INVESTIGADORES
MIGUEL Virginia
congresos y reuniones científicas
Título:
USING VIRTUAL SCREENING FOR THE DISCOVERY OF NEW GABAERGIC INSECTICIDES: ASSESSMENT OF RDL HOMOLOGY MODELS AND DOCKING SCORING FUNCTIONS
Autor/es:
FELSZTYNA, I; MIGUEL V; VILLARREAL MA; GARCÍA DA
Lugar:
La PLata, Argentina
Reunión:
Congreso; XLVII Reunión Anual de la Sociedad Argentina de Biofísica; 2018
Institución organizadora:
Sociedad Argentina de Biofísica
Resumen:
Insect nervous system is the main target of the most used insecticides, as is the case of non-competitive antagonists (NCAs) that block GABAA receptor (GABA-R). In insects, thehomopentamer formed by the so-called Rdl subunit is the most representative type of GABA-R. Docking based virtual screening (VS) is a computational method that allows to explore libraries of chemical compounds in order to identify those which are most likely to bind to a protein target. We aim to validate Rdl structural models and scoring functions to carry out a VS of NCAs that bind to Aedes aegypti Rdl homopentamer (Rdl homo5 ). Since there are no experimental 3D structures of any insect Rdl homo5 , we performed homology modeling to obtain our protein structure. GABA-R, as well as the rest of Cys-loop receptors, is a transmembrane ion channel whose conformation varies among three pharmacological states: open, closed and desensitized. Considering that the NCAs binding site is located inside the channel pore, and that its conformation could affect the docking calculation of the binding of these ligands, we obtained homology models based on Cys-loop receptors templates in the three pharmacological states. A set of ligands which are known to act as NCAs were docked in the proposed binding site of all our models, and the  binding energy obtained by docking was compared with experimental data for each  compound. Also, different docking scoring functions were used to select the combination of protein structure and scoring function that results in the best correlation with the experimental data. Although in previous NCAs binding studies an open channel template was chosen to model an insect Rdl homo5, we obtained the best correlation to experimentaldata using a model based on a structure crystalized in the desensitized conformation (human GABA-R β3 homopentamer), using the Vinardo scoring function. Afterwards, this scoring function was tuned in order to improve the obtained correlation coefficient.