Simple enantioselective total synthesis of glabrescol, a chiral C2-symmetric pentacyclic oxasqualenoid [17]

Full text of DOI:10.1021/ja0024901

9328
J. Am. Chem. Soc. 2000, 122, 9328-9329
Scheme 1a
Simple Enantioselective Total Synthesis of
Glabrescol, a Chiral C2-Symmetric Pentacyclic
Oxasqualenoid
Zhaoming Xiong and E. J. Corey*
Department of Chemistry and Chemical Biology
HarVard UniVersity, Cambridge, Massachusetts 02138
ReceiVed July 10, 2000
Glabrescol, the first pentacyclic member of the oxasqualenoid
family which includes diverse and structurally novel members
such as teurilene,1,2 thyrsiferol (venustatriol)3,4 longilene peroxide,5
quassiol A,6 and eurylene,7 was originally assigned the novel CS-
symmetric structure 1,8 the origin of which could reasonably be
explained as cascade pentacyclization of the precursor 2. However,
it was recently demonstrated by unequivocal synthesis of 1 that
glabrescol did not possess this structure.9 Furthermore, it was
shown by synthesis of the three other CS-symmetric diastereomers
of 1 which could result from a similar cascade pentacyclization
that none of these compounds correspond to glabrescol.9 We
surmised9 that the true structure of glabrescol might be generated
by a bidirectional pair of double cyclizations which could generate
either a novel CS-symmetric or a C2-symmetric structure, either
of which would be consistent with the reported8 spectroscopic
data for glabrescol. Although the reported lack of rotation of
glabrescol would seem to indicate that it is CS-symmetric, a C2-
symmetric structure would be possible if the optical rotation of
such a structure were to be very small or if the reported optical
rotation of glabrescol were to be in error. In this paper we describe
an enantioselective total synthesis of the C2-symmetric structure
3 and its identity with glabrescol. The synthesis is outlined in
Scheme 1.10
a Conditions: (a) Rieke barium, THF, -78 °C, 1 h (56%). (b) 3:1
HOAc-H2O, 50 °C, 3 h (97%). (c) Oxone, ketone 7 (cat.), K2CO3, buffer
pH 10.5, (MeO)2CH2-CH3CN-H2O, 0 °C, 1.5 h (66%). (d) Camphor-
10-sulfonic acid (6 equiv), CH2Cl2, -94 °C, 3 h (44%). (e) CH3SO2Cl
(5 equiv), C5H5N (10 equiv), DMAP (2 equiv), 0 °C for 1 h then 23 °C
for 1 h (50% yield at 60% conversion). (f) 0.1 M NaOAc in HOAc at 40
°C for 12 h (65%).
(c 1.2, CHCl3) (97%), which has been isolated previously from
various natural sources.13,14 Epoxidation of 6 using the Shi chiral
dioxirane from ketone 715 afforded tetraepoxide 8, [R]23D +38.1
(c 1.3, EtOH), in 66% yield (estimated diastereomeric purity ca.
The starting point for the synthesis was the bromo acetonide
4,11 which upon stirring with Rieke barium at -78 °C for 1 h
gave the bis-acetonide 5,12 [R]23 +3.5 (c 1.5, MeOH) (56%).
D
Ketal cleavage of 5 produced the chiral tetraol 6, [R]23 +18.2
D
1
80% by H NMR analysis) with an estimated R/S selectivity at
each double bond of >20:1.9 Exposure of 8 to camphor-10-
sulfonic acid in CH2Cl2 at -94 °C effected bidirectional tetra-
cyclization to 9, which was isolated in pure condition by column
(1) Suzuki, T.; Suzuki, M.; Furusaki, A.; Matsumoto, T.; Kato, A.; Imanaka,
Y.; Kurosawa, E. Tetrahedron Lett. 1985, 26, 1329.
(2) (a) Hashimoto, M.; Harigaya, H.; Yanagiya, M.; Shirahama, H. J. Org.
Chem. 1991, 56, 2299. (b) Morimoto, Y.; Iwai, T.; Kinoshita, T. J. Am. Chem.
Soc. 1999, 121, 6792.
(3) Sakemi, S.; Higa, T.; Jefford, C. W.; Bernardinelli, G. Tetrahedron
Lett. 1986, 27, 4287.
(4) Corey, E. J.; Ha, D.-C. Tetrahedron Lett. 1988, 29, 3171.
(5) Itokawa, H.; Kishi, E.; Morita, H.; Takeya, K.; Iitaka, Y. Chem. Lett.
1991, 2221.
(6) Tinto, W.; McLean, S.; Reynolds, W. F.; Carter, C. A. G. Tetrahedron
Lett. 1993, 34, 1705.
(7) Itokawa, H.; Kishi, E.; Morita, H.; Takeya, K.; Iitaka, Y. Tetrahedron
Lett. 1991, 32, 1803.
(8) Harding, W. W.; Lewis, P. A.; Jacobs, H.; McLean, S.; Reynolds, W.
F.; Tay, L.-L.; Yang, J.-P. Tetrahedron Lett. 1995, 36, 9137.
(9) Xiong, Z.; Corey, E. J. J. Am. Chem. Soc. 2000, 122, 4831.
(10) After the submission of this manuscript (July 7, 2000) a paper appeared
in which syntheses of 1 and 3 were reported by other, more lengthy routes;
see: Morimoto, Y.; Iwai, T.; Kinoshita, T. J. Am. Chem. Soc. 2000, 122,
7124.
chromatography on silica gel in 44% yield; [R]23 -6.1 (c 0.2,
D
CHCl3). Tetraol 9 could be selectively converted to the monome-
sylate 10 (at 60% conversion) despite an unusual lack of reactivity
(11) (a) This compound was prepared in four steps and 58% overall yield
from E,E-farnesyl acetate by the following sequence: (1) position and
enantioselective terminal dihydroxylation (as described by: Corey, E. J.; Noe,
M. C.; Lin, S. Tetrahedron Lett. 1995, 36, 8741), (2) bis-acetonide formation
(2,2-dimethoxypropane, 0.01 equiv of p-TsOH at 23 °C for 2 h), (3)
deacetylation (K2CO3, MeOH, at 23 °C for 2 h), and (4) mesylate formation
(1.3 equiv each of CH3SO2Cl and Et3N in THF, at -45 °C for 45 min) and
in situ displacement of mesylate by addition of 5 equiv of LiBr, and reaction
at 0 °C for 1 h. (b) See also: Huang, A. X.; Xiong, Z.; Corey, E. J. J. Am.
Chem. Soc. 1999, 121, 9999.
(12) (a) Corey, E. J.; Shieh, W.-C. Tetrahedron Lett. 1992, 33, 6435. (b)
Corey, E. J.; Noe, M. C.; Shieh, W.-C. Tetrahedron Lett. 1993, 34, 5995.
10.1021/ja0024901 CCC: $19.00 © 2000 American Chemical Society
Published on Web 09/07/2000

Communications to the Editor
J. Am. Chem. Soc., Vol. 122, No. 38, 2000 9329
of the two secondary (neopentyl type) hydroxyl groups of 9; for
10; [R]23D +10.0 (c 0.2, CHCl3).16 Because of the extreme steric
shielding at the backside of the mesyloxy carbon subunit of 10,
we anticipated that solvolysis of 10 in HOAc, a good ionizing
solvent, would occur via nucleophilic participation of the vicinal
tetrahydrofuran oxygen, forming the intermediate oxonium ion
11 (with inversion), which would then undergo a second internal
with those reported and measured by us for an authentic sample
of glabrescol.17 In addition the chromatographic mobilities
measured for synthetic 3 and natural glabrescol were identical in
several different solvent systems. Finally, the optical rotation of
3, [R]23 -25.2 (c 0.3, CHCl3), compares well with a recently
D
determined sample of natural glabrescol, [R]D about -28.18,19 On
the basis of all these data, as well as previous synthetic work,9 it
is clear that glabrescol possesses the C2-symmetric structure 3
and not a CS-symmetric alternative.20
The success of the bidirectional double cyclization of tetraol
tetraepoxide 8 to the tetracycle 9, a key feature in the synthesis
outlined in Scheme 1, can be ascribed to a considerably faster
rate of closure of rings A and B relative to ring C. Consequently,
the bidirectional formation of rings A/B and A′/B′ occurs rather
than a unidirectional formation of an A/B/C/B′ tetracycle. Of
course, in the case of the pentacyclization of 2 to form 1,9 the
unidirectional cyclization cascade is favored because there is a
single initiating secondary hydroxyl function.
This study suggests a number of interesting questions for further
research, for example: (1) Will the CS-symmetric pentatetrahy-
drofuran oxasqualenoids previously synthesized9 make their
appearance in due course as natural products? (2) Does the
biosynthetic route to glabrescol involve cyclizations analogous
to those used for the synthetic construction described herein? (3)
Do glabrescol and the other pentatetrahydrofurans display sig-
nificant bioactivity?
backside displacement by the secondary hydroxyl to generate a
new tetrahydrofuran ring. The resulting product would correspond
to overall ring closure of 10 with retention of configuration and,
more specifically, would therefore be the C2-symmetric chiral
product 3. We were pleased to find that the major product from
the acetolysis of 10 was indeed 3, as was clear from its optical
1
activity and the 2-fold simplification of the H and 13C NMR
spectra, due to C2-symmetry. The 500 MHz 1H NMR, 125 MHz
Acknowledgment. This research was assisted financially by a grant
from the National Science Foundation.
13C NMR, IR, and high-resolution mass spectra were identical
(13) (a) Nishiyama, Y.; Moriyasu, M.; Ichimaru, M.; Kato, A.; Mathenge,
S. G.; Nganga, J. N.; Juma, F. D. Phytochemistry 1999, 52, 1593. (b) Jonker,
S. A.; Nkunya, M. H. H.; Mwamtobe, L.; Geenevasen, J.; Koomen, G.-J.
Nat. Prod. Lett. 1997, 10, 245. (c) Nishiyama, Y.; Moriyasu, M.; Ichimaru,
M.; Tachibana, Y.; Kato, A.; Mathenge, S. G.; Nganga, J. N.; Juma, F. D.
Phytochemistry 1996, 42, 803.
(14) Rf values measured for compounds 3 and 5-10 using silica gel TLC
plates with the indicated elution solvent were as follows: 5, 0.30 (5:1 hex-
Et2O); 6, 0.15 (1:1 hex-EtOAc); 8, 0.24 (10:1 EtOAc-EtOH); 9, 0.26 (1:2
hex-EtOAc); 10, 0.19 (1:2 hex-EtOAc); 3, 0.37 (1:2 hex-EtOAc).
(15) Wang, Z.-X.; Tu, Y.; Frohn, M.; Zhang, J.-R.; Shi, Y. J. Am. Chem.
Soc. 1997, 119, 11224.
Supporting Information Available: Experimental procedures for the
synthesis of 3 (PDF). This material is available free of charge via the
Internet at http://pubs.acs.org.
JA0024901
(18) Personal communication to E.J.C. from Dr. Helen Jacobs, May 3, 2000.
We are indebted to Dr. Jacobs for this information and for her generous
assistance in this research.
(19) Additional values for the optical rotation of synthetic 3 as a function
of wavelength (Hg lines): -26.1 at 578 nm, -28.1 at 546 nm, -38.7 at 436
nm, and -43.2 at 365 nm (c 0.3, CHCl3).
(16) The two secondary hydroxyl groups of 9 are very resistant to SN2
substitution reactions, e.g., Mitsunobu-type displacement or reaction with
Ph3P-Br2.
(20) Measurement of the 1H-1H NOESY spectrum of synthetic 3 by us
confirmed the cis arrangement of the H and CH3 substituents at C(2) and
C(5) of the two equivalent A rings and also the two equivalent B rings, as
required by that structure.
(17) Kindly provided by Dr. Helen Jacobs, University of the West Indies.