Combined Ligand and Structure Guided Design and Virtual Screening of Modulators of Gardos Channel

Abstract

Red Cell deformability refers to the ability of red blood cells (erythrocytes) to change their shape to pass through narrow capillaries and small vessels in the circulatory system. In red cell deformability, an abnormality can result in impaired blood flow and decreased oxygen delivery, which can contribute to a variety of diseases such as Sickle cell anemia and hereditary xerocytosis. The Gardos Channel, also known as the calcium-activated potassium Channel (KCNN4/KCa3.1), is found in the membrane of red blood cells. In diseased conditions, hyperactivation of the Gardos Channel allows potassium ions to exit the cell, which leads to water loss and reduces the volume of the cell. In previous studies, Gardos Channel has been identified as a potential drug target because of its role in regulating red Cell volume. Inhibition of the Gardos Channel has been shown to increase the hydration of cells and maintain their normal shape which can be beneficial for diseases of red cell deformability. Previously, a structure- guided protocol was opted to explore the binding hypothesis of Gardos Channel (KNCC4) which could facilitate the design of the KNCC4 modulators. Three-point mutations R352H, V282M and V282E have been identified previously in Gardos Channel which subsequently led to the development of Senicapoc, Clotrimazole, Tram-34, UCL 1684 and 1- EBIO drugs. However, none of these drugs showed efficacy against reported mutations. Therefore, in this study, a combined ligand and structure-guided protocol was opted for virtual screening of modulators of Gardos Channel. The protocol consists of building ligand-based pharmacophore models, molecular docking, model evaluation, virtual screening, GRIND analysis, docking of hits and Molecular dynamic simulation. The data of known ligands of Gardos Channel and the mutant model of Gardos Channel was retrieved from the literature. Two types of pharmacophore models were developed in this study. The docking of 50 known ligands was done at the Calmodulin binding domain of Gardos Channel, generating an internal database for model evaluation. Virtual screening of unknown drug data was performed using the final selected pharmacophore model. The GRIND analysis was used for the prediction of IC50 values of hits. The predicted IC50 values for the top five hits fall within the range of 5.85nM to 12.67nM. The top five compounds with the highest IC50 values were docked at the Calmodulin binding domain, and a Molecular Dynamic simulation of the top pose of the highest active compound was carried out to probe the stability. The molecular dynamics simulations result showed the presence of residues MET282, ALA297, ARG287, TYR199, LEU198 and VAL195 within vicinity of ligand. The presence of MET282 and ALA297 is important as they formed a hydrophobic pocket, thus providing optimal fit of ligand and MET282 is one of our target mutations. This compound exhibits superior interactions with the 13 target, as no previously identified ligand has demonstrated interactions with the specific point mutation. The predicted IC50 of this compound is 5.85nM while the IC50 of the top active compound from known data is 8.99nM, therefore it has better binding potential against the Gardos channel than the already known compound. The Lipophilic efficiency value of this compound is 7.47 which qualifies the FDA criteria of an average oral drug.