Estrogen (GPR30) Receptors

Finally, hydrogen atoms were added, followed by a minimization step with the AMBER99 forcefield in MOE [7]

Finally, hydrogen atoms were added, followed by a minimization step with the AMBER99 forcefield in MOE [7]. 3.4.2. (green) to the least comparable pairs (red). For the quantitative analysis of SAS maps, the structure similarity was also evaluated with MACCS keys Josamycin (166-bits) and PubChem fingerprints as implemented in Activity Scenery Plotter [7]. A DAD map was generated plotting around the X- and Y-axis, the absolute value of the activity difference of compounds tested with G9a and DNMT1, respectively. To analyze the DAD maps threshold value of one logarithmic unit were used. 3.4. Molecular Docking 3.4.1. Protein Preparation The crystallographic structures of human G9a (PDB ID: 3RJW) and DNMT1 (PDB ID: 3SWR) were retrieved from the Protein Data Lender ( [16,17]. Co-crystal ligands were removed (quinazoline-4-amine CIQ and sinefungin, respectively) were removed. Missing loops and side-chains were added with YASARA [18]. Finally, hydrogen atoms were Rabbit Polyclonal to TOR1AIP1 added, followed by a minimization step with the AMBER99 forcefield in MOE [7]. 3.4.2. Ligand Preparation The ligands were built and energy-minimized in MOE using the MMFF94x pressure field (Chemical Computing Group, Montreal, QC, Canada). The more stable protomers at physiological pH were identified [19]. 3.4.3. Molecular Docking AutoDock 4 was used to add the solvent model and assign the atomic charges of Gasteiger to proteins and ligands [9]. For G9a, Josamycin the grid was centered on the carbon atom of the carboxyl group of ASP 1088 (chain A) with a size of 45 45 45 ?3, and for DNMT1 around the carbon atom of the carboxyl group of GLU 1266 (chain A) with a size of 65 65 65 ?3. A grid spacing of 0.375 ? was used. Using the Lamarckian-Genetic algorithm, the binding compounds were subjected to 20 search actions using 2,500,000 energy evaluations, and the default values of the other parameters. The ten best binding poses of the different clusters were generated. 3.4.4. Search for Ideal Conditions Based on studies that describe the interactions involved Josamycin in the molecular recognition of G9a and DNMT1. Key residues in the binding pocket for G9a are TYR 1067, ASP 1078, ASP 1074, ASP 1083, ASP 1088, LEU 1086, TYR 1152 and TYR 1154. For DNMT1 the key residues in the binding pocket are SER Josamycin 1233, MET 1235, TYR 1243, GLU 1269, ARG 1315, ARG 1576 and ASN 1580 [1,20,21]. Compound 6 (not present around the PDBs structures reported) was used to guide the development of a protocol that captured the interactions reported for other active compounds. To this end, different grid sizes were evaluated for G9a (i.e., 20, 30, 40, 45 and 50 ?3) and DNMT1 (i.e., 20, 30, 40, 50, 60, 65 and 70 ?3). The grid sizes selected were 45 45 45 ?3 for G9a, and 65 Josamycin 65 65 ?3 for DNMT1. 3.5. Molecular Dynamics MD studies of the protein-ligand complexes were performed using Desmond (version 2018-3, Schr?dinger, New York, NY, USA) with the OPLS 2005 forcefield [10]. The most representative docking pose for each ligand was used as starting point to initiate the MD simulations. The complexes were prepared with the System Builder Utility in a buffered orthomobic box (10 10 10 ?), using the transferable intermolecular potential with 3-point model for water (TIP3P). The complexes were neutralized and NaCl was added in a 0.15 M concentration. Complexes were minimized using the steep-descent (SD) algorithm followed by the Broyden-Fletcher-Goldfarb-Shanno (LBFGS) method in three stages. In the first stage water heavy atoms were restrained with a pressure constant of 1000 kcal mol?1 ??2 for 5000 actions (1000 SD, 4000 LBFGS) with a convergence criterion of 50 kcal mol?1 ??2; for the second stage, backbones were constrained with a 10 kcal mol?1 ??2 force constant using a convergence.