|Abstract or Summary
- Avermectins are a group of potent anthelmintic agents produced by Streptomyces avermitilis. They are active at extremely low dosages and are now in commercial use. In this study, the synthesis of two segments of avermectin B[subscript 1a] are described. The synthesis of the hexahydrobenzofuran subunit (segment A) began with the Diels-Alder reaction between maleic anhydride and 3-methyl-1-[(trimethylsilyl)oxyl-113-butadiene which gave the synthon 79. Cleavage of the silyl protecting group of 79 afforded the acid-lactone 81 which was resolved using (R)-(+)-methylbenzylamine. The absolute configuration of the resolved acid was determined by X-ray crystallographic analysis of its salt. Conversion of 81 to diazoketone 86, followed by acid catalyzed cyclization, then gave the key hexahydrobenzofuran building block 88. Stereoselective epoxidation of the C-617 olefin of 88 furnished the β-epoxide 94. Opening of the epoxide moiety of 94 to form the required C-5,6 olefin was attempted using several methods. Hydroiodic acid, followed by further manipulations, led to the undesired, conjugated isomer 104. The problem was solved by the use of triethylsilyl trifluoromethanesulfonate, which afforded a mixture of endo and exo olefins 118 and 119 with concomitant formation of the silyl ether of the ketone. Regioselective epoxidation of this mixture furnished two alcohols, the major product being the exo isomer 120. Attempted epimerization at C-4 of 94 with sodium methoxide gave epoxy lactone 126 which, upon treatment with triethylsilyl trifluoromethanesulfonate, afforded exclusively the endo olefin 127. Acidic hydrolysis of 127 gave 133 with the desired (natural) configuration at C-4. However, introduction of an angular hydroxyl group into 133 using the same methodology as before was unsuccessful in this case. The synthesis of segment B was initiated with the enantioselective epoxidation of allylic alcohol 141. Cuprate-mediated methylation of 142 gave exclusively the (2S,3S)pentanediol 143. Subsequent oxidation of the primary alcohol, followed by Wittig olefination with (carbomethoxymethyl)triphenylphosphorane, provided the required carbon skeleton. Reduction of the ester function of 149 furnished the allylic alcohol 150 which was transformed to the corresponding phosphonium salt. An attempted Wittig reaction of the ylid derived from 152 with 120 failed to give the desired product, presumably due to steric hindrance of the ketone function.