Total synthesis of (-)-gambierol

Article abstract of DOI:10.1021/ja028167a

The first total synthesis of (-)-gambierol (1), a marine polycyclic ether toxin, has been achieved. Key features of the successful synthesis include (1) a convergent union of the ABC and EFGH ring fragments (5 and 6, respectively) via our developed B-alky

Full text of DOI:10.1021/ja028167a

Published on Web 11/20/2002
Total Synthesis of (-)-Gambierol
Haruhiko Fuwa,† Noriko Kainuma,† Kazuo Tachibana,† and Makoto Sasaki*,‡
Contribution from the Department of Chemistry, Graduate School of Science,
The UniVersity of Tokyo, and CREST, Japan Science and Technology Corporation,
Hongo, Bunkyo-ku, Tokyo 113-0033, Japan, and Laboratory of Biostructural Chemistry,
Graduate School of Life Science, Tohoku UniVersity, Tsutsumidori-Amamiya,
Aoba-ku, Sendai 981-8555, Japan
Received August 16, 2002. Revised Manuscript Received October 3, 2002
Abstract: The first total synthesis of (-)-gambierol (1), a marine polycyclic ether toxin, has been achieved.
Key features of the successful synthesis include (1) a convergent union of the ABC and EFGH ring fragments
(5 and 6, respectively) via our developed B-alkyl Suzuki-Miyaura cross-coupling strategy leading to the
octacyclic polyether core 4 and (2) a late-stage introduction of the sensitive triene side chain by use of
Pd(PPh3)4/CuCl/LiCl-promoted Stille coupling. The ABC ring fragment 5 was synthesized in a linear manner
(B f AB f ABC), wherein the A ring was formed by intramolecular hetero-Michael reaction and the C ring
was constructed via 6-endo cyclization of hydroxy epoxide 7. An improved synthetic entry to the EFGH
ring fragment 6 is also described, in which SmI2-induced reductive cyclization methodology was applied to
the stereoselective construction of the F and H rings, leading to 6 with remarkable overall efficiency.
Stereoselective hydroboration of 5 and subsequent Suzuki-Miyaura coupling with 6 provided endocyclic
enol ether 45 in high yield, which was then converted to octacyclic polyether core 4. Careful choice of the
global deprotection stage was a key element for the successful total synthesis. Functionalization of the H
ring and global desilylation gave (Z)-vinyl bromide 2. Finally, cross-coupling of 2 with (Z)-vinyl stannane 3
under Corey’s Pd(PPh3)4/CuCl/LiCl-promoted Stille conditions completed the total synthesis of (-)-gambierol
(1).
Introduction
The fused polycyclic ether class of marine natural products,
exemplified by brevetoxins, ciguatoxins, and maitotoxin, has
attracted a great deal of attention among chemists due to their
complex molecular architecture as well as potent and diverse
biological acitivities.1,2 In 1993, Yasumoto and co-workers
reported the isolation of gambierol (1) as a toxic constituent
from the cultured cells of the ciguatera causative dinoflagellate
Gambierdiscus toxicus. The gross structure and relative stereo-
chemistry have been established by extensive NMR studies,3
and the absolute configuration was subsequently determined by
an application of a chiral anisotropic reagent.4 The structure
consists of a trans-fused octacyclic polyether core containing
18 stereogenic centers and a partially skipped triene side chain
including a conjugated (Z,Z)-diene system (Figure 1). Gambierol
exhibits potent toxicity against mice at 50 µg/kg (ip), and its
Figure 1. Structure of gambierol (1).
symptoms caused in mice resemble those shown for ciguatox-
ins,5 the principal toxin which is a very widespread seafood
poisoning. This finding implies that gambierol is also responsible
for ciguatera fish poisoning. However, the extremely limited
availability of this toxin from natural sources has hampered
detailed biological studies, including the precise biochemical
mode of action. Therefore, supply of useful quantities of this
natural product by practical chemical synthesis is strongly
demanded, and a number of substantial efforts toward the total
synthesis of gambierol have been reported to date.6,7 Herein we
describe a full account of the first total synthesis of gambierol,8
featuring our B-alkyl Suzuki-Miyaura coupling strategy for the
convergent synthesis of polycyclic ethers.9,10 The synthesis
* To whom correspondence should be addressed: e-mail msasaki@
bios.tohoku.ac.jp.
† The University of Tokyo and CREST, JST.
‡ Tohoku University.
(1) For reviews on marine polycyclic ether toxins, see (a) Yasumoto, T.; Murata,
M. Chem. ReV. 1993, 93, 1897. (b) Scheuer, P. J. Tetrahedron 1994, 50,
3. (c) Murata, M.; Yasumoto, T. Nat. Prod. Rep. 2000, 17, 293. (d)
Yasumoto, T. Chem. Rec. 2001, 1, 228.
(5) (a) Murata, M.; Legrand, A.-M.; Ishibashi, Y.; Fukui, M.; Yasumoto, T. J.
Am. Chem. Soc. 1989, 111, 8929. (b) Murata, M.; Legrand, A.-M.; Ishibashi,
Y.; Fukui, M.; Yasumoto, T. J. Am. Chem. Soc. 1990, 112, 4380. (c) Murata,
M.; Legrand, A.-M.; Scheuer, P. J.; Yasumoto, T. Tetrahedron Lett. 1992,
33, 525. (d) Satake, M.; Morohashi, A.; Oguri, H.; Oishi, T.; Hirama, M.;
Harada, N.; Yasumoto, T. J. Am. Chem. Soc. 1997, 119, 11325. (d)
Yasumoto, T.; Igarashi, T.; Legrand, A.-M.; Cruchet, P.; Chinain, M.; Fujita,
T.; Naoki, H. J. Am. Chem. Soc. 2000, 122, 4988 and references therein.
(2) For reviews on synthesis of polycyclic ethers, see (a) Alvarez, E.; Candenas,
M.-L.; Pe´rez, R.; Ravelo, J. L.; Mart´ın, J. D. Chem. ReV. 1995, 95, 1953.
(b) Mori, Y. Chem. Eur. J. 1997, 3, 849. (c) Marmsa¨ter, F. P.; West, F. G.
Chem. Eur. J. 2002, 8, 4347.
(3) Satake, M.; Murata, M.; Yasumoto, T. J. Am. Chem. Soc. 1993, 115, 361.
(4) Morohashi, A.; Satake, M.; Yasumoto, T. Tetrahedron Lett. 1999, 40, 97.
9
10.1021/ja028167a CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 14983-14992
14983

A R T I C L E S
Fuwa et al.
Scheme 1. Synthesis Plan of Gambierol (1)
Scheme 2. Synthesis Plan of the ABC Ring System of Gambierol
Scheme 1. We planned to construct the triene side chain at a
late stage of the synthesis due to its expected labile nature. A
Stille coupling protocol11 for the C33-C34 bond formation
between (Z)-vinyl bromide 2 and the known (Z)-vinyl stannane
312 reported as a model study by Kadota, Yamamoto, and co-
workers6c was chosen as a promising candidate for the construc-
tion of the triene side chain. The vinyl bromide 2 should be
available via functionalization of the H ring from the precursor
octacyclic polyether core 4. The key intermediate 4 was en-
visaged to be synthesized by a convergent union of the ABC
ring exocyclic enol ether 5 and the EFGH ring ketene acetal
phosphate 6 via the Suzuki-Miyaura coupling tactic developed
in our laboratory.10
We planned to synthesize the ABC ring fragment 5 in a linear
manner (B f AB f ABC) as shown in Scheme 2. Thus,
formation of the C ring would be performed via an acid-induced
6-endo cyclization of hydroxy epoxide 7.13 The stereoselective
construction of the A ring was envisioned to be achieved by an
intramolecular hetero-Michael reaction of 8, which should be
available from the known compound 9.14
Our previous approach to the EFGH ring fragment 6 relied
on the B-alkyl Suzuki-Miyaura coupling of the F and G rings.7b
However, the earlier synthesis lacked overall efficiency (42 steps
in the longest linear sequence and 6% overall yield from the
corresponding methyl ester of 14), which prompted us to explore
an alternative route for the synthesis of 6. Recently, Nakata and
co-workers have reported an iterative and efficient approach
for the stereoselective construction of trans-fused six- and seven-
membered ether rings based on an SmI2-induced reductive
cyclization.6h,15 The most attractive feature of the Nakata
protocol is that 1,3-diaxial angular methyl groups could be
stereoselectively introduced. We thus envisioned a second-
generation strategy, in which this SmI2-induced cyclization
methodology was utilized for the construction of the F and H
rings (Scheme 3). As previously reported,7b the E ring of 6 was
to be constructed as a lactone form, and construction of the F
ring was envisioned to be accessible via an SmI2-induced
involves as another key feature a late-stage installation of the
sensitive triene side chain through Pd(PPh3)4/CuCl/LiCl-
promoted Stille coupling. Our convergent and flexible strategy
employed in the present total synthesis will provide easy access
to structural analogues of gambierol for biological evaluation.
Retrosynthetic Analysis. Retrosynthetic analysis that we
employed for the total synthesis of gambierol (1) is outlined in
(6) (a) Kadota, I.; Park, C.-H.; Ohtaka, M.; Oguro, N.; Yamamoto, Y.
Tetrahedron Lett. 1998, 39, 6365. (b) Kadota, I.; Kadowaki, C.; Yoshida,
N.; Yamamoto, Y. Tetrahedron Lett. 1998, 39, 6369. (c) Kadota, I.; Ohno,
A.; Matsukawa, Y.; Yamamoto, Y. Tetrahedron Lett. 1998, 39, 6373. (d)
Kadowaki, C.; Chan, P. W. H.; Kadota, I.; Yamamoto, Y. Tetrahedron
Lett. 2000, 41, 5769. (e) Kadota, I.; Ohno, A.; Matsuda, K.; Yamamoto,
Y. J. Am. Chem. Soc. 2001, 123, 6702. (f) Kadota, I.; Takamura, H.; Sato,
K.; Yamamoto, Y. Tetrahedron Lett. 2001, 42, 4729. (g) Sakamoto, Y.;
Matsuo, G.; Matsukura, H.; Nakata, T. Org. Lett. 2001, 3, 2749. (h) Cox,
J. M.; Rainier, J. D. Org. Lett. 2001, 3, 2919. (i) Kadota, I.; Park, C.-H.;
Sato, K.; Yamamoto, Y. Tetrahedron Lett. 2001, 42, 6195. (j) Kadota, I.;
Kadowaki, C.; Takamura, H.; Yamamoto, Y. Tetrahedron Lett. 2001, 42,
6199. (k) Kadota, I.; Kadowaki, C.; Park, C.-H.; Takamura, H.; Sato, K.;
Chan, P. W. H.; Thorand, S.; Yamamoto, Y. Tetrahedron 2002, 58, 1799.
(l) Kadota, I.; Ohno, A.; Matsuda, K.; Yamamoto, Y. J. Am. Chem. Soc.
2002, 124, 3562. (m) Kadota, I.; Takamura, H.; Sato, K.; Yamamoto, Y.
J. Org. Chem. 2002, 67, 3494.
(7) (a) Fuwa, H.; Sasaki, M.; Tachibana, K. Tetrahedron Lett. 2000, 41, 8371.
(b) Fuwa, H.; Sasaki, M.; Tachibana, K. Tetrahedron 2001, 57, 3019. (c)
Fuwa, H.; Sasaki, M.; Tachibana, K. Org. Lett. 2001, 3, 3549.
(8) Part of this work has been communicated: see Fuwa, H.; Sasaki, M.; Satake,
M.; Tachibana, K. Org. Lett. 2002, 4, 2981.
(9) For reviews on Suzuki-Miyaura cross-coupling reaction, see: (a) Suzuki,
A.; Miyaura, N. Chem. ReV. 1995, 95, 2457. (b) Suzuki, A. In Metal-
catalyzed Cross-coupling Reactions; Diederich, F., Stang, P. J., Eds.; Wiley-
VCH: Weinheim, Germany, 1998; pp 49-97. (c) Suzuki, A. J. Organomet.
Chem. 1999, 576, 147. (d) Chemler, S. R.; Trauner, D.; Danishefsky, S. J.
Angew. Chem., Int. Ed. 2001, 40, 4544.
(10) (a) Sasaki, M.; Fuwa, H.; Inoue, M.; Tachibana, K. Tetrahedron Lett. 1998,
39, 9027. (b) Sasaki, M.; Fuwa, H.; Ishikawa, M.; Tachibana, K. Org. Lett.
1999, 1, 1075. (c) Sasaki, M.; Noguchi, K.; Fuwa, H.; Tachibana, K.
Tetrahedron Lett. 2000, 41, 1425. (d) Takakura, H.; Noguchi, K.; Sasaki,
M.; Tachibana, K. Angew. Chem., Int. Ed. 2001, 40, 1090. (e) Sasaki, M.;
Ishikawa, M.; Fuwa, H.; Tachibana, K. Tetrahedron 2002, 58, 1889. (f)
Sasaki, M.; Tsukano, C.; Tachibana, K. Org. Lett. 2002, 4, 1747. (g)
Takakura, H.; Sasaki, M.; Honda, S.; Tachibana, K. Org. Lett. 2002, 4,
2771.
(11) (a) Stille, J. K. Angew. Chem., Int. Ed. Engl. 1986, 25, 508. (b) For a
comprehensive review, see Farina, V.; Krishnamurthy, V.; Scott, W. J. In
Organic Reactions; Paquette, L. A., Ed.; John Wiley and Sons: New York,
1997; Vol. 50, pp 1-652.
(12) (a) Shirakawa, E.; Yamasaki, K.; Yoshida, H.; Hiyama, T. J. Am. Chem.
Soc. 1999, 121, 10221. (b) Matsukawa, Y.; Asao, N.; Kitahara, N.;
Yamamoto, Y. Tetrahedron 1999, 55, 3779.
(13) Nicolaou, K. C.; Prasad, C. V. C.; Somers, P. K.; Hwang, C.-K. J. Am.
Chem. Soc. 1989, 111, 5330.
(14) Nicolaou, K. C.; Nugiel, D. A.; Couladouros, E.; Hwang, C.-K. Tetrahedron
1990, 46, 4517.
(15) (a) Hori, N.; Matsukura, H.; Matsuo, G.; Nakata, T. Tetrahedron Lett. 1999,
40, 2811. (b) Hori, N.; Matsukura, H.; Nakata, T. Org. Lett. 1999, 1, 1099.
(c) Matsuo, G.; Hori, N.; Nakata, T. Tetrahedron Lett. 2000, 41, 7673.
9
14984 J. AM. CHEM. SOC. VOL. 124, NO. 50, 2002

Total Synthesis of (−)-Gambierol
A R T I C L E S
Scheme 3. Second-Generation Synthesis Plan of the EFGH Ring
Scheme 4. Synthesis of Olefin 20a
System of Gambierol
reductive cyclization of methyl ketone 10. The G ring of the
precursor 11 could be readily constructed via a 6-endo cycliza-
tion of hydroxy epoxide 12, which was again conceived to be
prepared by an SmI2-induced cyclization of 13. In turn, the
aldehyde 13 should be easily derived from the known ester 14.16
Synthesis of the ABC Ring Fragment 5. The synthesis of
the ABC ring fragment 5 commenced with the known olefin
914 (Scheme 4). Oxidative cleavage of the double bond [OsO4,
N-methylmorpholine N-oxide (NMO), and then NaIO4], Hor-
ner-Wadsworth-Emmons reaction, and reduction with di-
isobutylaluminum hydride (DIBALH) led to allylic alcohol 15
in 87% overall yield. Sharpless asymmetric epoxidation of 15
with (-)-diethyl tartrate as a chiral auxiliary gave hydroxy
epoxide, which, upon treatment with Red-Al, afforded 1,3-diol
1617 in good overall yield as a single stereoisomer. The
differentiation of the two resultant hydroxyl groups was
accomplished via anisylidene acetal formation followed by
regioselective reductive cleavage with DIBALH (CH2Cl2, -40
f 0 °C), giving 17 in 80% yield for the two steps. Oxidation
of alcohol 17 with SO3‚pyridine gave an aldehyde, which was
then subjected to methyl triphenylphosphoranylidene acetate to
give R,â-unsaturated ester 18 in quantitative yield over two
steps. Subsequent desilylation of 18 turned out to be somewhat
problematic due to the sensitive functionalities and protective
groups present in 18. After some experimentation, it was found
that the desired desilylated product 8 could be obtained in high
yield by treatment of 18 with tetra-n-butylammonium fluoride
(TBAF) buffered with acetic acid.
a Reagents and conditions: (a) OsO4, NMO, 1:1 THF/H2O, rt; then
NaIO4, rt. (b) (i-PrO)2P(O)CH2CO2Et, KOt-Bu, THF, -78 f 0 °C. (c)
DIBALH, CH2Cl2, -78 °C, 87% (three steps). (d) (-)-DET, Ti(Oi-Pr)4,
t-BuOOH, 4 Å MS, CH2Cl2, -28 °C. (e) Red-Al, THF, -40 f 0 °C, quant
(two steps). (f) p-MeOC6H4CH(OMe)2, PPTS, CH2Cl2, rt. (g) DIBALH,
CH2Cl2, -40 f 0 °C, 80% (two steps). (h) SO3‚pyridine, Et3N, 1:1 DMSO/
CH2Cl2, 0 °C. (i) Ph3PdCHCO2Me, toluene, 80 °C, quant (two steps). (j)
TBAF, HOAc, THF, rt f 35 °C, 91%. (k) NaH, THF, 0 °C f rt, 86%. (l)
DIBALH, CH2Cl2, -78 °C. (m) Ph3P+CH3Br-, NaHMDS, THF, 0 °C, 91%
(two steps).
Figure 2. NOE experiments on compound 20. The PMB group is omitted
for clarity.
in 91% overall yield. At this stage, the relative stereochemistry
was unambiguously determined by NOE experiments as shown
in Figure 2.
We next executed an intramolecular hetero-Michael reaction
to construct the A ring. This process could be easily ac-
complished by exposure of 8 to sodium hydride in THF at room
temperature, giving the desired tricylic ether 19 in 86% yield
as the sole product. Partial reduction of the ester moiety of 19
(1.1 equiv of DIBALH, CH2Cl2, -78 °C, 20 min) gave an
aldehyde, which upon Wittig methylenation furnished olefin 20
Hydroboration of 20 with 9-BBN followed by oxidative
workup gave an alcohol, which was protected as its benzyl ether
to afford 21 in 94% yield for the two steps (Scheme 5).
Oxidative removal of the p-methoxybenzyl (PMB) group and
reprotection as the benzyl ether gave bis(benzyl ether) 22 in
93% yield for the two steps. Removal of the benzylidene acetal
under acidic conditions led to diol 23. Bis-silylation followed
by selective liberation of the C1218 primary hydroxyl under
acidic conditions led to alcohol 24. Oxidation of 24 with a
(16) Nicolaou, K. C.; Wallace, P. A.; Shi, S.; Ouellette, M. A.; Bunnage, M.
E.; Gunzner, J. L.; Agrios, K. A.; Shi, G.-Q.; Yang, Z. Chem. Eur. J. 1999,
5, 618.
(17) Finan, J.; Kishi, Y. Tetrahedron Lett. 1982, 23, 2719.
9
J. AM. CHEM. SOC. VOL. 124, NO. 50, 2002 14985

A R T I C L E S
Fuwa et al.
Scheme 5. Synthesis of AB Ring Fragment 25a
Scheme 6. Synthesis of ABC Ring Fragment 5a
a Reagents and conditions: (a) 9-BBN, THF, rt; then aq NaHCO3, 30%
H2O2, rt. (b) BnBr, KOt-Bu, n-Bu4NI, THF, rt, 94% (two steps). (c) DDQ,
pH 7 phosphate buffer, CH2Cl2, rt. (d) BnBr, KOt-Bu, n-Bu4NI, THF, rt,
93% (two steps). (e) p-TsOH, 9:1 MeOH/CHCl3, rt, 95%. (f) TBSOTf, 2,6-
lutidine, CH2Cl2, 0 °C. (g) CSA, MeOH, rt, 92% (two steps). (h) TPAP,
NMO, 4 Å MS, CH2Cl2, rt, 95%. (i) Tebbe reagent, THF, 0 °C, 90%.
catalytic amount of tetra-n-propylammonium perruthenate (TPAP)
and NMO19 gave an aldehyde, which was then treated with
Tebbe reagent (THF, 0 °C)20 to afford olefin 25 in 86% overall
yield from 24. In this reaction, conventional Wittig methylena-
tion resulted in a poor yield of 25.
Construction of the C ring and completing the synthesis of
the ABC ring fragment 5 are summarized in Scheme 6.
Hydroboration of 25 with 9-BBN followed by oxidative workup
gave an alcohol, which was oxidized to the aldehyde and
subsequently homologated by Horner-Wadsworth-Emmons
reaction to give R,â-unsaturated ester 26 in 90% overall yield.
DIBALH reduction led to allylic alcohol 27 in nearly quantita-
tive yield. Subsequent Sharpless asymmetric epoxidation of 27
with (+)-diethyl tartrate was somewhat unsatisfactory with
regard to its diastereoselectivity and gave an inseparable mixture
(ca. 6:1) of diastereomers, favoring the desired R-epoxide 28.
On the other hand, oxidation with m-chloroperbenzoic acid
(CH2Cl2, 0 °C) led exclusively to the desired 28 in almost
quantitative yield.14 Oxidation of 28 was followed by Wittig
methylenation of the derived aldehyde to afford olefin 29 in
87% yield for the two steps. Removal of the TBS group with
TBAF gave hydroxy epoxide 7, which was exposed to mild
acidic conditions (PPTS, CH2Cl2, room temperature) to effect
6-endo cyclization, giving tricyclic ether 30 in 96% yield for
the two steps. The relative stereochemistry of 30 was unambigu-
ously established on the basis of NOE experiments and a large
coupling constant of J13,14 ) 9.2 Hz (Figure 3).
a Reagents and conditions: (a) 9-BBN, THF, rt; then aq NaHCO3,
30% H2O2, rt. (b) SO3‚pyridine, Et3N, 1:1 DMSO/CH2Cl2, 0 °C. (c)
(i-PrO)2P(O)CH2CO2Et, KOt-Bu, THF, -78 f 0 °C, 90% (three steps).
(d) DIBALH, CH2Cl2, -78 °C, 98%. (e) mCPBA, CH2Cl2, 0 °C, 99%. (f)
SO3‚pyridine, Et3N, 1:1 DMSO/CH2Cl2, 0 °C. (g) Ph3P+CH3Br-, NaHMDS,
THF, 0 °C, 87% (two steps). (h) TBAF, THF, rt, 98%. (i) PPTS, CH2Cl2,
rt, 98%. (j) PMBCl, KOt-Bu, n-Bu4NI, THF, rt. (k) OsO4, NMO, 1:1 THF/
H2O, rt; then NaIO4, rt. (l) NaBH4, MeOH, 0 °C f rt, 86% (three steps).
(m) I2, PPh3, imidazole, benzene, rt, 95%. (n) KOt-Bu, THF, 0 °C, 91%.
Figure 3. Determination of stereochemistry of compound 30. Benzyl groups
are omitted for clarity.
double bond (OsO4, NMO, and then NaIO4), and subsequent
reduction of the derived aldehyde led to alcohol 31 in good
overall yield. Iodination of the primary alcohol under standard
conditions followed by base treatment (KOt-Bu, THF, 0 °C)
afforded the ABC ring exocyclic enol ether 5 in 86% yield for
the two steps. The overall sequence proceeded in 36 steps from
9 and in 18% overall yield, and thus the ABC ring fragment 5
could be synthesized in multigram quantities.
Conversion of tricyclic ether 30 to the ABC ring fragment 5
was carried out in a straightforward manner. Protection of 30
as the PMB ether was followed by oxidative cleavage of the
(18) The numbering of carbon atoms of all compounds in this paper corresponds
to that of gambierol.
(19) Ley, S. V.; Norman, J.; Griffith, W. P.; Marsden, S. P. Synthesis 1994,
639.
(20) Tebbe, F. N.; Parshall, G. W.; Reddy, G. S. J. Am. Chem. Soc. 1978, 100,
3611.
9
14986 J. AM. CHEM. SOC. VOL. 124, NO. 50, 2002

Total Synthesis of (−)-Gambierol
A R T I C L E S
Scheme 7. Synthesis of GH Ring Fragment 11a
Figure 4. Structure determination of compound 11.
Scheme 8. Synthesis of EFGH Ring Fragment 6a
a Reagents and conditions: (a) O3, 1:1 MeOH/CH2Cl2, -78 °C; then
NaBH4, 0 °C, 96%. (b) I2, PPh3, imidazole, THF, rt, 99%. (c) 1,3-Dithiane,
n-BuLi, THF, -20 °C; then 33, 0 °C. (d) TBAF, THF, rt, 95% (two steps).
(e) Ethyl propiolate, NMM, CH2Cl2, rt. (f) MeI, NaHCO3, aq CH3CN, rt,
94% (two steps). (g) SmI2, MeOH, THF, rt, 70%. (h) DIBALH, CH2Cl2,
-78 °C; (i) Ph3PdC(Me)CO2Et, toluene, 80 °C, 97% (two steps). (j)
TBSOTf, 2,6-lutidine, CH2Cl2, 0 °C, quant. (k) DIBALH, CH2Cl2, -78
°C. (l) (-)-DET, Ti(Oi-Pr)4, t-BuOOH, 4 Å MS, CH2Cl2, -20 °C, 97%
(two steps). (m) SO3‚pyridine, Et3N, 1:1 DMSO/CH2Cl2, 0 °C. (n)
Ph3P+CH3Br-, NaHMDS, THF, 0 °C, 94% (two steps). (o) TBAF, THF,
rt. (p) PPTS, CH2Cl2, rt, 88% (two steps).
a Reagents and conditions: (a) TBSOTf, 2,6-lutidine, CH2Cl2, 0 °C. (b)
EtSH, Zn(OTf)2, NaHCO3, CH2Cl2, rt, 85% (two steps). (c) Me2C(OMe)2,
PPTS, CH2Cl2, rt, 92%. (d) 9-BBN, THF, rt; then aq NaHCO3, 30% H2O2,
rt. (e) SO3‚pyridine, Et3N, 1:1 DMSO/CH2Cl2, 0 °C. (f) MeMgBr, toluene,
-78 °C. (g) TPAP, NMO, 4 Å MS, CH2Cl2, rt, 86% (four steps). (h) TBAF,
THF, rt. (i) Ethyl propiolate, NMM, CH2Cl2, rt, 99% (two steps). (j) SmI2,
MeOH, THF, 0 °C, 87%. (k) TMSOTf, 2,6-lutidine, CH2Cl2, 0 °C, 96%.
(l) DIBALH, CH2Cl2, -78 °C. (m) Ph3PdCHCO2Bn, toluene, rt, 95% (two
steps). (n) TBAF, HOAc, THF, rt, 93%. (o) H2, Pd/C, 2:1 MeOH/THF, rt.
(p) 2,4,6-trichlorobenzoyl chloride, Et3N, 1:1 THF/toluene, rt; then DMAP,
toluene, 110 °C, 99% (two steps). (q) KHMDS (3 equiv), (PhO)2P(O)Cl
(10 equiv), 10:1 THF/HMPA (10 mM), -78 °C, quant.
Second-Generation Synthesis of EFGH Ring Fragment 6.
An improved second-generation synthesis of the EFGH ring
fragment 6 started with the known ester 1416 (Scheme 7).
Ozonolysis followed by reductive workup with NaBH4 afforded
alcohol 32, which was then iodinated under the standard
conditions to give 33 in 99% yield. Treatment of iodide 33 with
lithiated dithiane (1,3-dithiane, n-BuLi, THF, -20 °C) and
subsequent removal of the silyl protective group gave alcohol
34 in 95% yield for the two steps. The â-alkoxyacrylate moiety
was next incorporated by treatment with ethyl propiolate and
N-methylmorpholine (NMM), and subsequently the dithioacetal
moiety was removed to give aldehyde 13 (91% yield for the
two steps). As expected, reductive cyclization of 13 with SmI2
in the presence of methanol (THF, room temperature) proceeded
stereoselectively and the desired γ-lactone 35 was obtained in
70% yield along with a minute amount of the hydroxy ester
(3% yield, not shown).15 Lactone 35 was then converted to R,â-
unsaturated ester 36 in high overall yield by DIBALH reduction
and Wittig reaction of the derived lactol. Protection of the
secondary hydroxyl within 36 as the TBS ether gave 37
quantitatively. DIBALH reduction of the ester moiety gave
allylic alcohol, which was subjected to Sharpless asymmetric
epoxidation with (-)-diethyl tartrate as a chiral auxiliary to
afford hydroxy epoxide 38 in high overall yield as a single
stereoisomer. Oxidation with SO3‚pyridine followed by Wittig
methylenation of the derived aldehyde afforded vinyl epoxide
39 in 94% yield for the two steps. After desilylation, 6-endo
cylization of the resultant alcohol 12 was performed by exposure
to PPTS in CH2Cl2 at room temperature to give the GH ring
system 11 in 88% yield for the two steps. The relative
stereochemistry of 11 was determined on the basis of 1H NMR
coupling constants and NOE experiments (Figure 4).
Protection of the secondary hydroxyl within 11 as its TBS
ether followed by replacement of the benzylidene acetal with
the acetonide group led to 40 in 78% overall yield (Scheme 8).
Olefin 40 was then converted to methyl ketone 41 by a routine
four-step sequence of reactions. Thus, hydroboration of 40 with
9-BBN followed by oxidative workup gave primary alcohol,
which was oxidized with SO3‚pyridine to the aldehyde. Treat-
ment with MeMgBr (toluene, -78 °C) and further oxidation of
the derived secondary hydroxyl with TPAP/NMO afforded
methyl ketone 41 (86% overall yield from 40). Desilylation
followed by treatment with ethyl propiolate and NMM afforded
9
J. AM. CHEM. SOC. VOL. 124, NO. 50, 2002 14987

A R T I C L E S
Fuwa et al.
Scheme 9. Synthesis of Octacyclic Polyether Core 4 via B-alkyl
â-alkoxy acrylate 10 in nearly quantitative yield for the two
steps. Exposure of 10 to SmI2 (methanol, THF, 0 °C) again
effected stereoselective ring closure of the F ring furnished with
1,3-diaxial angular methyl groups to give the FGH ring system
42 in 87% yield as a single stereoisomer. Protection of the
tertiary hydroxyl as the TMS ether (TMSOTf, 2,6-lutidine),
followed by half-reduction of the ester moiety and Wittig
reaction with benzyl triphenylphosporanylidene acetate, afforded
R,â-unsaturated benzyl ester 43 in excellent overall yield (93%).
Desilylation by TBAF buffered with acetic acid was followed
by concomitant removal of the benzyl group and saturation of
the double bond to provide hydroxy acid, which was then
subjected to Yamaguchi lactonization.21 These sequences pro-
vided lactone 44 in 93% overall yield. Spectroscopic data (1H
and 13C NMR, [R]D, IR, and HRMS) of 44 thus prepared were
exactly matched with those of the previously reported product.7b
Finally, conversion of 44 into the EFGH ring ketene acetal
phosphate 6 was accomplished by using the modified Nicolaou
protocol [3 equiv of KHMDS, 10 equiv of (PhO)2P(O)Cl, 10:1
THF/HMPA (10 mM), -78 °C, quant].22 Thus, we completed
an improved synthesis of the EFGH ring fragment 6 in 33 steps
and 22% overall yield from the known compound 14. This
remarkably efficient synthesis allowed preparation of 6 in
multigram quantities.
Suzuki-Miyaura Couplinga
Convergent Union of the Two Key Intermediates: Syn-
thesis of the Octacyclic Polyether Core. With the requisite
key intermediates 5 and 6 in hand, we next turned our attention
to the crucial coupling of these components via the B-alkyl
Suzuki-Miyaura coupling strategy (Scheme 9). Hydroboration
of the ABC ring exocyclic enol ether 5 with 9-BBN in THF at
room temperature produced the corresponding alkylborane,
which was in situ reacted with the EFGH ring ketene acetal
phosphate 6 (1.4 equiv) in the presence of aqueous Cs2CO3 (3
equiv) and PdCl2(dppf) (50 mol %) in DMF at 50 °C for 22 h.
The desired cross-coupled product 45 was obtained in gratifying
86% yield. This remarkable yield represents the power and
feasibility of our B-alkyl Suzuki-Miyaura coupling chemistry.
Treatment of 45 with BH3‚THF (THF, room temperature, 1
h) followed by oxidative workup led to the desired alcohol 46r
in 87% yield as the sole product. Oxidation of 46r with TPAP/
NMO gave rise to ketone 47r in 98% yield. At this stage, the
relative stereochemistry of ketone 47r was unambiguously
established by NMR analyses. The large coupling constant of
J13,14 ) 9.0 Hz confirmed the trans relationship of H13 and
H14 protons, whereas the syn relationship of H16 and C21
methyl group was established by NOESY experiments. In
contrast, exposure of 45 with BH3‚SMe2 (THF, room temper-
ature, 3 h) followed by oxidative workup gave a mixture of
diastereomers 46r and 46â (71% and 11% yields, respectively),
which were readily separable by flash chromatography (Scheme
10). Oxidation of diastereomer 46â with TPAP/NMO gave
epimeric ketone 47â in 90% yield,23 which upon treatment
with DBU (toluene, 110 °C, 1 day) gave a 3:2 mixture of 47r
and 47â, favoring the thermodynamically more stable isomer
a Reagents and conditions: (a) 9-BBN, THF, rt; then 6, aq Cs2CO3,
PdCl2(dppf)‚CH2Cl2, DMF, 50 °C, 86%. (b) BH3‚THF, THF, rt; then aq
NaOH, 30% H2O2, rt, 87%. (c) TPAP, NMO, 4 Å MS, CH2Cl2, rt, 98%.
(d) DDQ, pH 7 phosphate buffer, CH2Cl2, rt. (e) EtSH, Zn(OTf)2, CH2Cl2,
rt. (f) Ac2O, Et3N, DMAP, CH2Cl2, rt, 75% (three steps). (g) Ph3SnH, AIBN,
toluene, 110 °C, 95%.
47r. From these results, we reached an intriguing conclusion
that hydroboration of endocyclic enol ether 45 occurred from
the sterically more hindered R-face of the molecule to afford
46r.
Oxidative removal of the PMB protective group within 47r
gave the corresponding hemiketal, which was then exposed to
ethanethiol in the presence of zinc trifluoromethanesulfonate
to effect mixed-thioketal formation with concomitant loss of
the acetonide group (Scheme 9). The liberated hydroxyl groups
were then acylated to afford diacetate 48 in 75% yield over the
three steps. Finally, desulfurization under radical reduction
conditions (Ph3SnH, AIBN, toluene, 110 °C)24 proceeded cleanly
to furnish the octacyclic polyether core 4 in 95% yield.
Model Studies for the Construction of Triene Side Chain.
Having succeeded in obtaining the key intermediate 4, all that
is necessary to complete the total synthesis of 1 is functional-
ization of the H ring and stereoselective installation of the
sensitive triene side chain. Especially, stereoselective construc-
tion of a partially skipped triene side chain that includes a
(21) (a) Yamaguchi, M.; Inanaga, J.; Hirata, K.; Sasaki, H.; Katsuki, T. Bull.
Chem. Soc. Jpn. 1979, 52, 1989. (b) Mulzer, J.; Mareski, P. A.; Bushmann,
J.; Luger, P. Synthesis 1992, 215.
(22) Nicolaou, K. C.; Shi, G.-Q.; Gunzner, J. L.; Ga¨rtner, P.; Yang, Z. J. Am.
Chem. Soc. 1997, 119, 5467.
(23) The 1H NMR spectrum of epimeric ketone 47â was different from that of
ketone 47r, indicating that we can rule out the possibility of epimerization
during the oxidation with TPAP/NMO.
(24) Nicolaou, K. C.; Prasad, C. V. C.; Hwang, C.-K.; Duggan, M. E.; Veale,
C. A. J. Am. Chem. Soc. 1989, 111, 5321.
9
14988 J. AM. CHEM. SOC. VOL. 124, NO. 50, 2002

Total Synthesis of (−)-Gambierol
A R T I C L E S
Scheme 10 a
Scheme 11. Synthesis of Fully Functionalized FGH Ring Systema
a Reagents and conditions: (a) BH3‚SMe2, THF, rt; then aq NaOH, 30%
H2O2, rt; 46r, 71%; 46â, 11%. (b) TPAP, NMO, 4 Å MS, CH2Cl2, rt,
90%. (c) DBU, toluene, 110 °C, quant (47r:47â ) 3:2).
conjugated (Z,Z)-diene system represents a formidable synthetic
challenge. As described in the synthetic plan, a modified Stille
coupling protocol for the C33-C34 bond formation is an
appropriate candidate for the construction of the triene side
chain.
a Reagents and conditions: (a) CSA, MeOH, rt. (b) TBSCl, imidazole,
DMF, 0 °C, 96% (two steps). (c) TPAP, NMO, 4 Å MS, CH2Cl2, rt, 92%.
(d) LiHMDS, TMSCl, Et3N, THF, -78 °C. (e) Pd(OAc)2, MeCN, rt, 92%
(two steps). (f) MeMgBr, toluene, -78 °C, 96%. (g) TBAF, THF, rt, 87%.
(h) SO3‚pyridine, Et3N, 1:1 DMSO/CH2Cl2, 0 °C. (i) CBr4, PPh3, Et3N,
CH2Cl2, 0 °C, 84% (two steps). (j) TMSOTf, Et3N, CH2Cl2, 0 °C, 93%.
(k) n-Bu3SnH, Pd(PPh3)4, benzene, rt, 81%. (l) 3, Pd(PPh3)4, CuCl, LiCl,
1:1 DMSO/THF, 60 °C, 95%.
Initially, we tried to synthesize model compound 59 in order
to establish a feasible route for the completion of the total
synthesis (Scheme 11). Removal of the acetonide group of 497b
under acidic conditions followed by selective protection of the
liberated primary hydroxyl gave alcohol 50 (96% yield for the
two steps), which was oxidized with TPAP/NMO to afford
ketone 51 in 92% yield. Subsequent conversion of 51 into enone
52 turned out to be somewhat problematic. Conventional
selenium-based methodology failed to give 52 (Table 1, entry
1). o-Iodoxybenzoic acid- (IBX-) mediated dehydrogenation
reaction, recently reported by Nicolaou and co-workers,25 was
also ineffective in the present case (entry 2). Exposure of the
kinetically formed lithium enolate of 51 to N-(tert-butyl)-
phenylsulfimidoyl chloride (THF, -78 °C)26 gave 52 in only a
modest yield (entry 3). The Ito-Saegusa protocol27 also gave
a comparable result. Thus, 51 was treated with LiHMDS and
the derived enolate was trapped with TMSCl and Et3N to give
the corresponding enol silyl ether, which, without purification,
was treated with Pd(OAc)2 in acetonitrile at room temperature
to afford 52 in 57% yield (entry 4). After several experiments,
it was discovered that treatment of 51 with LiHMDS in the
presence of TMSCl and Et3N cleanly gave the corresponding
enol silyl ether. Upon oxidation of the enol silyl ether with
Table 1. Conversion of Ketone 51 to Enone 52
entry
reagents and conditions
% yield
1
(a) LiHMDS, THF, -78 °C; then PhSeBr,
-78 °C f rt; (b) 30% H2O2, THF, rt
IBX, 1:1 DMSO/toluene, 55 °C
0
2
3
0
61
LiHMDS, THF, -78 °C;
then PhS(Cl)dNt-Bu, -78 °C
4
5
(a) LiHMDS, THF, -78 °C; then TMSCl,
Et3N, -78 °C; (b) Pd(OAc)2, MeCN, rt
(a) LiHMDS, TMSCl, Et3N, THF,
-78 °C; (b) Pd(OAc)2, MeCN, rt
57
92
Pd(OAc)2, the desired enone 52 was obtained in 92% yield
(entry 5). This remarkable improvement may be ascribed to the
lability of ketone 51 and/or the corresponding lithium enolate
under the reaction conditions. Stereoselective introduction of
the C30 methyl group was accomplished by treatment of 51
(25) Nicolaou, K. C.; Zhong, Y.-L.; Baran, P. S. J. Am. Chem. Soc. 2000, 122,
7596.
(26) Mukaiyama, T.; Matsuo, J.; Kitagawa, H. Chem. Lett. 2000, 1250.
(27) Ito, Y.; Hirao, T.; Saegusa, T. J. Org. Chem. 1978, 43, 1011.
9
J. AM. CHEM. SOC. VOL. 124, NO. 50, 2002 14989

A R T I C L E S
Fuwa et al.
Table 2. Stille Coupling of (Z)-Vinyl Bromide 57 and (Z)-Vinyl Stannane 3a
entry
3 (equiv)
catalyst
ligandb
cocatalystc (equiv)
additive (equiv)
solvent
temp
rt
60 °C
rt
60 °C
60 °C
60 °C
rt
% yield
1
2
3
4
5
6
7
8
9
2.0
2.0
1.5
1.5
1.5
2.0
2.0
2.0
2.0
Pd2(dba)3CHCl3
Pd2(dba)3CHCl3
Pd(PPh3)4
PdCl2(MeCN)2
Pd2(dba)3CHCl3
Pd2(dba)3CHCl3
Pd2(dba)3CHCl3
Pd(PPh3)4
TFP
TFP
CuI (1.6)
CuI (1.0)
DMSO/THF (1:1)
DMSO/THF (1:1)
DMF
49e
50
i-Pr2NEt (10)
NDd
39e
28e
69
TFP
Ph3As
TFP
DMF
DMF
CuTC (10)
CuTC (10)
CuCl (10)
CuCl (10)
DMSO/THF (1:1)
DMSO/THF (1:1)
DMSO/THF (1:1)
DMSO/THF (1:1)
TFP
75
LiCl (12)
LiCl (12)
rt
60 °C
49e
81
Pd(PPh3)4
a All reactions were carried out using Pd catalyst (20 mol%) and ligand (80 mol%) for 2 days. b TFP ) tri(2-furyl)phosphine. c TC ) thiophene-2-
1
carboxylate. d Yield not determined (18% conversion based on H NMR analysis). e Vinyl bromide 57 was not consumed completely and the yield was
1
estimated by H NMR analysis of an inseparable mixture of 57 and 58.
2, entries 1 and 2). Therefore, we reinvestigated the cross-
coupling reaction to establish the optimal conditions. Not
unexpectedly, in the absence of copper(I) salt, only poor yield
and low conversion were observed (entries 3-5). Use of copper-
(I) thiophene-2-carboxylate (CuTC)34 or CuCl advantageously
accelerated the reaction and improved the yield of 58 (entries
6-9). Observed improvements of the Stille coupling by the
addition of CuTC or CuCl could be ascribed to facile in situ
generation of a more reactive organocopper species as a
nucleophile that facilitates the transmetalation step in the
Figure 5. NOE experiments on compound 53. The benzyl and TBS groups
catalytic cycle. Especially, the Pd(PPh3)4/CuCl/LiCl-promoted
are omitted for clarity.
Stille conditions (1:1 DMSO/THF, 60 °C) developed by Corey
and co-workers35 were quite suitable for this process (entry 9).
Notably, under these conditions the cross-coupling reaction
proceeded with retention of olefin stereochemistry. Application
of these optimal conditions to the Stille coupling of (Z)-vinyl
bromide 56 with 3 produced the desired product 59 in 95%
isolated yield (Scheme 11). Again, none of the other isomer
was detected in the 1H NMR spectrum of the reaction mixture.
with MeMgBr in toluene at -78 °C28 to afford tertiary alcohol
53 as a single stereoisomer. Stereochemistry at the C30
quaternary stereocenter was determined by NOE experiments
as shown in Figure 5. Subsequent desilylation with TBAF gave
diol 54 in 84% overall yield from 52. Selective oxidation of
the primary hydroxyl followed by Corey-Fuchs reaction (CBr4,
PPh3, Et3N)29 of the derived aldehyde gave dibromoolefin 55
Completion of Total Synthesis of (-)-Gambierol. With a
reliable precedent for the stereoselective construction of the
(84% yield over two steps). After protection of the tertiary
alcohol as its TMS ether, stereoselective hydrogenolysis of the
30
sensitive triene side chain established, we proceeded forward
to complete the total synthesis with confidence. Removal of
the acetyl groups within 4 followed by selective protection of
the C32 primary hydroxyl and subsequent oxidation of the
remaining secondary hydroxyl with TPAP/NMO then produced
ketone 60 in 69% overall yield (Scheme 12). Treatment of 60
with LiHMDS in the presence of TMSCl and Et3N as described
above gave the corresponding enol silyl ether, which upon im-
mediate exposure to Pd(OAc)2 in acetonitrile at room temper-
ature afforded enone. Subsequent Grignard reaction with
MeMgBr (toluene, -78 °C) furnished tertiary alcohol 61 in
94% overall yield as a single stereoisomer. After protection of
the tertiary hydroxyl as its TBS ether,36 reductive cleavage of
the benzyl ethers by exposure to excess lithium di-tert-
butylbiphenylide (LiDBB),37 followed by protection of the C1
primary hydroxyl as the tert-butyldiphenylsilyl (TBDPS) ether,
dibromoolefin with n-Bu3SnH and Pd(PPh3)4 afforded (Z)-
vinyl bromide 56 in 81% yield, setting the stage for the
construction of the triene side chain.
The Stille coupling of 56 with the known (Z)-vinyl stannane
312 was quite difficult due to low reactivity of the sterically
hindered (Z)-vinyl stannane as well as the (Z)-vinyl bromide.
Moreover, an attempt to bring about the coupling of a simple
model substrate (Z)-vinyl bromide 5731 with 3 under the
modified Stille conditions [Pd2(dba)3‚CHCl3, (2-furyl)3P, CuI,
DMSO/THF32], which was originally reported by Liebeskind
et al.33 and successfully utilized in Yamamoto’s model study,6c
gave the cross-coupled product 58 in only a modest yield (Table
(28) Feng, F.; Murai, A. Chem. Lett. 1992, 1587.
(29) Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 13, 3769.
(30) (a) Uenishi, J.; Kawahama, R.; Shiga, Y.; Yonemitsu, O.; Tsuji, J.
Tetrahedron Lett. 1996, 37, 6759. (b) Uenishi, J.; Kawahama, R.;
Yonemitsu, O.; Tsuji, J. J. Org. Chem. 1996, 61, 5716. (c) Uenishi, J.;
Kawahama, R.; Yonemistu, O.; Tsuji, J. J. Org. Chem. 1998, 63, 8965.
(31) The corresponding iodide was too unstable to be used.
(32) A mixed solvent system of 1:1 DMSO/THF was used in the present study
due to the low solubility of (Z)-vinyl stannane 3 in DMSO.
(33) (a) Liebeskind, L. S.; Fengl, R. W. J. Org. Chem. 1990, 55, 5359. (b)
Farina, V.; Kapadia, S.; Krishnan, B.; Wang, C.; Liebeskind, L. S. J. Org.
Chem. 1994, 59, 5905.
(34) Allred, G. D.; Liebeskind, L. S. J. Am. Chem. Soc. 1996, 118, 2748.
(35) Han, X.; Stolz, B. M.; Corey, E. J. J. Am. Chem. Soc. 1999, 121, 7600.
(36) TMS and TES protective groups could not survive the next reductive
debenzylation with LiDBB.
(37) (a) Freeman, P. K.; Hutchinson, L. L. J. Org. Chem. 1980, 45, 1924. (b)
Ireland, R. E.; Smith, M. G. J. Am. Chem. Soc. 1988, 110, 854.
9
14990 J. AM. CHEM. SOC. VOL. 124, NO. 50, 2002

Total Synthesis of (−)-Gambierol
A R T I C L E S
Scheme 13. Synthesis of Fully Protected Gambierola
Scheme 12. Synthesis of Alcohol 63a
a Reagents and conditions: (a) TPAP, NMO, 4 Å MS, CH2Cl2, rt. (b)
CBr4, PPh3, Et3N, CH2Cl2, 0 °C. (c) n-Bu3SnH, Pd(PPh3)4, benzene, rt,
82% (three steps). (d) 3, Pd(PPh3)4, CuCl, LiCl, 1:1 DMSO/THF, 60 °C,
66% (82% based on recovered 64).
Scheme 14. Completion of Total Synthesis of Gambierola
a Reagents and conditions: (a) NaOMe, 1:1 MeOH/CH2Cl2, rt. (b)
TBSCl, imidazole, DMF, 0 °C. (c) TPAP, NMO, 4 Å MS, CH2Cl2, rt, 69%
(three steps). (d) LiHMDS, TMSCl, Et3N, THF, -78 °C. (e) Pd(OAc)2,
MeCN, rt. (f) MeMgBr, toluene, -78 °C, 94% (three steps). (g) TBSOTf,
Et3N, CH2Cl2, rt. (h) LiDBB, THF, -78 f -45 °C. (i) TBDPSCl, Et3N,
DMAP, CH2Cl2, rt, 99% (three steps). (j) TBSOTf, Et3N, CH2Cl2, rt. (k)
CSA, 1:1 MeOH/CH2Cl2, 0 °C, 93% (two steps).
afforded alcohol 62 (99% yield for the three steps). Further
silylation of the C6 alcohol with TBSOTf and Et3N followed
by selective removal of the TBS group at C32 led to alcohol
63 in 93% yield for the two steps.
Alcohol 63 was then transformed by a three-step sequence
as described before into dibromoolefin 64, setting the stage for
the introduction of the triene side chain (Scheme 13). The Stille
coupling of 64 with (Z)-vinyl stannane 3 was carried out under
the optimized conditions described before [Pd(PPh3)4, CuCl,
LiCl, 1:1 DMSO/THF, 60 °C] to furnish fully protected
gambierol 65 in 66% yield (82% yield based on recovered
64).38 Global silyl ether deprotection proved to be much more
difficult than expected. The sterically hindered C30 TBS
ether could not be cleaved under a variety of conditions, by
TBAF, HF‚pyridine, or tris(dimethylamino)sulfur difluorotri-
methylsilicate (TASF).39 Also, all attempts to cleave the C30
silyl group by extending the reaction time or forcing the reaction
a Reagents and conditions: (a) HF‚pyridine, THF, rt, quant. (b) 3,
Pd(PPh3)4, CuCl, LiCl, 1:1 DMSO/THF, 60 °C, 43%.
conditions resulted in isomerization or loss of the labile triene
moiety.
After extensive experimentation, it was found that exposure
of (Z)-vinyl bromide 64 to excess HF‚pyridine (THF, room
temperature, 6 days) cleanly afforded triol 2 in quantitative yield.
Finally, the Stille coupling of 2 with 3 under the established
Pd(PPh3)4/CuCl/LiCl-promoted conditions (1:1 DMSO/THF, 60
°C) furnished (-)-gambierol (1) in 43% isolated yield (Scheme
14 ). The synthetic gambierol was identical to the natural sample
1
by H NMR, 13C NMR, and HRMS spectra. Also, the CD
(38) In this reaction, (Z)-vinyl bromide 64 was not consumed completely, and
an inseparable 4:1 mixture of 65 and 64 was obtained in 82% yield. The
yield of 65 was estimated to be 66% by the 1H NMR analysis of the purified
mixture.
(39) (a) Noyori, R.; Nishida, I.; Sakata, J.; Nishizawa, M. J. Am. Chem. Soc.
1980, 102, 1223. (b) Scheidt, K. A.; Chen, H.; Follows, B. C.; Chemler,
A. R.; Coffey, D. S.; Roush, W. R. J. Org. Chem. 1998, 63, 6436.
spectra measured for synthetic 1 showed close agreement with
that for the natural material. Moreover, mouse lethality of
synthetic gambierol (ip, 50-75 µg/kg) was equipotent to that
of the natural toxin. Thus, the structure of gambierol including
9
J. AM. CHEM. SOC. VOL. 124, NO. 50, 2002 14991

A R T I C L E S
Fuwa et al.
the absolute configuration was unambiguously confirmed as
shown in Figure 1.
described herein would provide easy access to a variety of
structural analogues of gambierol. Studies toward clarification
of the biological profile of gambierol as well as preparation of
its analogues are currently underway and will be reported in
due course.
Conclusion
We have achieved the first total synthesis of the marine
polycyclic ether toxin gambierol. Efficient and practical syn-
thetic routes to the ABC and EFGH ring fragments (5 and 6,
respectively) allowed preparation of these advanced intermedi-
ates in multigram quantities. Convergent union of these key
fragments has been successfully accomplished by utilizing our
B-alkyl Suzuki-Miyaura coupling strategy, which realized a
highly convergent synthesis of the octacyclic polyether core of
gambierol. We are now confident that our strategy is feasible
and powerful enough for the convergent synthesis of complex
polycyclic ethers. We have also demonstrated the feasibility of
the Pd(PPh3)4/CuCl/LiCl-promoted Stille conditions for the
coupling of less reactive and sterically hindered (Z)-vinyl
bromide and (Z)-vinyl stannane. The present total synthesis is
highly convergent and flexible when one considers the structural
complexity and size of the targeted molecule; its longest linear
sequence from commercially available 2-deoxy-D-ribose con-
sisted of 71 steps with an overall yield of 0.57%. The chemistry
Acknowledgment. We thank Professor Masayuki Satake of
Tohoku University for measurement of NMR spectra and mouse
lethality assay of synthetic gambierol, NMR and CD spectra of
natural gambierol, and fruitful discussions. A postdoctoral
fellowship to H.F. from the Japan Society for the Promotion of
Science (JSPS) is gratefully acknowledged. This work was
financially supported in part by a grant from Suntory Institute
for Bioorganic Research (SUNBOR).
Supporting Information Available: Experimental procedures
and spectroscopic data for all new compounds, 1H and 13C NMR
spectra for selected compounds, and NMR and CD comparison
data for natural and synthetic gambierol (PDF). This material
is available free of charge via the Internet at http://pubs.acs.org.
JA028167A
9
14992 J. AM. CHEM. SOC. VOL. 124, NO. 50, 2002