Abstract A1: The physiological form of MetAP2 can be inhibited through binding to either of the two active-site metals

Methionine aminopeptidases (MetAP) are metalloenzymes that remove the N‐terminal initiator methionine from newly synthesized polypeptides allowing essential post‐translational modifications such as acetylation and myristoylation to take place. MetAP2, one of the two eukaryotic forms of the enzyme, was identified as the target of fumagillin, a natural product with anti‐angiogenic properties that inhibits the proliferation of endothelial cells. Clinical activity has been seen for a semi‐synthetic analogue of fumagillin, TNP470, suggesting MetAP2 is a good target for inhibiting angiogenesis. In vitro, MetAP2 appears to have sites for two divalent metal ions within its active site but there has been much discussion around the identity and number of metal ions actually present in the physiological states of the various MetAPs. An understanding of the physiologically relevant metalloform of the enzyme is essential for designing inhibitors that are active in cells. We have used tool compounds that bind the active site metals in diverse ways to investigate the relevance of the two potential metal binding sites in MetAP2. Using our fragment‐based screening approach, Pyramid™, we screened the manganese‐form of the MetAP2 enzyme. We identified multiple low-molecular weight fragment hits and confirmed their modes of binding to the two metals in the active site of MetAP2 by X‐ray crystallography. Three hit series, which bound metal 1 only, metal 2 only or both metals 1 and 2 were chosen for further optimization using structure‐based drug design. Optimized lead compounds had potent inhibitory activity against the in vitro MetAP2 enzyme (∼100 nM) and in HUVEC proliferation assays, whilst also showing greater than 1000‐fold selectivity for MetAP2 over MetAP1. Examples from each series, representing different active site metal binding modes, were used as tool compounds to investigate the mechanism of action in cells. The levels of the MetAP2 substrate, 14‐3‐3, were monitored by western blot in HUVECs treated with these compounds. Levels of methionylated 14‐3‐3 increased upon treatment with compounds from each of our series indicating the substrate was not being processed and that in each case the compound was inhibiting MetAP2 in these cells. These data indicate that the physiological form of MetAP2 can be inhibited by compounds which bind solely to either of the two active‐site metals, suggesting that both metals must be present in the intra‐cellular form of MetAP2 and allowing multiple approaches to inhibiting this key angiogenic target. The lead series identified here provide chemically diverse scaffolds for further optimization of drug like properties. Citation Information: Mol Cancer Ther 2009;8(12 Suppl):A1.