Biochemical Mechanisms of Phosphodiesterase 6D Inhibition in K-Ras-Driven Cancers.

Targeting K-Ras-driven oncogenesis remains a clinical challenge due to its elusive structure and high affinity for GTP and GDP. Phosphodiesterase 6D (PDE6D), a trafficking chaperone for K-Ras membrane localization, is an actionable indirect therapeutic target. In this study, a large-scale computational strategy was employed to identify novel PDE6D inhibitors with a similar structure to the control compound, Deltasonamide 1. Virtual screening of 11,221 PubChem compounds (≥ 80% Tanimoto similarity) resulted in the identification of three promising candidates: 135,235,500, 153,283,984, and 156,588,640, which exhibited improved docking scores. Machine learning-based regression models predicted high pIC50 values (7.55, 7.52, and 7.47) for these compounds, which were more significant than those of the control, indicating strong inhibitory activity. DFT analysis revealed distinct electronic properties: 156,588,640 had the smallest HOMO–LUMO gap (0.368 eV), indicating higher reactivity, while 135,235,500 and 153,283,984 exhibited better stability. Molecular redocking confirmed the strong and stable interactions of these compounds within the PDE6D binding pocket, involving important residues Arg61, Gln78, Glu88, and Tyr149. Molecular dynamics simulations (1 µs) revealed that all three compounds maintained structural stability in the active site, particularly compound 135,235,500, which exhibited the lowest RMSD values (approximately 0.25 nm). MM/GBSA analysis validated high binding affinities for all candidates, with 135,235,500 exhibiting a ΔG_total of -100.53 kcal/mol. Network pharmacology analysis revealed a higher gene-targeting potential, reaffirming the therapeutic importance of the compounds. Together, these findings recognize 135,235,500, 153,283,984, and 156,588,640 as potent inhibitors of PDE6D, warranting additional preclinical verification for the treatment of K-Ras-driven cancer.