文摘
In silico chemical library screening (virtual screening) was used to identify a novel lead compound capableof inhibiting S-adenosylmethionine decarboxylase (AdoMetDC). AdoMetDC is intimately involved in thebiosynthesis of polyamines, which are essential for tumor progression and are elevated in numerous typesof tumors. Therefore, inhibition of this enzyme provides an attractive target for the discovery of novelanticancer drugs. We performed virtual screening using a computer model derived from the X-ray crystalstructure of human AdoMetDC and the National Cancer Institute's Diversity Set (1990 compounds). Ourdocking study suggested several compounds that could serve as drug candidates since their docking modesand scores revealed potential inhibitory activity toward AdoMetDC. Experimental testing of the top-scoringcompounds indicated that one of these compounds (NSC 354961) possesses an IC50 in the low micromolarrange. A search of the entire NCI compound collection for compounds similar to NSC 354961 yielded twoadditional compounds that exhibited activity in the experimental assay but with significantly diminishedpotency relative to NSC 354961. In this report, we disclose the activity of NSC 354961 against AdoMetDCand its probable binding mode based on computational modeling. We also discuss the importance of virtualscreening in the context of enzymes that are not readily amenable to high-throughput assays, therebydemonstrating the efficacy of virtual screening, combined with selective experimental testing, in identifyingnew potential drug candidates.