Identification of non-lipid LPA3 antagonists by virtual screening Original Research Article

In the present study, we utilized virtual screening to identify LPA3 antagonists. We have developed a three-point structure-based pharmacophore model based on known LPA3 antagonists. This model was used to mine the NCI database. Docking, pharmacophore development, and database mining produced new, non-lipid leads. Experimental testing of seven computationally selected pharmacophore hits produced one potentiator and three antagonists, one of which displays both LPA3 selectivity and nanomolar potency. Similarity searching in the ChemBridge database using the most promising lead as the search target produced four additional LPA3 antagonists and a potent dual LPA1&2 antagonist.