The long term objective of this research is to develop analogs of mitomycin C that are less toxic and more efficaceous in the treatment of cancer. Our earlier studies have emphasized the synthesis of large series of analogs that might be useful for defining structure-activity relationships. Newer directions include computer-assisted design of analogs and the use of bioactivation mechanisms to define new prodrugs. Based on a recently derived computer model for the binding of mitomycins to B-DNA, analogs will be proposed and evaluated for non-covalent binding using the computer model and the AMBER molecular mechanics program. Promising compounds will be synthesized and submitted for testing. The resulting activities will be used for linear free energy based correlations, with calculated binding energy, log P, and quinone-reduction potential as the independent variables. Nucleotide derivatives of 2,7-diaminomitosene, in which the nucleotide phosphate is substituted at C-1, will be synthesized. The nucleotides will be FdUMP, ara-CMP, 6-mercaptopurine RP, and 6-thioguanine RP. The derivatives will be evaluated for liberation of the nucleotides and DNA alkylation by the mitosene moiety on catalytic reduction. If they pass this test and they are shown to be taken into cancer cells in culture, they will be screened widely against mouse tumors, and against human tumors in cell culture. The corresponding 7-methoxy analogs also will be prepared to ensure bioreductive activation. 2,7-Diaminomitosene, which has no alkylating functionality at C- 1, will be prepared and tested by computer modeling and an experimental binding study for its ability alkylate DNA through the loss of the carbamoyloxy substituent at C-10. This theoretically important process has never been demonstrated. A new set of compounds that represent hybrid structures between mitomycins and Anderson's bis(hydroxymethylpyrrole) biscarbamates will be synthesized and tested against tumors. They are designed to make a good fit to the computer-based mitomycin binding site. Antitumor testing will receive increased emphasis in colon cancer. New analogs will be submitted to a panel of six human colon cancers in cell culture. Derivation of structure-activity relationships in this panel will be attempted.
