Molecular Modeling Studies to Probe the Binding Hypothesis of Novel Lead Compounds against Multidrug Resistance Protein ABCB1

Abstract

The expression of drug efflux pump ABCB1/P-glycoprotein (P-gp), a transmembrane protein belonging to the ATP-binding cassette superfamily, is a leading cause of multidrug resistance (MDR). We previously curated a dataset of structurally diverse and selective inhibitors of ABCB1 to develop a pharmacophore model that was used to identify four novel compounds, which we showed to be potent and efficacious inhibitors of ABCB1. Here, we dock the inhibitors into a model structure of the human transporter and use molecular dynamics (MD) simulations to report the conformational dynamics of human ABCB1 induced by the binding of the inhibitors. The binding hypotheses are compared to the wider curated dataset and those previously reported in the literature. Protein-ligand interactions and MD simulations are in good agreement and, combined with LipE profiling, statistical and pharmacokinetic analyses, are indicative of potent and selective inhibition of ABCB1.

Keywords: ABCB1; MD simulation; ligand–ABCB1 interactions; molecular docking simulation.

Figure 1

2D structures of the lead compounds ‘A’, ‘D’, ‘E’ and ‘F’ along with their chemical formula, name, molecular weight, clogP, IC₅₀ for inhibition of calcein-AM transport and LipE values [21].

Figure 2

Detailed workflow of the computational methodology adopted to probe the 3D features of ligands of ABCB1. The dataset of 58 inhibitors was docked into the binding pocket of human ABCB1 (6QEX), followed by ligand–protein interaction analysis, SAR-guided pose analysis, PLIF, IF, statistical analyses and molecular dynamics simulation for 100 ns. Finally, pharmacokinetic analyses of the selected inhibitors was performed to identify likely non-toxic and efficacious inhibitors of ABCB1.

Figure 3

Binding solutions of the lead compounds ‘A’ (maroon), ‘D’ (yellow), ‘E’ (cyan) and ‘F’ (orange) within the binding pocket of ABCB1 (6QEX). The interacting residues are shown in stick form. Green dotted lines represent π-stacking, grey dotted lines show hydrophobic interactions and blue dotted lines represent hydrogen bonding.

Figure 4

RMSD vs. time (100 ns) of the simulated ABCB1–lead complexes with protein-RMSD displayed in blue while the RMSD of the ABCB1–lead compounds is displayed in maroon. (1) RMSD of ABCB1 and ‘ABCB1-A’; (2) RMSD of ABCB1 and ‘ABCB1-D’; (3) RMSD of ABCB1 and ‘ABCB1-E’; (4) RMSD of ABCB1 and ‘ABCB1-F’ complexes.

Figure 5

RMSF for the simulated ABCB1–ligand complexes for 100 ns. (1) RMSF of ‘ABCB1-A’; (2) RMSF of ‘ABCB1-D’; (3) RMSF of ‘ABCB1-E’; (4) RMSF of ‘ABCB1-F’. Residues that interact with the compounds are marked with green bars.

Figure 6

Comparison of ligand–protein interactions of lead compounds complexed with ABCB1. (1–4) Interaction pattern of leads ‘A’, ‘D’, ‘E’ and ‘F’, respectively in the binding pocket of ABCB1 immersed in a full membrane-aqueous environment at 0 ns. The interacting residues along with the lead compounds are shown in the stick form and hydrophobic, hydrogen bonding and π-stacking interactions are depicted by grey, blue and green dotted lines, respectively. (5–8) Interaction pattern of lead ‘A’, ‘D’, ‘E’ and ‘F’, respectively in the ABCB1-binding pocket, which was entrenched in a complete membrane-aqueous system after 100 ns.

Figure 7

Binding solutions of the derivatives of 6,7-dimethoxy-2-phenethyl-1,2,3,4-tetrahydroisoquinoline (orange), galloyl (raspberry), and propafenone (cyan) within the binding pocket of human ABCB1 (6QEX). The interacting amino acids are displayed in stick form (blue).

Figure 8

(A) Protein–ligand interaction pattern of binding solutions of the derivatives of 6,7-dimethoxy-2-phenethyl-1,2,3,4-tetrahydroisoquinoline and 6QEX. (B) Binding solutions of the derivatives of the galloyl showing interactions in the binding pocket of ABCB1. (C) Binding solutions of the propafenone derivatives showing interactions with ABCB1.

Figure 9

Selective inhibitors of ABCB1 and potential leads are represented as points in 3D space mapped according to activity values, molecular weight and clogP values (points are color-coded according to the PIC₅₀ values). The ellipse shows our lead compounds (red color of leads is for the representation of LipE > 5).