Synthesis and conformational analysis of (α-D-Galactosyl) phenylmethane and α-,β-difluoromethane analogues: interactions with the plant lectin viscumin

Article abstract of DOI:10.1002/chem.200801394

(aα-D-Galactosyl)phenylmethane (1), (α- and β-D-galactosyl)-(difluoro)phenylmethane (2 and 3) have been prepared and their conformations in solution were described by using a combination of force-field calculations and NMR spectroscopic studies. Galactosi

Full text of DOI:10.1002/chem.200801394

FULL PAPER
DOI: 10.1002/chem.200801394
Synthesis and Conformational Analysis of (a-d-Galactosyl)phenylmethane
and a-,b-Difluoromethane Analogues: Interactions with the Plant Lectin
Viscumin
Maria Kolympadi,[a] Marco Fontanella,[b] Chiara Venturi,[b] Sabine Andrꢀ,[c]
Hans-Joachim Gabius,[c] Jesffls Jimꢀnez-Barbero,*[b] and Pierre Vogel*[a]
1
Abstract: (a-d-Galactosyl)phenylme-
thane (1), (a- and b-d-galactosyl)-
galactosides. The X-ray crystal struc-
ture of its difluoromethylene derivative
2 similarly shows a C1 chair conforma-
chair and S3 skew boat conformations
and significant flexibility around the
pseudoglycosidic linkage when in solu-
tion. The b-stereoisomer 3 adopts a
4
A
have been prepared and their confor-
mations in solution were described by
using a combination of force-field cal-
culations and NMR spectroscopic stud-
ies. Galactoside 1 adopts a 4C1 chair
conformation and an exo anomeric ori-
entation, as is the case for natural a-
tion. Surprisingly, compound 2 exhibits
a different equilibrium between 1C4
major C1 chair conformation. Interest-
4
ingly, C-galactosides 1, 2, and 3 bind to
viscumin (VAA), a galactoside-specific
lectin, which is confirmed by NMR ex-
periments and docking calculations.
Keywords: carbohydrate mimics ·
lectin · molecular modeling · NMR
spectroscopy · protein structures
Introduction
The protein superfamily, by virtue of its effector functionali-
ty, is thus becoming an attractive target for drug design, for
example, to block harmful effects of pro-inflammatory medi-
ators or interfere with cell binding of toxic lectins. In the
latter case, viscumin, also referred to as Viscum album ag-
glutinin (VAA), is a potent biohazard akin to ricin, whose
toxicity can be neutralized by suitable inhibitors competing
with the cell-surface glycans.[3] This has made the toxin a
model for drug design, for example, applying library ap-
proaches in search of new binding partners.[4] On a broad
scale, the study of potent lectin counter-receptors will not
only provide a source for new pharmaceuticals, but also val-
uable insights into the mechanisms of protein–carbohydrate
interactions.[2b,5]
Recognition processes involving carbohydrate moieties of
cellular glycoconjugates have a significant impact on differ-
ent aspects of cell biology as they are involved in mediating
communication between cells and their environment.[1] De-
coding of the glycan signals is performed by lectins (carbo-
hydrate-binding proteins without enzymatic activity, exclud-
ing antibodies and transport proteins for free glycans).[1,2]
[a] M. Kolympadi, Prof. Dr. P. Vogel
Laboratoire de Glycochimie et Synthꢀse Asymꢁtrique (LGSA)
Swiss Federal Institute of Technology (EPFL)
Batochime, 1015 Lausanne (Switzerland)
Fax : (+41)216-939-375
Carbohydrate mimics such as C-glycosides[6] in which the
anomeric oxygen has been replaced by a methylene or a
substituted methylene unit, exhibit stability against chemical
and enzymatic hydrolysis and, depending on substitution,
improved pharmacokinetic properties over the natural sac-
charides. Beyond these advantages, of concern is whether
these pseudosaccharides maintain an affinity towards the
same receptors, taking into account that such an alteration
of the glycosidic bond can lead to conformational changes
of the intersaccharide torsions and may also affect the sac-
charide ring. As we previously reported,[7] C-analogues are
more flexible, possibly owing to the absence of the exo-
E-mail: pierre.vogel@epfl.ch
[b] Dr. M. Fontanella, Dr. C. Venturi, Prof. Dr. J. Jimꢁnez-Barbero
Centro de Investigaciones Biolꢂgicas, CSIC
Ramiro de Maeztu 9, 28040 Madrid (Spain)
Fax : (+34)915-360-432
E-mail: jjbarbero@cib.csic.es
[c] Priv.-Doz. Dr. S. Andrꢁ, Prof. Dr. H.-J. Gabius
Institute of Physiological Chemistry
Faculty of Veterinary Medicine, Ludwig-Maximilians-University
Veterinꢃrstrasse 13, 80539 Munich (Germany)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200801394.
Chem. Eur. J. 2009, 15, 2861 – 2873
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2861

anomeric effect, and show a distribution of different con-
formers. Nevertheless, if the energy barriers between these
conformers are small, a global-minimum-bound conforma-
tion with the receptor protein might be achieved, thus ex-
ceeding the entropic penalty.[2b,8] Whilst in cases where the
bound conformation resembles the ground state of O-glyco-
sides, C-glycosides with closer conformation behavior to O-
glycosides would be desirable.[9]
Introduction of fluorine atoms at the pseudoanomeric
center of a C-saccharide can “correct” the polarity of this
position relative to the native glycosidic oxygen and increase
the lipophilicity of the entire molecule.[10] As a matter of
fact, CF2 and CHF groups have been examined as isosteres
of oxygen in biologically active compounds[11] and the
gauche conformation of 2-fluoroethanol and related struc-
tures is well documented.[12] In the course of our search for
sugar mimics, we are interested in the preparation of CF2-
galactopyranosides, and the exploration of their conforma-
tion profile. Synthetic methods towards CF2-furanosides[13]
and pyranosides[14] have been described by different re-
search groups. We have reported the preparation of a CHF-
disaccharide of glucose whose study in solution revealed un-
usual conformations to exist in equilibrium with “normal”
conformations for C-glycosides.[14b] Mootoo and co-workers
developed a de novo synthesis of b-CF2- and b-CHF-galacto-
pyranosides,[14h] thus preparing three fluorinated C-disac-
charides. These mimics of sialyl LeX and their conformation
in solution, along with their activity towards the C-type
lectin P-selectin involved in inflammation were investigat-
their capacity as bioactive ligands, interactions of 1, 2, and 3
with the galactoside-binding VAA lectin, which has two
combining sites and is active with a and b anomers,[18] were
monitored by NMR experiments and docking calculations.
Results and Discussion
Synthesis: Key intermediates in the parallel synthesis of 1, 2,
and 3 are the a- and b-d-C-galactosyl carbaldehydes 7 and
8, respectively,[19] which were prepared by using a modified
sequence described by Bednarski and co-workers[19a]
(Scheme 1). Reaction of galactose pentaacetate with propar-
gyltrimethylsilane and Lewis acids led exclusively to the a-
allene 4.[20] The exchange of acetate protecting groups to
benzyl ethers and ozonolysis afforded the a-aldehyde 7 that
isomerized to the b-stereoisomer 8 under basic condi-
AHCTUNGTRENNUNG
tions.[19a] Both crude aldehydes 7 and 8 were converted to
the corresponding phenyl ketones 9 and 10[21] by addition of
phenylmagnesium bromide followed by oxidation.[16b,22] De-
protection of the benzyl ethers and reduction of the ketone
group of the intermediate 9 by catalytic hydrogenolysis[16b]
gave the (a-d-galactosyl)phenylmethane 1. Difluorination of
a-phenyl ketone 9 was achieved by using neat Deoxo-Fluor
ACHTUNGTRENNUNG
ed.[14i,15] Moreover, biologically useful pseudoglycopeptides of
CF2-galactopyranosylesters were reported by Leclerc, Quiri-
on, and co-workers,[14a,c,f,g,k] as well as CF2-sialylgalactose by
Sodeoka and co-workers.[14j]
Herein, we present the syn-
thesis and the conformational
analysis of three simple phenyl
C-galactopyranosides: (a-d-gal-
actosyl)phenylmethane 1, (a-
and
b-d-galactosyl)-
ACHTUNGTRENNUNG
3, respectively. Syntheses of the
glucose and mannose series of 1
are already known.[16] To the
best of our knowledge, 2 is the
first example of a-CF2-galacto-
sides,[17] however, a protected
form of 3 was also recently de-
scribed.[14h] We performed
conformational analysis
a
of
these molecules by using a com-
bination of molecular dynamics
and NMR spectroscopy and
have shown that chemical
changes at the pseudoanomeric
center have important effects
on the behavior of these C-gal-
actosides in solution. To study
Scheme 1. Parallel Synthesis of compounds 1, 2, and 3. a) propargyltrimethylsilane, BF3·OEt2, TMSOTf,
CH3CN, 80%; b) NaOMe, MeOH, 90%; c) NaH, BnBr, nBu4NI (cat.), DMF, 78%; d) O3, CH2Cl2; e) Et3N,
iPrOH, CH2Cl2; f) i) PhMgBr, Et2O, then aqueous work-up, ii) PCC, CH2Cl2 9: 50% for three steps, 10: 25%
for four steps; g) H2, Pd/C 10% (cat.), MeOH, EtOAc, AcOH, 77%; h) neat Deoxo-Fluor, HF-pyridine (cat.),
808C, 30 h, 77% (+8% SM); i) solution 50% Deoxo-Fluor in THF, HF-pyridine (cat.), 50–708C, 20 h, 23%
(+28% SM); j) H2, Pd/C 10%, EtOH 2: 70%, 3: 67%. Deoxo-Fluor=bis(2-methoxyethyl)aminosulfur tri-
fluoride, DMF=N,N-dimethylformamide, PCC=pyridinium chlorochromate, TMSOTf=trimethylsilyl
trifluoroACTHNUTRGNEUmGN ethanesulfonate.
2862
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2861 – 2873

Analysis of (a-d-Galactosyl)phenylmethane and a-,b-Difluoromethane Analogues
FULL PAPER
reagent,[23] a catalytic amount of HF-pyridine, and heating
to 808C for 30 h to afford the a-CF2-protected sugar 11 in
good yield. Under the same conditions of fluorination, the
b-phenyl ketone 10 afforded the desired b-CF2 analogue 12
in only 25% yield along with 5% unreacted starting materi-
al. No other product from the reaction mixture could be
identified. In an attempt to improve the yield, b-phenyl
ketone 10 was dissolved in a solution of 50% Deoxo-Fluor
in THF and was slowly heated from 508C to 708C to give
the b-CF2 analogue 12 in 23% yield along with 28% of
starting material. Diluted Deoxo-Fluor in CH2Cl2 at 258C
resulted in a very slow conversion of 10 to 12. All further at-
tempts to improve the yield of the above-mentioned reac-
tion were unsuccessful. Finally,
Table 1. Chemical shifts and coupling constants of compound
CD3OD (500 MHz, 298 K) and D2O (500 MHz, 298 K).
1 in
H
d(D2O)
J
(D2O)
d
G
JACHTUNGTRENNUNG
[ppm]
[Hz]
ACHTUNGTRENNUNG
H1
H2
H3
H4
4.261
4.040
3.896
4.031
4.021
3.643
3.597
4.5 (H2), 3.5, 11.0
4.5, 10.0
3.5 (H4), 10.0 (H2) 3.844
overlap
overlap
5.0 (H5), À11.5
7.5 (H5), À11.5
11.0, À14.5
3.5,À14.5
–
4.235
3.983
5.0 (H2), 4.5, 9.5
5.0, 8.8 (H3)
8.8, 3.0 (H4)
4.102
4.021
3.829
3.684
3.046
3.021
7.34
3.0, <0.5 (H5)
H5
H6a
H6b
CH2a 2.991
CH2b 2.978
Harom 7.33
5.2 (H6a), 7.0 (H6b)
5.2, À11.3 (H6b)
7.0, À11.3
9.5 (H1), À14.5
4.5 (H1), À14.5
–
both 11 and 12 were benzyl de-
protected by hydrogenolysis to
Table 2. Comparison between the experimental vicinal coupling constant values of 1 in CD3OD (Jexp) and
1
1
those calculated for the most stable conformations of the six-membered ring (4C1, C4, S3) according to MM3*
give (a- and b-d-galactosyl)-
calculations (Jtheor).
ACHTUNGTRENNUNG(difluoro)phenylmethane 2 and
Coupling
constant
1C4
H-C-C-H
torsion[b]
4C1
H-C-C-H
torsion[b]
1S3
H-C-C-H
torsion[b]
3, respectively, which together
with compound 1 were used for
NMR spectroscopy studies in
solution.
DE[a]
9.81
Jtheor
0.00
Jtheor
3.17
Jtheor
AHCTUNGTERNNUNG ]
[kJmolÀ1
Jexp[c] [Hz]
CD3OD
[b]
[b]
[b]
[Hz]
[Hz]
[Hz]
JH1,H2
JH2,H3
JH3,H4
JH4,H5
JH5,H6b
JH5,H6a
JH1,CH2a
JH1,CH2b
5.0
8.8
3.0
<0.5
7.0
5.2
2.4
1.4
5.6
3.4
9.5
4.2
11.0
1.7
À55
À67
À53
49
À171
À64
À177
63
3.9
8.7
5.0
0.1
10.5
4.2
43
172
54
À57
À176
À65
174
54
2.1
7.2
3.9
2.4
9.5
4.2
10.9
1.5
À33
À152
62
À33
À171
À66
À175
64
Conformational analysis
Compound 1: The analysis of
the vicinal proton–proton cou-
pling constants (Table 1 and
Table 2) for the pyranose ring
of 1 indicates a major presence
of the 4C1 chair conformation,
as expected for natural galacto-
sides.[24] The coupling values are
very similar in CD3OD and in
9.5
4.5
11.4
2.6
[a] The relative steric energy difference calculated by MM3* between the global minimum (4C1) and the other
two major conformers is also given. [b] The theoretical values (Jtheor) as well as the torsion angles were ob-
tained by applying the generalized Karplus equation proposed by Altona to the geometries calculated by
MM3* calculations. The gt conformer around the C5-C6 torsion was employed. [c] The comparison between
the coupling constants indicates a predominant presence of the 4C1 chair conformation. Moreover, there is a
predominant conformation around the pseudoglycosidic linkage and around the C5-C6 torsion. In the latter
case, there is also evidence for a non-negligible contribution of the alternative tg rotamer.
4
D2O solutions. Only this C1 ge-
ometry can explain the ob-
served couplings, whereas the
calculated predictions for the
1
1
alternative C4 chair or the S3 skew boat forms are far from
the observed values. Furthermore, the presence of one large
and one small–medium coupling constant between H1 and
the two methylene protons indicates a major orientation
around the glycosidic F torsion angle (defined as H1-C1-
CH2-Cipso). The molecular mechanics MM3* calculations[25]
also support these observations with the 4C1 chair form
Harom and H5/Harom NOE values for the aromatic protons,
can only be explained by a geometry in which the F torsion
is À608 (Figure 2). Nevertheless, the comparison between
the observed JH1, CH2a and JH1,CH2b with those calculated for a
pure F À608 geometry is not perfect (Table 2), indicating
the presence of rotamers (approximately 10–20%) with the
alternative F +608 geometry, which corresponds to a non-
exo-anomeric orientation. These geometries have been also
found in other C-glycosyl compounds.[28]
being more than 3 kJmolÀ1 more stable than the S3 skew
boat and much more stable than the alternative C4 chair
1
1
conformer. Regarding the orientation around the C5-C6 tor-
sional bond, the intermediate observed JH5,H6a and JH5,H6b
values (5.0–5.2 and 7.0–7.5 Hz) are in agreement with the
typical gt:tg equilibrium observed for galactoside deriva-
tives.[26]
Molecular dynamics simulations of 5 ns (with the MM3*
force field) were performed by using the global-minimum
geometry as the input (Figure 2 and the Supporting Infor-
mation). The results show that this conformer is fairly
stable, with fluctuations along the well-defined F torsion
value. Two sets of values (around +608 and À1208) are
found for the Y angle (defined as C1-CH2-Cipso-Cothro). These
values correspond to the symmetry that is inherent to a
monosubstituted phenyl ring, for which the two ortho posi-
tions are chemically equivalent. The calculation of the ex-
The analysis of the NOE values[27] (Figure 1) supports the
4
existence of the C1 chair geometry as does that of an exo-
anomeric-like orientation of the aromatic aglycon. The
strong H1/CH2b, H3/CH2a, and H5/CH2a NOE values for the
methylene protons, together with the medium–strong H1/
Chem. Eur. J. 2009, 15, 2861 – 2873
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2863

P. Vogel, J. Jimꢁnez-Barbero et al.
Thus, the recognition mainly in-
volved the galactose moiety.
Moreover, competitive STD ex-
periments by using lactose, a
natural substrate for VAA,
were performed. It was shown
that 1 competed with lactose
for the lectin site. The STD ef-
fects on 1 were clearly dimin-
ished when the haptenic inhibi-
tor (lactose) was added to the
NMR spectroscopy tube con-
taining the synthetic compound
and VAA. The binding affinity
of 1 to VAA was estimated to
be of the same order of magni-
tude to that of lactose, in the
millimolar range, as the STD
peak intensities on lactose were
similarly affected upon addition
of 1.
Figure 1. Section of the NOESY spectrum (500 MHz, 700 ms mixing time, D2O, 298 K) showing the key NOE
peaks and the 1D trace of the 1H NMR spectrum. The cross-peaks indicate a major 4C1 chair conformation as
well as a major rotamer around F angle (H1-C1-CH2-Cipso). The cross-peaks are positive (different sign as di-
agonal peaks) as is expected for a small molecule. Indeed, the observed cross-peaks can be quantitatively
fitted by using a correlation time of 50 ps.
Exchange transferred NOE
experiments (trNOE) experi-
ments were also applied to the
Table 3. Experimental NOE values for compound 1 in CD3OD in com-
parison with those estimated by applying a full relaxation matrix ap-
proach to the major conformation.
Proton
pair
NOEs%[a]
experimental
(700 ms)
calculated
(tc, 50 ps)
N
H1/Harom
H1/H2
H1/CH2b
H5/CH2a
H3/CH2a
H3/H4+H5
H5/Harom
H6a/H4+H5
H6b/H4+H5
1.6
5.0
2.9
4.2
4.3
7.3
1.3
4.0
2.4
3.5
5.0
3.2
3.4
3.7
6.9
1.1
2.5
1.6
Figure 2. The major conformer of 1 (depicted with and without hydrogen
atoms) according to the experimental NMR spectroscopic data and
MM3* molecular mechanics and dynamics calculations. This orientation
of the aromatic ring (defined by the Y angle) defines both orientations
deduced by the MD simulations (+608 and À1208) owing to the intrinsic
symmetry of the aromatic ring.
[a] The experiments (298 K) and the calculations (tc 50 ps, matched for
the intra-residual H1/H2) were performed at 500 MHz. The agreement
between experimental and theoretical values is satisfactory.
pected NOE values for the ensemble deduced by the MD
simulations permitted us to quantitatively explain the ob-
served NOE values (Table 3).[29]
According to the combined NMR spectroscopy and simu-
lation data, 1 behaves as a mimetic of a natural galactose
compound with a major C1 chair conformation and an exo-
1:VAA sample, which had a molar ratio of 25:1. Strong neg-
ative signals were observed for this sample, even for short
mixing times (100 and 200 ms), in contrast with the observa-
tion for the free state for which positive cross-peaks were
observed in the NOESY spectrum. This clearly indicates the
existence of binding and that the NOE peaks contain infor-
mation on the bound state. The pattern of NOE peaks is ba-
sically identical to that measured for the free ligand (see
Figure 1 and Figure 4), thus indicating that the lectin selects
the 4C1 major chair conformation of 1. Selective experiments
were carried out on specific signals of the ligand, namely the
aromatic protons, H1, and the combined H2+H4+H5 sig-
nals (Figure 4). In particular, inversion of the anomeric
proton H1 showed NOE signals at H2 and the methylene
4
anomeric orientation around F. To investigate its bioactivity,
saturation transfer difference (STD) NMR spectra[30] of 1 to-
gether with the galactoside-specific VAA lectin were carried
out in the molar ratio 1:VAA of 50:1 (Figure 3). VAA
lectin, which was isolated from the extracts of mistletoe,
generally recognizes galactose moieties in a- and b-anome-
ric position and tolerates an aromatic aglycon well.[3a,18a,31]
The STD experiments enabled us to deduce conclusions for
the interactions between 1 and VAA lectin. Major satura-
tion transfer to the protons in the pyranose ring was ob-
served with basically no transfer to the methylene protons.
Similar STD effects on the galactose peaks were observed.
2864
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2861 – 2873

Analysis of (a-d-Galactosyl)phenylmethane and a-,b-Difluoromethane Analogues
FULL PAPER
Figure 4. Selective 1D-trNOE experiments performed for a 25:1 mixture
of 1:VAA. The inverted signal in each case displays the highest intensity.
A) Inversion of the aromatic protons. B) Inversion of the anomeric
proton H1. C) Inversion of the resonance signal caused by H2, H4, and
H5. The NOE signals have the same sign as the inverted signal, indicat-
ing binding of 1 to the lectin. The NOE pattern is basically identical to
that observed in the free state (see NOESY in Figure 1), suggesting the
Figure 3. Compound 1 interacts with VAA lectin in an analogous manner
to that of lactose, and with similar affinity, in the millimolar range. The
concentration of the protein was approximately 25 mm. A) Section of the
500 MHz 1H NMR spectrum of 1 B) STD spectrum of 1 (2 s saturation
time) in the presence of VAA. The 1:VAA molar ratio is 50:1. C) The
STD spectrum of a 3:1 mixture of lactose:1 added to the NMR tube con-
taining VAA. The 1:VAA molar ratio is again 50:1. D) The STD spec-
trum of a 100:1 mixture of lactose:VAA. The off-resonance frequency
was set at 50 ppm while the on-resonance frequency was established at
À0.5 ppm. The STD percentages are shown above in the corresponding
caption of 1.
4
presence of a bound C1 conformation.
Table 4. Chemical shifts and coupling constants of compound
CD3OD (800 MHz, 298 K).[a]
2 in
H
d [ppm]
J [Hz]
H1
H2
H3
H4
4.401
3.880
3.852
4.161
3.993
3.834
3.685
7.583
7.442
3.2 (H2), 14.4 (F1), 16.8 (F2)
3.2, 6.4 (H3)
6.4, 3.2 (H4)
3.2, 4.8 (H5)
1
protons, but not at H6ab, suggesting that the C4 chair is not
H5
4.8, 4.0 (H6b), 7.2 (H6a)
bound. On the other hand, inversion of the aromatic protons
showed negative NOE peaks at H1 and H2 protons, but
these signals were very weak. This observation suggests that
the aromatic protons still display flexibility in the bound
state, probably owing to the lack of contact with the protein.
This hypothesis was supported by docking calculations (see
below).
H6a
H6b
Hortho
Hmeta+para
7.2, À12.0
4.0, À12.0
–
–
[a] The 1H NMR spectrum of 2 in D2O contains significant overlapping
for the key peaks of the pyranose ring even at 800 MHz.
To rationalize the interaction on the molecular level, the
low-energy conformer, as observed by NMR spectroscopy,
was docked into the VAA binding sites by the program AU-
TODOCK. Two sites are known in the lectin dimer, charac-
terized by Trp38 in the 1a subdomain and by Tyr249 in the
1
1
fit is found when a combination of major C4 and minor S3
conformers are considered. Strikingly, the contribution of
the regular 4C1 present in natural galactosides is basically
negligible.[32] This result is in complete contrast with the ob-
servations for the related GalaCH2Ph analogue 1. No con-
clusions about the conformation around the glycosidic link-
age could be drawn owing to the lack of hydrogen atoms at
the pseudoglycosidic CF2 moiety. However, the presence of
two similar JH1,F coupling constants (14.4 and 16.8 Hz) is in
agreement with a high degree of conformational averaging
around this F linkage (defined as H1-C1-CF2-Cipso). The
conformation around F for a similar ManaCF2R analogue
was also recently described as a 77:23 exo/non-exo equilibri-
um, giving two coupling values of 14.3 and 19.4 Hz.[15] In
4
2g subdomain.[31] The C1 conformer fitted very well in both
1
cases (see Figure 9), much better than the alternative C4 or
skew boat conformers. In all cases, no contacts between the
pseudoanomeric center and the lectin were predicted. Thus,
docking analysis was in accordance with the NMR-derived
observations.
Compound 2: The analysis of the vicinal proton–proton cou-
pling constants (Table 4 and Table 5) for 2 indicates that no
single conformation can fit the experimental data. The best
Chem. Eur. J. 2009, 15, 2861 – 2873
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2865

P. Vogel, J. Jimꢁnez-Barbero et al.
15.0 kcalmolÀ1 in CD3OD and
16.7 kcalmolÀ1 in CD3CN, prob-
ably related to changes around
the glycosidic torsions.
Table 5. Comparison between the experimental vicinal coupling constant values of 2 in CD3OD (Jexp) and
those expected for the most stable conformations of the six-membered ring (4C1, 1C4, 1S3) according to
MM3*calculations (Jtheor).
Coupling
constant
1C4
H-C-C-H
torsion[b]
4C1
H-C-C-H
torsion[b]
1S3
H-C-C-H
torsion[b]
Furthermore, 5 ns MM3* mo-
DE[a]
3.8
4.8
0.0
ACHTUNGTRENNUNG ]
[kJmolÀ1
lecular-dynamics
simulations
Jexp[c] [Hz]
Jtheor
Jtheor
Jtheor
[b]
[b]
[b]
were performed by using the
geometry of the three minima
as input values (Figure 5 and
the Supporting Information).
The results indicated that the
three conformers were fairly
stable, with fluctuations along
CD3OD
[Hz]
[Hz]
[Hz]
JH1,H2
JH2,H3
JH3,H4
JH4,H5
JH5,H6b
JH5,H6a
JH1,F1
JH1,F2
3.2
5.6
3.2
4.8
7.2
4.0
14.4
16.8
2.4
3.0
3.4
5.5
9.5
1.9
–
À55.2
À67.1
À54.0
49.1
4.7
8.7
5.0
0.1
10.0
2.5
–
43.7
À172.5
54.8
À57.2
176.6
65.1
3.2
7.4
3.9
2.4
9.5
2.0
-
À33.8
À152.5
62.9
À33.0
À172.0
69.6
À171.6
70.3
À177
63
174
54
À175
the well-defined
F
torsion
–
–
–
64
value. Two sets of values
(around +608 and À1208) were
found for the Y angle (defined
as C1-CF2-Cipso-Cortho). These
values corresponded to the
symmetry inherent to a mono-
substituted phenyl ring for
which the two ortho positions
are chemically equivalent. The
[a] The relative steric energy difference calculated by MM3* between the global minimum (1S3) and the other
two basic conformers is also given. [b] The theoretical values (Jtheor), as well as the torsion angles, were ob-
tained by applying the generalized Karplus equation proposed by Altona to the geometries calculated by
MM3* calculations. The gt conformer around the C5–C6 torsion was employed. [c] The calculated couplings
for the 1C4 and the 1S3 forms (italicized) are much closer to those observed experimentally than those expected
4
4
for the C1 geometry. The comparison between the couplings indicates that the contribution of the C1 confor-
1
mation is basically negligible. The best fit is given by an approximate 1:1 distribution between the C4 and the
1S3 conformers. Moreover, there is a predominant conformation around the pseudoglycosidic linkage and
around the C5–C6 torsion. In the latter case, a non-negligible contribution of the alternative tg rotamer is tan-
gible.
contrast, a 50:50 exo/non-exo equilibrium for an analogous
molecule with a GalbCF2R linkage produced two identical
JH1,F coupling values of 14.3 Hz.[15] Although no Karplus-like
relationship exists for this coupling pathway, the relatively
similar JH1,F1 and JH1,F2 coupling constants observed for 2 in-
dicate, beside the major exo-anomeric orientation, the pres-
ence of rotamers with the alternative F +60 non-exo-
anomeric geometry. This molecule therefore displays a wide
range of conformational flexibility, not only for the pyranose
ring but also around the glycosidic F torsion.
Figure 5. The major conformers of 2 (left, 1C4; center, 1S3) according to
the experimental NMR data and MM3* molecular mechanics and dy-
namics calculations. The non-exo-anomeric form around F is depicted
for the 1C4 chair and the exo-anomeric geometry for the 1S3 skew boat,
respectively. The minor 4C1 chair conformer (non-exo-anomeric orienta-
tion around F) is presented at the right side. This orientation of the aro-
matic ring (defined by Y angle) defines both orientations deduced by the
MD simulations (+608 and À1208) owing to the intrinsic symmetry of
the aromatic ring. NOE signals and J couplings are quantitatively ac-
Next, we probed into the issue on the origin and nature of
the conformational flexibility. As the conformational entro-
1
1
1
py of six-membered S3 skew-boat form is higher than those
counted for by a 1:1 mixture of the C4 and S3 conformers.
of the corresponding chair geometries, we addressed the
issue as to whether there is an entropic origin for the detect-
1
ed flexibility. A series of H NMR spectra was recorded at
calculation of predicted NOE values for a mixture of geo-
metries computed from MD simulations starting from a mix-
ture of the C4 and S3 conformers enabled us to quantita-
tively explain the observed NOE peaks (Table 6).
high field (800 MHz, to avoid interference from strong cou-
pling effects) and at different temperatures and solvents to
monitor the possible changes of coupling constants. In prin-
ciple, the decreasing temperature should lead to a higher
population of the enthalpically favored form, with a con-
comitant change in coupling-constant values. However, no
change was observed for any of the intra-ring H/H couplings
nor for the F-related JH1,F values (all differences were below
0.5 Hz). This indicates that the observed conformational
flexibility does not necessarily correspond to intrinsic effects
related to the conformational entropy of the molecule. Nev-
ertheless, the change in temperature leads to a clear coales-
cence in the 19F NMR spectra (see the Supporting Informa-
tion), both in CD3OD (at 273 K) and CD3CN (at 243 K).
The associated activation energy was approximately
1
1
According to the combined NMR spectroscopy and simu-
lation data, 2 does not appear to behave as a true mimetic
4
of the natural compound. The natural major C1 chair con-
formation is not present in solution and the exo-anomeric
orientation around F coexists with the unusual non-exo-
anomeric conformation.
Compound 2 was obtained as single crystals and its con-
formation determined in the solid state by X-ray crystallog-
raphy (Figure 6). Strikingly, its conformation was that which
4
is not present in methanol solution, that of the “natural” C1
chair conformation. This is not a completely novel case. In
several instances,[33] different conformations have been ob-
2866
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2861 – 2873

Analysis of (a-d-Galactosyl)phenylmethane and a-,b-Difluoromethane Analogues
FULL PAPER
Table 6. Experimental NOE values for compound 2 in CD3OD in com-
parison with those estimated by applying a full relaxation matrix ap-
1
1
proach to a 1:1 distribution of the C4 and S3 conformers.
Proton
pair
NOEs%[a]
experimental (700 ms)
calculated 1:1
1C4:1S3 distribution
50 ps correlation time)
H1/Harom
H1/H2
H1/H6a
H5/H3
H3/H4
H4/H5
H5/Harom
H6a/H5
H6b/H5
1.3
4.4
4.2
0.3
4.9
4.2
0.6
1.1
2.5
3.5
4.4
6.1
1.0
3.9
5.1
0.4
0.7
3.0
[a] The experiments (298 K) and the calculations (tc 50 ps, matched for
the intraresidual H1/H2) were performed at 500 MHz. The agreement
between calculation and experimental result is satisfactory.
Figure 7. Compound
2
interacts with VAA lectin. A) This spectrum
shows a section of the 500 MHz 1H NMR spectrum of 2 in D2O. B) STD
spectrum of 2 (2 s saturation time) in the presence of VAA (approxi-
mately 25 mm). The 2:VAA molar ratio is 50:1. The general STD spec-
trum also shows transfer to the aromatic protons. The off-resonance fre-
quency was set at 50 ppm and the on-resonance frequency was estab-
lished at À0.5 ppm. C) STD effects on the different proton atoms of 2, as
deduced from the STD experiments. 100% is given to the highest intensi-
ty (H4+H5), and all the other values are referred to this one. Analogous
numbers are observed for all the compounds.
4
Figure 6. The C1 chair geometry is obtained in the solid state after X-ray
analysis of single crystals of 2. The orientation of the aromatic ring corre-
sponds to a non-exo-anomeric conformation.[34]
tional perspective, it has been previously shown that lactose
is selected by VAA in the syn-conformation.[32,33]
tained for sugar derivatives in solution than in the solid
state. The key point is that this finding reveals intrinsic flexi-
bility of glycomimetics of the C-glycosyl family, not only for
the pseudoglycosidic linkages, but also at the level of the
six-membered ring.
To define which conformation of compound 2 VAA lectin
binds preferentially, a trNOE experiment (data not shown)
was performed. As for the CORCEMA analysis[35] of the
STD effects, the trNOE did not lead to conclusions on this
issue. Owing to the signal overlapping of the key protons of
2 in D2O, along with the small size of the ligand and the
similarity of the observed STDs for the ring protons, only
ambiguous conclusions could be drawn. Nevertheless, dock-
ing calculations (as for compound 1) with the three standard
Having described structural properties, we next used
STD NMR spectroscopic experiments to test the possible
binding of 2 to VAA lectin. The lectin binds 2, as observed
in the STD experiment (Figure 7). Major saturation transfer
is detected to the H2–H5 region, followed by the H1 and
H6 areas. This indicates a clear interaction of this molecule
with the lectin. Glycomimetic 2 also competed with lactose
for the lectin site. The STD effects of 2 were clearly dimin-
ished when lactose was added to the VAA/2 solution. In this
case, the STD peaks of lactose were also affected upon addi-
tion of 2 to the corresponding VAA/lactose mixture, al-
though to a lesser extent than that for the alternative experi-
ment. Thus, it seems feasible to assume that the binding af-
finity of 2 is smaller (but still in the millimolar range) than
that of the natural analogue, lactose. From the conforma-
4
1
1
conformations C1, C4, and S3 of 2 suggested that the chair
4
conformation C1 fits VAA better than others, in which case
compound 2 behaves like compound 1 when binding to
VAA occurs.
Compound 3: The chemical shifts and J coupling values of
the b-linked CF2 analogue 3 are compiled in Table 7. In this
4
case, all the data are in close agreement with a major C1
chair conformation in CD3OD and D2O solution. The two
coupling values to the fluorine atom are in the medium
Chem. Eur. J. 2009, 15, 2861 – 2873
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2867

P. Vogel, J. Jimꢁnez-Barbero et al.
Table 7. Comparison between the experimentally observed coupling con-
stants in CD3OD for 3 (Jexp) and those expected for a C1 chair conforma-
tion according to the MM3* force field (Jtheor). Chemical shifts (d, ppm)
of 3 in CD3OD (500 MHz, 298 K) are also given.
flexibility of the six-membered ring in solution. The confor-
mation of the ring observed in the solid state of this mole-
cule had an almost negligible population in solution. The in-
herent flexibility of ligands can have implications from the
molecular recognition viewpoint and, when interacting with
protein receptors, conformational selection processes usually
take place, given the relatively low-energy barriers required
for conformer interconversion.[33] However, despite the dif-
ferent conformational behavior, certain C-galactosyl ana-
logues might still be recognized by galactose-binding pro-
teins, as exemplified here by using VAA. The detection of
4
Coupling
constant
Jexp CD3OD
[Hz]
4C1 Jtheor
[Hz]
H-C-C-H
torsion
H
d
ACHTUNGTNER[NUNG ppm]
JH1,H2
JH2,H3
JH3,H4
JH4,H5
JH1,F1
JH1,F2
10.1
8.4
3.6
<1
9.4
10.5
13.2
8.7
2.5
174.9
À171.1
54.5
À56.4
H1
H2
H3
H4
H5
H6ab
3.70
3.73
3.48
3.87
3.46
3.62, 3.59
0.6
range, 9.4 and 10.5 Hz, indicating the possible presence of a
conformational equilibrium around F torsion between the
exo and non-exo orientations. Indeed, the steric energy dif-
ference (MM3* with GB/SA) between both forms amounts
to only 1 kcalmolÀ1. Strong NOE values are observed from
the ortho aromatic protons to both H1 and H2, with a small-
er NOE to H5. The simultaneous presence of these NOE
values indicates the presence of an equilibrium between the
exo and non-exo orientations, and the stronger NOE to H2
suggests that the second is the major conformer in solution.
(see Figure 8 and the Supporting Information).
STD experiments performed under the same experimen-
tal conditions as for 1 and 2 in-
Figure 8. Compound 3 shows a major 4C1 conformation in solution, in
contrast with the behavior of its a-linked analogue, compound 2. The ob-
served NOE peaks suggest that the orientation of the aromatic ring is
that shown at the lefthand side of the Figure (non-exo-anomeric), al-
though the alternative one (exo-anomeric, right) seems to be also present
in solution.
dicated ligand binding to VAA
lectin (data not shown), where-
as the AUTODOCK calcula-
tions predicted the preferential
4
binding of the major C1 confor-
mer (Figure 9). No preference
for any of the three possible
orientations around F, exo, non
exo, and anti, was deduced.
Conclusion
The conformations of two a-
linked and one b-linked C-gly-
cosyl compounds were studied
by using
NMR spectroscopic and molec-
ular-modeling procedures.
a combination of
These compounds, especially
the a-linked analogues, harbor
conformational flexibility. This
appears around the pseudo-gly-
cosidic linkages, but also in dif-
ferent geometries that are
adopted by the six-membered
ring, depending on the substitu-
tion at the pseudo-glycosidic
carbon. It was found that CF2
substitution at the a-anomeric
Figure 9. Different views of the putative binding mode of compounds 1–3 to VAA, according to AUTODOCK
calculations. A) Expansion of the Trp site of VAA, showing the interaction with 1 in the preferred 4C1 chair
conformation. B) Expansion of the Tyr-site of VAA, showing the interaction with 2 in the C1 chair conforma-
4
tion. C) Compound 3 bound to the complete VAA lectin, showing both sites. D) Expansion of the Trp site of
4
position resulted in significant VAA, showing the interaction with 3 in the preferred C1 chair conformation.
2868
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2861 – 2873

Analysis of (a-d-Galactosyl)phenylmethane and a-,b-Difluoromethane Analogues
FULL PAPER
The binding of the galactose analogues was evaluated by STD experi-
ments. STD experiments were performed without saturation of the resid-
ual HDO signal for molar ratios 50:1 of compound:VAA. The concentra-
tion of the protein was approximately 25 mm. A series of Gaussian-
shaped pulses of 50 ms each was employed with a total saturation time
for the protein envelope of 2 s and a maximum B1 field strength of
60 Hz. An off-resonance frequency of d=40 ppm and on-resonance fre-
quency of d=À1.0 ppm (protein aliphatic signals region) were applied.
In the case of the CH2-analogue 1, competition experiments with lactose
were also performed. In this case, a threefold excess of lactose was added
to the NMR tube of the sample containing VAA/1, and the STD experi-
ment was performed by using the same experimental conditions de-
scribed above. For compounds 2 and 3, analogous experiments were per-
formed.
For the VAA/1 sample, exchange trNOE experiments[41] were performed
by using selective 1D experiments with the double-pulse field-gradient
spin-echo (DPFGSE) module.[42] Measurements were done with a freshly
prepared ligand/lectin mixture, with mixing times of 100 and 200 ms, at
an approximately 20:1 molar ratio of ligand/protein. A concentration of
2 mm of the ligand was employed in all cases. No purging spin-lock
period was employed to remove the NMR signals of the macromolecule
background. Strong negative NOE cross-peaks were observed, which is
in contrast with the free state and indicates binding of the sugars to the
lectin preparation. trNOE experiments were also attempted for the
VAA/2 and VAA/3 samples, but they did not afford additional informa-
tion owing to the extensive overlapping of the key signals.
binding will give further research a clear direction in the
design of potent non-hydrolyzable mimics for medically rel-
evant lectins.
Experimental Section
Modeling: The low energy conformers were calculated by using the
MM3* force field[25] in the program MAESTRO.[36]
The torsion angle F is defined as H1Gal-C1Gal-C-Cipso and Y as C1Gal-C-
Cipso-Cortho. The three possible staggered rotamers around F combined
with those for Y (every 1208) were built (nine conformers in total) with
MAESTRO and submitted to exhaustive energy minimization. The prob-
ability distribution was calculated from the obtained energy values ac-
cording to a Boltzmann function at 300 K. In all the molecular mechanics
and dynamics calculations, the GB/SA solvation model for water was
used.[37]
3
For the a-CF2 analogue 2, for which the JH,H coupling analysis showed
the presence of distinct six-membered ring geometries present in solu-
tion, three different starting geometries were considered, corresponding
to the 4C1 chair, 1C4 chair, and 1S3 skew boat. Thus, in total, 27 conform-
ers were calculated. For the a-CH2 analogue 1 and the b-CF2 analogue 3,
3
only the 4C1 chair conformation was considered as the JH,H coupling
values were in agreement with the exclusive presence of this conformer.
Docking calculations: The 4C1 and 1C4 chair conformers of 1–3 (as ob-
served by NMR spectroscopy) were manually docked into the two carbo-
hydrate-binding sites of VAA in the 1a and 2g subdomains of the B sub-
The molecular dynamics simulations were also performed by using the
MM3* force field. For molecular dynamics simulations, the global-mini-
mum geometry was used as input. A temperature of simulation of 300 K
was employed with a time step of 1.5 fs and an equilibration time of
100 ps. The total simulation times for each compound were 5 ns.
AHCTUNGTRENNUNG
unit of VAA[43] by superimposing the terminal Gal residue (protein data-
base code 1PUM). Then, different possibilities of arranging the side
chain of the glycan were used as input geometries for AutoDock 3.0 sim-
ulations[44] with the multiple Lamarckian Genetic Algorithm. Only local
searches were performed centered in the two experimental galactoside-
specific VAA X-ray sites. Grids of probe atom interaction energies and
electrostatic potential were generated by the AutoGrid program present
in AutoDock 3.0. Grid spacings of 0.6 and 0.375 ꢆ were used for the
global and local searches, respectively. For each calculation, 100 docking
runs were performed by using a population of 200 individuals and an
energy evaluation number of 3x106. The best scoring was always obtained
with the 4C1 chair conformer, independently of the anomeric configura-
tion of 1–3.
For the C5-C6 torsion, only the gt geometry (w, defined as C4-C5-C6-O4
ca. 1808) was considered as it has been demonstrated to be the major
conformation for galactose derivatives (O4 in axial orientation).[26]
1
NMR experiments: H NMR (800 MHz) spectra of compound 2 were re-
corded on a Bruker AVANCE 800 spectrometer (CryoProbe). Three dif-
ferent solvents were used: CD3OD, CD3CN, and D2O. For every solvent,
a series of spectra at variable temperatures ranging from 08C to +508C
(for D2O: +108C to +508C) were measured at concentrations of 30 mm.
Chemical shifts are reported in ppm, after calibration of the residue peak
of each solvent: d=3.31 ppm for CD3OD, d=1.94 ppm for CD3CN, and
d=4.79 ppm for D2O. Vicinal proton–proton coupling constants were es-
timated from first-order analysis of the spectra recorded at 800 MHz to
minimize strong-coupling effects.
Synthesis
General: Commercial reagents (Fluka, Aldrich, Acros) were used with-
out purification. CH2Cl2 and THF were dried from activated alumina col-
umns (IT technology). The other anhydrous solvents were purchased and
stored over molecular sieves. CAUTION! Neat Deoxo-Fluor reagent is
thermally unstable[23a] over 1508C and it reacts violently with water,
therefore it should be handled carefully, under a well ventilated hood,
and behind a safety shield. Flash column chromatography (FC): Fluka
silica gel, No.60752, 230–400 mesh. TLC for reaction monitoring: Merck
silica gel 60 F254 plates; detection by UV light; Pancaldi reagent or
KMnO4 solution. Melting Points (m.p.): Bꢇchi SMP-20; uncorrected. Op-
tical rotations ([a]D25): Jasco P-1020; lamp Na, 589 nm; 258C. IR spectra:
Perkin Elmer Paragon 1000 or Perkin Elmer Spectrum One FT-IR Spec-
trometer. NMR spectra: Bruker ARX-400 or DPX-400 Spectrometer,
1H NMR (500 MHz) spectra were recorded at 308C in D2O and in
CD3OD on a Bruker AVANCE 500 spectrometer. Concentrations of ap-
proximately 2 mm of 1, 2, and 3 were used. Chemical shifts are reported
in ppm by using external TSP (2,2,3,3-tetradeutero-3-trimethylsilylpro-
pionic acid, 0 ppm) as the reference. The 2D-TOCSY experiment (30 ms
mixing time) was performed by using a data matrix of 256ꢅ2 K to digi-
tize a spectral width of 4000 Hz. Four scans were used per increment
with a relaxation delay of 2 s. 2D-NOESY (600 and 1000 ms) and 2D-T-
ROESY experiments (400 and 500 ms) used the standard sequences. Dis-
tances were estimated from NOESY/ROESY experimental data by using
the average values of the isolated spin-pair approximation[38] for the data
with the shorter mixing time. Estimated errors are below 10%.
400 MHz for 1H, 100.6 MHz for 13C and 376.7 MHz for 19F;
d for
A comparison between the observed NOE values and those estimated
for the different conformations was performed by using a full relaxation
matrix approach and a home-made program, which is available from the
authors upon request.[39] For all molecules, an average effective correla-
tion time of 50 ps was estimated for the best adjustment of the observed
and calculated cross-peaks.
1H NMR and 13C NMR in ppm relative to the solventꢈs residual signal as
the internal reference [CDCl3, d(H) 7.26 and d(C) 77.0, CD3OD, d(H)
3.31 and d(C) 49.0, D2O d(H) 4.79], d for 19F NMR in ppm relative to
signal of external reference [CFCl3, d(F) 0]; all 1H and 13C assignments
were confirmed by 2D-COSY and 2D-HSQC spectra. Mass spectra (MS):
Shimadzu Axima CFRplus for MALDI-TOF or Waters CapLC-coupled
Micromass Ultima for ESI QqTOF. Elemental analyses: Ilse Beetz,
D-96301 Kronach, Germany.
Interaction studies with VAA lectin: The lectin was isolated from the ex-
tracts of dried mistletoe leaves by using affinity chromatography on lac-
tosylated Sepharose 4B as crucial step. Purity and quaternary structure
were ascertained by one- and two-dimensional gel electrophoresis, gel fil-
tration and ultracentrifugation and carbohydrate-dependent activity was
tested by haemagglutination as well as solid-phase/cell assays.[3c,39,40]
The atom numbering in NMR spectra follows the standard numbering
for sugars, whereas the numbers in the compound name are generated
according to IUPAC rules.
Chem. Eur. J. 2009, 15, 2861 – 2873
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2869

P. Vogel, J. Jimꢁnez-Barbero et al.
CCDC-683094 (2) contains the supplementary crystallographic data for
this paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_re-
quest/cif.
cal spectroscopic data to those previously reported.[19a] [a]D25 =+95 (c=
0.18, CHCl3).
2,6-Anhydro-3,4,5,7-tetra-O-benzyl-d-glycero-l-gluco-heptose (7) and
2,6-anhydro-3,4,5,7-tetra-O-benzyl-d-glycero-l-manno-heptose (8): Com-
pounds 7 and 8 were prepared as previously reported[19a] and used in the
next steps as crude mixtures.
2,6-Anhydro-1,3,4,5-tetra-O-acetyl-7,8,9-trideoxy-d-glycero-l-galacto-
non-7,8-dienitol (4): BF3·Et2O (35.0 mL, 39.6 g, 0.279 mol) was added
dropwise to a stirred solution of b-d-galactose pentaacetate (36.2 g, 0.093
mol) and propargyltrimethylsilane (28.0 mL, 21.0 g, 0.187 mol) in anhy-
drous CH3CN (200 mL) at 08C under argon. This was then followed by
the dropwise addition of TMSOTf (33.6 mL, 41.3 g, 0.186 mol). The reac-
tion mixture was stirred at the same temperature for 2.5 h. Then, it was
diluted with EtOAc (200 mL) and aqueous HCl 1m (200 mL) was added.
The two phases were separated and the aqueous phase was extracted
with EtOAc (150 mLꢅ2). The combined organic phases were washed
with saturated aqueous NaHCO3 and brine until a neutral pH value was
obtained, dried over Mg2SO4, and concentrated in vacuo. FC (25%
EtOAc in petroleum ether) afforded 4 as a yellowish solid (27.5 g, 80%).
Part of this solid was recrystallized from Et2O and n-hexane to afford
colorless crystals of 4. m.p. 77–798C; [a]D25 =+186 (c=0.185, MeOH);
IR (neat): n˜ =1956 (allene), 1736, 1724 (C=O of AcO), 1227, 1210, 1091,
2,6-Anhydro-3,4,5,7-tetra-O-benzyl-1-C-phenyl-aldehydo-d-glycero-l-
gluco-heptose (9): PhMgBr solution in Et2O (ca. 3m, 1.5 mL, 4.5 mmol)
at À788C under argon was added dropwise to a stirred solution of the
crude a-aldehyde 7 (1.73 mmol) in anhydrous THF (22 mL). The reaction
mixture was kept at the same temperature for 1.5 h. It was warmed up to
À208C over an additional 2 h. Then, the reaction was quenched by pour-
ing the mixture into a pH 7 phosphate buffer solution 0.05m (30 mL) and
the mixture was filtered through Celite. The filtrate was extracted with
Et2O (4ꢅ25 mL) and the combined organic phases were dried over
MgSO4 and concentrated in vacuo to afford a mixture of alcohols that
were then used without purification. Activated 4 ꢆ molecular sieves
(ꢀ1.6 g) and pyridinium chlorochromate (PCC; 1.87 g, 8.67 mmol) were
added in one portion to a vigorously stirred solution of the previous
crude mixture in anhydrous CH2Cl2 (24 mL) at 258C and under argon.
The reaction mixture was stirred at the same temperature for 2 h. Then
Et2O (100 mL) was added and the mixture was left without stirring.
After 30 min, the upper solution was filtered through a pad of florisil.
The black tar that precipitated was washed several times with Et2O and
filtered as well. The filtrate was concentrated in vacuo and purified by
FC (15% Et2O in petroleum ether) to afford ketone 9 as a yellowish oil
(0.54 g, 50% for three steps starting from the a-allene 6). [a]D25 =+35
(c=0.184, CHCl3); IR (neat): n˜ =1694, 1682 (C=O), 1597, 1496, 1455,
1049, 1018, 910, 878 cmÀ1
;
1H NMR (400 MHz, CDCl3, 298 K): d=5.43
3
(dd, 1H, 3JH4,H3 =3.3 Hz, 3JH4,H5 =1.5 Hz, H4), 5.35 (dd, 1H, JH2,H3
=
10.5 Hz, 3JH2,H1 =5.5 Hz, H2), 5.23–5.28 (m, 1H, HC=), 5.25 (dd, 1H,
3JH3,H2 =10.7 Hz, 3JH3,H4 =3.3 Hz, H3), 4.91–4.95 (m, 3H,=CH2, H1), 4.24
3
3
3
(ddd, 1H, JH5,H6a =6.8 Hz, JH5,H6b =6.2 Hz, JH5,H4 =1.2 Hz, H5), 4.12 (dd,
1H, 2JH6b,H6a =11.1 Hz, 3JH6b,H5 =6.2 Hz, H6b), 4.07 (dd, 1H, JH6a,H6b
=
2
11.1 Hz, 3JH6a,H5 =7.4 Hz, H6a), 2.15, 2.06, 2.04, 2.00 ppm (4 s, 12H, 4ꢅ
CH3 of AcO); 13C NMR (100 MHz, CDCl3, 298 K): d=209.2 (=C=),
170.3, 170.1, 170.0, 169.7 (4ꢅC=O of AcO), 84.4 (HC=), 77.5 (=CH2),
70.7 (C1), 68.2 (C5), 68.0 (C3, C4), 67,6 (C2), 61.8 (C6), 20.6 ppm (CH3
of AcO); ESI QqTOF-MS: calcd for ([M+Na]) 393.1162, found 393.1167.
1372, 1217, 1094, 732 cmÀ1 1H NMR (400 MHz, CDCl3, 298 K): d=7.90–
;
7.93 (m, 2H), 7.50–7.55 (m, 1H), 7.24–7.40 (m, 17H), 7.16–7.22 (m, 3H),
7.03–7.07 (m, 2H) Harom, 5.20 (d, 1H, 3JH1,H2 =4.3 Hz, H1), 4.74 (d, 1H,
2J=11.7 Hz), 4.73 (d, 1H, 2J=12.0 Hz), 4.66 (d, 1H, 2J=12.0 Hz), 4.57
(d, 1H, 2J=11.7 Hz), 4.55 (d, 1H, 2J=12.0 Hz), 4.47 (d, 1H, 2J=
12.0 Hz), 4.47 (brs, 2H) 4ꢅCH2 of BnO, 4.37 (ddd, 1H, 3JH5,H6a =8.0 Hz,
3JH5,H6b =4.5 Hz, 3JH5,H4 =3.5 Hz, H5), 4.20 (dd, 1H, 3JH2,H3 =6.8 Hz,
3JH2,H1 =4.1 Hz, H2), 4.14 (dd, 1H, 3JH3,H2 =6.8 Hz, 3JH3,H4 =2.5 Hz, H3),
2,6-Anhydro-7,8,9-trideoxy-d-glycero-l-galacto-non-7,8-dienitol (5):
A
solution of 30% MeONa in MeOH (1.80 mL, 0.52 g, 0.0097 mol) was
added dropwise at 258C to a stirred solution of allene 4 (30.0 g, 0.081
mol) in MeOH (450 mL). After the mixture had been stirred at the same
temperature for 2 h, Dowex 50WX8 (50–100 mesh) was added in small
portions until the pH value of the reaction mixture became neutral.
After 15 min, it was filtered and concentrated in vacuo. The residue was
recrystallized from MeOH to afford 5 as white crystals (14.7 g after three
recrystallizations, 90%). M.p. 153–1578C; [a]D25 =+243 (c=0.175,
MeOH); IR (neat): n˜ =3675, 3429, 3227 (br, HO), 1948 (allene), 1077,
4.10 (dd, 1H, 3JH4,H5 =3.6 Hz, 3JH4,H3 =2.8 Hz, H4), 3.88 (dd, 1H,
2
2JH6a,H6b =10.6 Hz, 3JH6a,H-5 =7.6 Hz, H6a), 3.66 ppm (dd, 1H, JH6b,H6a
=
10.6 Hz, 3JH6b,H5 =4.8 Hz, H6b); 13C NMR (100 MHz, CDCl3, 298 K): d=
197.6 (C=O), 138.5, 138.4, 138.2, 137.9, 136.5 (Cipso of Ph and BnO),
133.0, 129.0, 128.4, 128.3, 128.2, 127.95, 127.9, 127.7, 127.67, 127.6, 127.54,
127.5 (Carom), 77.0 (C3), 76.2 (C2), 74.2 (C5), 73.9 (C4, CH2 of BnO),
73.2, 73.1, 73.0 (3ꢅCH2 of BnO), 72.3 (C1), 66.9 ppm (C6); MALDI-
TOF MS: calcd for ([M+K]) 667.2462, found 667.2457; elemental analysis
calcd (%) for C41H40O6 C 78.32, H 6.41; found: C 78.36, H 6.47.
1060, 1026, 999, 786 cmÀ1
;
1H NMR (400 MHz, D2O, 298 K): d=5.47
4
(ddd, 1H, JHC CH2a =4JHC CH2b =3JHC ,H1 =6.7 Hz, HC=), 4.96 (ddd, 1H,
= =
,
= =
,
=
2J=11.8 Hz, 4J=
==6.8 Hz, 5J=CH2a,H1 =2.7 Hz, =CH2a), 4.92 (ddd,
CH2a,HC
2
4
5
1H, J=11.7 Hz, J=
==6.6 Hz, J=CH2b,H1 =2.7 Hz, =CH2b), 4.65–4.69
CH2b,HC
(m, 1H, H1), 4.04 (dd, 1H, 3JH2,H3 =10.3 Hz, 3JH2,H1 =6.2 Hz, H2), 3.96–
4.01 (m, 2H, H5, H4), 3.81 (dd, 1H, 3JH3,H2 =10.3 Hz, 3JH3,H4 =3.2 Hz,
H3), 3.70 ppm (d, 2H, 3JH6,H5 =6.0 Hz, H6a,b); 13C NMR (100 MHz,
CD3OD in D2O, 298 K): d=210.8 (=C=), 85.1 (HC=), 77.5 (=CH2), 75.3
(C1), 73.7 (C5 or C4), 71.0 (C3), 70.4 (C5 or C4), 69.0 (C2), 62.2 ppm
(C6); ESI QqTOF-MS: calcd for ([M+Na]) 225.0739, found 225.0741.
2,6-Anhydro-3,4,5,7-tetra-O-benzyl-1-C-phenyl-aldehydo-d-glycero-l-
manno-heptose (10): Compound 10 was prepared in the same way as de-
scribed for its a-epimer 9, starting from the crude b-galactosyl carbalde-
hyde 8 (0.94 mmol). FC (7.5% EtOAc in petroleum ether) gave 10 as a
white solid (0.15 g, 25% for four steps starting from the a-allene 6),
which was recrystallized from cyclohexane. M.p. 104–1058C [ref. [21]
103–1058C]; [a]D25 =+6.2 (c=0.13, CHCl3) [ref. [21] +7.5 (c=0.8,
CHCl3)]. The spectroscopic data are identical to those previously report-
ed.[21]
2,6-Anhydro-1,3,4,5-tetra-O-benzyl-7,8,9-trideoxy-d-glycero-l-galacto-
non-7,8-dienitol (6): The solid allene 5 (1.00 g, 4.94 mmol) was added in
small portions to a vigorously stirred suspension of NaH (60% of purity,
0.90 g, 22.5 mmol, previously washed with anhydrous pentane under
argon) in anhydrous N,N-dimethylformamide (DMF; 20 mL) at 08C
under a flow of argon. After completion of the addition, the thick sus-
pension was stirred for 45 min at 258C. Catalytic amount of nBu4NI
(0.20 g, 0.54 mmol) was added, followed by the dropwise addition of
BnBr (3.2 mL, 4.6 g, 27 mmol) at 08C. After the completion of the addi-
tion, the reaction mixture was stirred at 258C for an additional 16 h.
Then, the reaction was quenched by careful addition of MeOH (3 mL) at
08C. The mixture was partitioned between Et2O (50 mL) and water
(50 mL), the two phases were separated and the aqueous phase was ex-
tracted with Et2O (3ꢅ30 mL). The combined organic phases were dried
over MgSO4 and concentrated in vacuo. Most of the DMF was removed
under high vacuum and the residue was purified by FC (10% Et2O in pe-
troleum ether) to afford 6 (2.17 g, 78%) as a colorless oil that has identi-
2,6-Anhydro-7-deoxy-7-C-phenyl-d-glycero-l-galacto-heptitol (1): To
a
solution of ketone 9 (0.141 g, 0.22 mmol) in a mixture of MeOH (4 mL)
and EtOAc (4 mL) was added Pd/C 10% (0.060 g) under argon, along
with three drops of AcOH. Argon was exchanged for H2 gas and the re-
action mixture was stirred under an H2 atmosphere (1 atm) at 258C for
two days. Then, the mixture was filtered carefully through Celite and the
filtrate was concentrated in vacuo. The residue was purified by FC (10%
MeOH in CH2Cl2) to give 1 (0.044 g, 77%) as a colorless oil. [a]D =+78
(c=0.175, MeOH); IR (neat): n˜ =3306, 2922 (br, OH), 1452, 1361, 1071,
1031, 978, 865, 778, 740, 697 cmÀ1 1H NMR (400 MHz, CD3OD, 298 K):
;
d=7.23–7.28 (m, 4H, Hortho, Hmeta), 7.14–7.18 (m, 1H, Hpara), 4.14 (ddd,
3
1H, 3JH1,CH2a =8.6 Hz,3JH1,H2 ꢀ JH1,CH2b ꢀ5.4 Hz, H1), 4.02 (brdd, 1H,
3
3JH4,H3 ꢀ3.2 Hz, JH4,H5 ꢀ2.7 Hz, H4), 3.94 (ddd, 1H, 3JH5,H6a =6.5 Hz,
3JH5,H6b =5.4 Hz, 3JH5,H4 =2.7 Hz, H5), 3.90 (dd, 1H, 3JH2,H3 =8.6 Hz,
2870
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2861 – 2873

Analysis of (a-d-Galactosyl)phenylmethane and a-,b-Difluoromethane Analogues
FULL PAPER
3JH2,H1 =4.9 Hz, H2), 3.74 (dd, 1H, 3JH3,H2 =8.6 Hz, 3JH3,H4 =3.2 Hz, H3),
248 Hz, CF2), 84.5 (C3), 79.6 (t, 2JC1,Fa =2JC1,Fb =29.7 Hz, C1), 77.4 (C5),
75.0 (CH2 of BnO), 74.8 (C2), 74.1, 73.4 (2ꢅCH2 of BnO), 73.3 (C4),
72.6 (CH2 of BnO), 68.9 (C6); 19F NMR (376 MHz, CDCl3 CFCl3, 298 K):
d ppm À97.6 (dd, 1F, 2JFa,Fb =257 Hz, 3JFa,H1 =4.4 Hz, Fa), À108.6 ppm
(dd, 1F, 2JFb,Fa =258 Hz, 3JFb,H-1 =14.4 Hz, Fb); ESI QqTOF MS: calcd for
([M+Na]) 673.2742, found 673.2739; elemental analysis calcd (%) for
C41H40F2O5: C 75.67, H 6.20; found: C 75.60, H 6.34.
3
3.67–3.74 (m, 1H, H6a), 3.67 (dd, 1H, 2JH6b,H6a =11.3 Hz, JH6b,H5 =5.4 Hz,
H6b), 2.91–2.97 ppm (m, 2H, CH2); 13C NMR (100 MHz, CD3OD,
298 K): d=140.9 (Cipso), 130.4, 129.2 (Cortho, Cmeta), 127.0 (Cpara), 77.5 (C1),
74.4 (C5), 72.0 (C3), 70.1 (C2), 69.9 (C4), 61.7 (C6), 32.2 ppm (CH2); ESI
QqTOF MS: calcd for ([M+H]) 255.1232, found 255.1231; elemental
analysis calcd (%) for C13H18O5: C 61.40, H 7.14; found: C 61.30, H 7.14.
2,6-Anhydro-1-deoxy-1,1-difluoro-3,4,5,7-tetra-O-benzyl-1-C-phenyl-d-
glycero-l-gluco-heptitol (11): In a PET vial containing the a-ketone 9
(0.50 g, 0.80 mmol), was added neat Deoxo-Fluor reagent (1.50 mL, 1.8 g,
8.1 mmol), followed by two drops of HF-pyridine as catalyst at 258C
under argon. The reaction was heated to 808C for 10 h. More reagent
was then added (0.30 mL, 0.36 g, 1.6 mmol) and the mixture was heated
to 808C for an additional 10 h. Further reagent was then added (0.15 mL,
0.18 g, 0.8 mmol) and heated to the same temperature for an additional
10 h. Then, the reaction mixture was added dropwise to a mixture of sa-
turated aqueous NaHCO3 and ice. It was allowed to reach 258C and was
extracted with Et2O (3ꢅ30 mL). The combined organic phases were
washed with brine, dried over MgSO4, and concentrated in vacuo. The
residue was purified by FC (15% Et2O in petroleum ether) to afford 11
(0.40 g, 77%, 8% of starting material was also recovered) as a pale
yellow oil. [a]D =+59 (c=0.165, CHCl3); IR (neat): n˜ =1496, 1454, 1372,
2,6-Anhydro-1-deoxy-1,1-difluoro-1-C-phenyl-d-glycero-l-gluco-heptitol
(2): Pd/C 10% (0.80 g) under argon was added to a solution of the
benzyl-protected difluoride 11 (0.372 g, 0.57 mmol) in EtOH (45 mL).
Argon was exchanged by H2 gas and the reaction mixture was stirred
under an H2 atmosphere (1 atm) at 258C for 38 h. Then, it was filtered
carefully through Celite and the filtrate was concentrated in vacuo. The
residue was purified by FC (10% MeOH in CH2Cl2) to give 2 (0.116 g,
70%) as a colorless oil that solidified after drying and standing in the
freezer. Some of the oil was recrystallized from iPrOH, affording color-
less crystals of 2. m.p. 128–1308C; [a]D25 =+34 (c=0.085, MeOH); IR
(neat): n˜ =3345, 2925, 2483 (br, OH), 1451, 1271, 1062, 1002, 971, 763,
700 cmÀ1 1H NMR (400 MHz, CD3OD, 298 K): d=7.56–7.59 (m, 2H,
;
3
3
Hortho), 7.42–7.45 (m, 3H, Hmeta, Hpara), 4.41 (ddd, 1H, JH1,Fa ꢀ JH1,Fb
3
ꢀ14.3 Hz, 3JH1,H2 =2.8 Hz, H1), 4.16 (dd, 1H, 3JH4,H5 =4.9 Hz, JH4,H3
=
3.0 Hz, H4), 3.99 (ddd, 1H, 3JH5,H6a =7.0 Hz, JH5,H4 ꢀ JH5,H6b ꢀ4.6 Hz,
3
3
3
3
1270, 1205, 1094, 1028, 914, 735 cmÀ1 1H NMR (400 MHz, CDCl3,
;
H5), 3.89 (dd, 1H, JH3,H2 ꢀ6.6 Hz, JH3,H4 ꢀ3.4 Hz, H3), 3.86 (dd, 1H,
3JH2,H3 ꢀ6.5 Hz, JH2,H1 ꢀ3.2 Hz, H2), 3.83 (dd, 1H, 2JH6a,H6b =12.0 Hz,
3
298 K): d=7.53–7.56 (m, 2H), 7.23–7.37 (m, 17H), 7.15–7.21 (m, 4H),
3JH6a,H5 =7.2 Hz, H6a), 3.68 ppm (dd, 1H, 2JH6b,H6a =12.0 Hz, JH6b,H5
=
=
=
3
7.11–7.14 (m, 2H) Harom, 4.59 (d, 1H, 2J=12.0 Hz), 4.53 (d, 1H, 2J=
4.2 Hz, H6b); 13C NMR (100 MHz, CD3OD, 298 K): d=137.5 (t, JC,Fa
2
2
12.9 Hz), 4.51 (brs, 2H), 4.45 (d, 1H, J=12.0 Hz) CH2 of BnO, 4.39–4.42
2JC,Fb =25.3 Hz, Cipso), 130.9 (Cpara), 129.2 (Cmeta), 126.7 (t, 3JC,Fa =3JC,Fb
2
2
(m, 1H, H5), 4.40 (d, 1H, J=12.0 Hz), 4.27 (d, 1H, J=12.0 Hz) CH2 of
6.8 Hz, Cortho), 123.3 (t, 1JC,Fa =1JC,Fb =249 Hz, CF2), 78.2 (C5), 73.6 (t,
2JC1,Fa =2JC1,Fb =27.3 Hz, C1), 72.2 (C3), 69.9 (C2), 67.4 (C4), 60.3 ppm
(C6); 19F NMR (376 MHz, CD3OD CFCl3, 298 K): d=À99.5 (brd, 1F,
2
BnO, 4.23–4.29 (m, 1H, H1), 4.16 (d, 1H, J=12.0 Hz, CH2 of BnO), 4.11
(dd, 1H, 3JH4,H5 =5.5 Hz, 3JH4,H3 =2.8 Hz, H4), 3.88 (dd, 1H, JH2,H3
3
ꢀ4.6 Hz, JH2,H1 ꢀ1.9 Hz, H2), 3.83 (dd, 1H, 2JH6a,H6b =11.8 Hz, JH6a,H5
=
3
3
2JFa,Fb =254 Hz, Fa), À101.1 ppm (brdd, 1F, 2JFb,Fa =254 Hz, JFb,H1
3
8.2 Hz, H6a), 3.74 (dd, 1H, 3JH3,H2 =4.9 Hz, 3JH3,H4 =2.8 Hz, H3),
3.71 ppm (dd, 1H, 2JH6b,H6a =12.0 Hz, 3JH6b,H5 =2.8 Hz, H6b); 13C NMR
(100 MHz, CDCl3, 298 K): d=138.5, 138.2, 138.1, 137.8 Cipso of BnO,
ꢀ15.3 Hz, Fb); ESI QqTOF MS: calcd for ([M+Na]) 313.0863, found
313.0866; elemental analysis calcd (%) for C13H16F2O5: C 53.79, H 5.56;
found: C 53.79, H 5.54.
136.1 (t, 2JC,Fa =2JC,Fb =25.7 Hz, Cipso of Ph), 129.8, 128.35, 128.3, 128.2,
3
128.1, 127.8, 127.7, 127.6, 127.5, 127.3 Carom of BnO, 125.7 (t, JC,Fa
=
2,6-Anhydro-1-deoxy-1,1-difluoro-1-C-phenyl-d-glycero-l-manno-heptitol
(3): Benzyl-protected difluoride 12 (0.021 g, 0.032 mmol) was submitted
to hydrogenation under the conditions described above for the prepara-
tion of compound 2. FC (10% MeOH in CH2Cl2) afforded 3 as a yellow-
ish oil (0.006 g, 67%). [a]D25 =À14 (c=0.07, MeOH); IR (neat): n˜ =
3661, 3423 (br, OH), 1449, 1393, 1236, 1144, 1100, 1075, 1066, 1051, 983,
3JC,Fb =6.4 Hz, Cortho of Ph), 120.8 (dd, 1JC,Fa =253 Hz, 1JC,Fb =245 Hz,
CF2), 75.6 (C5), 75.0 (C3), 74.4 (C2), 73.3 (CH2 of BnO), 73.2 73.15 (C4,
CH2 of BnO), 72.9, 72.0 (2ꢅCH2 of BnO), 70.8 (dd, 2JC1,Fa =33.7 Hz,
2JC1,Fb =25.7 Hz, C1), 65.7 ppm (C6); 19F NMR (376 MHz, CDCl3 CFCl3,
298 K): d=À100.8 (brd, 1F, 2JFa,Fb =257 Hz, Fa), À106.2 ppm (brd, 1F,
2JFb,Fa =256 Hz, Fb); ESI QqTOF MS: calcd for ([M+Na]) 673.2742,
found 673.2744; elemental analysis calcd (%) for C41H40F2O5: C 75.67, H
6.20; found: C 75.60, H 6.32.
862, 760, 692 cmÀ1
(m, 2H, Hortho), 7.38–7.43 (m, 3H, Hmeta, Hpara), 3.87 (brd, 1H, JH4,H3
;
1H NMR (400 MHz, CD3OD, 298 K): d=7.55–7.59
3
=
3.2 Hz, H4), 3.57–3.74 (m, 4H, H1, H2, H6a,b), 3.44–3.51 ppm (m, 2H,
H3, H5); 13C NMR (100 MHz, CD3OD, 298 K): d=136.8 (Cipso), 130.8
2,6-Anhydro-1-deoxy-1,1-difluoro-3,4,5,7-tetra-O-benzyl-1-C-phenyl-d-
glycero-l-manno-heptitol (12): In a PET vial containing the b-ketone 10
(0.040 g, 0.064 mmol), was added a solution of 50% Deoxo-Fluor in THF
(0.60 mL, 0.62 g, 1.40 mmol) followed by two drops of HF-pyridine as the
catalyst at 258C under argon. The reaction mixture was heated to 508C
for 5 h, 608C for 8 h, 658C for 8 h, and 708C for 8 h (in this way THF
was evaporated slowly out of the reaction mixture). The reaction was
quenched by pouring the mixture dropwise into a mixture of saturated
aqueous NaHCO3 and ice. It was allowed to reach 258C and was then ex-
tracted with Et2O (3ꢅ15 mL). The combined organic layers were washed
with brine, dried over MgSO4, and concentrated in vacuo. The residue
was purified by FC (5% EtOAc in petroleum ether) to give 12 (0.009 g,
23%) as a yellowish oil (28% of starting material was also recovered).
[a]D25 =+12 (c=0.155, CHCl3); IR (neat): n˜ =1453, 1406, 1394, 1380,
3
1
(Cpara), 128.9 (Cmeta), 127.3 (t, JC,Fa =3JC,Fb =6.4 Hz, Cortho), 121.9 (t, JC,Fa
=
1JC,Fb =247 Hz, CF2), 81.5 (t, 2JC1,Fa =2JC1,Fb =29.0 Hz, C1), 80.3 (C5), 76.2
(C3), 70.1 (C4), 68.4 (C2), 62.2 ppm (C6); 19F NMR (376 MHz, CD3OD
CFCl3, 298 K): d=À96.4 (brdd, 1F, 2JFa,Fb =260 Hz, JFa,H1 ꢀ4.9 Hz, Fa),
3
À105.9 ppm (brdd, 1F, 2JFb,Fa =259 Hz, JFb,H1 ꢀ12.0 Hz, Fb); ESI QqTOF
3
MS: calcd for ([M+Na]) 313.0863, found 313.0883; elemental analysis
calcd (%) for C13H16F2O5: C 53.79, H 5.56; found: C 53.42, H 5.30.
Acknowledgements
1250, 1075, 1066, 1057, 733, 696 cmÀ1 1H NMR (400 MHz, CDCl3,
;
Financial support by the Swiss National Science Foundation (Bern) is ac-
knowledged by the Lausanne group. The group at Madrid thanks the
298 K): d=7.58–7.60 (m, 2H), 7.24–7.39 (m, 21H), 7.15–7.18 (m, 2H)
Harom, 4.95 (d, 1H, 2J=12.0 Hz), 4.84 (d, 1H, 2J=9.9 Hz), 4.75 (d, 1H,
Ministry of Science and Innovation of Spain for
a research grant
2
2
2J=11.7 Hz), 4.70 (d, 1H, J=11.4 Hz), 4.67 (d, 1H, J=9.9 Hz), 4.59 (d,
(CTQ2006–10874-C02—01) and the Munich group acknowledges gener-
ous support by the research initiative LMUexcellent and an EC Marie
Curie Research Training network (MRTN-CT-2005—019561). We also
thank M. Rey, Dr. A. Razaname, A. Gillig, Dr. R. Scopelliti for technical
help, and C. Flowers for revising the manuscript (EPFL).
1H, 2J=12.0 Hz), 4.27 (brs, 2H) 4ꢅCH2 of BnO, 4.14 (dd, 1H, JH2,H1
=
3
9.6 Hz, 3JH2,H3 =9.2 Hz, H2), 3.92 (d, 1H, 3JH4,H3 =2.8 Hz, H4), 3.86 (ddd,
1H, 3JH1,Fb =14.5 Hz, 3JH1,H2 =9.6 Hz, 3JH1,Fa =5.5 Hz, H1), 3.67 (dd, 1H,
3JH3,H2 =9.2 Hz, 3JH3,H4 =2.8 Hz, H3), 3.47–3.53 ppm (m, 3H, H5, H6a,b);
13C NMR (100 MHz, CDCl3, 298 K): d=138.7, 138.2, 138.1, 137.9 Cipso of
BnO, 135.4 (t, 2JC,Fa =2JC,Fb =25.7 Hz, Cipso of Ph), 129.7, 128.4, 128.32,
128.3, 128.2, 128.16, 127.9, 127.73, 127.68, 127.6, 127.56, 127.4 Carom of
[1] The Sugar Code. Fundamentals of Glycosciences (Ed.: H.-J. Gabius),
Wiley-VCH, Weinheim, 2008.
BnO, 126.1 (t, 3JC,Fa =3JC,Fb =6.4 Hz, Cortho of Ph), 120.3 (t, 1JC,Fa =1JC,Fb
=
Chem. Eur. J. 2009, 15, 2861 – 2873
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2871

P. Vogel, J. Jimꢁnez-Barbero et al.
[2] a) D. Solꢉs, J. Jimꢁnez-Barbero, H. Kaltner, A. Romero, H.-C. Sie-
bert, C.-W. von der Lieth, H.-J. Gabius, Cells Tissues Organs 2001,
168, 5–23; b) H.-J. Gabius, H.-C. Siebert, S. Andrꢁ, J. Jimꢁnez-Bar-
bero, H. Rꢇdiger, ChemBioChem 2004, 5, 740–764.
[3] a) H.-J. Gabius, Biochimie 2001, 83, 659–666; b) F. Stirpe, Toxicon
2004, 44, 371–383; c) M. Jimꢁnez, J. L. Sꢊiz, S. Andrꢁ, H.-J. Gabius,
D. Solꢉs, Glycobiology 2005, 15, 1386–1395.
[4] a) S. Andrꢁ, Z. Pei, H.-C. Siebert, O. Ramstrçm, H.-J. Gabius,
Bioorg. Med. Chem. 2006, 14, 6314–6326; b) S. Andrꢁ, C. E. P. Mal-
jaars, K. M. Halkes, H.-J. Gabius, J. P. Kamerling, Bioorg. Med.
Chem. Lett. 2007, 17, 793–798.
[5] a) P. Sears, C.-H. Wong, Angew. Chem. 1999, 111, 2446–2471;
Angew. Chem. Int. Ed. 1999, 38, 2300–2324; b) L. A. Marcaurelle,
C. R. Bertozzi, Chem. Eur. J. 1999, 5, 1384–1390.
[6] a) J.-M. Beau, B. Vauzeilles, T. Skrydstrup in Glycoscience: Chemis-
try and Chemical Biology, Vol. 3 (Eds.: B. Fraser-Reid, K. Tatsuta, J.
Thiem), Springer, Berlin, 2001, pp. 2679–2724; b) Glycomimetics:
Modern Synthetic Methodologies (Ed.: R. Roy), ACS, Washington,
2005; c) D. E. Levy in The Organic Chemistry of Sugars (Eds.: D. E.
Levy, P. Fꢇgedi), CRC, Boca Raton, 2006, pp. 269–348.
[7] a) J. L. Asensio, F. J. CaÇada, X. Cheng, N. Khan, D. R. Mootoo, J.
Jimꢁnez-Barbero, Chem. Eur. J. 2000, 6, 1035–1041; b) V. Garcꢉa-
Aparicio, M. d. C. Fernꢊndez-Alonso, J. Angulo, J. L. Asensio, F. J.
CaÇada, J. Jimꢁnez-Barbero, D. R. Mootoo, X. Cheng, Tetrahedron:
Asymmetry 2005, 16, 519–527.
Chem. Soc. 2007, 129, 15420–15421; k) F. Poulain, A.-L. Serre, J.
Lalot, E. Leclerc, J.-C. Quirion, J. Org. Chem. 2008, 73, 2435–2438.
[15] J. Pꢁrez-Castells, J. J. Hernꢊndez-Gay, R. W. Denton, K. A. Tony,
D. R. Mootoo, J. Jimꢁnez-Barbero, Org. Biomol. Chem. 2007, 5,
1087–1092.
[16] a) R. R. Schmidt, H. Dietrich, Angew. Chem. 1991, 103, 1348–1349;
Angew. Chem. Int. Ed. Engl. 1991, 30, 1328–1329; b) H. Dietrich,
R. R. Schmidt, Carbohydr. Res. 1993, 250, 161–176; c) C. R. John-
son, B. A. Johns, Synlett 1997, 1406–1408; d) P. Blom, B. Ruttens, S.
Van Hoof, I. Hubrecht, J. van der Eycken, B. Sas, J. Van hemel, J.
Vandenkerckhove, J. Org. Chem. 2005, 70, 10109–10112.
[17] Access to a-CF2-galactosides was shown by Quirion and co-workers,
see ref. [14a].
[18] a) O. E. Galanina, H. Kaltner, L. S. Khraltsova, N. V. Bovin, H.-J.
Gabius, J. Mol. Recognit. 1997, 10, 139–147; b) M. Jimꢁnez, S.
Andrꢁ, H.-C. Siebert, H.-J. Gabius, D. Solꢉs, Glycobiology 2006, 16,
926–937.
[19] For glycosyl carbaldehydes, see: a) W. R. Kobertz, C. R. Bertozzi,
M. D. Bednarski, Tetrahedron Lett. 1992, 33, 737–740; b) A. Dondo-
ni, M.-C. Scherrmann, J. Org. Chem. 1994, 59, 6404–6412; c) A.
Dondoni, Pure Appl. Chem. 2000, 72, 1577–1588; d) A. Dondoni,
A. Marra, Chem. Rev. 2004, 104, 2557–2599; e) J. Zeitouni, S. Norsi-
kian, A. Lubineau, Tetrahedron Lett. 2004, 45, 7761–7763; f) G.
Chen, M. Chien, M. Tsuji, R. W. Franck, ChemBioChem 2006, 7,
1017–1022.
[8] a) M. S. Searle, D. H. Williams, J. Am. Chem. Soc. 1992, 114, 10690–
10697; b) J. L. Asensio, J. F. Espinosa, H. Dietrich, F. J. CaÇada,
R. R. Schmidt, M. Martꢉn-Lomas, S. Andrꢁ, H.-J. Gabius, J. Jimꢁ-
nez-Barbero, J. Am. Chem. Soc. 1999, 121, 8995–9000.
[9] J.-F. Espinosa, M. Bruix, O. Jarreton, T. Skrydstrup, J.-M. Beau, J. Ji-
mꢁnez-Barbero, Chem. Eur. J. 1999, 5, 442–448.
[10] a) B. E. Smart, J. Fluorine Chem. 2001, 109, 3–11; b) J. C. Biffinger,
H. W. Kim, S. G. DiMagno, ChemBioChem 2004, 5, 622–627.
[11] a) M. J. Tozer, T. F. Herpin, Tetrahedron 1996, 52, 8619–8683; b) D.
OꢈHagan, H. S. Rzepa, Chem. Commun. 1997, 645–652; c) H.-J.
Bçhm, D. Banner, S. Bendels, M. Kansy, B. Kuhn, K. Mꢇller, U.
Obst-Sander, M. Stahl, ChemBioChem 2004, 5, 637–643; d) Y. Xu,
L. Qian, A. V. Pontsler, T. M. McIntyre, G. D. Prestwich, Tetrahe-
dron 2004, 60, 43–49.
[20] a) A. Giannis, K. Sandhoff, Tetrahedron Lett. 1985, 26, 1479–1482;
b) L. Krçger, D. Henkensmeier, A. Schꢃfer, J. Thiem, Bioorg. Med.
Chem. Lett. 2004, 14, 73–75.
[21] A. Dondoni, N. Catozzi, A. Marra, J. Org. Chem. 2005, 70, 9257–
9268.
[22] A. Dondoni, N. Catozzi, A. Marra, J. Org. Chem. 2004, 69, 5023–
5036.
[23] a) G. S. Lal, G. P. Pez, R. J. Pesaresi, F. M. Prozonic, H. Cheng, J.
Org. Chem. 1999, 64, 7048–7054; b) R. P. Singh, J. M. Shreeve, Syn-
thesis 2002, 2561–2578.
[24] NMR Spectroscopy of Glycoconjugates (Eds.: J. Jimꢁnez-Barbero, T.
Peters), Wiley-VCH, Weinheim, 2002.
[25] N. L. Allinger, Y. H. Yuh, J. H. Lii, J. Am. Chem. Soc. 1989, 111,
8551–8566.
[12] a) N. C. Craig, A. Chen, K. H. Suh, S. Klee, G. C. Mellau, B. P. Win-
newisser, M. Winnewisser, J. Am. Chem. Soc. 1997, 119, 4789–4790;
b) C. R. S. Briggs, D. OꢈHagan, H. S. Rzepa, A. M. Z. Slawin, J. Flu-
orine Chem. 2004, 125, 19–25.
[26] K. Bock, J. O. Duus, J. Carbohydr. Chem. 1994, 13, 513–543.
[27] D. Neuhaus, M. P. Williamson, The Nuclear Overhauser Effect in
Structural and Conformational Analysis, Wiley-VCH, Weinheim,
1989.
[13] For CF2-furanosides see: a) W. B. Motherwell, B. C. Ross, M. J.
Tozer, Synlett 1989, 68–70; b) J. S. Houlton, W. B. Motherwell, B. C.
Ross, M. J. Tozer, D. J. Williams, A. M. Z. Slawin, Tetrahedron 1993,
49, 8087–8106; c) T. F. Herpin, W. B. Motherwell, M. J. Tozer, Tetra-
hedron: Asymmetry 1994, 5, 2269–2282; d) T. F. Herpin, W. B.
Motherwell, J.-M. Weibel, Chem. Commun. 1997, 923–924; e) J. Ko-
vensky, D. Burrieza, V. Colliou, A. F. Cirelli, P. Sinaꢋ, J. Carbohydr.
Chem. 2000, 19, 1–12; f) H. Berber, T. Brigaud, O. Lefebvre, R.
Plantier-Royon, C. Portella, Chem. Eur. J. 2001, 7, 903–909.
[14] For CF2- and CHF-pyranosides see: a) S. Marcotte, F. DꢈHooge, S.
Ramadas, C. Feasson, X. Pannecoucke, J. C. Quirion, Tetrahedron
Lett. 2001, 42, 5879–5882; b) J. Jimꢁnez-Barbero, R. Demange, K.
Schenk, P. Vogel, J. Org. Chem. 2001, 66, 5132–5138; c) A. B.
Cuenca, F. DꢈHooge, V. Gouge, G. Castelot-Deliencourt, H. Oulya-
di, E. Leclerc, P. Jubault, X. Pannecoucke, J.-C. Quirion, Synlett
2005, 2627–2630; d) A. Wegert, M. Hein, H. Reinke, N. Hoffmann,
R. Miethchen, Carbohydr. Res. 2006, 341, 2641–2652; e) J. Picard,
N. Lubin-Germain, J. Uzie, J. Auge, Synthesis 2006, 979–982; f) N. P.
Karche, C. Pierry, F. Poulain, H. Oulyadi, E. Leclerc, X. Panne-
coucke, J.-C. Quirion, Synlett 2007, 123–126; g) B. Moreno, C.
Quehen, M. Rose-Hꢁlꢀne, E. Leclerc, J.-C. Quirion, Org. Lett. 2007,
9, 2477–2480; h) K. A. Tony, R. W. Denton, A. Dilhas, J. Jimꢁnez-
Barbero, D. R. Mootoo, Org. Lett. 2007, 9, 1441–1444; i) R. W.
Denton, K. A. Tony, J. J. Hernꢊndez-Gay, F. J. CaÇada, J. Jimꢁnez-
Barbero, D. R. Mootoo, Carbohydr. Res. 2007, 342, 1624–1635; j) G.
Hirai, T. Watanabe, K. Yamaguchi, T. Miyagi, M. Sodeoka, J. Am.
[28] J. L. Asensio, F. J. CaÇada, A. Garcꢉa-Herrero, M. T. Murillo, A.
Fernꢊndez-Mayoralas, B. A. Johns, J. Kozak, Z. Zhu, C. R. Jonson, J.
Jimꢁnez-Barbero, J. Am. Chem. Soc. 1999, 121, 11318–11329.
[29] A. Poveda, J. L. Asensio, M. Martꢉn-Pastor, J. Jimꢁnez-Barbero, J.
Biomol. NMR 1997, 10, 29–43.
[30] M. Mayer, B. Meyer, J. Am. Chem. Soc. 2001, 123, 6108–6117.
[31] R. T. Lee, H.-J. Gabius, Y. C. Lee, J. Biol. Chem. 1992, 267, 23722–
23727.
[32] J. L. Asensio, M. Martꢉn-Pastor, J. Jimꢁnez-Barbero, Int. J. Biol.
Macromol. 1995, 17, 137–148.
[33] J. Jimꢁnez-Barbero, J. F. Espinosa, J. L. Asensio, F. J. CaÇada, A.
Poveda, Adv. Carbohydr. Chem. Biochem. 2000, 56, 235–284.
[34] CCDC-683094 (2) contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge from The
Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/
data_request/cif
[35] J. Angulo, B. Langpap, A. Blume, T. Biet, B. Meyer, N. R. Krishna,
H. Peters, M. M. Palcic, T. Peters, J. Am. Chem. Soc. 2006, 128,
13529–13538.
[36] Maestro, A powerful, all-purpose molecular modelling environment,
Schrodinger LLC, 2005.
[37] W. C. Still, A. Tempczyk, R. C. Hawley, T. Hendrickson, J. Am.
Chem. Soc. 1990, 112, 6127–6129.
[38] J. W. Keepers, T. L. James, J. Magn. Reson. 1984, 57, 404–426.
[39] J. M. Alonso-Plaza, M. A. Canales, M. Jimꢁnez, J. L. Roldꢊn, A.
Garcꢉa-Herrero, L. Iturrino, J. L. Asensio, F. J. CaÇada, A. Romero,
2872
www.chemeurj.org
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 2861 – 2873

Analysis of (a-d-Galactosyl)phenylmethane and a-,b-Difluoromethane Analogues
FULL PAPER
H.-C. Siebert, S. Andrꢁ, D. Solꢉs, H.-J. Gabius, J. Jimꢁnez-Barbero,
Biochim. Biophys. Acta Gen. Subj. 2001, 1568, 225–236.
H. C. Jarrell, Biochemistry 2007, 46, 36–44; b) T. Haselhorst, H.
Blanchard, M. Frank, M. J. Kraschnefski, M. J. Kiefel, A. J. Szyczew,
J. C. Dyason, F. Fleming, G. Holloway, B. S. Coulson, M. von Itz-
stein, Glycobiology 2007, 17, 68–81.
[40] a) H.-J. Gabius, H. Walzel, S. S. Joshi, J. Kruip, S. Kojima, V. Gerke,
H. Kratzin, S. Gabius, Anticancer Res. 1992, 12, 669–676; b) H.-J.
Gabius, F. Darro, M. Remmelink, S. Andrꢁ, J. Kopitz, A. Danguy, S.
Gabius, I. Salmon, R. Kiss, Cancer Invest. 2001, 19, 114–126; c) S.
Andrꢁ, H. Kaltner, T. Furuike, S.-I. Nishimura, H.-J. Gabius, Bio-
conjugate Chem. 2004, 15, 87–98; d) J. C. Manning, K. Seyrek, H.
Kaltner, S. Andrꢁ, F. Sinowatz, H.-J. Gabius, Histol. Histopathol.
2004, 19, 1043–1060; e) V. Garcꢉa-Aparicio, M. Sollogoub, Y. Blꢁri-
ot, V. Colliou, S. Andrꢁ, J. C. Asensio, F. J. GaÇada, H.-J. Gabius, P.
Sinaꢋ, J. Jimꢁnez-Berbero, Carbohydr. Res. 2007, 342, 1918–1928.
[41] For recent examples in the carbohydrate field, see: a) R. S. Houlis-
ton, N. Yuki, T. Hirama, N. H. Khieu, J.-R. Brisson, M. Gilbert,
[42] K. Stott, J. Stonehouse, J. Keeler, T.-L. Hwang, A. J. Shaka, J. Am.
Chem. Soc. 1995, 117, 4199–4200.
[43] M. Jimꢁnez, S. Andrꢁ, H.-C. Siebert, H.-J. Gabius, D. Solꢉs, Glycobi-
ology 2006, 16, 926–937.
[44] G. M. Morris, D. S. Goodsell, R. S. Halliday, R. Huey, W. E. Hart,
R. K. Belew, A. J. Olson, J. Comput. Chem. 1998, 19, 1639–1662.
Received: July 9, 2008
Published online: February 2, 2009
Chem. Eur. J. 2009, 15, 2861 – 2873
ꢄ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
2873