Combined Structure and Ligand Guided Protocols to Target the Gain-of-function Mutation of Gárdos Channel in Red Cell Deformability /

Abstract

Hereditary Xerocytosis (HX) is an autosomal dominant, hereditary hemolytic anemia characterized by red cells dehydration. In HX patients, the erythrocytes exhibit abnormal intracellular cation content and water loss. These factors cause an elevated erythrocyte mean hemoglobin concentration (MCHC) and decreased erythrocyte osmotic fragility. HX is linked with gain-of-function mutations of the Gárdos Channel (KCNN4 or KCa3.1). Gárdos channel is an intermediate conductance Ca2+ dependent K+ channel. It is Ca2+ sensitive and present in several cell types, including erythrocytes. The Gárdos channel is inactive under steady-state conditions. Under the external stimulus, intracellular Ca2+ increases and binds with Calmodulin molecules that are firmly attached to each of the four-channel subunits of KCNN4. In HX patients, even a ten times lesser concentration of calcium activates the channel and leads towards pathogenicity. The abnormal functioning of the Gárdos channel is due to three different types of mutations i.e. R352H (p.Arg352His); located in the calmodulin-binding domain within the long COOH-terminal cytoplasmic tail, V282M (p.Val282Met), and V282E (p.Val282Glu) in the pore region of the transmembrane. Numerous drugs such as Senicapoc and TRAM-34 have been designed to block the activity of the Gárdos channel, but none of them are completely efficient against all three mutations. Therefore, in the present project, a combined structure and ligand-guided protocol was opted to explore the binding hypothesis of KNCC4 which could facilitate the design optimization of the KNCC4 modulators. The protocol consists of modeling of mutant KCNN4, molecular dynamic simulation of mutant KCNN4, molecular docking, common scaffold clustering and GRIND analysis for the prediction of 3D features of the inhibitors of Gárdos Channel. The modeling of mutant KCNN4 structure introduced loops in the start (residue 1- 10) and end (residue 386 – 427) of the structure that were not present in the wild KCNN4. Therefore, those loops were spliced out. The docking of mutant KCNN4 with 51 ligands generated different conformations that were clustered using Common Scaffold Clustering. The clusters showed mainly pi-Hydrogen and Hydrogen-acceptor interaction with the Lysine at 309 position of the mutant protein. Moreover, the GRIND analysis generated features that were consistent with the docking output and complementary with the residues in the binding pocket. The features that contributed positively towards the biological activity of KCNN4 were Hydrophobic XIII – Hydrogen bond acceptor feature at a distance 9.60 – 10.00, Hydrogen – bond acceptor and Hydrogen bond donor feature at a distance of 8.40 – 8.80, Hydrogen- bond acceptor and shape based feature at a distance of 10.40 – 10.80, and Hydrogen- bond donor and shape based feature at a distance of 10.40 – 10.80 in the virtual receptor site. These features were complementary with the Glutamine353, Methionine302, Lysine360, Glutamine306 and Arginine359 residues in the actual receptor site. The predicted 3D features of the Gárdos Channel can help predict new drug-like compounds with optimized molecular properties.