Design of Druglike Small Molecules with LYN-Specific Binding

Abstract

Glioblastoma multiforme (GBM), a very aggressive brain tumor, has a median survival of only 14 months. LYN, an important kinase involved in regulation of hematopoietic cells, is significantly overexpressed in GBM. It is believed that LYN promotes migration of cancer cells, thus advancing the malignancy. This research addresses computational design of small druglike molecules that could potentially inhibit LYN and thus stave off the cancer advancement. LYN has a very similar binding site to the polo-box domain (PBD) in Polo-like kinase 1 (Plk1). Plk1 is a main regulator of mitosis. Considering the key cellular roles of both LYN and Plk1, it is important to design inhibitors that will specifically bind to LYN. In this work, physical and chemical properties of the binding sites of LYN and Plk1 were investigated and compared. Pertinent atomic distances within the LYN binding site were found to be smaller than those within the PBD of Plk1. The two sites also differed in their flexibilities. By utilizing the differences, novel molecules were designed that could potentially bind LYN with higher affinities than they could Plk1. Previously designed molecules that bonded both LYN and Plk1 were used as initial templates to design more specific inhibitors. Potential toxicities and drug-likeness of the molecules were evaluated. Molecules with no implied toxicities and optimal druglike properties were used for docking studies. Molecules that made the most stable docking configurations with LYN and with no other kinases were identified as LYN-specific. Binding energies of the stable complexes that these molecules formed with LYN were calculated. Possible utilization of the designed molecules against tumors with overexpressed LYN is discussed.