

The nature of the broad substrate specificity phenomenon, as manifested by multidrug resistance proteins, is not yet understood. In the Escherichia coli multidrug transporter, MdfA, the hydrophobicity profile and PhoA fusion analysis have so far identified only one membrane-embedded charged amino acid residue (E26). In order to determine whether this negatively charged residue may play a role in multidrug recognition, we evaluated the expression and function of MdfA constructs mutated at this position. Replacing E26 with the positively charged residue lysine abolished the multidrug resistance activity against positively charged drugs, but retained chloramphenicol efflux and resistance. In contrast, when the negative charge was preserved in a mutant with aspartate instead of E26, chloramphenicol recognition and transport were drastically inhibited; however, the mutant exhibited almost wild-type multidrug resistance activity against lipophilic cations. These results suggest that although the negative charge at position 26 is not essential for active transport, it dictates the multidrug resistance character of MdfA. We show that such a negative charge is also found in other drug resistance transporters, and its possible significance regarding multidrug resistance is discussed.     

A New Critical Conformational Determinant of Multidrug Efflux by an MFS Transporter

Secondary multidrug (Mdr) transporters utilize ion concentration gradients to actively remove antibiotics and other toxic compounds from cells. The model Mdr transporter MdfA from Escherichia coli exchanges dissimilar drugs for protons. The transporter should open at the cytoplasmic side to enable access of drugs into the Mdr recognition pocket. Here we show that the cytoplasmic rim around the Mdr recognition pocket represents a previously overlooked important regulatory determinant in MdfA. We demonstrate that increasing the positive charge of the electrically asymmetric rim dramatically inhibits MdfA activity and sometimes even leads to influx of planar, positively charged compounds, resulting in drug sensitivity. Our results suggest that unlike the mutants with the electrically modified rim, the membrane-embedded wild-type MdfA exhibits a significant probability of an inward-closed conformation, which is further increased by drug binding. Since MdfA binds drugs from its inward-facing environment, these results are intriguing and raise the possibility that the transporter has a sensitive, drug-induced conformational switch, which favors an inward-closed state.     

The Staphylococcal QacR Multidrug Regulator Binds a Correctly Spaced Operator as a Pair of Dimers

Expression of the Staphylococcus aureus plasmid-encoded QacA multidrug transporter is regulated by the divergently encoded QacR repressor protein. To circumvent the formation of disulfide-bonded degradation products, site-directed mutagenesis to replace the two cysteine residues in wild-type QacR was undertaken. Analysis of a resultant cysteineless QacR derivative indicated that it retained full DNA-binding activities in vivo and in vitro and continued to be fully proficient for the mediation of induction of qacA expression in response to a range of structurally dissimilar multidrug transporter substrates. The cysteineless QacR protein was used in cross-linking and dynamic light-scattering experiments to show that its native form was a dimer, whereas gel filtration indicated that four QacR molecules bound per DNA operator site. The addition of inducing compounds led to the dissociation of the four operator-bound QacR molecules from the DNA as dimers. Binding of QacR dimers to DNA was found to be dependent on the correct spacing of the operator half-sites. A revised model proposed for the regulation of qacA expression by QacR features the unusual characteristic of one dimer of the regulatory protein binding to each operator half-site by a process that does not appear to require the prior self-assembly of QacR into tetramers.