Predictions of Ligand Selectivity from Absolute Binding Free Energy Calculations

Abstract

Binding selectivity is a requirement for the development of a safe drug, and it is a critical property for chemical probes used in preclinical target validation. Engineering selectivity adds considerable complexity to the rational design of new drugs, as it involves the optimization of multiple binding affinities. Computationally, the prediction of binding selectivity is a challenge, and generally applicable methodologies are still not available to the computational and medicinal chemistry communities. Absolute binding free energy calculations based on alchemical pathways provide a rigorous framework for affinity predictions and could thus offer a general approach to the problem. We evaluated the performance of free energy calculations based on molecular dynamics for the prediction of selectivity by estimating the affinity profile of three bromodomain inhibitors across multiple bromodomain families, and by comparing the results to isothermal titration calorimetry data. Two case studies were considered. In the first one, the affinities of two similar ligands for seven bromodomains were calculated and returned excellent agreement with experiment (mean unsigned error of 0.81 kcal/mol and Pearson correlation of 0.75). In this test case, we also show how the preferred binding orientation of a ligand for different proteins can be estimated via free energy calculations. In the second case, the affinities of a broad-spectrum inhibitor for 22 bromodomains were calculated and returned a more modest accuracy (mean unsigned error of 1.76 kcal/mol and Pearson correlation of 0.48); however, the reparametrization of a sulfonamide moiety improved the agreement with experiment.

Figure 1

Ligands and proteins considered in this study. (a) Phylogenetic tree of the human bromodomain family; in green are the BRDs included in the study. (b) Conserved fold of human bromodomains (PDB ID 2OSS). Depicted as red spheres is the conserved water network found in most BRD binding pockets, and as blue sticks the conserved hydrogen bond donor, an asparagine residue in the majority of BRDs. A transparent surface shows the cavity in between the ZA and BC loops forming the acetyl-lysine binding pocket. Two clefts, placed in between the WP residues (conserved in subfamilies I and II) and the ZA and BC loops and exploited by many bromodomain inhibitors are indicated. (c) Chemical structures of the compounds.

Figure 2

Comparison of the X-ray structures of RVX-208 and RVX-OH in the first (BD1) and second (BD2) BET bromodomains binding pockets, and the corresponding docked structures. RVX-208 binds with the same pose to both BD1 and BD2, while RVX-OH adopts two distinct binding orientations. The two ligands have been docked into an apo structure of BRD4(1) (PDB ID 2OSS), and the docking poses best representing the crystal structures are shown. These poses are taken from previous work. BD1s are represented by white cartoons, and BD2s by gray cartoons; RVX-208 is highlighted in light blue, and RVX-OH in light green; the conserved Asn residue and network of waters are shown as reference points.

Figure 3

Scatter plot of calculated versus experimental affinities for the compounds RVX-208 and RVX-OH. The shaded gray areas indicate where the 1 and 2 kcal/mol error boundaries lie. On the right-hand side are the distributions of the calculated binding free energies for the two ligands, binding to BD1s (light blue and light green) and BD2s (dark blue and dark green); Gaussian curves have been fitted to the data.

Figure 4

Docking poses for bromosporine in the binding pocket of BRD4(1). (a) Comparison of the pose with highest predicted affinity to the X-ray structure of bromosporine in complex with BRPF1B (PDB-ID 5C7N). (b–e) The five diverse bromosporine poses suggested by docking, with their predicted binding affinity and RMSD with respect to the binding mode found in the crystal structure.

Figure 5

Scatter plots of calculated versus experimental binding free energies for bromosporine. (a) Calculated binding free energies for the initial bromosporine model. (b) Calculated binding free energies for the bromosporine model with optimized benzensulfonamide torsions. The shaded gray areas indicate where the 1 and 2 kcal/mol error boundaries lie.