Comparative Computational Screening of Natural-based Partial Agonists for PPARγ Receptor

The nuclear transcription factor PPARγ, which can modulate cell growth via proliferation and apoptosis-related mechanisms, is a promising target in cancer therapy. This study aims to focus on PPARγ as the target and use virtual screening to find hits.

A set of 5,677 flavonoid compounds were filtered by subjecting them to descriptor-based drug-likeness and ADMET strategies to discover drug-like compounds. The candidates' modes of binding to PPARγ were then evaluated using docking and MD simulation. PharmMapper was used to identify the potential targets of selected hits. The pharmacological network was constructed based on the GO and KEGG pathway analysis.

In primary screening, 3,057 compounds met various drug-likeness criteria and docked well as partial agonists in the PPARγ-LBD. Five compounds (euchrenone b₁, kaempferol-7-Orhamnoside, vincetoxicoside B, morusin, and karanjin) were selected with the use of ADMET profiles for further MD simulation investigation. Based on the PharmMapper findings, 52 proteins were then submitted to GO and KEGG enrichment analysis. As expected by GO and KEGG pathway enrichment studies, core targets were enriched in the PI3K-Akt signaling pathway (p < 0.01), indicating that certain chemicals may be involved in cancer processes.

Our results suggested that the selected compounds might have sufficient drug-likeness, pharmacokinetics, and in silico bioactivity by acting as PPARγ partial agonists. Although much work remains to illuminate extensive cancer therapeutic/ chemopreventive efficacy of flavonoids in vivo, in silico methodology of our cheminformatics research may be able to provide additional data regarding the efficacy and safety of potential candidates for therapeutic targets.