Scavenging byproducts in the sulfoxide glycosylation reaction: Application to the synthesis of ciclamycin 0

Article abstract of DOI:10.1021/ja990690a

We have developed a direct and efficient route for the synthesis of ciclamycin 0 using the sulfoxide glycosylation reaction to form the glycosidic linkages. In the course of completing the synthesis of ciclamycin 0, we developed new glycosylation conditions that improve the outcome of the sulfoxide reaction. The conditions involve the addition of agents that scavenge phenylsulfenyl triflate, a highly reactive byproduct that forms following activation of anomeric sulfoxides with triflic anhydride.

Full text of DOI:10.1021/ja990690a

6176
J. Am. Chem. Soc. 1999, 121, 6176-6182
Scavenging Byproducts in the Sulfoxide Glycosylation Reaction:
Application to the Synthesis of Ciclamycin 0
Jeff Gildersleeve, Andri Smith, Kaori Sakurai, Subharekha Raghavan, and Daniel Kahne*
Contribution from the Department of Chemistry, Princeton UniVersity, Princeton, New Jersey 08544
ReceiVed March 3, 1999
Abstract: We have developed a direct and efficient route for the synthesis of ciclamycin 0 using the sulfoxide
glycosylation reaction to form the glycosidic linkages. In the course of completing the synthesis of ciclamycin
0, we developed new glycosylation conditions that improve the outcome of the sulfoxide reaction. The conditions
involve the addition of agents that scavenge phenylsulfenyl triflate, a highly reactive byproduct that forms
following activation of anomeric sulfoxides with triflic anhydride.
Introduction
The anthracycline antibiotics are a large class of glycosylated
natural products which possess a wide range of biological
activity.1,2 Two of the most notable members, adriamycin (1)
and daunomycin (2) (see Figure 1), have been used clinically
to treat certain types of cancer. Like many other glycosylated
natural products, the carbohydrates on the anthracyclines are
essential for biological activity. Furthermore, even subtle
changes to the carbohydrates can significantly improve biologi-
cal activity and/or reduce toxicity. While this class of natural
products has been studied for over 40 years, their mechanism
of action and the roles of the carbohydrates are still not clear.
Our group has been interested in studying the roles of
carbohydrates on glycosylated natural products for a number
of years, and we were naturally drawn to the anthracycline
antibiotics. We are especially interested in ciclamycin, a complex
of anthracyclines isolated from Streptomyces capoamus that
possesses good activity against tumors both in vitro and in
vivo.3,4 Since most anthracyclines in the ciclamycin complex
contain the same aglycon, differences in activity and toxicity
are due to differences in the oligosaccharides. Therefore, we
wanted to develop a synthetic route to this class of anthracy-
clines that would allow us to systematically vary the oligosac-
charides to shed more light on the role of the carbohydrates
and the mechanisms by which these compounds work.
Figure 1. Anthracycline antibiotics.
for the synthesis of ciclamycin 0, a member of the ciclamycin
complex that has particularly low toxicity.6,7 Because ciclamycin
0 has been synthesized once previously by the Danishefsky
group using one of the best glycosylation methods available,8,9
we had a standard which would allow us to compare our
synthesis and glycosylation method.
Background
Several years ago, we reported the synthesis of the trisac-
charide of ciclamycin 0 using the sulfoxide glycosylation
reaction.7,10 Our synthesis of the trisaccharide used a polym-
erization reaction in which the trisaccharide was constructed in
As part of our program to study the roles of carbohydrates
on glycosylated natural products, we have been interested in
improving methods to synthesize oligosaccharides to make these
compounds more readily available. In 1989 we reported the use
of anomeric sulfoxides as glycosyl donors and have found
several advantages to using this glycosylation method: activa-
tion occurs at low temperature under mild conditions, and the
activated glycosyl donor produced is extremely reactive.5 In
1993, we began investigating the use of the sulfoxide reaction
(5) Kahne, D.; Walker, S.; Cheng, Y.; Van Engen, D. J. Am. Chem. Soc.
1989, 111, 6881.
(6) For the isolation and characterization of ciclamycin 0, see: Bieber,
L. W.; Da Silva Filho, A. A.; De Mello, J. F.; De Lima, O. G.; Do
Nascimento, M. S.; Veith, H. J.; Von Der Saal, W. J. Antibiot. 1987, 40,
1335.
(7) Raghavan, S.; Kahne, D. J. Am. Chem. Soc. 1993, 115, 1580.
(8) Suzuki, K.; Sulikowski, G. A.; Friesen, R. W.; Danishefsky, S. J. J.
Am. Chem. Soc. 1990, 112, 8895.
(9) For recent reviews on the glycal method for the synthesis of
oligosaccharides, see: (a) Thiem, J.; Klaffke, W. Top. Curr. Chem. 1990,
154, 285. (b) Danishefsky, S. J.; Bilodeau, M. T. Angew. Chem., Int. Ed.
Engl. 1996, 35, 1381.
(10) For lead references to the syntheses of related oligosaccharides,
see: (a) Monneret, C.; Martin, A.; Pais, M. J. Carbohydr. Chem. 1988, 7,
417. (b) Thiem, J. In Trends in Synthetic Carbohydrate Chemistry; Horton,
D., Hawkins, L. D., McGarvey, G. J., Eds.; American Chemical Society:
Washington, DC, 1989; Chapter 8. (c) Horton, D.; Priebe, W.; Sznaidman,
M. L. J. Antibiot. 1993, 46, 1720. (d) Kolar, C.; Bosslet, K.; Czech, J.;
Gerken, M.; Hermentin, P.; Hoffmann, D.; Sedlacek, H. Chapter 4 in ref 2.
(e) Animati, F.; Arcamone, F.; Berettoni, M.; Cipollone, A.; Fanciotti, M.;
Lombardi, P. J. Chem. Soc., Perkin Trans. 1 1996, 1327.
(1) Lown, W. Ed. Anthracycline and Anthracenedione-Based Anticancer
Agents; Elsevier Science Publishers: Amsterdam, 1988.
(2) Priebe, W., Ed. Anthracycline Antibiotics: New Analogues, Methods
of DeliVery, and Mechanisms of Action; American Chemical Society:
Washington, DC, 1995.
(3) Lima, V. Q. G.; Albert, C. A.; Lima, O. G. An. Acad. Bras. Cienc.
1964, 36, 317.
(4) Lyra, F. D. A.; Lima, O. G.; Coelho, J. S. B.; Albuquerque, M. M.
F.; Maciel, G. M.; Oliviera, L. L.; Maciel, M. C. N. An. Acad. Bras. Cienc.
1964, 36, 323.
10.1021/ja990690a CCC: $18.00 © 1999 American Chemical Society
Published on Web 06/16/1999

Sulfoxide Glycosylation Reaction
J. Am. Chem. Soc., Vol. 121, No. 26, 1999 6177
Scheme 1. One-Step Synthesis of Trisaccharides
Scheme 2. Formation of Trisaccharide 13 during the AB Glycosylation
a single reaction from three monomers, 4a, 5a, and 6 (see
In a model study, we found that p-methoxybenzyl (PMB)
Scheme 1).11 The reactivities of the sulfoxides and nucleophiles
were tuned such that the linkage between the A ring and B ring
would occur first, followed by formation of the linkage to the
C ring. Using the polymerization strategy, trisaccharide 7a could
be synthesized in 25% yield in a single reaction. Oxidation of
the anomeric sulfide produced trisaccharide sulfoxide 8 in 80%
yield. To complete the synthesis of ciclamycin 0, we simply
had to couple the trisaccharide to the aglycone ꢀ-pyrromycinone
(9)12 and deprotect the final product (10). Unfortunately, the
yield for glycosylating the aglycon was disappointing (16%),
and we were unable to deprotect the benzyl ethers because the
hydrogenation conditions also cleaved the benzylic glycosidic
linkage to the aglycon.13 To develop a good synthesis of
ciclamycin 0, we needed to improve the coupling of the
trisaccharide to the aglycon and modify the protecting groups
on the trisaccharide.
ethers could be deprotected with DDQ without cleaving the
trisaccharide from the aglycon. Therefore, the A and B ring
monomers were resynthesized with PMB protecting groups (4b
and 5b) and subjected to the conditions for the one-step
synthesis. However, the modified monomers did not work as
well in the one-step coupling reaction, and trisaccharide 7b could
only be isolated in 10% yield. A number of parameters were
varied, but the one-step reaction could not be optimized further.
The substrate-dependent coupling yields suggested that the
one-step reaction would not be amenable to the synthesis of
large numbers of trisaccharide derivatives of ciclamycin 0.
Therefore, we decided to build the trisaccharide in a stepwise
manner. Our studies on the individual glycosylations led to some
unexpected results which prompted a series of mechanistic
investigations. On the basis of new insights into the sulfoxide
reaction, we have been able to improve the glycosylation
conditions, develop a better synthesis of the trisaccharide, and
complete the synthesis of ciclamycin 0.
(11) For other one-pot syntheses of oligosaccharides, see: (a) Yamada,
H.; Harada, T.; Takahashi, T. J. Am. Chem. Soc. 1994, 116, 7919 and
references therein. (b) Chenault, H. K.; Castro, A. Tetrahedron Lett. 1994,
49, 9145. (c) Green, L.; Hinzen, B.; Ince, S. J.; Langer, P.; Ley, S. V.;
Warriner, S. L. Synlett 1998, 440 and references therein. (d) Zhang, Z.;
Ollmann, I. R.; Ye, X.-S.; Wischnat, R.; Baasov, T.; Wong, C. H. J. Am.
Chem. Soc. 1999, 121, 734.
(12) Pyrromycinone, 9, was obtained by acidic hydrolysis of marcello-
mycin, a generous gift from Bristol-Myers Squibb. Pyrromycinone can also
be obtained through known synthetic routes or through degradation of other
readily available anthracyclines; see: (a) Hauser, F. M.; Mal, D. J. Am.
Chem. Soc. 1984, 106, 1098. (b) Hoshino, T.; Setoguchi, Y.; Fujiwara, A.
J. Antibiot. 1984, 37, 1469. (c) Krohn, K.; Kilmars, M.; Kohle, H.-J.;
Ebeling, E. Tetrahedron 1984, 40, 3677.
Results and Discussion
AB Glycosylation and Mechanistic Studies. We began by
investigating the coupling of A ring derivative 4b with B ring
sulfoxide 11 (see Scheme 2). To prevent oligomerization, the
C-4 hydroxyl of the B ring sulfoxide was protected as an acetate
instead of the more labile TMS ether used in our original
synthesis of the trisaccharide. In addition to this structural
modification, we utilized an inverse addition procedure to
activate the sulfoxide because we have found that this strategy
(13) Raghavan, S. Ph.D. Thesis, Princeton University, 1994.

6178 J. Am. Chem. Soc., Vol. 121, No. 26, 1999
GildersleeVe et al.
Scheme 3. Potential Mechanism for Trisaccharide Formation
minimizes the formation of anomeric sulfenates, thereby
improving the yield of glycosylations.14,15 Glycosylation of 4b
with 11 produced disaccharide 12 in 33% yield along with a
large number of other products. Surprisingly, one of the other
products formed in the reaction was a trisaccharide, 13. Since
acetates are stable to the conditions of the sulfoxide reaction, it
was not immediately obvious how the trisaccharide was being
formed.
which react rapidly with phenylsulfenyl triflate at -78 °C. In
addition, the trap must react with phenylsulfenyl triflate
selectively in the presence of other reactive electrophiles, such
as triflic anhydride and the activated glycosyl intermediate(s).
Given the high reactivity of the activated glycosyl intermedi-
ate(s), it was not obvious that any reagent would satisfy these
requirements.
An extensive search of the literature suggested that unacti-
vated alkenes might have the appropriate reactivity.20-22 How-
ever, reaction of an alkene with phenylsulfenyl triflate would
produce an episulfonium ion which might react with the alcohol
acceptor (see Figure 2). To determine whether the addition of
an alkene would affect the outcome of the reaction, we repeated
the glycosylation of 4b with 11 in the presence of 10 equiv of
4-allyl-1,2-dimethoxybenzene.23 In the presence of this alkene,
the glycosylation reaction appeared much cleaner by TLC and
the yield of disaccharide improved considerably (from 33% to
71%). No alkylation of the starting alcohol was observed. In
addition, we were able to isolate compound 14, demonstrating
that the alkene is sulfenylated in the reaction. The isolation of
14 supports the hypothesis that phenylsulfenyl triflate causes
trisaccharide formation and suggests that the alkene improves
the outcome by removing phenylsulfenyl triflate from the
reaction mixture.
One potential explanation for the formation of trisaccharide
in the glycosylation of 4b with 11 came from consideration of
the byproducts that may form in the sulfoxide reaction.
Activation of an anomeric sulfoxide with triflic anhydride is
likely to release phenylsulfenyl triflate (see Scheme 3).13,16,17
Others have found that alkyl and aromatic sulfenyl triflates
activate anomeric sulfides rapidly at low temperature.18,19 We
reasoned that if phenylsulfenyl triflate were formed in significant
amounts during the sulfoxide reaction, it might activate the
anomeric sulfide of 12 and lead to the production of trisaccharide
13 (see Scheme 3). Since 4b, 12, and 13 each contain an
anomeric sulfide, this activation process could lead to the
production of a variety of oligomers which might explain why
so many products were formed in the glycosylation reaction.
The hypothesis that phenylsulfenyl triflate was causing problems
in the reaction led us to investigate two different strategies to
block phenylsulfenyl triflate mediated activation of the anomeric
sulfides. The first strategy involved the addition of reagents to
trap phenylsulfenyl triflate, while the second involved reducing
the nucleophilicity of the sulfur atom on the A ring. As we
demonstrate below, these strategies greatly improve the gly-
cosylation reactions.
(1) Methods To Suppress Sulfide Activation: Phenylsul-
fenyl Triflate Scavengers. Identification of specific reagents
to trap phenylsulfenyl triflate is complicated. To be effective, a
trap must be reactive enough to outcompete anomeric sulfides
Figure 2. Sulfenylation of an alkene.
Although the addition of 4-allyl-1,2-dimethoxybenzene im-
proved the glycosylation reaction considerably, trisaccharide
(14) Gildersleeve, J.; Pascal, R. A.; Kahne, D. J. Am. Chem. Soc. 1998,
120, 5961.
(15) For a recent application of inverse addition, see: Ge, M.; Thompson,
C.; Kahne, D. J. Am. Chem. Soc. 1998, 120, 11014.
(16) Crich, D.; Sun, S. J. Am. Chem. Soc. 1997, 119, 11217.
(17) There may be more than one highly reactive sulfenylating agent
produced in the reaction.
(18) Dasgupta, F.; Garegg, P. J. Carbohydr. Res. 1988, 177, C13.
(19) For the use of phenylsulfenyl triflate to activate sulfides, see: (a)
Martichonok, V.; Whitesides, G. M. J. Am. Chem. Soc. 1996, 118, 8187.
(b) Martichonok, V.; Whitesides, G. M. Carbohydr. Res. 1997, 302, 123.
(c) Crich, D.; Sun, S. J. Am. Chem. Soc. 1998, 120, 435.
(20) Hopkins, P. B.; Fuchs, P. L. J. Org. Chem. 1978, 43, 1208.
(21) Jones, G. A.; Stirling, C. J. M.; Bromby, N. G. J. Chem. Soc., Perkin.
Trans. 2 1983, 385.
(22) Patai, S., Ed. The Chemistry of Sulphenic Acids and Their DeriVa-
tiVes; John Wiley and Sons, Ltd.: Chichester, 1990.
(23) This specific alkene was chosen for a number of reasons: it is
commercially available and inexpensive, and has a sufficiently high boiling
point to permit azeotroping with toluene along with the other starting
materials prior to reaction.

Sulfoxide Glycosylation Reaction
J. Am. Chem. Soc., Vol. 121, No. 26, 1999 6179
Scheme 4. Glycosylation of the Modified A Ring
Formation of the ABC Trisaccharide and Completion of
Ciclamycin 0. The next step in the synthesis of ciclamycin 0
involved coupling the AB disaccharide to the C ring. Gly-
cosylations with the C ring sulfoxide have proven difficult in
the past. In a model study using our original glycosylation
conditions, coupling of A ring derivative 4b with 2 equiv of C
ring sulfoxide 6 produced the corresponding AC disaccharide
in 33% yield. We initially thought that the low yield of this
glycosylation was a result of decomposition of the C ring
sulfoxide prior to activation with triflic anhydride. However,
increasing the number of equivalents of sulfoxide to four resulted
in a decrease in yield (24%), indicating that the problem was
not simply due to instability of the C ring.
Our studies on the mechanism of the sulfoxide reaction
suggested that other factors were responsible for the low yield.
We have recently shown that anomeric sulfoxides can rearrange
to sulfenates under the reaction conditions and that this process
decreases the efficiency of a glycosylation reaction. In fact, we
find that the C ring sulfoxide rearranges to sulfenate rapidly at
-78 °C, indicating that sulfenate formation can compete with
glycoside formation for this sulfoxide. However, sulfenate
formation does not appear to be the only problem since
increasing the number of equivalents of sulfoxide leads to a
decrease in yield. One explanation for this unexpected observa-
tion comes from our studies on byproduct-mediated side
reactions. As the amount of sulfoxide is increased, more
byproducts are produced. Since the A ring and AC disaccharide
each contain an anomeric sulfide, increased production of
byproducts could lead to more extensive destruction of the
starting materials and products, resulting in a decrease in yield
for the reaction.
To suppress sulfenate formation and byproduct-mediated side
reactions, the glycosylation of disaccharide 18 with C ring
sulfoxide 6 was conducted using inverse addition in the presence
of a phenylsulfenyl triflate scavenger, 4-allyl-1,2-dimethoxy-
benzene (Scheme 5). Using these glycosylation conditions,
trisaccharide 19 was obtained in 68% yield (5:1 R/â). This
glycosylation reaction provides a good illustration of how the
mechanistic studies can be used to improve a difficult gly-
cosylation reaction.
Completion of ciclamycin 0 required oxidation of the
trisaccharide, attachment to the aglycon, and deprotection.
Oxidation of trisaccharide 19 with mCPBA resulted in Baeyer-
Villiger rearrangement of the C ring ketone. Interestingly, this
problem did not occur during the oxidation of trisaccharide 7a
which contains a simple thiophenyl group on the anomeric
carbon of the A ring. While mCPBA can oxidize a 2,6-
dichlorophenyl sulfide (e.g., synthesis of monosaccharide 17),
the oxidation requires much warmer temperatures than usual
(0 °C vs -45 °C). Apparently, the decreased rate of sulfur
oxidation of trisaccharide 19 results in preferential reaction of
mCPBA with the ketone of the C ring. Happily, we found that
dimethyldioxirane will selectively oxidize trisaccharide 19 to
sulfoxide 20 in 90% yield.26
The last obstacle for the synthesis of ciclamycin 0 involved
coupling the trisaccharide to the aglycon. Our initial work on
the coupling of 7a to the aglycon using our original glyco-
sylation conditions produced the benzyl-protected ciclamycin
derivative 10 in 16% yield. However, the glycosylation of the
aglycon 9 with trisaccharide 20 using our modified conditions
proceeded smoothly to generate PMB-protected ciclamycin
derivative 21 in 75% yield. Removal of the PMB ethers using
DDQ completed the synthesis of ciclamycin 0.
formation was not completely prevented. Several other traps
were investigated, but none was found to be more effective than
4-allyl-1,2-dimethoxybenzene.24 Therefore, we began investigat-
ing a second method for blocking sulfide activation.
(2) Methods To Suppress Sulfide Activation: Hindered
Phenyl Sulfides. The second strategy for blocking sulfide
activation involved the use of hindered sulfides. We expected
that a less nucleophilic sulfide would react more slowly with
phenylsulfenyl triflate, providing a better opportunity for the
scavenger to remove phenylsulfenyl triflate from the reaction
mixture before it could cause deleterious side reactions.
Therefore, a modified A ring with a 2,6-dichlorophenyl sulfide
(15) was synthesized; the two chlorine atoms positioned ortho
to the sulfur atom were expected to decrease the nucleophilicity
of the sulfur atom through steric and inductive effects. Gly-
cosylation of the modified A ring, 15, with sulfoxide 11 using
inverse addition in the presence of 10 equiv of 4-allyl-1,2-
dimethoxybenzene produced disaccharide 16 in 82% yield; no
trisaccharide could be detected (see Scheme 4).25
Before proceeding with the completion of ciclamycin 0, we
needed to establish that the 2,6-dichlorophenyl sulfide could
be oxidized and used as a glycosyl donor. We were particularly
concerned that the 2,6-dichlorophenyl group would cause
problems during the coupling of the trisaccharide sulfoxide to
the aglycon. The steric congestion and electron-withdrawing
effects of the 2,6-dichlorophenyl group could decrease the rate
of activation of the trisaccharide sulfoxide significantly, permit-
ting side reactions such as anomeric sulfenate formation to
occur.14 To address this possibility, we decided to reinvestigate
the AB glycosylation using a 2,6-dichlorophenyl-substituted B
ring sulfoxide, 17. This modified B ring sulfoxide was synthe-
sized from the A ring derivative 15 by acetylation followed by
oxidation with mCPBA. Glycosylation of 15 with sulfoxide 17
produced AB disaccharide 16 in 81% yield (see Scheme 4).
The yield for this glycosylation is virtually identical with the
yield obtained with the less hindered B ring sulfoxide 11,
indicating that the 2,6-dichlorophenyl substituent does not
disrupt the AB glycosylation. These results suggested that the
2,6-dichlorophenyl substituent would not be a problem during
the final coupling of the trisaccharide sulfoxide to the aglycon.
(24) Norbornylene was slightly less effective than 4-allyl-1,2-dimethoxy-
benzene, while vinylnorbornene, vinylnaphthalene, cyclopentene, and
vinyltrimethylsilane were significantly less effective. No product was formed
in the presence of triphenylphosphine or thioanisole.
(25) For comparison, glycosylation of 15 with sulfoxide 11 in the absence
of a scavenger produced 16 in 45% yield along with 18% of the
corresponding trisaccharide, showing that most of the improvement in yield
is a result of the scavenger. However, similar improvements in yield (10-
15%) are observed when the C ring sulfoxide is coupled. These modest
improvements in yield achieved by incorporating the 2,6-dichlorophenyl
sulfide become important when multiple glycosylations are conducted.
(26) Murray, R. M.; Jeyaraman, R.; Pillay, M. K. J. Org. Chem. 1987,
52, 746.

6180 J. Am. Chem. Soc., Vol. 121, No. 26, 1999
Scheme 5. Completion of Ciclamycin 0a
GildersleeVe et al.
a Reaction conditions: (a) LAH/Et2O, 0 °C, 30 min (92%); (b) Tf2O, DTBMP, 4-allyl-3,4-dimethoxybenzene, DCM, -78 °C, then 6 (68%, 5:1
R/â); (c) dimethyldioxrane, -72 to -42 °C (90%); (d) pyrromycinone (9) Tf2O, DTBMP, 4-allyl-3,4-dimethoxybenzene, DCM, -78 °C, then 20
(75%); (e) DDQ, 25 °C, 30 min (60%).
Implications for the Sulfoxide Glycosylation Method.
Recent studies on the mechanism of the sulfoxide reaction have
demonstrated that three different intermediates can be formed,
an oxonium ion, a glycosyl triflate, and a glycosyl sulfenate.14,16
The identification and study of these intermediates has led to
new methods to control stereoselectivity and improve efficiency
in the reaction. The studies on byproduct formation reported in
this paper provide a more complete understanding of the factors
that affect the outcome of a glycosylation reaction. For example,
we have found that byproducts produced during the activation
of a glycosyl sulfoxide will react with anomeric sulfides, causing
undesirable side reactions. This is a fairly serious problem given
that sulfides are versatile intermediates for oligosaccharide
synthesis and that sulfides are convenient precursors to sulfoxide
donors. Fortunately, we find that alkenes added to the reaction
will scavenge the byproducts and prevent decomposition of the
sulfide. We should note that sulfenylating agents react with a
wide range of other functional groups.27,28 Therefore, we think
the scavengers developed for the synthesis of ciclamycin 0 will
be useful in many other sulfoxide glycosylation reactions. By
improving our mechanistic understanding of the sulfoxide
reaction, we now have a better foundation for analyzing and
improving difficult glycosylation reactions.
Experimental Section
General Methods. NMR spectra were recorded on a JEOL GSX
270 or a Varian Inova 500 Fourier transform NMR spectrometer. Proton
chemical shifts are reported in parts per million (ppm) downfield from
tetramethylsilane (TMS) unless otherwise noted. Carbon chemical shifts
are reported in parts per million (ppm) downfield from TMS using the
solvent CD3COCD3 as an internal reference unless otherwise noted.
Coupling constants (J) are reported in hertz (Hz). Multiplicities are
abbreviated as follows: singlet (s), doublet (d), triplet (t), quartet (q),
multiplet (m), and broadened (br). Mass spectra were obtained on a
VG ZAB, VG 7070, or HP 5989A (University of California, Riverside,
Mass Spectrometry Facility).
Analytical thin-layer chromatography (TLC) was performed using
silica gel 60 F254 precoated plates (0.25 mm thickness) with a
fluorescent indicator. Flash column chromatography was performed
using silica gel 60 (230-400 mesh) from EM Science/Bodman.
All reactions were carried out under argon atmosphere with dry,
freshly distilled solvents under anhydrous conditions unless otherwise
noted. All reagents were purchased from commercial suppliers and used
without further purification unless otherwise noted.
2-Hydroxy-3-(3,4-dimethoxyphenyl)propyl Phenyl Sulfide (14).
Rf ) 0.32 (7% EtOAc/methylene chloride); 1H NMR (CDCl3, 270 MHz)
δ 7.29 (m, 5H), 6.81 (d, J ) 7.9 Hz, 1H), 6.73 (d, J ) 7.9 Hz, 1H),
6.72 (s, 1H), 3.92, (m, 1H), 3.86 (s, 3H), 3.84 (s, 3H), 3.13 (dd, J )
3.9, 13.5 Hz, 1H), 2.90 (dd, J ) 7.9, 13.5 Hz, 1H), 2.82 (m, 2H), 2.36
(m, 1H); 13C NMR (CD3COCD3, 125.8 MHz) δ 150.7, 149.5, 138.6,
132.8, 130.3, 130.0, 127.0, 122.9, 115.0, 113.4, 72.5, 56.7, 56.6, 43.5,
41.2; HRFABMS calcd for C17H20O3S (M+) 304.1133, found 304.1116.
Synthesis of the 2,6-Dichlorophenyl-Substituted A Ring Mono-
mer (15). To a solution of 1,3,4-tri-O-acetyl-2,6-dideoxy-R-L-galac-
topyranoside (421 mg, 1.54 mmol) in 20 mL of dichloromethane at
-42 °C was added 2,6-dichlorobenzenethiol (826 mg, 4.63 mmol). The
reaction mixture was stirred at -78 °C for 2 h and then allowed to
warm to 0 °C over 2 h. The reaction mixture was quenched by pouring
it into saturated NaHCO3 solution (20 mL) and allowing it to stir for
2 h. The aqueous layer was extracted with dichloromethane (3 × 10
mL). The organic layers were combined, dried over Na2SO4, filtered,
and concentrated in vacuo. The crude product was purified by flash
chromatography (20% EtOAc/petroleum ether) to afford 2,6-dichlo-
rophenyl 3,4-di-O-acetyl-2,6-dideoxy-1-thio-R-L-galactopyranoside (512
mg, 85%): Rf ) 0.29 (20% EtOAc/petroleum ether); 1H NMR (CDCl3,
500 MHz) δ 7.40 (d J ) 8.0 Hz, 2H), 7.21 (t, J ) 8.0 Hz, 1H), 5.82
(d, J ) 6.0 Hz, 1H), 5.34 (ddd, J ) 3.0, 5.0, 13.0 Hz, 1H), 5.25 (d,
J ) 3.0 Hz, 1H), 4.68 (q, J ) 6.0 Hz, 1H), 2.43 (ddd, J ) 6.0, 13.0,
13.0 Hz, 1H), 2.14 (s, 3H), 2.12 (m, 1H), 2.01 (s, 3H), 1.07 (d J ) 6.5
Hz, 1H); 13C NMR (CDCl 3, 125.8 MHz) δ 170.7, 170.1, 141.7, 131.4,
130.6, 128.9, 84.1, 69.8, 67.4, 67.3, 30.3, 21.0, 20.9, 16.5; HRFABMS
calcd for C16H18O5SCl2 (MNa+) 415.0149, found 415.0128.
Conclusions
Our mechanistic studies on byproducts produced in the
sulfoxide reaction have led to the development of improved
conditions for the reaction. By simply including an alkene such
as 4-allyl-1,2-dimethoxybenzene in the reaction, the yield of a
glycosylation reaction can be improved dramatically. The
modified conditions for the sulfoxide reaction have been used
to develop a short and efficient synthesis of ciclamycin 0. The
synthesis only requires six steps from the starting monomers
and proceeds with an overall yield of 17%. The directness and
efficiency of our synthesis should permit the rapid assembly of
derivatives which can be used to probe the relationships between
structure and activity for ciclamycin 0 and related anthracyclines.
(27) For example, sulfenylating agents are known to react with alcohols,
amines, and electron-rich aromatic rings. See ref 22.
(28) In addition to activating glycosyl sulfides, sulfenylating agents have
been used to activate glycosyl xanthates and glycals. For activation of
xanthates, see: (a) Marra, A.; Sinay, P. Carbohydr. Res. 1990, 195, 303.
(b) Birberg, W.; Lonn, H. Tetrahedron Lett. 1991, 32, 7453. (c) Mar-
tichonok, V.; Whitesides, G. M. J. Org. Chem. 1996, 61, 1702. For activation
of glycals, see: (d) Ito, Y.; Ogawa, T. Tetrahedron Lett. 1987, 28, 2723.
(e) Ramesh, S.; Kaila, N.; Grewal, G.; Franck, R. W. J. Org. Chem. 1990,
55, 5. (f) Grewal, G.; Kaila, N.; Franck, R. W. J. Org. Chem. 1992, 57,
2084. (g) Wang, J.; Wurster, J. A.; Wilson, L. J.; Liotta, D. Tetrahedron
Lett. 1993, 34, 4881.
To a solution of 2,6-dichlorophenyl 3,4-di-O-acetyl-2,6-dideoxy-1-
thio-R-L-galactopyranoside (512 mg, 1.30 mmol) in methanol (20 mL)
was added sodium methoxide (91 mg, 1.69 mmol). The reaction mixture

Sulfoxide Glycosylation Reaction
J. Am. Chem. Soc., Vol. 121, No. 26, 1999 6181
was stirred at room temperature for 30 min and then neutralized by
addition of Amberlite I-20 acid resin. The resin was removed by
filtration, and the filtrate was concentrated in vacuo to afford 2,6-
dichlorophenyl 2,6-dideoxy-1-thio-R-L-galactopyranoside as a white
min. After 15 min at -78 °C, the reaction was quenched with
diethylamine (100 µL), filtered into saturated aqueous NaHCO3 (30
mL), and extracted with methylene chloride (3 × 20 mL). The organic
layers were combined, dried over Na2SO4, decanted, and concentrated
in vacuo. The product was purified by flash chromatography (33%
EtOAc/petroleum ether) to afford disaccharide 16 (56 mg, 81%).
1
solid (386 mg, 96%): Rf ) 0.34 (10% EtOAc/dichloromethane); H
NMR (CD3OD, 500 MHz) δ 7.47 (d, J ) 8.0 Hz, 2H), 7.30 (t, J ) 8.0
Hz, 1H), 5.72 (d, J ) 6.0 Hz, 1H), 4.44 (q, J ) 6.5 Hz, 1H) 3.99 (ddd,
J ) 3.0, 5.0, 12.5 Hz, 1H), 3.62 (d, J ) 3.0 Hz, 1H), 2.31 (ddd, J )
5.8, 13.8, 13.8 Hz, 1H), 1.96 (dd J ) 5, 13.5 Hz, 1H), 1.12 (d, J ) 6.5
Hz, 1H); 13C NMR (CD3OD, 125.8 MHz) δ 142.6, 133.1, 131.8, 129.9,
86.0, 72.2, 70.2, 67.6, 33.6, 17.0; HRFABMS calcd for C12H14O3SCl2
(MNa+) 330.9923, found 330.9938.
2,6-Dichlorophenyl [2,6-Dideoxy-3-O-(4-methoxybenzyl)-r-L-ga-
lactopyranosyl]-(1f4)-2,6-dideoxy-3-O-(4-methoxybenzyl)-1-thio-
r-L-galactopyranoside (18). The disaccharide 16 (58 mg, 0.081 mmol)
was taken up in diethyl ether (4 mL) and cooled to 0 °C. Lithium
aluminum hydride (150 µL of 1 M LAH/ether, 0.15 mmol) was added.
The reaction was stirred for 30 min and then quenched by slow addition
of ethyl acetate (250 µL). After 15 min, the reaction was diluted with
ethyl acetate (20 mL) and washed with water (20 mL) and saturated
aqueous NaHCO3 (20 mL). The aqueous layers were reextracted with
ethyl acetate (20 mL), and then the organic layers were combined, dried
over Na2SO4, decanted, and concentrated in vacuo. The product was
purified by flash chromatography (50% EtOAc/petroleum ether) to
afford disaccharide 18 (50 mg, 92%): Rf ) 0.40 (50% EtOAc/petroleum
A combined solution of 2,6-dichlorophenyl 2,6-dideoxy-1-thio-R-
L-galactopyranoside (604 mg, 1.96 mmol) and dibutyltin oxide (536
mg, 2.16 mmol) in benzene (25 mL) was fitted with a Dean-Stark
apparatus and refluxed overnight. The reaction was cooled to room
temperature, and 4-methoxybenzyl chloride (1.06 mL, 7.82 mmol) and
tetrabutylammonium bromide (635 mg, 1.97 mmol) were added. The
reaction was refluxed for 3.5 h, cooled to room temperature, and
concentrated in vacuo. The crude reaction mixture was purified by silica
gel chromatography (25% ethyl acetate/petroleum ether) to afford 2,6-
dichlorophenyl 2,6-dideoxy-3-O-(4-methoxybenzyl)-1-thio-R-L-galac-
topyranoside (15) (833 mg, 99%): Rf ) 0.30 (33% EtOAc/petroleum
1
ether); H NMR (CDCl3, 270 MHz) δ 7.37 (d, J ) 7.9 Hz, 2H), 7.24
(m, 5H), 6.88 (m, 4H), 5.79 (d, J ) 5.6 Hz, 1H), 5.03 (d, J ) 2.6 Hz,
1H), 4.63 (d, J ) 11.9 Hz, 1H), 4.53 (d, J ) 11.9 Hz, 1H), 4.51 (d,
J ) 11.2 Hz, 1H), 4.46 (d, J ) 11.2 Hz, 1H), 4.32 (m, 2H), 3.87 (m,
3H), 3.82 (s, 3H), 3.79 (s, 3H), 3.71 (br s, 1H), 2.39 (ddd, J ) 5.6,
12.6, 13.2 Hz, 1H), 2.10 (m, 2H), 1.95 (m, 2H), 1.13 (d, J ) 6.6 Hz,
3H), 1.06 (d, J ) 6.6 Hz, 3H); 13C NMR (CD3COCD3, 125.8 MHz) δ
160.7, 160.66, 142.7, 133.4, 132.5, 132.4, 132.3, 132.2, 130.6, 130.3,
115.1, 115.0, 100.5, 86.5, 75.7, 74.6, 74.5, 71.2, 70.7, 70.2, 69.3, 69.2,
67.8, 56.1, 32.7, 31.4, 18.2, 18.0; HRFABMS calcd for C34H40O8NaSCl2
(MNa+) 701.1719, found 701.1744.
1
ether); H NMR (CDCl3, 270 MHz) δ 7.38 (d, J ) 8.2 Hz, 2H), 7.25
(m, 3H), 6.90 (d, J ) 8.2 Hz, 2H), 5.76 (d, J ) 5.6 Hz, 1H), 4.56 (s,
2H), 4.46 (q, J ) 6.6 Hz, 1H), 3.90 (m, 1H), 3.84 (m, 1H), 3.82 (s,
3H), 2.33 (ddd, J ) 5.6, 7.9, 11.9 Hz, 1H), 2.15 (m, 2H), 1.22 (d, J )
6.6 Hz, 3H); 13C NMR (CD3COCD3, 125.8 MHz) δ 160.7, 142.7, 133.4,
132.3, 132.2, 130.6, 130.3, 115.0, 86.5, 74.9, 70.4, 70.3, 69.1, 56.0,
31.8, 17.7; HRFABMS calcd for C20H22O4NaSCl2 (MNa+) 451.0514,
found 451.0503.
2,6-Dichlorophenyl {(2,3,6-Trideoxy-4-ulo-r-L-hexopyranosyl)-
(1f4)-[2,6-dideoxy-3-O-(4-methoxybenzyl)-r-L-galactopyranosyl]}-
(1f4)-2,6-dideoxy-3-O-(4-methoxybenzyl)-1-thio-r-L-galactopyran-
oside (19). Alcohol 18 (50 mg, 0.074 mmol), 4-allyl-1,2-dimeth-
oxybenzene (317 µL, 1.84 mmol), and 2,6-di-tert-butyl-4-methylpyri-
dine (168 mg, 0.82 mmol) were combined, and residual water was
removed by azeotropic distillation with toluene (3 × 10 mL). To the
residue in methylene chloride (5.0 mL) was added 4 Å molecular sieves
(500 mg), and the resulting suspension was stirred at room temperature
for 1 h. The suspension was cooled to -78 °C, and a solution of triflic
anhydride (28 µL, 0.168 mmol) in methylene chloride (350 µL) was
added over 1-2 min. A solution of sulfoxide 629 (40 mg, 0.168 mmol)
in methylene chloride (3.0 mL) was added via syringe over 10-15
min. After 15 min at -78 °C, the reaction was filtered into saturated
aqueous NaHCO3 (30 mL) and extracted with methylene chloride (3
× 20 mL). The organic layers were combined, dried over Na2SO4,
decanted, and concentrated in vacuo. The product was purified by flash
chromatography (33% EtOAc/petroleum ether) to afford a mixture of
trisaccharides (40 mg, 68%, R:â ) 5:1). R-Glycoside 19: Rf ) 0.30
2,6-Dichlorophenyl [4-O-Acetyl-2,6-dideoxy-3-O-(4-methoxyben-
zyl)-r-L-galactopyranosyl]-(1f4)-2,6-dideoxy-3-O-(4-methoxyben-
zyl)-1-thio-r-L-galactopyranoside (16). Alcohol 15 (42 mg, 0.098
mmol), 4-allyl-1,2-dimethoxybenzene (250 µL, 0.098 mmol), and 2,6-
di-tert-butyl-4-methylpyridine (87 mg, 0.42 mmol) were combined, and
residual water was removed by azeotropic distillation with toluene (3
× 10 mL). To the residue in methylene chloride (4.5 mL) was added
4 Å molecular sieves (500 mg), and the resulting suspension was stirred
at room temperature for 1 h. The suspension was cooled to -78 °C,
and a solution of triflic anhydride (25 µL, 0.148 mmol) in methylene
chloride (350 µL) was added over 1-2 min. A solution of sulfoxide
11 (62 mg, 0.148 mmol) in methylene chloride (2.5 mL) was added
via syringe over 10-15 min. After 15 min at -78 °C, the reaction
was quenched with diethylamine (100 µL), filtered into saturated
aqueous NaHCO3 (30 mL), and extracted with methylene chloride (3
× 20 mL). The organic layers were combined, dried over Na2SO4,
decanted, and concentrated in vacuo. The product was purified by flash
chromatography (33% EtOAc/petroleum ether) to afford disaccharide
16 (58 mg, 82%): Rf ) 0.35 (33% EtOAc/petroleum ether); 1H NMR
(CDCl3, 270 MHz) δ 7.39 (d, J ) 7.8 Hz, 2H), 7.21 (m, 5H), 6.89 (d,
J ) 7.8 Hz, 2H), 6.83 (d, J ) 8.9 Hz, 2H), 5.80 (d, J ) 5.3 Hz, 1H),
5.25 (d, J ) 2.0 Hz, 1H), 5.04 (m, 1H), 4.58 (m, 3H), 4.36 (m, 3H),
3.88 (m, 3H), 3.82 (s, 3H), 3.78 (s, 3H), 2.38 (ddd, J ) 5.3, 12.5, 12.8
Hz, 1H), 2.16 (m, 1H), 2.12 (s, 3H), 1.98 (m, 2H), 1.13 (d, J ) 6.6
Hz, 3H), 0.88 (d, J ) 6.6 Hz, 3H); 13C NMR (CD3COCD3, 125.8 MHz)
δ 171.5, 160.8, 160.7, 142.7, 133.4, 132.4, 132.2, 132.1, 130.8, 130.4,
130.3, 115.0, 114.9, 100.5, 86.5, 76.2, 74.6, 72.6, 71.0, 70.9, 70.8, 70.7,
66.5, 56.1, 56.0, 32.7, 32.6, 21.4, 18.1, 17.7; HRFABMS calcd for
C36H42O9NaSCl2 (MNa+) 743.1824, found 743.1792.
1
(7% EtOAc/methylene chloride); H NMR (CDCl3, 270 MHz) δ 7.38
(d, J ) 7.9 Hz, 2H), 7.22 (m, 7H), 6.86 (m, 4H), 5.78 (d, J ) 5.3 Hz,
1H), 5.06 (br s, 1H), 5.01 (t, J ) 4.3 Hz, 1H), 4.66 (q, J ) 6.6 Hz,
1H), 4.51 (m, 4H), 4.34 (q, J ) 6.6 Hz, 1H), 4.23 (q, J ) 6.6 Hz, 1H),
3.88 (m, 4H), 3.82 (s, 3H), 3.78 (s, 3H), 2.62 (m, 1H), 1.95-2.45 (m,
7H), 1.13 (d, J ) 6.6 Hz, 3H), 0.95 (d, J ) 6.6 Hz, 6H); 13C NMR
(CD3COCD3, 125.8 MHz) δ 211.1, 160.6, 160.5, 142.6, 133.3, 132.3,
132.2, 132.1, 130.3, 130.2, 130.1, 115.0, 114.8, 100.3, 99.0, 86.4, 76.7,
75.6, 74.5, 73.7, 72.4, 71.0, 70.7, 70.6, 68.6, 56.0, 34.9, 32.5, 32.0,
31.1, 18.1, 18.0, 15.8; HRFABMS calcd for C40H48O10NaSCl2 (MNa+)
813.2243, found 813.2221. â-Glycoside: Rf ) 0.28 (7% EtOAc/
1
Glycosylation Using the 2,6-Dichlorophenyl-Substituted B Ring
Sulfoxide. Alcohol 15 (41 mg, 0.096 mmol), 4-allyl-1,2-dimethoxy-
benzene (230 µL, 1.34 mmol), and 2,6-di-tert-butyl-4-methylpyridine
(104 mg, 0.51 mmol) were combined, and residual water was removed
by azeotropic distillation with toluene (3 × 10 mL). To the residue in
methylene chloride (4.5 mL) was added 4 Å molecular sieves (500
mg), and the resulting suspension was stirred at room temperature for
1 h. The suspension was cooled to -78 °C, and a solution of triflic
anhydride (23 µL, 0.134 mmol) in methylene chloride (350 µL) was
added over 1-2 min. A solution of sulfoxide 17 (65 mg, 0.134 mmol)
in methylene chloride (3.0 mL) was added via syringe over 10-15
methylene chloride); H NMR (CDCl3, 270 MHz) δ 7.38 (d, J ) 7.9
Hz, 2H), 7.23 (m, 5H), 6.87 (m, 4H), 5.79 (d, J ) 5.3 Hz, 1H), 5.30
(t, J ) 3.6 Hz, 1H), 5.06 (br s, 1H), 4.63 (d, J ) 11.6 Hz, 1H), 4.52
(d, J ) 11.6 Hz, 1H), 4.50 (s, 2H), 4.31 (m, 2H), 3.88 (m, 5H), 3.82
(s, 3H), 3.77 (s, 3H), 2.0-2.7 (m, 8H), 1.29 (d, J ) 6.6 Hz, 3H), 1.13
(d, J ) 6.3 Hz, 3H), 1.01 (d, J ) 6.6 Hz, 3H); 13C NMR (CD3COCD3,
125.8 MHz) δ 209.2, 160.7, 142.8, 133.4, 132.4, 132.2, 130.5, 130.4,
130.35, 130.3, 115.1, 100.7, 100.5, 86.5, 76.5, 75.7, 75.5, 75.1, 74.6,
(29) The known C ring sulfoxide 6 can be synthesized from rhamnose
in seven steps (58% overall yield); see ref 7.

6182 J. Am. Chem. Soc., Vol. 121, No. 26, 1999
GildersleeVe et al.
71.2, 71.0, 70.8, 68.0, 56.1, 35.8, 32.7, 31.9, 31.7, 18.5, 18.2, 16.4;
HRFABMS calcd for C40H48O10NaSCl2 (MNa+) 813.2243, found
813.2287.
phosphate-buffered saline (30 mL, pH 7) and extracted with methylene
chloride (3 × 20 mL). The organic layers were combined, dried over
Na2SO4, decanted, and concentrated in vacuo. The product was purified
by flash chromatography (40% EtOAc/petroleum ether + 1% acetic
acid) to afford protected ciclamcyin 0 (21) (9 mg, 75%): Rf ) 0.45
{(2,3,6-Trideoxy-4-ulo-r-L-hexopyranosyl)-(1f4)-[2,6-dideoxy-
3-O-(4-methoxybenzyl)-r-L-galactopyranosyl]}-(1f4)-2,6-dideoxy-
3-O-(4-methoxybenzyl)-1-(2,6-dichlorophenylsulfinyl)-r-L-galacto-
pyranoside (20). A solution of trisaccharide 19 (41 mg, 0.052 mmol)
and sodium bicarbonate (200 mg, 2.38 mmol) in methylene chloride
(3 mL) was cooled to -72 °C. Dimethyldioxirane30 (2 mL of ∼0.05
M dimethyldioxirane in acetone, 0.10 mmol) was added, and the
reaction was warmed slowly to -45 °C. The reaction was quenched
with dimethyl sulfide (3 drops) and then poured into saturated NaHCO3
(30 mL) and extracted three times with methylene chloride (3 × 30
mL). The combined organic layers were dried over Na2SO4, decanted,
and concentrated in vacuo. The crude mixture was purified by silica
column chromatography (50% ethyl acetate/petroleum ether) to afford
trisaccharide sulfoxide 20 (38 mg, 90%): Rf ) 0.45 (50% EtOAc/
1
(50% EtOAc/petroleum ether + 1% acetic acid); H NMR (CDCl3,
270 MHz) δ 12.97 (s, 1H), 12.83 (s, 1H), 12.25 (s, 1H), 7.73 (s, 1H),
7.30 (m, 4H), 7.13 (d, J ) 8.6 Hz, 2H), 6.88 (d, J ) 8.6 Hz, 2H), 6.76
(d, J ) 8.6 Hz, 2H), 5.26 (d, J ) 2.0 Hz, 1H), 5.08 (br s, 1H), 4.99 (t,
J ) 4.0 Hz, 1H), 4.35-4.75 (m, 6H), 4.32 (s, 1H), 4.20 (q, J ) 6.6
Hz, 1H), 4.12 (s, 1H), 4.02 (q, J ) 6.6 Hz, 1H), 3.5-3.98 (m, 3H),
3.82 (s, 3H), 3.74 (s, 3H), 3.70 (s, 3H), 1.92-2.70 (m, 9H), 1.40-
1.92 (m, 4H), 1.28 (d, J ) 6.2 Hz, 3H), 1.26 (s, 1H), 0.96 (d, J ) 6.6
Hz, 3H), 0.86 (d, J ) 6.6 Hz, 3H); 13C NMR (CD3COCD3, 125.8 MHz)
δ 211.2, 192.3, 187.3, 172.4, 163.5, 160.7, 160.5, 159.8, 159.1, 144.3,
134.2, 133.1, 132.4, 132.3, 131.5, 131.3, 130.5, 130.3, 129.5, 121.3,
116.2, 115.0, 114.9, 114.1, 114.0, 104.9, 103.5, 100.4, 99.1, 76.9, 75.6,
74.2, 72.6, 72.4, 72.2, 70.9, 70.8, 69.6, 68.6, 58.6, 56.2, 56.0, 53.4,
35.6, 35.0, 33.6, 32.4, 32.3, 25.0, 18.5, 18.3, 15.9, 7.7; HRMS calcd
for C56H64O19 (M+) 1040.4043, found 1040.4013.
Ciclamycin 0 (3). To a solution of PMB-protected ciclamycin 0
(20) (0.007 g, 0.00672 mmol) in 4.0 mL of methylene chloride were
added deionized water (3 drops) and 2,3-dichloro-5,6-dicyano-1,4-
benzoquinone (0.015 g, 0.0672 mmol). The reaction mixture was stirred
at room temperature for 30 min. The reaction mixture was washed with
water (5 mL), and the organic layer was dried over Na2SO4, filtered,
and concentrated. The crude product was purified by preparative thin-
layer chromatography (70% EtOAc/petroleum ether + 1% acetic acid)
to afford 0.003 g (60%) of ciclamycin 0 (3) as an orange solid.31
1
petroleum ether); H NMR (CDCl3, 270 MHz) δ 7.38 (m, 3H), 7.27
(m, 4H), 6.88 (d, J ) 8.6 Hz, 2H), 6.84 (d, J ) 8.6 Hz, 2H), 5.84 (d,
J ) 5.3 Hz, 1H), 5.05 (br s, 1H), 5.01 (t, J ) 4.3 Hz, 1H), 4.65 (q,
J ) 6.6 Hz, 1H), 4.59 (s, 2H), 4.55 (d, J ) 11.6 Hz, 1H), 4.48 (d, J )
11.6 Hz, 1H), 4.24 (q, J ) 6.3 Hz, 1H), 4.08 (q, J ) 6.6 Hz, 1H), 3.88
(m, 4H), 3.82 (s, 3H), 3.77 (s, 3H), 2.77 (m, 1H), 2.61 (m, 1H), 2.1-
2.5 (m, 4H), 2.03 (m, 2H), 1.13 (d, J ) 6.6 Hz, 3H), 0.96 (d, J ) 6.6
Hz, 3H), 0.94 (d, J ) 6.3 Hz, 3H); 13C NMR (CD3COCD3, 125.8 MHz)
δ 211.2, 160.7, 138.8, 137.6, 134.8, 132.3, 132.1, 130.5, 130.4, 115.1,
115.0, 100.5, 99.1, 92.5, 76.8, 75.3, 74.8, 73.8, 73.7, 72.6, 70.9, 70.8,
68.8, 56.1, 35.0, 32.2, 25.0, 18.5, 18.3, 15.9; HRFABMS calcd for
C40H48O11NaSCl2 (MNa+) 829.2192, found 829.2163.
7-O-{(2,3,6-Trideoxy-4-ulo-r-L-hexopyranosyl)-(1f4)-[2,6-dideoxy-
3-O-(4-methoxybenzyl)-r-L-galactopyranosyl]-(1f4)-2,6-dideoxy-3-
O-(4-methoxybenzyl)-r-L-galactopyranosyl}-E-pyrromycinone (21).
ꢀ-Pyrromycinone (9)12 (5 mg, 0.012 mmol), 4-allyl-1,2-dimethoxyben-
zene (60 µL, 0.35 mmol), and 2,6-di-tert-butyl-4-methylpyridine (32
mg, 0.16 mmol) were combined, and residual water was removed by
azeotropic distillation with toluene (3 × 10 mL). To the residue in
methylene chloride (2.5 mL) was added 4 Å molecular sieves (200
mg), and the resulting suspension was stirred at room temperature for
1 h. The suspension was cooled to -78 °C, and a solution of triflic
anhydride (6 µL, 0.036 mmol) in methylene chloride (200 µL) was
added over 1-2 min. A solution of trisaccharide sulfoxide 20 (28 mg,
0.035 mmol) in methylene chloride (1.5 mL) was added via syringe
over 10 min. After 20 min at -78 °C, the reaction was filtered into
Acknowledgment. We thank Professor Samuel Danishefsky
for providing us with an authentic sample of ciclamycin 0 and
Bristol-Myers Squibb for a gift of marcellomycin. In addition,
we thank the National Institutes of Health for financial support
of this work.
Supporting Information Available: Experimental proce-
dures for the synthesis of 4b, 11, 12, and 17, as well as 1H and
13C NMR spectra for 4b and 11-21 (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
JA990690A
(30) Adam, W.; Bialas, J.; Hadjiarapoglou, L. Chem. Ber. 1991, 124,
2377.
(31) The 1H, 13C, and mass spectra for synthetic ciclamycin 0 were
identical with the spectra reported in the literature (see refs 6 and 8).