Cap and capture-release techniques applied to solid-phase synthesis of oligosaccharides

Article abstract of DOI:10.1021/jo060255g

This paper reports a new strategy for oligosaccharide synthesis by combining solid-phase methods with cap and capture-release separation techniques, using the p-(5-(ethoxycarbonyl)pentyloxy)benzyl group (CPB) as a tag for the capture of desired oligosacch

Full text of DOI:10.1021/jo060255g

carbohydrate synthesis are much less effective,10 which is even
more so for the two-phase reactions in solid-phase synthesis.
Consequently, after a few steps of glycosylations, a significant
amount of side products with deficient carbohydrate sequences
will accumulate on the resin, whereas only a low yield of the
desired oligosaccharide will be formed. After the products are
released from the resin, the desired oligosaccharide has to be
isolated from a large quantity of impurities, which can be
especially challenging. Meanwhile, the removal of an array of
side products that are different by just one sugar unit is another
notable difficulty.11 Therefore, an efficient method to purify the
desired final product is one of the key issues that needs to be
addressed to achieve practical, automated solid-phase carbohy-
drate synthesis.
Cap and Capture-Release Techniques Applied to
Solid-Phase Synthesis of Oligosaccharides
Jian Wu and Zhongwu Guo*
Department of Chemistry, Wayne State UniVersity,
5101 Cass AVenue, Detroit, Michigan 48202
zwguo@chem.wayne.edu
ReceiVed February 7, 2006
In recent years, many new and interesting strategies have been
developed for product isolation and purification in organic
synthesis,12,13 among which the capture-release technique using
various unique “capture” reagents has been extensively inves-
tigated in combinatorial, solution-phase, and solid-phase syn-
thesis of peptides, carbohydrates, and other molecules.14,15 The
basic principle of this strategy is to label the desired product
with a tag during the synthetic process. After the synthesis is
finished, the tagged structures are captured by materials, such
as resin or column, which can specifically bind or react with
the tag to facilitate the separation of the desired products.
Insoluble polymers are commonly used as capture materials,
as they enable separation by filtration.
The polymer-based capture-release technique has brought
up many brilliant designs for carbohydrate synthesis.15 For
example, in soluble polymer-supported carbohydrate synthesis,
Ito and co-workers10,16,17 used the chloroacetyl (ClAc) group
as a temporary protection of hydroxyl groups and as the tag for
solid-phase capture-release as well. After each glycosylation
reaction, the product was treated with a capture resin having
free thio groups. Although the unreacted glycosyl acceptor was
inert to capture resin, the elongated sugar chains, which
contained ClAc, would link to the resin via the thio groups.
The loaded resin could be easily separated from the side products
by filtration. Thereafter, the oligosaccharide on the capture resin
was released and then subjected to the next cycle of polymer-
supported reaction and capture-release purification. A useful
feature of this approach is that each glycosylation product was
purified. Another elegant approach was developed by Fukase
and co-workers18,19 using 3-chloro-4-azidobenzyl (ClAzb) as a
temporary protecting group and the tag of incoming sugar units.
After the elongation was completed and the carbohydrates were
detached from the solid-phase support, the desired oligo-
This paper reports a new strategy for oligosaccharide
synthesis by combining solid-phase methods with cap and
capture-release separation techniques, using the p-(5-
(ethoxycarbonyl)pentyloxy)benzyl group (CPB) as a tag for
the capture of desired oligosaccharides. After a complex
carbohydrate mixture was obtained by solid-phase synthesis,
the desired oligosaccharide containing a free carboxyl group
derived from CPB was attached to an amino resin. The
loaded resin was readily separated from side products by
filtration and finally treated with acid to release the pure
oligosaccharide product.
Among the three major classes of biooligomers, which include
oligosaccharides, oligonucleotides, and oligopeptides, the struc-
ture of oligosaccharides is by far the most complex and diverse.
As a result, the chemical synthesis of oligosaccharides posts
an important challenge in organic chemistry,1 despite the
significant advancement in carbohydrate chemistry in the past
three decades.2-5 For instance, although automated solid-phase
synthesis of oligonucleotides and oligopeptides using synthesiz-
ers has become routine, the attempt to automate solid-phase
oligosaccharide synthesis has just started.6 Solid-phase carbo-
hydrate synthesis has shown great promise,7-9 as it can
significantly simplify the oligosaccharide assembly by eliminat-
ing the time-consuming process of intermediate purification.
However, it is exactly this feature that often complicates the
final product purification.
(10) Ando, H.; Manabe, S.; Nakahara, Y.; Ito, Y. Angew. Chem., Int.
Ed. 2001, 40, 4725.
Compared to the coupling reactions used in peptide and
nucleotide synthesis, glycosylation reactions established for
(11) Palmacci, E. R.; Hewitt, M. C.; Seeberger, P. H. Angew. Chem.,
Int. Ed. 2001, 40, 4433.
(12) Curran, D. P. Angew. Chem., Int. Ed. 1998, 37, 1174.
(13) Tzschucke, C. C.; Markert, C.; Bannwarth, W.; Roller, S.; Hebel,
A.; Haag, R. Angew. Chem., Int. Ed. 2002, 41, 3964.
(14) Kirschning, A.; Monenschein, H.; Wittenberg, R. Chem.-Eur. J.
2000, 6, 4445.
(15) Ito, Y.; Manabe, S. Chem.-Eur. J. 2002, 8, 3076.
(16) Hanashima, S.; Manabe, S.; Ito, Y. Synlett 2003, 979.
(17) Hanashima, S.; Manabe, S.; Ito, Y. Angew. Chem., Int. Ed. 2005,
44, 4218.
(1) Paulsen, H. Chem. Soc. ReV. 1984, 15.
(2) Boons, G. J. Contemp. Org. Synth. 1996, 3, 173.
(3) Koeller, K. M.; Wong, C.-H. Chem. ReV. 2000, 100, 4465.
(4) Demchenko, A. V. Synlett 2003, 1225.
(5) Boons, G.-J. Tetrahedron 1996, 54, 1095.
(6) Plante, O. J.; Palmacci, E. R.; Seeberger, P. H. Science 2001, 1523.
(7) Seeberger, P. H.; Danishefsky, S. J. Acc. Chem. Res. 1998, 31, 685.
(8) Seeberger, P. H.; Haase, W.-C. Chem. ReV. 2000, 100, 4349.
(9) Sears, P.; Wong, C. H. Science 2001, 291, 2344.
(18) Egusa, K.; Kusumoto, S.; Fukase, K. Synlett 2001, 777-780.
(19) Egusa, K.; Kusumoto, S.; Fukase, K. Eur. J. Org. Chem. 2003, 3435.
10.1021/jo060255g CCC: $33.50 © 2006 American Chemical Society
Published on Web 07/29/2006
J. Org. Chem. 2006, 71, 7067-7070
7067

SCHEME 1
saccharide was captured by a resin that could selectively react
with the azido group. One potential problem that may affect
the efficiency of the capture-release purification is that the
unreacted glycosyl acceptors may participate in subsequent chain
elongation.
Another useful and closely related technique in organic
synthesis is the employment of a scavenger resin to remove
byproducts or the excessive reagents.20-22 Seeberger and co-
workers11 have made creative use of this technique by combining
it with automated carbohydrate synthesis and the cap technique.
After each glycosylation, the unreacted glycosyl acceptor was
capped by the 2-azido-2-methylpropionyl (AMP) group. AMP
also served as a tag to facilitate the scavenge technique. Once
the chain elongation was complete and the products were
detached from the solid-phase support, side products were
removed by a resin via its selective reaction with AMP and
filtration. The use of the cap technique reduced the complexity
of the resultant product mixture.
SCHEME 2a
a Reagents and conditions: (a) LevOH, DIPC, DMAP, CH2Cl2, 0 °C to
room temperature, 4 h, 95%; (b) TBAF, HOAc, THF, o/n, 95%; (c) succinic
anhydride, DMAP, pyridine, o/n, quantitative.
In the present paper, we described a new strategy for
carbohydrate synthesis that combines solid-phase synthesis with
cap and capture-release techniques. We expect that it will
significantly facilitate the final product purification and improve
carbohydrate synthesis efficiency.
hydrazine at room temperature.23 Third, the reaction used to
couple the desired oligosaccharide and the capture resin has to
be simple and effective. To fulfill this condition, we designed
the p-(5-carboxypentyloxy)benzyl (CPB) group as a tag of the
terminal sugar unit. The carboxyl group of CPB can react with
amino resins effectively enabling the product capture. Mean-
while, CPB should be easily removable under mild oxidative
or moderate acidic conditions to realize the release of attached
oligosaccharides later on.
We chose an R-linked trimannoside 1 as our synthetic target
to test the new strategy (Scheme 1). This trisaccharide appears
in many natural glycoconjugates24,25 and is therefore the focus
of a number of synthetic studies.11 Following the discussion
above, 1 was disconnected into three logical building blocks
2-4. As 2 contained a free carboxylic group that enabled its
attachment to the solid-phase support, its 2-OH was temporarily
protected by a Lev group. Glycosyl donor 3 was the typical
building block used to elongate carbohydrate chains for the new
strategy, and 4 was the terminal sugar unit having the unique
tag. For glycosylations, we planned to use the Schmidt method26
because the reagents and byproducts of this method can be easily
washed off after the reaction is finished.
Compound 2 was prepared according to a route shown in
Scheme 2 with methyl D-mannoside (5) as the starting material.
The conversion of 5 to 6 followed a reported procedure.27 The
introduction of Lev to 6 was realized by reaction with levulinic
acid and N,N′-diisopropylcarbodiimide (DIC), to afford an
excellent yield (95%) of the desired ester 7. This condition was
also used to install Lev in the synthesis of 3. Thereafter, the
tert-butyldimethylsilyl (TBS) group was selectively removed
with tetrabutylammonium fluoride (TBAF), which was followed
by reaction with succinic anhydride to afford 2. The free
carboxyl group in 2 enabled its coupling to the polymer support
for solid-phase synthesis.
Our strategy is outlined as follows. After each glycosylation,
we use a cap to block the remaining hydroxyl groups, which
can prevent the deficient structures from further elongation.
Next, the temporary protecting group of the installed sugar unit
will be removed to expose a free hydroxyl group and subject
to the next cycle of glycosylation, capping, and deprotection.
The same protocol will be repeated to introduce other mono-
saccharide units. The last glycosyl donor is designed to bear a
unique functionality. Thus, the final synthetic target, and only
the desired one, will contain this unique group to facilitate its
capture by a resin. After the side products, which remain in
solution, are removed through filtration, the desired oligosac-
charide will be released from the capture resin. Theoretically,
the product thus obtained should be homogeneous.
This new strategy will benefit the final product purification
in several aspects. First, the cap technique will minimize the
number and complexity of side products formed on the resin.
Second, no matter how complex or big the desired oligosac-
charide is, only the last glycosyl donor needs a special design.
This can simplify the selection of protection tactics and the
synthesis of building blocks. Third, the capture resin will catch
or react with the desired product only, because in the mixture
only it contains the tag. Therefore, even if the coupling reaction
is less than perfect, only the recovery rate of the desired product
will be compromised and not its purity.
For the new strategy to work well, we need to address several
issues. First, the cap reaction must be efficient and easily
achievable, and the cap must be stable to the reactions involved
in carbohydrate chain elongation. In this respect, the acetyl group
seems to be an ideal choice. Second, the reaction conditions
used to remove the temporary protecting group must be
compatible with the cap as well as the linker between the solid
support and the sugar chain. We chose the levulinoyl group
(Lev) for this use, as it can be easily removed by reaction with
(23) Van Boom, J. H.; Burgers, P. M. J. Recl. TraV. Chim. Pays-Bas.
1978, 93, 73.
(24) Ferguson, M. A. J.; Williams, A. F. Annu. ReV. Biochem. 1988, 57,
(20) Booth, R. J.; Hodges, J. C. Acc. Chem. Res. 1999, 32, 18
(21) MacCoss, R. N.; Brennan, P. E.; Ley, S. V. Org. Biomol. Chem.
2003, 1, 2029.
(22) Ito, Y.; Kanie, O.; Ogawa, T. Angew. Chem., Int. Ed. 1996, 35,
2510.
285.
(25) Kornfeld, R.; Kornfeld, S. Annu. ReV. Biochem. 1985, 54, 631.
(26) Schmidt, R. R.; Kinzy, W. AdV. Carbohydr. Chem. Biochem. 1994,
50, 21.
(27) Gurjar, M. K.; Talukdar, A. Tetrahedron 2004, 60, 3267.
7068 J. Org. Chem., Vol. 71, No. 18, 2006

SCHEME 3a
SCHEME 4a
a Reagents and conditions: (a) LevOH, DIPC, DMAP, CH2Cl2, 0 °C to
room temperature, 4 h, g95%; (b) NIS, CH3CN/H2O (9:1), 5 min, >99%;
(c) NaH, CCl3CN, CH2Cl2, 0 °C, 20 min, g90%; (d) NaH, ethyl 6-(4-
bromomethylphenyloxyl)hexanoate, DMF, -5 °C, 1 h, 90%.
Compounds 3 and 4 were prepared from a common inter-
mediate 9 (Scheme 3).28 After the introduction of a Lev group
to the 2-OH of 9 and oxidative hydrolysis at the anomeric
position of 10, the hemiacetal product 11 was transformed into
the Schmidt donor 3 in an excellent yield. In the preparation of
4, prior to the oxidative hydrolysis and subsequent imidation
of the anomeric hydroxyl group, the 2-OH of 9 was protected
by CPB first. All transformations shown in Scheme 3 were very
efficient.
For solid-phase oligosaccharide assembly (Scheme 4), we
used commercially available poly(ethylene glycol) (PEG)-
grafted polystyrene with an amino headgroup (0.77 mmol/g)
as the solid-phase support. The coupling of 2 to this resin was
realized with DIC/HOBt as the condensation reagent. The
reaction was effective, judging from the reaction yield (81%)
calculated on the basis of the weight gain following the coupling
reaction. The remaining free amino groups were capped by
treating the resin with acetic anhydrate in pyridine (1:2).
Ultimately, the carbohydrate loading of resin 14 was ca. 0.46
mmol/g of dry resin. Thereafter, the carbohydrate chain was
elongated by repeated deprotection, glycosylation, and capping,
using 3 and 4 as glycosyl donors. The deprotection to remove
the Lev group was achieved with hydrazine.23 To ensure
complete deprotection, the resin was treated with hydrazine for
a prolonged time (overnight). For Schmidt glycosylation,
trimethylsilyl triflate (TMSOTf) was used as the promoter, and
the reactions were performed at -20 °C using excessive donors
(3 equiv). This reaction did not affect the protecting groups or
the linkers. We found that the glycosylation using 3 gave a 52%
yield calculated on the basis of the weight gain of the resin.
Similar results were obtained with 4, and the reaction yield was
50%. During carbohydrate chain elongation, the cap of unreacted
hydroxyl groups was realized by treating the resin with acetic
anhydride in pyridine (1:2).
After two cycles of chain elongation, the carbohydrate chains
were released from the polymer support by NaOMe in methanol/
dichloromethane, which also removed the acetyl caps. To make
sure that the carboxyl group of CPB was deprotected, the
product was further treated with NaOH. These operations
eventually resulted in a complex mixture (17) as proved by its
NMR spectrum (Figure 1A). The free hydroxyl groups in 17
were then blocked by acetyl groups, followed by treatment with
the fishing resin. The coupling between the fishing resin and
CPB was achieved with DIC/HOBt. After washing with
solvents, the loaded resin was treated with 10% trifluoroacetic
acid (TFA) in dichloromethane. The resin was filtered off and
washed with dichloromethane. The filtrate and the washing were
a Reagents and conditions: (a) DIC, HOBt, DMF/CH2Cl2 (1:1), room
temperature, 24 h; (b) Ac2O/pyridine (1:2), room temperature, 6 h; (c)
hydrazine monohydrate, Py/HOAc (3:2), room temperature, o/n; (d)
TMSOTf (0.1 equiv), CH2Cl2, -20 °C, 2 h; (e) NaOMe, CH2Cl2/MeOH
(9:1), room temperature, 6 h; (f) NaOH, THF/MeOH (5:1), room temper-
ature, o/n; (g) Ac2O/pyridine (1:10); (h) TFA/CH2Cl2 (1:9), room temper-
ature, 2 h.
combined, rinsed, dried, and condensed to give 1 that was
positively identified by NMR and MS spectroscopy. The NMR
spectrum of 1 (Figure 1B) suggested its homogeneity. To prove
the purity, it was further subjected to silica gel column
chromatography. The NMR spectrum (Figure 1C) of chroma-
tography-purified 1 was essentially identical to that of the
product released from the fishing resin.
We have proposed and tested in this paper a new strategy
for carbohydrate synthesis, which combines solid-phase syn-
thesis with the cap and polymer-based capture-release tech-
niques for reducing side-product formation and facilitating final
product purification, respectively. The basic concept is that
during solid-phase elongation of the carbohydrate chains acetic
anhydride was employed to block free hydroxyl groups of
unreacted glycosyl acceptors after each glycosylation, so that
the side products with deficient sequences would not grow
further. Meanwhile, the last glycosyl donor used was modified
(28) Paulsen, H.; Heume, M. Carbohydr. Res. 1990, 200, 127.
J. Org. Chem, Vol. 71, No. 18, 2006 7069

envision that the more complex the synthetic targets are, the
more advantageous the new strategy can be.
Experimental Section
Detachment of Carbohydrates from the Solid-Phase (Prepa-
ration of 17). Dry resin 16 (215 mg) was mixed with CH2Cl2 (2.7
mL), and the resultant suspension was shaken under argon for 10
min. A solution of MeONa (0.46 mmol) in MeOH (0.3 mL) was
added, and the mixture was shaken under an N2 atmosphere for 6
h. TLC analysis confirmed the formation of a carbohydrate mixture
in solution. The resin was filtered off and rinsed with CH2Cl2/MeOH
(9:1, 3 × 5 mL). This detachment procedure was repeated to
examine if the cleavage was complete. Then, cation-exchange resin
(HP+ form) was added to the combined filtrate and washed to
neutralize the solution. After the ion-exchange resin was filtered
off, the filtrate was concentrated to yield a mixture (70 mg). It was
dissolved in a 5:1 mixture of THF and MeOH (5 mL), and an
aqueous NaOH solution (4 M, 0.1 mmol) was added dropwise at
room temperature. After stirring overnight, the reaction mixture
was diluted with CH2Cl2 and washed with H2O. The organic layer
was dried (Na2SO4) and concentrated in a vacuum to offer 17 as a
mixture.
Fishing Out the Desired Oligosaccharide from 17 (Prepara-
tion of 18). After a mixture of 17, Ac2O (0.2 mL), and pyridine (2
mL) was stirred at room temperature for 6 h, it was diluted with
CH2Cl2, followed by washing with 1 M HCl aq. solution and H2O.
The organic layer was dried over Na2SO4 and concentrated in a
vacuum. The residue, HOBt (32 mg, 0.24 mmol), and DIC (0.037
mL, 0.24 mmol) were dissolved in dry CH2Cl2 (0.5 mL). The
resultant solution was added to NovaGel-NH2 (60 mg, 0.048 mmol)
swollen in DMF (0.5 mL) under an argon atmosphere dropwise at
room temperature. After the mixture was shaken for 24 h, the solid
material was filtered off, washed with H2O (2 × 15 mL), DMF (2
× 5 mL), THF (2 × 10 mL), and CH2Cl2 (2 × 10 mL), and finally
dried under vacuum to give 18 (72 mg).
Release of the Desired Oligosaccharide 1 from the Fishing
Resin. The resin 18 obtained above (72 mg) was mixed with a
solution of 10% TFA in CH2Cl2 (1 mL), and the resultant
suspension was shaken under argon at room temperature for 2 h.
Thereof, TLC analysis confirmed the release of a single carbohy-
drate product. The resin was then filtered off and rinsed with
CH2Cl2. This procedure was repeated once to confirm the complete
release of carbohydrates from the resin. The combined filtrate and
washing were washed with NaHCO3 and H2O. The organic layer
was dried over Na2SO4 and concentrated in a vacuum to afford the
desired product 1 (10 mg). Rf ) 0.34 (toluene/EtOAc ) 10:1).
[R]D20 +26.5° (c 0.2, CHCl3). 1H NMR (CDCl3, 400 MHz): 7.40-
7.14 (m, 40 H), 5.18 (bs, 1 H), 5.13 (bs, 1 H), 4.86-4.78 (m, 4 H),
4.69-4.41 (m, 12 H), 4.36 (d, J ) 12.4 Hz, 1 H), 4.32 (bd, J )
10.0 Hz, 1 H), 4.17 (dd, J ) 11.6, 4.0 Hz, 1 H), 4.11 (m, 2 H),
3.99-3.74 (m, 9 H), 3.71-3.61 (m, 4 H), 3.57 (m, 1 H), 3.20 (s, 3
H), 2.39 (d, 1 H), 1.86 (s, 3 H). 13C NMR (125 MHz, CDCl3) 171.3,
138.8, 138.6, 138.6, 138.5, 138.4, 138.4, 138.2, 133.2, 129.9, 128.7,
128.6, 128.6, 128.5, 128.3, 128.2, 128.1, 128.0, 128.0, 127.9, 127.8,
127.8, 127.7, 127.6, 123.8, 116.1, 101.2, 101.0, 99.8, 80.2, 79.5,
75.6, 75.3, 75.0, 74.53, 74.2, 73.6, 73.4, 72.4, 72.4, 72.3, 71.8, 71.7,
69.7, 69.7, 69.4, 69.2, 68.8, 64.2, 63.5, 55.0, 32.2, 31.8, 29.9, 29.9,
29.6, 21.0. FAB MS (m/e): calcd for C77H84NaO17 (M + Na+),
1303.6; found 1304.0.
FIGURE 1. 1H NMR spectra of (A) 17; (B) 1 released from the fishing
resin; and (C) 1 after further purification by silica gel column
chromatography.
with CPB that contained a carboxyl group. After all products
were detached from the solid-phase support, the CPB-containing
carbohydrates were able to couple to a second resin, i.e., the
capture resin, that contained free amino groups. Although a
complex carbohydrate mixture was derived from solid-phase
synthesis, only the desired oligosaccharide had CPB and could
be captured by this resin. The loaded capture resin was
conveniently isolated by filtration and then treated with TFA
to release the captured oligosaccharide. Our results showed that
the oligosaccharide released from the capture resin was pure
and that no further purification was necessary. Because solid-
phase carbohydrate synthesis almost always gives complex
mixtures, the isolation of the desired oligosaccharide from a
large quantity of similar impurities is extremely challenging.
The synthetic strategy described herein can significantly simplify
and accelerate the final product purification process and result
in pure oligosaccharides. Thus, the new strategy should be
especially useful for promoting the use of a solid-phase strategy
in oligosaccharide synthesis. It is worth noting that our designs
are very simple with regard to the synthesis of the tagged sugar
unit and the cap and capture-release separation techniques,
which makes this strategy practical. The results have unequivo-
cally demonstrated the effectiveness of all designs and have
thus provided a proof of principle. We predict that this strategy
should be applicable to more complex structures. In fact, we
Acknowledgment. We thank the NSF for the financial
support of this work through a research grant, CHE-0554777.
Supporting Information Available: Experimental details and
selected NMR spectra of intermediates and the final product. This
material is available free of charge via the Internet at http://
pubs.acs.org.
JO060255G
7070 J. Org. Chem., Vol. 71, No. 18, 2006