Synthesis of oligosaccharides on soluble high-molecular-weight branched polymers in combination with purification by nanofiltration

Article abstract of DOI:10.1021/ol0352355

[Equation presented] An efficient approach for polymer-supported oligosaccharide synthesis is described whereby branched and high-molecular- weight PEG derivatives are used in combination with purification by nanofiltration. This methodology was applied t

Full text of DOI:10.1021/ol0352355

ORGANIC
LETTERS
2003
Vol. 5, No. 20
3591-3594
Synthesis of Oligosaccharides on
Soluble High-Molecular-Weight Branched
Polymers in Combination with
Purification by Nanofiltration
Debatosh Majumdar, Tong Zhu, and Geert-Jan Boons*
Complex Carbohydrate Research Center, The UniVersity of Georgia,
220 RiVerbend Road, Athens, Georgia, 30602-4712
gjboons@ccrc.uga.edu
Received July 3, 2003
ABSTRACT
An efficient approach for polymer-supported oligosaccharide synthesis is described whereby branched and high-molecular-weight PEG derivatives
are used in combination with purification by nanofiltration. This methodology was applied to the preparation of a tetraglucoside and the
tumor-associated antigen Lex.
Solid-supported synthesis provides a powerful means of
preparing oligopeptides, oligonucleotides, and many small
molecules.1 These procedures eliminate time-consuming
workup steps and allow the use of excess reagents to drive
reactions to completion, as well as providing an opportunity
for automation. Consequently, complex compounds can be
synthesized within a matter of days rather than weeks or
months.
The development of polymer-supported oligosaccharide
synthesis has been slow, and, although the first attempts were
reported in the 1970s, their success was limited and only
simple di- and trisaccharides could be obtained.2,3 These
disappointing results were primarily a consequence of the
absence of efficient methods for glycosidic bond formation.
During the last 15 years, many powerful glycosylation
approaches have been developed,4-6 and this progress has
allowed the preparation of more complex oligosaccharides
by solid-supported procedures. These developments have
resulted in the first example of an automated synthesis of a
heptaglucoside.7,8
Despite these impressive advancements, polymer-sup-
ported oligosaccharide synthesis is still afflicted by many
shortcomings. For example, glycosylations on a solid support
often require repetitive glycosylations with large excesses
of glycosyl donors or acceptors to drive reactions to
completion. This represents a serious drawback since oli-
gosaccharide chemistry requires elaborate and expensive
glycosyl donors and acceptors. In addition, the rate of a
reaction on a solid support is often considerably slower
compared to similar solution-based reactions, making it
difficult to extrapolate solution-phase conditions to solid-
supported procedures.
The problems of slow reactivity and the required use of a
large excess of reagents have been addressed by replacing
(1) Brown, R. C. D. J. Chem. Soc., Perkin Trans. 1 1998, 3293-3320.
(2) Seeberger, P. H.; Haase, W. C. Chem. ReV. 2000, 100, 4349-4394.
(3) Osborn, H. M. I.; Khan, T. H. Tetrahedron 1999, 55, 1807-1850.
(4) Toshima, K.; K. Tatsuta Chem. ReV. 1993, 93, 1503-1531.
(5) Boons, G. J. Tetrahedron 1996, 52, 1095-1121.
(7) Sears, P.; Wong, C. H. Science 2001, 291, 2344-2350.
(8) Plante, O. J.; Palmacci, E. R.; Seeberger, P. H. Science 2001, 291,
1523.
(6) Boons, G.-J. Contemp. Org. Synth. 1996, 3, 173-200.
10.1021/ol0352355 CCC: $25.00
© 2003 American Chemical Society
Published on Web 08/29/2003

insoluble cross-linked resins with soluble polymeric sup-
ports.9 In this way, the reaction conditions typical of classical
organic reactions are employed, while product purification
can be facilitated by taking advantage of the macromolecular
properties of the polymer. Poly(ethylene glycol) methyl ether
(MPEG) is the most widely used polymer for liquid-phase
oligosaccharide syntheses. This polymer is soluble under
glycosylation and protecting group manipulation conditions
but can be made insoluble during a workup procedure by
the simple addition of diethyl ether or tert-butylmethyl
ether.10-25 Several problems are, however, associated with
MPEG-supported approaches; for example, this methodology
is hampered by low loading of saccharide onto MPEG and
difficulties associated with selective precipitation when large
saccharides are attached to the polymer, and it is not
amenable to automation. These problems need to be urgently
addressed if soluble-polymer-based methods are to be
competitive.
Figure 1. Structures of 4- and 8-armed PEG derivatives.
Da, which is derivatized with an oxy benzoic acid function
for attachment and cleavage of saccharides (Scheme 1).15,16
Here we report a new and more efficient approach for
soluble polymer-supported oligosaccharide synthesis whereby
branched and high-molecular-weight PEG derivatives are
used in combination with purification by nanofiltration.
Innovation in membrane technology has given rise to
advanced filtration equipment containing specific pore sizes.
Membrane filters are now available in a range of pore sizes,
suitable for a specific application. The development of
membranes compatible with organic solvents has been slower
than those used for aqueous systems. However, there are now
several commercially available membranes that are compat-
ible with a range of organic solvents, and the most promising
one is a Biomax membrane with a nominal molecular weight
limit of 5000 Da. Thus, it was anticipated that oligosaccha-
rides attached to a polymer of sufficiently high molecular
weight can be retained by nanofiltration, and it was expected
that branched poly(ethylene glycol) derivatives would be the
most promising for this purpose (Figure 1).
Scheme 1
To develop the new methodology, we set out to synthesize
tetrasaccharide 10 using 4-armed PEG 4a of MW ) 10 000
(9) Gravert, D. J.; Janda, K. D. Chem. ReV. 1997, 97, 489-509.
(10) Douglas, S. P.; Whitfield, D. M.; Krepinsky, J. J. J. Am. Chem.
Soc. 1991, 113, 5095-5097.
(11) Douglas, S. P.; Whitfield, D. M.; Krepinsky, J. J. J. Am. Chem.
Soc. 1995, 117, 2116-2117.
(12) Verduyn, R.; Klein, P. A. M. v. d.; Douwes, M.; Marel, G. A. v.
d.; Boom, J. H. v. Recl. TraV. Chim. Pays-Bas 1993, 112, 464-466.
(13) Ito, Y.; Kanie, O.; Ogawa, T. Angew. Chem., Int. Ed. Engl. 1996,
35, 2510-2512.
(14) Ito, Y.; Ogawa, T. J. Am. Chem. Soc. 1997, 119, 5562-5566.
(15) Zhu, T.; Boons, G. J. J. Am. Chem. Soc. 2000, 122, 10222-10223.
(16) Zhu, T.; Boons, G. J. Chem.sEur. J. 2001, 7, 2382-2389.
(17) Geurtsen, R.; Boons, G. J. Eur. J. Org. Chem. 2002, 1473-1477.
(18) Mehta, S.; Whitfield, D. Tetrahedron Lett. 1998, 39, 5907-5910.
(19) Schmidt, D.; Thiem, J. Chem. Commun. 2000, 1919-1920.
(20) Belogi, G.; Zhu, T.; Boons, G. J. Tetrahedron Lett. 2000, 41, 6969-
6972.
(21) Yan, F. Y.; Gilbert, M.; Wakarchuk, W. W.; Brisson, J. R.;
Whitfield, D. M. Org. Lett. 2001, 3, 3265-3268.
Thus, glucoside 3, which has a selectively removable
levulinoyl ester (Lev) at C-6 and an anomeric p-hydroxy-
benzyl moiety, was coupled with 4a using DCC and DMAP
in dichloromethane to give immobilized 6a (Scheme 1).
Unfortunately, polymer 6a could not be retained by nano-
filtration using a Biomax filter with a nominal molecular
weight limit of 5000 Da. However, polymer 6a could be
(22) Ito, Y.; Manabe, S. Chem.sEur. J. 2002, 8, 3077-3084.
(23) Yan, F. Y.; Mehta, S.; Eichler, E.; Wakarchuk, W. W.; Gilbert,
M.; Schur, M. J.; Whitfield, D. M. J. Org. Chem. 2003, 68, 2426-2431.
(24) Porcheddu, A.; Ruda, G. F.; Sega, A.; Taddei, M. Eur. J. Org. Chem.
2003, 907-912.
(25) Brinkmann, N.; Malissard, M.; Ramuz, M.; Romer, U.; Schumacher,
T.; Berger, E. G.; Elling, L.; Wandrey, C.; Liese, A. Bioorg. Med. Chem.
Lett. 2001, 11, 2503-2506.
3592
Org. Lett., Vol. 5, No. 20, 2003

purified by either selective precipitation using diethyl ether
or size exclusion column chromatography using Sephadex
LH-20. 1H NMR analysis of 6a showed the absence of excess
reagent and any side product, and furthermore, integration
of key signals indicated that the coupling had proceeded to
completion. The Lev group of 6a could easily be removed
by treatment with hydrazine acetate in dichloromethane to
give polymer-bound acceptor 7a. N-Iodosuccinimide (NIS)/
trimethylsilyl triflate (TMSOTf)26-mediated coupling of 7a
with thioglycosyl donor 5 gave, after purification by pre-
cipitation or size exclusion column chromatography, polymer-
bound disaccharide 8a (n′ ) 1). It is important to note that
only 1.5 equiv of glycosyl donor was required to achieve
quantitative glycosylation. Furthermore, only the â-glycoside
was formed due to neighboring group participation of the
C-2 benzoyl ester of 5. The process of Lev removal and
glycosylation with 5 was repeated twice to give a polymer-
bound tetrasaccharide, which was cleaved from the support
by hydrolysis of the phenolic ester linkage using H2O2 and
Et3N in THF for 24 h15,16 to give, after purification by
Sephadex LH-20 size exclusion column chromatography,
compound 10 in an overall yield of 24%.
in pyridine, it was decided to explore another phenolic ester
linker that can be cleaved under milder conditions. Recently,
we found that a phenolic succinyl ester linker can be cleaved
within a few minutes, whereas the corresponding hydroxy-
benzoyl oxybenzoate linker requires 18 h for cleavage.15,16
Thus, it was anticipated that the use of succinyl-modified
branched PEG 12 (Scheme 2) would allow more facile
Scheme 2
Although four-armed PEG 4a has the advantage of double
the loading capacity of traditionally used MPEG (MW )
5000 Da), it did not meet our requirement of being
compatible with purification by nanofiltration. To address
this problem, the use of branched PEG derivative 4b that
has eight arms and a molecular weight of 40 000 Da was
explored. In this case, coupling of 4b with 3 using standard
conditions gave crude 6b, which was dissolved in a mixture
of methanol/water (1/9, v/v) and applied to a Millipore high-
pressure reactor that was fitted with a Biomax filter with a
MW cutoff of 5000 Da. In this case, the polymer-bound
saccharide could be retained and excess reagent could
conveniently be removed by washing with a mixture of
methanol/water. The Lev group of 6b could be removed by
treatment with hydrazine acetate in dichloromethane, and
after purification by nanofiltration, the resulting polymer-
bound acceptor 7b was coupled with thioglycosyl donor 5
to give polymer-bound disaccharide 8b (n′ ) 1). In this case,
the polymer 8b was dissolved in acetonitrile and purified
by nanofiltration over a Biomax 5000 Da filter. The process
of Lev removal and glycosylation was repeated twice, and
after each reaction step the product was purified by nano-
filtration. Finally, the resulting polymer-bound tetrasaccharide
(9b, n′ ) 3) was released from the polymeric support by
treatment with H2O2 and Et3N in THF, but in this case, the
reaction took more than 7 days to complete and mass
spectrometric analysis of the product showed that part of
the compound had lost one or more benzoyl groups. It
appears that the phenolic ester linkage of 9b is sterically
more hindered than that of the four-armed counterpart 9a,
slowing the hydrolysis of the phenolic ester linkage to give
10.
cleavage of a final compound from the polymer support.
Thus, polymer 12 was coupled with 11 to give 14, which
was purified by nanofiltration using water/methanol (9/1, v/v)
as solvents. The TBDMS group of 14 was cleaved by
treatment with HBF4 in acetonitrile, and after purification
by nanofiltration, the polymer-bound acceptor 15 was
coupled with thioglycosyl donor 13 using NIS/TMSOTf as
the promoter system26 to give 16 (n ) 1), which was purified
by nanofiltration using acetonitrile as the solvent system. The
9-fluorenylmethoxycarbonyl (Fmoc) protecting group of 16
was removed by treatment with Et3N in CH2Cl2,27 and the
resulting glycosyl acceptor 17 (n ) 1) was coupled with 13
to give a polymer-bound trisaccharide 16 (n ) 2).
Although, the partially debenzoylated 10 could be con-
verted into one compound by treatment with benzoyl chloride
Selection of Fmoc as a temporary protecting group was
important because conditions for the removal of Lev resulted
(26) Veeneman, G. H.; van Leeuwen, S. H.; van Boom, J. H. Tetrahedron
Lett. 1990, 31, 1331-1334.
(27) Zhu, T.; Boons, G. J. Tetrahedron: Asymmetry 2000, 11, 199-
205.
Org. Lett., Vol. 5, No. 20, 2003
3593

derivative 1815 was attached to the eight-arm branched PEG
polymer 19, having a phenolic succinyl linker, by NIS/
TMSOTf-mediated glycosylation26 to give polymer-bound
sugar 20, which was immediately used in the next glycosy-
lation without workup by addition of thiogalactosyl donor
2115 and another quantity of NIS/TMSOTf.26 The glyco-
polymer was purified by nanofiltration, and the resulting 22
was treated with Et3N in dichloromethane to remove the
Fmoc group27 to give glycosyl acceptor 23, which was
purified by nanofiltration and then glycosylated with fucoysl
donor 2415,28 to give polymer bound trisaccharide 25. The
use of 1.5 equiv of glycosyl donor 24 resulted only in partial
glycosylation, and the reaction had to be repeated to drive it
to completion. The polymer-bound material could conve-
niently be cleaved from the polymer support using standard
conditions to give, after purification, 26 in an overall yield
of 35%. Oxidative removal of the p-hydroxyl benzyl moiety
of 26 with DDQ gave a lactol,15 which can be easily
converted into a glycosyl donor or deprotected as described
previously (Scheme 3).
Scheme 3
In conclusion, it has been demonstrated that commercially
available branched PEG derivatives are convenient for
polymer-supported oligosaccharide synthesis. The use of the
four-armed PEG derivative 4a (MW ) 10 000 Da) is
attractive because it has a higher loading capacity than
traditionally used MPEG (MW ) 5000 Da), and, in this case,
synthetic intermediates could be purified by precipitation or
size exclusion column chromatography. Compounds syn-
thesized on an eight-armed PEG derivative that has an
average molecular weight of 40 000 Da could be purified
by nanofiltration over a Biomax filter that has a nominal
molecular weight limit of 5000 Da. Nanofiltration is more
convenient than precipitation and will provide a future
opportunity for automation. This new technology allowed
the preparation of a branched trisaccharide of biological
importance that has R- and â-glycosidic linkages.
in partial cleavage of the succinyl linker. The process of
Fmoc removal, glycosylation, and Fmoc removal was
repeated and the resulting tetrasaccharide was released from
the polymeric support by treatment with NH2NH2‚CH3-
COOH, and in this case, 10 was obtained as a single
compound in an overall yield of 40%.
Having established that an eight-armed PEG derivative in
combination with a phenolic succinyl linker is convenient
for oligosaccharide synthesis, attention was focused on the
preparation of the tumor-associated antigen Lewisx using a
previously reported glycosylation strategy.15 Thus, in this
case, a trichloroethoxycarbonyl (Troc)-protected glucosamine
Acknowledgment. This work was supported by the NIH
Resource Center for Biomedical Complex Carbohydrates
(P41-RR-5351) and NIH National Cancer Institute
(CA088986).
Supporting Information Available: Experimental pro-
cedures, 1H NMR spectra, and HRMS data. This material is
available free of charge via the Internet at http://pubs.acs.org.
OL0352355
(28) Kiyoi, T.; Nakai, Y.; Kondo, H.; Ishida, H.; Kiso, M.; Hasegawa,
A. Bioorg. Med. Chem. 1996, 4, 1167-1176.
3594
Org. Lett., Vol. 5, No. 20, 2003