Synthesis of an F-H gambierol subunit using a C-glycoside-centered strategy

Article abstract of DOI:10.1021/ol034100w

(Matrix presented) This manuscript describes our synthesis of the F-H subunit of gambierol. In addition to the synthesis of the tricycle, of note is an interesting protecting group influence on the generation of a C(23) C-glycoside as well as the use of r

Full text of DOI:10.1021/ol034100w

ORGANIC
LETTERS
2003
Vol. 5, No. 6
913-916
Synthesis of an F−H Gambierol Subunit
Using a C-Glycoside-Centered Strategy
,†
Utpal Majumder,† Jason M. Cox, and Jon D. Rainier*
Department of Chemistry, UniVersity of Utah, 315 South 1400 East,
Salt Lake City, Utah 84112, and Department of Chemistry, The UniVersity of Arizona,
Tucson, Arizona 85721
rainier@chem.utah.edu
Received January 20, 2003
ABSTRACT
This manuscript describes our synthesis of the F−H subunit of gambierol. In addition to the synthesis of the tricycle, of note is an interesting
protecting group influence on the generation of a C(23) C-glycoside as well as the use of ring-closing metathesis to generate a tetrasubstituted
enol ether.
Gambierol, a member of the marine ladder toxin family of
natural products, was isolated in 1993 by Yasumoto and co-
workers from the marine dinoflagellate Gambierdiscus
toxicus.1 As part of this initial work, the Yasumoto group
determined gambierol’s relative configuration; they subse-
quently elucidated its absolute structure.2 Not surprisingly,
gambierol’s polycyclic ether architecture and intriguing
biological activity has attracted the attention of chemists
interested in its synthesis. To date, this has led to a number
of important studies3 and the total synthesis of gambierol
by the Sasaki and Yamamoto groups.4
Figure 1.
† Current address: Department of Chemistry, University of Utah, 315
South 1400 East, Salt Lake City, UT 84112.
(1) Satake, M.; Murata, M.; Yasumoto, T. J. Am. Chem. Soc. 1993, 115,
Our interest in gambierol stems from our program that
361.
targets the chemical synthesis of polycyclic ether-containing
natural products. Central to our approach has been the
generation of carbon C-glycosides from the single flask
coupling of glycal anhydrides with carbon nucleophiles.5-7
As applied to gambierol, we recently described the use of
(2) Morohashi, A.; Satake, M.; Yasumoto, T. Tetrahedron Lett. 1999,
40, 97.
(3) (a) Marmsater, F. P.; West, F. G. J. Am. Chem. Soc. 2001, 123, 5144.
(b) Kadota, I.; Ohno, A.; Matsuda, K.; Yamamoto, Y. J. Am. Chem. Soc.
2001, 123, 6702. (c) Marmsater, F. P.; Vanecko, J. A.; West, F. G.
Tetrahedron 2002, 58, 2027. (d) Kadota, I.; Ohno, A.; Matsuda, K.;
Yamamoto, Y. J. Am. Chem. Soc. 2002, 124, 3562. (e) Sasaki, M.; Tsukano,
Tachibana, K. Org. Lett. 2002, 4, 1747.
(4) (a) Fuwa, H.; Kainuma, N.; Tachibana, K.; Sasaki, M. J. Am. Chem.
Soc. 2002, 124, 14983. (b) Kadota, I.; Takamura, H.; Sato, K.; Ohno, A.;
Matsuda, K.; Yamamoto, Y. J. Am. Chem. Soc. 2003, 125, 46.
(5) For a recent monograph on C-glycoside synthesis, see: Postema, M.
H. D. In Glycochemistry Principles, Synthesis and Applications; Bertozzi,
C., Wang, P. G., Eds.; Marcel Dekker: New York, 2000; pp 77-131.
10.1021/ol034100w CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/19/2003

this strategy to synthesize the A-D subunit.8 Outlined herein
is the synthesis of the F-H subunit.
nesium chloride gave C-glycosides 4a or 4b, respectively,
as single diastereomers in high yield.13,14
Our approach to the F-H subunit was to utilize differ-
entially protected glucal as the G-ring, introduce the F-ring
using a C-glycoside-forming reaction, and then to pursue the
synthesis of the H-ring. Although glucal contains an ad-
ditional hydroxyl substituent (C(25) in gambierol), we
believed that the additional functionality would be beneficial
in that it would direct the facial selectivity in the epoxidation
reaction and ultimately the formation of the C(24) and C(23)
stereocenters.
With this plan in mind, our synthesis of the F-H subunit
began with D-glucal derivative 2.9 Introduction of the C(23)
methyl group (gambierol numbering system) gave C-glyco-
side precursor 3.10 To determine the feasibility of the
oxidation/coupling sequence, we examined the reaction of
the epoxide from glycal 3 with a number of nucleophiles
(Table 1). To our immense pleasure, the epoxidation of 3
To generate the gambierol skeleton, methyl substitution
on the nucleophile was required; unfortunately, all attempts
at coupling the epoxide from 3 with 2-methylpropenylmag-
nesium chloride were unsuccessful and led to pinacol
rearrangement product 5 (entry 3). As we had previously
found that the success of some glycal anhydride coupling
reactions was highly dependent on the nature of the coun-
terion on the nucleophile,6c we examined the reaction of
2-methylpropenylmagnesium bromide with the epoxide from
3. We were extremely pleased to find that this coupling was
successful to give 4c in 90% yield.
Having achieved the synthesis of the desired C-glycoside,
we had the opportunity to examine the influence of C(25)
substitution on the coupling reaction. Unexpectedly, the tert-
butyldiphenylsilyl (TBDPS) group proved to be important
not only in the DMDO oxidation but also for the subsequent
C-C bond-forming sequence. That is, when TBDPS was
replaced by tert-butyldimethylsilyl (TBDMS), the selectivity
in the coupling diminished significantly (eq 1). While the
nature of the TBDPS effect is presently unclear to us, we
do not believe that it can be simply attributed to steric
interactions, as the TBDPS group sits on the face of the
epoxide that undergoes attack.
Table 1.
With an efficient route to C-glycoside 4c in hand, our next
challenge was the generation of the tetrasubstituted enol ether
required for the synthesis of the F-ring. From the outset, our
plan had been to use an enol ether-olefin ring-closing
metathesis (RCM) reaction. This was clearly a daunting task;
while a number of related transformations have taken place,15
to the best of our knowledge, there was very little precedent
1
a Minor product was not observed by H NMR.
with dimethyl dioxirane (DMDO)11,12 followed by the
addition of propenylmagnesium chloride or propynylmag-
(12) To obtain reproducible yields on a large scale, we used Messeguer’s
“acetone-free” conditions. The use of this protocol allowed us to avoid the
concentration of the epoxide. See the Experimental Section and: Ferrer,
M.; Gibert, M.; Sa´nchez-Baeza, F.; Messeguer, A. Tetrahedron Lett. 1996,
37, 3585.
(13) Our previous work with similarly substituted epoxides had resulted
in poor diastereoselectivity and/or low yields. The major products from
these reactions were generally rationalized as coming from the formation
of intermediate oxocarbenium ions. See refs 6 and 8.
(6) (a) Rainier, J. D.; Allwein, S. P. J. Org. Chem. 1998, 63, 5310. (b)
Rainier, Cox, J. M. Org. Lett. 2000, 2, 2707. (c) Allwein, S. P.; Cox, J. M.;
Howard, B. E.; Johnson, H. W. B.; Rainier, J. D. Tetrahedron 2002, 58,
1997.
(7) Among the more challenging structural features of gambierol are the
C(7), C(11), C(21), and C(23) stereocenters. These centers contain angular
methyl groups oriented in a 1,3-disposition to one another and, for our
approach, require the addition of a carbon nucleophile to the more substituted
end of a trisubstituted glycal anhydride. For synthetic work by others toward
these centers, see refs 3 and 4 and: Suzuki, K.; Matsukura, H.; Matsuo,
G.; Koshino, H.; Nakata, T. Tetrahedron Lett. 2002, 43, 8653.
(8) Cox, J. M.; Rainier, J. D. Org. Lett. 2001, 3, 2919.
(9) Parker, K. A.; Georges, A. T. Org. Lett. 2000, 2, 497.
(10) Lesimple, P.; Beau, J.-M.; Jaurand, G.; Sinay, P. Tetrahedron Lett.
1986, 27, 6201.
(11) For the preparation of DMDO, see: Adam, W.; Bialas, J.;
Hadjiarapoglou, L. Chem. Ber. 1991, 124, 2377.
(14) We have also coupled Al nucleophiles with trisubstituted glycal
anhydrides. See refs 6b and 6c.
(15) For selected examples of enol ether-olefin RCM, see: (a) Fujimura,
O.; Fu, G. C.; Grubbs, R. H. J. Org. Chem. 1994, 59, 4029. (b) Clark, J.
S.; Kettle, J. G. Tetrahedron Lett. 1997, 38, 123. (c) Postema, M. H. D. J.
Org. Chem. 1999, 64, 1770. (d) Clark, J. S.; Hamelin, O. Angew. Chem.,
Int. Ed. 2000, 39, 372. (e) Chatterjee, A. K.; Morgan, J. P.; Scholl, M.;
Grubbs, R. H. J. Am. Chem. Soc. 2000, 122, 3783. (f) Rainier, J. D.; Cox,
J. M.; Allwein, S. P. Tetrahedron Lett. 2001, 42, 179.
914
Org. Lett., Vol. 5, No. 6, 2003

for the formation of a tetrasubstituted enol ether.16 With this
in mind, 4c was converted into RCM precursor 11. Esteri-
fication of the C(24) hydroxyl group required a large excess
of acid 9, DCC, and DMAP for success. The subsequent
enol ether-forming reaction using the Takai protocol17 was
even more problematic; despite considerable effort, we were
able to generate enol ether 11 in only 35% yield.18 In light
of the steric crowding about C(24) in 4c, it is probably not
surprising that these conversions gave us difficulty.
that had been successful for us previously (rt, benzene, 15%
catalyst),6c,15f we only recovered starting material when 11
was subjected to 14 (entry 2). The enhanced stability of 14
at elevated temperatures turned out to be critical;21 subjecting
11 to 14 at 65 °C resulted in the generation of a small amount
(5%) of the tetrasubstituted enol ether 12 (entry 3). We were
pleasantly surprised to find that we could isolate 12 in 82%
yield by simply increasing the temperature of the reaction
to 80 °C (entry 4).22
While pleased with the formation of 12, we were not
satisfied with the reactions leading up to 12. From the notion
that the C(25) TBDPS ether might be hindering functional-
ization at C(24), we opted to postpone the RCM reaction
until after C(25) deoxygenation. Toward this goal, selective
removal of the TBDPS group gave the corresponding
C(24),C(25) diol.23,24 Conversion of the diol into the C(25)
TMS ether provided 15 after acylation with 9 and removal
of the TMS group using HOAc. We were encouraged to find
that the esterification of the C(25) TMS analogue of 4c was
much easier than it had been for 4c itself. Methyl xanthate
formation and free radical-induced deoxygenation25 provided
enol ether precursor 16. By subjecting 16 to the Takai
With 11 in hand, we investigated the generation of the
F-ring using ring-closing metathesis (RCM). Because of its
generally higher reactivity, we initially examined Schrock
catalyst 13;19 to our disappointment, we only recovered
starting material when 11 was subjected to 13 (Table 2, entry
Scheme 1
Table 2.
procedure, we isolated 17 in 83% yield. This experiment
clearly validates the notion that our difficulties with the
derivatization of 4c and 10 had been due to the steric
(20) Scholl, M.; Ding, S.; Lee, C. W.; Grubbs, R. H. Org. Lett. 1999, 1,
953.
(21) See ref 20 and: Sanford, M. S.; Love, J. A.; Grubbs, R. H. J. Am.
Chem. Soc. 2001, 123, 6543.
(22) We believe that there could be two reasons for the need for elevated
temperatures: (1) the inherent difficulty associated with generating tetra-
substituted enol ethers; (2) the reaction may be proceeding through a
relatively unreactive Fischer carbene intermediate prior to cyclization. For
an excellent study on the formation and reactivity of Fischer carbenes from
enol ethers and 14, see: Louie, J.; Grubbs, R. H. Organometallics 2002,
21, 2153.
1). From the notion that the transformation of 11 into 12
might be catalyst dependent, we turned to Grubbs’ second
generation Ru imidazole catalyst 14.20 Using the conditions
(16) To the best of our knowledge, the only other example of the
formation of a tetrasubstituted enol ether using metathesis also came from
our laboratory. See: Cox, J. M. Ph.D. Dissertation, The University of
Arizona, Tucson, Arizona, 2002.
(17) Takai, K.; Kakiuchi, T.; Kataoka, Y.; Utimoto, K. J. Org. Chem.
1994, 59, 2668.
(18) Other enol ether-forming protocols (Tebbe, Petasis) were even less
successful in our hands.
(19) Schrock, R. R.; Murdzek, J. S.; Bazan, G. C.; Robbins, J.; DiMare,
M.; O’Regan, M. B. J. Am. Chem. Soc. 1990, 112, 3875.
(23) Shekhani, M. S.; Khan, K. M.; Mahmood, K.; Shah, P. M.; Malik,
S. Tetrahedron Lett. 1990, 31, 1669.
(24) Attempts to selectively remove the TBDPS group from 10 were
unsuccessful.
(25) Paquette, L. A.; Oplinger, J. A. J. Org. Chem. 1988, 53, 2953.
Org. Lett., Vol. 5, No. 6, 2003
915

environment about C(24). RCM utilizing the conditions that
had been successful for 11 provided tetrasubstituted enol
ether 18.26
Scheme 2
To complete the gambierol F-ring, it remained to oxidize
C(21) and reduce C(20). Although hydroboration and oxida-
tion would be the conventional method for carrying out this
transformation,27 we targeted a single flask dioxirane oxida-
tion/DIBAL reduction approach.28 Oxidation of 18 with
DMDO and reduction of the resulting glycal anhydride (i.e.,
19) using DIBAL provided 20 as a 10:1 mixture of
diastereomers in 91% yield.29 With the successful reduction,
we had completed the synthesis of the F-ring and the C(21)
and C(23) angular methyl groups.
unoptimized 37% overall yield by first converting 23 into a
mixture of the corresponding cyclic and acyclic acetals and
then by subjecting the mixture to PPTS, pyridine, and heat
according to the protocol that we had developed earlier.32
In summary, we have synthesized the F-H subunit of the
marine ladder toxin gambierol utilizing a C-glycoside-
centered strategy. Of note in these studies is our discovery
that TBDPS substitution at C(25) influenced the formation
of a C(23) C-glycoside. Also noteworthy was the use of the
second generation Grubbs’ catalyst in the synthesis of
tetrasubstituted enol ethers 12 and 18 using enol ether-olefin
RCM.
Our final challenge for the F-H coupling precursor was
the synthesis of the H-ring. From the possibilities,30 we opted
to examine an acid-mediated cyclization sequence. To this
end, we removed the cyclic silylene group and then
sequentially converted the primary alcohol into the corre-
sponding triflate and the secondary alcohol into the corre-
sponding TBDMS ether to give 21. The remaining carbon
atoms for the H-ring were introduced by coupling triflate
21 with allyl cuprate.31 Hydroboration/oxidation and Swern
oxidation of the resulting primary alcohol gave aldehyde 23.
Finally, tricyclic oxepene 24 was generated from 23 in an
Acknowledgment. We are grateful to the National
Institutes of Health, General Medical Sciences (GM56677),
for support of this work. We would like to thank Dr. Charles
Mayne for help with NMR experiments and Mr. Elliot M.
Rachlin for help in obtaining mass spectra.
(26) We also attempted to cyclize 17 using Schrock’s catalyst 13 without
success.
(27) Hydroboration of cyclic enol ethers has been applied to the synthesis
of polycyclic ether natural products. For a selected example, see: Nicolaou,
K. C.; Bunnage, M. E.; McGarry, D. G.; Shi, S.; Somers, P. K.; Wallace,
P. A.; Chu, X.-J.; Konstantinos, A.; A.; Gunzner, J. L.; Yang, Z. Chem.
Eur. J. 1999, 5, 599.
(28) For other examples of this transformation, see ref 8 and: Fujiwara,
K.; Awakura, D.; Tsunashima, M.; Nakamura, A.; Honma, T.; Murai, A.
J. Org. Chem. 1999, 64, 2616.
(29) Minor product from this coupling is diastereomeric at C(20). We
believe that this material comes from the direct reduction of 19.
(30) Another possibility would involve an RCM sequence. We have
previously demonstrated the generation of oxepenes using RCM; see ref
6c.
Supporting Information Available: Experimental pro-
cedures and spectroscopic data for compounds 2-5, 15-
18, and 20-24. This material is available free of charge via
the Internet at http://pubs.acs.org.
OL034100W
(31) Winkler, J. D.; Rouse, M. B.; Greaney, M. F.; Harrison, S. J.; Jeon,
Y. T. J. Am. Chem. Soc. 2002, 124, 9726.
(32) Rainier, J. D.; Allwein, S. P. Tetrahedron Lett. 1998, 39, 9601.
916
Org. Lett., Vol. 5, No. 6, 2003