Stereoselective N-glycosylation by Staudinger ligation

Article abstract of DOI:10.1021/ol048271s

(Chemical equation presented) Stereoselective methods for the chemical synthesis of β-N-glycosyl amides are needed to generate glycopeptides and glycoproteins. Here, we report that the Staudinger ligation can be used to form glycosylated asparagine deriva

Full text of DOI:10.1021/ol048271s

ORGANIC
LETTERS
2004
Vol. 6, No. 24
4479-4482
Stereoselective N-Glycosylation by
Staudinger Ligation
Yi He,† Ronald J. Hinklin,† Jiyoung Chang,† and Laura L. Kiessling*,†,‡
Departments of Chemistry and Biochemistry, UniVersity of Wisconsin-Madison,
Madison, Wisconsin 53706
kiessling@chem.wisc.edu
Received August 29, 2004
ABSTRACT
Stereoselective methods for the chemical synthesis of â-N-glycosyl amides are needed to generate glycopeptides and glycoproteins. Here, we
report that the Staudinger ligation can be used to form glycosylated asparagine derivatives. The reaction proceeds with high stereoselectivity,
and a variety of glycosyl azides can function as substrates. Our results provide precedence for the use of this powerful amide-bond-forming
reaction for N-glycopeptide synthesis.
N-Glycosylation is an important posttranslational modifica-
tion of proteins.1 Most N-glycoproteins share a common core
structure in which an oligosaccharide is attached via a
â-glycosyl amide bond to an Asn residue located within the
consensus sequence Asn-Xxx-Thr/Ser (Xxx is any amino
acid except Pro).2 N-Glycoproteins and glycopeptides ob-
tained from physiological sources are typically hetero-
geneous, thereby complicating an analysis of their biological
roles.3 Chemical synthesis can provide access to homo-
geneous samples of these products.4,5
bond linking the glycan and the protein. The most convergent
approach is to append the carbohydrate(s) of interest to a
synthetic peptide. This approach is advantageous because it
avoids carrying a protected glycosylated peptide sequence
through multiple subsequent synthetic steps. The glycosyl
amide is commonly synthesized by condensation of a
glycosylamine and an activated carboxylic acid derivative.6
Though valuable, this approach often results in the production
of anomeric mixtures due to the propensity of the glycosyl-
amine to isomerize. Other sugar-derived precursors have been
explored as alternatives to glycosylamines, including glycosyl
isothiocyanates,7 pentenyl glycosides,8 and glycosyl sulfox-
ides.9 Each of these has some utility, yet milder methods
for the stereoselective formation of glycosyl amides are still
needed.
One challenge in the chemical synthesis of N-glycopep-
tides is the stereoselective formation of the glycosyl amide
† Department of Chemistry.
‡ Department of Biochemistry.
(1) Recent reviews on glycoproteins: (a) Bertozzi, C. R.; Kiessling, L.
L. Science 2001, 291, 2357-2364. (b) Helenius, A.; Aebi, M. Science 2001,
291, 2364-2369. (c) Ritchie, G. E.; Moffatt, B. E.; Sim, R. B.; Morgan,
B. P.; Dwek, R. A.; Rudd, P. M. Chem. ReV. 2002, 102, 305-319. (d)
Imperiali, B.; Tai, V. W.-F. Chemistry and Biochemistry of Asparagine-
Linked Protein Glycosylation. In Carbohydrate-based Drug DiscoVery;
Wong, C. H., Ed. Wiley-VCH Verlag GmbH Publisher: Weinheim, 2003;
pp 281-302.
(2) (a) Marshall, R. D. Biochem. Soc. Symp. 1974, 40, 17-26. (b) Hart,
G. W.; Brew, K.; Grant, G. A.; Bradshaw, R. A.; Lennarz, W. J. J. Biol.
Chem. 1979, 254, 9747-9752.
(3) Schachter, H. Biochem. Cell Biol. 1986, 64, 163-181.
(4) Recent reviews on the chemical synthesis of glycoproteins: (a)
Arsequell, G.; Valencia, G. Tetrahedron: Asymmetry 1999, 10, 3045-3094.
(b) Mizuno, M. Trends Glycosci. Glycotechnol. 2001, 13, 11-30. (c) Hang,
H. C.; Bertozzi, C. R. Acc. Chem. Res. 2001, 34, 727-736. (d) Davis, B.
G. Chem. ReV. 2002, 102, 579-601.
(5) Recent examples of chemical synthesis of glycoproteins: (a) Warren,
J. D.; Miller, J. S.; Keding, S. J.; Danishefsky, S. J. J. Am. Chem. Soc.
2004, 126, 6576-6578. (b) Macmillan, D.; Bertozzi, C. R. Angew. Chem.,
Int. Ed. 2004, 43, 1355-1359.
(6) Examples of N-glycoamino acids/peptides synthesized with glyco-
sylamines: (a) Marks, G. S.; Neuberger, A. J. Chem. Soc. 1961, 4872-
4879. (b) Dudkin, V. Y.; Miller, J. S.; Danishefsky, S. J. J. Am. Chem.
Soc. 2004, 126, 736-738.
(7) (a) Gunther, W.; Kunz, H. Angew. Chem., Int. Ed. Engl. 1990, 29,
1050-1051. (b) Ho, H. L.; Baptista, J. A. B.; Krepinsky, J. J. Can. J. Chem.
1990, 68, 953-957.
(8) (a) Ratcliffe, A. J.; Konradsson, P.; Fraser-Reid, B. J. Am. Chem.
Soc. 1990, 112, 5665-5667. (b) Ratcliffe, A. J.; Konradsson, P.; Reid, B.
F. Carbohydr. Res. 1991, 216, 323-335. (c) Handlon, A. L.; Fraser-Reid,
B. J. Am. Chem. Soc. 1993, 115, 3796-3797.
10.1021/ol048271s CCC: $27.50
© 2004 American Chemical Society
Published on Web 10/30/2004

Two- and three-component Staudinger ligation reactions
of different types have been used to modify carbohydrates.
For example, three-component Staudinger ligation reactions
that employ an azide, a carboxylic acid (or derivative), and
a phosphine have been used to generate glycosyl amides,10-13
but anomerization is often problematic.14 Two-component
Staudinger ligation reactions have been developed in which
carbohydrate azide derivatives selectively engage in amide
bond formation in complex environments.15 Specifically, cells
treated with a carbohydrate-containing azide can undergo
reaction with a phosphinoester, thereby modifying the cell
surface.15d In this version of the Staudinger ligation, an
activated ester and phosphine reside in the same molecule.
Simple glycosyl amide bonds have been formed using
substrates of this type;16 however, the products were obtained
as isomeric mixtures. The Staudinger ligation had not been
used to generate more complex N-glycopeptide precursors.
A version of the two-component Staudinger ligation
reaction in which a phosphinothioester reacts with an azide-
containing amino acid has been shown to afford peptide
bonds.17 Indeed, reactions of this type occur under mild
conditions. Novel proteins of interest and small molecules
have been immobilized on surfaces.17e,18 We sought to
examine the scope of the Staudinger ligation by evaluating
the coupling of glycosyl azides and asparagine-derived
phosphinothioesters.
Scheme 1. Proposed Mechanism for the Formation of the
Glycosyl Amide Bond via Staudinger Ligation
properties of the resulting iminophosphorane intermediate
and therefore the reaction outcome. Thus, we needed a
convenient synthetic route to phosphinothioesters with alkyl
as well as aryl phosphine substituents.
Diphenylphosphinothioesters were used in previous stud-
ies. We investigated whether the most effective route to
diphenylphosphinothioesters17c could afford the more nu-
cleophilic dialkylphosphine derivatives. Although the dialkyl
species could be generated, the phosphinothiol precursors
were very unstable, presumably because the dialkyl deriva-
tives are more prone to oxidation. Thus, we developed a
general route to phosphinothioesters.
Phosphine-boranes19 and trialkylphosphonium salts20 can
protect oxygen-sensitive phosphine compounds. Air-stable
diphenylphosphine-borane has been used to efficiently
generate diphenylphosphinothioesters.17c We reasoned that
this protecting group might prove to be valuable in the
synthesis of a wide range of phosphinothioesters. To this
end, we generated phosphine-borane complexes and added
them to formaldehyde to yield alcohols 2a-c (Scheme 2).
These alcohols were converted to thioacetates 3a-c in two
steps. The acetyl groups were removed, and the resulting
thiols were coupled to Boc-Asp-OBz to afford the borane-
protected phosphinothioesters 4a-c in six steps. The borane
group was efficiently removed by heating 4a-c with 1,4-
diazabicyclo[2.2.2]octane (DABCO) in toluene under argon.
The direct addition of the substituted phosphine-borane to
formaldehyde, which occurs under mild conditions, is a key
step in our scheme. We anticipate that this reaction will be
useful for the synthesis of a wide range of functionalized
phosphines. Overall, the revised synthetic route affords the
dialkyl phosphinothioesters in higher yields than previous
schemes. With phosphinothioesters 5a-c in hand, we tested
their reactivity in the coupling reaction.
Our initial goal was to generate an appropriate phosphino-
thioester that would serve as a precursor to glycosylated
asparagine derivatives. We anticipated that the putative
iminophosphorane intermediate (Scheme 1), which must
undergo intramolecular transacylation to form the glycosyl
amide, would be less nucleophilic than those generated in
the peptide coupling reactions explored previously. The
phosphine substituents will influence the steric and electronic
(9) Kahne, D.; Walker, S.; Cheng, Y.; Vanengen, D. J. Am. Chem. Soc.
1989, 111, 6881-6882.
(10) (a) Maunier, V.; Boullanger, P.; Lafont, D. J. Carbohydr. Chem.
1997, 16, 231-235. (b) Mizuno, M.; Muramoto, I.; Kobayashi, K.;
Yaginuma, H.; Inazu, T. Synthesis 1999, 162-165. (c) Boullanger, P.;
Maunier, V.; Lafont, D. Carbohydr. Res. 2000, 324, 97-106. (d) Mizuno,
M.; Muramoto, I.; Kobayashi, K.; Yaginuma, H.; Inazu, T. Phosphorus
Sulfur Silicon Relat. Elem. 2002, 177, 1945-1945.
(11) Damkaci, F.; DeShong, P. J. J. Am. Chem. Soc. 2003, 125, 4408-
4409.
(12) Shangguan, N.; Katukojvala, S.; Greenberg, R.; Williams, L. J. J.
Am. Chem. Soc. 2003, 125, 7754-7755.
(13) Malkinson, J. P.; Falconer, R. A.; Toth, I. J. Org. Chem. 2000, 65,
5249-5252.
(14) Kova´cs, L.; OÄ sz, E.; Domokos, V.; Holzer, W.; Gyo¨rgydea´k, Z.
Tetrahedron 2001, 57, 4609-4621.
(15) (a) Saxon, E.; Armstrong, J. I.; Bertozzi, C. R. Org. Lett. 2000, 2,
2141-2143. (b) Saxon, E.; Bertozzi, C. R. Science 2000, 287, 2007-2010.
(c) Saxon, E.; Luchansky, S. J.; Hang, H. C.; Yu, C.; Lee, S. C.; Bertozzi,
C. R. J. Am. Chem. Soc. 2002, 124, 14893-14902. (d) Vocadlo, D. J.;
Hang, H. C.; Kim, E. J.; Hanover, J. A.; Bertozzi, C. R. Proc. Natl. Acad.
Sci. U.S.A. 2003, 100, 9116-9121.
(16) Bianchi, A.; Bernardi, A. Tetrahedron Lett. 2004, 45, 2231-2234.
(17) (a) Nilsson, B. L.; Kiessling, L. L.; Raines, R. T. Org. Lett. 2000,
2, 1939-1941. (b) Nilsson, B. L.; Kiessling, L. L.; Raines, R. T. Org. Lett.
2001, 3, 9-12. (c) Soellner, M. B.; Nilsson, B. L.; Raines, R. T. J. Org.
Chem. 2002, 67, 4993-499. (d) Nilsson, B. L.; Hondal, R. J.; Soellner, M.
B.; Raines, R. T. J. Am. Chem. Soc. 2003, 125, 5268-5269. (e) Soellner,
M. B.; Dickson, K. A.; Nilsson, B. L.; Raines, R. T. J. Am. Chem. Soc.
2003, 125, 11790-11791.
(19) Recent reviews on phosphine-borane complexes: (a) Brunel, J.
M.; Faure, B.; Maffei, M. Coord. Chem. ReV. 1998, 180, 665-698. (b)
Ohff, M.; Holz, J.; Quirmbach, M.; Borner, A. Synthesis 1998, 1391-1415.
(c) Carboni, B.; Monnier, L. Tetrahedron 1999, 55, 1197-1248.
(20) Netherton, M. R.; Fu, G. C. Org. Lett. 2001, 3, 4295-4298.
(18) Ko¨hn, M.; Wacker, R.; Peters, C.; Schro¨der, H.; Soule`re. L.;
Breinbauer, R.; Niemeyer, C. M.; Waldmann, H. Angew. Chem., Int. Ed.
2003, 42, 5830-5834.
4480
Org. Lett., Vol. 6, No. 24, 2004

finding is advantageous, as this stereochemistry is that of
natural N-glycoprotein linkages. Our results are consistent
with those from related reactions in which conditions for
reductive acylation that largely preserve glycosyl azide
stereochemistry have been identified.16 Still, the anomeriza-
tion of glycosyl iminophosphoranes is known, and this
anomerization can give rise to isomeric products.21 Thus,
we sought to explore further the basis for the observed
stereochemical outcome.
Scheme 2. Synthetic Route to Phosphinothioesters 5a-ca
There are two explanations for the selective production
of â-glycosyl amide from compounds 6-9: (1) anomeriza-
tion of the â-iminophosphorane intermediate does not
compete with transacylation or (2) anomerization occurs but
transacylation of the â-iminophosphorane is favored (Scheme
3). To distinguish between these, we generated R-glycosyl
a Yields of each step are presented for each compound.
We initially explored the reaction of glycosyl azides 6-9
using phosphinothioesters 5a-c (see: Scheme 1 and Table
1, entries 1-4). The coupling proceeded smoothly to afford
the desired glycosylated amino acid in moderate yield.
Solvent strongly influences the reaction yields. Transforma-
tions carried out in dimethylformamide (DMF) typically
afforded the products in higher yields than did those
conducted in tetrahydrofuran (THF). This solvent effect is
pronounced for reactions of the electron-deficient glycosyl
azides 8-10; no product was observed when reactions were
carried out in THF. Overall, the more nucleophilic phosphi-
nothioesters 5a and 5b afforded the glycosylated asparagine
derivatives in yields similar to those obtained with 5c. Still,
subtle differences in phosphinothioester reactivity are mani-
fested in the outcome of the coupling reaction (Table 1).
Scheme 3. Glycosyl Iminophosphorane Anomerization
azide 10 and exposed it to each of the phosphinothioesters
5a-c (entry 5, Table 1). If 10 affords the R-glycosyl amide,
we would conclude that transacylation occurs more rapidly
than glycosyl iminophosphorane anomerization. Alterna-
tively, if both isomeric glycosyl amides are generated, we
would conclude that isomerization of the intermediate
R-iminophosphorane occurs and that the isomeric intermedi-
ates both undergo transacylation. Finally, if the â-glycosyl
amide is the sole product, from 10, we would conclude that
the anomerization of the R-iminophosphorane occurs and that
only the â-iminophosphorane undergoes transacylation. We
were surprised to find that the glycosyl amide was generated
only when phosphinothioester 5a was employed. Compound
10 reacted with 5a to afford the â-glycosyl amide exclusively
(Table 1). Thus, regardless of glycosyl azide configuration,
all of the phosphinothioester coupling reactions examined
proceed stereoselectively to deliver the desired â-glycosyl
amide products.
Table 1. Staudinger Ligation of Glycosyl Azides 6-10 and
Phosphinothioesters 5a-ca
The coupling of azide 10 provides insight into the ligation
reaction. The process exhibits a marked dependence on
phosphinothioester: only the diethyl derivative 5a reacts.
This phosphine is more nucleophilic than its diphenyl
counterpart 5c. The high electron density of the glycosyl
iminophosphorane generated from 5a may facilitate its
a Yields of the â-glycosyl asparagine derivatives are for the isolated
product. The R-isomer could not be detected by 1H NMR spectroscopy.
Conditions: DMF, rt. b THF, rt.
(21) (a) Paulsen, H.; Gyo¨rgydea´k, Z.; Friedmann, M. Chem. Ber. 1974,
107, 1590-1613. (b) Paulsen, H.; Gyo¨rgydea´k, Z.; Friedmann, M. Chem.
Ber. 1974, 107, 1568-1578.
Strikingly, the â-glycosyl amide was the only product
observed when â-glycosyl azides 6-9 were employed. This
Org. Lett., Vol. 6, No. 24, 2004
4481

anomerization (Scheme 3). The difference in reactivity
between 5a and 5b may be due to the steric hindrance of
the latter. Thus, the steric and electronic properties of the
phosphinothioester can dramatically influence the reaction
outcome.
Our findings extend the utility of the Staudinger ligation
to generate a new class of amide bonds. The reaction of
phosphinothioesters with glycosyl azides provides a general
method for the stereoselective formation of â-glycosyl
amides. In the course of these studies, we also developed a
new route for the synthesis of functionalized phosphines. The
stereoselectivity of the reaction provides insight into the
requirements for ligation. Further mechanistic studies and
their application to N-glycopeptide synthesis are underway.
Acknowledgment. This research was supported by the
NIH (GM49975). We thank Prof. R. T. Raines, Dr. B. L.
Nilsson, M. B. Soellner, and Prof. H. J. Reich for helpful
conversations.
Supporting Information Available: Complete experi-
mental details. This material is available free of charge via
the Internet at http://pubs.acs.org.
OL048271S
4482
Org. Lett., Vol. 6, No. 24, 2004