Control of disaccharide conformation by π-stacking

Article abstract of DOI:10.1139/v03-062

The conformations of a series of derivatives of the disaccharide α-L-fucopyranosyl-(1→3)-2-acetamido-2-deoxy-D-glucopyranoside, part of the Le^x determinant, were studied by molecular modelling using the MM3* forcefield and by ¹H NMR spectroscopy. Unusually shielded O- benzyl protons were observed in the ¹H NMR spectrum of phenyl 2,3,4-tri-O-benzyl-α-L-fucopyranosyl-(1→3)- 2-deoxy-2-phthalimido-1-thio-α-D-glucopyranoside and assigned to the 2-O-benzyl group. This observation was explained by a shift in the population of the conformational mixture present about the glycosidic linkage from the positive Ψ region in the unsubstituted disaccharide to the negative Ψ region induced by π-stacking between the phthalimide and the 2-O-benzyl phenyl ring. The experimental nuclear Overhauser enhancements confirm the accuracy of the calculations.

Full text of DOI:10.1139/v03-062

364
Control of disaccharide conformation by
␲-stacking
Jonathan Watts, Jesús Jiménez-Barbero, Ana Poveda, and T. Bruce Grindley
Abstract: The conformations of a series of derivatives of the disaccharide α-L-fucopyranosyl-(1→3)-2-acetamido-2-
deoxy-D-glucopyranoside, part of the Lex determinant, were studied by molecular modelling using the MM3* forcefield
1
1
and by H NMR spectroscopy. Unusually shielded O-benzyl protons were observed in the H NMR spectrum of phenyl
2,3,4-tri-O-benzyl-α-L-fucopyranosyl-(1→3)-2-deoxy-2-phthalimido-1-thio-α-D-glucopyranoside and assigned to the 2-O-
benzyl group. This observation was explained by a shift in the population of the conformational mixture present about
the glycosidic linkage from the positive Ψ region in the unsubstituted disaccharide to the negative Ψ region induced by
π-stacking between the phthalimide and the 2-O-benzyl phenyl ring. The experimental nuclear Overhauser enhance-
ments confirm the accuracy of the calculations.
Key words: disaccharide, conformation, π-stacking, Lex determinant, NOE measurements, MM3 calculations.
Résumé : Faisant appel à la modélisation moléculaire à l’aide du champ de force MM3* et à la spectroscopie WRaMttNs et
1
al.
du H, on a étudié les conformations d’une série de dérivés du disaccharide α-L-fucopyranosyl-(1→3)-2-acétamido-2-
1
désoxy-D-glucopyranoside, une portion du déterminant Lex. Dans le spectre RMN du H du 2,3,4-tri-O-benzyl-α-L-fuco-
pyranosyl-(1→3)-2-désoxy-2-phtalimido-1-thio-α-D-glucopyranoside de phényle, on a observé des protons O-benzyliques
exceptionnellement blindés qui ont été attribués au groupe 2-O-benzyle. Cette observation est expliquée par un déplace-
ment dans la population du mélange conformationnel présent autour de la liaison glycosidique par rapport à la région
Ψ positive dans le disaccharide non substitué à une région Ψ négative induite par un empilement π entre le phtalimide
et le noyau phényle du groupe 2-O-benzyle. Des rehaussements expérimentaux d’effet Overhauser nucléaire confirment
l’exactitude des calculs.
Mots clés : disaccharide, conformation, empilement π, déterminant Lex, mesures d’effets Overhauser nucléaire, calculs
MM3.
[Traduit par la Rédaction] 375
Introduction
the oligosaccharides present on mamalian zona pellucidas as
potential immunocontraceptives, we prepared a disaccharide,
phenyl 2,3,4-tri-O-benzyl-α-L-fucopyranosyl-(1→3)-2-deoxy-
2-phthalimido-1-thio-α-D-glucopyranoside (4). Fucose-containing
oligosaccharides were targeted because of a report that the
presence of fucose markedly enhanced sperm binding to the
murine zona pellucida glycoprotein ZP3 (10). The confor-
mation adopted by this disaccharide (4) appeared to be rather
unusual because a signal of one of its benzyl protons ap-
peared at 3.38 ppm in its 1H NMR spectrum in chloroform-d,
far from the normal region for benzyl protons, ~4.4–4.9 ppm
(11). This report summarizes our determination of the cause
of this unusual observation and provides additional informa-
tion about the conformational energy surface for the very
important disaccharide 8.
Recognition of carbohydrate epitopes by protein depends
on the conformation adopted by the carbohydrate chains.
Normally, the most populated conformations are recognized
(1–5), but often less populated conformers are the epitopes
(6–8). Therefore, it is important to fully define the factors
that influence conformational stability and also to determine
the conformational potential energy surface of biologically
active oligosaccharides completely.
The sequence α-L-fucopyranosyl-(1→3)-2-acetamido-2-deoxy-
D-glucose (8) is a principal constituent of Lex-bearing glycocon-
jugates, which mediate protein binding to the selectins and
other proteins in several processes: inflammation (9), cancer,
and others. As part of a project to synthesize conjugates of
Received 22 January 2002. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on 20 May 2003.
Results and discussion
Dedicated to the late Professor R.U. Lemieux.
Synthesis
J. Watts and T.B. Grindley.1 Department of Chemistry,
Dalhousie University, Halifax, NS B3H 4J3, Canada.
J. Jiménez-Barbero. Centro de Investigaciones Biológicas,
CSIC, Velázquez 144, 28006 Madrid, Spain.
A. Poveda. Universidad Autonoma Madrid, 28049
Cantoblanco, Spain.
A summary of the synthetic procedures is shown in
Scheme 1. Benzylation of phenyl 1-thio-α-L-fucopyranoside
(12) gave the 2,3,4-tri-O-benzyl derivative (1), whose NMR
spectra had not been fully assigned previously. Reaction of
phenyl 2-deoxy-2-phthalimido-1-thio-α-D-glucopyranoside with
benzaldehyde dimethyl acetal and p-toluenesulfonic acid in
DMF afforded the 4,6-O-benzylidene acetal (2) in 80% yield,
1Corresponding author (e-mail: bruce.grindley@dal.ca).
Can. J. Chem. 81: 364–375 (2003)
doi: 10.1139/V03-062
© 2003 NRC Canada

Watts et al.
365
Scheme 1.
a compound prepared previously by a different method (13).
Oxidation of 1 to a mixture of sulfoxide diastereomers with
3-chloroperbenzoic acid (14, 15) and glycosidation with 2
and triflic acid yielded the de-O-benzylidenated disaccharide
4 in 31% yield. Direct reaction of 1 with 2 catalyzed by N-
iodosuccinimide and triflic acid gave a good yield of the
succinimide derivative of the glycosyl donor. Such products
have been observed previously when active glycosyl donors
were reacted with inactive glycosyl acceptors in the pres-
ence of succinimide (16). Catalysis of the reaction of 1 and
2 by dimethyl(methylthio)sulfonium triflate (DMTST) (17,
18), a cationic catalyst with a non-nucleophilic counterion,
yielded the expected disaccharide 3 in 73% yield. The N-
acetyl analogs of 3 and 4, compounds 5 and 6, respectively,
were obtained by removal of the phthalimide group from 4
with ethylenediamine, N-acetylation to give 5, then de-O-
benzylidenation with aqueous sulphuric acid to give 6.
Methyl 2,3,4-tri-O-benzyl-α-L-fucopyranoside (7) was ob-
tained from 1 by reaction with methanol and iodine (19).
These latter three compounds were prepared in order to dis-
cover which interactions were critical in determining the con-
formations that caused the unusual chemical shift for the
benzyl proton of 4 (Scheme 1).
NMR spectroscopy and conformational analysis
Conformations about glycosidic bonds are discussed in
terms of the Φ and Ψ angles, defined as shown in Scheme 2
(20). In the following discussion, protons on the fucosyl res-
idue are indicated by following the ring position by F, e.g.,
H1F is the fucosyl anomeric proton, while those on the N-
acetylglucosaminyl residue are indicated by a following G.
NMR chemical shifts of the compounds discussed are shown
in Table 1, while details of the calculations and the NOE re-
sults are listed in Table 2.
Adiabatic potential energy surfaces for α-L-fucopyranosyl-
(1→3)-2-acetamido-2-deoxy-D-glucopyranoside (8) derivatives
© 2003 NRC Canada

366
Can. J. Chem. Vol. 81, 2003
Scheme 2.
For the Ψ-positive minimum, the π-stacking between the
phthalimide moiety and the benzyl group at position O2F
disappears with a concomitant change in the interproton dis-
tances. The short interresidue distances from the global mini-
mum conformer are calculated to be longer in this conformer
of 4: H1F–H3G is 3.0 Å; H5F–H4G is 4.0 Å; H6F–OH4G is
4.7 Å; and H1F–OH4G is 4.8 Å; and the two hydrogen atoms
in the potentially hydrogen-bonded pair OH4G–O5F are now
far apart (4.1 Å). In this conformer, the short interproton dis-
tances are H5F–H3G (2.5 Å) and H5F–OH4G (2.6 Å). As
mentioned above, the calculated relative energy difference
between the minima amounts to 10 kJ mol–1. This difference
in stability results in the prediction that the population of the
Ψ-positive minimum will be negligible.
NMR spectral data for 4 were recorded and scrutinized to
assess the possible existence of both conformers (or con-
formational families). The most striking feature was the dou-
blet in the spectrum of 4 due to an unusually shielded benzyl
proton, which sparked this investigation (see Table 1); it was
assigned to the O2F benzyl group through HMBC and NOESY
experiments. This shielded position indicated that this benzy-
lic proton is in the vicinity of an aromatic moiety. Obviously,
this moiety could correspond to its vicinal OBn moiety at
position C3F or to the interresidue phthalimide ring attached
to C2G. The observation of normal chemical shifts for the
benzyl groups of methyl 2,3,4-tri-O-benzyl-α-L-fucopyran-
oside (7) suggests that the vicinal OBn moiety at position
C3F is not responsible for the shielding. As will be shown
below, by using non-aromatic substituents at C2G, this latter
hypothesis is valid.
have been explored previously using a variety of force fields
(21, 22). The major low-energy region is centered in the so-
called (23, 24) syn-Φ/Ψenergy region, where both Φ and Ψ
have absolute values <60°. For 3 and 4, MM3* calculations
indicate that the global minimum is located in the syn-Φ/Ψre-
gion and has dihedral angles of ca. Φ = 32° and Ψ = –52° (see
Table 2). A second minimum was also found for 3 and 4 with
positive Ψ angles, with Φ ca. 65° and Ψ ca. 40° — about 9 kJ
mol–1 higher in energy than the global minimum. Obviously,
low-amplitude motions around both glycosidic linkages may
take place, as commonly found for oligosaccharide molecules.
The computed Φ/Ψ values are in agreement with those re-
ported for the natural deprotected disaccharide 8 and its thio-
analogue (22). However, for 8, the Ψ-positive conformer was
Relevant interproton NOEs were observed for 4 (see Ta-
ble 2). Thus, the methyl group of the Fuc residue gives an
NOE to OH4G (see Fig. 2), and the H1F gives NOEs to
H3G and to OH4G. These NOEs are consistent with the ex-
istence of a major conformation in the negative Ψ area, as
predicted by the calculations. There was a peak that could
also correspond to the fourth expected NOE, according to
the short distances explained above, H5F–H4G, but the sig-
nal of H5F overlaps with the more deshielded of the
benzylic protons at O2F and that of H4G overlaps with that
of the other O2F benzylic proton (see Figs. 2 and 3), and the
geminal O2F benzylic protons must have a large NOE. No
NOEs were observed for H5F–H3G and H5–OH4G, consis-
tent with the predicted negligible contribution from con-
formers around the positive Ψ area. The presence of the
intramolecular O4G—H···O5F hydrogen bond was indirectly
deduced from the appearance of the OH signal. This reso-
nance is an unusually deshielded, slightly broadened signal
(4.3 ppm) with a vicinal coupling smaller than 1.5 Hz. There-
fore, the hydroxyl proton adopts a particular orientation with
respect to its vicinal H4G and does not freely rotate in solu-
tion. Otherwise, a medium-size coupling of ca. 5–7 Hz
would have been expected. This particular orientation may
be due to its involvement in an intramolecular hydrogen
bond to O5F, as deduced from the calculations.
calculated (22) to be more stable by about 3 kJ mol–1, the re-
verse of the preferences calculated for 3 and 4. Ψ-positive
conformers are also calculated (25, 26) and observed (25) to
be more stable for the α-L-fucopyranosyl linkage to O-3 of N-
acetyl-D-glucosamine in Lex. Both conformers have Φ values
in the +syn region, in agreement with expectations based on the
exo-anomeric effect (20). For 3 and 4, the calculations indicate
that the anti minima were, respectively, 24 and 31 kJ mol–1
less stable than the global minimum, in contrast to those for 8
(22), where it was only 10 kJ mol–1 less stable than the global
minimum.
Analysis of the key structural factors present in both min-
ima for 3 and 4 was performed. For the global minimum (Ψ
negative), it is evident that the aromatic rings of the phthali-
mide moiety and the benzyl group at position C-2′ are per-
fectly stacked (see Fig. 1). Short interresidue proton–proton
distances between H1F–H3G (2.3 Å), H5F–H4G (2.5 Å),
H6F–OH4G (2.7 Å), and H1F–OH4G (2.7 Å) are predicted
for 4 (see Table 2). In addition, for 4, a short distance was
calculated between the hydroxyl hydrogen OH4G and O5F
(2.7 Å), which suggests that a hydrogen bond may exist be-
tween them.
1
In the H NMR spectrum of the 4,6-O-benzylidene deriva-
tive of 4, compound 3, a single benzyl proton (assigned to the
2-O-benzyl group) is again markedly shielded (3.84 ppm).
This indicates that 3 also prefers Ψ-negative conformers, al-
though probably not to the same extent as for 4. The fucose
methyl group was also shielded, appearing at 0.87 ppm,
© 2003 NRC Canada

Watts et al.
367
Table 1. NMR spectral parameters for benzyl groups in chloroform-d.
C-2 OCH2Ph C-3 OCH2Ph
C-4 OCH2Ph
Compound
δHA
δHB
JA,B
δHA
δHB
JA,B
δHA
δHB
JA,B
1
7
3
4
5
6
4.73
4.69
3.84
3.38
4.61
4.64
4.79
4.82
4.26
4.13
4.93
4.84
10.3
12.1
12.6
13.0
11.5
11.5
4.732
4.73
4.38
4.64
4.75
4.75
4.745
4.87
4.43
4.84
4.75
4.78
12.0
11.9
11.6
11.5
—
12.6
4.67
4.65
4.49
4.55
4.56
4.62
5.01
4.98
4.79
4.89
4.93
4.96
11.6
11.6
11.6
11.4
11.5
11.5
a
aChemical shift difference too small to measure JA,B
.
compared to a shift of 1.11 ppm for the methyl of α-L-
fucopyranose (11), presumably because it is in the shielding
cone of the benzylidene phenyl (see Fig. 1). In this case, the
MM3* calculations indicate again that the Ψ-negative region
contains the global minimum, while the Ψ-positive and anti
conformers are less stable by 8.6 and 23.9 kJ mol–1, respec-
tively (see Table 2).
chemical shifts, similar to those observed for the other CH2
protons of the benzyl groups at positions 3 and 4 and very
similar to those of the monosaccharide 7. However, the
methyl group of the acetamide unit is strongly shielded
(1.61 ppm), in contrast with the chemical shift of 2.02 ppm
for the acetamido methyl of phenyl 2-acetamido-2-deoxy-1-
thio-β-D-glucopyranoside, the related monosaccharide (27).
These features provide evidence about two points, namely
that the phthalimide moieties of 3 and 4 are π-stacked with
the fucose O2 benzyl groups and that a conformer with an
analogous Me–aromatic interaction is present for 6. The
NOESY experiments showed the presence of the short H1F–
H3G, H5F–H4G, and H1F–OH4G distances, as expected for
the presence of the negative Ψ conformer. However, addi-
tional crosspeaks appear in the spectrum that cannot be ex-
plained by the exclusive presence of conformers from the
negative Ψ area. These are H5F–H3G and H1F–HNG NOEs
that can only take place when conformers with positive Ψ
values coexist in the conformational mixture.
Similar observations were made for the benzylidene deriva-
tive 5. The acetamide methyl group was shielded (1.66 ppm),
as was the methyl group of the fucose moiety (0.83 ppm),
indicating the presence of conformers having interactions of
these methyl groups with aromatic moieties. H1F–HNG and
H1F–H3G NOEs were observed, indicating the presence of
conformers in both positive and negative Ψ areas. The analy-
sis of the geometry of the MM3*-derived conformers indi-
cate that the shielding of the methyl group of the fucose
moiety is owing to the benzylidene unit of the GlcNAc moi-
ety for both positive and negative Ψ angles, while that of the
acetamide methyl group is owing to the 2-O-benzyl group of
the Fuc unit only when Ψ adopts negative values.
In conclusion, a number of NMR observations assisted by
MM3* calculations have shown the importance of stacking
interactions for the existence of conformational variations
around the glycosidic torsion angles of oligosaccharides. It
has been shown that the existence of π-stacking Ar–Ar inter-
action in the phthalimide derivatives (3 and 4) limits the
populated conformers to the Ψ-negative region of the com-
plete Φ/Ψ map. In comparison with the unsubstituted
disaccharide (22), this is a change of about 13 kJ mol–1. The
stabilization provided by π-stacking of benzene rings is
about 10 ± 5 kJ mol–1 (28–31), similar in magnitude to this
change. The change of the phthalimide unit to an acetamide
shifts the conformational equilibrium in a way that the posi-
tive Ψ area starts to be populated. Although an Me–Ar inter-
action is still evident and stabilizes the negative Ψ area, the
removal of the Ar–Ar interaction changes the conform-
As a further step, the existence and importance of the π-
stacking interaction between both aromatic rings located at
the different GlcN and Fuc moieties was explored. Thus, the
GlcNAc-containing analogs of compounds 3 and 4 were stud-
ied — disaccharides 5 and 6, respectively. The molecular me-
chanics calculations on 5 and 6 also predict the occurrence
of conformers with negative and positive Ψ values (see
Table 2). However, in contrast with results of the MM3* cal-
culations for 4, the energy preference for the Ψ-negative con-
former is calculated to be much less, ca. 3 kJ mol–1 for the
unprotected derivative 6 and only 1 kJ mol–1 for compound
5. The anti conformers are also destabilized to a lesser ex-
tent, 14.1 and 12.8 kJ mol–1 for 5 and 6 with respect to the
global minima, respectively, but are still too energetic to
contribute perceptibly to the conformational mixture. These
relatively smaller values, in comparison to those of 3 and 4,
are probably owing to the significant stabilization provided
by the Ar–Ar interaction in the phthalimide-containing mol-
ecules. Therefore, for both 5 and 6, significant populations
of the Ψ-positive conformers are calculated to be present in
equilibrium mixtures in which the Ψ-negative conformers
are most populated. The local minimum geometries in this
area have Φ/Ψ 62°/43° for the unprotected derivative 6 and
slightly different values, Φ/Ψ 45°/17°, for the benzylidene
analogue 5. For both compounds, the Φ/Ψ values and the
corresponding geometrical features in the Ψ-negative region
are basically identical to those described above for 4 (Φ/Ψ
30°/–50°). In both cases, interactions between the aromatic
rings of the 2-O-benzyl groups of the Fuc moieties and the
methyl groups of the acetamide units of the GlcNAc resi-
dues are evident. The H1F–HNAc distance is greater than
4 Å.
The expected distances for the secondary Ψ-positive mini-
mum of the non-benzylidenated derivative 6 are analogous
to those described above. One additional interresidue contact
does appear: H1F–HNAc (2.5 Å). Some changes in the dis-
tances take place for the benzylidene analogue 5: H1F–H3G
(2.4 Å); H5F–H4G (3.6 Å); H5F–H3G (3.4 Å); and H1F–
HNAc (2.5 Å).
1
In the H NMR spectrum of the unprotected analogue 6,
the 2-O-benzylic protons of the Fuc residue have normal
© 2003 NRC Canada

368
Can. J. Chem. Vol. 81, 2003
© 2003 NRC Canada

Watts et al.
369
Fig. 1. From the left, positive, anti, and negative Ψ conformers of compound 3. The more important stacking interactions take place in
the negative Ψ conformer, although the positive Ψ conformer also shows a less energetically relevant 6-Me–benzylidene interaction.
The conformers of compound 4 are similar. Torsion angle values and energies are given in Table 2.
ational distribution towards the “natural” one (more positive
than negative) (22). The inclusion of a 4,6-O-benzylidene
group on the GlcNAc moiety further shifts the equilibrium
towards the positive Ψ area, possibly attributable to an addi-
tional stabilizing Me–Ar interaction, which is stronger for
conformations within this region. Although the MM3* force
field used in this work only uses van der Waals interactions
to model the Me–Ar or the Ar–Ar interactions, it provides
models that satisfactorily explain the experimental observa-
tions.
The fact that a shift in conformational population to the
negative Ψ area occurs on introducing benzyl substituents
onto fucose provides experimental support for the conclu-
sion, based on MM3* calculations (22), that the potential
energy surface for the natural disaccharide 8 contains a min-
imum in the negative Ψ area only slightly higher in energy
than the global minimum.
enhanced sequence and an inverse probe. A data matrix of
256 * 2K was used to digitize a spectral width of 4000 Hz in
F2 and 15 000 Hz in F1. Eight scans were used per increment
with a relaxation delay of 1 s and a delay corresponding to a
J value of 145 Hz. 13C decoupling was achieved by the
WALTZ scheme. HMBC experiments were recorded using
the gradient-enhanced sequence; 16 scans were used per in-
crement (256), adjusting the long-range coupling evolution
delay for values of J = 3, 4, 6, 8, 10, and 12 Hz. 2D NOESY
and 2D-T-ROESY (32) experiments were performed with
the standard sequences and with 600 ms of mixing time.
Data matrixes of 256 * 2K were employed. The exact masses
of compounds 1 and 2 were measured on a CEC 21–110B
mass spectrometer using electron ionization (70 eV). The
others were measured on a Micromass ZAB mass spectrom-
eter at the University of Alberta using electrospray ioniza-
tion. Thin layer chromatography was performed on 0.25-mm
thick Whatman G/UV and Silicycle silica gel aluminum
plates. Components were visualized by spraying with a 2%
ceric sulfate solution in 1 M H2SO4, followed by heating on
a hot plate until coloured spots formed. They were purified
by flash chromatography on TLC standard grade (230–400
mesh) silica gel using mixtures of hexanes and ethyl acetate
as eluents unless otherwise specified. Melting points were
determined with a Fisher-Johns melting point apparatus and
are uncorrected. Optical rotations were measured on a Digi-
Pol model 781 polarimeter.
Experimental
General methods
NMR spectra for synthetic purposes were recorded on
Bruker AC-250 and Bruker AMX-400. The samples were
made up in concentrations of approximately 10–15 mM in
chloroform-d unless otherwise specified. Chemical shifts are
given in parts per million (ppm) (±0.01 ppm) relative to
1
TMS (tetramethylsilane) in the case of H NMR spectra and
to the central line of CDCl3 (δ 77.16) for the 13C NMR
spectra. H-1 chemical shifts and coupling constants were ob-
tained by first-order analyses. NMR experiments for con-
formational studies were recorded on Varian Unity 500 and
Bruker DRX-500 spectrometers, using an approximately
5 mg mL–1 solution of the disaccharides at different temper-
atures. Chemical shifts are reported in ppm, using external
TMS (0 ppm) as references. CDCl3 and (CD3)2CO were
used as solvents. The COSY spectra were performed with a
data matrix of 256 * 1K to digitize a spectral width of
4000 Hz. 16 scans were used with a relaxation delay of 1 s.
The 2D TOCSY experiments were performed using a data
matrix of 256 * 2K to digitize a spectral width of 4000 Hz; 8
scans were used per increment with a relaxation delay of 2 s.
MLEV 17 was used for the 70 ms isotropic mixing time.
The one-bond proton–carbon correlation experiment was col-
Pyridine was dried by reflux and distillation over calcium
hydride. Dry methanol was obtained by distillation from mag-
nesium turnings activated by iodine. N,N-Dimethylformamide
was dried by distillation under reduced pressure from 4 Å
molecular sieves onto fresh 4 Å molecular sieves. Toluene
was stored over calcium hydride for 3 h and distilled onto 4 Å
molecular sieves. Dichloromethane was distilled from calcium
hydride and stored over 4 Å molecular sieves. All organic ex-
tracts were dried with anhydrous MgSO4.
Molecular mechanics calculations
Molecular mechanics calculations were performed using
the MM3* force field as implemented in MACROMODEL
4.5 (33). The MM3* force field implemented in MACRO-
MODEL differs from the original MM3 force field (34) in
the treatment of the electrostatic term since it uses charge–
charge instead of dipole–dipole interactions. Atoms from the
1
lected in the H-detection mode using the HSQC gradient-
© 2003 NRC Canada

370
Can. J. Chem. Vol. 81, 2003
Fig. 2. Part of the 500 MHz NOESY spectrum of phenyl 2,3,4-tri-O-benzyl-α-L-fucopyranosyl-(1→3)-2 deoxy-2-phtalimido-1-thio-β-D-
glucopyranoside (4).
non-reducing monosaccharide are dubbed F and those of the
reducing one, G. The torsional angle Φ across the disacchar-
ide linkage is defined as H1F-C1F-O-C3G and Ψ as C1F-O–
C3F-H3F.
© 2003 NRC Canada

Watts et al.
371
Fig. 3. Part of the 500 MHz HSQC spectrum of phenyl 2,3,4-tri-O-benzyl-α-L-fucopyranosyl-(1→3)-2-deoxy-2-phthalimido-1-thio-β-D-
glucopyranoside (4).
MM3* calculations either in vacuo or with the GB/SA sol-
vent model for chloroform were performed. The minimizations
were performed using 10 000 conjugate gradient iterations until
the rms derivative was better than 0.03 kJ Å–1 mol–1. Eighteen
different combinations of geometries of 4 and 6, corresponding
to the combination of the nine possible staggered rotamers
© 2003 NRC Canada

372
Can. J. Chem. Vol. 81, 2003
around Φ/Ψ with the two major rotamers around C5–C6 (gg
and gt) of the reducing end were built and submitted to energy
minimizations under both dielectric conditions. For 5, only the
nine possible rotamers around Φ/Ψ were considered. The S-
phenyl group was set at a Φ value of –60°, in agreement with an
exo-anomeric orientation (H1G-C1G-S-CPh). The phthalimide
and the benzyl groups were set at positions where steric inter-
actions with their corresponding rings were minimized and let
free during the minimization process, after several trials with
the corresponding monosaccharides. In particular, the starting
position of the phthalimide group defined as H2G-C2G-N-CO
was given an initial value of 30° and H2G-C2G-N-CO′ was
started at –150°. For the 2-O-benzyl group, the starting orienta-
tion H2F-C2F-O-CH2 was given a value of 60° and C2F-O-
CH2-CPh was started in an anti arrangement. For the 3-O-
benzyl group, the starting orientation H3F-C3F-O-CH2 was 60°
and C3F-O-CH2-CPh was also oriented anti. For the 4-O-benzyl
group, the starting orientation H4F-C4F-O-CH2 was –60° and
C4F-O-CH2-CPh was again anti. Depending on the compound,
displacements of these groups from the original positions were
found during the energy minimization steps, particularly for the
2-O-benzyl group. The HO-4 hydroxyl group of compounds 4
and 6 was set with a torsion angle (H-C-O-H) of 60°, pointing
away from the glycosidic linkage and let free during the
minimization process. From the Φ/Ψ viewpoint, and for all of
3–6, three conformers were always the most preferred ones:
those with exo-anomeric orientations for the Φ angle of F,
combined with the three possible Ψ+, Ψ–, and Ψanti orientations.
In all cases, the Ψanti conformers displayed the highest energy
values, in agreement with the experimental NOE results (see
text). The non-anomeric (Φ ca. –60°) starting conformers were
not local minima and converged to the exo-anomeric region.
The Φanti conformers were stable, but highly destabilized
(more than 50 kJ mol–1) with respect to the global minimum
and were not considered for subsequent analysis. Values of Φ/
Ψ for the global and local minima oscillated ±10° between the
GB/SA and the ε = 80 values, and between the gg and gt
rotamers of the hydroxymethyl group of the G moiety of 4 and
6. Additionally, the relative steric energy values varied less
than 4 kJ mol–1 between the GB/SA and the ε = 80 values, and
between the gg and gt rotamers of the hydroxymethyl group of
the G moiety of 4 and 6. The most stable conformers, which
agreed with the experimental NOEs, are depicted in the corre-
sponding figures.
3.53 (dq, 1H, J5,6 = 6.4 Hz, J4,5 = 0.9 Hz, H-5), 3.59 (dd,
1H, J3,4 = 2.8 Hz, J2,3 = 9.2 Hz, H-3), 3.63 (dd, 1H, J4,5
=
0.9 Hz, J3,4 = 2.8 Hz, H-4), 4.60 (d, 1H, J1,2 = 9.6 Hz, H-1),
4.67 (d, 1H, J = 11.6 Hz, 1 H of -CH2Bn on O-3), 4.73 and
4.79 (2 d (leaning), 2H, J = 10.2 Hz, -CH2Bn on O-2), 4.74
(s with second-order side peaks, 2H, -CH2Bn on O-4), 5.01
(d, 1H, J = 11.5 Hz, 1 H of -CH2Bn on O-3), 7.0–7.7 (m,
20H, 4 Ph). 13C NMR (100.62 MHz, CDCl3) δ: 17.43 (C6),
72.92 (CH2 of 4-O Bn), 74.69 (C5 + CH2 of 3-O Bn), 75.66
(CH2 of 2-O Bn), 76.68 (C4), 77.19 (C2), 84.61 (C3), 87.60
(C1). EI-HR-MS calcd. for C33H34O4S ([M+]): 526.2178;
found: 526.2178.
Phenyl 4,6-O-benzylidene-2-deoxy-2-phthalimido-1-thio-
β-D-glucopyranoside (2)
Anhydrous phenyl 2-deoxy-2-phthalimido-1-thio-β-D-gluco-
pyranoside (13) (31.1 g, 77.5 mmol) was dissolved in dry
DMF (150 mL) in a 500-mL, round-bottomed flask. Benzal-
dehyde dimethyl acetal (12.3 g, 80.7 mmol) and para-
toluenesulfonic acid hydrate (0.11 g, 0.58 mmol) were added.
The solution was attached to a rotary evaporator, evacuated
with a water aspirator, rotated, and lowered into a water bath
at 60 ± 5°C. After 140 min, another aliquot of benzaldehyde
dimethyl acetal (10.1 g, 66.3 mmol) was added. After 220
min, about half the solvent was removed by reducing the
pressure further with an oil pump, then a solution of NaHCO3
(0.43 g) in water (50 mL) was added. A precipitate formed,
and the mixture was concentrated to near-dryness. The reaction
mixture was partitioned between dichloromethane (500 mL)
and water (300 mL). The water layer was extracted with
dichloromethane (3 × 100 mL) and discarded. The combined
organic layers were washed with water (2 × 100 mL) and
dried. These extractions were relatively difficult owing to the
recurring formation of an emulsion layer. The organic layers
were dried and concentrated, and the residue was purified by
flash chromatography (5:2 Hex:EtOAc to 2:1 Hex:EtOAc),
yielding 2 (30.2 g, 79.5%) as an amorphous solid: [α]D 35.0°
1
(c 2.04, CHCl3), lit. (13) [α]D 34.2° (c 1.3, CHCl3). H NMR
(CDCl3) δ: 2.84 (d, 1H, JH-3,OH = 3.75 Hz, OH), 3.54 (t, 1H,
J3,4 ≈ J4,5 = 9.2 Hz, H-4), 3.63 (dt, 1H, J4,5 ≈ J5,6ax = 9.3 Hz,
J5,6eq = 4.6 Hz, H-5), 3.78 (t, 1H, J6ax,6eq ≈ J5,6 = 9.9 Hz, H-
6ax), 4.30 (t, 1H, J2,3 = J1,2 = 10.3 Hz, H-2), 4.35 (dd, 1H,
J5,6eq = 4.5 Hz, J6ax,6eq = 10.2 Hz, H-6eq), 4.59 (ddd, 1H,
JH3,OH = 3.8 Hz, J3,4 = 8.5 Hz, J2,3 = 9.9 Hz, H-3), 5.53 (s,
1H, PhCH(O-)2), 5.65 (d, 1H, J1,2 = 10.5 Hz), 7.23–7.84 (m,
2 Ph + Phth). 13C NMR (CDCl3) δ: 55.6 (C-2), 68.6 (C-6),
69.7 (C-3), 70.3 (C-5), 81.9 (C-4), 84.3 (C-1), 102.0
(PhCH(OR)2). EI-HR-MS: calcd. for C27H23O6NS ([M+]):
489.1246; found: 489.1241.
Phenyl 2,3,4-tri-O-benzyl-1-thio-β-L-fucopyranoside (1)
A solution of phenyl 1-thio-β-L-fucopyranoside (12) (1.00 g,
3.9 mmol) in DMF (30 mL) was cooled to 0°C. Sodium hy-
dride (60% w/w oil suspension, 30 mmol) was added, and
the mixture was allowed to stir for 30 min. Benzyl bromide
(2.0 mL, 2.9 g, 17 mmol) was added, and the mixture was
allowed to warm to room temperature, then stirred for a fur-
ther 2 h. Methanol (1 mL) was added, and after 15 min, the
mixure was concentrated under reduced pressure, then parti-
tioned between ethyl acetate and water; the water layers
were washed with ethyl acetate and discarded, and the com-
bined ethyl acetate layers dried and concentrated. Recrystal-
lization (ether–hexane) yielded colourless prisms (1.26 g,
61%): mp 97–103°C (lit. (12) mp 107–109°C); [α]D –7.0 (c
First attempt to prepare phenyl 2,3,4-tri-O-benzyl- -L-
fucopyranosyl-(1→3)-4,6-O-benzylidene-2-deoxy-2-
phthalimido-1-thio-β-D-glucopyranoside (3)
Anhydrous compounds 1 (175 mg, 0.332 mmol) and 2
(115 mg, 0.236 mmol), along with N-iodosuccinimide (NIS,
85 mg, 0.378 mmol) and powdered 4 Å molecular sieves
were placed in a 2-necked, round-bottomed flask, which was
fitted with a rubber septum and an argon-filled balloon. The
mixture was cooled to –40°C, and dichloromethane (8 mL)
was added via syringe. The stirred solution was allowed to
warm up to –20°C over 35 min, and triflic acid (0.3 ␮L,
1
1.58, CHCl3), (lit. (12) [α]D –14.0 (c 0.7, CHCl3). H NMR
(400.14 MHz, CDCl3) δ: 1.27 (d, 3H, J5,6 = 6.4 Hz, 3 H-6),
© 2003 NRC Canada

Watts et al.
373
0.5 mg, 0.0035 mmol) was added via syringe at 65 min. The
cooling bath was removed after 100 min, and the reaction
continued to stir and was monitored by TLC.
Phenyl 2,3,4-tri-O-benzyl-α-L-fucopyranosyl-(1→3)-2-
deoxy-2-phthalimido-1-thio-β-D-glucopyranoside (4)
Compound 1 (473 mg, 0.955 mmol) was dissolved in di-
chloromethane (30 mL) in a 50-mL, round-bottomed flask
fitted with a drying tube and a magnetic stirrer. The solution
was cooled to –78°C using an ethyl acetate – liquid nitrogen
slurry. m-Chloroperbenzoic acid (227 mg of 85% reagent,
1.12 mmol) was added to the stirring solution, and the reac-
tion was monitored by TLC (eluent 2.5% acetone in CH2Cl2;
Rf sulfide = 0.92, Rf sulfoxide = 0.51, 0.39 (R and S iso-
mers), Rf sulfone = 0.82). After 3.5 h, the reaction was al-
lowed to warm up somewhat, then saturated aqueous sodium
bicarbonate (15 mL) and dichloromethane (20 mL) were
added. The aqueous layer was extracted with dichlorome-
thane (2 × 20 mL). The combined organic layers were wash-
ed with water (30 mL), dried, and concentrated, yielding a
residue that was purified by flash chromatography (2.5% ac-
etone in CH2Cl2). The R and S isomers of phenyl 2,3,4-tri-
O-benzyl-1-thio-α-L-fucopyranosyl sulfoxide (1a) were col-
lected together and used for the next step.
After 175 min, the reaction mixture was diluted with di-
chloromethane (10 mL) and filtered, then washed with satu-
rated aqueous sodium bicarbonate (3 × 50 mL) and brine
(50 mL), dried, and concentrated. The residue was purified
by flash chromatography (4:1 to 2:1 Hex:EtOAc) to give N-
(2,3,4-tri-O-benzyl-β-L-fucopyranosyl)succinimide as a syrup
(132 mg, 77%). 1H NMR (CDCl3) δ: 1.09 (d, J5,6 = 6.4 Hz, 3
H-6), 2.55 (s, 2 CH2 of succinimide), 3.71 (bd, J3,4 = 2.6 Hz,
H-4), 4.28 (bq, J5,6 = 6.4 Hz, H-5), 4.49 (dd, J1,2 = 7.5 Hz,
J2,3 = 9.8 Hz, H-2), 4.58 (dd, J2,3 = 9.8 Hz, J3,4 = 2.9 Hz, H-
3), 4.48, 4.67 (2 leaning d, J = 11.7 Hz, CH2Bn), 4.66, 4.99
(2 leaning d, J = 11.6 Hz, CH2Bn), 4.76, 4.86 (2 leaning d, J =
11.8 Hz, CH2Bn), 6.13 (d, J1,2 = 7.5 Hz, H-1), 7.25–7.33 (m,
3Ph). 13C NMR (CDCl3) δ: 17.5 (C6), 28.3 (2 CH2 of
succinimide), 72.8 (C5), 73.3 (CH2Bn), 73.6 (CH2Bn), 74.9
(C2 + CH2Bn), 76.7 (C1), 77.7 (C4), 80.6 (C3), 127.6,
128.3, 128.4, 138.1, 138.6, 139.1 (aromatic), 178.0 (2 C=O
of succinimide).
Compounds 1a (313 mg, 0.577 mmol) and 2 (200 mg,
0.408 mmol) were combined with 2,6-di-tert-butyl-4-methyl-
pyridine (168 mg, 0.812 mmol) in a 2-necked, round-
bottomed flask. One neck was stoppered and the mixture
was azeotroped with toluene (3 × 10 mL), keeping the tem-
perature low to avoid decomposition of the sulfoxide. The
flask was fitted with a rubber septum and an argon-filled
balloon. Dichloromethane (25 mL) was added by syringe, and
the solution was cooled to –78°C using an ethyl acetate –
liquid nitrogen slurry. To the cold, stirred solution was added
triflic anhydride (0.08 mL, 130 mg, 0.45 mmol), dropwise, by
syringe. Stirring was continued at –78°C, and the reaction
was monitored by TLC. When the reaction neared comple-
tion, it was removed from the ethyl acetate slurry. The
solution was extracted with saturated aqueous sodium bicar-
bonate (25 mL, 2 × 50 mL) and water (50 mL), then dried
and concentrated. The residue was purified using flash chro-
matography (3:1 to 3:2 Hex:EtOAc). The main product was
identified as phenyl 2,3,4-tri-O-benzyl-α-L-fucopyranosyl-
(1→3)-2-deoxy-2-phthalimido-1-thio-β-D-glucopyranoside in
Phenyl 2,3,4-tri-O-benzyl-α-L-fucopyranosyl-(1→3)-4,6-
O-benzylidene-2-deoxy-2-phthalimido-1-thio-β-D-
glucopyranoside (3)
Anhydrous compounds 1 (670 mg, 1.28 mmol) and 2
(420 mg, 0.851 mmol) were dissolved in dichloromethane
(40 mL) in a 2-necked, round-bottomed flask, which was
fitted with a rubber septum and an argon-filled balloon. This
solution was cooled to 0°C and a solution of dimethyl-
(methylthio)sulfonium triflate (0.88 g, 3.4 mmol) in dichloro-
methane (40 mL) was added via syringe, with stirring. The re-
action was monitored by TLC. After 1 h, the reaction was
quenched by adding triethylamine (5 mL), then filtered
through Celite. Dichloromethane (70 mL) was added, and the
solution extracted with saturated aqueous sodium bicarbonate
(2 × 100 mL) and water (100 mL), then dried and concen-
trated. The residue was separated by flash chromatography
(5:1 hexanes: ethyl acetate) to give the title compound
(0.47 g, 61%) plus a mixed fraction (0.18 g) in which 3 con-
stituted 50%: total yield 0.56 g, 73%; [α]D 7.1° (c 2.04,
1
31% yield. H NMR (500.13 MHz, CDCl3) δ: 1.09 (d, 3H,
CHCl3). 1H NMR (400.13 MHz, CDCl3) δ: 0.87 (d, 3H, J5,6
6.6 Hz, 3 H-6 Fuc), 3.47 (dd, 1H, J3,4 = 2.37 Hz, J4,5
1.27 Hz, H-4 Fuc), 3.66–3.76 (m, 4H, H-4 Glc, H-5 Glc, H-
2 Fuc, H-3 Fuc), 3.83 (t, 1H, J5,6ax ≈ J6ax,6eq = 9.5 Hz, H-6ax
Glc), 3.84, 4.26 (2 d, 2H, J = 12.6, CH2Ph), 4.04 (dq, 1H,
=
=
J5,6 = 6.5 Hz, 3 H-6 Fuc), 3.38, 4.13 (2 leaning d, 2H, J =
13.0 Hz, -CH2Bn), 3.53 (dd, 1H, J3,4 = 2.5 Hz, J4,5 = 1.2 Hz,
H-4 Fuc), 3.57 (bt, 1H, J3,4 ≈ J4,5 = 9.1 Hz, H-4 Glc), 3.64
(ddd, 1H, J5,6 = 4.9 Hz, J5,6′ = 3.4 Hz, J4,5 = 8.3 Hz, H-5
Glc), 3.74 (dd, 1H, J1,2 = 3.3 Hz, J2,3 = 10.3 Hz, H-2 Fuc),
3.78 (dd, 1H, J2,3 = 10.3 Hz, J3,4 = 2.5 Hz, H-3 Fuc), 3.87
(dd, 1H, J5,6 = 4.9 Hz, J6,6′ = 11.8 Hz, H-6 Glc), 4.00 (dd,
1H, J5,6′ = 3.4 Hz, J6,6′ = 11.8 Hz, H-6′ Glc), 4.07 (dq, 1H,
J4,5 = 1.1 Hz, J5,6 = 6.6 Hz, H-5 Fuc), 4.38, 4.43 (2 d, 2H, J =
11.6 Hz, CH2Ph), 4.43 (dd, J1,2 = 10.6 Hz, J2,3 = 9.9 Hz, H-2
Glc), 4.43 (dd, 1H, H-6eq Glc), 4.49, 4.79 (2 d, 2H, J =
11.6 Hz, CH2Ph), 4.81 (d, 1H, J1,2 = 2.9 Hz, H-1 Fuc), 5.55
(s, 1H, PhCH(OR)2), 5.74 (d, 1H, J1,2 = 10.7 Hz, H-1 Glc),
7.0–7.8 (m, aromatics). 13C NMR (100.62 MHz, CDCl3) δ:
16.57 (C6 Fuc), 54.98 (C4 Glc), 67.51 (C5 Fuc), 68.77 (C6
Glc), 70.79 (C5 Glc), 72.93 (CH2Bn), 73.29 (CH2Bn), 74.92
(CH2Bn), 75.84 (C4 Glc), 76.72 (C3 Glc), 78.28 (C4 Fuc),
79.83 (C2 Fuc), 82.12 (C3 Fuc), 84.45 (C1 Glc), 99.64 (C1
Fuc), 101.36 (PhCH(OR)2), 123.50, 126.30, 127.61, 127.66,
128.05, 128.28, 128.38, 128.40, 128.48, 128.57, 128.99,
129.15, 133.04, 134.06, 138.55 (aromatics), 167.76, 168.52
(2 C=O of phthalimide). HR-MS calcd. for C54H51NO10S +
Na: 928.3131; found: 928.3130.
J4,5 = 1.0 Hz, J5,6 = 6.6 Hz, H-5 Fuc), 4.25 (dd, 1H, J2,3
10.4 Hz, J3,4 = 8.1 Hz, H-3 Glc), 4.33 (t, 1H, J1,2 = J2,3
=
=
10.3 Hz, H-2 Glc), 4.55, 4.89 (2 leaning d, 2H, J = 11.4 Hz,
-CH2Bn), 4.56, 4.63 (2 leaning d, 2H, J = 11.8 Hz, -CH2Bn),
4.62 (d, 1H, J1,2 = 3.3 Hz, H-1 Fuc), 5.78 (d, 1H, J1,2
=
10.3 Hz, H-1 Glc), 6.9–7.9 (m, aromatics). 13C NMR
(62.9 MHz, CDCl3) δ: 16.58 (C6 Fuc), 53.67 (C2 Glc),
62.77 (C6 Glc), 68.70 (C5 Fuc), 71.66 (C4 Glc), 72.55
(CH2Bn), 73.54 (CH2Bn), 73.88 (C2 Fuc), 74.86 (CH2Bn),
77.92 (C4 Fuc), 79.01 (C3 Fuc), 79.40 (C5 Glc), 83.33 (C1
Glc), 84.03 (C3 Glc), 101.04 (C1 Fuc), 123.15, 123.33, 127.65,
127.74, 128.29, 128.38, 128.49, 128.56, 129.12, 132.88,
© 2003 NRC Canada

374
Can. J. Chem. Vol. 81, 2003
133.90, 138.16, 138.36, 138.81 (aromatics), 167.80, 168.76 (2
C=O of phthalimide). HR-MS calcd. for C47H47NO10S + Na:
840.2818; found: 840.2824.
2.6 Hz, J2,3 = 10.2 Hz, H-3 Fuc), H-07–4.17 (complex m, H-
2, H-5 Fuc), 4.618, 4.963 (2 d, 2H, J = 11.5 Hz, -CH2Ph),
4.750, 4.779 (AB quartet, 2H, J = 12.6 Hz, -CH2Bn), 4.636,
4.840 (2 d, 2H, J = 11.5, -CH2Ph), 4.946 (d, 1H, J1,2
=
3.2 Hz, H-1 Fuc), 5.056 (d, 1H, J1,2 = 9.4 Hz, H-1 Glc),
5.530 (d, 1H, J1,2 = 7.02 Hz, NH), 7.2–7.5 (m, aromatics).
13C NMR (100.62 MHz, CDCl3) δ: 16.88 (C-6 Fuc), 23.23
(NCOCH3), 54.55 (C-2 Glc), 63.00 (C-6 Glc), 68.63 (C-5
Fuc), 71.19 (C-4 Glc) 72.67, 74.32, 75.04 (3 OCH2Ph),
76.65 (C-3 Glc), 76.48 (C-2 Fuc), 77.55 (C-4 Fuc), 79.33
(C-5 Glc), 79.39 (C-3 Fuc), 85.96 (C-3 Glc), 86.38 (C-1
Glc), 99.96 (C-1 Fuc), 127.9–129.2, 131.9 (PhCH), 133.66,
138.51, 138.45,138.67 (quat C Ph), 171.16 (C=O). HR-MS
calcd. for C41H47NO9S + Na: 752.2869; found: 752.2877.
Phenyl 2,3,4-tri-O-benzyl-α-L-fucopyranosyl-(1→3)-2-
acetamido-4,6-O-benzylidene-2-deoxy-1-thio-β-D-
glucopyranoside (5)
To a stirred solution of 3 (0.43 g, 0.47 mmol) in anhydrous
1-butanol (80 mL) was added ethylenediamine (15 mL,
13.5 g, 220 mmol). Argon was bubbled through the solution
and stirring continued for 5.5 h, monitoring the reaction by
TLC with a ninhydrin spray reagent (0.25% in ethanol). The
mixture was then evaporated to dryness. Tolune (2 × 25 mL)
and ethanol (25 mL) were in turn added and evaporated. To
the residue was added methanol (60 mL), resulting in a het-
erogeneous mixture, which dissolved upon addition of trie-
thylamine (1 mL, 7.2 mmol) and acetic anhydride (8 mL,
85 mmol) with stirring. After continued stirring for 19 h, the
reaction was added to 250 mL ice water, which melted and
was extracted with dichloromethane (3 × 100 mL). The or-
ganic layers were dried and concentrated. Purification by
Methyl 2,3,4-tri-O-benzyl-α-L-fucopyranoside (7)
To a stirred solution of 1 (250 mg, 0.47 mmol) in dry meth-
anol (2 mL) was added iodine (190 mg, 0.75 mmol). The
flask was fitted with a silica gel drying tube and stirred for
4.5 h. Methanol (5 mL) and Rexyn 201 (OH) ion-exchange
resin (4 mL) was added to the flask, and stirring was contin-
ued until the solution became virtually colorless. Filtration,
concentration, and purification by flash chromatography
yielded a syrup (150 mg, 70%). 1H NMR (400.14 MHz,
CDCl3) δ: 1.114 (d, 3H, J5,6 = 6.6 Hz, 3 H-6 Fuc), 3.350 (s,
3H, OMe), 3.630 (dd, 1H, J3,4 = 2.8 Hz, J4,5 = 1.1 Hz, H-4),
3.828 (qd, 1H, H-5), 3.924 (dd, 1H, J2,3 = 10.1 Hz, H-3),
4.035 (dd, 1H, J1,2 = 3.7 Hz, H-2), 4.649, 4.978 (2 d, 2H, J =
11.6 Hz, C-4 -OCH2Ph), 4.653 (d, 1H, H-1), 4.687, 4.821 (2d,
2H, J = 12.1 Hz, C-2-OCH2Ph), 4.730, 4.872 (2d, 2H, J =
11.9 Hz, C-3-OCH2Ph), 7.22–7.41(m, Ph H). 13C NMR
(62.9 MHz, CDCl3) δ: 16.63 (C-6), 55.31 (OMe), 66.08 (C-5),
73.36 (C3- OCH2Ph), 73.48 (C2- OCH2Ph), 74.83 (C4-
OCH2Ph), 76.33 (C-2), 77.83 (C-4), 79.41 (C-3), 98.81 (C-1),
127.5–128.4 (PhC), 138.57 (C-2 quat PhC), 138.62 (C-4 quat
PhC), 138.97 (C-3 quat PhC). HR-MS calcd. for C28H32O5 +
Na: 471.2147; found: 471.2147.
1
flash chromatography yielded a white solid (0.22 g, 57%). H
NMR (400.14 MHz, CDCl3) δ: 0.831 (d, 3H, J5,6 = 6.4 Hz, 3
H-6 Fuc), 1.659 (s, 3H, NAc), 3.5–3.65 (complex m, 4H, H-4
Fuc, H-2, H-4, H-5 Glc), 3.774 (t, 1H, J5,6ax = J6ax,6eq
10.0 Hz, H-6ax Glc), 3.934 (dd, J2,3 = 10.1 Hz, J3,4
=
=
2.7 Hz, H-3 Fuc), 4.084 (dd, H-2 Fuc), 4.093 (br q, H-5
Fuc), 4.289 (br t, J2,3 = J3,4 = 9.0 Hz, H-3 Glc), 4.363 (dd,
J5,6eq = 4.5 Hz, J6ax,6eq = 10.5 Hz, H-6eq), 4.562, 4.605 (2 d,
2H, J = 11.5, 2 CH2Ph H), 4.754 (s, 2H, CH2Ph), 4.931 (d,
2H, J = 11.5, 2 CH2Ph H), 5.102 (d, 1H, J1,2 = 3.5 Hz, H-1
Fuc), 5.193 (broad d, 1H, J1,2 = 10.4 Hz, H-1 Glc), 5.518 (s,
1H, O2CHPh), 5.741 (d, 1H, J1,2 = 7.7 Hz, NH), 7.2–7.6 (m,
Ph H). 13C NMR (62.9 MHz, CDCl3) δ: 16.42 (C-6 Fuc),
23.53 (NCOCH3), 56.84 (C-2 Glc), 67.03 (C-5 Fuc), 68.79
(C-6 Glc), 70.64 (C-5 Glc), 72.67, 74.32, 75.04 (3 OCH2Ph),
76.65 (C-3 Glc), 77.16 (C-2 Fuc), 77.55 (C-4 Fuc), 79.93
(C-3 Fuc), 80.64 (C-4 Glc), 86.70 (C-1 Glc), 98.39 (C-1
Fuc), 101.61 (O2CHPh), 126.3–129.0, 132.5 (PhCH), 137.3,
138.7 (quat C Ph), 170.6 (C=O). HR-MS calcd. for
C48H51NO9S + Na: 840.3182; found: 840.3188.
Acknowledgments
JW thanks the Natural Sciences and Engineering Research
Council of Canada (NSERC) for the award of an NSERC
summer research assistantship. TBG is grateful to NSERC
for a research grant and to the Atlantic Region Magnetic
Resonance Centre for NMR time, and to Dr. Angelina
Morales-Izquierdo at the University of Alberta for determin-
ing accurate masses. JJB thanks Direccion General de
Investigacion Cientifica y Tecnica for financial support
(BQU2000–1501-C02–01) and AP is grateful to Servicio
Interdepartamental de Investigacion de la Universidad
Autonoma de Madrid for the facilities provided.
Phenyl 2,3,4-tri-O-benzyl-α-L-fucopyranosyl-(1→3)-2-
acetamido-2-deoxy-1-thio-β-D-glucopyranoside (6)
Compound 5 (0.13 g, 0.16 mmol) was dissolved in a solu-
tion of ethanol (150 mL) and water (1 mL). Sulfuric acid
(0.18 g, 1.8 mmol) was added with stirring. The resulting so-
lution was heated to reflux for 2 h, then cooled. Solid
NaHCO3 (0.31 g, 3.7 mmol) was added and the solvent re-
moved under reduced pressure. The residue was partitioned
between water (100 mL) and CH2Cl2 (100 mL) and the
aqueous layer extracted with two further portions of CH2Cl2
(100 mL each). Combined organic layers were dried and
concentrated, and the product was purified by flash chroma-
tography (2:1 to 1:1 hexanes:ethyl acetate) to yield a white
solid (0.10 g, 86%). 1H NMR (250.13 MHz, CDCl3) δ: 1.155
(d, 3H, J5,6 = 6.6 Hz, 3 H-6 Fuc), 1.612 (s, 3H, Nac), 3.42
(complex m, H-5 Glc), 3.47 (t, 1H, J3,4 = J4,5 = 9.5 Hz, H-4
Glc), 3.58–3.73 (complex m, 3H, H-2, H-3 Glc, H-4 Fuc),
References
1. L. Poppe, G.S. Brown, J.S. Philo, P.V. Nikrad, and B.H. Shah.
J. Am. Chem. Soc. 119, 1727 (1997).
2. A. Poveda and J. Jiménez-Barbero. Chem. Soc. Rev. 27, 133
(1998).
3. J. Jiménez-Barbero, J.L. Asensio, F.J. Cañada, and A. Poveda.
Curr. Opin. Struct. Biol. 9, 549 (1999).
3.769 (dd, 1H, J5,6 = 4.7 Hz, H-6 Glc), 3.907 (dd, J5,6′
2.9 Hz, J6,6′ = 11.8 Hz, H-6′ Glc), 3.932 (dd, 1H, J3,4
=
=
© 2003 NRC Canada

Watts et al.
375
4. R. Harris, G.R. Kiddle, R.A. Field, M.J. Milton, B. Ernst, J.L.
Magnani, and S.W. Homans. J. Am. Chem. Soc. 121, 2546
(1999).
5. J.F. Espinosa, F.J. Cañada, J.L. Asensio, H. Dietrich, M. Martín-
Lomas, R.R. Schmidt, and J. Jiménez-Barbero. Angew. Chem.
Int. Ed. Eng. 35, 303 (1996).
20. T.B.Grindley. Structure and conformation of carbohydrates. In
Glycoscience: Chemistry and chemical biology. Edited by
B.Fraser-Reid, K.Tatsuta, and J.Thiem. Springer–Verlag, Hei-
delberg. 2001. p. 3.
21. A. Imberty, E. Mikros, J. Koca, R. Mollicone, R. Oriol, and S.
Pérez. Glycoconjugate J. 12, 331 (1995).
6. J.F. Espinosa, E. Montero, A. Vian, J.L. Garcia, H. Dietrich,
R.R. Schmidt, M. Martín-Lomas, A. Imberty, F.J. Cañada, and
J. Jiménez-Barbero. J. Am. Chem. Soc. 120, 1309 (1998).
7. M.J. Milton and D.R. Bundle. J. Am. Chem. Soc. 120, 10 547
(1998).
8. T. Haselhorst, J.F. Espinosa, J. Jiménez-Barbero, T. Sokolowski,
P. Kosma, H. Brade, L. Brade, and T. Peters. Biochemistry, 38,
6449 (1999).
22. B. Aguilera, J. Jiménez-Barbero, and A. Fernández-Mayoralas.
Carbohydr. Res. 308, 19 (1998).
23. J. Dabrowski, T. Kozar, H. Grosskurth, and N.E. Nifant′ev. J.
Am. Chem. Soc. 117, 5534 (1995).
24. J.L. Asensio and J. Jiménez-Barbero. Biopolymers, 35, 55
(1995).
25. T. Haselhorst, T. Weimar, and T. Peters. J. Am. Chem. Soc.
123, 10 705 (2001).
9. L.A. Lasky. Science (Washington, D.C.), 258, 964 (1992).
10. D.S. Johnston, W.W. Wright, J.H. Shaper, C.H. Hokke, D.H.
Van den Eijnden, and D.H. Joziasse. J. Biol. Chem. 273, 1888
(1998).
11. K. Bock and H. Thøgersen. Annu. Rep. NMR Spectrosc. 13, 1
(1982).
26. G.I. Csonka, C.P. Sosa, and I.G. Csizmadia. J. Phys. Chem. A,
104, 3381 (2000).
27. B. Guilbert, N.J. Davis, M. Pearce, R.T. Aplin, and S.L.
Flitsch. Tetrahedron: Asymmetry, 5, 2163 (1994).
28. C. Chipot, R. Jaffe, B. Maigret, D.A. Pearlman, and P.A.
Kollman. J. Am. Chem. Soc. 118, 11 217 (1996).
29. F. Tran, J. Weber, and T.A. Wesolowski. Helv. Chim. Acta, 84,
1489 (2001).
12. S. Komba, H. Ishida, M. Kiso, and A. Hasegawa. Bioorg. Med.
Chem. 4, 1833 (1996).
13. R.K. Jain and K.L. Matta. Carbohydr. Res. 226, 91 (1992).
14. D. Kahne, S. Walker, Y. Cheng, and D. Van Engen. J. Am.
Chem. Soc. 111, 6881 (1989).
15. L. Yan and D. Kahne. J. Am. Chem. Soc. 118, 9239 (1996).
16. Z. Zhang, I.R. Ollmann, X.-S. Ye, R. Wischnat, T. Baasov, and
C.-H. Wong. J. Am. Chem. Soc. 121, 734 (1999).
17. M. Ravenscroft, R.M.G. Roberts, and J.G. Tillett. J. Chem.
Soc., Perkin Trans. 2, 1569 (1982).
18. F. Andersson, P. Fügedi, P.J. Garegg, and M.A. Nashed. Tetra-
hedron Lett. 27, 3919 (1986).
19. K.P.R. Kartha, M. Aloui, and R.A. Field. Tetrahedron Lett. 37,
5175 (1996).
30. V. Spirko, O. Engvist, P. Soldan, H.L. Selzle, E.W. Schlag, and
P. Hobza. J. Chem. Phys. 111, 572 (1999).
31. P. Hobza, H.L. Selzle, and E.W. Schlag. J. Phys. Chem. 100,
18 790 (1996).
32. T.L. Hwang and A.J. Shaka. J. Am. Chem. Soc. 114, 3157
(1992).
33. F. Mohamadi, N.G.J. Richards, W.C. Guida, R. Liskamp, M.
Lipton, C. Caufield, G. Chang, T. Hendrickson, and W.C. Still.
J. Comput. Chem. 11, 440 (1990).
34. N.L. Allinger, Y.H. Yuh, and J.-H. Lii. J. Am. Chem. Soc. 111,
8551 (1989).
© 2003 NRC Canada