Computational Analysis of Phase I Metabolism in Clinically Studied Flavoalkaloids: Molecular Docking, MMGBSA Binding Energy Calculations, and Molecular Dynamics Simulations

Abstract Flavoalkaloids, a potent class of cyclin‐dependent kinase inhibitors, have undergone extensive clinical trials across various cancer types. Among them, flavopiridol, riviciclib, and voruciclib are the most advanced candidates. These molecules are lipophilic and possess functional groups conducive to both phase I and phase II metabolism in the liver. This study employs in‐silico methods to identify the potential cytochrome P450‐mediated sites of phase I metabolism (SOM) in these compounds. According to two web‐server‐based predictions, the N ‐methyl group was identified as the primary SOM in these compounds, showing the highest probability score for involvement of CYP3A4. Docking studies conducted using both Glide and the induced‐fit method further confirmed N ‐demethylation as a significant metabolic pathway predominantly catalyzed by CYP3A4. The distance between the N ‐methyl group of the docked ligands and the iron atom of the heme ranged from 3 to 7 Å, with flavopiridol's N ‐methyl group positioned closest to the heme iron at 3.16 Å. The 100 ns molecular dynamic simulations additionally validated the stability of the interaction between flavoalkaloids and the heme Fe. Similarly, the docking‐based SOM predictions matched the experimental SOM for other approved drugs. Thus, employing docking‐based SOM predictions holds potential for enhancing lead optimization efforts in future medicinal chemistry research.