Molecular determinants of ligand selectivity for the human multidrug and toxin extruder proteins MATE1 and MATE2-K

Abstract

The present study compared the selectivity of two homologous transport proteins, multidrug and toxin extruders 1 and 2-K (MATE1 and MATE2-K), and developed three-dimensional pharmacophores for inhibitory ligand interaction with human MATE1 (hMATE1). The human orthologs of MATE1 and MATE2-K were stably expressed in Chinese hamster ovary cells, and transport function was determined by measuring uptake of the prototypic organic cation (OC) substrate 1-methyl-4-phenylpyridinium (MPP). Both MATEs had similar apparent affinities for MPP, with K(tapp) values of 4.4 and 3.7 μM for MATE1 and MATE2-K, respectively. Selectivity was assessed for both transporters from IC(50) values for 59 structurally diverse compounds. Whereas the two transporters discriminated markedly between a few of the test compounds, the IC(50) values for MATE1 and MATE2-K were within a factor of 3 for most of them. For hMATE1 there was little or no correlation between IC(50) values and the individual molecular descriptors LogP, total polar surface area, or pK(a). The IC(50) values were used to generate a common-features pharmacophore, quantitative pharmacophores for hMATE1, and a bayesian model suggesting molecular features favoring and not favoring the interaction of ligands with hMATE1. The models identified hydrophobic regions, hydrogen bond donor and hydrogen bond acceptor sites, and an ionizable (cationic) feature as key determinants for ligand binding to MATE1. In summary, using a combined in vitro and computational approach, MATE1 and MATE2-K were found to have markedly overlapping selectivities for a broad range of cationic compounds, including representatives from seven novel drug classes of Food and Drug Administration-approved drugs.

Fig. 1.

Kinetic characteristics of transport mediated by hMATE1 and hMATE2-K expressed in CHO cells. A and B, kinetics of MPP transport mediated by hMATE1 (A) or hMATE2-K (B). C and D, effect of extracellular [H⁺] on MPP transport mediated by hMATE1 (C) or hMATE2-K (D). In all experiments, 5-min uptakes of [³H]MPP (∼13 nM) were measured in the presence of increasing concentrations of unlabeled MPP (A and B) at an external pH of 8.5 or increasing extracellular H⁺ concentrations (C and D). Each point is the mean (± S.E.M.) of uptakes measured in three wells of a 24-well plate from single representative experiments. Kinetic values shown represent the average of three to five experiments.

Fig. 2.

Range of inhibition of transport mediated by the human orthologs of MATE1 (A) and MATE2-K (B) produced by the battery of test compounds used in this study. The height of the gray bars indicates the degree of inhibition of mediated uptake (5 min) of [³H]MPP (∼13 nM) produced by 10 μM concentrations of 59 test compounds. The horizontal dashed lines indicate 50% inhibition of transport.

Fig. 3.

The effect on the rate of hMATE1- and hMATE2-K-mediated MPP transport of increasing concentration of representative test inhibitors. ● represent hMATE1, ○ represent hMATE2-K, and they show inhibition produced by quinidine (A), agmatine (B), nialamide (C), and allopurinol (D). Each point represents the mean value (± S.E.M.; determined in three separate experiments) of the 5-min uptake of [³H]MPP (∼13 nM) measured in the presence of increasing concentration of test inhibitor; uptakes were normalized to that measured in the absence of inhibitor. NI, no interaction.

Fig. 4.

Comparison of hMATE1 and hMATE2-K IC₅₀ values. A, hMATE1 IC₅₀ values were graphed as a function of hMATE2-K IC₅₀ values for the 59 test compounds measured at pH 8.5 (left) and as calculated for pH 7.4 (right). The solid lines represent equal IC₅₀ values for the two transporters; the dashed lines indicate 3-fold ± differences in these values. B, inhibitory profiles for several test ligands (atropine, amantadine, and APMI, left to right) against MPP transport mediated by hMATE1 (●) and hMATE2-K (○). Each point is the mean of 5-min uptakes (± S.E.M.; normalized to uptake measured in the absence of inhibitor) determined in three separate experiments, each run in duplicate. IC₅₀ values are average values from three separate experiments.

Fig. 5.

A, C, and D, relationship between hMATE1 IC₅₀ values and the molecular descriptors LogP (A), TPSA (C), and pK_a (D). B, the relationship between LogP and the IC₅₀ values for a structurally constrained n-tetraalkylammonium series (TMA, tetramethylammonium; TEA; TPrA, tetrapropylammonium; TBA, tetrabutylammonium; and TPeA).

Fig. 6.

Inhibitory profiles for the five test ligands (A) used to generate a common-features pharmacophore (B). Each point is the mean of 5-min uptake (normalized to the transport measured in the absence of inhibitor) determined in single representative experiments with PYR (▵), quinidine (▴), histamine (♦), caffeine (■), and chloramphenicol (●). The common-features pharmacophore (displayed with the structure of PYR) includes two hydrophobic regions (cyan), one H-bond donor (magenta), and one H-bond acceptor (green).

Fig. 7.

A and B, kinetics of inhibition of hMATE1-mediated MPP transport produced by cinchonidine (A) and ethohexadiol (B). Each point is mean (± S.E.M.) of 5-min uptake measured in triplicate in single representative experiments. IC₅₀ values are average values from three separate experiments. C and D, common-features pharmacophores are shown for cinchonidine (C) and ethohexadiol (D).

Fig. 8.

A, quantitative pharmacophore generated from analysis of data obtained by using the first round of hMATE1 inhibitors (see “First Iteration: Quantitative Pharmacophore Development for hMATE1” in Results). Twenty four of the initial 26 compounds were used in an analysis that resulted in a model containing two hydrophobic features (cyan), one hydrogen bond donor (magenta), and one positive ionizable feature (red), shown here with the structure of cinchonidine. B, the relationship between measured and predicted IC₅₀ values based on the model shown in A (r = 0.68; p < 0.0001). C, quantitative pharmacophore generated from analysis of the data that incorporated the second round of hMATE1 inhibitors. Analysis of 43 compounds (including the initial 24 plus the test set of 15 compounds that probed the common features model) resulted in a model that included two hydrophobes (cyan), two hydrogen bond acceptors (green), and an ionizable feature (red). D, the relationship between measured and predicted IC₅₀ values based on the model shown in C (r = 0.71; p < 0.0001). E, quantitative pharmacophore generated from analysis of 46 of the 59 test ligands (see “Final Iteration: Quantitative Pharmacophore Development for hMATE1” in Results for inclusion criteria). The model included two hydrophobes (cyan), a hydrogen bond acceptor (magenta), and an ionizable feature (red). F, the relationship between measured and predicted IC₅₀ values based on the N46 model (r = 0.73; p < 0.0001). Quinidine is mapped to all pharmacophores.

Fig. 9.

A, FCFP_6 features associated with hMATE1 inhibitors–pH 8.5 N46 model. B, FCFP_6 features associated with hMATE1 noninhibitors–pH 8.5 N46 model. Each panel shows the naming convention for one fragment, the numbers of compounds containing the fragment, and the Bayesian score for the fragment.

Fig. 10.

hMATE1 quantitative pharmacophores based on the inhibition of MATE1-mediated transport of ASP (A) or MPP (B). The ASP-derived pharmacophore (with ondansetron) had an r = 0.95 and was generated by using a data set obtained from Kido et al. (2011) that consisted of six compounds with IC₅₀ values ranging from 0.15 to 66 μM. The MPP-derived N46 pharmacophore (with quinidine) was developed as described in “Computational Modeling” in Materials and Methods, and in Fig. 8E for the other pharmacophores and in Fig. 8E. Pharmacophore features are as described in Fig. 8 (with the addition of gray features indicating excluded volumes).