| Abstract: | The total synthesis of (12S,13S)‐trans‐epothilone A ( 1a ) was achieved based on two different convergent strategies. In a first‐generation approach, construction of the C(11)? C(12) bond by Pd0‐catalyzed Negishi‐type coupling between the C(12)‐to‐C(15) trans‐vinyl iodide 5 and the C(7)‐to‐C(11) alkyl iodide 4 preceded the (nonselective) formation of the C(6)? C(7) bond by aldol reaction between the C(7)‐to‐C(15) aldehyde 25 and the dianion derived from the C(1)‐to‐C(6) acid 3 . The lack of selectivity in the aldol step was addressed in a second‐generation approach, which involved construction of the C(6)? C(7) bond in a highly diastereoselective fashion through reaction between the acetonide‐protected C(1)‐to‐C(6) diol 31 (‘Schinzer's ketone') and the C(7)‐to‐C(11) aldehyde 30 . As part of this strategy, the C(11)? C(12) bond was established subsequent to the critical aldol step and was based on B‐alkyl Suzuki coupling between the C(1)‐to‐C(11) fragment 40 and C(12)‐to‐C(15) trans‐vinyl iodide 5 . Both approaches converged at the stage of the 3‐O, 7‐O‐bis‐TBS‐protected seco acid 27 , which was converted to trans‐deoxyepothilone A ( 2 ) via Yamaguchi macrolactonization and subsequent deprotection. Stereoselective epoxidation of the trans C(12)? C(13) bond could be achieved by epoxidation with Oxone ® in the presence of the catalyst 1,2 : 4,5‐di‐O‐isopropylidene‐L ‐erythro‐2,3‐hexodiuro‐2,6‐pyranose ( 42a ), which provided a 8 : 1 mixture of 1a and its (12R,13R)‐epoxide isomer 1b in 27% yield (54% based on recovered starting material). The absolute configuration of 1a was established by X‐ray crystallography. Compound 1a is at least equipotent with natural epothilone A in its ability to induce tubulin polymerization and to inhibit the growth of human cancer cell lines in vitro. In contrast, the biological activity of 1b is at least two orders of magnitude lower than that of epothilone A or 1a . |
