Designing and optimization of novel human LMTK3 inhibitors against breast cancer - a computational approach.

Estrogen receptor-α (ERα) is expressed more in patients with breast cancer and its level correlated with endocrine resistance. LMTK3 is reported as breast cancer target with regulation of estrogen receptor-α (ERα) through phosphorylation. In this computational study, structure-based inhibitor screening was performed on human LMTK3 using ZINC database. ATP-binding cavity with critical residues involved in the LMTK3 phosphorylation was used as target site for the screening. From the large ligand library, the best compounds were screen with three-phase virtual screening methods in Dockblaster, AutoDock Vina and AutoDock, respectively. The evaluation of ligands was carried out by binding energy and weak interactions, such as hydrogen bond interactions and hydrophobic contacts, in the target site that favors LMTK3 inhibition. Top compounds were found to be more effective in druglikeness activity by ADME prediction. The stability and binding affinity of ligand complexes were optimized by trajectory analysis such as RMSD, Rg, SASA and interhydrogen bonds from molecular dynamics simulations. The behavior of protein motion after ligand binding was illustrated by eigenvectors from principal component analysis (PCA). In addition, binding free energy of the LMTK3-ligand complexes were calculated by MM/PBSA methods and results supported the strong binding in dynamic system. Thus, the computational studies illustrated the structural insights on LMTK3 inhibition mechanism by ligands ZINC04670539, ZINC05607079 and ZINC04344028, also proposed as potent lead candidates. Our findings step towards developing novel LMTK3 inhibitors and identified lead candidates can be future breast cancer drugs with further experimental studies.