Homology modelling, virtual screening and molecular dynamics of the MARK3 KA1 domain for cancer drug design

Structural homology modelling was done with the software AMPS, MODELLER, PROCHECK, WHATIF AND VERIFY-3D to generate a quality model of human MARK3. Macromolecular docking simulations seem to confirm recent data in the literature and in MARK3 there does not occur intramolecular interactions between the associated kinase domain KA1 and the catalytic domain. Using virtual screening, we were able to identify and suggest the principal residues of MARK3 which interact with the ligands in addition to those reported in the literature. The pharmacophoric model obtained from Discovery Studio coincides with those obtained by molecular interaction fields, indicating the principal ligand residues of the MARK3 KA1 domain. Using virtual screening with pharmacophoric constraints as well as molecular dynamics, the most stable compounds in the ligand site as well as their potential toxicities were used to select potential inhibitors for further in vitro and in vivo investigations of human MARK3 KA1 domain, which could eventually pass to the market to be used for the treatment of head and neck cancer.