The focus of our analyses of the MD trajectories was coordination to the iron atom in the heme group and the possibility for the interactions with Arg239 and/or Asp298. group. The other compounds showed predominantly only one binding pose. Compounds 1d and 1e with only one nitrogen atom available for coordination were positioned accordingly and compound 1b was coordinating with benzimidazole moiety (Supplementary Materials S4). We submitted each complex for 100 ns molecular dynamics (MD) simulations without any constraints with different initial velocities to explore if the docking poses represented the optimal binding mode. The focus of our analyses of the MD trajectories was coordination to the iron atom in the heme group and the possibility for the interactions with Arg239 and/or Asp298. Simulations of compounds 1a and 1c, which displayed mixed binding modes, were most revealing and generally showed a preference for pyridine-iron coordination. This is in agreement with previous DFT calculations showing stronger heme affinity towards pyridine than benzimidazole [27]. A comparison of the average free energies of binding calculated by the MM/GBSA method (Table 2) shows that compounds 1a, 1c and 1e should bind better to the CYP17A1 enzymes than 1b and 1d. However compound 1a suffers a penalty due to unfavourable nitrogen atom position in the pyridine ring which agrees with the experimentally determined binding (Table 1). It is also noteworthy to add that, although binding through pyridine nitrogen atom seems to be preferred, the relatively low binding energy of compound 1d together with high affinity demonstrates that binding through benzimidazole cannot be ruled out. However, in the case of compound 1d, there is additional interaction where NH between the benzene linker and pyridine ring forms a hydrogen bond to Asp298 and pi-cation interaction between benzene ring and the charged sidechain of Arg239. Table 2 Average binding free energy of compounds 1a-1e with two possible binding modes. n/anon applicable. = 3; DMSO wells = 15C30) and normalized growth rate inhibition (GR) metrics were calculated according to Hafner et al. 2016 [19]. GR50 curves were visualized using the SB 271046 Hydrochloride GR-calculator SB 271046 Hydrochloride webserver (www.grcalculator.org) [29]. 3.9. Molecular Modelling The Protein Preparation Wizard in Maestro Software version 11.1 was used to prepare the proteins structures [30]. The cytochrome P450 17A1 structures were obtained from Protein Data Bank [31] (PDB protein codes 3SWZ [10] and 5IRQ [11]). Bond orders were assigned, hydrogens were added, and zero-order bonds to metals were created. For protein structures the A chains were selected, and all water molecules were removed. The formal charge on heme iron was set to +3 and non-protonated ligand state was used. The hydrogen bonding network was optimized at pH 7.0. A SB 271046 Hydrochloride restrained protein minimization was performed using OPLS3 [32] force field with convergence of heavy atoms to RMSD 0.30 ?. Ligands preparation was performed with LigPrep in Maestro [30]. Possible tautomers and protonation states were generated at pH 7.0 2.0. The Epik program was used to predict pKa values of ligands [33]. Docking was performed with GOLD (Genetic Optimisation for Ligand Docking) program version 5.6 [34]. Proteins prepared by Protein Preparation Wizard were used without additional modifications in GOLD. The co-crystalized ligand was extracted, and the binding site was defined around the center of the mass of the co-crystalized ligand within 15 ?. Ligands prepared by LigPrep were exported from Maestro. Ligands were docked 10 times with slow genetic algorithm and with ChemScore as the scoring SB 271046 Hydrochloride function [35]. For constrained docking Mouse monoclonal to CD3.4AT3 reacts with CD3, a 20-26 kDa molecule, which is expressed on all mature T lymphocytes (approximately 60-80% of normal human peripheral blood lymphocytes), NK-T cells and some thymocytes. CD3 associated with the T-cell receptor a/b or g/d dimer also plays a role in T-cell activation and signal transduction during antigen recognition the distance between the heme iron and the atom expected to be coordinated to Fe was constrained between 1.5 and 3.5 ?. The Desmond system builder was used to create the molecular dynamics (MD) systems with the protein-ligand complex embedded in a SPC water model yielding an orthorhombic box with a buffer size of 10 ? between the protein and the box boundary. The final system comprised close to 70,000 atoms including approximately 7500 atoms for the protein including the heme group, 36 atoms for the ligand (in the case of 1a), one chloride ion to neutralize the system, and approximately 21,000 water molecules. The MD simulations were performed with the Desmond program (version 3.6) using SB 271046 Hydrochloride the OPLS-2005 force field [36]. For equilibration of the system prior to the production runs, the Desmond default equilibration protocol was used. Subsequently, the systems were simulated for 100 ns and 1000 frames collected. A total of 36 simulations were performed based on the.