Identification of the Potential of Plant-Derived Agents to Act as KRAS G12D Inhibitors Through Molecular Docking, Molecular Dynamics Simulations, and MMGBSA.
INTRODUCTION/OBJECTIVE The KRAS G12D mutation is a significant oncogenic factor that promotes cell proliferation and tumor development, particularly in pancreatic cancer. Inhibiting the mutated form of the KRAS protein is a critical strategy in cancer treatment. While molecules like MRTX1133 show promise, the limited efficacy and safety of current treatment options highlight the need to discover new and safer therapeutic candidates. METHODS Approximately 1000 plant-derived bioactive compounds downloaded from the literature and Dr. Duke's database were filtered according to Lipinski's Rules and pharmacokinetic properties to create a set of 425 candidate molecules. Geometry optimization and QSAR parameters were calculated using Spartan software, and molecular docking and binding affinities were calculated using AutoDock Vina. MRTX1133 was used as a reference in the evaluations. Protein-ligand interaction maps were generated, and key interaction residues were identified using BIOVIA Discovery Studio. Molecular dynamics simulations of 200 ns were performed using Schrödinger-Desmond for the six compounds with the best binding affinities, and MMGBSA binding free energy calculations were performed. RESULTS The results showed that Isochlorogenic acid, Coniferin, Neochlorogenic acid, Cryptochlorogenic acid, Gamma-L-glutamyl-L-phenylalanine, and Chlorogenic acid demonstrated good bonding affinity to KRAS G12D with -11.0, -9.8, -9.5, -9.1, -9, and -8.9 Kcal.mol-1, respectively, compared to MRTX1133 (-8.3 Kcal.mol-1), exhibiting favorable physicochemical profiles and revealing their potential as KRAS G12D inhibitors. Based on bonding scores and QSAR data, four of the six selected molecules (Coniferin, Gamma-Lglutamyl- L-phenylalanine, Isochlorogenic acid, and Neochlorogenic acid) formed stable protein-ligand complexes with mean RMSD values < 2.5 Å during 200 ns MD simulations. MMGBSA analysis confirmed the binding free energies, and the compound giving the strongest interaction was determined to be Coniferin with ΔG_bind = -72.53 ± 5.24 Kcal.mol-1. Superior docking scores and stable MD trajectories demonstrate that the selected plant-derived compounds, especially Coniferin, effectively interact with the KRAS G12D binding pocket via suitable hydrophobic and hydrogen-bond contacts. These findings suggest that natural compounds can provide structurally diverse and potentially safer alternatives compared to synthetic inhibitors; however, experimental validation is necessary to confirm their inhibitory potential and selectivity. DISCUSSION According to the MMGBSA results, these compounds have suitable binding free energy values comparable to the reference, with Coniferin as the strongest plant-derived candidate with ΔGbind = -72.53 ± 5.24 Kcal/mol-. CONCLUSION As a result of this in silico study, coniferin, neochlorogenic acid, and chlorogenic acid have been identified as potential precursor molecules for KRAS G12D inhibition. Isochlorogenic acid was excluded due to its dynamic instability. These plant-derived compounds require further experimental validation as potential anticancer agents targeting KRAS G12D mutations.