Validation of KCa3.1 Channel Small Molecules Interaction Sites Predicted by Rosetta

Abstract

The calcium-activated intermediate-conductance KCa3.1 channel is an effective regulator of intracellular calcium and an attractive pharmacological target for immunosuppression, fibroproliferative disorders, hypertension and various neurological diseases. However, the development of drugs for this medically relevant channel is hindered by the unavailability of a crystal structure useful for structure-guided drug design. Using the Rosetta modeling method we generated a homology model of the KCa3.1 channel transmembrane region using the Kv1.2-Kv2.1 channel structure (pdb id: 2R9R) as a template. Docking of known KCa3.1 small molecule blockers into the KCa3.1 model, identified, surprisingly, two independent sites for the dihydropyridine nifedipine and for its isoster methyl-5-acetyl-4-(4-chloro-3-(trifluoromethyl)phenyl)-2,6-dimethyl-4H-pyran-3-carboxylate. While nifedipine is predicted to bind in the fenestration between the pore-lining S5 and S6 segments, the pyran isoster has its lowest energy binding configurations in the inner pore region. We validated these predictions via site-directed mutagenesis and patch-clamp recording. Blocking of KCa3.1 by nifedipine was significantly reduced by replacing the side chains at the fenestration positions L209, T212 and V272 with either alanine or valine. Replacement with bulkier phenylalanines confirmed T212 and V272 as the main interacting sites for nifedipine without compromising the affinity of inner pore blockers like TRAM-34 or the 4-phenyl pyran. In contrast, the inner pore T250S and V275A mutants, which are known to nullify TRAM-34 binding disrupted binding of the 4-phenyl pyran but did not interfere with nifedipine binding. We conclude that the Rosetta modeling approach can be useful to distinguish the molecular mechanisms of action of KCa channel modulators and has promising potential in guiding the development of clinically relevant drugs.