Docking study

To rationalize the relationship between antibacterial activity and conformational preference for synthesized quinolones, docking calculations were performed with the Glide module of Mestro 11.5 (Schrödinger LLC, New York) (Friesner et al., 2006). The X-ray structures of DNA gyrase (PDB ID: 2XCT) and topoisomerase IV (PDB ID: 4KPF) with co-crystallized ligands were downloaded from the RCSB Protein Data Bank and preprocessed by fixing missing side-chain atoms, ionizing and tautomerizing het groups, optimizing hydrogen bonding networks, and removing water molecules with the Protein Preparation Wizard panel (Sastry et al., 2013). The binding site was confined to a 20 × 20 × 20 Å cuboid enclosing box with the centroid of coordinates (2XCT: x = 2.74, y = 44.49, z = 67.95; 4KPF: x = -40.34, y = 78.32, z = -10.99). The compounds were drawn using a 2D sketcher and prepared using the Ligprep module for low-energy 3D conformers. With the default parameters, flexible ligand docking was carried out using the extra precision (XP) mode (Bahekar et al., 2007). Docking accuracy was evaluated by extracting the bound conformation of the co-crystallized ligand and redocking into the same binding site. From the obtained results, docking score, glide energy, and glide ligand efficiency were finally chosen to rank the docking poses per ligand. After that, the binding free energy of each docking pose was calculated using the Prime MM-GBSA approach (Kekenes-Huskey et al., 2004), and the graphic models of compound 2-e in complex with 2XCT/4KPF were generated by the 2D ligand interaction diagram and 3D Pose Viewer (Gupta et al., 2017) for detailed analysis of the receptor-ligand interaction.