gem-diacetates as carbonyl surrogates for asymmetric synthesis. Total syntheses of sphingofungins E and F

Article abstract of DOI:10.1021/ja0118338

The equivalent of an asymmetric addition to a carbonyl group with a stabilized anion is accomplished by discriminating between the enantiotopic C-O single bonds of a gem-diacetate. In this way, enantioselective total syntheses of two antifugal agents, sph

Full text of DOI:10.1021/ja0118338

J. Am. Chem. Soc. 2001, 123, 12191-12201
12191
gem-Diacetates as Carbonyl Surrogates for Asymmetric Synthesis.
Total Syntheses of Sphingofungins E and F
Barry M. Trost* and Chulbom Lee†
Contribution from the Department of Chemistry, Stanford UniVersity, Stanford, California 94305
ReceiVed July 27, 2001
Abstract: The equivalent of an asymmetric addition to a carbonyl group with a stabilized anion is accomplished
by discriminating between the enantiotopic C-O single bonds of a gem-diacetate. In this way, enantioselective
total syntheses of two antifugal agents, sphingofungins E and F, have been accomplished. The synthetic strategy
is based on a series of catalytic processes whereby all of the chiral centers are created with high stereoselectivities.
The first two stereocenters are introduced by an asymmetric allylic alkylation reaction of gem-diacetate 9 with
azlactone 10. The complex of Pd(0) and ligand 14 efficiently catalyzes this key reaction, which differentiates
both the enantiotopic leaving groups of a gem-diacetate and enantiotopic faces of the enolate of an azlactone
in high enantiomeric excess and diastereomeric excess. From these two stereocenters, the configurations of
the remaining two centers are set by a diastereoselective Os(VIII)-catalyzed dihydroxylation reaction with
excellent stereocontrol. The trans-alkene is established by Cr(II)-mediated olefination, and a subsequent B-alkyl
Suzuki coupling reaction conjoins the polar head unit and the nonpolar, 13-carbon lipid tail. The efficiency of
our strategy is illustrated by the completion of syntheses of sphingofungins F and E in 15 and 17 steps, and
in 17% and 5% overall yields, respectively.
Sphingosines constitute the backbone of sphingolipids, ubiq-
uitous membrane components that are involved in a number of
cellular events. The discovery of the ability of sphingosine to
inhibit protein kinase C (PKC) prompted a wider examination
of the possible roles of sphingosines and their derived products
in signal transduction.1 These investigations have led to a
recognition that sphingosines and various biosynthetic inter-
mediates, albeit in low cellular concentrations, are important
Figure 1. Sphingosine analogues of fungal origin.
second messengers and regulatory molecules.2 While the
structure of sphingosines varies among organisms, a variety of
D), they were found to block the biosynthesis of sphingolipids,
leading to apoptosis in both yeast and mammalian cells. These
cellular effects are due to their potent inhibitory activities against
serine palmitoyltransferase (SPT), an essential committed
enzyme involved in the first step of sphingosine biosynthesis.5
Particularly noteworthy in these compounds is their striking
similarity to myriocin6 (4, also known as thermozymocidin and
ISP-1) and mycestericin7 (cf. 5), which are known to be
remarkable immunosuppressive agents with potencies equivalent
to and 10- to 100-fold higher than those of clinically used
FK5068 and cyclosporin A9 (CsA), respectively. This series
apparently inhibits T cell proliferation by a completely new
mode of action, suggesting the prospect of developing a novel
therapeutic agent on the basis of modulation of sphingolipid
novel analogues have been isolated mainly from the cultivation
of fungi (Figure 1).3 These sphingosine-like compounds typically
possess more functionalized head and lipid tail units as compared
to common erythro-sphingosine.
Sphingofungins E (1) and F (2) were isolated as antifungal
agents from fermentation of Paecilomyces Variotii (ATCC
74079) by Merck in 1992.4 Along with other congeners (A-
† Present address: Department of Chemistry, Princeton University,
Princeton, NJ 08544.
(1) (a) Hannun, Y. A.; Loomis, C. R.; Merrill, A. H.; Bell, R. M. J.
Biol. Chem. 1986, 261, 2604. (b) Hannun, Y. A. Science 1989, 243, 500.
(2) (a) Hannun, Y. A. Sphingolipid-Mediated Signal Transduction; R.
G. Landes: Austin, TX, 1997. (b) Merrill, A. H., Jr.; Sweeley, C. C.
Sphingolipids: Metabolism and Cell Signaling; Elsevier Science: Amster-
dam, 1996. (c) Merrill, A. H., Jr.; Sweeley, C. C. In Biochemistry of Lipids,
Lipoproteins and Membranes; Vance, D. E., Vance, J., Eds.; Elsevier
Science: Amsterdam, 1996; pp 309-339. (d) Hakomori, S. In Handbook
of Lipid Research; Kafner, J. N., Hakomori, S., Eds.; Plenum: New York,
1983; Vol. 3, pp 1-164. (e) Ariga, T.; Jarvis, W. D.; Yu, R. K. J. Lipid
Res. 1998, 39, 1. (f) Igarashi, Y. J. Biochem. 1997, 122, 1080. (g) Hannun,
Y. A. Science 1996, 274, 1855. (h) Shayman, J. A. J. Am. Soc. Nephrol.
1996, 7, 171. (i) Spiegel, S.; Milstein, S. J. Membr. Biol. 1995, 146, 225.
(3) (a) Merrill, A. H., Jr.; Liotta, D. C.; Riley, R. E. In Handbook of
Lipid Research; Bell, R. M., Ed.; Plenum: New York, 1996; Vol. 8, pp
205-237. (b) Mandala, S. M.; Harris, G. H. Methods Enzymol. 2000, 311,
335.
(5) Zweerink, M. M.; Edison, A. M.; Well, G. B.; Pinto, W.; Lester, R.
L. J. Biol. Chem. 1992, 267, 25032.
(6) Fujita, T.; Inoue, K.; Yamamoto, S.; Ikumoto, T.; Sasaki, S.; Toyama,
R.; Chiba, K.; Hoshino, Y.; Okumoto, T. J. Antibiot. 1994, 47, 208.
(7) Sasaki, S.; Hashimoto, R.; Kiuchi, M.; Inoue, K.; Ikumoto, T.; Hirose,
R.; Chiba, K.; Hoshino, Y.; Okumoto, T.; Fujita, T. J. Antibiot. 1994, 47,
420.
(8) (a) Kino, T.; Hatanaka, H.; Hashimoto, M.; Nishiyama, M.; Goto,
T.; Okuhara, M.; Kohsaka, M.; Aoki, H.; Imanaka, H. J. Antibiot. 1987,
40, 1249. (b) Kino, T.; Hatanaka, H.; Miyata, S.; Inamura, N.; Nishiyama,
M.; Yajima, T.; Goto, T.; Okuhara, M.; Kohsaka, M.; Aoki, H.; Ochiai, T.
J. Antibiot. 1987, 40, 1256.
(9) Borel, J. F. In Cyclosporin A; Borel, J. F., Ed.; Elsevier Biochemical
Press: Amsterdam, 1982; pp 5-17.
(4) Horn, W. S.; Smith, J. L.; Bills, G. F.; Raghoobar, S. L.; Helms, G.
L.; Kurtz, M. B.; Marrinan, J. A.; Frommer, B. R.; Thornton, R. A.;
Mandala, S. M. J. Antibiot. 1992, 45, 1692.
10.1021/ja0118338 CCC: $20.00 © 2001 American Chemical Society
Published on Web 11/14/2001

12192 J. Am. Chem. Soc., Vol. 123, No. 49, 2001
Trost and Lee
biosynthesis.10 Recently, two murine myriocin-binding proteins
have been isolated and found to be genetically linked to
sphinogolipid biosynthesis.11 However, as manifested by the
seemingly independent nature of the SAR,12 a direct connection
between immunosuppression and SPT inhibition remains to be
established.
Scheme 1. Retrosynthetic Analysis
Due to their biological importance and novel structures, the
sphingosine analogues have stimulated a number of synthetic
efforts.13 While these studies largely made use of the “chiral
pool”, the structural homology imparted in these compounds
led us to explore a more general synthetic strategy. Using a
new asymmetric catalytic methodology, we hoped to achieve
unambiguous enantiocontrol in the creation of the C2 quaternary
center, which had been a major difficulty in the synthesis of
this series.14 In contrast to the chiral pool approaches, our
flexible strategy was anticipated to provide ready access to not
only this important class of molecules but also their analogues
by simple modifications in the synthetic scheme. Herein, we
report a full account on the successful realization of concise
syntheses of sphingofungins E and F.15
anticipated elimination problems in using the enolate of 10 (R
) OH), an OH surrogate will be required.
This synthetic plan required an asymmetric addition of a
rather stabilized anion to a carbonyl groupsa feature that suffers
from an unfavorable equilibrium as well as no precedent for
such asymmetric addition. Both of these issues are tackled by
the key aldol-like reaction whose motif came from our continued
studies on the asymmetric allylic alkylation (AAA) reaction. It
has been demonstrated that gem-diesters can serve as “chiral
carbonyl surrogates” in the Pd-catalyzed AAA reactions, giving
rise to an aldol-type product with high enantiomeric excess
(ee).16 In a separate set of investigations, azlactones have proved
to be potent nucleophiles to provide various R-alkylated amino
acid derivatives in excellent ee.17 These reactions, when
combined together, achieved differentiation not only between
enantiotopic leaving groups of a gem-diacetate but also between
enantiotopic faces of an azlactone enolate. For example, the
reaction of gem-diacetate 11 and azlactone 12 catalyzed by the
complex of (η3-C3H5PdCl)2 (13) and ligand (R,R)-14 gave
readily separable diastereomers 15 and 16 with high stereose-
lectivities (eq 1). The relative stereochemistry of the major
Synthetic Plan
The conspicuous structural feature of the sphingofungins led
us to a retrosynthetic analysis dividing these compounds into
the two major parts, nonpolar lipid tail 6 and polar headgroup
7 (Scheme 1). We envisaged that these two fragments could be
conjoined by a cross-coupling reaction at a late stage of the
synthesis. While the preparation of 6 appeared straightforward,
the presence of polyhydroxyl groups and a quaternary center
in 7 made the stereocontrolled synthesis of this part more
challenging. In further analysis, the polyhydroxyl groups of 7
were envisioned to derive from a stereoselective oxygenation
of 8 wherein the allylic directing effect would require the
inverted configuration at the allylic center based upon the known
diastereoselectivity of osmium-catalyzed dihydroxylation. If the
key intermediate 8 were to emanate from our newly developed
asymmetric alkylation, then gem-diacetate 9 and serine, or
alanine, derived azlactone 10 (R ) OH or H) could serve as
precursors for the E and F systems, respectively. Because of
(10) Miyake, Y.; Kozutsumi, Y.; Nakamura, S.; Fujita, T.; Kawasaki,
T. Biochem. Biophys. Res. Commun. 1995, 211, 396.
(11) Chen, J. K.; Lane, W. S.; Schreiber, S. L. Chem. Biol. 1999, 6,
221.
(12) Fujita, T.; Hirose, R.; Yoneta, M.; Sasaki, S.; Inoue, K.; Kiuchi,
M.; Hirase, S.; Chiba, K.; Sakamoto, H.; Arita, J. J. Med. Chem. 1996, 39,
4451.
(13) For syntheses of myriocin, see: (a) Banfi, L.; Bretta, M. G.;
Colombo, L.; Gennari, C.; Scolastico, C. J. Chem. Soc., Chem. Commun.
1982, 488. Banfi, L.; Bretta, M. G.; Colombo, L.; Gennari, C.; Scolastico,
C. J. Chem. Soc., Perkin Trans. 1 1983, 1613. (b) Yoshikawa, M.;
Yokokawa, Y.; Okuno, Y.; Murakami, N. Chem. Pharm. Bull. 1994, 42,
994. Yoshikawa, M.; Yokokawa, Y.; Okuno, Y.; Murakami, N. Tetrahedron
1995, 51, 6209. (c) Sano, S.; Kobayashi, Y.; Kondo, T.; Takebayashi, M.;
Maruyama, S.; Fujita, T.; Nagao, Y. Tetrahedron Lett. 1995, 36, 2097. (d)
Hatakeyama, S.; Yoshida, M.; Esumi, T.; Iwabuchi, Y.; Irie, H.; Kawamoto,
T.; Yamada, H.; Nishizawa, M. Tetrahedron Lett. 1997, 45, 7887. (e) Rao,
A. V. R.; Gurjar, M. K.; Devi, T. R.; Kumar, K. R. Tetrahedron Lett. 1993,
34, 1653. (f) Deloisy, S.; Thang, T. T.; Olesker, A.; Lukas, G. Tetrahedron
Lett. 1994, 35, 4783. For synthesis of sphingofungin E, see: (g) Nakamura,
T.; Shiozaki, M. Tetrahedron Lett. 2001, 42, 2701. For syntheses of
sphingofungin F, see: (h) Kobayashi, S.; Matsumura, M.; Furuta, T.;
Hayashi, T.; Iwamoto, S. Synlett 1997, 301. Kobayashi, S.; Furuta, T.;
Hayashi, T.; Nishijima, M.; Hanada, K. J. Am. Chem. Soc. 1998, 120, 908.
Kobayashi, S.; Furuta, T.; Hayashi, T.; Nishijima, M.; Hanada, K.
Tetrahedron 1998, 54, 10275. (i) Liu, D.-G.; Wang, B.; Lin, G.-Q. J. Org.
Chem. 2000, 65, 9114.
product 15 corresponded well to the C2 and C3 stereocenters
of the key intermediate 8. Thus, simple structural variations of
the two reaction partners would afford proper intermediates of
type 8 for the synthesis of sphinogfungins E and F.
A notable advantage of this approach is the installation of
two stereogenic centers, including the C2 quaternary carbon,
by using a single, unambiguous asymmetric reaction. Once the
configurations of C2 and C3 are established, the remaining
stereocenters are set with reference to this stereochemistry by
a diastereoselective oxygenation. Another point of practical
consequence is that the requisite gem-diacetate 9 and azlactone
10 are readily available from the corresponding R,â-unsaturated
aldehyde and R-amino acids, respectively.
(16) (a) Trost, B. M.; Lee, C. B.; Weiss, J. M. J. Am. Chem. Soc. 1995,
117, 7247. (b) Trost, B. M.; Lee, C. B. J. Am. Chem. Soc. 2001, 123, 3687.
(c) Trost, B. M.; Lee, C. B. J. Am. Chem. Soc. 2001, 123, 3671.
(17) (a) Trost, B. M.; Ariza, X. Angew. Chem., Int. Ed. Engl. 1997, 36,
2635. (b) Trost, B. M.; Ariza, X. J. Am. Chem. Soc. 1999, 121, 10727.
(14) For reviews, see: (a) Martin, S. F. Tetrahedron 1980, 36, 419. (b)
Fuji, K. Chem. ReV. 1993, 93, 2037. (c) Corey, E. J.; Guzman-Perez, A.
Angew. Chem., Int. Ed. 1998, 37, 388.
(15) Trost, B. M.; Lee, C. B. J. Am. Chem. Soc. 1998, 120, 6818.

Total Syntheses of Sphingofungins E and F
J. Am. Chem. Soc., Vol. 123, No. 49, 2001 12193
Table 1. Alkylation of Azlactone 10a with gem-Diacetate 9 (Eq 2)
% yielda (% ee)b
entry
catalyst
solvent
temp, time
21
22
ratio 21:22c
1
2
3
4
5
6
2% Pd(PPh3)4
DME
DME
CH3CNd
THF
25 °C, 5 h
0 °C, 3 h
28 (-)
73 (82)
71 (10)e
68 (89)
70 (89)
70 (87)
32 (-)
17 (55)
10 (60)
5 (86)
5 (89)
6 (89)
1:1.6
5.3:1
2.0% 13, 6.0% 14
1.0% 13, 3.0% 14
2.0% 13, 6.0% 14
0.5% 13, 1.5% 14
0.2% 13, 0.6% 14
0 °C, 2 h
3.5:1
0 °C, 3 h
9.6:1
10.5:1
11.2:1
THF
-5 °C, 4 hf
THF
-5 °C, 6 hf
1
a Isolated yields. b HPLC analysis of the corresponding methyl esters 23 and 24 (eqs 3 and 4). c Diastereomeric ratio determined by H NMR
integration on the crude mixture. d Triethylamine was used as base. e The ee of the opposite enantiomer, (ent)-21. f Slow addition (see text).
Scheme 2. Preparation of Azlactone 10a and gem-Diacetate
9
as the product in a 1.6:1 ratio as measured by 1H NMR
integration of the signal of the R-acetoxy methine protons, which
resonated at δ 5.61 (d, J ) 6.5 Hz) for the major isomer and
5.57 (d, J ) 6.9 Hz) for the minor isomer. The IR absorptions,
at 1824 and 1827 cm-1, indicated the presence of an intact
azlactone moiety in both products. The major and minor
products could be easily separated by silica gel chromatography
and were isolated in 32% and 28% yields, respectively.
We then set out to examine the asymmetric reaction. Accord-
ingly, the reaction was carried out with 2 mol % 13 and 6 mol
% 14 in DME at 0 °C to provide a mixture of diastereomers in
a 5.3:1 ratio (entry 2). Interestingly, the major product from
this asymmetric reaction turned out to be the minor product of
the racemic reaction. The absolute and relative configurations
of products were tentatively assigned by analogy to previous
results (cf. eq 1). To determine the ee of the alkylation products,
21 and 22 were separated and each converted to the corre-
sponding methyl esters, 23 and 24, in quantitative yields by
treatment with methanol under acid catalysis (Vide infra, eq’s
3 and 4). Chiral HPLC analysis successfully determined the ee’s
Total Synthesis of Sphingofungin F
The first stage of the synthesis involved the preparation of
gem-diacetate 9 and methyl-substituted azlactone 10a (Scheme
2). While 10a was synthesized from N-benzoyl-D,L-alanine
(17),18 9 was prepared by FeCl3-catalyzed addition19 of acetic
anhydride to aldehyde 19,20 which was readily available from
cis-2-butene-1,4-diol (18a) by monosilylation (TBDPSCl, imi-
dazole, 98%) and PCC oxidation. Alternatively, aldehyde 19
could be directly prepared from monosilylated 2-butyne-1,4-
diol (20b) in comparable yield by an internal atom economical
redox reaction under ruthenium catalysis without having re-
course to stoichiometric reductants and oxidants.21 Since the
cis-2-butene-1,4-diol derives from 2-butyne-1,4-diol, this latter
route also reduces the actual number of steps.
of 23 and 24, and thus, those of 21 and 22, to be 82% and
55%, respectively. When the azlactone enolate was generated
in acetonitrile using triethylamine as base, the reaction generated
a 3.5:1 mixture of diastereomers whose major product turned
out to be (ent)-21 with an ee of only 10% (entry 3). This result
presents a sharp contrast to the previous results wherein the
combination of triethylamine base and acetonitrile solvent
proved most effective, leading to high ee’s and diastereomeric
excesses (de’s) in a number of cases.16 Significant improvement
was realized by switching the solvent to tetrahydrofuran (THF)
and the counterion back to sodium (entry 4). The de jumped to
9.6:1, and ee’s of 89% and 86% were obtained for 21 and 22,
respectively. When a solution of 9 was slowly added at -5 °C
over 4 h by a syringe pump to a solution containing 0.5 mol %
13, 1.5 mol % 14, and the sodium enolate of 10a, the two
diastereomers were formed in a 10.5:1 ratio with an ee of 89%
for both isomers (entry 5). Using this slow addition protocol,
the load of 13 and 14 could be further reduced to 0.2 mol %
and 0.6 mol %, respectively, with little effect on the stereo-
chemical outcome (entry 6).
With the two substrates readily prepared, our initial attention
focused on the Pd-catalyzed alkylation (eq 2, Table 1). The
viability of this alkylation was first tested by using achiral
(PPh3)4Pd as the catalyst. The nucleophile was generated from
2.0 equiv of sodium hydride and 2.5 equiv of azlactone 10a in
dimethyl ether (DME) at ambient temperature (entry 1). The
reaction proceeded smoothly to give a mixture of diastereomers
(18) Chen, F. M. F.; Kuroda, K.; Beneiton, N. L. Synthesis 1979, 230.
(19) Kochhar, K. S.; Bal, B. S.; Deshpande, R. P.; Rajadhyaksha, S. N.;
Pinnick, H. W. J. Org. Chem. 1983, 48, 1765.
(20) Quimpere, M.; Ruest, L.; Deslongchamps, P. Synthesis 1992, 132.
(21) Trost, B. M.; Livingston, R. C. J. Am. Chem. Soc. 1995, 117, 9586.
As the asymmetric reaction could consistently be conducted
with satisfactory ee and yield, we turned to dihydroxylation as
a means for introducing the syn-hydroxyl groups at C4 and C5.

12194 J. Am. Chem. Soc., Vol. 123, No. 49, 2001
Scheme 3. Dihydroxylation after Inversion at C3
Trost and Lee
Scheme 4. Dihydroxylation of Transposed Allylic Acetate
32
Figure 2. X-ray crystal structure of 29.
We first explored a sequence that involved inversion of C3
configuration prior to the introduction of the two hydroxyl
groups (Scheme 3). The inversion at C3 was achieved by
hydrolysis of 23 and subsequent activation via mesylate 26,
which led to the formation of oxazoline 27. Exercising the
dihydroxylation reaction, we noted that the anti directing effect22
of allylic heteroatoms had often been reversed in rigid cyclic
systems.23 In particular, a syn-dihydroxylation seemed viable
if OsO4 would approach the alkene from the face opposite the
sterically encumbering C2 center in the “H-eclipsed conforma-
tion”.24 However, the dihydroxylation of 27 gave an inseparable
mixture of lactones 28 and 29 in a 1:5 ratio, favoring the
undesired anti,syn-isomer over the desired syn,syn-isomer. The
attempt to override this stereochemical bias by using an
asymmetric method (AD-mix-R)25 provided little improvement,
generating a diastereomeric mixture in a 1:2 ratio. The results
seemed to indicate that the steric hindrance to OsO4 attack can
be avoided if the olefin adopts an alternate “O-eclipsed
conformation” that led to the anti-attack. Although chromato-
graphic separation of the two isomers was not feasible,
gratifyingly, the major product 29 spontaneously crystallized
from the mixture. An X-ray crystallographic analysis was
performed on a single crystal of this isomer to establish the
stereostructure of 29 (Figure 2). Thus, the relative stereochem-
istry of the two hydroxyl groups (C4, C5) and the tentative
structural assignment (C2, C3) for the asymmetric alkylation
products were verified at this point.
Kishi’s empirical rule22 for predicting the stereoselectivity of
dihydroxylation reactions (eqs 5 and 6). Hence, azlactone 21
was treated with catalytic OsO4 and NMO at 25 °C in methylene
chloride. The reaction furnished an 8:1 mixture of diastereomeric
lactones 30 and 31 from which the major isomer 30 could be
isolated in 86%. This rather encouraging selectivity led us to
examine other substrates in the hope of further improvement.
When methyl ester 23 was subjected to the same dihydroxylation
conditions, the desired lactone 30 was generated in 98% yield
as a single diastereomer! No trace of the minor isomer 31 could
be detected in a number of repeated experiments.
Intrigued by these results, we took a brief excursion to further
explore the cooperative effect of the two chiral centers on the
stereoselectivity of the dihydroxylation reaction. One interesting
substrate for this study was the isomeric allylic acetate 32
derived from a Pd(II)-catalyzed allylic transposition of 21 in
79% yield with complete transfer of chirality (Scheme 4).26 By
contrast, the transposed allylic acetate 32, bearing an alkene
flanked by two stereogenic centers, exhibited only a mediocre
selectivity (dr ) 2.3:1), despite proceeding with an excellent
94% yield.
The stereochemical outcome of the dihydroxylation reaction
of 27 reverted our strategy back to the original plan based on
(22) (a) Cha, J. K.; Christ, W. J.; Kishi, Y. Tetrahedron Lett. 1983, 24,
3943; Christ, W. J.; Cha, J. K.; Kishi, Y. Tetrahedron Lett. 1983, 24, 3947.
(b) Cha, J. K.; Christ, W. J.; Kishi, Y. Tetrahedron 1984, 40, 2247.
(23) (a) Brimacombe, J. S.; Kabir, A. K. M. S. Carbohydr. Res. 1986,
150, 35. (b) DeNinno, M. P.; Danishefsky, S. J.; Schulte, G. J. Am. Chem.
Soc. 1988, 110, 3925.
(24) Cha, J. K.; Kim, N.-S. Chem. ReV. 1995, 95, 1761.
(25) Kolb, H. C.; Van Nieuwenhze, M. S.; Sharpless, K. B. Chem. ReV.
1994, 94, 2483.
The remarkable diastereoselectivity of the dihydroxylation
of 23 seems to be a consequence of a strong cooperative effect
of the two vicinal stereogenic centers on the conformation. To
(26) (a) Overman, L. E.; Knoll, F. M. Tetrahedron Lett. 1979, 321. (b)
Also see ref 16c.

Total Syntheses of Sphingofungins E and F
J. Am. Chem. Soc., Vol. 123, No. 49, 2001 12195
Scheme 5. Completion of the Synthesis of Polar Headgroup 41 for Sphingofungin F
For setting the proper configuration at C3, lactone 30 was
protected as a PMB ether, and the C3 acetate was hydrolyzed.
Activation of the C3-O bond of 36 via the formation of triflate
37b induced cyclization to afford bicyclic oxazoline 38 in high
yield. This oxazoline-forming process efficiently adjusted the
C3 stereochemistry and provided simultaneous protection for
both the hydroxyl and benzamide groups as well. At this
juncture, we established all of the required stereogenic centers.
To append a one-carbon unit for C7, the primary alcohol was
liberated by fluoride-induced removal of the silyl protecting
group, and the resulting alcohol 39 was oxidized32 to afford
the unstable aldehyde 40. Upon subjection of 40 to a CrCl2-
mediated iodomethylenation reaction,33 the desired polar head-
group, vinyl iodide 41, was generated as a single isomer. The
formation of an (E)-alkene was verified later by 1H NMR
analysis on a more advanced intermediate. The exclusive (E)-
selectivity in this case mirrors a previously reported literature
example in which a similar Takai-Utimoto olefination of an
R-alkoxy-aldehyde exhibited complete (E)-selectivity.34
Figure 3. MM2 calculation and stable conformations of 35.
the extent that the dihydroxylation process occurs through an
early transition state, the relative energy of the ground-state
conformation will be reflected in the facial selectivity. To
examine the conformation of 23, an MM2* calculation was
performed wherein the TBDPSO group was simplified to a
methyl ether (Figure 3).27 A computer simulation revealed two
low-energy conformers for 35, while other conformations were
found to be at least 2 kcal/mol higher in energy. In the model
of the lowest energy conformer 35a, one olefinic face is much
more exposed to external reagents, whereas the acetoxy group
effectively shields the other face. The osmylation onto the more
open face of the double bond, however, leads to the unobserved
diastereomer 31. On the other hand, the second lowest energy
conformation 35b presents one face of its alkene readily
accessible to osmylation and corresponds to the formation of
30. Interestingly, 35b places the NH proton within a hydrogen-
bonding distance (rO-H ) 1.90 Å) of the carbonyl oxygen of
the acetate, thus efficiently blocking the other face of the alkene
from the approach of OsO4. Although the effect of hydrogen
bonding may be small due to the distortion of the N-H-O
dihedral angle and the presence of water in the reaction medium,
it appears to be a contributing factor to such an unusually high
degree of acyclic stereocontrol.28 An alternative rationale arises
from the possible directing ability of a remote NH group rather
than the effect of the allylic substitutent.29 Considering the
evidences that suggest the possible interaction between OsO4
and a nitrogen group,30 the observed selectivity may be due to
the NH templating effect or, at the minimum, a combination of
such an effect and conformational preference thereof.31
To establish the stereochemistry of the C-O bonds, correla-
tion studies were conducted. First, the PMB ether 38, obtained
as in Scheme 5, was treated with DDQ to give the corresponding
alcohol (eq 7).35 The product of this reaction was found to be
identical in all aspects to the minor diastereomer 28 from the
previous route (cf. Scheme 3). On the other hand, the O-
(30) (a) Walsh, P.; Bennani, Y. L.; Sharpless, K. B. Tetrahedron Lett.
1993, 34, 5545. (b) Xu, D.; Park, C. Y.; Sharpless, K. B. Tetrahedron Lett.
1994, 35, 2495.
With the two C-O bonds at C4 and C5 installed, we set out
to complete the synthesis of the polar headgroup (Scheme 5).
(31) However, the NH syn-directing model may require a more rigorous
assessment of such effects: (a) Davis, F. A.; Chen, B.-C. In StereoselectiVe
Synthesis; Helmchen, G., Hoffmann, R. W., Mulzer, J., Schaumann, E.,
Eds.; Houben-Weyl: Stuttgart, 1996; Vol. 8, pp 4592-4593. Also see: (b)
Poli, G. Tetrahedron Lett. 1989, 30, 7385. (c) Donohoe, T. J.; Blades, K.;
Helliwell, M.; Moore, P. R.; Winter, J. J. G. J. Org. Chem. 1999, 64, 2980.
(32) (a) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4156. (b)
Dess, D. B.; Martin, J. C. J. Am. Chem. Soc. 1991, 113, 7277.
(33) Takai, T.; Nitta, K.; Utimoto, K. J. Am. Chem. Soc. 1986, 108, 7408.
(34) Kende, A. S.; DeVita, R. J. Tetrahedron Lett. 1990, 31, 307. For a
related study of the solvent effect on the (E)-selectivity, see: Evans, D.
A.; Black, W. C. J. Am. Chem. Soc. 1993, 115, 4497.
(27) The force field calculation was carried out by using Macromodel
6.0 on a Silicon Graphics System. Mohamadi, F.; Richards, N. G. J.; Guida,
W. C.; Liskamp, R.; Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.;
Still, W. C. J. Comput. Chem. 1990, 11, 440.
(28) For example, see: Ikemoto, N.; Schreiber, S. L. J. Am. Chem. Soc.
1990, 112, 9657.
(29) (a) Haiza, M.; Lee, J.; Snyder, J. K. J. Org. Chem. 1990, 55, 5008.
(b) Trost, B. M.; Guo, G.-H.; Benneche, T. J. Am. Chem. Soc. 1988, 110,
621. (c) Houser, F. M.; Ellennberger, S. R. J. Am. Chem. Soc. 1984, 106,
2458. (d) Johnson, C. R.; Barbarchyn, M. R. J. Am. Chem. Soc. 1984, 106,
2459. (e) Hacksell, U.; Daves, G. D. J. Org. Chem. 1983, 48, 4144.
(35) Oikawa, Y.; Yoshioka, T.; Yonemitsu, O. Tetrahedron Lett. 1982,
33, 671.

12196 J. Am. Chem. Soc., Vol. 123, No. 49, 2001
Trost and Lee
Scheme 6. B-Alkyl Suzuki Merger and Completion of the Total Synthesis of Sphingofungin F
alkylation36 of lactone 29 afforded PMB ether 42 whose
spectroscopic properties were different from those of 38. Since
alcohol 29 had been analyzed by an X-ray crystal structure (cf.
Figure 3), these chemical correlations globally confirmed our
stereochemical assignment on the alkylation, dihydroxylation,
and inversion reactions.
Scheme 7. Synthesis of Silyl Analogue of Serine Azlactone
Upon preparation of the polar headgroup 41, we set out to
synthesize the lipid tail unit 44 via a three-step sequence starting
from heptanoyl chloride (43) (eq 9). After formation of a
from which sphingofungin F was isolated, in that the pH affects
the degree of protonation of the amino acid.13i,40 The chemical
shift of C10 (δ 25.6) reported by Merck’s group is different
compared to the values from Kobayashi et al. (δ 23.6), Liu et
al. (δ 23.9), and the present work (δ 23.6). While a plausible
explanation for this disparity should await direct comparisons,
unfortunately, a sample or the data therefrom of the natural
product was not available.41
Weinreb amide,37 the Grignard addition reaction with 5-hex-
enylmagnesium bromide gave ketoalkene 6. Subsequent treat-
ment of 6 with ethylene glycol and catalytic p-TsOH under
azeotropic dehydration conditions provided ketal 44.
The final stage of the synthesis involved a cross-coupling of
the polar headgroup 41 with the nonpolar lipid tail 44 and
deprotection (Scheme 6). The Pd-catalyzed B-alkyl Suzuki
reaction of 41 with organoborane 45, prepared in situ from 44
by hydroboration with 9-BBN-H, provided the fully protected
46 in excellent yield.38 In the 1H NMR spectrum of alkene 46,
a characteristic trans-coupling constant (J ) 15.6 Hz) between
the two olefinic protons (δ 6.02 and δ 5.54) was observed, thus
providing support for an (E)-geometry. Removal of the PMB
group by CAN effected simultaneous hydrolysis of the ethylene
ketal and oxazoline to give ketolactone 47. Finally, a base-
promoted hydrolysis cleaved the lactone and the amide groups,
and subsequent neutralization with an acidic resin (Amberlite
IRC-76) afforded sphingofungin F (2).
Total Synthesis of Sphingofungin E
Having accomplished the synthesis of sphingofungin F (2),
we then sought to extend this strategy to the total synthesis of
sphingofungin E (1). For this synthesis, however, an additional
issue had to be addressed due to the presence of a C21 hydroxyl
group. Namely, the alkylation required an “enolate of serine”.42
Mindful of the potential problem in the direct use of azlactone
10 (R ) OH in Scheme 1) as nucleophile, we sought to prepare
a silyl analogue of azlactone (R ) SiR′3) that would serve as a
serine enolate equivalent, avoiding the â-elimination. It was
envisioned that the requisite C-O bond would emerge from
the C-Si bond by Tamao-Fleming-type oxidation after the
alkylation.43 Accordingly, methyl hippurate (48, methyl N-
benzoylglycine), prepared from hippuric acid or methyl glyci-
nate, was converted to a silyl analogue of serine and advanced
to the corresponding azlactone 10b (Scheme 7). The sequence
started with photoinitiated bromination of 48, which gave the
unstable bromide 49. Treatment of bromide 49 with 2.5 equiv
The synthetic compound was spectroscopically (1H and 13C
NMR, IR, and MS) in good agreement with the natural4 and
synthetic13h,i sphingofungin F.39 Our melting point, 143-145
°C, also agreed with that reported, mp 142-144 °C13h and 145-
147 °C.13i In the 13C NMR data, the chemical shifts of C2 and
C10 showed some difference among the reported values. For
C2, the 13C shifts were δ 66.2 (Merck), 67.7 (Kobayashi et
al.13h), 67.1 (Liu et al.13i), and 66.7 (the present work). It is
possible that the variability in the chemical shift of C2 could
be a consequence of the difference in the pH of the medium
(40) A similar observation has been made: Ohfune, Y.; Shimamoto, K.;
Ishida, M.; Shinozaki, H. Bioorg. Med. Chem. Lett. 1993, 3, 15.
(41) Although the differences between the values reported for the natural
and synthetic samples might result from a pH effect as well, it could be
simply a typographical error.
(42) For related examples, see: (a) Meyers, A. I.; Mihelich, E. D. Angew.
Chem., Int. Ed. Engl. 1976, 15, 270. (b) Seebach, D.; Aebi, J. D. Tetrahedron
Lett. 1984, 25, 2545. (c) Yu, L.-C.; Helquist, P. J. Org. Chem. 1981, 46,
4536. (d) Adam, W.; Ehrig, V. Synthesis 1976, 817.
(43) (a) Tamao, K.; Ishida, N.; Tanaka, T.; Kumada, M. Organometallics
1983, 2, 1694. (b) Fleming, I.; Henning, R.; Plaut, H. J. Chem. Soc., Chem.
Commun. 1984, 29. For a review, see: (c) Jones, G. R.; Landais, Y.
Tetrahedron 1996, 52, 7599.
(36) Dueno, E. E.; Chu, F. X.; Kim, S.-I.; Jung, K. W. Tetrahedron Lett.
1999, 40, 1843.
(37) Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815.
(38) (a) Miyaura, N.; Suzuki, A. Chem. ReV. 1995, 95, 2457. (b) Miyaura,
N.; Ishiyama, T.; Sasaki, H.; Ishikawa, M.; Satoh, M.; Suzuki, A. J. Am.
Chem. Soc. 1989, 111, 314. For modified conditions, see: (c) Johnson, C.
R.; Braun, M. P. J. Am. Chem. Soc. 1993, 115, 11014. (d) Ohba, M.;
Kawase, N.; Fujii, T. J. Am. Chem. Soc. 1996, 118, 8250.
(39) Although the melting point and optical rotation are in good
agreement among synthetic samples, those for the natural compound had
not been reported.

Total Syntheses of Sphingofungins E and F
J. Am. Chem. Soc., Vol. 123, No. 49, 2001 12197
Table 2. Alkylation of Azlactone 10b with gem-Diacetate 9 (Eq 10)
% yielda (% ee)b
entry
catalyst
solvent
temp, time
>53
>54
ratio 53:54c
1
1.0% 13, 3.0% 14
1.0% 13, 3.0% 14
0.5% 13, 1.5% 14
1.0% 13, 3.0% 14
THF
DME
THF
THF
0 °C, 3 h
0 °C, 2 h
0 °C, 5 h
0 °C, 2 h
68 (96)
40 (94)
60 (96)
49 (93)
23 (96)
22 (94)
27 (96)
26 (96)
2.4:1
1.7:1
1.6:1
1.5:1
2
3d
4f
a
Isolated yields. b HPLC analysis of the corresponding methyl esters 55 and 56 (eqs 11 and 12). c Diastereomeric ratio determined by 1H NMR
integration on the crude mixture. d Slow addition via syringe pump. e Not determined. f n-BuLi was used as base (see text).
of the Grignard reagent afforded the silyl analogue of serine 51
in moderate yield through the intermediacy of glycine imine
50.44 After hydrolysis of the methyl ester, the resulting acid 52
was converted to azlactone 10b by a DCC-promoted cyclode-
hydration.18
With the silylazlactone 10b in hand, we then examined its
Pd-catalyzed asymmetric alkylation with gem-diacetate 9 (eq
10, Table 2). Following the previous procedure, 9 was reacted
Figure 4. Model for the double asymmetric induction.
(entry 2). Reducing the amount of the catalyst and implementing
a slow addition procedure gave a similar stereochemical outcome
(entry 3).
In the Pd-catalyzed alkylation of azlactone nucleophiles,
increasing the size of the alkyl chains of azlactones resulted in
a dramatic increase in the diastereomeric ratio.17 The observed
trend was explained by using a simplified transition-state model
in which the chiral pocket of the catalyst discriminated two
prochiral faces of the azlactone enolate, as illustrated in Figure
4. In this model, the interaction between the oxyanion of the
azlactone enolate and the positively charged π-allylpalladium
complex is considered to be important.45 As a result, the major
with the sodium enolate of 10b in the presence of 1 mol % 13
isomer arises from the preferred approach (57), in which the
alkyl group is oriented away from steric congestion while the
phenyl group is positioned near the “flap” of the chiral pocket.
It is difficult, however, to apply this model to explain the
rather surprising results obtained from the alkylation of 10b,
because the sterically demanding silylalkyl chain gave rise only
to poor diastereomeric ratios. The possibility is that the effective
size of the silylmethyl group may become much smaller if the
anion of silylazlactone 10b adopts a bicyclic structure due to
the ability of a silicon atom to accommodate hypervalent
coordination (eq 13). The interaction between the oxyanion and
and 3 mol % 14 at 0 °C in THF (entry 1). The reaction was
completed in 3 h, giving rise to a 2.4:1 mixture of diastereomers
1
53 and 54 as determined by H NMR integration. The signals
for R-acetoxy methine protons of the major and minor products
had chemical shifts of δ 5.47 (d, J ) 7.3 Hz) and 5.53 (d, J )
8.3 Hz), respectively, providing signals for the measurement
of the diastereomeric ratio.
The alkylation products were separated by chromatography
and converted to the corresponding methyl esters 55 and 56 by
acid-catalyzed methanolysis (eqs 11 and 12). In contrast to
silicon atom in enolate 59 could facilitate the formation of
bicyclic oxazole 60, in which the pentavalent silicon atom adopts
a trigonal bipyramidal geometry.
If the bicyclic structure 60 constituted the more reactive
limiting structure or, at least, contributed to the reactivity of
the enolate, the model established for the previous alkylation
may no longer be applicable. Delocalization of the anionic
charge by hypercoordination and the pseudo-C2 symmetry
imparted in the relatively flat bicyclic oxazole may allow the
nucleophile to adopt four orientations with respect to the
π-allylpalladium intermediate (Figure 5). Consequently, the
differentiation of the two prochiral faces of the nucleophile by
the chiral pocket would occur in a much less selective fashion.
On the basis of this hypothesis, the enolate with more oxophilic
methyl-substituted azlactones 21 and 22, the methanolysis of
these silylazlactones was very sluggish and required additional
amounts of acid catalyst (1% vs. 5%) and a longer reaction time
(8 h vs. 48 h) for completion. Nevertheless, the reactions
furnished the methyl esters in nearly quantitative yields, and
chiral HPLC analyses successfully measured the ee of both
products to be 96%. Changing the solvent from THF to DME
had a slightly adverse effect, as lower de and ee were obtained
(44) (a) Kober, R.; Hammes, W.; Steglich, W. Angew. Chem., Int. Ed.
Engl. 1982, 21, 203. (b) Kober, R.; Steglich, W. Liebigs Ann. Chem. 1983,
599. (c) Munster, P.; Steglich, W. Synthesis 1987, 223. (d) Bretschneider,
T.; Miltz, W.; Munster, P.; Steglich, W. Tetrahedron 1988, 44, 5403.
(45) (a) Reference 16. (b) For detailed discussions, see: Trost, B. M.;
Toste, F. D. J. Am. Chem. Soc. 1999, 121, 4545.

12198 J. Am. Chem. Soc., Vol. 123, No. 49, 2001
Trost and Lee
Scheme 8. Installation of 21-OH Group and C3
Configuration
Our next concern was installation of a primary hydroxyl group
at C21 and adjustment of the C3 configuration (Scheme 8). The
former was readily achieved by activation of the C-Si bond of
lactone 65 by using a one-pot, modified Fleming oxidation
procedure.46 Upon treatment of 65 with a mixture of peracetic
acid and sodium bromide, the C-Si bond was smoothly
oxidized to give diol 67 in good yield. It is worth noting that
all other functional groups including the tert-butyldiphenylsilyl
(TBDPS) moiety, a diarylsilane, remained unaffected under these
conditions. As a prelude to inversion of the configuration at
C3, the two hydroxyl groups at C5 and C21 were converted to
PMB ethers. After the hydrolysis of the C3 acetoxy group, the
resulting alcohol was subjected to a cyclodehydration reaction
with triflic anhydride to provide bicyclic oxazoline 69 in high
yield. To determine the relative stereochemistry, an NOE study
on 69 was conducted. Irradiation of Hc at δ 5.28 (d, J ) 4.9
Hz) induced an enhancement of the intensity of the Hd signal
at δ 5.08 (dd, J ) 7.3, 4.9 Hz) and the Hb signal at δ 3.67 (d,
J ) 8.8 Hz) by 11% and 3%, respectively. Interestingly, Ha at
δ 4.28 (d, J ) 8.8 Hz) did not exhibit any NOE upon irradiation
of Hc or Hd while showing strong geminal correlation with Hb.
It appears that only one of two C21 methylene protons is in
proximity of Hc, presumably because of hindered rotation. Based
on these observations, the syn relationship among carbon-
oxygen bonds was reasonably established.
With all of the necessary stereocenters introduced, we set
out to finish the synthesis of sphingofungin E (1) using a
sequence similar to the synthesis of sphingofungin F (2)
(Scheme 9). After removal of the TBDPS group, primary alcohol
70 was oxidized with Dess-Martin periodinane to aldehyde 71,
which was directly subjected to Takai-Utimoto olefination to
afford iodide 72 as a single geometric isomer. To this fully
elaborated polar headgroup was attached the nonpolar lipid tail
by the B-alkyl Suzuki coupling reaction. Under palladium
catalysis, the vinyl iodide 72 smoothly reacted with the B-alkyl
borane 45 to afford alkene 73 in excellent yield. The large
coupling constant (J ) 15.5 Hz) between the two olefinic
protons (δ 6.00 and 5.53) established the olefin geometry to be
of the E-configuration. The final stage of the synthesis required
sequential deprotection reactions. Treatment of the alkene 73
with CAN in aqueous acetonitrile effected removal of the two
PMB groups with simultaneous cleavage of the ketal and
oxazoline functionalities to give triol 74. Alternatively, this
transformation could be carried out in comparable yield by a
stepwise sequence involving removal of the PMB groups by
Figure 5. Model for the chiral induction at a bicyclic nucleophile.
lithium as counterion was expected to disfavor formation of
the hypervalent structure (entry 4, Table 2). Whereas lithium
hydride failed to deprotonate azlactone 10b, treatment with
n-butyllithium at -78 °C in THF generated a lithium enolate
without effecting the carbonyl addition. The subsequent Pd-
catalyzed reaction with gem-diacetate 9 afforded the desired
alkylation products. However, the ratio (dr ) 1.5:1) turned out
to be similar to that obtained when a sodium enolate was
employed as nucleophile. The ee’s of both products were still
excellent and were determined to be 93% and 96% for the major
and minor products, respectively. Although the structure of the
reacting enolate remains unclear, the enolate derived from
silylazlactone 10b appeared to be a kinetically potent nucleophile
in that high chiral induction was achieved at the electrophile
(C3 center), regardless of the counterion.
The major product 53, though of unproven stereochemistry,
was carried on to the next step of the synthetic sequence in the
hope that more advanced intermediates would provide an
opportunity to determine the stereochemistry. The introduction
of the C4 and C5 hydroxyl groups was first attempted using an
osmium-catalyzed dihydroxylation reaction (eq 14). Treatment
of azlactone 53 with 1% OsO4 and NMO in moist CH2Cl2 at
25 °C generated a 14:1 mixture of diastereomeric lactones in
89% yield. Decreasing the reaction temperature to 0 °C
improved the ratio to 21:1, furnishing 65 and 66 in isolated
yields of 93% and 3%, respectively. These selectivities are
significantly higher than those observed with the methyl-
substituted azlactone 21 (dr ) 8:1), attesting to rather strong
facial guidance exerted by the C2 stereogenic center in the
dihydroxylation reaction. Methyl ester 55 was also used as
substrate to examine the possibility of achieving complete
selectivity as encountered previously (eq 15). Indeed, the
dihydroxylation of 55 under identical conditions at 25 °C for
12 h produced lactone 65 as a single diastereomer in 94% yield.
Thus, we secured two routes to the desired lactone 65 in which
the C4 and C5 hydroxyl groups were installed with high
diastereoselectivity.
(46) Fleming, I.; Sanderson, P. E. J. Tetrahedron Lett. 1987, 28, 4229.

Total Syntheses of Sphingofungins E and F
J. Am. Chem. Soc., Vol. 123, No. 49, 2001 12199
Scheme 9. B-Alkyl Suzuki Merger and Completion of the Synthesis of Sphingofungin E (1)
DDQ and of the oxazoline and ketal groups by subsequent acidic
hydrolysis.47 Finally, hydrolysis of triol 74 in methanolic sodium
hydroxide followed by neutralization and desalting with an
acidic resin afforded sphingofungin E (1) in 64% yield. Our
synthetic sample showed spectroscopic (1H and 13C NMR, IR)
characteristics that agreed very well with those reported for the
natural compound and those recently reported for a synthetic
sample.13g As was the case with the data of sphingofungin F,
small discrepancies in the chemical shifts of C1 [δ 174.2
(Merck),4 173.4 (Nakamura et al.),13g 173.2 (present work)] and
C2 [δ 71.8 (Merck), 71.1 (Nakamura et al.), 71.2 (present work)]
were noted in the 13C NMR spectrum. These inconsistencies
may again be ascribed to the pH difference which arose from
the disparity in the isolation procedures.
the biological properties suggest that the absolute stereostructure
of sphingofungins E and F is likely to follow those of other
fungal-originated congeners.
Given the significant biochemical role of sphingolipids,
various sphingosine analogues, which are themselves potent
inhibitors of sphingolipid biosynthesis, may serve as novel tools
for many areas of sphingosine-related research that otherwise
have proven recalcitrant to biochemical and molecular ap-
proaches. In particular, the difficulty involved in accessing these
compounds through fermentation,48 let alone analogues, makes
an efficient synthesis valuable. The route described herein is
useful for preparation of a number of analogues. For example,
the C2 and C3 centers are controlled by the chiral ligand. The
C3 epimer could be derived by carrying out the synthesis
without inversion. Dihydroxylation after inversion at C3 also
provides access to the C4 and C5 epimers. The use of different
azlactones would allow variation of the C21 alkyl substituent.
Finally, the lipid tail can be readily varied prior to the cross-
coupling reaction.
Summary
An efficient synthetic strategy for sphingosine analogues has
been developed. This strategy, based on a series of transition-
metal-catalyzed reactions, achieved relevant stereocontrol of the
key transformations in a highly selective manner and culminated
in concise total syntheses of antifungal agents, sphingofungins
E and F. The Pd-catalyzed asymmetric allylic alkylation of a
gem-diacetate and azlactones provided the key aldol-like
products bearing two stereogenic centers, including the C2
quaternary carbon, with high ee. These reactions achieved the
equivalent of an asymmetric carbonyl addition of an R-amino
acid enolate by differentiation of the two prochiral elements
present in both the electrophile and nucleophile, enantiotopic
leaving groups of a gem-diacetate and enantiotopic faces of an
azlactone enolate. While the synthesis of sphingofungin F
utilized an alanine azlactone, the azlactone derived from a novel
silyl analogue of serine proved an efficient synthon for a serine
enolate in the synthesis of the E congener. Once the first two
stereocenters were created by an asymmetric reaction, the
additional centers were set by a directed Os-catalyzed dihy-
droxylation reaction with high diastereoselectivity. The requisite
(E)-alkene was established by a Cr-mediated iodomethylenation
reaction, setting the final stage which involved merger of the
polar head and the lipid tail units by a Pd-catalyzed cross-
coupling reaction. The efficiency of our approach is readily
noted in the syntheses of sphingofungins E and F which have
been accomplished in only 17 steps (5.1% overall yield) and
15 steps (17% overall yield), respectively. Due to the lack of
data for the natural compounds, the present syntheses could not
confirm the absolute stereochemistry. However, the origin and
Experimental Section
Pd-Catalyzed Asymmetric Alkylation: Preparation of 21 and
22. To a suspension of sodium hydride (95% powder, 290 mg, 11.5
mmol) in THF (3 mL) was added azlactone 10a18 (2.50 g, 14.3 mmol)
in THF (10 mL). The resulting solution was stirred at 25 °C until
hydrogen gas evolution ceased. The resulting red solution was briefly
(∼0.5 h) sonicated while being purged with a stream of nitrogen. To
this solution was added a mixture of 13 (17.4 mg, 0.0476 mmol, 0.5
mol %) and the (R,R)-14 (98.6 mg, 0.143 mmol, 1.5 mol %) in THF
(2 mL) via a cannula. After 5 min, a solution of diacetate 9 (4.05 g,
9.52 mmol) in THF (10 mL) was added to the mixture at -5 °C via a
pump-driven syringe over 4 h. After the addition was complete, the
stirring was continued for an additional 1 h. The reaction mixture was
then poured into 10% aqueous NaH2PO4 (50 mL) and extracted three
times with ether (50 mL × 3). The combined organic layers were
washed with brine, dried over MgSO4, filtered, and concentrated.
Separation on a SiO2 column (petroleum ether/ethyl acetate ) 15:1)
yielded 3.610 g (70%) of the less polar major diastereomer 21 and
0.258 g (5%) of the more polar minor diastereomer 22, both as colorless
sticky oils. The ee of major diastereomer 21 could be directly
determined by chiral HPLC analysis (89%) and corroborated by the ee
of methyl ester 23 (89%). The ee of 22 was indirectly determined to
be 89% via the corresponding methyl ester 24.
(1′R,4S,2′E)-4-[1′-Acetoxy-4′-(tert-butyldiphenylsilyloxy)but-2′-en-
1′-yl]-4-methyl-2-phenyl-4,5-dihydro-5-oxazolone (21): tr (2S,3R) )
7.56 min, tr (2R,3S) ) 9.43 min (Chiralpak AD, λ ) 230 nm, Hep:i-
PrOH ) 99:1); [R]D -51.3 (c 3.90, CHCl3, 89% ee); IR (film) 1827,
(47) See experimental details in the Supporting Information.
(48) Riley, R. T.; Plattner, R. D. Methods Enzymol. 2000, 311, 348.

12200 J. Am. Chem. Soc., Vol. 123, No. 49, 2001
Trost and Lee
1750, 1656, 1224, 1113, 1007, 701 cm-1; 1H NMR (300 MHz, CDCl3)
δ 8.07-8.04 (m, 2H), 7.72-7.67 (m, 4H), 7.63-7.36 (m, 9H), 6.13-
5.98 (m, 2H), 5.57 (d, J ) 6.9 Hz, 1H), 4.36-4.22 (m, 2H), 2.00 (s,
3H), 1.49 (s, 3H), 1.09 (s, 9H); 13C NMR (75 MHz, CDCl3) δ 178.9,
169.3, 161.0, 137.7, 135.6, 135.5, 133.5, 133.4, 132.9, 129.8, 128.8,
128.2, 127.8, 125.9, 122.0, 76.6, 72.1, 63.1, 26.6, 20.8, 20.2, 19.1;
HRMS calcd for C28H26NO5Si (M+ - t-C4H9) 484.1581, found
484.1587.
7.64 (m, 4H), 7.55-7.38 (m, 9H), 6.42 (s, 1H), 5.89 (d, J ) 8.8 Hz,
1H), 4.99 (dd, J ) 8.8, 1.2 Hz, 1H), 3.93-3.86 (m, 1H), 3.8-3.73 (m,
2H), 2.38 (d, J ) 4.2 Hz, 1H), 2.17 (s, 3H), 1.63 (s, 3H), 1.07 (s, 9H);
13C NMR (75 MHz, CDCl3) δ 174.6, 170.6, 167.4, 135.6, 132.9, 132.8,
132.6, 132.3, 130.0, 128.7, 128.0, 127.3, 76.8, 75.5, 68.6, 64.7, 58.8,
26.7, 20.6, 19.1, 17.9; HRMS calcd for C32H38NO7Si (MH+) 576.2418,
found 576.2422.
(1′S,6R,3aS,6aR)-6-[2′-tert-Butyldiphenylsilyloxy-1′-(4′′-methoxy-
benzyloxy)ethyl]-3a-methyl-2-phenyl-6,6a-dihydro-3aH-furo[3,4-d]-
oxazol-4-one (38). To a mixture of alcohol 37a (0.175 g, 0.268 mmol),
DMAP (1.6 mg, 0.013 mmol, 5%), and pyridine (0.11 mL, 1.36 mmol)
in 1,2-dichloroethane (2.5 mL) was added dropwise trifluoromethane-
sulfonic anhydride (0.065 mL, 0.386 mmol). The mixture was stirred
at 25 °C for 10 min and heated under reflux for 2 h. The mixture was
cooled to room temperature, diluted with ethyl acetate (10 mL), and
poured into 2 N HCl (10 mL). The organic phase was separated, and
the aqueous phase was extracted with ethyl acetate (10 mL × 2). The
combined organic phases were washed with 10% aqueous NaHCO3
and brine and dried over MgSO4. Purification by flash chromatography
(petroleum ether/ethyl acetate ) 10:1) on a SiO2 column afforded 0.164
g (96%) of oxazoline 38 as a colorless, viscous liquid: [R]D +25.0 (c
(1′R,4R,2′E)-4-[1′-Acetoxy-4′-(tert-butyldiphenylsilyloxy)but-2′-
en-1′-yl]-4-methyl-2-phenyl-4,5-dihydro-5-oxazolone (22): [R]D -33.3
(c 2.81, CHCl3, 89% ee); IR (film) 1824, 1750, 1656, 1228, 1113, 1007,
1
701 cm-1; H NMR (300 MHz, CDCl3) δ 8.02-7.99 (m, 2H), 7.66-
7.33 (m, 13H), 6.03-6.02 (m, 2H), 5.61 (d, J ) 6.5 Hz, 1H), 4.23-
4.21 (m, 2H), 1.98 (s, 3H), 1.50 (s, 3H), 1.01 (s, 9H); 13C NMR (75
MHz, CDCl3) δ 177.9, 169.6, 161.4, 137.4, 135.5, 133.5, 133.4, 133.0,
129.79, 129.75, 128.9, 128.1, 127.8, 125.7, 121.0, 76.4, 71.7, 63.0,
26.6, 20.8, 20.2, 19.1; HRMS calcd for C28H26NO5Si (M+ - t-C4H9)
484.1580, found 484.1603.
Methyl (2S,3R,4E)-3-Acetoxy-2-benzamido-2-methyl-6-(tert-bu-
tyldiphenylsilyloxy)-4-hexenoate (23). A mixture of azlactone 21 (5.86
g, 10.8 mmol) and p-toluenesulfonic acid monohydrate (21 mg, 0.11
mmol) in methanol (15 mL) was stirred at 25 °C for 8 h. The mixture
was concentrated and separated on a SiO2 column (petroleum ether/
ethyl acetate ) 10:1) to yield 6.15 g (99%) of methyl ester 23 as a
colorless, sticky oil: tr (2S,3R) ) 7.37 min, tr (2R,3S) ) 17.99 min
(Chiralpak AD, λ ) 230 nm, Hep:i-PrOH ) 9:1); [R]D -17.0 (c 1.25,
CHCl3, 89% ee); IR (film) 3385, 1749, 1669, 1522, 1488, 1234, 1113
1.65, CHCl3); IR (film) 1787, 1645, 1514, 1299, 1249, 1112, 701 cm-1
;
1H NMR (300 MHz, CDCl3) δ 7.89-7.87 (m, 2H), 7.75-7.71 (m,
4H), 7.54-7.36 (m, 9H), 7.14 (d, J ) 8.7 Hz, 2H), 6.79 (d, J ) 8.7
Hz, 2H) 5.03 (dd, J ) 7.0, 5.0 Hz, 1H), 4.95 (d, J ) 5.0 Hz, 1H),
4.54-4.46 (m, 2H), 4.08-3.99 (m, 2H), 3.86-3.81 (m, 1H), 3.78 (s,
3H), 1.72 (s, 3H), 1.10 (s, 9H); 13C NMR (75 MHz, CDCl3) δ 175.1,
164.5, 159.1, 135.7, 135.6, 133.1, 132.7, 132.3, 130.1, 130.0, 129.1,
128.7, 128.5, 127.92, 127.86, 126.0, 113.6, 84.6, 81.2, 78.3, 76.5, 72.5,
63.1, 55.1, 26.7, 20.6, 19.1. Anal. Calcd for C38H41NO6Si: C, 71.78;
H, 6.50; N, 2.20. Found: C, 71.83; H, 6.48; N, 2.04.
1
cm-1; H NMR (300 MHz, CDCl3) δ 7.78-7.75 (m, 2H), 7.67-7.63
(m, 4H), 7.54-7.34 (m, 9H), 7.24 (s, 1H), 5.92-5.91 (m, 2H), 5.73
(d, J ) 4.7 Hz, 1H), 4.22 (s, 2H), 3.70 (s, 3H), 2.16 (s, 3H), 1.72 (s,
3H), 1.06 (s, 9H); 13C NMR (75 MHz, CDCl3) δ 171.6, 171.1, 167.5,
136.2, 135.5, 134.1, 133.4, 133.3, 131.8, 129.9, 128.7, 127.8, 127.1,
121.7, 77.3, 63.6, 63.1, 52.7, 26.7, 21.1, 19.1, 17.6. Anal. Calcd for
C33H39NO6Si: C, 69.08; H, 6.85; N, 2.44. Found: C, 68.81; H, 6.71;
N, 2.43.
Preparation of 30 and 31. Procedure A (from 21). To a mixture
of azlactone 21 (0.167 g, 0.308 mmol) and NMO (54 mg, 0.461 mmol)
in CH2Cl2 (3 mL) was added a 4% aqueous solution of OsO4 (0.0038
mL, 0.0062 mmol). The mixture was stirred at 25 °C for 24 h. Upon
complete consumption of the starting material, the reaction mixture
was diluted with ethyl acetate (5 mL) and an aqueous solution of Na2-
SO3 (∼0.1 g in 3 mL of water) and stirred for 2 h. The organic phase
was separated, and the aqueous phase was extracted with ethyl acetate
(5 mL × 2). The combined organic layers were washed with 2 N HCl,
10% NaHCO3, and brine, dried over MgSO4, and concentrated.
Purification by flash column chromatography (petroleum ether/ethyl
acetate ) 7:1) on SiO2 gave 0.152 g (white solid, 86%) of alcohol 30
and 0.0178 g (sticky oil, 10%) of alcohol 31.
Procedure B (from 23). Following the same procedure as above,
methyl ester 23 (1.98 g, 3.45 mmol) was reacted with NMO (0.606 g,
5.17 mmol) and OsO4 (4% in water, 0.42 mL, 0.069 mmol) in CH2Cl2
(10 mL) at 25 °C for 16 h. Purification by flash column chromatography
(petroleum ether/ethyl acetate ) 7:1) on SiO2 gave 2.02 g of crude
alcohol 30 as a sticky oil which, upon recrystallization from methylene
chloride/pentane (1:10), afforded 1.97 g (98%) of white solid.
(1′S,3S,4S,5R)-3-Methyl-3-benzamido-4-acetoxy-5-(2′-tert-butyl-
diphenylsilyloxy-1′-hydroxyethyl)oxolan-2-one (30): mp 90-91 °C
(CH2Cl2-pentane); [R]D +24.6 (c 1.98, CHCl3); IR (film) 3486, 3302,
1790, 1760, 1738, 1634, 1538, 1228, 1113 cm-1; 1H NMR (300 MHz,
CDCl3) δ 7.79-7.76 (m, 2H), 7.69-7.64 (m, 4H), 7.52-7.35 (m, 9H),
6.70 (s, 1H), 5.96 (d, J ) 7.3 Hz, 1H), 4.65 (dd, J ) 7.3, 2.5 Hz, 1H),
3.99-3.90 (m, 1H), 3.84 (d, J ) 6.3 Hz, 2H), 3.35 (d, J ) 6.9 Hz,
1H), 2.13 (s, 3H), 1.50 (s, 3H), 1.07 (s, 9H); 13C NMR (75 MHz, CDCl3)
δ 173.8, 170.5, 167.7, 135.6(2), 133.2, 133.1, 132.5, 132.4, 129.9, 128.7,
127.9, 127.4, 78.5, 75.3, 70.9, 63.7, 59.9, 26.7, 20.5, 19.1, 17.5. Anal.
Calcd for C32H37NO7Si: C, 66.76; H, 6.48; N, 2.43. Found: C, 66.50;
H, 6.61; N, 2.42.
(2E,1′S,6R,3aS,6aR)-6-[10′,10′-Ethylenedioxy-1′-(4′′-methoxyben-
zyloxy)hexadec-2′-en-1′-yl]-3a-methyl-2-phenyl-6,6a-dihydro-3aH-
furo[3,4-d]oxazol-4-one (46). To a solution of 44 (10.7 mg, 0.445
mmol) in THF (0.5 mL) was added a solution of 9-BBN-H (0.5 M in
THF, 0.10 mL, 0.050 mmol) at 25 °C. After being stirred for 1 h, the
solution was treated with degassed water (0.025 mL, 1.4 mmol) and
transferred to a mixture of vinyl iodide 41 (17.4 mg, 0.0341 mmol),
[1,1′-bis(diphenylphosphino)ferrocene]palladium dichloride (1.25 mg,
0.0017 mmol, 5 mol %), triphenylarsine (0.52 mg, 0.0017 mmol, 5
mol %), and Cs2CO3 (14.4 mg, 0.442 mmol) in dimethylformamide
(0.5 mL). After being stirred at 25 °C for 4 h, the dark brown mixture
was poured into brine (5 mL). Extraction with ether (5 mL × 3),
evaporation of solvent, and flash chromatography (petroleum ether/
ethyl acetate ) 10:1) on a SiO2 column afforded 20.4 mg (94%) of
olefin 46 as a pale yellow oil: [R]D +19.2 (c 2.25, CHCl3); IR (film)
1
1788, 1645, 1515, 1453, 1298, 1249, 1084, 696 cm-1; H NMR (300
MHz, CDCl3) δ 7.92 (dd, J ) 7.1, 1.2 Hz, 2H), 7.52 (t, J ) 7.4 Hz,
1H), 7.41 (t, J ) 7.2 Hz, 2H), 7.20 (d, J ) 8.7 Hz, 2H), 6.82 (d, J )
8.6 Hz, 2H), 6.02 (dt, J ) 15.6, 6.6 Hz, 1H), 5.54 (dd, J ) 15.6, 7.5
Hz, 1H), 4.85 (d, J ) 4.6 Hz, 1H), 4.58 (dd, J ) 8.1, 5.2 Hz, 1H),
4.54 (d, J ) 11.2 Hz, 1H), 4.35 (d, J ) 11.0 Hz, 1H), 4.15 (t, J ) 7.7
Hz, 1H), 3.90 (s, 4H), 3.77 (s, 3H), 2.15 (q, J ) 7.0 Hz, 2H), 1.69 (s,
3H), 1.62-1.54 (m, 4H), 1.47-1.27 (m, 16H), 0.87 (t, J ) 6.7 Hz,
3H); 13C NMR (75 MHz, CDCl3) δ 175.2, 164.7, 159.3, 138.3, 132.3,
130.1, 129.2, 128.8, 128.5, 126.2, 125.6, 124.5, 113.8, 111.9, 84.6,
82.6, 78.2, 76.5, 70.5, 64.8, 55.2, 37.1, 32.4, 31.7, 30.2, 29.64, 29.59,
29.5, 29.1, 28.9, 23.7, 22.5, 20.3, 13.0; HRMS calcd for C38H52NO7
(MH+) 634.3744, found 634.3764.
(6E,2S,3R,4R,5S)-2-Amino-2-methyl-3,4,5-trihydroxy-14-oxoeicos-
6-enoic Acid (Sphingofungin F, 2). A mixture of benzamidolactone
47 (14.2 mg, 0.0291 mmol) and 1 N NaOH (1 mL) was heated under
reflux for 8 h. The reaction mixture was cooled to room temperature,
and Amberlite IRC-76 resin (pretreated and washed with 2 N HCl and
deionized water) was added until the pH of the solution reached
approximately 7. The resin was filtered off, and the filtrate was
concentrated in vacuo to give a white solid. The white solid was
dissolved in a minimal amount of methanol-water and separated on a
SiO2 column (2.5 g; CHCl3/CH3OH/H2O ) 20:5:1) to yield 9.1 mg
(78%) of pure sphingofungin F (2) as a white solid: mp 143-145 °C
(CH3OH-H2O); [R]D +0.99 (c 0.25, CH3OH); IR (KBr) 3423, 1718,
(1′R,3S,4S,5S)-3-Methyl-3-benzamido-4-acetoxy-5-(2′-tert-butyl-
diphenylsilyloxy-1′-hydroxyethyl)oxolan-2-one (31): [R]D +30.5 (c
1.15, CHCl3); IR (film) 3358, 1772, 1750, 1645, 1530, 1233, 1113,
1051 cm-1; 1H NMR (300 MHz, CDCl3) δ 7.79-7.76 (m, 2H), 7.68-

Total Syntheses of Sphingofungins E and F
J. Am. Chem. Soc., Vol. 123, No. 49, 2001 12201
1
1629, 1459, 1408, 1113, 790 cm-1; H NMR (500 MHz, CD3OD) δ
5.91 (dd, J ) 15.4, 3.7 Hz, 1H), 5.53 (d, J ) 8.3 Hz, 1H), 4.30-4.18
(m, 2H), 1.88 (s, 3H), 1.55 (d, J ) 14.9 Hz, 1H), 1.41 (d, J ) 14.9 Hz,
1H), 1.03 (s, 9H), 0.32 (s, 3H), 0.31 (s, 3H); 13C NMR (75 MHz, CDCl3)
δ 178.1, 169.3, 160.9, 137.5, 137.1, 135.4, 133.7, 133.4, 133.0, 132.7,
129.7, 129.2, 128.7, 128.0, 127.74, 127.67, 125.6, 120.8, 127.74, 127.67,
125.6, 120.8, 77.8, 73.4, 63.1, 26.7, 21.6, 20.9, 19.2, -1.8, -2.2; HRMS
calcd for C40H46NO5Si2 (MH+) 676.2915, found 676.2904.
5.77 (dt, J ) 15.4, 6.8 Hz, 1H), 5.45 (dd, J ) 15.5, 7.8 Hz, 1H), 4.10
(t, J ) 7.5 Hz, 1H), 3.86 (d, J ) 0.7 Hz, 1H), 3.67 (d, J ) 7.3 Hz,
1H), 2.44 (t, J ) 7.3 Hz, 2H), 2.44 (t, J ) 7.5 Hz, 2H), 2.06 (q, J )
6.8 Hz, 2H), 1.57-1.50 (m, 4H), 1.48 (s, 3H), 1.42-1.38 (m, 2H),
1.36-1.23 (m, 10H), 0.90 (t, J ) 7.0 Hz, 3H); 13C NMR (100 MHz,
CD3OD) δ 214.4, 175.2, 135.7, 130.2, 76.2, 75.7, 72.4, 66.7, 43.5(2),
33.5(2), 32.8, 30.2(2), 30.0(2), 24.9(2), 23.6, 21.8, 14.4; LRMS calcd
for C21H39NNaO6 (M + Na+) 423.3, found 423.3.
Pd-Catalyzed Asymmetric Alkylation: Preparation of 53 and
54. To a suspension of sodium hydride (95%, 0.110 g, 4.35 mmol) in
THF (5.0 mL) was added azlactone 10b (1.50 g, 4.85 mmol). The
resulting solution was stirred at 25 °C until the hydrogen gas evolution
ceased. After brief sonication (0.5 h), a mixture of 13 (8.0 mg, 0.022
mmol, 1 mol %) and (R,R)-14 (45.4 mg, 0.066 mmol, 3 mol %) in
THF (1 mL) was added. After the mixture was purged with a stream
of nitrogen, a solution of geminal diacetate 9 (0.936 g, 2.19 mmol) in
THF (6.0 mL) was added at 0 °C. After the addition was complete, the
stirring was continued at 0 °C for 3 h. The reaction mixture was poured
into 10% aqueous NaH2PO4 (20 mL) and extracted with ether (30 mL
× 3). The combined organic layers were washed with brine, dried over
MgSO4, and concentrated. The diastereomeric ratio was determined to
(6E,2R,3R,4R,5S)-2-Amino-2-hydroxymethyl-3,4,5-trihydroxy-14-
oxoeicos-6-enoic acid (Sphingofungin E, 1). To a solution of
benzamidolactone 74 (10.2 mg, 0.0203 mmol) in methanol (0.5 mL)
was added 1 N NaOH (1 mL) at 25 °C. The mixture was heated at 65
°C for 2 h and cooled to room temperature. Amberlite IRC-76 resin
(∼1 g, pretreated with 2 N HCl and washed with deionized water)
was added, and the mixture was stirred until the pH of the solution
reached approximately 7. The resin was filtered with the aid of methanol
(2 mL), and the filtrate was concentrated in vacuo to give a pale yellow
solid. The solid was adsorbed onto SiO2 that was suspended in a
minimal amount of methanol. The suspension was carefully loaded on
a SiO2 column (2.5 g) and eluted with a chloroform-methanol-water
(20:5:1) mixture to give 5.4 mg (64%) of pure sphingofungin E (1) as
a white flaky powder: mp 154-155 °C (CH3OH-H2O); [R]D -3.5 (c
0.31, CH3OH); IR (KBr) 3538, 3357 (br), 3200 (br), 3113 (br), 2929,
1
1
be 2.4:1 by H NMR analysis of the crude mixture. Separation on a
2853, 1708, 1639, 1519, 1466, 1397, 1071, 971, 722 cm-1; H NMR
SiO2 column (petroleum ether/ethyl acetate ) 20:1) yielded 1.01 g
(68%) of the less polar major diastereomer 53 and 0.340 g (23%) of
the more polar minor diastereomer 54, both as colorless, sticky oils.
The ee of both products was determined to be 96% by chiral HPLC
analysis of the corresponding methyl esters 55 and 56.
(300 MHz, CD3OD) δ 5.76 (dt, J ) 15.6, 6.3 Hz, 1H), 5.44 (dd, J )
15.4, 7.6 Hz, 1H), 4.10 (t, J ) 7.3 Hz, 1H), 3.97 (d, J ) 11.0 Hz, 1H),
3.94 (br s, 1H), 3.84 (d, J ) 10.7 Hz, 1H), 3.63 (d, J ) 7.1 Hz, 1H),
2.44 (t, J ) 7.2 Hz, 4H), 2.05 (q, J ) 6.6 Hz, 2H), 1.54 (br s, 4H),
1.41-1.28 (m, 12H), 0.90 (t, J ) 6.6 Hz, 3H); 13C NMR (75 MHz,
CD3OD) δ 214.4, 173.2, 135.7, 130.2, 76.3, 75.6, 71.2, 70.1, 64.9, 43.5-
(2), 33.4, 32.8, 30.18, 30.15, 30.0(2), 24.89, 24.87, 23.6, 14.4; HRMS
calcd for C21H36O7 (M+ - NH3) 400.2460, found 400.2483; LRMS
calcd for C21H36O7 (M+ - H) 416.3, found 416.3.
(1′R,4S,2′E)-4-[1′-Acetoxy-4′-(tert-butyldiphenylsilyloxy)but-2′-en-
1′-yl]-4-(dimethylphenylsilyl)methyl-2-phenyl-4,5-dihydro-5-oxazolo-
ne (53): [R]D -59.5 (c 5.12, CHCl3, 96% ee); IR (film) 1819, 1750,
1657, 1428, 1225, 1114, 1045, 1023, 973, 839, 738, 702 cm-1; 1H NMR
(300 MHz, CDCl3) δ 7.88-7.85 (m, 2H), 7.68-7.64 (m, 4H), 7.61-
7.56 (m, 1H), 7.49-7.17 (m, 13H), 6.06 (dt, J ) 15.6, 3.2 Hz, 1H),
5.97 (dd, J ) 15.6, 7.3 Hz, 1H), 5.47 (d, J ) 7.3 Hz, 1H), 4.30-4.18
(m, 2H), 1.94 (s, 3H), 1.62 (d, J ) 14.7 Hz, 1H), 1.45 (d, J ) 14.6 Hz,
1H), 1.06 (s, 9H), 0.29 (s, 3H), 0.28 (s, 3H); 13C NMR (75 MHz, CDCl3)
δ 179.1, 169.2, 160.5, 138.1, 136.9, 135.4, 133.7, 133.4, 133.3, 132.6,
129.6(2), 129.2, 128.6, 128.1, 127.7, 125.8, 121.9, 78.3, 74.0, 63.2,
26.7, 21.4, 20.9, 19.2, -1.9, -2.4. Anal. Calcd for C40H45NO5Si2: C,
71.07; H, 6.71; N, 2.07. Found: C, 71.21; H, 6.89; N, 2.03.
Acknowledgment. We thank the National Science Founda-
tion and the National Institute of Health, General Medical
Sciences, for their generous support of our programs. Mass
spectra were provided by the Mass Spectrometry Regional
Center of the University of California-San Francisco, supported
by the NIH Division of Research Resources. Dr. Athar Masood
is gratefully acknowledged for the X-ray analysis.
(1′R,4R,2′E)-4-[1′-Acetoxy-4′-(tert-butyldiphenylsilyloxy)but-2′-
en-1′-yl]-4-dimethylphenylsilylmethyl-2-phenyl-4,5-dihydro-5-ox-
azolone (54): [R]D -5.5 (c 1.12, CHCl3, 96% ee); IR (film) 1818,
Supporting Information Available: Detailed descriptions
of experimental procedures and spectral data for all new
compounds (PDF). This material is available free of charge via
the Internet at http://pubs.acs.org.
1
1751, 1654, 1428, 1226, 1113, 1024, 980, 824, 701 cm-1; H NMR
(300 MHz, CDCl3) δ 7.88-7.85 (m, 2H), 7.67-7.64 (m, 4H), 7.60-
7.51 (m, 1H), 7.48-7.19 (m, 13H), 6.06 (dt, J ) 15.4, 7.9 Hz, 1H),
JA0118338