Homology modeling and substrate binding studies of human P-glycoprotein

Multidrug-resistance (MDR) in cancer cells often relates to the overexpression of P-glycoprotein (P-gp), resulting in increased efflux of chemotherapy from cancer cells. A clear understanding of P-gp substrate binding will lead to the development of selective P-gp inhibitors, which resensitize cancer cells to standard chemotherapy. Unfortunately, the three-dimensional structure of human P-glycoprotein has not yet been available. In this investigation, homology model of human P-gp was developed based on a recent refined structure of mouse P-gp (PDB: 4M1M). The models were further assessed by Ramachandran plot in PROCHECK, ProSA-web Z-score and QMEAN score. The results indicated that the proposed models were reliable for further binding site and docking studies. Using AutoDock-based blind docking protocol, the probable binding sites for the known substrates rhodamine B, daunorubicin, colchicine, and Hoechst 33342 were identified and in very good agreement with the available side-directed mutagenesis studies. The binding location of the cytotoxic drug vinblastine was identified and characterized. The docking result indicated that vinblastin and verapamil shared overlapping sites on P-gp, composed of residue Leu65, Met69, Ile340, Phe983, Tyr953, and Met986. This might aid in understanding how verapamil, an inhibitor of P-gp, effectively enhanced cytotoxicity of vinblastine against P-gp-mediated MDR. Our observations suggested that human P-gp model derived from 4M1M could better explain the binding of human P-gp substrates/ inhibitors. This model served as a starting point to gain knowledge of P-gp drug-binding region(s) and to identify novel P-gp inhibitors that might have a potential to overcome MDR in cancer therapy. Keyword: P-glycoprotein, Homology model, Molecular docking, Drug binging sites, Interaction.