Small molecules provide powerful tools to interrogate biological pathways but many important pathway participants remain refractory to inhibitors. For example, Cdc25 dual-specificity phosphatases regulate mammalian cell cycle progression and are implicated in oncogenesis, but potent and selective inhibitors are lacking for this enzyme class. Thus, we evaluated 10,070 compounds in a publicly available chemical repository of the National Cancer Institute for in vitro inhibitory activity against oncogenic, full-length, recombinant human Cdc25B. Twenty-one compounds had mean inhibitory concentrations of Ͻ1 M; Ͼ75% were quinones and Ͼ40% were of the paranaphthoquinone structural type. Most notable was NSC 95397 (2,3-bis-[2-hydroxyethylsulfanyl]-[1,4]naphthoquinone), which displayed mixed inhibition kinetics with in vitro K i values for Cdc25A, -B, and -C of 32, 96, and 40 nM, respectively. NSC 95397 was more potent than any inhibitor of dual specificity phosphatases described previously and 125-to 180-fold more selective for Cdc25A than VH1-related dual-specificity phosphatase or protein tyrosine phosphatase 1b, respectively. Modification of the bis-thioethanol moiety markedly decreased enzyme inhibitory activity, indicating its importance for bioactivity. NSC 95397 showed significant growth inhibition against human and murine carcinoma cells and blocked G 2 /M phase transition. A potential Cdc25 site of interaction was postulated based on molecular modeling with these quinones. We propose that inhibitors based on this chemical structure could serve as useful tools to probe the biological function of Cdc25.

Wild-type p53 is frequently mutated in late-stage ovarian cancer and has been proposed as a determinant of cisplatin chemosensitivity. We have therefore established a human ovarian cancer cell line differing only in p53 status and characterized its response after treatment with different platinum complexes. The wild-type p53-expressing cell line A2780 was stably transfected with HPV-16 E6 (E6) or an empty vector (VC) as control. Parental A2780 and VC had similar cisplatin sensitivities, whereas E6 was 3-to 4-fold more sensitive as measured by sulforhodamine B and clonogenic assay. E6 was 2-to 3-fold more sensitive to transplatin and the novel cisplatin analog ZD0473 than VC, whereas the trans-platinum analog JM335 was approximately equitoxic. Platinum uptake was similar for all of the cell lines after cisplatin. The removal of platinum-DNA adducts, as measured by atomic absorption spectroscopy, was reduced in E6 compared with VC after cisplatin but similar after JM335. After 10 M cisplatin, the G 1 population (0 -96 h) was reduced in E6 cells compared with VC, whereas the S phase (8 -48 h) and G 2 phase (48 -96 h) were increased. Similar proportions of VC and E6 cells died by apoptosis, as detected by annexin V binding, but more E6 cells died by necrosis relative to VC. Our results suggest that the loss of functional p53 can increase cisplatin cytotoxicity in A2780, with loss of G 1 /S checkpoint control and decreased cisplatin-DNA adduct repair, but these effects can be circumvented by the use of JM335, which forms different DNA-platinum adducts.

No abstract