Molecular Modeling Studies of N-phenylpyrimidine-4-amine Derivatives for Inhibiting FMS-like Tyrosine Kinase-3

Abstract

Overexpression and frequent mutations in FMS-like tyrosine kinase-3 (FLT3) are considered risk factors for severe acute myeloid leukemia (AML). Hyperactive FLT3 induces premature activation of multiple intracellular signaling pathways, resulting in cell proliferation and anti-apoptosis. We conducted the computational modeling studies of 40 pyrimidine-4,6-diamine-based compounds by integrating docking, molecular dynamics, and three-dimensional structure–activity relationship (3D-QSAR). Molecular docking showed that K644, C694, F691, E692, N701, D829, and F830 are critical residues for the binding of ligands at the hydrophobic active site. Molecular dynamics (MD), together with Molecular Mechanics Poison–Boltzmann/Generalized Born Surface Area, i.e., MM-PB(GB)SA, and linear interaction energy (LIE) estimation, provided critical information on the stability and binding affinity of the selected docked compounds. The MD study suggested that the mutation in the gatekeeper residue F691 exhibited a lower binding affinity to the ligand. Although, the mutation in D835 in the activation loop did not exhibit any significant change in the binding energy to the most active compound. We developed the ligand-based comparative molecular field analysis (CoMFA) and comparative molecular similarity index analysis (CoMSIA) models. CoMFA (q2 = 0.802, r2 = 0.983, and QF32 = 0.698) and CoMSIA (q2 = 0.725, r2 = 0.965 and QF32 = 0.668) established the structure–activity relationship (SAR) and showed a reasonable external predictive power. The contour maps from the CoMFA and CoMSIA models could explain valuable information about the favorable and unfavorable positions for chemical group substitution, which can increase or decrease the inhibitory activity of the compounds. In addition, we designed 30 novel compounds, and their predicted pIC50 values were assessed with the CoMSIA model, followed by the assessment of their physicochemical properties, bioavailability, and free energy calculation. The overall outcome could provide valuable information for designing and synthesizing more potent FLT3 inhibitors.

Structure–activity relationship and in silico development of c‐Met kinase inhibitors

“…The cutoff and Particle mesh Ewald (PME) schemes were used to handle the van der Waals (vdW) and coulombic interactions. Additional methodological details can be found in our previously studies 33–34 . The built‐in “ gmx rms ” and “ gmx hbond ” functionalities were used to calculate RMSD and h‐bond distances.…”

“…We have calculated χ 2 , RMSE, and MAE for the internal validation of the model. For external validation, we have calculated k , k ', | r 0 2 ‐ r ' 0 2 |, ( r 2 ‐ r 0 2 )/ r 2 , r m, 2 , Δ r m 2 , r 2 pred , , , , and matric according to our previous study 34 . In addition, the progressive scrambling stability test was used to evaluate the sensitivity and robustness of the developed models.…”

“…Additional methodological details can be found in our previously studies. [33][34] The built-in "gmx rms" and "gmx hbond" functionalities were used to calculate RMSD and h-bond distances.…”

“…, and Q 2 ccc matric according to our previous study. 34 In addition, the progressive scrambling stability test was used to evaluate the sensitivity and robustness of the developed models. As described in this research, 39 we performed the applicability domain (AD) analysis using the distance-based Williams plot.…”

Binding Studies and Lead Generation of Pteridin-7(8H)-one Derivatives Targeting FLT3

“…Computational drug design is a popular choice for discovering small molecules targeting kinase receptors. In our previous study [ 12 ], we performed the molecular modeling of pyrimidine4,6-diamine derivatives against inactive FLT3 as type II inhibitors. In this paper, we conduct the modeling study of 35 pteridin-7(8 H )-one compounds as type I inhibitors targeting the active FLT3 conformer.…”