Combined ligand and structure-based approaches on HIV-1 integrase strand transfer inhibitors

HIV-1 integrase (IN) is an essential enzyme in the viral replication cycle and represents a promising target for anti-HIV drug design. In the present study, pharmacophore modeling and atom-based 3D-QSAR studies were carried out on a series of compounds belonging to dihydroxy isoindole derivatives as HIV-1 IN strand transfer inhibitors. The best pharmacophore model generated consists of six features AADHRR: two hydrogen bond acceptors (A), a hydrogen bond donor (D), a hydrophobic group (H) and two aromatic rings (R). Based on the best pharmacophore model, a statistically valid atom-based 3D-QSAR model was developed. The obtained atom-based 3D-QSAR model has an excellent correlation coefficient value (R(2)=0.87) and also exhibited good predictive power (Q(2)=0.72). The best pharmacophore model was further validated through enrichment calculations and it shows strong predictive power with a high performance in identifying active ligands from the total hits (actives+decoys). QM-polarized ligand docking and molecular dynamics simulations of selected active compounds in the active site of prototype foamy virus intasome gave important insights into the chemical and structural basis involved in the molecular recognition process. The O,O,O donor atom triad of compounds show metal chelation with divalent Mg(2+) ions bound to the three catalytic amino acids in the enzyme's active site and π-stacking interaction with the viral DNA residue DA17. The results might have implications for rational design of specific HIV-1 integrase strand transfer inhibitors with improved affinity and selectivity.