Abstract
The characterization of Multidrug and Toxic Compound Extrusion (MATE) transporters has led to the profound discovery of the mediation of vacuolar transport in plants and the resistance of multidrug in mammals and bacteria. MATE transporters were first identified by Tsuchiya and his colleagues in Vibrio parahaemolyticus and Escherichia coli, named NorM and YdhE, respectively. MATE transporters are organic cation transporters that belong to the multidrug transporter family. The multidrug family is comprised of six main families such as the small multidrug resistance (SMR) family, the major facilitator superfamily (MFS), the resistance nodulation cell division (RND) family, the ATP-binding cassette (ABC) family, the MATE family, and now recently, the Proteobacterial Antimicrobial Compound Efflux (PACE). MATE transporters extrude chemically distinct drugs, including xenobiotics and endogenous compounds, across the plasma membrane. Thus, MATE transporters have been advantageous for the breakthrough of safer drugs for the improvement of drug therapy. This paper will discuss the structural and functional significance of MATE transporters as well as its molecular mechanisms through which the H+ and Na+ electrochemical gradients pump cytotoxic chemicals out of the cell.
Introduction
Multidrug and Toxic Compound Extrusion (MATE) are found in all three domains of life. In plants, MATE proteins are responsible for the metabolite transport, affecting crop yields all over the world. MATE transporters can also facilitate multidrug resistance in mammals and bacteria through a mechanism in which it exports cytotoxic molecules out of the cell, thus impeding the improvement of pharmaceutical drugs for various diseases. MATE transporters were first discovered in 1998 by Tsuchiya and his colleagues. Based on their experimental data, they concluded that Vibrio parahaemolyticus, NorM, and its homolog YdHe of E. coli are Na+-driven Na+/drug antiporters (Fig 1). Initially, these drug efflux systems, NorM and YdhE were assigned to the major facilitator superfamily (MFS) because of their possession of 12 to 14 transmembrane domains. However, a year later, Skurray and his colleagues discovered that neither of these proteins have a significant sequence similarity with any members of the major facilitator superfamily. Thus, NorM and YdhE were reassigned to a new family of transporters known as the Multidrug and Toxic Compound Extrusion (MATE) family.
The families of multidrug resistance transporters include the small multidrug resistance (SMR) family, the major facilitator superfamily (MFS), the resistance nodulation cell division (RND) family, the ATP-binding cassette (ABC) family, the MATE family, and now recently, the Proteobacterial Antimicrobial Compound Efflux (PACE) (Fig 2). The ATP-binding cassette (ABC) family primarily use the energy of ATP hydrolysis and binding to transport substrates across the cell membrane, whereas the rest of the multidrug resistance transporter families require proton movement down the electrochemical gradient to export drugs out of the cell. Specifically, the RND, SMR, and MFS multidrug transporters mainly utilize preexisting H+ gradient to pump cytotoxic chemicals across the membrane, while MATE transporters can use either Na+ or H+ electrochemical gradient to drive transport.
Molecular Function
ATP-binding cassette transporters
The ABC superfamily is comprised of four domains: two cytoplasmic, nucleotide-binding domains (NBDs) that hydrolyze and bind ATP and possess a particular protein fold than that of other ATP-binding transporters and two membrane-spanning domains (MSDs) which consist of typically membrane-spanning -helices that serve as a means for small molecules to transport across the cell membrane. ABC transporter are usually specific for a certain ligand such as amino acids, sugar, polypeptide or an inorganic ion. However, some ABC transporters, like the Caenorhabditis elegans P-glycoprotein (P-gp) have a broader specificity for hydrophobic molecules (Mousa et al, 2016). When overexpressed, P-glycoprotein can implicate resistance of cancer and bacterial cells to chemotherapeutic drugs including paclitaxel, doxorubicin, etoposide, and actinomycin. ABC transporters can also confer drug resistance in parasitic protozoa and fungi and herbicide resistance in plants.
Sav1866 from Staphylococcus aureus was the first homologous ABC bacterial multidrug transporter to be identified in the ATP-bound state. In the structure of Sav1866, two MSDs form an extracellularly open chamber that allows one large molecule or two small molecules to bind at the ATP-binding pocket. Several transmembrane -helices contribute to the aromatic and hydrophobic amino acids within the chamber. Typically, NBDs interact with one another in an outward facing conformation and are taken apart in the inward conformation. Like all other ABC transporters, ATP hydrolysis allows for the NBD to move back to its intracellular-facing conformation where another drug molecule can be attached.
P-glycoprotein is believed to have a similar molecular function to that of Sav1866. The P-gp, solved in apo-form, was crystallized in three various inward facing conformations which exemplified its substrate diversity. X-ray structured revealed that NBD and MSD are linked together via a “ball-and-socket†bridge, similar to a salt bridge in ABC importers. Small molecules bind to the drug-binding pocket developed by MSD’s in the cell membrane. Drugs tend to bind to high-affinity sites on transporter proteins based on symmetry, thus this symmetry effectively increases transport of another drug (Higgins et al., 2007). Another groundbreaking discovery is the recent identification of the heterodimeric ABC transporter structures, TMrAB and TM287/288. The two NBDs stayed together unlike other transporters where NBDs are eventually separated in the inward-facing conformation. Even when AMP-PNP is not present, the NBDs remain intact and thereby provides an allosteric crosstalk mechanism. This mechanism was observed at the D-loops of two NBDs. D-loops connect the two NBDs together at the ATP-binding site even when there are no nucleotides involved.
NaAtm1, a heavy metal detoxification transporter, also demonstrates the vast transport capabilities of ABC transporters. NaAtm1 utilizes its highly conserved residues to transport Ag- and Hg-glutathione molecules. Additionally, the human transporter, ABCB10 was also identified. This transporter consists of similar folds in the inwards state to that of other transporters within the same superfamily. ABCB10 also validated the observation that nucleotides can still bind without the need for a bound substrate.
The structure of a peptidase-containing ATP-binding cassette transporter (PCAT) characterized the mechanism of a peptidases cleaving off a leader sequence of another peptide then exporting it. This structure was classified in the ATP-free and bound state in the Clostridium thermocellum. In the ATP-free state, the structure consisted of MSDs, NBDs, and the peptidase domain which is responsible for cleaving the peptide structure and transporting it to the MSDs. Conversely, the ATP-bound conformation, the peptidase domain is not acknowledged. Furthermore, an another observation is acknowledged that the peptidase domain recruits the peptide substrate, thus providing a remarkable perception into the peptidase-coupled ABC transporter.
Resistance nodulation cell division transporters
AcrAB-TolC of E. coli is the best characterized RND transporter protein that can implicate the resistance to a multitude of drugs. RND transporters consist of three protein complexes. In the case of AcrAB-TolC, the complex is comprised of a cytoplasmic membrane AcrB, a pore-like outer membrane molecule TolC that is comprised of a100 Å α-helical pore, and AcrA that serves at a bridge between TolC and AcrB. The complex is believed to exist in 3:6:3 AcrB: AcrA: TolC ratio. The periplasmic domain in each subunit of the AcrAB trimer has a slightly different conformation that contains a substrate recognition side and drugs attached to the structures (Figure 3). Recently, the AcrAB-TolC tripartite complex has been distinguished structurally using cryo-electron microscopy (EM). AcrZ, a 49-residue protein, was demonstrated to influence the complex’s resistance to antibiotics by binding to AcrB across the cytoplasmic membrane (Higgins et al., 2007). Although TolC adopts a closed conformation in isolation, it is open in the cryo-EM representation. This open conformation is likely due to TolC interacting with AcrA. The AcrAB-TolC complex can transport a wide range of drugs out of the cell that have little to no structural or sequence similarity. This complex essentially allows E. coli to remove bile salts including glycocholate and taurocholate and to resist tetraphenylphosphonium, ethidium, and the β-lactam family.
Pyridopyrimidine-derivative, a bound inhibitor, was bound to a pocket consisting of phenylalanine residues, or the “hydrophobic trapâ€. Binding in the hydrophobic trap forms – stacking interactions and keeps the proteins from rotating which is essential to the extrusion process. Using various inhibitors led to the observation that inhibitor potency was increased when acrylamide or acetamide were added in hydrogen bonding interactions with AcrB. This development allows for antibiotic generation and testing to happen in a controlled setting.
Small multidrug resistance transporters
Small multidrug resistance (SMR) proteins are only restricted to prokaryotic cells. They only have four transmembrane α-helices, making them the smallest multidrug transporters. SMR proteins only have one structure in the Protein Data Bank (PDB), EmrE from E.coli. EmrE is a kDa electrogenic transporter that can promote the resistance to a wide range of hydrophobic molecules such as ethidium and methyl violegen at a neutral pH. EmrE has been hypothesized to exist as a homodimer based on EM and X-ray structural data of EmrE . Data from NMR and FRET spectroscopy illustrate that EmrE is also an antiparallel homodimer in which the two subunits are oriented in opposite directions (Mousa et al., 2016). Essentially, EmrE is a cluster of eight transmembrane α-helices that form a hydrophobic and aromatic amino acid filled pathway (Figure 4). Glu (Th1) is known to be essential for cationic drug binding and the movement of protons. This residue’s involvement in proton and substrate binding suggests a different mechanism of proton exchange for RND and MRS transporters.
Major facilitator superfamily transporters
Major facilitator superfamily (MFS) transporters are the largest and structurally diverse multidrug transporters. The MFS are similar to MATEs in that they both are comprised of 12-transmembrance helices, hence why MATEs were believed to belong to the MFS. However, MFS are much more functionally and structurally unique. Although some MFS proteins have a high specificity for certain molecules, the MFS can transport a wide range of molecules including amino acids, ions, peptides, nucleotides, drugs, and many others. The first structures of MFS transporters were identified in 2003, termed LacY and G1pT from E.coli. This led to the subsequent discovery of EmrD, MdfA, and YajR—all derived from E.coli. EmrD is a protein consisting of twelve membrane-spanning-α-helices and can export a diverse group of cytotoxic molecules form the cell including SDS, 3-chlorophenylhydrazone, and benzylalkonium chloride (Higgins et al., 2007). The EmrD homolog MdfA appears to transport erythromycin, benzalkonium, tetraphenylphosphonium, ciprofloxacin, thiamphenicol, and chloramphenicol (Figure 4).
As previously mentioned YarjR is a derivative of E. coli; however, it has structurally unique components that differentiates it from other MFS transporters. YarjR is an antiporter and is believed to have a break in TH4 that increases the size of its internal cavity thereby affecting its specificity for substrates. YarjR’s internal cavity is lined with numerous hydrophilic residues facing the central cavity, thus making it highly charged. Additionally, YarjR’s C-terminus is displayed on to a 65-residue domain in a ferredoxin manner. The lack of interaction between this domain and the YarjR cytoplasmic peptides suggests that the C-terminus may have more of a regulatory role than a structural one (Mousa et al., 2016). The structure of YarjR also provides evidence of the first MFS “motif Aâ€. Motif A is a conserved region that plays an integral role in the loop that attaches TH2 and TH3 to one another, thus stabilizing them in an outward-facing state. The outward state of YajR consists of Gly69, motif A residues Asp73 and Arg77, an interdomain charge-helix with residues TH11 and Asp73, and a charge-relay system with Asp126(TH4). Studies suggest that the removal of the charge-helix dipole lock by virtue of the outward to inward conversion provides the energy for that very conversion.
MFS can use multiple mechanisms to transport molecules such as the symporter process, where H+ is imported into the cell with other small molecules. One example of a transport mechanism is the disruption of an Asp(TH1)-Arg(TH4) salt bridge once Asp is protonated. In the LacY structure, Glu(TH10) and Arg(TH9) plays an essential role in proton translocation. Upon protonation of Glu(TH4) or Asp(TH1), an Asp/Glu proton translocation process begins. This causes the structure to change conformations for the FucP transport mechanism. Peptide transporters also has a similar mechanism. Although peptide transporters contain two Arg-Glu salt bridges, only one salt bridge, Arg(TH1) and Glu(TH7), is involved in proton translocation. In GkPOT, Glu(TH1) attaches itself to the carboxyl group of alafosfalin peptide substrate (Higgins et al., 2007). This leads to deprotonation which induces the release of the substrate and the development of the salt bridge with Arg(TH1) to stabilizes the structure. For this very reason, bacterial PeptSo uses His(TH2). In NRT1.1, a nitrate transporter, His(TH7) is protonated which allows for the nitrate to bind and induce a change in conformation. This conformation change destroys the Lys(TH4)-Glu(TH10) salt bridge thereby allowing the phosphorylation of Thr(TH4) to increase the flexibility of the N-terminal.
The cation-coupled melibiose symporter comprised of two Asp residues on TH2 and on the TH4 is another specially characterized mechanism of MFS transporters. The cation binding around the symporter site is thought to bind to either Na+, H3O+, or H+. Additionally, the Asp(TH1) and Arg(TH5) salt bridge binds to the substrate. This structure has supported the evidence for other symporter mechanisms.
Cellular Localization of MATE Transporters
In humans, Multidrug and Toxic Compound Extrusion (MATE) transporters are widely expressed in the liver and kidney, although it can also be expressed in testis, skeletal muscle, first semester trimester, and adrenal gland. MATE2 and MATE2-K are typically localized in the kidney; however, MATE2 has also been observed in the placenta. MATE transporters work in conjunction with organic cation transporter to regulate the transport of their substrates. In the liver and kidneys, OTC-MATE serves as a pathway to export a vast range of structurally and functionally diverse molecules such as cytotoxic molecules and endogenous compounds. OCTs are found in the basolateral membrane (apical, urine side) of proximal tubules of the kidney and canalicular membrane (apical, bile side) of hepatocytes where they induce the movement of their substrates into the cell (Mousa et al., 2016). Furthermore, MATEs are localized in the apical membrane of the polarized cells in which they export small molecules out of the cell. MATEs are also often co-expressed with multidrug transporters from the ABC family, specifically the breast cancer protein, P-glycoprotein, or MRP2.
Structures of MATE Transporters and their Molecular Mechanism
MATE transporters can be further divided into subfamilies, NorM, eukaryotic, and DniF (DNA damage-inducible protein F), based on their amino-acid sequence similarity. Transporters belonging to the NorM and DinF subfamilies use the Na+ or H+ electrochemical gradient whereas eukaryotic MATE proteins are H+ dependent. Additionally, NorM and DinF contain both eubacterial and archaeal groups that have low amino-acid sequence similarity. Similar to the rest of the multidrug family, MATE transporters can push out positively charged drugs at physiological pH (Lu, 2016). Thus, MATE transporters are poly-specific, meaning they cannot export negatively charged compounds. Nonetheless, MATE serve as xenobiotic efflux pumps that induce the resistance to anti-cancer, anti-diabetic, and many other antibiotics.
The structure of Cation-bound NorM-VC
NorM-VC from Vibro cholera provides a fundamental understand on the mechanism of MATE transporters. The structure of NorM-VC was solved at a resolution of 3.65 Å which portrays the transporter in an outward-facing, substrate-free conformation, revealing its two clusters of six transmembrane helices. NorM-VC covers approximately 50 Å in the plane of the lipid bilayer and develops a large internal cavity. A “V†shape is formed by the NorM-VC structure with each arm of the “V†consisting of 6 transmembrane helices (Figure 5). This divides the transporter in two domains: the N domain from transmembrane 1 to transmembrane 6 and the C domain from transmembrane 7 to transmembrane 12 (Lu, 2016). The two domains appear to fold symmetrically along an axis where they are then laid on to top of each other to achieve a root mean squared deviation of < 3 Å.
NorM-VC attributes the export of drugs out of the cell to the influx of cations such as Na+ and the cation-binding site at the C terminus of the transporter, shown by the 4.2 Å resolution structure of NorM-VC bound to Rb+. Multiple amino acids were hypothesized to coordinate Na+, based on this structure. However, the NorM-VC doesn’t reveal much on how a MATE transporter induces drug-Na+ exchange or how it binds to Na+. Although Rb+ and Cs+ share similarities to that of Na+, they cannot mediate the export of drugs out of the cell. Still, Rb+ or Cs+ have been critical in localizing Na+ binding sites in membrane transporters (Figure 6).
The Structure of Substrate-bound NorM-NG
The structure of a substrate-bound NorM transprter from Neisseria gonorrhea was solved at a resolution of 3.5 Å. The structure appeared to be in an outward-facing, substrate bound confirmation that shows the MATE transport interacting with polyaromatic and cationic molecules. NorM-NG is comprised of 12 transmembrane helices, similarly to that of NorM-VC. The N and C domain also adopt a “V†shape near the center of the membrane bilayer. Like NorM-VC, the two domains also have a two-fold symmetry around the membrane. Moreover, the central cavity is found between the N and C domains where it is only slightly closed toward the extracellular space, allowing for ions and other solvents to transport easily into the cavity.
Structures without Na+ were used to identify the substrate-binding site of NorM-NG. These structure reveal that the central cavity between the N and C domains is makes up the substrate-binding site in NorM-NG which provides evidence on the mechanism of MATE transporters. The solvent on the extracellular side of the membrane is able to access about 30% surface area of the bounds substrates (Lu, 2016).. Thus, these structures show the transporter in an extracellularly-open, substrate bound conformation. Additionally, three different substrates interact closely with the amino acids of both N and C domains in the central cavity. Remarkably, the NorM-NG structure produces more H-bonding and ionic contacts with lipophilic and cationic molecules. This is ironic because hydrophobic interactions usually dominate the interactions between multidrug transporters and their drug substrate. This then suggests that NorM-NG’s transport mechanism has adapted over time from other multidrug transporters.
Cation-bound NorM-NG structure
The structure of Cs+ attached to NorM-NG was also identified at a resolution of 3.7 Ã…. It is very similar to the structure of the substrate-bound NorM-NG due to the unknown ligan that serves a surrogate substrate to stabilized the substrate-bound protein . The cation-bound NorM-NG structure provides a new perspective on the coordination of Na+ in the Na= coupled MATE transporter. Furthermore, the cation-bound NorM-NG structure is involved with 2 carboxylate amino acids (NorM-NGD377 and NorM-NGE261) and one aromatic compound (NorM-NGY294) (Lu, 2016). These 3 amino acids are significantly conserved in the eukaryotic and NorM subfamilies and are essential for the functionality of multidrug efflux.
NorM-NGY294 provides evidence that a membrane protein can coordinate through a Na+- interaction. This interaction allows for NorM-NG to bind to Na+ during the transport process. Additionally, the Na+- interaction is also believed to be involved with BetP, a secondary Na+ coupled symporter. This interaction can possibly be an integral part in Na+ binding by membrane antiporters and symporters. NorM-NG’s cation-bound structure also implies that the transporter uses a specific set of amino acid that can interact with drugs and cations and bind them at the same time (Lu, 2016). Thus, the interaction of a drug and Na+ is facilitated by the conformational changes in the arrangement of NorM-NG’s transmembrane helices.
PfMATE structure and its H+ coupled Mechanisms
H+ coupled MATE transporters from Pyrococcus furious (PfMATE) were first identified at 2.4-3.0 Å resolutions. Although PfMATE shares structure similarities to that of NorM-NG or NorM-VC, they lack amino acids that make up the cation-binding site in NorM-NG. PfMATE also belong to the Bacillus halodurans (DiNF) subfamily which was determined as a proton-coupled transporter with resistance to rhoadamine 6G, tetraphenylphosphodium, and ethidium. PfMATE is described in to different apo-states: TH1 is “bent†in the low pH, protonated state and “straight in the high pH, deprotonated state (Lu, 2016). Because TH1 is straight in the PfMATE complex with a substrate analog, studies suggest that the protonation of a conserved aspartate results in the collapse of the ligand-binding pocket and yields the bent conformation of TH1. In comparison to other cation-coupled transporters, PfMATE is believed to be unable to bind metal cations.
Structure of the substrate-bound DinF-BH
After the PfMATE complexes were published, the structure of Bacillus halodurans were solved at resolutions between 3.2 – 3.7 Å. Specifically, the 3.2 Å resolution substrate-free and 3.7 Å substrate-bound DinF-BH structures both exemplify an asymmetric conformation of the 12 transmembrane helices. TH7 and TH8 are kinked around two proline residues, resulting in a broken pseudo-twofold symmetry. The asymmetric structure within DinF-BH gives rises to a substrate-binding chamber and blocks TM1 from the solvent. Inside the substrate-binding chamber, DinF-BH interacts with a cationic substrate in single charge-charge manner (Mousa et al, 2016). The protonated mutant of DinF-BH, DinG-BHD40N establishes a hydrogen bond with DinF-BHD184. The molecular mechanism on DinF-BH is reported to begin with a H+ competing with a bound substrate which then upsets drug binding. The protonation of DinF-BHD40 disrupts the interaction between DinF-BHD40N and the ligand, thus activating the release of the drug from the transporter. The protonated protein adopts an intracellular conformation and TH7 and TH8 shift by approximately 20.
Recently, X-ray structured of NorM-NG and DinF-BH in complexes with verapamil were established. Verapamil is a drug and ion-channel blocker that inhibits both hMATE1, hMATE2, ABC and MFS families. This drug was also proven to reduce bacterial growth in point mutants and inhibit ethidium export in transporters. In DinF-BH and NorM-NG structures bound to verapamil, verapamil is located in a space that is overlapping with the multidrug-binding sites. In vitro, verapamil can also inhibit a radio-labeled ligand from binding to either NorM-NG or DinF-BH (Mousa et al, 2016). Thus, verapamil inhibits MATE-facilitated multidrug efflux by binding to the multidrug binding site, which is akin to the inhibitory mechanism of the inhibitors of RND multidrug transporters. Although DinF-BH and NorM-NG have a similar inhibitory mechanism, they have very distinct molecular interactions with verapamil. First, verapamil adopts a linear state within DinF-BH and a horse-shoe like, folded conformation in NorM-NG, which can increase favorable interactions between verapamil and the MATE transporter. Second, the DinF-BH is significantly unaffected when bound to verapamil, while the two extracellular loops within NorM-NG undergo a conformational change upon binding with verapamil (Lu, 2016). These differences exemplify the mechanical and structural variance between NorM-NG and DinF-BH.
Biological Implications
As discussed before, multidrug resistance transporters play a significant role in the export of cytotoxic chemicals out of the cell. Over the past two decades, studies have been conducted to overcome drug resistance in biology. Scientists have worked considerably to chemically alter the composition of antibiotic and drugs so that they no longer bind multidrug transporters. Unfortunately, little changes made to a drug seldom cause a decrease in the affinity for transport. Thus, these efforts have been proven to be futile. Researchers now understand that this is due to large, flexible drug-binding sites and because perfect alignment of an amino acid to a substrate is not needed for high-affinity binding. Also, any alteration to an antibiotic that would decrease its affinity for a transporter would also read its ability to cross the plasma membrane and attach to its target. Another method to prevent drug resistance is the use of inhibitors of transporters which also has been unsuccessful. Since numerous multidrug transporters can resist a specific drug, more than one inhibitor is necessary. Therefore, trial and error methods were utilized to develop several inhibitors of human P-glycoprotein (Higgins et al., 2007). Although human P-glycoproteins are successful in the lab, they have various side effects caused by inhibiting P-glycoprotein’s physiological functions.
Conclusion
The past several years has shed light upon the multiple mechanisms utilized for multidrug resistance transporters, yet there is still work to be done to completely understand the kinetic studies of each transport cycle. Although not much is known about the structure of the MATE family, MATE transport inhibitors seem likely as cyclic peptide inhibitors of PfMATE were developed. Inhibition targets for the MATE family would probably include cation or proton binding sites where disruption of electrostatic interactions would be necessary to prevent the transportation of drugs.