Identification of novel C-15 fluoro isosteviol derivatives for GABA-AT inhibition by in silico investigations

Context The treatment of epilepsy is associated with the inhibition of γ-aminobutyric acid-aminotransferase (GABA-AT), which suppresses the concentration of a key neurotransmitter GABA. Isosteviol, a natural bioactive molecule, has been reported to possess an anticonvulsant property. In this work, we first reported a series of C-15 fluoro isosteviol analogs which are bearing different functional groups at C-16 to investigate the interactions with GABA-AT by applying molecular docking and molecular dynamic simulation approach. The results revealed that all fluoro isosteviol analogs displayed a greater binding affinity than references vigabatrin, an FDA-approved GABA-AT inactivator, and CPP-115, which has Orphan Drug Designation status, and positioned at the same binding site as references. Furthermore, molecular dynamic (MD) simulation studies on minimum (A1), maximum (E1) binding energy score of fluoro isosteviol analogs, and isosteviol (G1) revealed their stable complex formation in terms of RMSD, RMSF, RG, and hydrogen bond formation. All analogs were found to have drug-like nature, non-toxic, >80% absorption, and the majority tend to penetrate brain-blood-barrier (BBB). The investigations found in this study can help in the development of isosteviol derivatives as drugs for the treatment of epilepsy. Methods The two-dimensional (2D) ligand structures were drawn using ChembioDraw Ultra 14.0. Molecular docking with Autodock4 and molecular dynamic simulation with GROMACS version 2020.1 were performed. The CHARMM27 all-atom force field was applied for writing the topology. Biovia Discovery Studio DS2021 was used for viewing and analyzing the protein-ligand complexes. The data generated from molecular dynamic simulation trajectories were plotted using the Origin® 8 software. The Open Babel software was utilized for extracting SMILEs files of all the fluoro isosteviol analogs. The drug-likeness and ADMET of the molecules were evaluated by SwissADME and ADMETlab 2.0 web tools.