Stereoselective glycosylation reactions with chiral auxiliaries

Article abstract of DOI:10.1002/anie.200461745

Good neighbors: A chiral auxiliary is used for the first time to control the anomeric selectivity of glycosylation reactions. In this approach, upon treatment with an activator A⁺, neighboring-group participation pation of an (S)-ethoxycarbonyl

Full text of DOI:10.1002/anie.200461745

Angewandte
Chemie
Anomeric Selectivity
Stereoselective Glycosylation Reactions with
Chiral Auxiliaries**
Jin-Hwan Kim, Hai Yang, and Geert-Jan Boons*
In memory of Professor Jacques van Boom
It is now well recognized that protein- and lipid-bound
saccharides play essential roles in many molecular processes
that have an impact on eukaryotic biology and disease.[1–3]
Examples of such processes include fertilization, embryo-
genesis, neuronal development, hormone activities, and the
proliferation of cells and their organization into specific
tissues. The interactions of saccharides with proteins or lipids
are also important in health science and are involved in the
invasion and attachment of pathogens, inflammation, metas-
tasis, blood-group immunology, and xenotransplantation. A
major obstacle to advances in glycobiology is the lack of pure
and structurally well-defined carbohydrates and glycoconju-
gates. These compounds are often found in low concentra-
tions and in microheterogeneous forms, thus greatly compli-
cating their isolation and characterization. In many cases,
well-defined oligosaccharides can only be obtained by organic
synthesis.[4]
Scheme 1. Conventional and new approaches for stereoselective
glycosylation. A=activating group, Nu=nucleophile, X=leaving
Although much progress has been made in methods for
oligosaccharide synthesis, the construction of complex carbo-
hydrates remains time consuming.[5] One-pot multistep gly-
cosylations[6,7] and polymer-supported syntheses[8–10] are two
approaches being applied to streamline the preparation of
complex oligosaccharides. The usefulness of these methods is,
however, compromised by the fact that many glycosylations
give mixtures of the two possible anomers. If these anomers
are not separated after each glycosylation, complex mixtures
of products are obtained that cannot be used for biological
studies. Routine oligosaccharide synthesis will only be
possible when robust stereoselective glycosylations become
available.
The most reliable method for stereoselective glycosidic-
bond formation is based on the neighboring-group participa-
tion of a 2-O-acyl functionality (Scheme 1a).[11] In these
reactions, a promoter activates an anomeric leaving group,
thereby resulting in its departure and the formation of an
oxonium ion. Subsequent neighboring-group participation of
a 2-O-acyl protecting group will give a more stable acetoxo-
nium ion. An alcohol can attack the anomeric center of an
group.
acetoxonium ion from only one face, to provide a 1,2-trans
glycoside. Thus, b-linked products will be obtained in the case
of glucosyl-type donors, whereas mannosides will give a-
glycosides. The introduction of 1,2-cis glycosidic linkages,
such as a-glucosides and a-galactosides, requires glycosyl
donors with a non-assisting functionality at the C2 position.
Invariably, these glycosylations lead to mixtures of ano-
mers.[12] In general, reasonable anomeric selectivities will only
be obtained by extensive optimization of reaction conditions
such as solvent, temperature, promoter, leaving group, and
protecting-group pattern. Thus, the stereoselective formation
of 1,2-cis glycosides is the principal challenge in complex
oligosaccharide synthesis.
We describe herein a novel strategy for stereoselective
glycosylations in which a chiral auxiliary at the 2-position of a
glycosyl donor is used (Scheme 1b,c). The auxiliary is a
substituted ethyl moiety that contains a nucleophilic group.
Upon formation of an oxonium ion, participation of the
nucleophilic moiety of the auxiliary should lead to the
formation of either a trans- or a cis-decalin system. It was
expected that the use of an auxiliary with S stereochemistry
would lead only to the formation of the trans decalin, since
the alternate cis-fused system would place the phenyl
substituent in an axial position and induce unfavorable
steric interactions (Scheme 1b). Subsequent displacement of
the anomeric moiety of the trans-decalin intermediate would
then lead to the formation of a 1,2-cis glycoside. Alternatively,
the use of an auxiliary with R stereochemisty would lead to
the formation of a 1,2-trans glycoside, because in this case the
[*] Dr. J.-H. Kim, H. Yang, Prof. Dr. G.-J. Boons
Complex Carbohydrate Research Center
University of Georgia
315 Riverbend Road, Athens, GA 30602 (USA)
Fax: (+1)706-542-4412
E-mail: gjboons@ccrc.uga.edu
[**] This work was supported by the National Cancer Institute of the
National Institutes of Health (Grant: RO1 CA88986).
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2005, 44, 947 –949
DOI: 10.1002/anie.200461745
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
947

Communications
trans-decalin system would experience unfavorable
steric interactions, and therefore glycosylation
would only take place from the cis-decalin inter-
mediate (Scheme 1c).
Ethyl mandelate was explored as a first-gener-
ation chiral auxiliary (Scheme 2). This derivative was
deemed attractive because both enantiomers are
Scheme 2. Preparation of glycosyl donors 5R and 5S. Bn=benzyl,
DBU=1,8-diazabicyclo[5.4.0]undec-7-ene, TMSOTf=trimethylsilyl
trifluoromethanesulfonate.
readily available, esters are well established as
appropriate participating functionalities in glycosyl-
ations, and the benzylic nature of the auxiliary
should make its removal possible under reductive
conditions.
Glycosyl donors 5R and 5S, containing an (R)-
or (S)-ethoxycarbonylbenzyl moiety, could be pre-
pared by an efficient procedure from the readily
available epoxide 1.[13] Thus, the reaction of 1 with
ethyl (R)-mandelate in the presence of BF3·Et2O led
to a trans-diaxial opening of the epoxide to give 2R
in a yield of 48%. Next, acetolysis of the 1,6-anhydro
bridge of 2R with acetic anhydride and TMSOTf
gave compound 3R in an almost quantitative yield.
The anomeric acetyl group of 3R was selectively
removed with hydrazine acetate to give hemiacetal
4R, which was converted into trichloroacetimidate
5R by using standard conditions.[14] Glycosyl donor
5S was prepared by a similar protocol with ethyl (S)-
mandelate as the starting material.
Scheme 3. Stereoselective glycosylation with glycosyl donors 5R and 5S.
Bz=benzoyl.
With glycosyl donors 5R and 5S in hand,
attention was focused on the glycosylation of a
range of different glycosyl acceptors. Thus, coupling of 5S
with 6 by using a catalytic amount of TMSOTf in dichloro-
methane at À788C gave disaccharide 7S, mainly as the a-
glycoside, in an almost quantitative yield (Scheme 3). At this
low temperature, the reaction was completed within
15 minutes, which indicates that the glycosyl donor 5S is
highly reactive. Dilution of the reaction mixture led to a small
increase in anomeric selectivity, whereas higher reaction
temperatures led to decreases in selectivity. As expected,
coupling of 1R with 6 under similar reaction conditions gave
7R mainly as the b anomer.
support for the proposed mode of participation as outlined in
Scheme 1.
To demonstrate the generality of the approach, a range of
glycosyl acceptors were glycosylated with 5R and 5S. As can
be seen in Scheme 3, glycosylations of 5S with glycosyl
acceptors that have either a primary or a secondary hydroxy
group gave disaccharides with high a-anomeric selectivity. In
each case, the use of glycosyl donor 5R led to a reversal of
anomeric selectivity and mainly b-linked disaccharides were
isolated, albeit with somewhat lower selectivity than that
observed with 5S.
The fact that an inversion of configuration at the
asymmetric center of the auxiliary led to a reversal of the
stereochemical outcome of the glycosylation provides strong
The results of the glycosylations indicate that the use of
ethyl mandelate as an auxiliary at the 2-position provides a
reliable approach for the synthesis of a-linked glycosides. To
948
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 947 –949

Angewandte
Chemie
be useful for target synthesis, it is important that the auxiliary
can be removed under mild conditions. The auxiliary was
designed as a substituted benzyl ether, and thus should be
removable by a catalytic hydrogenation over Pd or by Birch
reduction. Indeed, saponification of the benzoyl and acetyl
esters of 9S by treatment with NaOMe in methanol followed
by removal of the benzyl ethers and the auxiliary by using
sodium in liquid ammonia led to the clean formation of the
deprotected compound 16 (Scheme 4).
[1] P. M. Rudd, T. Elliott, P. Cresswell, I. A. Wilson, R. A. Dwek,
Science 2001, 291, 2370.
[2] A. Varki, Glycobiology 1993, 3, 97.
[3] R. A. Dwek, Chem. Rev. 1996, 96, 683.
[4] C. R. Bertozzi, L. L. Kiessling, Science 2001, 291, 2357.
[5] P. Sears, C. H. Wong, Science 2001, 291, 2344.
[6] K. M. Koeller, C. H. Wong, Chem. Rev. 2000, 100, 4465.
[7] N. L. Douglas, S. V. Ley, U. Lucking, S. L. Warriner, J. Chem.
Soc. Perkin Trans. 1 1998, 51.
[8] O. J. Plante, E. R. Palmacci, P. H. Seeberger, Science 2001, 291,
1523.
[9] P. H. Seeberger, W. C. Haase, Chem. Rev. 2000, 100, 4349.
[10] D. Kahne, Curr. Opin. Chem. Biol. 1997, 1, 130.
[11] G.-J. Boons, Contemp. Org. Synth. 1996, 3, 173.
[12] A. V. Demchenko, Synlett 2003, 1225.
[13] J. Xue, Z. W. Guo, Tetrahedron Lett. 2001, 42, 6487.
[14] R. R. Schmidt, Angew. Chem. 1986, 98, 213; Angew. Chem. Int.
Ed. Engl. 1986, 25, 212.
[15] N. R. Williams, Adv. Carbohydr. Chem. Biochem. 1970, 25, 109.
Scheme 4. Deprotection of disaccharide 9S.
Although the glycosylations with 5S led to disaccharides
with high a-anomeric selectivity, the ultimate goal is the
development of an auxiliary that gives only one of the two
possible anomers in each glycosylation. The small amount of
unwanted anomer that is formed probably results from
glycosylation of the oxonium ion. In this respect, participation
by an ethoxycarbonylbenzyl moiety is probably slower than
that of a conventional acyl functionality because the forma-
tion of the six-membered ring is slower than that of the five-
membered ring. Thus, it is expected that a second-generation
auxiliary could be obtained by improving the nucleophilicity
of the auxiliary or reducing the flexibility of the rotatable
bonds.
In summary, it has been shown for the first time that the
anomeric selectivity of a glycosylation can be controlled by a
chiral auxiliary. This new method for anomeric control is
particularly suited for the introduction of a-glucosides.
However, it is expected that other glycosides can easily be
installed by a similar approach. For example, the use of the
readily available 1,6:2,3-dianhydro-b-d-talopyranoside[15] as
the starting material, instead of epoxide 1, would lead to a
galactosyl donor. Furthermore, preliminary studies have
shown that the methodology is also compatible with other
protecting-group patterns. It is to be expected that a system-
atic optimization of the structure of the auxiliary will lead to a
stereospecific glycosylation protocol. Only such a method will
make it possible for the potential of polymer-supported[8–10]
and one-pot multistep oligosaccharide[6,7] syntheses to be
realized, so that these methods can be used in a routine
manner for a large number of oligosaccharide targets.
Received: August 20, 2004
Published online: January 3, 2005
Keywords: anomeric control · carbohydrates · chiral auxiliaries ·
.
glycosides · glycosylation
Angew. Chem. Int. Ed. 2005, 44, 947 –949
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
949