Identification of potential small-molecule cancer drugs in E3 ubiquitin ligase systems
R. Conor Holton-Burke
May 9, 2012
Department or Program
Biochemistry, Biophysics and Molecular Biology
The rapidly expanding understanding of the molecular basis of cancer the function of human proteins and has lead to the identification of several human enzymes and receptors as potential targets for antineoplastic drugs. The current method for identifying leads for potential cancer drugs, High Throughput Screening (HTS), is too expensive and unreliable to test all such potential drug targets. Advancements in virtual screening technology, however, have now made it possible screen drug-like molecules against these proteins quickly and cheaply, using only a crystal structure and an online database of drug-like molecule structure files. In this experiment, Glide docking software (Schrodingler, New York, USA) was used to virtually screen molecules against two such potential targets, the E3 ubiquitin ligase F-box subunits Fbx4 and Cks1, to identify leads for cancer drugs. Additionally, the accuracy of the Glide software was assessed using enrichment screens on a separate protein with known inhibitors. The Glide software was found to be reasonably effective at enriching a virtual screen for known binders, identifying 12 of 16 known binders in the top 10% of the screen results. Whether or not Glide identified cancer drug leads awaits the results of an in vitro assay.