Abstract 3835: Identification and characterization of inhibitors of 5′-nucleotidase cN-II issued from virtual screening

Abstract

Abstract Clinical and preclinical observations have lead to the hypothesis that 5′-nucleotidase cN-II could constitute a therapeutic target in oncology, either per se, either to increase the activity of cytotoxic nucleoside analogues. We performed in silico screening of freely available chemical databases, in vitro enzymatic assays with recombinant full length cN-II, soaking experiments with crystals of truncated cN-II and biological evaluation of inhibitors alone or in combination with cytotoxic nucleoside analogues on cancer cells both in vitro and in vivo. The top ranked compounds from virtual screening contained an anthraquinone-derivative (A) and a trisubstituted triazine (B) that were selected for further studies. In vitro enzymatic assay experiments with recombinant protein showed that compound A is a competitive inhibitor with Ki of approximately 3 mM, whereas derivative B is a non-competitive inhibitor with Ki of 1.1 mM. We also obtained crystallographic data at a resolution of 2.0 Å after soaking experiments with crystals of truncated cN-II. These showed interaction between A and F354/N154 situated in the effector site 1 of cN-II, whereas B was not found in crystallographic data even though crystals were degraded in presence of B over a longer time. Derivative A showed different levels of cytotoxicity in vitro on several cancer cell lines such as Raji (IC50=782 µM), RL (IC50=210 µM), CCRF-CEM (IC50=824 µM), MCF-7 (IC50=654 µM), A549 (IC50=569 µM) and HCT-116 (IC50=793 µM), whereas B was much less cytotoxic. When used at 1 mM, A and B did not modify IC50 of the nucleoside analogues cladribine, clofarabine or fludarabine in CCRF-CEM cells. When used at 100 µM, compound A increased the induction of apoptosis in RL cells incubated with 0.5 or 1.5 µM cladribine, 0.05 µM clofarabine or 30 µM fludarabine. The administration of A (200 mg/kg, 5d/w, 4w) to mice xenografted with RL cells, induced a delay of tumor development that was higher than in mice treated with fludarabine (50 mg/kg, 1d/w, 4w). We showed that virtual screening can be used for the identification of potent cN-II inhibitors, and we produced clear evidences for the interaction of one inhibitor with the enzyme. Our biological evaluation indicated interesting activity for one lead compound that will be developed further in order to identify candidates with higher biological activity. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3835. doi:1538-7445.AM2012-3835