Study of binding of inhibitor molecules to the active site of protein kinases by method of molecular dynamics

1 - Pyatigorsk Medical and Pharmaceutical Institute - branch of the Volgograd State Medical University of the Ministry of Health of Russia; 2 - National Medical Research Centre for Oncology Russia, Rostov-on-Don

Eukaryotic protein kinases are represented by a large superfamily of homologous proteins similar in structure of the catalytic (kinase) domain, consisting of 250–300 amino acids [3]. Most human protein kinases share a common fold consisting of an N-terminal lobe, consisting of a five-stranded β-sheet with an α-helix called the C-helix, and a C-terminal lobe comprising six α-helices. The active center is located between the two lobes and forms binding sites for adenosine triphosphate (ATP) and protein [4]. Protein kinases modify the functions of other proteins by phosphorylation, acting as elements of signaling pathways. Disorders and abnormalities in the activity of these enzymes give rise to the pathogenesis of many malignant neoplasms.

Based on this fact, the targeted synthesis of some substances capable of selectively modifying the activity of protein kinases is an urgent task. The study of the process of interaction of molecules with the catalytic center of protein kinase by the method of molecular dynamics makes it possible to search for chemical structures of substances capable of inhibiting this enzyme [2]. Taking into account the high similarity of the structure of the catalytic domains of various human protein kinases, it is of scientific interest to develop a technique for modeling the molecular dynamics of the ligand-enzyme system, which allows predicting the spectrum of inhibitory activity against a set of protein kinases.

Using the method of molecular dynamics, the process of binding of inhibitor molecules to the active sites of phosphoinositide-dependent protein kinase-1 (PDPK1), mitogen-activated protein kinase-1 (MAPK1), and phosphoinositide-3-kinase (PI3K) was simulated. The PI3K kinase is not a protein kinase, however, this object is of interest in this study, because the substrate of the enzyme is also an ATP molecule. The rectangular modeling area included an active center of the enzyme and a region of space with water molecules and a ligand molecule. Modeling by the method of molecular dynamics was carried out with the Bioeureka software [1]. We used the quantum-chemical ab initio method, the density functional theory (ub3lyp) and the 6-311G** basis set to preliminarily calculate the charges of atoms in the ligand molecules. To increase the efficiency of scanning the conformational space, the approach of accelerated molecular dynamics was applied [5]. Temperature was controlled out separately for water, protein, and ligand molecules. The simulation was conducted for 500 ns with a step of 2 fs. During the modeling of molecular dynamics, the binding of ligands to amino acids of the active center in various conformations was observed. For each of the targets, a ligand was selected with a known location in the active center according to X-ray diffraction analysis (XRD).

The location of those ligands in the enzyme according to the simulation data was compared with the location according to the X-ray diffraction data by calculating the root-mean-square deviation (RMSD) of the atomic coordinates. The parameters of the modeling technique were chosen so that the RMSD value for individual conformations of the ligand was less than 1 nm. When the inhibitor was bound to the active site, the average value of RMSD was estimated. When the staurosporine molecule was bound to the PDPK1 active center, the RMSD value was 0.309 ± 0.028 nm, and the minimum was 0.236 nm. For the binding of the ulixertinib molecule to the MAPK1 active site, the RMSD was 0.468±0.163 nm, and the minimum was 0.178 nm. The average value of RMSD for the binding of the PQR530 inhibitor molecule to the PI3K active site was 0.973±0.253 nm, and the minimum value thereof was 0.496 nm. Thus, the developed technique allowed modeling the spatial arrangement of ligand molecules in the active center of the kinases selected for the study according to experimental XRD data.

This modeling technique can be successfully used to explore the spectrum of inhibitory activity of substances against protein kinases and design the structures of new targeted drugs with antitumor activity.