A program of research on the synthesis of pyrrolizidine alkaloids (PAs) and derivative structures will be continued. PAs are plant metabolites which, in many cases, are known to possess severely hepatotoxic properties but which also manifest promising antitumor activity. This activity has been shown to extend to extend in a limited clinical study to successful stabilization of breast and ovarian cancer. The long-range goal of this project is to discover through synthesis substances with desirable therapeutic properties for the treatment of solid tumors and for which dose-limiting hepatotoxic side effects are absent. A hypothesis is advanced that antitumor activity and hepatotoxicity of PAs are associated with two chemically distinct metabolic pathways. It is further postulated that antitumor action derives from a transformation (Polonovski reaction) of the pyrrolizidine N-oxide which renders the system more reactive as an electrophile. A primary objective of the present proposal is to devise chemistry which tests these ideas. The structural templates around which the chemistry is to be designed consists of five PAs-indicine N-oxide (INO), monocrotaline, swazine, retroisosenine, and loline. A new, general entry to the necine bases of PAs is proposed based on desymmetrization of an achiral (meso) pyrrolizidindione. Both enzymatic desymmetrization with liver alcohol dehydrogenase and chemical desymmetrization by stereoselective reduction will be investigated. Efficient synthesis of heliotridine, retronecine, crotanecine, otonecine, and loline in enanthiomerically pure form are envisioned by this means. A previously explored, asymmetric route to necic acids from the terpenes (R)- and (S)-Beta-citronellol will be developed further to accommodate trachelanthic (the necic acid of INO), monocrotalic, swazinecic, bulgarsenecic, and nemorensic acids. These substances will also provide the starting point for hybrid structures designed to test the possible participation of necic acids in the projected activation pathway of pyrrolizidine oxides. Coupling of necine bases with necic acids to form macrolactone PAs is envisioned from the corresponding pyrrolizidine- borane and features a recently devised lactonization strategy involving a catalytic, tin(IV)-mediated closure of a omega hydroxy trifluoroethyl ester. The efficacy of this plan for the design of anticancer agents based on PA structures will be tested by examination of the chemical and biological properties of their N-oxides. A screening regime conducted by NCI will be used for evaluation of synthesized compounds for tumor- inhibitory activity. Chemical studies will focus on the Polonovski reaction of PA N-oxide mediated by a Fe(II)FE(III) redox system, the purpose being to ascertain whether a putative azacyclooctenone is formed in this process and if so, whether it possesses properties which could account for the antitumor activity of INO and other PAs.