The First Entry to Complex Bakkanes: A Highly Effective Retroaldol–Aldol-Based Approach to (−)-Bakkenolides III, B, C, and H

Abstract

Dedicated to Prof. Jean Lhomme on the occasion of his retirementSince the first members were isolated and elucidated in the late 1960s, the bakkanes have rapidly increased in number, particularly recently, so as to count now approximately 50. [1] These hydrindane sesquiterpenes of both terrestrial and marine origin, characterized by their novel spiro g-butyrolactone function and significant biological effects, [1m,n, 2] have become popular synthetic targets. To date bakkenolide A, homogynolides A and B, palmosalide C, and 9-acetoxyfukinanolide have been prepared, most more than once. [3] The nonfunctionalized and monofunctionalized bakkenolides, however, are each oxygenated singly and taken together comprise only a relatively minor number of the known bakkanes. Much more representative (ca. 80 %) of this class are the densely functionalized and stereochemically challenging C1,C9-dioxygenated bakkenolides (I, R, R 1 H, acyl). These compounds, which have significant inhibitory activity toward Hep G2, Hep G2,2,15, and P-388 tumor cell lines, [1n] as well as PAF, arachidonic acid, and collagen-induced platelet aggregation, [1m] have been isolated from Petasites japonicus Maxim. and Petasites formosanus Kitamura (Compositae). [1b±f, m, n] Herein we illustrate the first entry to this large group of bakkanes through the preparation of the representative natural bakkenolides III, B, C, and H (1 ± 4). We note in advance that this highly stereocontrolled and efficient approach allows potentially total access to the group.With its six stereocenters in the proper configurations and two hydroxyl groups which could be, in principle, easily differentiated, diol 1 (bakkenolide III) appeared to be the ideal late intermediate for accessing the entire range of known C1,C9-difunctionalized bakkenolides (Scheme 1). It seemed this diol might be secured from epoxy ketone II through chemoselective reduction, followed by stereochemical adjustments.Scheme 1. Retrosynthesis of bakkenolide III (1). [4] revealed diol 1 to have the lowest global minimum energy of the four C7,C9 diastereomers, thus it could possibly be reached through a retroaldol ± aldol equilibration. Epoxy ketone II, in turn, was viewed as arising from keto lactone 5, an intermediate previously prepared in our laboratory from 1,6-dimethylcyclohexene. [3n, 5] Keto lactone 5 was synthesized stereoselectively from enantiopure 1,6-dimethylcyclohexene (6) [3d] in five steps and 27 % overall yield [6] in a manner similar to that used earlier. It was found, however, that ring expansion could be improved by using boron trifluoride as the catalyst and that neutral conditions benefitted the transesterification [7] (Scheme 2). Scheme 2. Synthesis of keto lactone 5. a) Ref. [3d, n]; b) Zn, CH 3 CO 2 H, 70 8C, 1 h (63 %, 2 steps); c) N 2 CHCO 2 C 2 H 5 , BF 3´O (C 2 H 5 ) 2 , (C 2 H 5 ) 2 O, 0 8C, 12 h (72 %); d) HCCCH 2 OH, C 6 H 5 CH 3 , reflux, 24 h (87 %); e) Mn(O 2 CCH 3 ) 3 , C 2 H 5 OH, 20 8C, 2.5 h (68 %). Molecular mechanics calculationsOf the numerous procedures exam...