Natural volatiles causing inhibition of mosquito biting behaviors: development of a virtual screening platform predicting antagonists of ORco function for accelerated discovery
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Insect olfactory receptors are heteromeric ligand-gated cation channels composed of an obligatory receptor subunit, ORco, and one of many variable subunits, ORx, in as yet undefined molar ratios. When expressed alone ex vivo , ORco forms homotetrameric channels gated by ORco-specific ligands acting as channel agonists. Using an insect cell-based system as a functional platform for expressing mosquito odorant receptors ex vivo, we identified small molecules of natural origin acting as specific ORco channel antagonists, orthosteric or allosteric relative to a postulated ORco agonist binding site, and causing severe inhibition of olfactory function in mosquitoes. In the present communication, we are reporting on the compilation of common structural features of such orthosteric antagonists and development of a ligand-based pharmacophore whose properties are deemed necessary for binding to the agonist binding site and inhibition of ORco’s biological function. In silico screening of an available collection of natural volatile compounds with this pharmacophore resulted in identification of several ORco antagonist hits. Cell-based functional screening of the same compound collection resulted in the identification of several compounds acting as orthosteric and allosteric antagonists of ORco channel function ex vivo and inducing anosmic behaviors to Aedes albopictus mosquitoes in vivo. Comparison of the in silico screening results with those of the functional assays revealed that the pharmacophore predicted correctly 7 out of the 8 confirmed orthosteric antagonists and none of the allosteric ones. Because the pharmacophore screen also produced additional hits that did not cause inhibition of the ORco channel function, we generated a Support Vector Machine (SVM) model based on two descriptors of all pharmacophore hits. Training of the SVM on the ex vivo validated compound collection resulted in the selection of the confirmed orthosteric antagonists with a very low cross-validation out-of-sample misclassification rate. Employment of the combined pharmacophore-SVM platform for in silico screening of a larger collection of olfaction-relevant volatiles produced several new hits. Functional validation of randomly selected hits and rejected compounds from this screen confirmed the power of this virtual screening platform as convenient tool for identifying novel vector control agents. To the best of our knowledge, this study is the first one that combines a pharmacophore with a SVM model for identification of AgamORco antagonists and specifically orthosteric ones.