Experimental measurement of carbohydrate-aromatic stacking in water by using a dangling-ended DNA model system

Article abstract of DOI:10.1002/chem.200800335

Protein-carbohydrate recognition is of fundamental importance for a large number of biological processes; carbohydrate-aromatic stacking is a widespread, but poorly understood, structural motif in this recognition. We describe, for the first time, the measurement of carbohydrate-aromatic interactions from their contribution to the stability of a dangling-ended DNA model system. We observe clear differences in the energetics of the interac-tions of several monosaccharides with a benzene moiety depending on the number of hydroxy groups, the stereochemistry, and the presence of a methyl group in the pyranose ring. A fucose-benzene pair is the most stabi-lizing of the studied series (-0.4 Kcal mol^-1) and this interaction can be placed in the same range as other more studied interactions with aromatic residues of proteins, such as Phe-Phe, Phe-Met, or Phe-His. The noncovalent forces involved seem to be dispersion forces and nonconventional hydrogen bonds, whereas hydrophobic effects do not seem to drive the interaction.

Full text of DOI:10.1002/chem.200800335

DOI: 10.1002/chem.200800335
Experimental Measurement of Carbohydrate–Aromatic Stacking in Water by
Using a Dangling-Ended DNA Model System
[
a]
[a]
[a]
[b]
[a]
Juan C. Morales,* JosØ J. Reina, Irene Díaz, Anna AviÇó, Pedro M. Nieto, and
[
b]
Ramón Eritja
Abstract: Protein–carbohydrate recog-
nition is of fundamental importance for
a large number of biological processes;
carbohydrate–aromatic stacking is a
widespread, but poorly understood,
structural motif in this recognition. We
describe, for the first time, the mea-
surement of carbohydrate–aromatic in-
teractions from their contribution to
the stability of a dangling-ended DNA
model system. We observe clear differ-
ences in the energetics of the interac-
tions of several monosaccharides with
a benzene moiety depending on the
number of hydroxy groups, the stereo-
lizing of the studied series (ꢀ0.4 Kcal
ꢀ1
mol ) and this interaction can be
placed in the same range as other more
studied interactions with aromatic resi-
dues of proteins, such as Phe–Phe,
Phe–Met, or Phe–His. The noncovalent
forces involved seem to be dispersion
forces and nonconventional hydrogen
bonds, whereas hydrophobic effects do
not seem to drive the interaction.
chemistry, and the presence of
a
methyl group in the pyranose ring. A
fucose–benzene pair is the most stabi-
Keywords: carbohydrate–aromatic
stacking · carbohydrates · DNA ·
molecular recognition
interactions
· stacking
Introduction
which different noncovalent forces have been suggested to
play a role, namely, CH–p interactions, Van der Waals
[4]
Protein–carbohydrate interactions play a fundamental role
within the living organisms in processes such as apoptosis,
infection, inflammation, or fertilization. Among the most
cited noncovalent forces that participate in carbohydrate–
protein recognition are hydrogen-bonding and ionic interac-
tions. The role of water has also been mentioned to be a de-
terminant driving force in this process. In addition, a wide-
spread, but poorly understood, structural motif found be-
tween carbohydrates and proteins is the stacking of apolar
forces, and hydrophobic effects.
Studies into carbohydrate–aromatic stacking interactions
[1]
[5]
have been approached by using molecular biology tools,
[3b,6]
[7]
NMR spectroscopy,
IR spectroscopy, computational
and model systems. Artificial receptors that
[4a–c,8]
[9]
methods,
incorporate aromatics in their structure and take advantage
of stacking interactions for carbohydrate recognition in
[2]
[10]
apolar solvents have been reported. Recently, a receptor
containing meta-terphenyl units has been shown to bind cel-
[3]
[11]
patches of oligosaccharides with aromatic residues, in
lobiose derivatives selectively in water.
Nevertheless, few experimental binding data on sugar–
arene stacking interactions have been reported and the
quantification of these interactions has been attempted only
in two specific examples. The interaction of a glucose–tyro-
sine pair as a transition-state stabilization was measured to
[
a] Dr. J. C. Morales, J. J. Reina, I. Díaz, Dr. P. M. Nieto
Department of Bioorganic Chemistry
Instituto de Investigaciones Químicas
CSIC—Universidad de Sevilla
4
9 Americo Vespucio, 41092 Sevilla (Spain)
ꢀ
1
be 0.7 kcalmol in the ꢀIV subsite of a 1,3–1,4-b-d-glucan-
Fax : (+34)954-460-565
E-mail: jcmorales@iiq.csic.es
4-glucanohydrolase system, in which cooperativity could be
[12]
[
b] Dr. A. AviÇó, Prof. R. Eritja
playing a role. Recently, a nonbiologically relevant tetraa-
cetylglucose–tryptophan interaction was found to stabilize a
Institute for Research in Biomedicine
Barcelona Science Park, IBMB—CSIC
CIBER—BBN Networking Centre on Bioengineering
Biomaterials and Nanomedicine
ꢀ1 [13]
hairpin-peptide model system by 0.8 kcalmol .
Herein, we describe the first model system in which car-
bohydrate–aromatic stacking interactions are measured and
systematically studied. To do so, we quantified the contribu-
tion of two sugar–arene pairs to the stability of a dangling-
Josep Samitier 1, 08028 Barcelona (Spain)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200800335.
7828
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 7828 – 7835

FULL PAPER
[
15]
ended DNA system. The energetic contributions of mono-
saccharide–benzene pairs in our DNA model system range
sides were placed at the end of a duplex in a “dangling”
position. The resulting stabilization of the duplex was mea-
sured and compared with the duplex lacking the added nu-
cleotides. Our model system consists of a “dangling” ben-
zene nucleoside with a covalently bound monosaccharide
placed at the 5’-ends of a DNA duplex (Scheme 1A). The
sugar is linked through an ethylene glycol spacer to a phos-
phate group located on the primary hydroxy group of the
benzene nucleoside. This allows the pyranose ring to be on
top of the benzene ring because it will be apart from the
phosphate by the same distance as that found in a nucleo-
side between the base and its corresponding phosphate
group (Scheme 1B). At the same time, this spacer allows
enough freedom for the sugar to be either immersed in bulk
water or in contact with the benzene ring. The DNA double
helix acts as an energetic probe that allows the carbohy-
ꢀ
1
from ꢀ0.15 to ꢀ0.40 kcalmol . The observed differences
are due to the influence of the number of hydroxy groups,
the stereochemistry, and the presence of a methyl group in
the pyranose ring. An upfield chemical shift of the NMR
signals of all sugar protons in the oligonucleotide–carbohy-
drate conjugates confirmed the proximity of the carbohy-
drate to the face of the arene ring.
Results and Discussion
Design of the DNA model system: Our design of a dan-
gling-ended DNA model is inspired by studies of aromatic
[14]
stacking in DNA in which natural or non-natural nucleo-
Scheme 1. A) Dangling-ended DNA model system designed to measure carbohydrate–aromatic interactions. B) Enlarged view of the structure of the
dangling-end area of the carbohydrate–oligonucleotide conjugate. C) Carbohydrate–oligonucleotide conjugates included in the study. Oligonucleotide
ꢀ
conjugate 1 is BCGCGCG, in which B is the benzene nucleoside. In conjugates 2–7, DNA seq represents OPO
2
ꢀBCGCGCG.
Chem. Eur. J. 2008, 14, 7828 – 7835
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7829

J. C. Morales et al.
drate–aromatic interaction to be quantified when the carbo-
hydrate–oligonucleotide conjugate is compared with the
control lacking the carbohydrate moiety. In addition, the
conjugate will allow the incorporation of different non-natu-
ral aromatic nucleosides and monosaccharides to quantify
the interaction.
were then prepared by standard phosphoramidite chemistry
(Scheme 2).
Finally, the monosaccharide controls (8–12, Scheme 1C),
used for structural comparison with the carbohydrate–oligo-
nucleotide conjugates, were prepared in high yield by depro-
tection of the corresponding acetylated derivatives (14, 17,
[19]
2
0, 23, and 24) by classical methanolysis.
Synthesis of the carbohydrate–oligonucleotide conjugates:
Preparation of the carbohydrate–oligonucleotide conjugates
Thermodynamic parameters of the carbohydrate–DNA
model systems: The measured thermodynamic parameters
for the carbohydrate–oligonucleotide conjugates under
physiological conditions (140 mm NaCl, 20 mm KCl, 10 mm
phosphate buffer, pH 7) are illustrated in Table 1. The ther-
(COCs, 1–7; Scheme 1C) was carried out by standard solid-
phase oligonucleotide automatic synthesis by using the
[16]
DMT-phosphoramidite benzene nucleoside and the corre-
sponding carbohydrate phosphoramidite (DMT: 4,4’di-
A
C
H
T
R
E
U
N
G
methoxytriphenylmethyl). b-d-Glucose, b-d-2-deoxyglucose,
a-d-2-deoxyglucose, b-d-galactose, and b-l-fucose were the
monosaccharides selected to study how the interaction with
a benzene ring is affected by different factors, such as the
number of hydroxy groups, the anomeric configuration, the
stereochemistry of the pyranose ring, and the presence of a
methyl group.
Table 1. Carbohydrate–aromatic stacking, as measured in a DNA-duplex
context.
Dangling
moiety
T
[8C]
m
ꢀDH8
ꢀDS8
ꢀDG837
DDG8
stacking
[a]
[b]
[c]
[c]
[c]
[d]
[Kcal]
[eu]
[Kcal]
–
51.1
50.0
51.6
50.4
51.8
50.7
58.2
56.1
63.1
59.1
67.4
61.0
156
150
171
159
185
166
9.7
9.4
9.9
9.7
10.1
9.7
–
–
HO C2
ꢀ
The synthesis of the carbohydrate phosphoramidites was
carried out in two steps. First, in the case of glucose, galac-
tose, and fucose, attachment of the ethylene glycol spacer to
the carbohydrate was carried out by classical glycosidation
chemistry with the peracetylated bromo monosaccharides to
obtain only the corresponding b isomers (Scheme 2). 2-Hy-
droxyethyl-2,3,4,6-tetra-O-acetyl-b-d-glucopyranoside (14)
and 2-hydroxyethyl-2,3,4,6-tetra-O-acetyl-b-d-galactopyra-
b-d-GlcꢀC2
b-d-GalꢀC2
b-l-FucꢀC2
ꢀ0.25
ꢀ0.15
ꢀ0.40
ꢀ0.15
b-d-2-deoxy-
GlcꢀC2
a-d-2-deoxy-
51.0
60.9
165
9.8
ꢀ0.20
GlcꢀC2
[a] Core sequence is BCGCGCG, in which B is the benzene nucleoside.
ꢀ
C2 indicates CH
2
ꢀCH
2
ꢀOPO
2
. [b] Melting temperature, T
m
, was mea-
[17]
sured at 10 mm COC concentration. [c] Estimated errors are ꢁ8% in
noside (17) were prepared as described previously, and 2-
hydroxyethyl-2,3,4-tri-O-acetyl-b-l-fucopyranoside (20) was
synthesized following the same methodology. The 2-deoxy-
glucose derivatives were prepared by electrophilic addition
to the corresponding acetyl-protected glucal and subsequent
separation of the a and b isomers by column chromatogra-
ꢂ
DH8, ꢁ8% in DS8, and ꢁ11% in DG . [d] DDG8 stacking per carbohy-
3
7
drate–benzene pair=(DG8 conjugate with carbohydrateꢀ(DG8 conjugate
with spacer)/2)).
modynamic parameters were calculated from the average
values obtained from melting curve fitting and linear plots
[
18]
phy.
The corresponding carbohydrate phosphoramidites
[20]
of 1/Tm versus ln[conjugate]. Calculation of the contribu-
tion of the carbohydrate–aromatic interaction to DNA stabi-
lization was carried out by comparison of the corresponding
carbohydrate–DNA conjugate with the control that lacks
the pyranose ring but still maintains the ethylene glycol
spacer (second row in Table 1). In fact, the presence of the
linker destabilizes the DNA conjugate, most probably due
to the entropy cost because of its high mobility, and this
factor must be taken into account.
The thermodynamic data show that all of the conjugates
that contain a carbohydrate–aromatic moiety show higher
stability than the control conjugate 2, most probably due to
the affinity of the monosaccharides to stack with the ben-
zene ring. It is important to note that the interaction is en-
thalpy driven, which seems to indicate the minor signifi-
cance of classical hydrophobic effects in the interaction. Ac-
tually, the glucose–benzene interaction showed a stabiliza-
ꢀ
1
tion of ꢀ0.25 kcalmol in our dangling-ended DNA system.
Next, we investigated the effect of the number of hydroxy
groups in the pyranose ring on the interaction. b-Glucose
stacks more strongly on the benzene ring than b-2-deoxyglu-
cose, a more hydrophobic sugar that lacks the hydroxy
Scheme 2. Synthetic routes used to prepare the carbohydrate phosphor-
amidite derivatives. a) Ethylene glycol, Ag CO CH Cl RT 18 h,
8–62%; b) 2-cyanoethyl-N,N’-diisopropylamino-chlorophosphoramidite,
N,N-diisopropylethylamine (DIEA), CH Cl , RT, 1.5 h, 74–92%; c) ethyl-
ene glycol, PPh ·HBr, CH Cl , RT , 7 h, 5% of23 and 20% of 24.
2
3
,
2
2
,
,
2
2
2
3
2
2
7830
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 7828 – 7835

Carbohydrate–Aromatic Interactions
FULL PAPER
group at the 2-position. This result seems to support the fact
that dispersion forces and nonconventional hydrogen bonds
are the main noncovalent forces involved in this interaction,
because the proton at the 2-position may be less polarized
in b-2-deoxyglucose and, therefore, the interaction may lose
one of the nonconventional hydrogen bonds. When 2-de-
oxyglucose is attached in the a-anomeric configuration, the
stacking is similar to that found for b-2-deoxyglucose. This
result could indicate a similar stacking geometry with the ar-
omatic moiety due to the high flexibility of the spacer, but
the different chemical shifts of the NMR signals of the sugar
protons seem to point to different approaches to the ben-
zene ring (see the structural features section below).
The effect of the stereochemistry of the carbohydrate on
the interaction was examined by comparison of b-glucose
and b-galactose in their interactions with benzene. Surpris-
ingly, the results show that stacking is stronger for glucose
than for galactose, although a more extended apolar area in
galactose could lead to the opposite result being predicted,
especially if dispersion forces are involved in the interaction.
A possible reason for this result could be that the ethylene
glycol spacer is not long enough to allow the galactose
moiety to adopt a more apical geometry and interact with
the arene ring through the apolar patch formed by its H3–5
protons. Actually, the chemical-induced shifts of the NMR
signals for the protons, obtained by comparing the sugar
unit in conjugate 4 with the b-galactose control 9, indicate a
higher population of parallel-stacking geometry between the
carbohydrate and the aromatic ring (see the structural fea-
tures section below). In addition, it should be pointed out
that the monosaccharides can approach the benzene unit
through both faces of the pyranose ring due to the high flex-
ibility of the linker. A stacking interaction with a parallel
geometry between the b face of galactose and the aromatic
ring is not possible due to the axial configuration of the hy-
droxy group in the 4-position and, consequently, the poten-
tial contribution to the stability due to contacts through that
face is lost in galactose but not in glucose (Figure 1).
of a hydrogen bond between the 4-OH group of glucose and
the arene moiety cannot be ruled out; this interaction is not
possible in galactose without disrupting the parallel stacking
with benzene, due to its intrinsic geometry. Similar reason-
ing should also be considered when comparing b-glucose
and b-2-deoxyglucose. Actually, an H-bond between a glu-
cose 4-OH group and toluene has been observed in the gas
[4a]
[7]
phase by IR spectroscopy, and similar H-bonds have been
observed in the X-ray crystal structures of proteins, such as
between the hydroxy group of Thr13 and the center of the
[21]
arene ring of Tyr6 in a class m glutathione transferase.
Finally, we studied the relevance of a methyl group in the
structure of the pyranose ring. The contribution to DNA
stabilization of the fucose–benzene interaction is ꢀ0.40 kcal
ꢀ
1
mol , which makes this pair the most stable of the series.
This carbohydrate–aromatic interaction is in the same range
as other more studied interactions with aromatic residues of
[22]
proteins, such as Phe–Phe, Phe–Met, or Phe–His.
The
presence of the methyl group considerably increases the
apolar surface of the carbohydrate and, therefore, dispersion
forces involved in the interaction may be favored.
Structural features of the carbohydrate–aromatic stacking:
1D and 2D NMR experiments were carried out to study the
geometry of the dangling moiety of the carbohydrate–oligo-
nucleotide conjugates. We compared the chemical shifts of
the resonances of the sugar protons in the carbohydrate–oli-
gonucleotide conjugates (3–7) with those of the correspond-
ing monosaccharide controls (8–12) that possess the linker
moiety but not the benzene nucleoside or the DNA strand.
The comparison of the chemical shifts of the DNA resonan-
ces along the series of carbohydrate–oligonucleotide conju-
gates indicates no structural changes in the nucleic acid frag-
ment. We observed that all of the proton resonances of the
pyranose ring in the carbohydrate–oligonucleotide conju-
gates exhibited upfield shifting relative to those of the corre-
sponding carbohydrate controls (Table 2). Comparison of
1
the H NMR spectrum of carbohydrate conjugate 5, which
It is important to note that the energetic data obtained
refer to differences in stability between carbohydrate–DNA
conjugates but these energetic values are an average of the
different conformations available for each carbohydrate–
DNA conjugate. Furthermore, the possibility of formation
contains the b-l-fucose unit, with that of the b-l-fucose con-
trol 10 clearly shows the upfield shift observed for all of the
protons in the carbohydrate moiety (Figure 2). These data
seem to confirm that all of the monosaccharides in the study
1
Table 2. H NMR chemically induced shift differences between the sugar
residues of carbohydrate–oligonucleotide conjugates 3–7 and the corre-
[a]
sponding monosaccharide controls 8–12 in D
2
O.
b-d-Glc b-d-Gal b-l-Fuc b-d-2-deoxy-Glc a-d-2-deoxy-Glc
H1
H2
H2b
H3
H4
H5
H6
ꢀ0.112
ꢀ0.056
–
ꢀ0.074
ꢀ0.048
ꢀ0.120
ꢀ0.069
ꢀ0.102
ꢀ0.059
–
ꢀ0.087
ꢀ0.051
ꢀ0.116
ꢀ0.058
ꢀ0.050
–
ꢀ0.144 ꢀ0.190
ꢀ0.084 ꢀ0.163
ꢀ0.094
ꢀ0.100
ꢀ0.087
ꢀ0.071
ꢀ0.074
ꢀ0.067
–
ꢀ0.127
ꢀ0.082 ꢀ0.149
ꢀ0.093 ꢀ0.086
[
[
b]
b]
ꢀ0.178
–
–
Figure 1. Model of the dangling-end area of carbohydrate–DNA conju-
gate 4, which contains a b-galactose unit. A) Possible stacking geometry
of b-galactose on top of the benzene ring through the a face of the sugar.
B) Possible stacking geometry of b-galactose on top of the benzene ring
through the b face of the sugar, with the steric clash between the 4-OH
group and the arene shown.
[b]
ꢀ
–
H6b ꢀ0.086
–
ꢀ0.118
ꢀ0.092
CH
3
–
ꢀ0.122
–
–
[a] Chemical-shift differences are given in ppm. [b] Unassigned proton
due to overlapping resonances.
Chem. Eur. J. 2008, 14, 7828 – 7835
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7831

J. C. Morales et al.
Figure 3. A) Upfield shifting of carbohydrate protons due to the proximi-
ty to the face of the aromatic ring. Differences are calculated between
the conjugate containing b-d-glucose on the sequence BCGCGCG, 3,
and 2-hydroxyethyl-b-d-glucopyranoside (8). B) Model of possible stack-
ing geometry of b-glucose on top of the benzene ring through the a face
of the sugar. C) Model of possible stacking geometry of b-glucose on top
of the benzene ring through the b face of the sugar.
1
Figure 2. H NMR spectra of carbohydrate–oligonucleotide conjugate 5
bottom) and the corresponding monosaccharide control 10 (top) in D
(
2
O
showing the chemical-shift differences for the protons of the b-l-fucose
residue.
prefer to stay close to the arene ring rather than just being
immersed in the bulk water molecules. The proximity of the
monosaccharides to the face of the arene ring could explain
the extra stability found in the carbohydrate–oligonucleotide
conjugates due to the presence of a potential carbohydrate–
aromatic stacking interaction.
is higher for b-l-fucose than for b-2-deoxyglucose during
the interaction with benzene. This points to the fact that
closer contact of b-2-deoxyglucose with the benzene ring
may be possible due to the lack of a hydroxy group at the 2-
position but this does not translate into a more favorable
A
H
R
U
G
Moreover, the monosaccharides can approach the ben-
zene unit through both faces of the pyranose ring, probably
due to the high flexibility of the linker. It is important to
note that the chemical-induced shifts show a similar trend in
the four sugars linked to the DNA conjugate through a b-
anomeric configuration. Upfield shift is more significant on
the a face of the pyranose moiety, where protons H1, H3,
and H5 are pointing down, as can be observed, for example,
in b-glucose (Figure 3). This seems to indicate a more favor-
able interaction of the b-linked monosaccharides with the
benzene ring through their a face. Actually, a similar geo-
metric tendency has been found by Waters and co-workers
between b-glucose and tryptophan in a b-hairpin-peptide
Conclusion
Our results indicate that a dangling-ended DNA model
system is a useful tool to measure carbohydrate–aromatic
stacking interactions. Clear differences have been observed
in the energetics of the interaction depending on the
number of hydroxy groups, the stereochemistry, and the
presence of a methyl group in the pyranose ring. The nonco-
valent forces involved seem to be dispersion forces and non-
conventional hydrogen bonds between the protons of the
pyranose ring and the benzene ring, whereas classical hydro-
phobic effects do not seem to drive the interaction. Investi-
gations into the influence of other aromatic rings in the in-
teraction and NMR spectroscopy and molecular modeling
studies are in progress to obtain a better understanding of
carbohydrate–aromatic stacking interactions by using our
dangling-ended DNA model system.
[13]
model.
In contrast to the results for sugars linked through a b-
anomeric configuration to the DNA conjugate, the chemi-
cal-induced shifts for a-2-deoxyglucose show similar magni-
tude on both sides of the pyranose moiety. This could sug-
gest that this sugar approaches the arene ring in a different
way to the b-2-deoxyglucose and, therefore, different confor-
mations are present during stacking with the aromatic
moiety.
Finally, the b-2-deoxyglucose unit shows the largest reso-
nance shifts of the series, even larger than those for b-l-
fucose. This result is surprising since the stability measured
in our DNA model for the carbohydrate–aromatic stacking
Experimental Section
General remarks: All chemicals were obtained from Aldrich Chemicals
and used without further purification, unless otherwise noted. All reac-
tions were monitored by TLC on precoated silica gel 60 F254 plates
7832
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 7828 – 7835

Carbohydrate–Aromatic Interactions
FULL PAPER
3
(
(
Merck), with detection by heating with Mostain (10% H
2
SO
4
(500 mL),
11.7 Hz, 1H; H6b), 3.82–3.59 (m, 5H; H5, 2”H7, 2”H8), 2.74 (t, J-
3
2
ACHTREUNG
NH Mo O (25 g), Ce(SO ·4H O (1 g)). Products were puri-
4
)
6
7
O
24·4H
2
A
H
R
U
G
4
)
2
2
A
H
R
U
G
A
H
R
U
2
fied by flash chromatography with Merck silica gel 60 (200–400 mesh).
High-resolution FAB (+) mass spectral analyses were obtained on a
J
A
C
H
T
R
E
U
N
G
(H,H)=
eq
3
1
3
11.0 Hz, 1H; H2 ); C NMR (75 MHz, CDCl ): b isomer 23: d=170.6,
ax
3
A
C
H
T
R
E
U
N
G
Micromass AutoSpec-Q spectrometer.
170.2, 169.7, 100.1, 72.7, 71.9, 70.3, 68.9, 62.4, 62.0, 36.0, 20.8, 20.6 ppm;
+
FAB HRMS: b isomer 23: m/z: calcd for C14
22 9
H O : 334.1264 [M] ;
NMR spectra were recorded on Bruker AVANCE 300, ARX 400, or Ad-
vance DRX 500 MHz spectrometers (300 or 400 MHz ( H), 75 or 100
+
1
1
found: 357.1165 [M+Na] . H NMR (300 MHz, CDCl ): a isomer 24: d=
3
3
1
3
5.25 (ddd,
H4), 4.94 (dd,
J(H,H)=5.3, 9.5, 11.5 Hz, 1H; H3) 4.96–4.89 (m, 2H; H1,
A
T
E
N
(
C); 500 MHz for the carbohydrate controls and for the carbohydrate–
DNA conjugate samples) at room temperature for solutions in CDCl
O, or CD OD. Chemical shifts were referenced to the solvent signal.
D experiments (COSY, TOCSY, ROESY, and HMQC) were done when
3
2
3
J
A
H
E
N
J(H,H)=12.4 Hz, 1H; H6a), 4.00 (dd, J-
A
H
R
U
G
3
,
2
ACHTREUNG
A
H
R
U
G
J
D
2
2
3
3
3
2
ACHTREUNG
necessary to assign resonances of the oligosaccharides or the carbohy-
drate–DNA conjugates. The carbohydrate–DNA samples for NMR spec-
troscopy were purified by HPLC and then ion-exchanged with Dowex
2.22 (dd, J
A
H
R
U
G
(H,H)=12.9 Hz, 1H; H2eq), 2.02, 1.97, 1.95 (3s,
3
2
ACHTREUNG
9H; 3”OCOCH
3
J(H,H)=3.6, J(H,H)=12.9 Hz;
A
H
R
U
G
1
3
3
5
D
D
0W resin. Finally, samples were lyophilized to dryness three times from
+
O to deuterate all exchangeable protons and dissolved in 99.999%
O just prior to recording of the NMR spectra. Sample concentration
HRMS: m/z: a isomer 24: calcd for C14
357.1167 [M+Na] .
H
22
O
9
: 334.1264 [M] ; found:
2
+
2
ranged from 0.7 to 2.8 mm depending on the conjugate. Chemical shifts
d) are given in ppm with respect to the acetone signal, which was used
b-Cyanoethoxy-b-(2,3,4,6-tetra-O-acetyl-b-d-glucopyranosyl)ethoxy-diiso-
propylamine phosphine (15): DIEA (0.49 mL, 3.72 mmol) and 2-cya-
(
1
as an external reference (volume of 1 mL). In all experiments, the H car-
rier frequency was kept at the water resonance. Data were processed by
using the manufacturer software, raw data were multiplied by a shifted
exponential window function prior to Fourier transformation, and the
baseline was corrected by using polynomial fitting.
noethyl-N,N’-diisopropylamino-chlorophosphoramidite
(170 mL,
0.76 mmoL) were added to a solution of 2-hydroxyethyl-2,3,4,6-tetra-O-
[
17]
acetyl-b-d-glucopyranoside (14; 0.2 g, 0.51 mmol) in anhydrous CH Cl
2
2
(4 mL) at room temperature under an argon atmosphere. After 1.5 h, the
solution was diluted with EtOAc (20 mL), and the organic phase was
Melting curves for the DNA conjugates were measured in a Perkin–
Elmer Lambda 12 UV/Vis spectrophotometer at 280 nm while the tem-
4
washed with brine (3”25 mL), dried over anhydrous MgSO , and fil-
tered; the solvent was then evaporated to dryness. The product was puri-
fied by silica gel column chromatography by using Hex/EtOAc (2:1 with
ꢀ1
perature was raised from 20 to 808C at a rate of 1.08Cmin . Curve fits
2
ꢀ6
were excellent, with c values of 10 or better, and the Vanꢁt Hoff linear
3
2% of NEt ) as the eluent to give compound 15 (215 mg, 74%) as a
2
1
3
fits were quite good (r =0.98) for all oligonucleotides. Differences of less
3
syrup. H NMR (300 MHz, CDCl ): mixture of isomers: d=5.11 (t, J-
3
3
than 10% were observed between thermodynamic parameters as deter-
A
H
R
U
G
J(H,H)=9.6 Hz, 1H; H4), 4.89 (t, J-
A
H
R
U
3
ACHTREUNG
3
mined by 1/T
DG errors were calculated as described previously.
Preparation of carbohydrate alcohols and phosphoramidites
-Hydroxyethyl-2,3,4-tri-O-acetyl-b-l-fucopyranoside (20): Na
spatula tips) and ethylene glycol (2.0 mL, 35.86 mmoL, 18 equiv) were
added to solution of tri-O-acetylfucopyranosyl bromide (0.68 g,
.93 mmol) in anhydrous CH Cl (40 mL). After the mixture had been
stirred at room temperature for 15 min under an argon atmosphere,
Ag CO (9.00 g, 16.31 mmol, 8.5 equiv) was added to the reaction mix-
ture and stirring was continued for 18 h (TLC: hexane (Hex)/EtOAc
:3). The mixture was filtered, diluted with CH Cl (40 mL), washed with
water (3”100 mL), and dried over anhydrous MgSO ; the solvent was
m
versus ln[conjugate] plots and curve fittings. DH, DS, and
A
H
R
U
G
[
20]
2
ACHTREUNG
A
H
R
U
G
3
2
1
2
2
CN, 2”CHisopropyl), 2.56 (m,
2
2
3
3
),
(H,H)=6.2 Hz, 12H; 4”CH3isopropyl); C NMR (75 MHz,
2
2
SO
4
(7
13
AHCTREUNG
CDCl
3
a
1
1
2
2
6
2
2.3, 62.2, 62.1, 61.7, 60.1, 58.4, 58.3, 58.1, 58.0, 42.9, 42.9, 42.8, 42.7, 24.4,
4.4, 24.3, 24.3, 20.8, 20.5, 20.4, 20.2, 20.2, 20.1, 20.1 ppm; FAB HRMS:
2
3
+
+
m/z: calcd for C25
41 2
H N O12P: 592.2397 [M] ; found: 615.2294 [M+Na] .
b-Cyanoethoxy-b-(2,3,4,6-tetra-O-acetyl-b-d-galactopyranosyl)ethoxy-di-
isopropylamine phosphine (18): DIEA (0.96 mL, 3.72 mmol) and 2-cya-
2
2
2
4
noethyl-N,N’-diisopropylamino-chlorophosphoramidite
(332 mL,
then evaporated under reduced pressure. The product was purified by
silica gel column chromatography by using Hex/EtOAc (2:3) as the
1
.49 mmoL) were added to a solution of 2-hydroxyethyl-2,3,4,6-tetra-O-
1
[17]
eluent to give compound 20 (0.4 g, 62%) as a syrup. H NMR (300 MHz,
acetyl-b-d-galactopyranoside (17;
2 2
0.39 g, 0.99 mmol) in dry CH Cl
3
3
ACHTREUNG
CDCl
3
): d=5.19 (dd, J
A
C
H
T
R
E
U
N
G
(H,H)=0.5, 3.0 Hz, 1H; H4), 5.14 (dd, J
(4 mL) at room temperature under an argon atmosphere. After 1.5 h, the
solution was diluted with EtOAc (20 mL), and the organic phase was
washed with brine (3”25 mL), dried over anhydrous MgSO , and fil-
4
tered; the solvent was then evaporated to dryness. The product was puri-
fied by silica gel column chromatography by using Hex/EtOAc (2:3 with
3
ACHTREUNG
7
.8, 10.5 Hz, 1H; H2), 4.98 (dd, J
3
J
A
C
H
T
R
E
U
N
G
3
(
brs, 1H; OH), 2.14, 2.04, 1.94 (3s, 9H; 3”OCOCH
3
), 1.19 ppm (d, J-
): d=170.4, 169.9,
69.6, 101.2, 71.9, 70.9, 69.9, 68.9, 68.7, 61.4, 20.5, 20.4, 20.3, 15.7 ppm;
1
3
A
C
H
T
R
E
U
N
G
3 3
(H,H)=6.3 Hz, 3H; CH ); C NMR (75 MHz, CDCl
1
3
2% of NEt ) as the eluent to give compound 18 (368 mg, 84%) as a
+
1
3
FAB HRMS: m/z: calcd for C14
H
22
O
9
: 334.1264 [M] ; found: 357.1167
syrup. H NMR (300 MHz, CDCl
3
): mixture of isomers: d=5.32 (d, J-
+
3
ACHTREUNG
[
M+Na] .
A
H
R
U
G
3
3
ACHTREUNG
4
.95 (dt,
J
A
H
R
U
G
2
2
-Hydroxyethyl-3,4,6-tri-O-acetyl-2-deoxy-a,b-d-glucopyranoside
4): A solution of PPh ·HBr (0.71 g, 2.07 mmoL) and ethylene glycol in
Cl (25 mL) was added to a solution of tri-O-acetyl-glucal (7.5 g,
7.5 mmoL) in dry CH Cl (8 mL). The reaction mixture was stirred at
08C. After 24 h, more PPh ·HBr (2.13 g, 6.21 mmoL) was added to the
(23,
H1), 4.15–4.01 (m, 2H; 2”H6), 3.95–3.45 (m, 9H; H5, 2”H7, 2”H8,
3
3
OCH
OCH
2
CH
2
CN,
2”CHisopropyl),
2.59
(t,
J
A
H
R
N
(H,H)=6.1 Hz,
2H;
),
(H,H)=6.5 Hz, 12H; 4”CH3isopropyl); C NMR (75 MHz,
): mixture of isomers: d=170.2, 170.1, 170.0, 169.3, 117.6, 101.1,
01.0, 70.7, 70.4, 69.4, 69.3, 69.1, 69.0, 68.6, 68.5, 66.9, 66.8, 62.4, 62.3,
dry CH
2
4
2
2
2
CH
2
CN), 2.09, 2.01, 2.00, 1.94, 1.91 (5s, 12H; 4”OCOCH
3
2
2
3
ACHTREUNG
13
1
.12 ppm (2d, J
3
CDCl
1
3
reaction mixture, and stirring was continued for 7 h. The reaction was
then stopped by addition of CH Cl (120 mL) and water (150 mL). The
organic phase was washed with saturated NaHCO (3”150 mL), dried
over anhydrous Na SO , and filtered; the solvent was then evaporated to
2
2
6
2.2, 62.1, 61.1, 61.0, 58.4, 58.3, 58.1, 58.1, 43.0, 42.9, 42.8, 24.5, 24.4, 24.4,
3
2
4.3, 20.7, 20.5, 20.5, 20.4, 20.3, 20.3, 20.2, 20.2 ppm; FAB HRMS: m/z:
2
4
+
+
calcd for C25
41 2
H N O12P: 592.2397 [M] ; found: 592.2397 [M+Na] .
dryness. The final product was purified by silica gel column chromatogra-
phy by using a solvent gradient (Hex/EtOAc 5:4, 1:1, 4:5, 2:3) to obtain
b-Cyanoethoxy-b-(2,3,4,-tri-O-acetyl-l-fucopyranosyl)ethoxy-diisopropyl-
amine phosphine (21): DIEA (0.51 mL, 2.92 mmol) and 2-cyanoethyl-
b isomer 23 (460 mg, 5%), a isomer 24 (1.84 g, 20%), and a fraction mix-
1
ture of 23 and 24 (1.38 g, 15%). H NMR (300 MHz, CDCl
3
): b isomer
N,N’-diisopropylamino-chlorophosphoramidite
(260 mL, 1.18 mmoL)
3
ACHTREUNG
2
3: d=5.03–4.86 (m, 2H; H3, H4), 4.58 (d,
J
(H,H)=9.6 Hz, 1H; H1),
were added to a solution of 2-hydroxyethyl-2,3,4-tri-O-acetyl-b-l-fucopyr-
3
2
2
ACHTREUNG
4
.18 (dd,
J
A
C
H
T
R
E
U
N
G
(H,H)=5.6,
J
A
H
R
U
G
J
(H,H)=
anoside (20; 0.26 g, 0.78 mmol) in dry CH
2
Cl
2
(4 mL) at room tempera-
Chem. Eur. J. 2008, 14, 7828 – 7835
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7833

J. C. Morales et al.
1
3
ACHTREUNG
ture under an argon atmosphere. After 1.5 h, the solution was diluted
with EtOAc (20 mL), and the organic phase was washed with brine (3”
white solid. H NMR (500 MHz, D
2
3
ACHTREUNG
O): d=4.49 (d, J
H1), 3.98 (m, 1H; H7a), 3.91 (dd, J
(m, 3H; H7b, 2”H8) 3.72 (dd, J
J(H,H)=9.2 Hz, 1H; H3), 3.45 (m, 1H; H5), 3.39 (t, J(H,H)=9.4 Hz,
1H; H4), 3.31 ppm (dd,
(100 MHz, CD OD): d=103.1, 76.6, 76.5, 73.8, 71.0, 70.2, 61.3, 60.9 ppm;
ESIMS: m/z: calcd for C25
3
ACHTREUNG
2
5 mL), dried over anhydrous MgSO
evaporated to dryness. The product was purified by silica gel column
chromatography by using Hex/EtOAc (2:1 with 2% NEt ) as the eluent
to give compound 21 as a syrup (200 mg, 48%). H NMR (300 MHz,
4
, and filtered; the solvent was then
3
3
ACHTREUNG
A
H
R
U
G
3
13
J(H,H)=8.0, 9.4 Hz, 1H; H6b); C NMR
A
H
R
U
G
3
1
3
3
+
CDCl
3
): mixture of isomers: d=5.18 (d, J
A
H
R
U
G
H
41
N
2
O
12P: 592.2397 [M] ; found: 592.2397
3
3
+
(
ddd,
J
A
C
H
T
R
E
U
N
G
(H,H)=2.4, 7.9, 10.4 Hz, 1H; H2), 4.96 (dt,
[M+Na] .
3
1
0.4 Hz, 1H; H3), 4.51 (m, J
CH
A
H
R
U
G
1
-Hydroxyethyl-b-d-galactopyranoside (9): NaOMe (27 mg, 0.50 mmol)
H5, 2”H7, 2”H8, OCH
H; OCH CH
(H,H)=6.3, 3H; CH
2
2
CN, 2”CHisopropyl), 2.60 (t, J
was added to a solution of 2-hydroxyethyl-2,3,4,6-tetra-O-acetyl-b-d-gal-
actopyranoside (17; 100 mg, 0.25 mmol) in MeOH (3 mL), and the reac-
tion was stirred at room temperature for 30 min. The solution was then
neutralized with Amberlite IRA-120 resin and filtered, and the solvent
was eliminated in vacuo to obtain compound 9 (56 mg, quant yield) as a
1
2
2 3
CN), 2.13, 2.03, 2.02, 1.93 (4s, 9H; 3”OCOCH
3
3
ACHTREUNG
J
A
C
H
T
R
E
U
N
G
3
), 1.14 ppm (2d,
J
1
3
CH3isopropyl); C NMR (75 MHz, CDCl
3
1
1
3
ACHTREUNG
6
8.9, 68.8, 68.7, 62.5, 62.4, 62.3, 62.2, 58.5, 58.2, 43.0, 43.0, 42.9, 42.8, 24.5,
white solid. H NMR (500 MHz, D O): d=4.42 (d, J
(H,H)=7.8 Hz, 1H;
2
3
2
4.5, 24.4, 20.8, 20.8, 20.6, 20.5, 20.3, 20.3, 15.9 ppm; FAB HRMS: m/z:
H1), 4.05 (m, 1H; H7a), 3.92 (d,
(m, 4H; H6a, H7b, 2”H8), 3.74 (dd, J(H,H)=4.2, 11.7 Hz, 1H; H6b),
3.70 (dd, J
1H; H3), 3.55 ppm (dd,
(100 MHz, CD OD): d=103.8, 75.3, 73.5, 71.3, 70.0, 68.9, 61.1, 61.0 ppm;
ESIMS: m/z: calcd for C
J(H,H)=3.4 Hz, 1H; H4), 3.81–3.75
A
H
R
U
G
+
+
3
ACHTREUNG
calcd for C23
H
41
N
2
O
10P: 534.2342 [M] ; found: 557.2240 [M+Na] .
3
3
ACHTREUNG
b-Cyanoethoxy-b-(3,4,6-tri-O-acetyl-2-deoxy-b-d-glucopyranosyl)ethoxy-
diisopropylamine phosphine (25): DIEA (0.39 mL, 2.24 mmol) and 2-cya-
A
H
R
U
G
3
13
J(H,H)=7.9, 10.0 Hz, 1H; H2); C NMR
A
H
R
U
G
noethyl-N,N’-diisopropylamino-chlorophosphoramidite
(200 mL,
.9 mmoL) were added to a solution of 2-hydroxyethyl-2,3,4-tri-O-acetyl-
-deoxy-b-d-glucopyranoside (23; 0.2 g, 0.6 mmol) in anhydrous CH Cl
3
+
+
0
2
8 16 7
H O : 224.1 [M] ; found: 247.1 [M+Na] .
2
2
1-Hydroxyethyl-b-l-fucopyranoside (10): NaOMe (32 mg, 0.60 mmol)
was added to a solution of 2-hydroxyethyl-2,3,4-tri-O-acetyl-b-l-fucopyr-
anoside (20; 100 mg, 0.30 mmol) in MeOH (3 mL), and the reaction was
stirred at room temperature for 30 min. The solution was then neutral-
ized with Amberlite IRA-120 resin and filtered, and the solvent was
eliminated in vacuo to obtain compound 10 (53 mg, 86%) as a white
solid. H NMR (500 MHz, D O): d=4.90 (d, J(H,H)=7.9 Hz, 1H; H1),
3.98 (m, 1H; H7a), 3.90 (dd,
1H; H5), 3.76 (m, 2H; 2”H8) 3.74 (m, 1H; H4), 3.64 (dd, J
10.0 Hz, 1H; H3), 3.50 (dd, J
(
4 mL) at room temperature under an argon atmosphere. After 1.5 h, the
solution was diluted with EtOAc (20 mL), and the organic phase was
washed with brine (3”25 mL), dried over anhydrous MgSO , and fil-
4
tered; the solvent was then evaporated to dryness. The product was puri-
fied by silica gel column chromatography by using Hex/EtOAc (2:1 with
1
3
ACHTREUNG
2
% of NEt
3
) as the eluent to give compound 25 (284 mg, 89%) as a
2
3
1
syrup. H NMR (300 MHz, CDCl ): mixture of isomers: d=4.98–4.79 (m,
3
J(H,H)=6.5, 7.6 Hz, 1H; H7b) 3.77(m,
A
T
U
G
3
3
3
ACHTREUNG
2
H; H3, H4), 4.56 (d, J
A
C
H
T
R
E
U
N
G
(H,H)=9.4 Hz, 1H; H1), 4.19 (dd, J
A
H
R
U
G
2
3
ACHTREUNG
J
A
C
H
T
R
E
U
N
G
(H,H)=12.1 Hz, 1H; H6a), 4.06–3.92 (m, 1H; H6b), 3.92–3.40 (m, 9H;
3
3
13
H5, 2”H7, 2”H8, OCH
H; OCH CH
H2eq), 1.97, 1.92, 1.91 (3s, 9H; 3”OCOCH
2
CH
2
CN, 2”CHisopropyl), 2.56 (t, J
A
H
R
U
G
A
H
R
U
G
3 3
J(H,H)=6.5 Hz, 1H; CH ); C NMR (100 MHz, CD OD): d=103.6,
73.6, 71.6, 71.0, 70.8, 70.6, 61.0, 15.3 ppm; ESIMS: m/z: calcd for
C H O : 208.1 [M] ; found: 231.1 [M+Na] .
8 16 7
3
2
ACHTREUNG
2
2
2
CN), 2.22 (dd,
J
A
H
R
U
G
J
(H,H)=10.0 Hz, 1H;
+
+
3
1
3
H2ax); C NMR (75 MHz, CDCl
3
1
0
2
-Hydroxyethyl-2-deoxy-b-d-glucopyranoside (11): NaOMe (16 mg,
.30 mmol) was added to a solution of 2-hydroxyethyl-2,3,4-tri-O-acetyl-
-deoxy-b-d-glucopyranoside (23; 50 mg, 0.15 mmol) in MeOH (2 mL),
and the reaction was stirred at room temperature for 30 min. The solu-
tion was then neutralized with Amberlite IRA-120 resin and filtered, and
the solvent was eliminated in vacuo to obtain compound 11 (31 mg,
quant yield) as a white solid. H NMR (500 MHz, D
(H,H)=1.5, 9.8 Hz, 1H; H1), 3.96 (m, 1H; H7) 3.92 (dd,
12.3 Hz, 1H; H6b), 3.74 (m, 4H; H6a, H7b, 2”H8) 3.71 (dd,
1.5, 9.8 Hz, 1H; H4), 3.38 (ddd, J
1
6
2
+
0.5 ppm; FAB HRMS: m/z: calcd for C25
39 2
H N O10P: 534.2342 [M] ;
+
found: 557.2240 [M+Na] .
b-Cyanoethoxy-b-(3,4,6-tri-O-acetyl-2-deoxy-a-d-glucopyranosyl)ethoxy-
diisopropylaminophosphine (26): DIEA (0.59 mL, 3.38 mmol) and 2-cya-
1
3
2
O): d=4.74 (dd, J-
3
noethyl-N,N’-diisopropylamino-chlorophosphoramidite
(0.3 mL,
.35 mmol) were added to a solution of 2-hydroxyethyl-2,3,4-tri-O-acetyl-
-deoxy-a-d-glucopyranoside (24; 0.3 g, 0.9 mmol) in dry CH Cl (4 mL)
A
H
R
U
G
J(H,H)=2.3,
A
H
R
U
G
3
1
2
J
A
C
H
T
R
E
U
N
G
(H,H)=
3
ACHTREUNG
(H,H)=2.0, 6.5, 9.0 Hz, 1H; H5), 3.26
2
2
3
3
2
ACHTREUNG
at room temperature under an argon atmosphere. After 1 h, the solution
was diluted with EtOAc (20 mL), and the organic phase was washed with
(t,
J
A
H
R
U
G
J
A
H
R
U
G
J
(H,H)=
(H,H)=12.1 Hz, 1H;
3
H2ax); C NMR (100 MHz, CD OD): d=100.0, 76.6, 71.7, 71.2, 70.5,
3
2
ACHTREUNG
12.1 Hz, 1H; H2eq), 1.52 ppm (dt, J
A
H
E
G
1
3
brine (3”25 mL), dried over anhydrous MgSO
was then evaporated to dryness. The product was purified by silica gel
column chromatography by using Hex/EtOAc (3:2 with 2% of NEt ) as
the eluent to give compound 26 (410 mg, 85%) as a syrup. H NMR
300 MHz, CDCl ): mixture of isomers: d=5.33–5.36 (m, 1H; H3), 5.02–
.88 (m, 2H; H1, H4), 4.25 (dd,
H6a), 4.08–3.91 (m, 2H; H5, H6b), 3.91–3.65 (m, 5H; H7a, 2”H8,
4
, and filtered; the solvent
+
61.5, 61.0, 38.9 ppm; ESIMS: m/z: calcd for C
231.1 [M+Na] .
8
H
16
O
6
: 208.1 [M] ; found:
+
3
1
1-Hydroxyethyl-2-deoxy-a-d-glucopyranoside (12): NaOMe (16 mg,
0.30 mmol) was added to a solution of 2-hydroxyethyl-2,3,4-tri-O-acetyl-
2-deoxy-a-d-glucopyranoside (24; 50 mg, 0.15 mmol) in MeOH (2 mL),
and the reaction was stirred at room temperature for 30 min. The solu-
tion was then neutralized with Amberlite IRA-120 resin and filtered, and
the solvent was eliminated in vacuo to obtain compound 12 (31 mg,
(
4
3
3
2
ACHTREUNG
J(H,H)=4.4, J(H,H)=12.3 Hz, 1H;
A
H
R
U
G
3
ACHTREUNG
OCH
2
CH
2
CN), 3.65–3.44 (m, 3H; H7b, 2”CHisopropyl), 2.63 (t,
J
3
2
ACHTREUNG
6
.2 Hz, 2H; OCH
2
CH
2
CN), 2.20 (dd,
J
A
H
R
U
G
J
3
1
3
1
J
H; H2eq), 2.04, 1.98, 1.95 (3s, 9H; 3”OCOCH
A
H
E
G
quant yield) as a white solid. H NMR (500 MHz, D O): d=4.74 (dd, J-
2
2
3
ACHTREUNG
A
C
H
T
R
E
U
N
G
(H,H)=12.2 Hz, 1H; H2ax), 1.14 ppm (d,
A
H
R
U
G
J(H,H)=2.3,
1
3
3
ACHTREUNG
CH3isopropyl); C NMR (75 MHz, CDCl
3
): mix of isomers: d=170.7, 170.7,
70.2, 169.9, 117.7, 97.1, 97.0, 69.3, 69.3, 69.1, 69.0, 67.9, 67.8, 67.6, 67.5,
7.4, 62.5, 62.4, 62.3, 62.2, 58.6, 58.4, 43.2, 43.1, 24.7, 24.6, 21.0, 20.8, 20.7,
J
(H,H)=
3
ACHTREUNG
1
6
2
;
1.5, 9.8 Hz, 1H; H4), 3.38 (ddd, J
(H,H)=9.6 Hz, 1H; H4), 2.29 (ddd,
12.1 Hz, 1H; H2 ), 1.52 ppm (dt, J
H2 ); C NMR (100 MHz, CD OD): d=100.0, 76.6, 71.7, 71.2, 70.5,
1.5, 61.0, 38.9 ppm; ESIMS: m/z: calcd for C
31.1 [M+Na] .
(H,H)=2.0, 6.5, 9.0 Hz, 1H; H5), 3.26
3
3
2
ACHTREUNG
(t,
J
A
H
R
U
G
J
A
H
R
U
G
J
(H,H)=
+
3
2
ACHTREUNG
0.4, 20.3 ppm; FAB HRMS: m/z: calcd for C23
found: 573.2120 [M+K] .
H
39
N
2
O
10P: 534.2342 [M]
A
H
R
U
G
(H,H)=12.1 Hz, 1H;
eq
+
13
ax
3
+
6
2
8
H
16
O
6
: 208.1 [M] ; found:
1
-Hydroxyethyl-b-d-glucopyranoside (8): NaOMe (27 mg, 0.50 mmol)
+
was added to a solution of 2-hydroxyethyl-2,3,4,6-tetra-O-acetyl-b-d-glu-
copyranoside (14; 100 mg, 0.25 mmol) in MeOH (3 mL), and the reaction
was stirred at room temperature for 30 min. The solution was then neu-
tralized with Amberlite IRA-120 resin and filtered, and the solvent was
eliminated in vacuo to obtain compound 8 (56 mg, quant yield) as a
Synthesis of carbohydrate–oligonucleotide conjugates 1–7: Carbohy-
drate–oligonucleotide conjugates were synthesized on an Applied Biosys-
tems 394 synthesizer by using standard b-cyanoethylphosphoramidite
chemistry. The benzene DMT-phosphoramidite derivative was prepared
7834
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 7828 – 7835

Carbohydrate–Aromatic Interactions
FULL PAPER
[
16]
as previously described. Oligonucleotide conjugates were synthesized
on a 0.2 or 1.0 mmol scale low-volume resin and in DMT-off mode. DMT-
ethane phosphoramidite was purchased from Chemgenes Corp. Oligonu-
cleotide supports were treated with 33% aqueous ammonia for 16 h at
[7] J. Screen, E. C. Stanca-Kaposta, D. P. Gamblin, B. Liu, N. A. Ma-
cleod, L. C. Snoek, B. G. Davis, J. P. Simons, Angew. Chem. 2007,
119, 3718; Angew. Chem. Int. Ed. 2007, 46, 3644.
[8] R. Palma, M. E. Himmel, J. W. Brady, J. Phys. Chem. B 2000, 104,
7228.
5
58C, then the ammonia solutions were evaporated to dryness and the
conjugates were purified by reversed-phase HPLC in a Waters Alliance
separation module with a PDA detector. HPLC conditions were as fol-
lows: Nucleosil 120 C18, 250”8 mm 10 mm column; flow rate:
[9] a) J. C. Morales, S. PenadØs, Angew. Chem. 1998, 110, 673; Angew.
Chem. Int. Ed. 1998, 37, 654; b) K. Kobayashi, Y. Asakawa, Y. Kato,
Y. Aoyama, J. Am. Chem. Soc. 1992, 114, 10307.
[10] a) A. P. Davis, R. S. Wareham, Angew. Chem. 1999, 111, 3160;
Angew. Chem. Int. Ed. 1999, 38, 2978; b) M. Mazik, H. Cavga, P. G.
Jones, J. Am. Chem. Soc. 2005, 127, 9045.
ꢀ
1
3
9
7
mLmin ; 27 min linear gradient 0!30%B (solvent A: 5% CH
3
CN/
5% 100 mm triethylammonium acetate (TEAA; pH 6.5); solvent B:
0% CH
3
CN/30% 100 mm TEAA (pH 6.5)).
[
[
11] Y. Ferrand, M. P. Crump, A. P. Davis, Science 2007, 318, 619.
12] K. Piotukh, V. Serra, R. Borriss, A. Planas, Biochemistry 1999, 38,
1
6092.
13] S. E. Kiehna, Z. R. Laughrey, M. L. Waters, Chem. Commun. 2007,
026.
14] a) M. Senior, R. A. Jones, K. J. Breslauer, Biochemistry 1988, 27,
879; b) S. Bommarito, N. Peyret, J. SantaLucia, Jr., Nucleic Acids
[
[
Acknowledgements
4
We thank the Ministry of Science and Education for funding (grant no.:
CTQ-2006-01123/BQU) and for a Ramón y Cajal contract (J.C.M.).
3
Res. 2000, 28, 1929.
[
15] a) E. T. Kool, J. C. Morales, K. M. Guckian, Angew. Chem. 2000,
1
12, 1046; Angew. Chem. Int. Ed. 2000, 39, 990; b) K. M. Guckian,
[
1] a) T. Feizi, Adv. Exp. Med. Biol. 1998, 435, 51; b) R. A. Dwek,
Chem. Rev. 1996, 96, 683; c) A. Varki, Glycobiology 1993, 3, 97;
d) H. J. Gabius, H. C. Siebert, S. Andre, J. Jimenez-Barbero, H. Ru-
diger, ChemBioChem 2004, 5, 740.
B. A. Schweitzer, R. X. F. Ren, C. J. Sheils, D. C. Tahmassebi, E. T.
Kool, J. Am. Chem. Soc. 2000, 122, 2213; c) J. C. Morales, E. T.
Kool, Biochemistry 2000, 39, 12979.
[16] a) U. Wichai, S. A. Woski, Org Lett. 1999, 1, 1173; b) S. Matsuda,
F. E. Romesberg, J. Am. Chem. Soc. 2004, 126, 14419.
[17] B. L. Sorg, W. E. Hull, H.-C. Kliem, W. Mier, M. Wiessler, Carbo-
hydr. Res. 2005, 340, 181.
[18] A. P. Rauter, S. Lucas, T. Almeida, D. Sacoto, V. Ribeiro, J. Justino,
A. Neves, F. V. M. Silva, M. C. Oliveira, M. J. Ferreira, M.-S. Santos,
E. Barbosa, Carbohydr. Res. 2005, 340, 191.
[19] M. Adinolfi, L. De Napoli, G. Di Fabio, A. Iadonisi, D. Montesarch-
io, G. Piccialli, Tetrahedron 2002, 58, 6697.
[20] a) M. Petersheim, D. H. Turner, Biochemistry 1983, 22, 256; b) T.
Ohmichi, S. Nakano, D. Miyoshi, N. Sugimoto, J. Am. Chem. Soc.
2002, 124, 10367.
[
[
2] R. U. Lemieux, Acc. Chem. Res. 1996, 29, 373.
3] a) A. B. Boraston, D. N. Bolam, H. J. Gilbert, G. J. Davies, Biochem.
J. 2004, 382, 769; b) H. Kogelberg, D. Solis, J. Jimenez-Barbero,
Curr. Opin. Struct. Biol. 2003, 13, 646; c) H. Lis, N. Sharon, Chem.
Rev. 1998, 98, 637.
[
4] a) M. C. Fernandez-Alonso, F. J. Canada, J. Jimenez-Barbero, G.
Cuevas, J. Am. Chem. Soc. 2005, 127, 7379; b) M. S. Sujatha, Y. U.
Sasidhar, P. V. Balaji, Biochemistry 2005, 44, 9554; c) V. Spiwok, P.
Lipovova, T. Skalova, E. Buchtelova, J. Hasek, B. Kralova, Carbo-
hydr. Res. 2004, 339, 2275; d) S. Elgavish, B. Shaanan, Trends Bio-
chem. Sci. 1997, 22, 462; e) E. J. Toone, Curr. Opin. Struct. Biol.
1
994, 4, 719.
[21] S. Liu, X. Ji, G. L. Gilland, W. J. Stevens, R. N. Armstrong, J. Am.
Chem. Soc. 1993, 115, 7910.
[22] a) L. K. Tsou, C. D. Tatko, M. L. Waters, J. Am. Chem. Soc. 2002,
124, 14917; b) E. Lacroix, A. R. Viguera, L. Serrano, J. Mol. Biol.
1998, 284, 173.
[
5] a) J. Flint, D. Nurizzo, S. E. Harding, E. Longman, G. J. Davies, H. J.
Gilbert, D. N. Bolam, J. Mol. Biol. 2004, 337, 417; b) L. Guan, Y.
Hu, H. R. Kaback, Biochemistry 2003, 42, 1377; c) M. Kasahara, M.
Maeda, J. Biol. Chem. 1998, 273, 29106; d) K. Maenaka, M. Mat-
sushima, H. Song, F. Sunada, K. Watanabe, I. Kumagai, J. Mol. Biol.
1
995, 247, 281.
[
6] G. Terraneo, D. Potenza, A. Canales, J. Jimenez-Barbero, K. K. Bal-
dridge, A. Bernardi, J. Am. Chem. Soc. 2007, 129, 2890.
Received: February 22, 2008
Published online: July 21, 2008
Chem. Eur. J. 2008, 14, 7828 – 7835
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7835