Modeling the binding of diverse ligands within the Ah receptor ligand binding domain

Abstract

The Ah receptor (AhR) is a ligand-dependent transcription factor belonging to the basic helix-loop-helix Per-Arnt-Sim (bHLH-PAS) superfamily. Binding to and activation of the AhR by a variety of chemicals results in the induction of expression of diverse genes and production of a broad spectrum of biological and toxic effects. The AhR also plays important roles in several physiological responses, which has led it to become a novel target for the development of therapeutic drugs. Differences in the interactions of various ligands within the AhR ligand binding domain (LBD) may contribute to differential modulation of AhR functionality. We combined computational and experimental analyses to investigate the binding modes of a group of chemicals representative of major classes of AhR ligands. On the basis of a novel computational approach for molecular docking to the homology model of the AhR LBD that includes the receptor flexibility, we predicted specific residues within the AhR binding cavity that play a critical role in binding of three distinct groups of chemicals. The prediction was validated by site-directed mutagenesis and evaluation of the relative ligand binding affinities for the mutant AhRs. These results provide an avenue for understanding ligand modulation of the AhR functionality and for rational drug design.

Figure 1

2D representation of the set of ligands analyzed in this study and the homology models of the mouse AhR (mAhR) PAS-B domain developed here. (a) Structures of AhR ligands used in this study. (b) Cartoon representation of all the mAhR PAS-B models superimposed. The region refined by loop modeling is colored according to the model. Secondary structure elements are labeled according to the PAS domain nomenclature.

Figure 2

Correlation between the relative affinity of each test compound to bind to the mAhR and their relative potency to stimulate AhR DNA binding (Pearson correlation coefficient R² = 0.56). The relative affinity (LogIC₅₀) and potency (LogEC₅₀) of the indicated test compounds was determined by [³H]TCDD competitive ligand binding and ligand-dependent DNA binding analysis as described under Materials and Methods.

Figure 3

Internal surfaces of the binding cavities for the HIF-2α templates (gray and transparent with the ligand inside) and for the mAhR models (colored and solid). Models are ordered by increasing cavity volumes (Table S3).

Figure 4

Dynamic view of a TCDD pose inside the binding cavity; ligand is shown as sticks and protein as gray cartoons. In blue, the pose previously obtained by docking; in dark green, one of the docking poses obtained in this work; in green, 10 snapshots taken from the last 8 ns of the MD simulation.

Figure 5

Ligands are gathered into three groups that were defined by the occupancy of different sites inside the AhR cavity and by characteristic residue interactions during MD simulations. (a) Per-residue decomposition of ΔG_bind. Values were obtained as averages of the last 8 ns of simulation. In the plots, only residues lining the internal cavity are shown. (b) Ten snapshots sampled during the last 8 ns of simulation are shown for each ligand belonging to the three groups; ligands are represented as sticks.

Figure 6

Dynamic view of the binding poses of ligands in group 1, (a) TCDD, (b) TCDF, (c) BaP and group 2, (d) 3MC, (e) PCB126, (f) DBA. Ligands are represented as sticks and 10 snapshots extracted during the last 8 ns of MD simulation are shown in transparency; solid sticks indicate the most sampled poses in the MD simulations. The most relevant residues identified by per-residue decomposition of ΔG_bind are shown as lines and labeled.

Figure 7

Dynamic view of the binding poses of ligands in group 3: (a) BNF, (b) FICZ, (c) IR, and (d) LEFL. Ligands are represented as sticks and 10 snapshots extracted during the last 8 ns of the MD simulation are shown in transparency; solid sticks are the most sampled poses in the MD simulations. The most relevant residues identified by per-residue decomposition of ΔG_bind are shown as lines and labeled.

Figure 8

Relative binding affinity for group 1, 2 and 3 ligands relative to wild-type and mutant AhRs. The relative affinity (−logIC₅₀) of each test chemical for the AhR ligand was determined from concentration-dependent inhibition curves obtained using [³H]TCDD ligand binding analysis, as described under Materials and Methods. The mean IC₅₀ ± standard deviation was determined using three-parameter non-linear regression of nine independent reactions. (*) Represents a significant (p ≤ 0.05) change in ligand binding affinity relative to wild-type mAhR in One-Way ANOVA Multiple Comparison Test.