Abstract
The past two decades have witnessed major breakthroughs in developing compounds that target the chloride channel CFTR for the treatment of patients with cystic fibrosis. However, further improvement in affinity and efficacy for these CFTR modulators will require insights into the molecular interactions between CFTR modulators and their binding targets. In this study, we use in silico molecular docking to identify potential binding sites for GLPG1837, a CFTR potentiator that may share a common mechanism and binding site with VX-770, the FDA-approved drug for patients carrying mutations with gating defects. Among the five binding sites predicted by docking, the two top-scoring sites are located at the interface between CFTR's two transmembrane domains: site I consists of D924, N1138, and S1141, and site IIN includes F229, F236, Y304, F312, and F931. Using mutagenesis to probe the importance of these sites for GLPG187 binding, we find that disruption of predicted hydrogen-bonding interactions by mutation of D924 decreases apparent affinity, while hydrophobic amino acids substitutions at N1138 and introduction of positively charged amino acids at S1141 improve the apparent affinity for GLPG1837. Alanine substitutions at Y304, F312, and F931 (site IIN) decrease the affinity for GLPG1837, whereas alanine substitutions at F229 and F236 (also site IIN), or at residues in the other three lower-scoring sites, have little effect. In addition, current relaxation analysis to assess the apparent dissociation rate of VX-770 yields results consistent with the dose-response experiments for GLPG8137, with the dissociation rate of VX-770 accelerated by D924N, F236A, Y304A, and F312A, but decelerated by N1138L and S1141K mutations. Collectively, these data identify two potential binding sites for GLPG1837 and VX-770 in CFTR. We discuss the pros and cons of evidence for these two loci and the implications for future drug design.