1949
(2S,3R,4R)-8 in 91% yield. Its stereochemistry was confirmed by NOESY spectra in which significant
NOEs were observed between 2-H and 3-H, 2-H and 4-H, and 3-H and 4-H.
Treatment of the pyrrolidinone 8 with tert-butyldimethylsilyl chloride followed by reaction with
phosphorus pentasulfide produced the thiolactam 9 in 75% yield. At this time we decided to try to
use the thiolactam 11 with a free hydroxyl group to carry out the Eschenmoser sulfide contraction.
Accordingly, the silyl protecting groups were removed from 9 under the action of TsOH in methanol,
and the phenol hydroxy group was reprotected with TBDPS to provide 10. The compound 10 was
isomerized to 11 by treatment with potassium carbonate in methanol for 30 min. It was observed that
S-alkylation of 11 with 1-bromo-2-octadecanone in methylene chloride and subsequent Eschenmoser
sulfide contraction mediated by triphenylphosphine and triethylamine worked well to afford the olefin
12 in 75% yield. Finally, deprotection of 12 with TBAF/HOAc in THF furnished (2S,3S,4R)-plakoridine
A in 71% yield. Its spectral data were the same as those of the natural product except for its optical
rotation (lit.1 [α]D20=−0.4 (c 0.5, CHCl3), our observed: ([α]D22=−43.0 (c 0.5, CHCl3)).9 Although the
reason for this difference is not clear, one possible explanation is that the natural plakoridine A isolated
by Kobayashi’s group may be racemic.
Initially, an acetyl group was chosen for protecting the two hydroxy groups in 8 because of its small
steric hindrance and easy deprotection. However, we found that S-alkylation8 of the thiolactam 13 with
1-bromo-2-octadecanone in methylene chloride with the assistance of silver triflate followed by treatment
with triphenylphosphine and triethylamine did not give the desired olefin but delivered a polysubstituted
pyrrole 14. This product might result from the elimination of the 4-OAc. When the thiolactam 10 was
used for the Eschenmoser transformation, a similar elimination product (involving the 4-hydroxy as a
leaving group) was obtained. No S-alkylation occurred when the thiolactam 9 reacted with 1-bromo-2-
octadecanone, perhaps because of the larger steric hindrance of the silyloxy group. When the thiolactam
11 was used as a substrate in which the 4-OH is cis to 3-H, the elimination was depressed thereby giving
the desired olefin.
In conclusion, we have developed a stereoselective route to (2S,3S,4R)-plakoridine A from the
conveniently available β-amino ester 6. The overall yield for the 14 steps is about 3%. It is notable
that the present synthesis also provides an unusual example of an Eschenmoser sulfide contraction of a
fully substituted thiolactam, which will be a guide for synthesizing other polysubstituted pyrrolidines or
piperidines.
Acknowledgements
The authors are grateful to the Chinese Academy of Sciences and National Natural Science Foundation
of China (grant 29725205) for their financial support. We thank Professor Kobayashi for providing NMR
spectrum of natural plakoridine A.
References
1. Takeuchi, S.; Ishibashi, M.; Kobayashi, J. J. Org. Chem. 1994, 59, 3712.
2. Takeuchi, S.; Kikuchi, T.; Tsukamoto, S.; Ishibashi, M.; Kobayashi, J. Tetrahedron 1995, 51, 5979.
3. (a) Roth, M.; Dubs, P.; Goschi, E.; Eschenmoser, A. Helv. Chim. Acta. 1971, 54, 710. (b) Tamaru, Y.; Harada, T.; Yoshida, I.
J. Am. Chem. Soc. 1980, 102, 2392. (c) Hart, D. J.; Kanai, K. J. Am. Chem. Soc. 1985, 107, 1255.
4. Stafford, J. A. Tetrahedron Lett. 1995, 36, 681.
5. (a) Davies, S. G.; Ichihara, O. Tetrahedron: Asymmetry 1991, 2, 183. (b) Davies, S. G.; Ichihara, O.; Walters, I. A. S. J.
Chem. Soc., Perkin Trans. 1 1994, 1141. (c) Davies, S. G.; Smyth, G. D. Tetrahedron: Asymmetry 1996, 7, 1001. (d) Davies,