Synthesis of the C31-C67 fragment of amphidinol 3

Article abstract of DOI:10.1002/anie.200602742

(Chemical Equation Presented) A tetrahydropyran intermediate, prepared from (-)-diethyl tartrate, was used to assemble the C31-C67 fragment of amphidinol 3. Both of the tetrahydropyran rings in the natural product were prepared from this intermediate. The polyene segment was assembled with high selectivity.

Full text of DOI:10.1002/anie.200602742

Communications
Polyketides
DOI: 10.1002/anie.200602742
Synthesis of the C31–C67 Fragment of
Amphidinol 3**
Javier de Vicente, John R. Huckins, and
Scott D. Rychnovsky*
The amphidinols are a structurally unique group of metab-
olites isolated from the marine dinoflagellates Amphidinium
klebsii and Amphidinium carterae.[1] Amphidinol 3 (AM3; 1)
is one of the most biologically active compounds in this class,
as it exhibits potent hemolytic activity against human
erythrocytes and antifungal activity against Aspergillus
niger.[1] Key structural features of AM3 include two highly
substituted tetrahydropyran (THP)rings, a long irregular
polyol domain, and a skipped polyene chain all distributed
over a flexible 67-carbon backbone. The absolute stereo-
chemistry of AM3 was elucidated by Murata and co-workers
by using a new and powerful J-based NMR spectroscopic
technique.[1d,2] Because of its complex molecular architecture
and substantial biological activity, AM3 has emerged as an
important synthetic target, and a number of recent efforts
toward AM3 have been reported.[3] Herein, we disclose the
synthesis of the fully protected C31–C67 fragment.
Our approach to AM3 dissects the molecule into three
main fragments: polyene 2, bisTHP core 3, and the lower
polyol fragment (Scheme 1). The union of these segments
relies on two sequential olefination reactions. To take
advantage of the partial symmetry present in the molecule,
both THP rings in the C31–C52 core fragment 3 will be
derived from a common intermediate. Synthesis of the C53–
C67 polyene fragment 2 will rely on Pd-catalyzed cross-
coupling and the Horner–Wadsworth–Emmons reaction to
establish the all-E-alkene geometries.
Synthesis of core fragment 3 began with known 2,6-cis-
THP 4 (Scheme 2), which is available in 11 steps from (d)-
(ꢀ)-diethyl tartrate.[3a] TMSOTf-mediated allylation of the
oxocarbenium ion derived from acetoxy ether 4 proceeded in
excellent yield and furnished the desired 2,6-trans-THP ring
as a 10:1 mixture of diastereomers, which were separated by
column chromatography. Axial addition predominated, as
expected for the addition to a THP oxocarbenium ion.[4]
Protection of the diol as the acetonide and subsequent
[*] Dr. J. de Vicente, J. R. Huckins, Prof. S. D. Rychnovsky
Department of Chemistry
University of California, Irvine
1102 Natural Sciences 2, Irvine, CA 92697-2025 (USA)
Fax: (+1)949-824-6379
E-mail: srychnov@uci.edu
[**] This study was supported by the National Institutes of General
Medicine (GM 43854). J.d.V. acknowledges the Spanish Ministry of
Education and Science for a postdoctoral fellowship.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
7258
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7258 –7262

Angewandte
Chemie
Scheme 1. Retrosynthetic analysis of amphidinol 3 (1). TBS=tert-butyldimethylsilyl; SEM=b-(trimethylsilyl)ethoxymethyl.
Scheme 2. Synthesis of the C31–C52 fragment 12. Reagents and conditions: a) allyltrimethylsilane, DTBMP, TMSOTf, CH2Cl2, ꢀ788C; b) K2CO3,
MeOH (82% over two steps); c) 2,2-DMP, TsOH, C6H6 (82%); d) 1. O3, sudan III, CH2Cl2; 2. P P 3h, ꢀ788C!RT (94%); e) 5, piperidine, CH3CN,
ꢀ58C (80%); f) SEMCl, iPr2NEt, CH2Cl2, 408C (85%); g) OsO4, NMO, acetone/H2O (8:1); h) NaIO4, THF/H2O (1:1; 82% over two steps);
i) TBSOTf, 2,6-lutidine, CH2Cl2, 08C!RT (97%); j) H2 (90 psi), Pd/C, THF; k) TBSOTf, 2,6-lutidine, CH2Cl2, 08C!RT (84% over two steps);
l) 1. Me(OMe)NH·HCl, nBuLi, THF, ꢀ78!08C; 2. 8, ꢀ788C; m) MeMgBr, THF, 08C (98% for two steps); n) 1. LDA, THF, ꢀ788C; 2. N-(5-chloro-
2-pyridyl)triflimide, ꢀ708C (74%); o) Me3SnSnMe3, LiCl, [Pd(PPh3)4], THF, 808C in a sealed tube (85%); p) 1. 10, nBuLi, THF, ꢀ788C; 2. 11;
q) Dess–Martin periodinane, pyridine, CH2Cl2; r) Zn(BH4)2, Et2O, ꢀ358C (43% over three steps); s) TBSOTf, 2,6-lutidine, CH2Cl2 (73%).
Bn=benzyl, 2,2-DMP=2,2-dimethoxypropane, DTBMP=2,6-di-tert-butyl-4-methylpyridine, LDA=lithium diisopropylamide, NMO=N-methylmor-
pholine-N-oxide, OTf=trifluoromethanesulfonate, TMS=trimethylsilyl, TsOH=para-toluenesulfonic acid.
ozonolysis provided aldehyde 5. Sulfoxide 6 was prepared as
reported[5] and coupled with aldehyde 5 under the hydrox-
ylative Knoevenagel conditions detailed by Trost and Mal-
lart.[6] This crucial [2,3] sulfoxide/sulfenate rearrangement
established the last stereogenic center in the THP core of
AM3 and furnished allylic alcohol 7 as a 85:15 mixture of
Angew. Chem. Int. Ed. 2006, 45, 7258 –7262
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7259

Communications
diastereomers at C39/C44. The configuration of the major
isomer was confirmed by using the advanced Mosherꢀs
analysis.[7]
With common intermediate 7 in hand, preparation of both
halves of the core section was initiated (Scheme 2). To arrive
at the nucleophilic coupling partner 10, the alcohol 7 was first
protected as the TBS ether. A strategic protecting-group
exchange was effected by hydrogenolysis of the benzyl group,
which was accompanied by reduction of the enoate alkene,
followed by silylation of the primary alcohol. The C44
epimers of ester 8, which arise from the hydroxylative
Knoevenagel reaction, were readily separated by column
chromatography. Methyl ketone 9 was obtained from ester 8
via the intermediate Weinreb amide.[8] Kinetic deprotonation
with LDA followed by trapping with Comins reagent[9]
furnished the enol triflate, which was then transformed into
vinyl stannane 10 by a Stille cross-coupling with hexamethyl-
ditin.[10] Stannane 10 serves as a precursor to a vinyl lithium
reagent that will be used to couple the fragments.
With the desired nucleophilic coupling partner in hand,
common intermediate 7 was next converted into the requisite
aldehyde 11 (Scheme 2). The free alcohol was protected as a
SEM ether, and the C39 diastereomers were separated by
medium-pressure liquid chromatography (MPLC). Oxidative
cleavage of the alkene by using the Johnson–Lemieux
protocol afforded aldehyde coupling partner 11.[11]
Scheme 3. Synthesis of the C53–C67 fragments 17 and 18. Reagents
and conditions: a) (trimethylsilyl)acetylene, [Pd(PPh3)4], pyrrolidine
(98%); b) K2CO3, MeOH (99%); c) 1. catecholborane (748C); 2. H2O;
d) 1. NaOH, Et2O, 08C; 2. I2, 08C (70% over two steps); e) TBAF, THF
(100%); f) Ac2O, DMAP, pyridine, CH2Cl2 (95%); g) 1. 13, tBuLi, THF,
ꢀ788C; 2. ZnCl2, ꢀ788C!RT; 3. 14, [Pd(PPh3)4], ꢀ20!78C (91%);
h) K2CO3, MeOH (100%); i) 1. (COCl)2, DMSO, CH2Cl2, ꢀ788C;
2. NEt3, ꢀ788C!RT (87%); j) diethylallyl phosphonate, nBuLi, HMPA,
THF, ꢀ78!ꢀ108C (62%); k) TBAF, THF (100%); l) SO3·pyridine,
DMSO, iPr2NEt, CH2Cl2 (87%); m) PPh3, LiBr, DIAD, THF, 08C!RT
(ca. 72%); n) benzenesulfinic acid sodium salt, NaI, NaHCO3, DMF,
458C (84%). DIAD=diisopropyl azodicarboxylate, DMAP=4-dimethyl-
aminopyridine, HMPA=hexamethylphosphoramide, TBAF=tetrabutyl-
ammonium fluoride.
Union of the two THP fragments 10 and 11 was
accomplished by treatment of stannane 10 with nBuLi to
afford the nucleophilic vinyl lithium species and subsequent
addition to aldehyde 11. This reaction furnished the C31–C52
bisTHP core of AM3 as a 1:1 mixture of diastereomers.
Oxidation of the allylic alcohol mixture with Dess–Martin
periodinane[12] followed by chelation-controlled reduction of
the resultant ketone with Zn(BH4)2 afforded the desired
R alcohol as the sole product. This transformation highlights
the selection of the SEM protecting group for the C44
alcohol, which can be removed with other silyl groups, but
participates in chelation. The yield over these three steps was
43%, and the stereochemistry at C43 was verified by using the
advanced Mosherꢀs analysis.[7] Protection of the alcohol as the
TBS ether afforded the complete, fully protected core
fragment 12.
group followed by a Swern oxidation delivered aldehyde
15.[16]
With the all-E triene portion of the polyene in place, a
semistabilized Horner–Wadsworth–Emmons olefination was
utilized to install the terminal E-1,3-diene moiety.[17] Treat-
ment of diethylallyl phosphonate with nBuLi in the presence
of HMPA, followed by addition of trienal 15 led to the
formation of the all-E pentaene in 62% yield. Removal of the
TIPS ether moiety furnished alcohol 16. Note that polyene 16
was synthesized with excellent control over the E/Z geometry
of each double bond. We anticipated that undesired alkene
isomers would likely be difficult or impossible to separate.
Alcohol 16 was further elaborated to aldehyde 17 and sulfone
18 as potential coupling partners with the bisTHP core.
A report from Paquette and Chang, which describes the
union of the polyene segment to one of the protected THP
rings of the core of AM3, drew our attention.[3e] They
described a highly selective Julia–Kocienski olefination
using aldehyde 17 and a b-alkoxy sulfone core fragment. We
were intrigued by this result because we too had explored this
strategy on several model systems, one of which was very
similar to the core fragment prepared by Paquette and Chang.
This strategy had proved unselective in our hands, and we
chose to pursue an alternate route. The disclosure was
interesting enough, however, to prompt us to attempt that
same coupling on the fully elaborated core.
Synthesis of the C53–C67 skipped polyene is detailed in
Scheme 3 and began from E vinyl iodide 13, available after
triisopropylsilyl (TIPS)etherification and hydrozirconation/
iodinolysis of the corresponding alkynol. Sonagashira cross-
coupling[13] of iodide 13 and (trimethylsilyl)acetylene fol-
lowed by methanolysis of the TMS group furnished the
terminal enyne. Selective hydroboration of the alkyne with
catecholborane and subsequent iodinolysis afforded the
E,E dienyl iodide with complete geometric selectivity.[14]
simple protecting-group interchange provided iodide 14.
A
A Negishi cross-coupling procedure was employed to
form the central all-E triene.[15] Treatment of another equiv-
alent of vinyl iodide 13 with two equivalents of tBuLi followed
by transmetallation to zinc furnished the requisite nucleo-
phile. Addition of vinyl iodide 14 in the presence of [Pd-
(PPh3)4] afforded the all-E triene with excellent (> 20:1)
geometric selectivity in 91% yield. Hydrolysis of the acetate
Further elaboration of bisTHP fragment 12 to the desired
sulfone nucleophile began with reductive debenzylation using
LiDBB to reveal the primary alcohol (Scheme 4).[18] No
reduction of the allylic ether at C43 was observed under the
reaction conditions. The phenyltetrazole sulfide was intro-
duced under conditions developed by Mitsunobu,[19] followed
7260 www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7258 –7262

Angewandte
Chemie
Scheme 4. The union of the polyene and core fragments. Reagents and conditions: a) LiDBB, THF, ꢀ788C (98%); b) Dess–Martin periodinane,
pyridine, CH2Cl2 (100%); c) PPh3, phenyltetrazole thiol, DIAD, THF, 08C!RT (86%); d) [Mo7O24(NH4)6]·4H2O, H2O2, EtOH (59%); e) 1. 18,
LDA, THF, ꢀ788C; 2. 20, ꢀ788C!RT (76%); f) Bz2O, DMAP, pyridine, CH2Cl2 (89%); g) Na(Hg), Na2HPO4, THF/MeOH (3:1), ꢀ208C (72%);
h) 1. 19, KHMDS, THF, ꢀ788C; 2. 17, ꢀ788C!RT (40%). Bz2O=benzoic anhydride, KHMDS=potassium bis(trimethylsilyl)amide,
LiDBB=lithium di-tert-butylbiphenylide.
by molybdate-mediated oxidation[20] to provide sulfone 19. A
suitable electrophile was generated by Parikh–Doering
oxidation of alcohol 16[21] to furnish aldehyde 17 (Scheme 3).
The Julia–Kocienski olefination[22] was carried out by
treating a solution of sulfone 19 in THF at ꢀ788C with
KHMDS, followed by addition of aldehyde 17. In accord with
our model studies, the coupling of these two fragments
proceeded in an unselective manner to furnish the C31–C67
fragment of AM3 (21)in 40% yield as a 1:1 mixture of E/
Z isomers. These results stand in contrast to those obtained by
Paquette and Chang. Subtle differences between substrates,
reagents, and even the scale of the reactions may be
responsible for these divergent outcomes.
An alternate strategy we explored for uniting the polyene
with the core relied on the Julia–Lythgoe olefination.[23] The
required sulfone 18 (Scheme 3)was synthesized from alcohol
16 by using the Mitsunobu reaction to provide the primary
bromide followed by displacement with the sodium salt of
benzenesulfinic acid. The requisite aldehyde 20 was available
by Dess–Martin oxidation of the alcohol revealed by deben-
zylation of core fragment 12 (Scheme 4).
In agreement with our earlier model studies, metalation of
sulfone 18 with LDA followed by addition of aldehyde 20
afforded the coupled b-hydroxy sulfone product in 76% yield
(Scheme 4). Acylation of the alcohol with benzoic anhydride
followed by elimination with sodium amalgam afforded the
C31–C67 section of AM3 (21)with a satisfying 11:1 E/Z ratio.
This sequence proceeded in good yield, was very selective,
and afforded the product cleanly. In our hands, the Julia–
Lythgoe procedure was reproducible and led to fewer by-
products than the Julia–Kocienski strategy for this coupling.
In summary, the fully protected C31–C67 fragment of
amphidinol 3 (21), the largest portion of this challenging
natural product yet to be disclosed, has been synthesized in a
very convergent manner. Key features of the synthesis are the
use of common intermediate 7 for the construction of both
halves of the bisTHP core (12), a highly stereoselective
polyene synthesis, and the union of those fragments by a
stereoselective Julia–Lythgoe olefination. Our efforts in the
total synthesis of amphidinol 3 are ongoing.
Received: July 10, 2006
Published online: October 2, 2006
Keywords: Julia olefination · natural products · polyenes ·
.
polyketides · tetrahydropyran
[1] a)G. K. Paul, N. Matsumori, M. Murata, K. Tachibana, Tetrahe-
dron Lett. 1995, 36, 6279; b)M. Satake, M. Murata, T. Yasumoto,
T. Fujita, H. Naoki, J. Am. Chem. Soc. 1991, 113, 9859; c)T.
Houdai, S. Matsuoka, M. Murata, M. Satake, S. Ota, Y. Oshima,
L. L. Rhodes, Tetrahedron 2001, 57, 5551; d)M. Murata, S.
Matsuoka, N. Matsumori, G. K. Paul, K. Tachibana, J. Am.
Chem. Soc. 1999, 121, 870; e)G. K. Paul, N. Matsumori, K.
Konoki, M. Murata, K. Tachibana, J. Mar. Biotech. 1997, 5, 124;
f)G. K. Paul, N. Matsumori, K. Konoki, M. Sasaki, M. Murata,
K. Tachibana in Harmful and Toxic Algal Blooms. Proceedings
of Seventh International Conference on Toxic Phytoplankton
(Eds.: T. Yasumoto, Y. Oshima, Y. Fukuyo), UNESCO, Sendai,
Japan, 1996, p. 503.
[2] N. Matsumori, D. Kaneno, M. Murata, H. Nakamura, K.
Tachibana, J. Org. Chem. 1999, 64, 866.
[3] a)J. de Vicente, B. Betzemeier, S. D. Rychnovsky, Org. Lett.
2005, 7, 1853; b)S. BouzBouz, J. Cossy, Org. Lett. 2001, 3, 1451;
c)E. M. Flamme, W. R. Roush, Org. Lett. 2005, 7, 1411; d)J. D.
Hicks, E. M. Flamme, W. R. Roush, Org. Lett. 2005, 7, 5509;
e)L. A. Paquette, S.-K. Chang, Org. Lett. 2005, 7, 3111; f)S.-K.
Chang, L. A. Paquette, Synlett 2005, 2915; g)C. Dubost, I. E.
Marko, J. Bryans, Tetrahedron Lett. 2005, 46, 4005.
[4] a)L. Ayala, C. G. Lucero, J. A. C. Romero, S. A. Tobacco, K. A.
Woerpel, J. Am. Chem. Soc. 2003, 125, 15521; b)R. V. Stevens,
A. W. M. Lee, J. Am. Chem. Soc. 1979, 101, 7032; c)R. V.
Stevens, Acc. Chem. Res. 1984, 17, 289; d)P. Deslongchamps,
Stereoelectronic Effects in Organic Chemistry, Pergamon, New
York, 1983, pp. 209 – 221.
Angew. Chem. Int. Ed. 2006, 45, 7258 –7262
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7261

Communications
[5] J.-M. Brunel, P. Diter, M. Deutsch, H. B. Kagan, J. Org. Chem.
1995, 60, 8086.
[6] B. M. Trost, S. Mallart, Tetrahedron Lett. 1993, 34, 8025.
[7] The major isomer of 7 was purified by MPLC before the Mosher
analysis: a)M. J. Rieser, Y. H. Hui, J. K. Rupprecht, J. F.
Kozlowski, K. V. Wood, J. L. McLaughlin, P. R. Hanson, Z.
Zhuang, T. R. Hoye, J. Am. Chem. Soc. 1992, 114, 10203; b)J. A.
Dale, H. S. Mosher, J. Am. Chem. Soc. 1973, 95, 512; c)G. R.
Sullivan, J. A. Dale, H. S. Mosher, J. Org. Chem. 1973, 38, 2143;
d)I. Ohtani, T. Kusumi, Y. Kashman, H. Kakisawa, J. Am.
Chem. Soc. 1991, 113, 4092.
[8] a)A. Basha, M. Lipton, S. M. Weinreb, Tetrahedron Lett. 1977,
18, 4171; b)J. I. Levin, E. Turos, S. M. Weinreb, Synth. Commun.
1982, 12, 989.
[9] D. L. Comins, A. Dehghani, Tetrahedron Lett. 1992, 33, 6299.
[10] W. J. Scott, J. K. Stille, J. Am. Chem. Soc. 1986, 108, 3033.
[11] R. Pappo, D. S. Allen, Jr., R. U. Lemieux, W. S. Johnson, J. Org.
Chem. 1956, 21, 478.
[12] D. B. Dess, J. C. Martin, J. Org. Chem. 1983, 48, 4155.
[13] a)K. Sonagashira, Y. Todah, N. Hagihara, Tetrahedron Lett.
1975, 16, 4467; b)M. Alami, F. Ferri, G. Linstrumelle, Tetrahe-
dron Lett. 1993, 34, 6403.
[14] a)H. C. Brown, S. K. Gupta, J. Am. Chem. Soc. 1972, 94, 4370;
b)H. C. Brown, T. Hamaoka, N. Ravindran, J. Am. Chem. Soc.
1973, 95, 5786.
[15] a)E.-I. Negishi, Acc. Chem. Res. 1982, 15, 340; b)E.-I. Negishi,
T. Takahashi, S. Baba, D. E. Van Horn, N. Okukado, J. Am.
Chem. Soc. 1987, 109, 2393.
[16] a)S. L. Huang, K. Omura, D. Swern, J. Org. Chem. 1976, 41,
3329; b)T. T. Tidwell, Org. React. 1990, 39, 297.
[17] Y. Wang, F. G. West, Synthesis 2002, 99.
[18] LiDBB is a reducing agent for arene radical anions, first
reported by Freeman et al.: P. K. Freeman, L. L. Hutchinson, J.
Org. Chem. 1980, 45, 1924.
[19] O. Mitsunobu, Synthesis 1981, 1.
[20] H. S. Schultz, H. B. Freyermuth, S. R. Buc, J. Org. Chem. 1963,
28, 1140.
[21] J. R. Parikh, W. v. E. Doering, J. Am. Chem. Soc. 1967, 89, 5505.
[22] a)P. R. Blakemore, W. J. Cole, P. J. Kocienski, A. Morley, Synlett
1998, 26; b)P. R. Blakemore, J. Chem. Soc. Perkin Trans. 1 2002,
2563.
[23] M. Julia, J. M. Paris, Tetrahedron Lett. 1973, 14, 4833.
7262 www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 7258 –7262