Virtual screening approaches for the identification of non-lipid autotaxin inhibitors.

Autotaxin (ATX, NPP-2) catalyzes the conversion of lysophosphatidyl choline (LPC) to lysophosphatidic acid (LPA), a mitogenic cell survival factor that stimulates cell motility. The high expression of both ATX and receptors for LPA in numerous tumor cell types has produced substantial interest in exploring ATX as an anticancer chemotherapeutic target. ATX inhibitors reported to date are analogs of LPA, a phospholipid, and are more hydrophobic than is typical of orally bioavailable drugs. This study applied both structure-based and ligand-based virtual screening techniques with hit rates of 20% and 37%, respectively, to identify a promising set of non-lipid, drug-like ATX inhibitors. Structure-based virtual screening necessitated development of a homology model of the ATX catalytic domain due to the lack of structural information on any mammalian NPP family member. This model provided insight into the interactions necessary for ATX inhibition, and produced a suitably diverse training set for the development and application of binary QSAR models for virtual screening. The most efficacious compound identified in this study was able to completely inhibit ATX-catalyzed hydrolysis of 1 microM FS-3 (a synthetic, fluorescent LPC analog) at a 10 microM concentration.