Green glycosylation using ionic liquid to prepare alkyl glycosides for studying carbohydrate-protein interactions by SPR

Article abstract of DOI:10.1039/b814171a

Several simple glycosides of d-glucose (Glc) and N-acetyl-d-galactosamine (GalNAc) were prepared in a single step glycosylation reaction using unprotected and unactivated sugar donors. The resulting GalNAc glycoside, containing a bifunctional linker, was used to immobilize this glycoconjugate to a self assembled monolayer on a gold biosensor chip. Surface plasmon resonance (SPR) experiments demonstrated that this immobilized glycoconjugate bound to GalNAc specific lectin, Viscum album agglutinin.

Full text of DOI:10.1039/b814171a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/greenchem | Green Chemistry
Green glycosylation using ionic liquid to prepare alkyl glycosides
for studying carbohydrate–protein interactions by SPR
F. Javier Mun˜oz,a Sabine Andre´,b Hans-Joachin Gabius,b Jose´ V. Sinisterra,a,c Mar´ıa J. Herna´iz*a,c,d and
Robert J. Linhardt*e
Received 14th August 2008, Accepted 25th November 2008
First published as an Advance Article on the web 24th December 2008
DOI: 10.1039/b814171a
Several simple glycosides of D-glucose (Glc) and N-acetyl-D-galactosamine (GalNAc) were
prepared in a single step glycosylation reaction using unprotected and unactivated sugar donors.
The resulting GalNAc glycoside, containing a bifunctional linker, was used to immobilize this
glycoconjugate to a self assembled monolayer on a gold biosensor chip. Surface plasmon
resonance (SPR) experiments demonstrated that this immobilized glycoconjugate bound to
GalNAc specific lectin, Viscum album agglutinin.
methods for the anomeric position (halides, thioglycosides,
Introduction
trichloroacetimidates, etc.) are required in traditional synthetic
Glycoconjugates including proteoglycans, glycoproteins and
chemistry to obtain new glycoconjugates.6 A major problem
glycolipids are involved in many different biological events,
in the synthesis of glycoconjugates is the need for multiple
such as molecular recognition for cell–cell interaction.1–3 Owing
protection, deprotection and activation steps. The crucial impor-
to their ubiquitous presence at the cell membrane surface,
tance of glycoconjugates in the study of glycobiology cries out
carbohydrates are located in an environment containing many
for alternative, simpler methods for their synthesis. Enzymatic
proteins, such as growth factors, cytokines, receptors, enzymes,
synthesis of carbohydrates can be used to avoid many protection,
and viruses and bacterial proteins. The numerous biological roles
deprotection and activation steps and the products are often
of carbohydrates are attributed to their interactions with these
obtained with high regio- and stereoselectivity.7–9 However,
proteins, called lectins. Membrane surface carbohydrates, called
enzymes are often less efficient for the functionalization of
glycoconjugates, modulate lectin activity at the cell-extracellular
glycoconjugates with linker chains required their immobilization
interface. Many studies have been directed towards understand-
to study their interaction with lectins.
ing the recognition of these cell surface glycoconjugates by
Glycosylation using unprotected and unactivated donors is
lectins from different sources.3,4 The most widely used techniques
(i.e., enzyme linked immunosorbent assays, microarrays, atomic
often preferable as it can reduce the number of steps, en-
hance reactivity, allow a different stereochemistry, and increase
force microscopy and surface plasmon resonance (SPR)) for the
prospects for further process modifications.10,11 The chemical
study of such molecular interactions require the immobilization
of either the lectin or the glycoconjugate.5
A number of strategies have been developed to mimic the
synthesis of unprotected carbohydrates possesses a number of
challenges, including their poor solubility in most conventional
solvents. It is important to investigate new solvent systems that
presentation of glycoconjugates on the membrane surface. In ad-
dissolve carbohydrates and support glycosidation reactions of
dition, many different methodologies have been investigated for
unprotected sugars.
the synthesis of glycoconjugates. Protecting groups for hydroxyl
Recently, room temperature ionic liquids (RTILs) have re-
moieities in carbohydrates (acyl, ethers, etc.) and activation
ceived attention as solvents for a wide range of chemical
processes.12 RTILs show exceptional solvent properties includ-
aBiotransformations Group, Departamento de Qu´ımica Orga´nica y
Farmace´utica, Facultad de Farmacia, Universidad Complutense
de Madrid, Pz/Ramo´n y Cajal s/n., 28040, Madrid, Spain.
E-mail: mjhernai@farm.ucm.es; Fax: +34 9139471822;
Tel: +34 913947208
ing thermal stability, low volatility and an ability to dissolve both
polar and non-polar compounds. RTILs also have been widely
researched as possible “green” replacements for organic solvents
because they have very low vapour pressures and may be used to
replace volatile organic solvents and may be easier to efficiently
reuse than organic solvents.13,14 Carbohydrate chemistry is not
an exception to this trend. In the past few years ionic liquids
have been used in both protection and glycosylation reactions,
resulting in new methodologies and enhanced yield.15–21
We report a convenient method for the modification of un-
protected and unactivated carbohydrates, of biological interest,
with a bifunctional linker. The use of RTIL solvent allowed
us to introduce a linker chain into carbohydrates in a single,
simple step. The resulting glycoconjugates were immobilized
bInstitut fu¨r Physiologische Chemie, Tiera¨rztliche Fakulta¨t,
Ludwig-Maximilians-Universita¨t, Mu¨nchen, Veterina¨rstr, 13, 80539,
Mu¨nchen, Germany
cServicio de Biotransformaciones Industriales, Parque Cient´ıfico de
Madrid C/Santiago Grisol´ıa, 28760, Tres Cantos, Spain
dServicio de Interacciones Biomoleculares, Parque Cient´ıfico de Madrid,
Unidad de Proteo´mica, Pz/Ramo´n y Cajal s/n, 28040, Madrid, Spain
eDepartments of Chemistry and Chemical Biology, Biology, Chemical
and Biological Engineering, Center for Biotechnology and
Interdisciplinary Studies, Rensselaer Polytechnic Institute, 110 8th St.,
Troy, NY 12180-3590, USA. E-mail: linhar@rpi.edu;
Fax: +1 518-276-3405; Tel: +1 518-276-3404
This journal is
The Royal Society of Chemistry 2009
Green Chem., 2009, 11, 373–379 | 373
©

View Article Online
Table 1 Functionalization of carbohydrates with different linker
Table 2 Glycosylation of Glc with different linker chainsa
chainsa
Compound Product
t
Yield a:b
Donor
Glc
Acceptor
Compound Yield
1
17 h 35%
77:23
Allyl alcohol
Benzylamine
Benzylmercaptan
4-Aminobenzyl alcohol
2-(2¢-Aminoethoxy)ethanol
tBoc-2-(2¢-Aminoethoxy)ethanol (6)
Allyl alcohol
1
2
3
4
5
7
8
9
35%
n.p.
n.p.
n.r.
n.p.
12%
n.p.
35%
7%
2
6 h n.p.
17:83
Lac
3
4
6 h n.p.
7 d n.r.
59:41
—
GalNAc Allyl alcohol
tBoc-2-(2¢-Aminoethoxy)ethanol (6) 10
a n.p.: no product isolated. n.r.: no reaction observed.
on a carboxylated functionalized matrix and used in lectin
interaction studies performed by SPR.
5
7
6 h n.p.
50:50
40:60
Results and discussion
24 h 12%
Synthesis of functionalized carbohydrates
In this study, we present the functionalization of non-activated
carbohydrates in a single step to give a glycoconjugate having
a free amino group. Thereby, it is possible to obtain various
kinds of functionalized glycoconjugates without the need of
protection, activation and deprotection steps required for the
chemical synthesis chemistry of carbohydrates. For this pur-
pose, we selected three substrates, D-glucose (Glc), N-acetyl-
D-galactose (GalNAc) and lactose as model carbohydrates for
study.
Glycosylation reactions were carried out in the presence of the
RTIL, 1-ethyl-3-methylimidazolium benzoate [emIm][ba], and
Amberlite IR-120 [H+] providing the acid milieu necessary to
promote glycosylation. The application of [emIm][ba] RTIL was
previously demonstrated in the protection and functionalization
of carbohydrate.21
This RTIL was also previously used for the glycosylation
of bulky acceptors, including monosaccharides and benzyl
alcohol.15 These initial successes prompted us to examine if the
same one-step glycosylation reaction could be similarly carried
out using linear alkyl chain acceptors. For these purposes we
initially tested the reactivity of an a,b-anomeric mixture of
Glc monosaccharide donor with the allyl alcohol as acceptor
(Table 1).
The reaction, carried out following the conditions described
in the Experimental section, led to formation of the desired com-
pound in 35% yield (Scheme 1, Table 2). Some stereoselectivity
was observed in this glycosylation reaction, giving primarily
the a-glycoside (a:b, 77:23), as previously described for similar
processes.15,22
a n.p.: no product isolated. n.r.: no reaction observed.
solvent posed a problem that was overcome by increasing the
reaction temperature to 50 ◦C. Under the conditions, allyl
alcohol afforded the complete conversion of Glc to the allyl
glycoside 1. However, the viscosity of this RTIL solvent posed
additional problems in product purification. While a nearly
quantitative yield was observed by TLC, the recovered yield
of purified glucoside product was only 35%. Despite a yield of
only 35%, the requirement of only one step makes this approach
promising for glycosylation reactions, as protection, activation
and deprotection steps would certainly decrease more greatly
the yield of isolated product.
Allyl alcohol was selected as we expected it would be an
excellent glycosidation acceptor due to its very small size when
compared to the benzylamine, benzylmercaptan, 4-aminobenzyl
alcohol or 2-(2¢-aminoethoxy)ethanol acceptors (Table 2). This
was demonstrated by the case when 4-aminobenzyl alcohol
was used as acceptor: no reaction was observed, probably
due to the poor solubility of the linker chain in the reaction
medium. In addition, benzylamine, benzylmercaptan, and 2-(2¢-
aminoethoxy)ethanol acceptors afforded a mixture of products
by TLC, possibly including the a,b-glycosides, the signals of
which correspond to the anomeric protons identified by
1H-NMR. Further isolation failed to afford the desired product
due to the high viscosity of the RTIL at room temperature
(Table 2). [emIm][ba] RTLs are relatively viscous RTLs with
a viscosity of 425 cP at 25 ◦C, making both stirring and product
recovery difficult.15
Next, a bifunctional linker chain was tested as a glycosylation
acceptor in [emIm][ba], the tert-butoxycarbonyl (tBoc) amino
protected 2-(2¢-aminoethoxy)ethanol. The advantage of this
linker is that once bound to the carbohydrate through the
glycosylation of the hydroxyl group, selective removal of tBoc
exposes an amino group that could be used for a covalent
immobilization of the resulting glycoconjugate onto surfaces.
Compound 7 was obtained as a mixture of a- and b-isomers
Scheme 1 General glycosylation reaction of Glc with various acceptors.
One of the most remarkable features of [emIm][ba] is its
high viscosity at room temperature. The viscous nature of this
374 | Green Chem., 2009, 11, 373–379
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Table 3 Glycosylation of non-modified monosaccharides with allyl
alcohol as linker chains in the presence of [emIm][ba]
Functionalization of a biosensor chip
The basic SPR sensor chip contains a gold surface to which a
ligand is immobilized. Commercially available biosensor chips,
such as the CM5 chip, typically contain a functionalized matrix
on the gold surface. Thus, when immobilizing low-molecular-
weight ligands onto these types of surfaces, the greater the
distance from the gold surface the lower the relative response
(measured as “response units”, RU), or sensitivity when a bind-
ing partner, in the solution phase, interacts with immobilized
ligand. The problem of low sensitivity has been reported by
others23 when immobilizing small molecules on a commercial
CM5 biosensor chips (Biacore).
Percentage of
Yield a-derivative
Compound Donor
Linker chain
=
1
9
Glc
GalNAc
HOCH2CH CH2
35%
35%
77%
80%
(40:60) in 12% yield (Table 2). The a and b configurations were
inferred from the chemical shifts and coupling constant of the
anomeric proton singnal.
The investigation of the glycosylation reaction in [emIm][ba]
was next extended to other glycosyl donors, including GalNAc
and lactose using allyl alcohol as acceptor (Table 1). Using
lactose as a disaccharide donor afforded a mixture of products
from which we were unable to detect the desired glycoconjugate
by 1H NMR.
Compound 11 was immobilized directly on a gold chip
through a self-assembled monolayer (SAM) prepared using
compound 14 (Fig. 1).
Using GalNAc as donor and carrying out reactions under
identical experimental conditions as described for Glc, afforded
the desired glycoside product in 35% yield (Table 3). Inter-
estingly, the conversion Glc and GalNAc afforded identical
35% isolated yields (Table 3); however, the Glc initial donor
completely disappeared in 17 h, while the GalNAc reaction
was terminated because no further evolution of product was
observed after 28 h. On the other hand, the a-stereoselectivity
observed using both Glc and GalNAc donors was surprising,
as neighboring group participation of the 2-acetamido group
typically affords the b-product using such donors. However, the
a-selectivity was also described for the glycosylation of Man,
Glc and GalNAc with benzyl alcohol by other authors.15,22
The bifunctional linker, tBoc amino protected 2-(2¢-
aminoethoxy) ethanol was also examined as a glycosylation
acceptor in [emIm][ba] using both Glc and GalNAc donors.
The conversion was very low in both cases, 12% for Glc and
7% for GalNAc (Table 4). The high polarity of the solvent may
explain the low reactivity towards this non-polar acceptor and
also may have complicated the isolation of pure products. Again,
compared to traditional glycosylation procedures, requiring
several intermediate steps, the low overall yields associated
with direct glycosylation in RTIL was considered acceptable.
Interestingly, glycosylation by GalNAc exclusively afforded the
a-glycoside (Table 4).
Fig. 1 Preparation of the chip with mixed chains and immobilization
of functionalized glycoconjugate 14.
This thiol-reactive heterobifunctional linker 14 was synthe-
sized as described in the Experimental section with good
yield (Scheme 2). Tetra(ethylene glycol) reacted with 1,8-
dibromooctane to afford compound 12 in 40% yield. Compound
12 was then oxidized with Jones reagent to afford 13 in good
yield (70%). And finally, reaction with thiourea followed by
base catalyzed hydrolysis afforded thiol group and carboxyl
containing compound 14 in 60% yield. The resulting monolayer
was very homogeneous judging from the similar RU values
obtained for both flowcells. The linker chains forming the SAM
contained a free carboxylic group for functionalization. The
SAM was formed quickly and was quite stable due to the avidity
of the thiol group for gold. Several washings with 50% aqueous
MeOH were required to eliminate non-covalently bound linker.
Deprotection of t-Boc with trifluoroacetic acid (TFA) quan-
titatively afforded the glycoconjugate 11. This reaction, in the
absence of water and at room temperature was complete within
3 min, and deprotected compound 11 decomposed to a complex
mixture unless residual TFA was quickly removed and the
sample was stored cold (< 4 ◦C). Thus, compound 11 was best
immediately immobilized on the surface of the biosensor chip
for carbohydrate–protein interaction studies.
Table 4 Glycosylation of monosaccharides with 2-(2¢-aminoethoxy)-
ethanol protected with different protecting groups
Percentage of
Yield a-derivative
Product Donor Linker chain
7
Glc
HOCH2CH2OCH2CH2NH- 12% 43%
tBoc
10
GalNAc
7% 99%
Scheme 2 Synthesis of compound 14.
This journal is
The Royal Society of Chemistry 2009
Green Chem., 2009, 11, 373–379 | 375
©

View Article Online
The GalNAc glycoconjugate 11 was immobilized to the SAM
on the biosensor chip using standard EDC/NHS activation
(Fig. 1). After several injections of glycoconjugate 11, the
response increased 600 RU and after the standard blocking of
the remaining activated binding sites with ethanolamine, the
increase of the response decreased to 374 RU. These results
compare favorably to those obtained for the immobilization of
the same concentration of glycoconjugate 11 on the carboxy
dextran matrix of a CM5 chip, where no increase in RU
was observed. Thus, the self-assembled monolayer considerably
improves the sensitivity of SPR detection using small molecule
ligands.
Experimental
The RTIL [emIm][ba] was synthesized according to the reported
procedure15 and all chemicals were purchased from Fisher
or Sigma–Aldrich. H-NMR and 13C-NMR experiments were
1
performed in a Varian 500 MHz (COSY or 1H-13C heteronuclear
experiments were also carried out as required for making
assignments). Chemical shifts (d) are indicated in parts per
million (ppm). LRMS was carried out with an Agilent 1100
series LC/MSD trap. TLC was performed on Kieselgel plates
60 F254 (Merck). Product detection on TLC plates was carried
out under a UV-lamp or with 5% H2SO4 in CH3OH and heating.
Column chromatography used silica gel 230–400 mesh (Natland
International Corporation). Sensor chips and solutions used for
the SPR assays were purchased from Biacore. SPR experiments
were carried out with a BIAcore-3000 and the sensorgrams
were analyzed with BIAEvaluation software version 4.1, 2003
(Biacore).
Interaction studies with lectin
The final step of this study was to determine the utility of
SAM containing glycoconjugates to study a carbohydrate–
protein interaction. VAA, a potent toxin and biohazard, was
selected as a model lectin to examine the bioactivity of this
surface. VAA recognizes and selectively binds to GalNAc
residues. Different concentrations of VAA were injected over
the GalNAc containing biochip and sensorgrams were obtained
and analyzed (Fig. 2). The affinity parameters obtained were
KA: 6.72 ¥ 104 M-1; KD: 1.49 ¥ 10-5 M, clearly demonstrating
binding in the micromolar range (Fig. 2). SPR, using the CM5
surface, showed no observable interaction.
The Viscum album agglutinin (VAA) lectin was purified from
mistletoe extracts of dried leaves by affinity chromatography on
lactosylated Sepharose 4B obtained by divinyl sulfone activation
and ligand coupling as crucial step and assays to ascertain
purity and activity were run by one- and two-dimensional
gel electrophoresis and gel filtration or haemagglutination and
solid-phase/cell binding.24,25
General method for glycosylation reactions in RTIL
Carbohydrate was dissolved in [emIm][ba] in the presence of
Amberlite IR-120 (H+) and a excess of the linker (Scheme 1 and
Table 1). The reaction mixture was stirred at 50 ◦C under argon
atmosphere and monitored by TLC.
When reaction was complete, or no further conversion of
substrates into products was observed, the Amberlite was
removed by filtration and washed with CH3OH. Solvent was
removed and product purified by column chromatography
(CH2Cl2–CH3OH in the appropriate ratio depending on the
polarity of compounds). If necessary, traces of RTIL could be
removed from the product using a Dowex MR-3 mixed bed
ion-exchange resin. All fractions containing carbohydrate were
analyzed by 1H-NMR.
Synthesis of 1-O-allyl-glucopyranoside (1)
Allyl alcohol (322 mL, 4.1 mmol) was added to a mixture of
Amberlite IR-120 (H+) (300 mg) and Glc (50 mg, 0.28 mmol)
[emIm][ba] (300 mg, 1.3 mmol) under the conditions described
above. Compound 1 was purified by column chromatography
eluting with CH2Cl2–CH3OH gradient (9:1 to 3:1) affording
21 mg of product in 35% yield (Table 1).
1H-NMR (500 MHz, D2O): Allyl: 5.97 (m, 1H, H-2), 5.36 (dd,
1H, J = 17.0, J = 1.8 Hz, H-3a), 5.26 (dd, 1H, J = 10.6, J =
0.7 Hz, H-3b), 4.23 (dd, 1H, J = 13.0, J = 5.7 Hz, H-1a), 4.07
(dd, 1H, J = 13.0, J = 6.2 Hz, H-1b). Glc: 4.96 (d, 1H, J =
3.8 Hz, H-1a), 3.85 (dt, 1H, J = 12.1, J = 1.9 Hz, H-6a), 3.75
(dd, 1H, J = 11.6, J = 5.3 Hz, H-6b), 3.69 (m, 1H, H-5), 3.68
(t, 1H, J = 10.2 Hz, H-3), 3.55 (dd, 1H, J = 9.8, J = 3.9 Hz,
H-2), 3.40 (t, 1H, J = 9.3 Hz, H-4). 13C-NMR (125 MHz, D2O):
133.69, 118.34, 97.47, 73.26, 71.99, 71.38, 69.81, 68.61, 60.66.
Fig. 2 A. Sensorgrams of the glycoconjugate–lectin interaction com-
pared to the control sensorgrams. B. Steady-state affinity study of
the interaction between B subunit from Viscum album lectin with
glycoconjugate 11, immobilized on an Au chip.
376 | Green Chem., 2009, 11, 373–379
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
MS: [M + Na] calcd: 243.1, found: 242.7. [M+ Cl] calcd: 255.6,
found: 256.9.
Synthesis of t-Boc-2¢-(2¢¢-aminoethoxy)ethyl-2-acetamido-
2-deoxy-galactopyranoside (10)
t-Boc-2-(2¢aminoethoxy)ethanol (232 mg, 1.13 mmol) was
added to a mixture of Amberlite IR-120 (H+) (150 mg) and
of GalNAc (25 mg, 0.11 mmol) in [emIm][ba] (150 mg,
0.65 mmol) under identical purification and reaction conditions
as described in the general procedure. Purification by column
chromatography CH2Cl2–CH3OH (5:1) afforded compound 10
(3 mg, 7%) (Table 1).
1H-NMR (500 MHz, D2O): linker chain: 3.61–3.59 (4H, H-1¢
and H-2¢), 3.53 (t, 2H, J = 5.5 Hz, H-1¢¢), 3.20 (t, 2H, J = 5.2
Hz, H-2¢¢), 1.27 (s, 9H, –(CH3)3). GalNAc: 4.96 (d, 1H, J =
2.1 Hz, H-1a), 4.11 (dd, 1H, J = 2.3, J = 4.4 Hz, H-2), 4.07
(dd, 1H, J = 4.5, J = 6.6 Hz, H-4), 3.96 (dd, 1H, J = 3.4,
J = 6.6 Hz, H-3), 3.79 (m, 2H, H-5 and H-6a), 3.64 (dd, 1H,
J = 3.7, J = 6.6 Hz, H-6b), 1.96 (s, 3H, –CH3). 13C-NMR
(125 MHz, D2O): 174.99, 156.08, 101.78, 80.22, 79.01, 72.83,
71.16, 70.02, 69.91, 64.07, 63.85, 41.00, 27.82, 23.20. MS: [M +
Na] calcd: 431.2 found: 430.8. [M + Cl] calcd: 443.5 found:
444.5.
Synthesis of t-Boc-2-(2¢-aminoethoxy)ethanol (6)
2-(2¢Aminoethoxy)ethanol (10 5 mL, 0.95 mmol) was dissolved
in 15 mL of anhydrous THF and cooled by stirring in an ice-
water bath, NEt3 (13 mL) was added and the reaction mixture
was stirred under argon atmosphere at 0 ◦C. (t-Boc)2O (311 mg,
1.42 mg) was slowly added to the reaction mixture at 0 ◦C, and
then allowed to warm to room temperature with stirring. When
reaction was complete, aqueous NaHCO3 was added and this
mixture was extracted with ethyl acetate (3 ¥ 50 mL). The organic
phase was dried with anhydrous MgSO4, filtered and the solvent
was removed by rotary evaporation. The final product was
purified by column chromatography (hexanes–EtOAc gradient
1:1 to 1:9) affording compound 6 (158 mg, 81%).
1H-NMR (500 MHz, CDCl3): 5.15 (s, 1H, –NH–), 3.66 (t,
2H, J = 4.8 Hz, H-2¢), 3.53 (t, 2H, J = 4.8 Hz, H-1¢), 3.50
(t, 2H, J = 5.5 Hz, H-2), 3.23 (t, 2H, J = 5.5 Hz, H-1), 2.73
(s, –OH), 1.44 (s, 9H,–(CH3)3). 13C-NMR (125 MHz, CDCl3):
156.10, 79.35, 72.16, 70.26, 61.62, 40.27. ESI-MS: [M + Na]
calcd: 228.1, found: 228.0.
Synthesis of 2¢-(2¢¢-aminoethoxy)ethyl-2-acetamido-2-deoxy-
galactopyranoside (11)
Synthesis of t-Boc-2-(2¢-aminoethoxy)ethyl-glucopyranoside (7)
Trifluoroacetic acid (100 mL) was added to 1 mg of compound
10 and reaction was stirred for 1 min. The mixture was dried
under vacuum and compound 11 was analyzed by NMR without
purification.
1H-NMR (500 MHz, D2O): 5.10 (d, 1H, J = 3.6, H-1a),
4.51 (d, 1H, J = 8.3, H-1b), 4.04–3.51 (H-2 to H-6, H-1¢
and H-2¢), 3.51 (t, 2H, J = 5.5 Hz, H-1¢¢), 3.09 (t, 2H, J =
5.5 Hz, H-2¢¢), 1.92 (s, 3H, –CH3). 13C-NMR (125 MHz, D2O):
173.54, 100.35, 81.22, 74.55, 73.71, 71.39, 70.04, 65.11, 64.54,
60.01, 40.89, 23.21. ESI-MS: [M+ Na] calcd: 331.2, found:
331.0.
t-Boc-2-(2¢aminoethoxy)ethanol (58 mg, 0.28 mmol) was added
to a mixture of Amberlite IR-120 (H+) (30 mg) and Glc(5 mg,
0.023 mmol) in [emIm][ba] (30 mg, 0.13 mmol) under identical
purification and reaction conditions as described in the general
procedure. Column chromatography was performed in CH2Cl2–
CH3OH (50:1 to 4:1) gradient, affording compound 7 (1 mg,
12% yield) (Table 1).
1H-NMR (500 MHz, D2O): 5.16 (d, 1H, J = 3.9, H-1a),
4.57 (d, 1H, J = 7.9, H-1b), 3.84–3.16 (14H, H-2 to H-6, and
sugar), 1.44 (s, 9H,–(CH3)3). 13C-NMR (125 MHz, D2O): 156.08,
107.78, 81.76, 79.38, 77.78, 75.83, 73.16, 70.15, 70.09, 65.30,
64.87, 40.54, 27.92. MS: [M + Na] calcd: 390.2, found: 389.1.
[M + Cl] calcd: 402.7, found: 403.5.
Synthesis of 20-bromo-3,6,9,12-tetraoxatrieicosan-1-ol (12)
NaH (160 mg, 6.5 mmol) was added to a solution of
tetraethyleneglycol (1 g, 5 mmol) in anhydrous THF and
th◦e reaction mixture was stirred under argon atmosphere at
0 C. After 3 h the reaction mixture, under argon atmosphere
at 0 ◦C, was added over a period of 4 h to a solution of
1,8-dibromooctane (1.4 g, 5 mmol). The NaBr formed was re-
moved by filtration and the solvent was evaporated. Purification
of final product was carried out by column chromatography
(CH2Cl2–CH3OH 20:1) affording 12 in 40% yield.
Synthesis of 1-O-allyl-2-acetamido-2-deoxy-
galactopyranoside (9)
Allyl alcohol (300 mL, 5.0 mmol) was added to a mixture of
Amberlite IR-120 (H+) (60 mg) and GalNAc (10 mg, 0.05 mmol)
in [emIm][ba] (60 mg, 0.26 mmol) under identical purification
and reaction conditions as described in the general procedure.
The solvent was removed with a high-vacuum pump to afford
compound 9 (4 mg, 35%) (Table 1).
1H-NMR (500 MHz, D2O): Allyl: 5.90 (m, 1H, H-2), 5.28 (dd,
1H, J = 17.2, J = 1.5 Hz, H-3a), 5.23 (dd, 1H, J = 10.6, J =
1.1 Hz, H-3b), 4.13 (1H, H-1a), 3.98 (dd, 1H, J = 13.0, J = 6.1
Hz, H-1b). GalNAc: 4.89 (d, 1H, J = 3.8 Hz, H-1a), 4.10 (dd,
1H, J = 10.9 J = 3.8 Hz, H-2), 3.94–3.90 (m, 2H, H-6), 3.86
(dd, J = 10.9 J = 3.2 Hz, H-3), 3.72–3.64 (m, 2H, H-4, H-5),
1.90 (s, 3H, –CH3). 13C-NMR (125 MHz, D2O): 175.04, 134.10,
122.20, 118.34, 96.66, 71.41, 68.89, 68.11, 61.65, 50.33, 22.34.
MS: [M + Na] calcd: 284.1, found: 283.7. [M + Cl] calcd: 296.6,
found: 296.9.
1H-NMR(300 MHz, CDCl3): 3.62 (m, 16H, H-1 a H-8), 3.43
(t, 2H, J = 6.56 Hz, H-9), 3.38 (t, 2H, J = 7.03 Hz, H-19), 1.
83 (q, 2H, J = 7.2 Hz, H-18), 1.55 (q, 2H, J = 6.90 Hz, H-10),
1.38 (m, 2H, H-11), 1.25 (m, 12H, H-12 and H-17). 13C-NMR
(125 MHz, CDCl3): 72.63 (C-9), 71.52 (C-7), 70.61 (C-3, C-4,
C-5, C-6), 70.28 (C-8), 70.05 (C-2), 61.50 (C-1), 34.00 (C-19),
32.89 (C-18), 29.56 (C-10, C-12, C-13, C-14, C-15), 28.72 (C-11),
28.18 (C-16), 26.13 (C-17). Analysis calculated for C16H33BrO5:
C: 49.87%; H: 8.63%; Found: C: 49.88%; H: 8.65%. ESI-MS:
[M + Na] Calcd: 407.2. Found: 407.3.
This journal is
The Royal Society of Chemistry 2009
Green Chem., 2009, 11, 373–379 | 377
©

View Article Online
Synthesis of 20-bromo-3,6,9,12-tetraoxatrieicosan-1-oic
acid (13)
Immobilization of glycoconjugate 11 on the Au chip (Biacore).
For the immobilization of GalNAc glycoconjugate (11), the
carboxylated groups of the self-assembled monolayer were
activated by the injection of the amino coupling kit (Biacore):
A mixture of 12 (650 mg, 1.7 mmol) in acetone (50 mL) was
stirred at room temperature and Jones reagent (2.4 mmol in
CrO3) was added dropwise over 15 min. The reaction was stirred
for an additional 30 min followed by the addition of few drops
of 2-propanol. Saturated NaCl solution (20 mL) was added
to the mixture and stirred for an additional 30 min followed
by removal of the acetone. The aqueous mixture was extracted
with CH2Cl2 (3 ¥ 30 mL) and the solvent was removed by rotary
evaporation. Purification of final product was carried out by
column chromatography (CH2Cl2–CH3OH 10:1) yielding 13
in 70%.
◦
EDC/NHS, 35 mL, at a flow rate of 5 mL min-1, at 25 C, and
HBS-P as running buffer. At the same conditions of flow and
temperature, a solution of 11 (2 mM, 150 mL) in HBS-P was
injected to get the maximum immobilization (605 RU), followed
by the injection of ethanolamine (1 M, 35 mL) to block the
possible remaining active sites and another injection of MeOH
(5 mL). On the second flow cell, the activation and blocking
standard procedures were carried out as described above, and it
was left as a negative control of the interaction.
1H-NMR (300 MHz, CDCl3): 6.71 (s, 1H, COOH), 4.13 (s,
2H, H-2), 3.56 (m, 12H, H-3 a H-8), 3.39 (t, 2H, J = 6.90 Hz,
H-9), 3.34 (t, 2H, J = 6.85 Hz, H-14), 1.81 (q, 2H, J = 7.16 Hz,
H-15), 1.52 (m, 2H, H-10), 1.34 (m, 2H, H-11), 1.28 (m, 6H,
H-12,H-13, H-14). 13C-NMR (125 MHz, CDCl3): 173.03 (C-1),
72.07 (C-9), 71.97 (C-8), 70.84 (C-7), 70.74 (C-6), 70.65 (C-5),
70.45 (C-4), 70.34 (C-3), 69.11(C-2), 34.44 (C-16), 33.10 (C-15),
29.70 (C-10), 29.55 (C-12), 29.00 (C-11), 28.41 (C-13), 26.24
(C-14). Analysis calculated for C16H31BrO6: C: 48.12%; H:
7.82%. Found: C: 48.20%; H: 7.81%. ESI-MS: [M + Na] Calcd:
421.1. Found: 421.0.
Immobilization of glycoconjugate 11 on the CM5 chip. The
immobilization of the functionalized GalNAc glycoconjugate
(11) was carried out on commercially available CM5 chips
(Biacore). The carboxylated dextran matrix was activated by the
standard procedure (Biacore), as described above. A solution of
glycoconjugate 11 (150 mL, 2 mM) in the running buffer (HBS-P)
was injected at a flow rate of 5 mL min-1 at 25 ◦C followed by the
injection of ethanolamine (1 M, 35 mL) to block the remaining
active sites. The same activation and blocking procedures were
repeated in a different flow cell and used as negative control for
the interaction studies.
Interaction with Viscum album lectin. The interaction exp◦er-
iments were carried at a constant flow of 5 mL min-1 at 25 C
over GalNAc glycoconjugate 11 immobilized on the Au chip
functionalized with the thiol chains as described above. For
that purpose, 15 mL of different solutions of the VAA in the
running buffer (HBS-P) were injected: 2.2, 5.4, 8.9, 17.9 and
35.7 mM. Sensorgrams obtained for the interaction of VAA with
11, were analyzed and kinetic parameters of the interaction were
determined with BIAEvaluation software (Biacore).
Synthesis of 20-mercapto-3,6,9,12-tetraoxatrieicosan-1-oic
acid (14)
A solution of 13 (230 mg, 0.5 mmol) and thiourea (120 mg,
1.5 mmol) in 30 mL of water was heated under reflux for 6 h.
After the addition of NaOH (600 mg, 15 mmol), the heating
was resumed for an additional period of 6 h. After the mixture
was cooled at room temperature, the pH was adjusted to 1
with concentrated HCl. The resulting solution was extracted
with CH2Cl2 (3 ¥ 20 mL). The combined extracts were dried
on MgSO4 and evaporated to dryness, leaving an oil, which was
purified by column chromatography (CH2Cl2–CH3OH 5:2.) with
a 60% yield.
1H-NMR (300 MHz, CDCl3): 8.23 (s, 1H, COOH), 4.0 (s, 2H,
H-2), 3.56 (m, 12H, H-3 and H-8), 3.38 (t, 2H, J = 6.93 Hz,
H-9), 2.58 (t,2H, J = 7.35 Hz, H-19), 1.58 (m, 4H, H-10,
H-18), 1.20 (m, 14H, H-11, H-12, H-13, H-14, H-15, H-16,
H-17). 13C-NMR (125 MHz, CDCl3): 173.03 (C-1), 72.07 (C-9),
71.97 (C-8), 70.84 (C-7), 70.74 (C-6), 70.65 (C-5), 70.45 (C-4),
70.34 (C-3), 69.11(C-2), 34.44 (C-16), 33.10 (C-15), 29.70 (C-10),
29.55 (C-12), 29.00 (C-11), 28.41 (C-13), 26.24 (C-14). Analysis
calculated for C16H32O6S: C: 54.52%; H: 9.15%; S: 9.10%. Found:
C: 54.53%; H: 9.17%; S: 9.09%. ESI-MS: [M + Na] Calcd: 375.2.
Found: 375.0.
Conclusions
Different glycoconjugates were prepared in the presence of ionic
liquids in order to carry out carbohydrate–protein interaction
studies.
The direct functionalization of carbohydrates (Glc, GalNAc,
lactose) with different linker chains (in the presence of ionic
liquids) was carried out. Despite the poor yields obtained, the
decrease in the number of steps made it a green alternative to
the traditional carbohydrate chemistry.
An amino functionalized GalNAc was directly synthesized
in the presence of ionic liquids and further immobilized on
an alkanethiol-coated surface. The interaction of immobilized
GalNAc derivative with a carbohydrate binding protein was
studied, and kinetic parameters of this procedure were deter-
mined.
Immobilization of glycoconjugates on the SPR chip
Immobilization of thiol chains on the gold surface. For the
immobilization of the thiol linker functionalized with a carboxyl
group, an Au chip (Biacore) was required. Flow was set at 2 mL
min-1 and temperature at 25 ◦C using HBS-P (0.01 M HEPES,
0.15 M NaCl, pH= 7.4) as running buffer. A solution of the
linker 14 was injected into two different flow cells until 721 and
1228 RU (relative response unit) were obtained.
Acknowledgements
This work was supported by Spanish MEC (CTQ2006–
09052/BQU) and EU grant (SOLVSAFE (Advanced safer
solvents for innovative industrial eco-processing, FP-62003-
NMP-SMF-3, proposal 011774–2), an EC Marie Curie Re-
search Training Network grant (MRTN-CT-2005–019561), the
378 | Green Chem., 2009, 11, 373–379
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
research initiative LMUexcellent and MEC-FPU predoctoral
fellowship (AP2003–4820) and US National Institutes of Health
(NIH grant AI065786).
12 N. Jain, A. Kumar, S. Chauhan and S. M. S. Chauhan, Tetrahedron,
2005, 61, 1015–1060.
13 J. M. DeSimone, Science, 2002, 297, 799–803.
14 S. Park and R. J. Kazlauskas, Curr. Opin. Biotechnol., 2003, 14, 32–
37.
15 T.-J. Park, M. We¨ıwer, X. Yuan, S. N. Baytas, E. M. Munoz, S.
Murugesan and R. J. Linhardt, Carbohydr. Res., 2007, 342, 614–620.
16 J.-Y. Huang, M. Lei and Y.-G. Wang, Tetrahedron Lett., 2006, 47,
3047–3050.
References
1 C. R. Bertozzi and L. L. Kiessling, Science, 2001, 291, 2357–2364.
2 A. Varki, Glycobiology, 1993, 3, 97–130.
3 H.-J. Gabius, S. Andre´, H. Kaltner and H.-C. Siebert, Biochim.
Biophys. Acta, 2002, 1572, 165–177.
17 A. Rencurosi, L. Lay, G. Russo, E. Caneva and L. Poletti, Carbohydr.
Res., 2006, 341, 903–908.
4 C. Bies, C.-M. Lehr and J. F. Woodley, Adv. Drug Delivery Rev., 2004,
56, 425–435.
5 K. Larsen, M. B. Thygesen, F. Guillaumie, W. G. T. Willats and K. J.
Jensen, Carbohydr. Res., 2006, 341, 1209–1234.
6 K. C. Nicolaou and H. J. Mitchell, Angew. Chem., Int. Ed., 2001, 40,
1576–1624.
18 K. Toshima, Carbohydr. Res., 2006, 341, 1282–1297.
19 S. Murugesan, N. Karst, T. Islam, J. M. Wiencek and R. J. Linhardt,
Synlett, 2003, 1283–1286.
20 S. Murugesen and R. J. Linhardt, Curr. Org. Synth., 2005, 2, 437–451.
21 L. Poletti, A. Rencurosi, L. Lay and G. Russo, Synlett, 2003, 2297–
2300.
7 D. H. G. Crout and G. Vic, Curr. Opin. Chem. Biol, 1998, 2, 98–111.
8 M. M. Palcic, Curr. Opin. Biotechnol., 1999, 10, 616–624.
9 M. Scigelova, S. Singh and D. H. G. Crout, J. Mol. Catal. B: Enzym.,
1999, 6, 483–494.
22 M. Sharma, G. G. Potti, O. D. Simmons and W. Korytnyk, Carbohydr.
Res., 1987, 162, 41–51.
23 C. Plath, T. Weimar, H. Peters and T. Peters, ChemBioChem, 2006,
7, 1226–1230.
24 M. Jime´nez, J. L. Sa´iz, S. Andre´, H.-J. Gabius and D. Sol´ıs,
Glycobiology, 2005, 15, 1386–1395.
10 S. Hanessian, Preparative Carbohydrate Chemistry, Marcel Dekker,
New York, 1997.
11 S. Hanessian and B. L. Lou, Chem. Rev., 2000, 100, 4443–
4463.
25 S. Andre´, H. Kaltner, T. Furuike, S. I. Nishimura and H.-J. Gabius,
Bioconjug. Chem., 2004, 15, 87–98.
This journal is
The Royal Society of Chemistry 2009
Green Chem., 2009, 11, 373–379 | 379
©