DESCRIPTION (provided by applicant): Microtubules, which are formed by the self-association of the alpha/beta-tubulin heterodimers, provide structural support for a cell and play key roles in cell motility, mitosis, and meiosis. They are also the targets of several anticancer agents, indicating their importance in maintaining cell viability. Currently available tubulin binding anticancer drugs, including new taxol derivatives and epothilones, interact with beta-tubulin subunit of the alpha/beta-tubulin heterodimers and have no effect on microtubule minus ends. Furthermore, cancer cells with an altered beta-tubulin expression profile may be resistant to these agents. We used a three-dimensional computer model of tubulin constructed based upon its recently resolved electron crystallographic structure for rational design of a novel mono-tetrahydrofuran (THF)-containing synthetic anticancer drug targeting a unique narrow binding cavity on the surface of alpha-tubulin. We discovered a previously unidentified region with a remarkable abundance of leucine residues, which is located between the GDP/GTP binding site and the taxol binding site. This unique region contains a narrow cavity with elongated dimensions, which could accommodate a fully stretched aliphatic chain with a length of up to twelve carbon atoms. Using this model, a comprehensive structure search of the organic compound files in the Parker Hughes Institute Drug Discovery Program led to the identification of the recently reported chiral THF-epoxides as potential molecular templates for the rational synthesis of novel anti-cancer drugs containing structural elements capable of hydrophobic binding interactions with this leucine-rich binding cavity of tubulin. Our lead compound designated as COBRA-1, inhibited GTP-induced tubulin polymerization in cell free turbidity assays. Treatment of human breast cancer and brain tumor (glioblastoma) cells with COBRA-1 caused destruction of microtubule organization and apoptosis. Like other microtubule-interfering agents, COBRA-1 activated the pro-apoptotic c-Jun N-terminal kinase (JNK) signal transduction pathway, as evidenced by rapid induction of c-jun expression. The further development of COBRA-1 as an anticancer agent will depend on in vivo efficacy, and toxicity studies in relevant animal models. We are now proposing to use the severe combined immunodeficiency (SCID) mouse model for detailed in vivo anticancer activity in SCID mice challenged with human breast cancer or glioblastoma cells. Our specific aims are: (i) To study the in vivo toxicity profile of COBRA-1 in BALB/c mice and (ii) To study the in vivo anti-cancer activity of COBRA-1 in a SCID mouse model of metastatic human breast cancer and glioblastoma. The knowledge gained from these studies described under Specific Aims 1-2 is expected to facilitate the design of innovative treatment regimens employing COBRA-1 for the treatment of metastatic solid tumors.