Homology Modeling Angiotensin Ⅱ Type1Receptor and Molecular Design for Its Antagonists
|School||Central South University|
|Course||Chemical Engineering and Technology|
|Keywords||Angiotensin Ⅱ type1receptor split-and-link modeling ligand-steered remodeling receptor-ligand complex modeling moleculardocking virtual screening molecular design|
Angiotensin Ⅱ type1receptor (AT1) antagonists serve as potent antihypertensives. Homology model of AT1could be of great help for the design of novel AT1antagonists, considering the absence of AT1’s crystal structure. Among previous studies on AT1’s homology modeling, the sequence identity between AT1and the template is smaller than30%, in which case conventional homology modeling approach can not provide models sufficient for rational drug design. To this end, the present work attempted to improve the current modeling procedure from the perspectives of templates and techniques in order to gain models that can satisfy rational drug design.Seeking for better templates, we appointed several subtypes of opioid receptor as templates for modeling AT1, whose sequence identities with AT1are higher than30%.Breakthroughs in the development of homology modeling techniques:split-and-link strategy, ligand-steered remodeling and protein-ligand complex modeling have been reported recently. Considering that applications of these techniques in AT1’s homology modeling have not been reported for now, they were separately utilized here to generate AT1models.(1) A hybrid model was generated from5initial models, each from a template, by split, pick and link, whose structural quality was apparently enhanced.(2) Losartan was docked into5initial models, and then the complexes obtained from docking were imported for remodeling including losartan, during which induced-fit between losartan and residues in binding sites was performed. The remodeling procedure provided5final models with better conformation of binding site and stronger interaction with losartan.(3) The accuracy of current protein-ligand complex homology modeling approach is dependent on the target-template ligand similarity. Since non-peptide AT1antagonists are quite structurally different from template-ligands, the current strategy is not suitable for AT1’s modeling. Herein, we demonstrated a novel sites-steered pre-docking strategy, which tried to dock the target-ligand into the template-receptor to form hybrid-template. The binding sites for pre-docking were defined by locating those template residues which were aligned to AT1’s hot residues during sequences alignment. Then the hybrid-template was utilized for complex modeling, during which the contained target-ligand was employed for induced-fit. In this way,10AT1-losartan/telmisartan models were generated from5templates, of which more than97%residues were within the allowed regions according to the Ramachandean φ-Ψ dihedral plots. Besides, ligands in complex models reappeared interactions with generally recognized hot residues (Lys199, His256, Va1108, et al) of AT1.These10AT1-losartan/telmisartan models were verified with molecular docking and virtual screening tests. Docking scores consistent with experimental activity data were observed in the docking of telmisartan, candesartan and losartan for all10models. They also enriched more than81%of active compounds in the top ranked10%of the entire database, among which tel-m2did best by reaching100%. Then, the model tel-m2was chosen to guide the design for AT1antagonists, providing candidates with potential AT1antagonistic activity.