CHEMICAL STUDIES OF TAXOL AND MICROTUBULES

紫杉醇是一种从太平洋红豆杉（Taxus brevifolia）中分离出来的二萜

Taxol, a diterpene isolated from the Pacific yew, Taxus brevifolia, has exhibited antineoplastic activity in phase I clinical trials against melanoma, adenocarcinoma, refractory ovarian carcinoma, non-small cell lung carcinoma, gastric, colon, and head and neck carcinomas, and lymphoblastic leukemias. In combination with cisplatin, activity has been observed against advanced large cell lung carcinoma, non-small cell lung carcinoma, breast, head and neck, and colon carcinomas, advanced ovarian carcinoma, and melanoma. Activity also has been displayed by taxol in phase II trials against advanced ovarian epithelial tumors in patients pretreated with radiation and chemotherapy, some of whom were considered resistant to cisplatin. Three areas associated with the development of taxol as an antitumor agent, and with the exploitation of its unique mode of action as a fundamentally new and general strategy for the management of neoplastic disease are addressed in this application. At the present time, the limited quantities of taxol available are insufficient for its broad application in cancer chemotherapy. Partial synthesis through the attachment of the critical A-ring side chain to more readily available but inactive natural substances can alleviate this problem. We will develop three new A-ring side chain attachment methodologies, two that involve 5,6-dihydro-6-keto-1,3 oxazine chemistry (including one that coalesces asymmetric side chain synthesis with attachment to the diterpenoid nucleus into a total of five steps), and one that involves ketene methodology. The success of the former strategy is suggested by preliminary results on the attachment of side chains with variable functionality. The binding site(s) of taxol on microtubules remain completely uncharacterized. This state of affairs prevents the understanding at the molecular level of taxol's unique mode of action, which, in turn, undermines its exploitation through the design of new drugs. Three areas of investigation aimed at the chemically relevant characterization of the taxol binding site will be pursued. First, we will evaluate experimentally (further structure-activity studies and the conformational analysis of taxol analogues through NOE and 13C T1 studies) and computationally (molecular mechanics modeling and dynamical calculations) the hypothesis that taxol and its biologically active analogues are preorganized for tight binding to microtubules. This notion is supported strongly by preliminary structure-activity and modeling results. Should this prove to be true, significant information on the nature of the binding site and for the optimization of drug design will be available in the short term. Second, we will engage in photoaffinity labeling of the binding site. These studies will reveal the domains and amino acid residues of microtubules responsible for the recognition of taxol. Third, we will combine the information from the first two studies with the available data on the structure of the tubulins in a computational model of the taxol binding site. The picture that will emerge will reveal the three-dimensional chemical features of the interaction between taxol and the important structural elements of tubulin at the binding site. Finally, we will pursue the preparation of taxol derivatives that will be useful at the cellular and subcellular levels for the investigation of the biology that underlies taxol's cytotoxicity. For example since the site on microtubules to which taxol binds cannot have evolved for that purpose, it is possible that taxol mimics an endogenous substance with similar biological activity. Antibodies generated to taxol will be employed in a search for such endogenous substances. Similar antibodies and fluorescent taxol analogues for the imaging of taxol bound to microtubules will be prepared.