Total synthesis of (+)-chinensiolide B via tandem allylboration/ lactonization

Article abstract of DOI:10.1021/ja9104478

(Chemical Equation Presented) The chinensiolides are a family of guaiane type α-methylene γ-lactone natural products recently isolated from Ixeris chinensis Nakai, a plant used in Chinese folk medicine. The first enantioselective total synthesis of (+)-chinensiolide B was accomplished in 15 steps for the longest linear sequence with an overall yield of 6.7% starting from inexpensive and readily available (R)-carvone. A highly stereoselective and E/Z-selective tandem allylboration/lactonization reaction between two highly functionalized partners was exploited as a key step. The synthesis also highlights several solutions to chemoselectivity issues arising from the reactive α-methylene γ-lactone. For instance, a ring-closing metathesis formed the requisite seven-membered ring in a chemoselective fashion while avoiding the reactivity of the conjugated α-methylene unit.

Full text of DOI:10.1021/ja9104478

Published on Web 01/12/2010
Total Synthesis of (+)-Chinensiolide B via Tandem Allylboration/Lactonization
Tim G. Elford and Dennis G. Hall*
Department of Chemistry, UniVersity of Alberta, Edmonton, AB, T6G 2G2, Canada
Received December 10, 2009; E-mail: dennis.hall@ualberta.ca
R-Methylene γ-lactones are privileged structures present in a
large number of natural products displaying a wide scope of
biological activities.1 The chinensiolides constitute a family of
guaiane type R-methylene γ-lactone natural products with a tricyclic
5,7,5-ring system (Figure 1). The various members of the chinen-
siolide family were recently isolated from Ixeris chinensis Nakai,2,3
a plant used in Chinese folk medicine. Chinensiolide B (1) has
been shown to display cytotoxic behavior against human primary
liver cancer (HepG2) and human lung fibroblast (WI-38 and VA-
13) cell lines.4 The activity of 1 against HepG2 is comparable to
that of paclitaxel; however, its selectivity remains to be explored.
Substituted R-methylene γ-lactones can be assembled in one step
via a thermal, Lewis, or Brønsted acid catalyzed addition of
2-alkoxycarbonyl allylboronates to aldehydes in tandem with
lactonization.5 This attractive process has never been applied to
substrates as densely functionalized as 4 and 5, and the critical
issue of diastereofacial selectivity is unclear. Moreover, early
introduction of the reactive R-methylene γ-lactone is risky due to
potential chemoselectivity issues later on in the sequence. Regard-
less, it was expected that aldehyde 4 would arise from (R)-carvone
via a Favorskii rearrangement. This enantioselective route would
provide an opportunity to confirm the absolute configuration of
natural chinensiolide B.
The synthesis began with a two-step protocol reported by Ley
and co-workers to convert carvone into ketone 66 (Scheme 1).
Using ketone 6, protection of the secondary alcohol was followed
by a Favorskii rearrangement6 to provide the desired tetrasub-
stituted cyclopentane 7 in an excellent yield of 85%. The ester
was fully reduced and then reoxidized under Swern conditions7
to provide aldehyde 4. Converting 4-pentyn-1-ol into allylbor-
onate 5 was achieved by standard literature protocols.8-10
Unfortunately, 5 could only be prepared as a mixture of alkene
isomers (Z/E ratio of ∼3.5:1) and these isomers could not be
separated on a gram scale. As a result, the isomeric mixture
was employed in the allylboration of aldehyde 4.
Figure 1. Structures of chinensiolides A-D.
No total synthesis of any of the chinensiolides has been reported
to date. The challenge of chinensiolide B is alluring. It contains
six contiguous stereocenters, including five along a flexible seven-
membered ring,2 and the sixth stereocenter may be subject to
epimerization due to a neighboring ketone. The R-methylene
γ-lactone itself is a reactive center, and reactions carried out after
its installation must be carefully optimized to avoid possible side
reactions. Keeping these potential pitfalls in mind, a retrosynthetic
analysis suggested that (+)-1 could be accessed via regioselective
reductive opening of epoxide 2 (Figure 2). The latter intermediate
would arise from a desilylation/elimination followed by a chemose-
lective ring-closing metathesis and diastereoselective epoxidation
of 3. Precursor 3 would be produced directly from the key tandem
allylboration/lactonization reaction between carvone-derived alde-
hyde 4 and allylboronate 5, which would be made in three steps
from 4-pentyn-1-ol.
Scheme 1
With the two key fragments 4 and 5 in hands, the key tandem
allylboration/lactonization step was attempted. After much experi-
mentation with thermal and catalytic methods, it was found that
using 2.5 mol % of BF3 · OEt2 at 0 °C for 48 h provided trans
γ-lactone product 3 in a remarkable yield of 87% (based on the
amount of Z-5) (Scheme 2). The observed diastereoselectivity
(>95% dr) was surprising since four isomers could be expected
from this reaction. Indeed, the reaction was E/Z-selective as
allylboronate E-5 proved to be essentially inert to the reaction
conditions. The trans diastereoselectivity in the allylboration step
can be explained with the usual six-membered chairlike transition
state of this reaction. The remarkable diastereofacial selectivity on
aldehyde 4 can be rationalized according to the Felkin model shown
in Scheme 2, with the large vinyl-bearing carbon placed opposite
to the approach of reagent 5.
Figure 2. Retrosynthetic analysis of chinensiolide B (+)-1.
9
1488 J. AM. CHEM. SOC. 2010, 132, 1488–1489
10.1021/ja9104478  2010 American Chemical Society

C O M M U N I C A T I O N S
Scheme 2
R-methylene γ-lactone group from undergoing conjugate addition
with the highly reactive LiEt3BH. Simple treatment of the crude
triol with MnO2 easily reformed the R-methylene γ-lactone. The
two diastereomers originating from the epoxidation could now be
separated to provide the desired γ-lactone 11. Finally, oxidative
cleavage of the secondary TBS protecting group16 was achieved
in one step to reveal the ketone, thus completing the total
synthesis.17
This first enantioselective total synthesis of (+)-chinensiolide B
(1) was achieved in 15 steps for the longest linear sequence with
an overall yield of 6.7% starting from inexpensive and readily
available (R)-carvone. The absolute configuration of the chinensi-
olides is thus confirmed. A highly stereoselective and E/Z-selective
tandem allylboration/lactonization reaction between two highly
functionalized partners was exploited as a key step. The synthesis
also highlights several solutions to chemoselectivity issues arising
from the reactive R-methylene γ-lactone. For instance, ring-closing
metathesis formed the requisite seven-membered ring chemoselec-
tively while avoiding the reactivity of the conjugated R-methylene
unit.
Acknowledgment. This work was funded by the Natural
Sciences and Engineering Research Council (NSERC) of Canada
and the University of Alberta. T.G.E. thanks NSERC and the
Alberta Ingenuity Fund for graduate scholarships. We thank Dr.
Robert McDonald for the X-ray crystallographic analysis of 8.13
Supporting Information Available: Full experimentals and NMR
spectral reproductions for all compounds, and complete ref 3. This
material is available free of charge via the Internet at http://pubs.acs.org.
Selective deprotection of the primary TBDPS group11 on 3 was
followed by a Grieco elimination12 of the primary alcohol to afford
desired triene 9, albeit with only a moderate yield of 60% for the
two steps. Careful control of the stoichiometry of both reagents
was important to minimize formation of byproduct 8 where the
cyanide anion had undergone conjugate addition to the enoate.10
Fortuitously, byproduct 8 proved to be crystalline and allowed for
X-ray crystal structure13 confirmation of the allylboration’s ste-
reoselectivity. Formation of medium rings by RCM can be
problematic when the final alkene is tri- or tetrasubstituted.14 In
spite of our apprehension, a chemoselective RCM of triene 9 using
5 mol % of Grubbs II catalyst provided the desired tricycle 10 in
high yield. Most likely, steric bulk around the lactone’s R-methylene
unit and formation of a bridgehead olefin helped suppress closure
to the possible tetrasubstituted six-membered enoate, allowing the
desired RCM pathway to proceed uncontested. The final stage of
the synthesis involves the diastereoselective epoxidation of this
newly formed alkene in 10. Nucleophilic epoxidation reagents could
not be used due to the electrophilic nature of the R-methylene
γ-lactone. Satisfactorily, treatment of 10 with mCPBA gave epoxide
2 as an unseparable 4:1 mixture of diastereomers favoring the
desired one. Regioselective opening of epoxides to give Markovni-
kov products is typically achieved through the use of nucleophilic
hydride reagents.15 However, with 2, over-reduction of the γ-lactone
occurred with LiAlH4 to give the fully saturated triol and conjugate
reduction of the R-methylene group took place preferentially to
epoxide opening with LiEt3BH. What proved ultimately successful
was a one-pot double reduction protocol whereby the γ-lactone
moiety of 2 was first reduced to the diol with DIBALH, and then
LiEt3BH was added to regio- and chemoselectively open the
epoxide. This unusual protocol allowed for protection of the
References
(1) (a) Hoffmann, H. M. R.; Rabe, J. Angew. Chem., Int. Ed. Engl. 1985, 24,
94–110. (b) Konaklieva, M. I.; Plotkin, B. J. Mini ReV. Med. Chem. 2005,
5, 73–95.
(2) Zhang, S.; Wang, J.; Xue, H.; Deng, Q.; Xing, F.; Ando, M. J. Nat. Prod.
2002, 65, 1927–1929.
(3) Zhang, S.; et al. J. Nat. Prod. 2006, 69, 1425–1428.
(4) Dai, J.; Yang, L.; Sakai, J.; Ando, M. J. Mol. Catal. B: Enzym. 2005, 33,
87–91.
(5) (a) Hall, D. G. Synlett 2007, 1644–1655. (b) Elford, T. G.; Arimura, Y.;
Yu, S. H.; Hall, D. G. J. Org. Chem. 2007, 72, 1276–1284. (c) Kennedy,
J. W. J.; Hall, D. G. J. Org. Chem. 2004, 69, 4412–4428.
(6) Oliver, S. F.; Hogenauer, K.; Simic, O.; Antonello, A.; Smith, M. D.; Ley,
S. V. Angew. Chem., Int. Ed. 2003, 42, 5996–6000.
(7) Omura, K.; Swern, D. Tetrahedron 1978, 34, 1651–1660.
(8) (a) Ramachandran, P. V.; Pratihar, D.; Biswas, D.; Srivastava, A.; Reddy,
M. V. R. Org. Lett. 2004, 6, 481–484.
(9) (b) Nyzam, V.; Belaud, C.; Villie´ras, J. Tetrahedron Lett. 1993, 34, 6899–
6902.
(10) See Supporting Information for details.
(11) Crouch, R. D. Tetrahedron 2004, 60, 5833–5871.
(12) Grieco, P. A.; Gilman, S.; Nishizawa, M. J. Org. Chem. 1976, 41, 1485–
1486.
(13) Crystallographic data (excluding structure factors) for the structure of 9
have been deposited with the Cambridge Crystallographic Data Center as
supplementary publication no. CCDC-756958. Copies of the data can be
obtained in the SI. (cif file) or via www.ccdc.cam.ac.uk/data_request/cif.
(14) For example, see: (a) Chatterjee, A. K.; Morgan, J. P.; Scholl, M.; Grubbs,
R. H. J. Am. Chem. Soc. 2000, 122, 3783–3784. (b) Hekking, K. F. W.;
van Delft, F. L.; Rutjes, F. P. J. T. Tetrahedron 2003, 59, 6571–6758.
(15) Brown, H. C.; Krishnamurthy, S. Tetrahedron 1979, 35, 567–607.
(16) Coss´ıo, F. P.; Aizpurua, J. M.; Palomo, C. Can. J. Chem. 1986, 64, 225–
231.
(17) The NMR spectral data and the optical rotation of synthetic 1 were in full
agreement with that of natural chinensiolide B (+)-1.3,9 We thank Prof.
M. Ando for supplying a copy of the 1H NMR spectrum of isolated (+)-1.
As it has been shown that (+)-1 obtained from the natural source can be
converted into chinensiolide C in three steps, this total synthesis of 1 also
constitutes a formal total synthesis of chinensiolide C in 18 steps for the
longest linear sequence.
JA9104478
9
J. AM. CHEM. SOC. VOL. 132, NO. 5, 2010 1489