A C-glycosidation approach to the central core of amphidinol 3: Synthesis of the C39-C52 fragment

Article abstract of DOI:10.1021/ol050477l

(Chemical Equation Presented) A concise route to an advance precursor (3) of the central core of amphidinol 3, a natural occurring polyketide, has been developed by applying a reductive lithiation as key step. The origin of the diastereoselectivity of thi

Full text of DOI:10.1021/ol050477l

ORGANIC
LETTERS
2005
Vol. 7, No. 9
1853-1856
A C-Glycosidation Approach to the
Central Core of Amphidinol 3:
Synthesis of the C39−C52 Fragment
Javier de Vicente, Bodo Betzemeier, and Scott D. Rychnovsky*
Department of Chemistry, 516 Rowland Hall, UniVersity of California,
IrVine, California 92697-2025
srychnoV@uci.edu
Received March 4, 2005
ABSTRACT
A concise route to an advance precursor (3) of the central core of amphidinol 3, a natural occurring polyketide, has been developed by
applying a reductive lithiation as key step. The origin of the diastereoselectivity of this reaction was comprehensively studied for nucleophilic
C-glycoside donor 5 and differently protected analogues.
The substance class of amphidinols are polyketide metabo-
lites isolated from the marine dinoflagellate Amphidinium
klebsii and carterae.1-3 Dinoflagellates are protist unicellular
algae and a rich source of structurally and biologically
interesting natural products, e.g., brevetoxins,4 maitotoxin,5
and okadaic acid.6 Amphidinols have shown potent hemolytic
activity against human erythrocytes and antifungal activity
against Aspergillus niger.1 The dinoflagellate A. klebsii was
initially isolated from washed seaweed collected from the
Aburatsubu Bay in Kanagawa, Japan. Processing of 440 L
of cell culture yielded 12 mg of amphidinol 3 with other
amphidinol homologues, rendering extraction from natural
sources impractical.7 The structure of amphidinol 3 was
elucidated by mass spectroscopic analysis and NMR spec-
troscopy7 using a potentially powerful new J-based method.8
The novel architecture and the bioactive properties of
amphidinol 3 led us to investigate its total synthesis9 with
special attention to the central core of amphidinol 3 (2, C31-
C52 segment), which can be disconnected from the rest of
the molecule by two olefination reactions.
Our approach to the central core of amphidinol 3 involves
sequential addition of two glycosyllithium reagents to an
aldehyde derived from alkene 3 and to the known epoxy
aldehyde 614 using the almost identical glycoside donors 4
(7) Murata, M.; Matsuoka, S.; Matsumori, N.; Paul, G. K.; Tachibana,
K. J. Am. Chem. Soc. 1999, 121, 870.
(8) Matsumori, N.; Kaneno, D.; Murata, M.; Nakamura, H.; Tachibana,
K. J. Org. Chem. 1999, 64, 866.
(9) There are no reported syntheses of amphidinol 3. There are a few
reported synthetic studies: (a) BouzBouz, S.; Cossy, J. Org. Lett. 2001, 3,
1451. (b) Flamme, E. M.; Roush, W. R. Org. Lett. 2005, 7, ASAP.
(10) Mukaiyama, T.; Suzuki, K.; Yamada, T.; Tabusa, F. Tetrahedron
1990, 46, 265.
(1) Paul, G. K.; Matsumori, N.; Murata, M.; Tachibana, K. Tetrahedron
Lett. 1995, 36, 6279.
(2) Satake, M.; Murata, M.; Yasumoto, T.; Fujita, T.; Naoki, H. J. Am.
Chem. Soc. 1991, 113, 9859.
(3) Houdai, T.; Matsuoka, S.; Murata, M.; Satake, M.; Ota, S.; Oshima,
Y.; Rhodes, L. L. Tetrahedron 2001, 57, 5551.
(4) Lin, Y.-Y.; Risk, M.; Ray, S. M.; Van Engen, D.; Clardy, J.; Golik,
J.; James, J. C.; Nakanishi, K. J. Am. Chem. Soc. 1981, 103, 6773.
(5) Murata, M.; Naoki, H.; Matsunaga, S.; Satake, M.; Yasumoto, T. J.
Am. Chem. Soc. 1994, 116, 7098.
(11) Brown, C. A.; Ahuja, V. K. J. Org. Chem. 1973, 38, 2226.
(12) Independent of our synthetic studies, Forsynth’s group has extended
the use of TEMPO oxidation of 1,5-diols to δ-lactones: Hansen, T. M.;
Florence, G. J.; Lugo-Mas, P.; Chen, J.; Abrams, J. N.; Forsyth, C. J.
Tetrahedron Lett. 2003, 44, 57.
(6) Yasumoto, T.; Murata, M. Chem. ReV. 1993, 93, 1897.
10.1021/ol050477l CCC: $30.25
© 2005 American Chemical Society
Published on Web 04/07/2005

Scheme 1. Retrosynthetic Analysis of Amphidinol 3 (1)
and 5, respectively (Scheme 1). This convergent approach
takes maximum advantage of the partial symmetry of
amphidinol 3.
dihydroxylation conditions. Osmium tetraoxide oxidation of
lactol acetate 12 afforded the desired diastereomeric diol 13
(96:4 dr) by a selective approach of the oxidant to the less
sterically hindered face of the alkene. This reductive acety-
lation/dihydroxylation sequence was more efficient than
reversing the order of steps, as osmium tetraoxide oxidation
of lactone 11 provided a 4:1 mixture of 1,2-cis diols. After
a strategic reprotection sequence, chemoselective phenylthio
acetal installation provided C-glycoside donor 5. Reductive
lithiation of phenylthio acetals 5 and subsequent coupling
to known epoxy aldehyde 614 led preferentially to the C31-
C52 segment of amphidinol 3 (3) but with poor diastereo-
selectivity (48:26:14:12 dr). The addition showed modest
The preparation of C-glycoside donor 5 commenced from
aldehyde 7,10 a known chiral synthon conveniently obtained
by standard refunctionalization of D-(-)-tartaric acid (Scheme
2). Aldehyde 7 was coupled with a titanium acetylide leading
to alcohol 9 with high diastereoselectivity (95:5 dr).10 Partial
reduction of the triple bond under P-2 nickel reduction
conditions11 and alcohol desilylation afforded diol 10.
Primary-selective TEMPO alcohol oxidation followed by
subsequent lactol oxidation efficiently provided lactone 11.12
Reductive acetylation of lactone 11 diastereoselectively
produced acetoxy ether 12,13 which was then subjected to
Scheme 2. Synthesis of the C39-C52 Fragment of Amphidinol 3 (3)
1854
Org. Lett., Vol. 7, No. 9, 2005

Scheme 3. Mechanism of the Reductive Lithiation Reaction
Scheme 4. Reductive Lithiations of Phenylthio Acetals 5
also tested at -50 °C for 30 min without any change of
selectivity after acetone trapping. Reductive lithiation of each
individual phenylthio acetal, 5ax and 5eq, in the presence
of a large excess of LiDBB (inverse addition) and subsequent
trapping with acetone gave again a ca. 6:4 mixture of adducts,
confirming the lack of selective formation of axial radical
intermediate 16ax or its subsequent reduction.
Felkin-Ahn selectivity.15 Unexpectedly, the selectivity for
axial lithiation of the 2-thiophenyl tetrahydropyrans 5 was
very low.16,17
Lithium di-tert-butylbiphenylide (LiDBB)18 reacts with
2-thiophenyltetrahydropyrans by a single-electron transfer
(SET) normally producing a dynamic mixture of anomeric
radicals, which equilibrate toward the thermodynamically
more stable axial radical.19 The preferred axial radical is then
reduced by a second SET providing a thermodynamically
less stable axial organolithium.16a
The loss of selectivity in the overall reductive lithiation
of 5 and coupling with an electrophile could have been
caused by (a) nonselective axial radical formation, (b) poor
configurational stability of the organolithium intermediates,
or (c) nonstereospecific electrophilic addition (Scheme 3).
Reductive lithiation of a mixture of phenylthio acetals 5 and
trapping with different electrophiles such acetone, CD3OD,
or Me3SnCl led to a ca. 6:4 mixture of adducts independent
of the identity of the electrophile (Scheme 4). Apparently,
the alkyllithium is formed with poor selectivity but reacts
with retention of configuration.
The thermodynamic preference of the radical intermediates
did not favor the formation of an axial radical stabilized by
pseudoanomeric effect. This preference is probably due to a
twist-boat conformation of the pyranosyl radicals generated.
Conformational analysis of the C-glycosidation adducts (3,
19ax, and 18ax) by 1H NMR suggested that protected
pyranosides with such substitution patterns have a tendency
to adopt twist-boat-like conformations.20
Conformational analysis of anomeric radical intermediate
16 would be helpful in understanding the stereochemical
outcome of its reductive lithiation and coupling reactions
with electrophiles. Computational studies of a model radical
intermediate 26, analogue to radical 16, revealed a preference
for boat conformations (Scheme 5). The calculated low
Scheme 5. Conformational Analysis of Model Pyranosyl
Radical Intermediate 26 at UB3LYP/6-31G*
Alternative preparation of the organolithium intermediates
(17ax and 17eq) via tin-lithium exchange and trapping with
acetone confirmed their configurational stability and their
stereospecific reaction with acetone as electrophile (Scheme
3). The configurational stability of the organolithium inter-
mediates generated by reductive lithiation at -78 °C was
(13) Kopecky, D. J.; Rychnovsky S. D, J. Org. Chem. 2000, 65, 191.
(14) Gao, Y.; Hanson, R. M.; Klunder, J. M.; Ko, S. Y.; Masamune, H.;
Sharpless, K. B. J. Am. Chem. Soc. 1987, 109, 5765.
(15) Nacro, K.; Baltas, M.; Escudier, J.-M.; Gorrichon, L. Tetrahedron
1996, 52, 9047.
(16) (a) Cohen, T.; Lin, M.-T. J. Am. Chem. Soc. 1984, 106, 1130. (b)
Boeckman, R. K.; Enholm, E. J.; Demko, D. M.; Charette, A. B. J. Org.
Chem. 1986, 51, 4743. (c) Cohen, T.; Bhupathy, M. Acc. Chem. Res. 1989,
22, 152.
(17) (a) Rychnovsky, S. D.; Mickus, D. E. Tetrahedron Lett. 1989, 30,
3011. (b) Rychnovsky, S. D.; Buckmelter, A. J.; Dahanukar, V. H.;
Skalitzky, D. J. J. Org. Chem. 1999, 64, 7678.
(18) LiDBB is an arene radical anion reducing agent first described by
Freeman: Freeman, P. K.; Hutchinson, L. L. J. Org. Chem. 1980, 45, 1924.
(19) Rychnovsky, S. D.; Powers, J. P.; LePage, T. J. J. Am. Chem. Soc.
1992, 114, 8375.
energy difference between trans and cis boat radicals was
only 0.2 kcal/mol and was in agreement with all our
experimental results.
To increase the diastereoselectivity on the reductive
lithiation of 5 and coupling with 6, we investigated a
conformational restriction strategy. The conformational influ-
ence of the protecting groups installed on the C-glycoside
(20) See Supporting Information.
Org. Lett., Vol. 7, No. 9, 2005
1855

and the alkyl chain substituent of the THP ring.21 Variation
in the protecting groups did influence the stereoselectivity,
but none of the protecting groups led to useful levels of
selectivity.
Scheme 6. Reductive Lithiation of Phenylthio Acetals 20, 21,
and 22
In summary, we have designed a convergent synthesis of
the central core of amphidinol 3. Our strategy was imple-
mented on the preparation of the C39-C52 fragment (3) in
16 steps, which also constitutes a formal synthesis of the
C31-C38 segment (4). The synthesis includes the use of a
cyclic R-acetoxy ether 12 as a transient intermediate capable
of directing a diastereoselective dihydroxylation reaction. The
nucleophilic C-glycosidation reaction of phenylthio acetal 5
and key coupling with aldehyde 6 was studied in order to
increase its diastereoselectivity. The anomalous conforma-
tional behavior of the derived radical intermediates was
influenced but not improved by alternate protecting group
patterns.
The modest selectivity in the key coupling reaction led
us to investigate nucleophilic additions to an oxocarbenium
ion derived from the lactol acetate 13. These oxocarbenium
ion additions are highly diastereoselective and form the crux
of our revised approach to amphidinol 3.
donor should affect the stereoselectivity.21 We identified the
cyclic acetal of 5 as potentially responsible for the lack of
selectivity on its reductive lithiation. Thus, we studied three
differently protected C-glycoside donors (20, 21, and 22)
that did not incorporate bridging protecting groups on the
glycoside (Scheme 5), all conveniently prepared from
R-acetoxy ether 13.20 Reductive lithiation of diol 21 and bis-
SEM analogue 22 followed by coupling with acetone
provided mixtures similar to those observed with bis-
acetonide 5. The exception was bis-TBS ether 20 that led to
a 33:66 ratio favoring the equatorial isomer. This exception
was probably due to a ring inversion of the pyranosyl ring
caused by a 1,2-diequatorial interaction of the OTBS groups
Acknowledgment. This work was supported in part by
the National Institute of General Medicine (GM 43854).
J.d.V. acknowledges the Spanish Ministry of Education and
Science for a postdoctoral fellowship. B.B. thanks the
Deutsche Forschungsgemeinschaft (DFG) for a postdoctoral
fellowship.
Supporting Information Available: Experimental pro-
cedures and characterization data of the described com-
pounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
(21) Matsuda has described a conformational restriction strategy for
controlling the stereochemical outcome of pyranosyl radical reactions: (a)
Abe, H.; Shuto, S.; Matsuda, A. J. Am. Chem. Soc. 2001, 123, 11870. (b)
Abe, H.; Terauchi, M.; Matsuda, A. Shuto, S. J. Org. Chem. 2003, 68,
7439.
OL050477L
1856
Org. Lett., Vol. 7, No. 9, 2005