Exploring specificity of glycosyltransferases: synthesis of new sugar nucleotide related molecules as putative donor substrates

Article abstract of DOI:10.1016/j.carres.2007.11.009

We investigated the specificity of glycosyltransferases toward donor substrates in two complementary directions. First we prepared simple N-acetyl-α-d-glucosamine 1-diphosphates: methyl-(2-acetamido-2-deoxy-α-d-glucopyranosyl)-diphosphate, benzyl-(2-acetamido-2-deoxy-α-d-glucopyranosyl)-diphosphate, 4-phenylbutyl-(2-acetamido-2-deoxy-α-d-glucopyranosyl)-diphosphate , by the coupling of the corresponding activated alkyl phosphates with N-acetyl-α-d-glucosamine 1-phosphate. These diphosphates as well as 2-acetamido-2-deoxy-α-d-glucopyranose 1-diphosphate, tested as donors of N-acetylglucosamine in a reaction catalyzed by Neisseria meningitidis N-acetylglucosaminyltransferase (LgtA), proved to be devoid of activity. Evaluated as inhibitors, only 2-acetamido-2-deoxy-α-d-glucopyranose 1-diphosphate showed some inhibitory activity with an IC50 value of 7 mM. In the second approach, we prepared sugar nucleotide mimics having the diphosphate bridge replaced by the oxycarbonylaminosulfonyl linker. The surrogate of GDP-Fuc was synthesized as a 9:1 α/β anomeric mixture, in 40% yield, starting from chlorosulfonyl isocyanate, perbenzylated l-fucopyranose, and a guanosine derivative, protected on the exocyclic amine and secondary hydroxyl functions of ribose. Then two deprotection steps, hydrogenolysis and enzymatic hydrolysis catalyzed by penicillin G amidase afforded the target molecule to be tested as fucose donor with recombinant human α-(1→3/4)-fucosyltransferase (FucT-III). Tested as a 4:1 α/β anomeric mixture, both in the absence and in the presence of cationic cofactors, this new guanosine fucose conjugate proved to be ineffective. Its inhibitory activity toward FucT-III evaluated through a competition fluorescence assay was very poor (IC50 value of 20 mM). The surrogate of UDP-GlcNAc that was already known as its protected acetylated derivative, tested as N-acetylglucosamine donor with LgtA in the presence of Mn²⁺ turned out not to be active either.

Full text of DOI:10.1016/j.carres.2007.11.009

Available online at www.sciencedirect.com
Carbohydrate Research 343 (2008) 167–178
Exploring specificity of glycosyltransferases: synthesis of new
sugar nucleotide related molecules as putative donor substrates
*
´
Amira Khaled, Olga Piotrowska, Katarzyna Dominiak and Claudine Auge
Glycochimie Mole´culaire et Macromole´culaire, Laboratoire de Chimie Organique Multifonctionnelle, UMR 8182,
ˆ
Institut de Chimie Mole´culaire et des Mate´riaux d’Orsay (ICMMO), Batiment 420, Univ Paris-Sud, F-91405 Orsay, France
Received 20 September 2007; received in revised form 23 October 2007; accepted 7 November 2007
Available online 17 November 2007
Abstract—We investigated the specificity of glycosyltransferases toward donor substrates in two complementary directions. First we
prepared simple N-acetyl-a-D-glucosamine 1-diphosphates: methyl-(2-acetamido-2-deoxy-a-D-glucopyranosyl)-diphosphate, benzyl-
(2-acetamido-2-deoxy-a-D-glucopyranosyl)-diphosphate, 4-phenylbutyl-(2-acetamido-2-deoxy-a-D-glucopyranosyl)-diphosphate,
by the coupling of the corresponding activated alkyl phosphates with N-acetyl-a-D-glucosamine 1-phosphate. These diphosphates
as well as 2-acetamido-2-deoxy-a-D-glucopyranose 1-diphosphate, tested as donors of N-acetylglucosamine in a reaction catalyzed
by Neisseria meningitidis N-acetylglucosaminyltransferase (LgtA), proved to be devoid of activity. Evaluated as inhibitors, only
2-acetamido-2-deoxy-a-D-glucopyranose 1-diphosphate showed some inhibitory activity with an IC50 value of 7 mM.
In the second approach, we prepared sugar nucleotide mimics having the diphosphate bridge replaced by the oxycarbonyl-
aminosulfonyl linker. The surrogate of GDP-Fuc was synthesized as a 9:1 a/b anomeric mixture, in 40% yield, starting from chloro-
sulfonyl isocyanate, perbenzylated L-fucopyranose, and a guanosine derivative, protected on the exocyclic amine and secondary
hydroxyl functions of ribose. Then two deprotection steps, hydrogenolysis and enzymatic hydrolysis catalyzed by penicillin G
amidase afforded the target molecule to be tested as fucose donor with recombinant human a-(1!3/4)-fucosyltransferase (FucT-III).
Tested as a 4:1 a/b anomeric mixture, both in the absence and in the presence of cationic cofactors, this new guanosine fucose
conjugate proved to be ineffective. Its inhibitory activity toward FucT-III evaluated through a competition fluorescence assay
was very poor (IC50 value of 20 mM). The surrogate of UDP-GlcNAc that was already known as its protected acetylated derivative,
tested as N-acetylglucosamine donor with LgtA in the presence of Mn2+ turned out not to be active either.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Keywords: Sugar nucleotides; Donor substrate; Surrogate; Fucosyltransferase; N-Acetylglucosaminyltransferase; Inhibitor
1. Introduction
Adding an His tag to the primary sequence of recom-
binant enzymes is a further advantage that allows to
immobilize enzymes on Ni2+–agarose and use them in
this form for synthetic purposes,3 a procedure that is
now widely applied in glycobiotechnology.4 Neverthe-
less the major drawback to the enzymatic approach
remains the requirement for sugar nucleotides, exhibiting
high affinity constant for glycosyltransferases. A signifi-
cant advance in this area would be the availability of do-
nor substrates with lower affinity but simpler structures
than sugar nucleotides.5 Research work in this direction
is rather limited. However a very recent report by
Withers and co-workers is very worthy of note; these
authors demonstrated that p-nitrophenyl a-sialoside
could act as an alternative donor substrate for a
Glycosyltransferases of Leloir-type that typically cata-
lyze sugar units transfer from a sugar nucleotide donor
to an unprotected saccharide acceptor with complete
regio- and stereoselectivity, have proved to be precious
tools in preparative oligosaccharide synthesis.1,2
Through cloning strategy these enzymes, originally
membrane-associated proteins, are now produced in
the secreted form, in expression systems such as bacte-
ria, yeasts, fungi, baculovirus-infected insect cells, mak-
ing glycosyltransferases easily available to chemists.
*
Corresponding author. E-mail: clauauge@icmo.u-psud.fr
0008-6215/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2007.11.009

168
A. Khaled et al. / Carbohydrate Research 343 (2008) 167–178
bacterial a-(2!3/2!8)-sialyltransferase in the presence
of catalytic amounts of CMP, thereby replacing CMP-
NeuAc, the natural donor.6 The validity of this elegant
strategy remains to be proved for other glycosyltransfer-
ases, especially those from animal sources.
ration (Chart 1). LgtA has an absolute requirement for
Mn2+, which co-ordinates the pyrophosphate group of
the donor substrate. Mn2+ binding is fulfilled by the res-
idues of the conserved D · D motif, which has been
identified in many glycosyltransferases.10
As for us we addressed the same topic according to
two different and complementary approaches. Our pre-
vious work has shown a relative flexibility of glycosyl-
transferases toward the nucleotide part of the
molecule, since we found that the ability of the sugar
nucleotide to act as a donor substrate is retained with
unnatural nucleobase.7 In continuation of this work
we further investigated the donor specificity of a bacte-
rial glycosyltransferase, Neisseria meningitidis UDP-
GlcNAc: b-D-Galp-(1!4)-b-D-Glcp-(1!30)-b-N-acetyl-
glucosaminyl-transferase (EC 2.4.1.56, LgtA) cloned
and expressed in Escherichia coli,8,9 first by testing
very simple 2-acetamido-2-deoxy-a-D-glucopyranose 1-
diphosphates. LgtA catalyzes transfer of N-acetylglucos-
amine from UDP-GlcNAc onto position-3 of the termi-
nal galactose of lactose unit with an inversion of
configuration at the anomeric carbon of glucosamine.
By analogy with glycosidase reactions the acceptor
hydroxyl group attacks the C-1 of glucosamine in an
SN2-like mechanism leading to the inversion of configu-
Here we report the synthesis of diphosphates 1, 2, 3,
and 4 (Chart 2), related to UDP-GlcNAc, where the
nucleotide part of the molecule was replaced by methyl,
benzyl, hydrogen, or 4-phenylbutyl groups, which were
tested as donors of N-acetylglucosamine in the reaction
catalyzed by LgtA.
In the second approach we were interested in devising
substrate analogues having the diphosphate bridge
replaced by a more easily available linkage that, like
diphosphate, could act as a leaving group too. As
surrogate of the diphosphate, we selected the
oxycarbonylaminosulfonyl linker, a non-charged iso-
stere of pyrophosphate, for three reasons:
(i) Both carbonyl and sulfonyl groups have been used
as phosphate substitutes in several examples.11–13
(ii) The carbamate group is cleaved by glycosidases
with the release of CO2;14 thereby in view of some
similarity between glycosidases and glycosyltrans-
ferases, we guessed there was a reasonable chance
HO
HO
HO
OH
O
HO
O
OR
O
OH
HO
LgtA
HO
OH
O
OH
HO
HO
HO
O
O
OR
O
+
O
O
HO
OH
NH
AcHN
HO
HO
HO
UDP
O
O
P
O
P
N
O
HO
O
AcHN
O
O
O
O-
O-
Mn2+
HO
OH
Chart 1. Reaction catalyzed by Neisseria meningitidis b-(1!3)-N-acetylglucosaminyltransferase (LgtA).
HO
HO
HO
HO
HO
HO
O
O
O
O
O
O
P
O-
O
O
O
HO
HO
HO
AcHN
P
O-
P
O-
CH3
(CH2 )n
AcHN
P
O-
AcHN
P
O-
P
O-
O
O
O
O
O
OH
O
O
O
3
2 n = 1
4 n = 4
1
O
O
N
N
N
NH
NH2
NH
RO
O
S
O
O
N
RO
RO
O
O
S
O
O
O
CH3
O
NH
O
AcHN
O
OH
O
NH
O
O
OH
O
HO
HO OH
5 R = Ac
6 R = H
HO OH
7
Chart 2. Structure of the sugar nucleotide related molecules that were synthesized and tested toward glycosyltransferases.

A. Khaled et al. / Carbohydrate Research 343 (2008) 167–178
169
that the carbamate group might also be substrate
for glycosyltransferases.
phosphates 8 and 9 were treated with carbonyldiimidaz-
ole in the presence of triethylamine. The formation of
the imidazolidate could be monitored by 31P NMR spec-
troscopy20: the 31P resonance signal is shifted from 0.14
(for compound 9) to ꢀ7.3 (for compound 11). The reac-
tion of the mono-cyclohexylammonium salt of 2-acet-
amido-2-deoxy-a-D-glucopyranose 1-diphosphate 12
and activated methyl phosphate 10 was carried out in
DMF (Scheme 1). Purification by anion exchange chro-
matography on DEAE-Sephadex afforded pure pyro-
phosphate 1 in 43% yield. Condensation of activated
benzyl phosphate 11 with glucosamine phosphate 12
was performed in DMF in the presence of zinc chlo-
ride21 reported as an activator for this coupling, but
no increased yield was observed and diphosphate 2
(iii) In chemical glycosylation, glycosyl carbamates15
and glycosyl sulfonyl carbamates16 can serve as
glycosyl donors after activation with an electro-
philic promoter.
The glucosaminyl derivative 5 belonging to this class
of compounds had been previously synthesized and
tested for its antiviral activity.11 Thus, in this second
part we report on the biochemical evaluation of the
deprotected derivative 6 as a donor substrate toward
LgtA and, in addition, the synthesis of the new GDP-
Fuc mimic 7 (Chart 2), which was tested as a fucose do-
nor toward recombinant human GDP-Fuc: b-D-Galp-
(1!3/4)-b-D-GlcpNAc-(1!40/30)-a-fucosyltransferase
(EC 2.4.1.65, FucT-III).3 FucT-III catalyzes the transfer
of fucose onto position-4 or position-3 of the N-acetyl-
glucosamine unit of disaccharides b-D-Galp-(1!3)-b-
D-GlcpNAc or b-D-Galp-(1!4)-b-D-GlcpNAc, respec-
tively. Like LgtA, FucT-III is an inverting enzyme;
b-linked GDP-Fuc is the donor substrate, giving rise
to a-linked fuco-oligosaccharides (Chart 3), but as
opposed to LgtA it does not show an absolute require-
ment for Mn2+ for activity, just exhibiting some activa-
tion by this cation, depending on the pH.17 Therefore, it
was of interest to test both sugar nucleotide mimics 6
and 7.
O
P
O
a
R
P
O-
N
R
HO
O
N
O
OH
8 R = CH3
9 R = Bn
10 R = CH3
11 R = Bn
15 R = (CH2)4Ph
O
(CH2)4
HO
b
(CH2)4
P
RO
O
OR
14 R = Bn
13 R = H
c
HO
HO
2. Results and discussion
HO
O
d or e
OH
O
O
O
O
HO
HO
HO
AcHN
P
O-
The synthesis of the diphosphates of N-acetylglucos-
amine relies on the coupling of 2-acetamido-2-deoxy-a-
D-glucopyranose 1-phosphate and an activated form of
the corresponding monoalkyl phosphates. Methyl and
benzyl alcohols were phosphorylated according to an
old procedure using inorganic phosphorous acid and
mercuric chloride in the presence of triethylamine.18
The monoalkyl phosphates 8 and 9 could be isolated
and crystallized as their bis(cyclohexylammonium) salts.
As an activated form of the phosphate, the imidazoli-
date, commonly used in the synthesis of nucleotides,
was chosen.19 Thus, the acidic forms of the monoalkyl
AcHN
P
O-
P
O-
R
O
O
O
O
12
1 R = CH3
2 R = Bn
3 R = H
f
4 R = (CH2)4Ph
Scheme 1. Reagents and conditions: (a) CDI 2 equiv, NEt3 1 equiv,
DMF, 3 h, rt, then MeOH 1 equiv; (b) DIAD 1.5 equiv, PPh3
1.5 equiv, dibenzyl phosphate 1.5 equiv, THF, 4 h, rt, 60%; (c) Pd
10%/C, EtOH–M NaHCO3, H2, overnight, 69%; (d) 10 1 equiv or 15
0.66 equiv, DMF, 3 d, rt, 43% for 1, 28% for 4; (e) 11 1 equiv, ZnCl2
8 equiv, DMF, 2 d, rt, 38% for 2; (f) Pd 10%/C, MeOH–0.1 M
NaHCO3ꢀNHEt3þ, H2, 3 h, 70%.
HO
HO
OH
O
HO
HO
CH3
OH
O
NHAc
OR
NHAc
OR
FucT-III
O
O
O
O
O
HO
O
N
OH
OH
HO
HO
N
N
NH
NH2
+
O
GDP
O
OH
O
P
O-
OH
P
O-
O
HO
O
CH3
O
O
O
OH
OH
HO
HO OH
Chart 3. Reaction catalyzed by a-(1,3/4)-fucosyltransferase (FucT-III).

170
A. Khaled et al. / Carbohydrate Research 343 (2008) 167–178
was finally isolated in 38% yield after two purification
steps. The benzyl-(2-acetamido-2-deoxy-a-D-glucopyr-
anosyl)-diphosphate 2 was hydrogenolyzed in the pres-
ence of palladium on charcoal giving the N-acetyl-a-D-
glucosamine diphosphate 3 as its triethylammonium salt
after purification on anionic exchange chromatography
in 70% yield.
100
80
60
40
20
0
Mitsunobu reaction22 was applied to the synthesis of
4-phenylbutylphosphate 13: 4-phenyl-1-butanol was
reacted with dibenzylphosphate in the presence of diiso-
propyl azodicarboxylate and triphenylphosphine to give
after purification 4-phenylbutyl-dibenzylphosphate 14 in
60% yield. Catalytic hydrogenation of compound 14 in
the presence of palladium on charcoal afforded com-
pound 13 as its sodium salt in 69% yield. Phosphate
13 was converted to imidazolidate 15 as previously
described, which in turn was reacted with the mono-
triethylammonium salt of 2-acetamido-2-deoxy-a-D-
glucopyranose 1-diphosphate 12 used in excess. Two
purification procedures, anionic exchange chromatogra-
phy and gel permeation were necessary to isolate pure
diphosphate 4 in 29% yield.
0
5
10
15
20
25
30
GlcNAcPP concentration (mM)
Figure 1. Inhibition of LgtA by 2-acetamido-2-deoxy-a-D-glucopyr-
anose 1-diphosphate (GlcNAcPP) 3 as a function of the concentration.
All synthesized N-acetyl-a-D-glucosamine diphos-
phates 1, 2, 3, and 4 were tested as N-acetylglucosamine
donor toward recombinant N. meningitidis UDP-
GlcNAc: b-D-Galp-(1!4)-b-D-Glcp-(1!30)-b-N-acetyl-
glucosaminyl-transferase (LgtA). For the evaluation of
these analogues, we applied a fluorescent assay utilizing
b-D-Galp-(1!4)-b-D-Glcp-O-(CH2)6-NH-dansyl disac-
charide23 as an acceptor substrate. This disaccharide
was incubated with each of diphosphates 1, 2, 3 or 4
at a concentration of 60 mM. We speculated that, within
the active site, simple juxtaposition of diphosphate 3
and uridine in the presence of Mn2+ might replace
UDP-GlcNAc and function as a donor. Thus compound
3 was also tested in the presence of uridine, both com-
pounds at the same concentrations (60 or 100 mM).
But no conversion of substrate into product could be
observed in any of these assays. Then, we studied the
inhibitory activity of the four diphosphates toward
LgtA through a competitive fluorescent assay. Only
diphosphate 3 showed some inhibition. Inhibitory activ-
ity of LgtA by 2-acetamido-2-deoxy-a-D-glucopyranose
diphosphate 3 is reported in Figure 1 exhibiting an
IC50 value of 7 mM. It would seem clear that the
diphosphate 3 is recognized even weakly by LgtA, as
opposed to the other three diphosphates 1, 2, 4 that
are not recognized.
by the in situ reaction of the unstable intermediate with
the protected nucleotide derivative free on the primary
hydroxyl function.
Whereas the synthesis of the protected N-acetyl-
glucosamine derivative
5
had been previously
reported,11 for biochemical tests we had to prepare the
deacetylated compound 6 that to our knowledge has
not been described yet. Removal of the acetyl protecting
groups of compound 5 by treatment with methanolic
ammonia afforded compound 6 in a high yield (90%).
The UDP-GlcNAc mimic 6 was tested as a N-acetylglu-
cosamine donor toward LgtA at a concentration of
60 mM. No conversion of substrate into product could
be observed and compound 6, tested at concentrations
up to 10 mM, through a competition fluorescence assay
turned out not to show any inhibitory activity.
Nevertheless we decided to synthesize the new GDP-
Fuc mimic 7 to be tested with FucT-III, an enzyme from
animal source that could not involve the same enzymatic
mechanism as LgtA, since in particular it does not show
an absolute requirement for Mn2+. We chose the benzyl
ether as protecting groups on the fucose residue.
Regarding the nucleoside part, as protection on the exo-
cyclic amine of guanosine, we selected the phenylacetyl
group, that can be cleaved off with penicillin G amidase
under very mild conditions.24 Thus, the formation of the
phenylacetamide was performed by the treatment of
guanosine 16 with phenylacetyl chloride after transient
trimethylsilylation of hydroxyl functions,25 affording
the guanosine derivative 17 in 70% overall yield (Scheme
2). Then, the phenylacetamide derivative of guanosine
was protected as its benzylidene acetal by the treatment
of 17 with benzaldehyde and zinc chloride under
standard conditions.26 Acetal 18 was obtained as an
In the second part of this work we focused on the
synthesis of sugar nucleotide mimics in which the
diphosphate bridge was replaced by the oxycarbonyl-
aminosulfonyl linker: compounds 6 and 7, analogues
of UDP-GlcNAc and GDP-Fuc, respectively. The strat-
egy of synthesis of these compounds was based on the
coupling of a protected sugar derivative, free on its ano-
meric position, with chlorosulfonyl isocyanate, followed

A. Khaled et al. / Carbohydrate Research 343 (2008) 167–178
171
O
N
O
N
O
N
N
NH
NH
N
N
O
NH
NH2
N
NH
O
a, b
c
O
N
HO
N
NH
O
HO
O
HO
17
18
16
O
O
HO OH
HO OH
Ph
O
N
N
N
NH
O
O
O
S
d
NHR1
CH3
BnO
OH
OBn
O
O
OBn
CH3
NH2
O
OBn
O
+
O
NH
O
OBn
O
BnO
20
CH3
O
19
OR2
R3O OR4
OR2
R2O
21 R1 = COCH2Ph, R2 = Bn, R3, R4 = CHPh
22 R1 = COCH2Ph, R2 = R3 = R4 = H
e
f
OCH3
OH
CH3
O
+
OH
HO
23
R1 = R3 = R4 = H
7
Scheme 2. Reagents and conditions: (a) TMSCl 7.5 equiv, pyridine, 8 h, rt, then PhCH2COCl 1.5 equiv, HOBt 1.6 equiv, CH3CN, overnight, rt; (b)
water, 25% NH4OH, 0 ꢁC, 30 min, 70% for two steps; (c) ZnCl2 3.4 equiv, PhCHO, 50 ꢁC, 1 h, 88%; (d) (1) O@C@N–SO2Cl 1 equiv, CH2Cl2,
ꢀ35 ꢁC, 2 h, then t!ꢀ10 ꢁC; (2) 18 1 equiv, pyridine 2.5 equiv, CH2Cl2, rt, 20 h, 21: 40%, a/b = 9:1, 20: 17%; (e) Pd 10%/C, THF–MeOH–water, H2
(40 psi), 4 d, 35%; (f) penicillin G amidase on Eupergit, 40 mM phosphate buffer pH 7.8, rt, overnight, 91%.
equimolar diastereomeric mixture in excellent yield. As
the occurrence of both diastereomers appeared as a
drawback, we tried another method to obtain one
single benzylidene diastereomer,27 but our effort was
unsuccessful.
both anomers prevented the complete elucidation of
1H NMR spectra. However, in the first fraction the a
anomeric configuration was demonstrated by the pres-
ence of two doublets (d 6.10, 6.06, J 3.5 Hz) assigned
to the fucose a anomeric proton; concomitantly the
H-8 guanosine proton appeared as two singlets (d
7.99, 7.91), the benzylidene proton as two singlets (d
6.09, 5.85), and the H-1 proton of the ribose as two dou-
blets (d 6.11, 6.02). Likewise the signal for the H-6 pro-
ton of the fucose residue was split into two doublets (d
1.06, 1.03, J 6 Hz). By chance the b anomer eluted later
than the a anomer from Sephadex LH-20, resulting in a
b enriched mixture for the second fraction. The occur-
rence of the b anomer was ascertained on the basis of
1H NMR data that revealed the presence of two extra
doublets (d 5.45, 5.43) with a typical large coupling
constant (J 8.0 Hz) assigned to the fucose b anomeric
proton, upfield shifted as compared to the a anomeric
proton. In parallel the 1H NMR spectrum showed three
extra singlets, two for the H-8 proton of guanosine (d
7.98, 7.95) and one for the benzylidene proton (d
5.91). 13C NMR data also exhibited two signals for
the fucose anomeric carbon (d 95.3, 91.0). Low ratios
in the b anomer have already been reported in such reac-
tions with 2-deoxy or 6-deoxyglucose derivatives29 and
seems to be reflecting the anomeric composition of the
starting glycopyranoside hemiacetal in dichloromethane
The perbenzylated fucopyranose 19 was first treated
with chlorosulfonyl isocyanate at low temperature. An
unstable intermediate was formed by the reaction of
the anomeric hydroxyl group with isocyanate, the most
reactive group.28 Then when all starting fucose had
reacted, the guanosine derivative 18 was added to the
reaction mixture together with pyridine, to allow the
in situ reaction of the chlorosulfonyl group with the 50
hydroxyl function of 18. The product was purified by
chromatography on silica gel allowing to isolate
first the carbamoyl fucose derivative 20 (17%) arising
from the hydrolysis of the chlorosulfonyl intermediate,
and the expected fucose guanosine conjugate still con-
taminated by byproducts. Then, a further purification
by gel permeation on Sephadex LH-20 column afforded
compound 21 as two fractions: the first eluted fraction
was pure a anomer and the second one an anomeric
mixture in 4:1 a/b ratio, estimated from 1H NMR. Alto-
gether compound 21 was obtained in 40% yield and 9:1
a/b overall ratio. None of both purification procedures
was successful in isolating the pure b anomer. The
occurrence of R and S benzylidene diastereomers for

172
A. Khaled et al. / Carbohydrate Research 343 (2008) 167–178
solution at low temperature. Despite this unfavorable
ratio, compound 21, as a 4:1 a/b anomeric mixture,
was immediately submitted to deprotection steps.
Then, we studied the inhibitory activity of compound
7 toward FucT-III through a competition fluorescence
assay. Compound 7 only exhibited a very poor inhibi-
tion with an IC50 of 20 mM. It is noteworthy that a
related analogue of GDP-Fuc had been previously
reported not to show any significant inhibition.32
Removal of the benzyl and benzylidene protecting
groups by hydrogenolysis proved to be particulary
problematic. The sluggish deprotection of benzyl ethers
had been previously observed with nucleosides and may
be the result of inhibition due to the nucleobase that
could adsorb onto the catalyst.30 Therefore, we first
studied hydrogenolysis of the guanosine derivative 18:
deprotection of 18 carried out under a hydrogen pres-
sure of 40 psi in THF–methanol–water mixture, in the
presence of 10% palladium on charcoal, required two
days to reach complete reaction. These conditions were
applied to compound 21 and the formation of the de-
sired deprotected compound was monitored by HPLC
(reverse phase column, elution with a gradient of aceto-
nitrile–water). After 4 days, the reaction was stopped,
but large decomposition of the fucose guanosine conju-
gate could not be avoided; finally compound 22 could
be isolated on a C18 Sep-Pak cartridge by elution with
a gradient of methanol–water in 35% yield. One major
byproduct could be identified in the fraction eluted with
water as methyl L-fucopyranoside 23, predominantly b,
arising from the nucleophilic attack of the predominant
compound 21a by methanol in the course of hydrogen-
olysis. The proportion of the b anomer in compound 22
was 20% as confirmed by 1D and 2D NMR experi-
ments (COSY, HMQC, HMBC) and in good agree-
ment with the a/b ratio of starting 21. Enzymatic
removal of the phenylacetyl group on the exocyclic
amine of the guanosine by the treatment of compound
22 with penicillin G amidase (EC 3.5.1.11) immobilized
on Eupergit, was straightforward. Purification on a C18
Sep-Pak cartridge eluted with 9:1 water–methanol
afforded the target compound 7 (91% yield), the 13C
In conclusion we prepared several new molecules
related to the sugar nucleotides UDP-GlcNAc and
GDP-Fuc. The series of N-acetyl-a-D-glucosamine
diphosphates tested as N-acetylglucosamine donors in
LgtA-catalyzed reaction turned out to be devoid of
activity, even at high substrate concentrations (120-fold
higher than the concentration commonly used with
UDP-GlcNAc). Expected hydrophobic interactions
between amino acid residues of the enzyme active site
and either the benzyl group of compound 2 or the
4-phenylbutyl chain of compound 4 are not able to
stabilize the molecule avoiding the anomeric carbon of
2-acetamido-2-deoxy-a-D-glucopyranose 1-diphosphate
to be attacked by the 3-hydroxyl of the galactose.
Furthermore each constituent of UDP-GlcNAc, that is
uridine and N-acetyl-a-D-glucosamine diphosphate 3,
does not exhibit by itself an affinity high enough to allow
the in situ formation of a functional intermediate able to
play the role of UDP-GlcNAc. These experimental
results confirm that the nucleotide part of the molecule
is essential for donor recognition.
In the second part, we explored the ability of the oxy-
carbonylaminosulfonyl linker to replace the diphos-
phate bridge. We synthesized new sugar nucleotide
conjugates that were supposed to be donor substrates
for glycosyltransferases. But neither 6, mimic of UDP-
GlcNAc, nor 7, mimic of GDP-Fuc, tested at a concen-
tration 100-fold higher than the natural substrates,
turned out to be functional. The non-inhibitory activity
of compound 6 as well as the poor inhibitory activity of
compound 7 prove that these mimics are unable to
accommodate the enzyme active site. Molecular model-
ing experiments together with the X-ray crystallographic
structure of the enzyme active sites would be essential to
understand the binding hindrance of these analogues. In
this regard the quite recent X-ray crystal structure of
Helicobacter pylori a-(1!3/4)-fucosyltransferase could
be helpful.33 Furthermore, directed mutagenesis of the
amino acid residues involved in the binding of the nucle-
otide part may provide an enzyme with a lower specifi-
city toward the donor substrate.
1
NMR, H NMR, and mass spectral data of which were
in complete agreement with the structure. The a/b mix-
ture was not viewed as a problem to test compound 7
as a fucose donor because the a-(1!3/4)-fucosyltrans-
ferase (FucT-III) was expected to select the required b
anomer.
We examined the ability of compound 7 to act as a
fucose donor using b-D-Galp-(1!3)-b-D-GlcpNAc-O-
(CH2)6-NH-dansyl as a fluorescent acceptor substrate.23
The biochemical tests were carried out at pH 6.0 using
soluble or immobilized FucT-III, with or without the
addition of a cationic cofactor: manganese chloride or
ytterbium triflate. Indeed we previously observed that
this lanthanide salt was able to replace manganese
chloride for increasing FucT-III activity.31 Compound
7 was not available in larger amount for assays at higher
concentrations than 5 mM in the b anomer. However,
this concentration is 100-fold higher than the Km for
GDP-Fuc. Unfortunately no conversion of substrate
into product could be observed in any of the assays.
3. Experimental
3.1. General methods
NMR spectra were recorded using Bruker AC-200, AC-
250 or DRX-400 spectrometers; the chemical shifts are
given relative to the signal of tetramethylsilane in

A. Khaled et al. / Carbohydrate Research 343 (2008) 167–178
173
1
CDCl3; for H NMR and 13C NMR spectra in D2O,
53.9 (d, JC,P 8.7 Hz, C-2), 53.5 (CH3), 46.9 (CH2–
CH3)3, 22.3 (CH3), and 8.6 (CH2–CH3)3; 31P NMR
(101 MHz, D2O): d ꢀ9.70 and ꢀ13.23 (JPP 22.9 Hz);
ESIMS (negative mode): m/z 394.0 [Mꢀ2NHEt3+H]ꢀ.
HRESIMS (negative mode) m/z: calcd for C9H18O12-
N1P2, 394.0299; found, 394.0303.
acetone (d 2.22 and 30.5 ppm) was used as an internal
reference. Optical rotations were measured with a Jasco
digital micropolarimeter. Mass spectra were performed
either on a Finigan MATT 95 apparatus using the ESI
technique or on a Perspective Biosystems DE-STR-
MALDI-TOF System. Reactions were monitored by
TLC on Silica Gel 60F254 with detection by charring
with 10% H2SO4 in EtOH or 2% orcinol in 10%
H2SO4. Reversed phase HPLC (RP-HPLC) analyses
were performed using a Waters 600 pump equipped with
a reversed phase column (C18, Nucleosil, 5 lm, 3.9 mm
id · 300 mm; mobile phase, water–MeCN: 73:27; flow
rate 1 mL/min), with a Luminescence Spectrometer
LS50B (Perkin–Elmer) as the detection device. Fluores-
cence of substrates and products was read at 385 nm
excitation/540 nm emission. For NMR spectroscopy
complete signal assignments were based on COSY,
HSQC, and HMBC correlations. Penicillin G amidase
immobilized on Eupergit C was purchased from Fluka,
ytterbium triflate from Aldrich.
3.3. Benzyl-(2-acetamido-2-deoxy-a-D-glucopyranosyl)-
diphosphate, bis(triethylammonium)salt (2)
Benzyl phosphate bis(cyclohexylammonium) salt
(1 mmol) was dissolved in a small amount of water
and passed through a column of Dowex-50, H+ resin
(20 cm · 1.8 cm). Benzyl phosphoric acid (9) was eluted
with 1:1 EtOH–water and the soln was freeze-dried. To
the dry acid, carbonyldiimidazole (CDI) (0.324 g,
2 mmol), Et3N (0.13 mL, 1 mmol), and DMF (5 mL)
were added and the mixture was vigorously stirred at
rt for 3 h. To neutralize the excess of CDI, MeOH
(30 lL) was added and the reaction mixture was stirred
additionally for 15 min. Solvents were evaporated
affording crude activated benzyl phosphorimidazolidate
11; 31P NMR (101 MHz, MeOD): d ꢀ7.3. To a soln of
activated benzylphosphate 11 (0.25 mmol) and 2-acet-
amido-2-deoxy-a-D-glucopyranosyl phosphate mono-
cyclohexylammonium salt 12 (200 mg, 0.5 mmol) in
dry DMF (6 mL) was added ZnCl2 (272 mg, 2 mmol)
and the mixture was stirred at rt under argon for
48 h. Then after evaporation to dryness the residue dis-
solved in 1:1 water–MeOH was purified by anion
3.2. Methyl-(2-acetamido-2-deoxy-a-D-glucopyranosyl)-
diphosphate, bis(triethylammonium)salt (1)
Methyl phosphate bis(cyclohexylammonium) salt
(1 mmol) was dissolved in a small amount of water
and passed through a column of Dowex-50, H+ resin
(20 cm · 1.8 cm). Methyl phosphoric acid (8) was eluted
with 1:1 EtOH–water and the soln was freeze-dried. To
the dry acid, carbonyldiimidazole (CDI) (0.324 g,
2 mmol), Et3N (0.13 mL, 1 mmol), and DMF (5 mL)
were added and the mixture was vigorously stirred at
rt for 3 h. Then to neutralize the excess of CDI, MeOH
(30 lL) was added and the reaction mixture was stirred
additionally for 15 min. Solvents were evaporated
affording crude methyl phosphorimidazolidate 10; 31P
NMR (101 MHz, MeOD): d ꢀ6.35. A soln of activated
methylphosphate 10 (0.50 mmol) and 2-acetamido-2-
deoxy-a-D-glucopyranosyl phosphate mono-cyclohexyl-
ammonium salt34 12 (200 mg, 0.5 mmol) in dry DMF
(6 mL) was stirred at rt under argon for 4 d. Then, the
reaction mixture was concentrated to dryness and the
residue dissolved in water was purified by anionic
exchange chromatography on DEAE-Sephadex A-25
(HCO3ꢀ form) column (2 · 23 cm). Elution with a gradi-
ent of 0 to M triethylammonium hydrogenocarbonate
buffer (pH 7.8) afforded 1 as a white solid (128 mg,
exchange chromatography on
a DEAE-Sephadex
ꢀ
A-25 (HCO3 form) column. Elution with a gradient
of 0 to M triethyl- ammonium hydrogenocarbonate
buffer (pH 7.8) afforded 2 as its triethylammonium salt,
but still contaminated by salts. After exchanging trieth-
ylammonium for sodium salt by passing through Dow-
ex-50 (Na+ form) resin, the product dissolved in water
was applied onto a Biogel P-2 column (<45 lm,
2 · 63 cm); elution with 6 mM triethylammonium
hydrogenocarbonate buffer (pH 7.8) afforded pure
29
compound 2 as a white powder (64 mg, 38%); ½aꢁD
+52 (c 0.66, water); 1H NMR (250 MHz, D2O): d
7.50–7.35 (m, 5H, Ph), 5.47 (dd, 1H, J1,2 3 Hz, J1,P
7 Hz, H-1), 5.00 (d, 2H, JH,P 7 Hz, CH2), 4.01–3.72
(m, 5H, H-2, H-4, H-5, H-6, H-60), 3.52 (t, 1H,
J2,3 = J3,4 9.5 Hz, H-3), 3.18 (q, 12H, CH2–CH3),
1.98 (s, 3H, NHAc), and 1.24 (t, 18H, CH2–CH3);
13C NMR (62.9 MHz, D2O): d 174.9 (CO), 137.6 (d,
JC,P 6 Hz, Ar-C), 128.9, 128.5, 128.1 (Ar-C), 94.7 (d,
JC1–P 5.7 Hz, C-1), 73.2, 71.2, 69.8 (C-3, C-4, C-5),
68.2 (d, JC,P 5.5 Hz, CH2), 60.6 (C-6), 53.9 (d, JC,P
8.8 Hz, C-2), 46.9 (CH2–CH3), 22.3 (CH3), and 8.5
(CH2–CH3); 31P NMR (101 MHz, D2O): d ꢀ11.17
and ꢀ13.20 (JP,P 23.2 Hz); ESIMS (negative mode):
m/z 470.0 [Mꢀ2NHEt3+H]ꢀ. HRESIMS (negative
29
1
43%); ½aꢁD +49 (c 0.79, water); H NMR (250 MHz,
D2O): d 5.44 (dd, 1H, J1,2 3 Hz, J1,P 7 Hz, H-1), 3.99–
3.71 (m, 5H, H-2, H-4, H-5, H-6, H-60), 3.60 (d, 3H,
JH,P 11.5 Hz, CH3), 3.52 (t, 1H, J2,3 = J3,4 9.5 Hz,
H-3), 3.10 (q, 12H, J 7 Hz, CH2–CH3), 2.03 (s, 3H,
NHAc), and 1.21 (t, 18H, CH2–CH3); 13C NMR
(62.9 MHz, D2O): d 175.0 (CO), 94.7 (d, JC,P 6.3 Hz,
C-1), 73.2, 71.1, 69.8, (C-2, C-3, C-4), 60.6 (C-6),

174
A. Khaled et al. / Carbohydrate Research 343 (2008) 167–178
mode) m/z: calcd for C15H22O12N1P2, 470.0612; found,
3.6. 4-Phenylbutylphosphate disodium salt (13)
470.0617.
To a soln of compound 14 (0.41 g, 1 mmol) in 5:1
EtOH–M NaHCO3 (12 mL), 10% Pd on charcoal
(200 mg) was added and the mixture was stirred at rt
under hydrogen atmosphere overnight. Catalyst was
filtered, rinsed with 1:1 EtOH–water; filtrate and
washing were evaporated, the residue was taken up in
water, extracted with EtOAc, then the aqueous phase
was concentrated and freeze-dried yielding pure 4-phenyl-
butylphosphate disodium salt 13 as a white solid
3.4. 2-Acetamido-2-deoxy-a-D-glucopyranosyl-diphos-
phate, bis(triethylammonium)salt (3)
To a soln of compound 2 (43 mg, 64 lmol) in 1:1
MeOH–0.1 M triethylammonium hydrogenocarbonate
buffer (pH 8) (2 mL), 10% Pd on charcoal (20 mg) was
added and the mixture was stirred at rt under hydrogen
pressure for 3 h. Catalyst was filtered, rinsed with MeOH
and water; filtrate and washing were concentrated under
diminished pressure and then freeze-dried. The residue
was purified by anion exchange chromatography on a
DEAE-Sephadex A-25 (HCO3 form) column. Elution
with a gradient of 0 to M triethylammonium hydrogeno-
carbonate buffer (pH 7.8) afforded 3 as a white solid
1
(189 mg, 69%); H NMR (360 MHz, MeOD): d 7.30–
7.10 (m, 5H, Ph), 3.83 (dt, 2H, JH,P = JH,H 6.5 Hz,
CH2), 2.63 (t, 2H, J 7 Hz CH2), and 1.67 (m, 4H,
2CH2); 13C NMR (62.9 MHz, MeOD): d 143.8, 129.4,
129.2, 126.6 (Ar-C), 65.0 (CH2Ph), 36.7 (CH2), 32.0 (d,
JC,P 7.8 Hz, CH2), and 29.1 (CH2); 31P NMR
(101 MHz, D2O): d 6.17; ESIMS (negative mode): m/z
229.1 [Mꢀ2Na+H]ꢀ.
29
(26 mg, 70%); ½aꢁD +58 (c 0.54, water); 1H NMR
(250 MHz, D2O): d 5.44 (br, 1H, H-1), 4.01–3.69 (m,
5H, H-2, H-4, H-5, H-6, H-60), 3.47 (t, 1H, J2,3 = J3,4
9.5 Hz, H-3), 3.15 (q, 12H, CH2–CH3), 2.03 (s, 3H,
NHAc), and 1.24 (t, 18H, CH2–CH3); 13C NMR
(62.9 MHz, D2O): d 175.1 (CO), 94.6 (C-1), 73.2, 71.2,
69.9 (C-3, C-4, C-5), 60.7 (C-6), 53.9 (C-2), 46.9 (CH2–
CH3), 22.3 (CH3), and 8.5 (CH2–CH3); 31P NMR
(101 MHz, D2O): d ꢀ9.20 and ꢀ12.80; ESIMS (negative
mode): m/z 380.0 [Mꢀ2NHEt3+2H]ꢀ. HRESIMS (neg-
ative mode) m/z: calcd for C8H16O12N1P2, 380.0142;
found, 380.0160.
3.7. 4-Phenylbutyl-(2-acetamido-2-deoxy-a-D-glucopyr-
anosyl)-diphosphate, bis(triethylammonium) salt (4)
4-Phenylbutylphosphate 13 as its disodium salt
(1 mmol) was dissolved in a small amount of water
and passed through a column of Bio-Rad AG 50W-X8
resin (H+ form). 4-Phenylbutylphosphoric acid was
eluted with 1:1 EtOH–water and the soln was freeze-
dried. To the dry acid, carbonyldiimidazole (CDI)
(0.324 g, 2 mmol), Et3N (0.13 mL, 1 mmol), and DMF
(5 mL) were added and the mixture was vigorously stir-
red at rt for 3 h. To neutralize excess of CDI, MeOH
(30 lL) was added and the reaction mixture was stirred
additionally for 15 min. Solvents were evaporated
affording crude imidazolide 15; 31P NMR (101 MHz,
(CD3)2SO): d ꢀ4.71. A soln of activated phosphate 15
(0.3 mmol) and 2-acetamido-2-deoxy-a-D-glucopyrano-
syl phosphate mono-triethylammonium salt 12
(180 mg, 0.45 mmol) in dry DMF (5 mL) was stirred
at rt under argon atmosphere for 3 d. Then after evap-
oration to dryness the residue dissolved in 1:1 water–
MeOH was purified by anionic exchange chꢀromato-
graphy on a DEAE-Sephadex A-25 (HCO3 form)
column. Elution with a gradient of 0 to M triethylam-
monium hydrogenocarbonate buffer (pH 7.8) afforded
4 as its triethylammonium salt, still contaminated by
salts. After exchanging triethylammonium for sodium
salt by passing through Dowex-50 (Na+ form) resin,
the product dissolved in water was applied onto a Biogel
P-2 column (<45 lm, 2 · 63 cm); elution with 6 mM
triethylammonium hydrogenocarbonate buffer (pH 7.8)
afforded pure compound 4 as a white powder (63 mg,
3.5. 4-Phenylbutyl-dibenzylphosphate (14)
Diisopropyl azodicarboxylate (1.2 mL, 6 mmol) was
added dropwise to a soln of 4-phenyl-1-butanol (0.6 g,
4 mmol), dibenzylphosphate (1.67 g, 6 mmol), and tri-
phenylphosphine (1.57 g, 6 mmol) in THF (4 mL). The
mixture was stirred at rt for 4 h. Then solvent was evap-
orated, the residue was taken up in CH2Cl2, washed with
satd aq NaHCO3, the organic phase was evaporated to
dryness and Et2O was added to the residue. After 1 h
at 0 ꢁC the precipitate was filtered, solvent was evapo-
rated, and the residue was purified by flash chromatog-
raphy (5:1 petroleum ether–ethyl acetate) affording pure
compound 14 as a colorless syrup (0.98 g, 60%); 1H
NMR (250 MHz, CDCl3): d 7.45–7.05 (m, 15H, Ph),
5.03 (d, 2H, JH,P 7.5 Hz, CH2), 5.02 (d, 2H, JH,P
7.5 Hz, CH2), 4.00 (m, 2H, H-2, CH2), 2.57 (m, 2H,
CH2), and 1.64 (m, 4H, 2CH2); 13C NMR (62.9 MHz,
CDCl3): d 141.6, 135.8, 135.6, 128.3, 128.2, 128.1,
127.7, 125.6 (Ar-C), 69.4 (CH2Ph), 68.9 (d, JC,P
6.9 Hz, CH2Ph), 67.4 (d, JC,P 6.9 Hz, CH2Ph), 34.9
(CH2), 29.4 (d, JC,P 9.2 Hz, CH2), and 26.9 (CH2); 31P
NMR (101 MHz, CDCl3): d ꢀ0.26; ESIMS (positive
mode): m/z 433.1 [M+Na]+. HRESIMS (positive mode)
m/z: calcd for C24H27O4Na1P1, 433.1539; found,
433.1543.
29
1
29%); ½aꢁD +32 (c 0.55, water); H NMR (250 MHz,
D2O): d 7.40–7.20 (m, 5H, Ph), 5.43 (dd, 1H, J1,2
3 Hz, J1,P 7 Hz, H-1), 4.00–3.71 (m, 5H, H-2, H-4, H-
5, H-6, H-60), 3.56 (q, 2H, CH2), 3.50 (t, 1H,

A. Khaled et al. / Carbohydrate Research 343 (2008) 167–178
175
J2,3 = J3,4 9.5 Hz, H-3), 3.16 (q, 12H, CH2–CH3), 2.65
(t, 2H, CH2), 2.01 (s, 3H, NHAc), 1.65 (m, 4H,
2CH2), and 1.24 (t, 18H, CH2–CH3); 13C NMR
(62.9 MHz, D2O): d 174.8 (CO), 143.2, 128.7, 126.0
(Ar-C), 94.5 (C-1), 73.1, 71.1, 69.6 (C-3, C-4, C-5),
66.6 (CH2), 60.5 (C-6), 53.8 (C-2), 46.8 (CH2–CH3),
34.7, 29.5, 27.1 (CH2), 22.3 (CH3), and 8.3 (CH2–CH3);
31P NMR (101 MHz, D2O): d ꢀ10.78 and ꢀ13.23
(JP,P 21.4 Hz); ESIMS (negative mode): m/z 512.3
[Mꢀ2NHEt3+2H]ꢀ. HRESIMS (negative mode) m/z:
calcd for C18H28O12N1P2, 512.1081; found, 512.1092.
crystallized. The crystals were filtered, washed with
29
water, and dried giving 17 (558 mg, 70%). ½aꢁD ꢀ10 (c
0.3, MeOH); mp 189 ꢁC. 1H NMR (250 MHz,
(CD3)2SO): d 11.9 (d, 1H, NH), 8.27 (s, 1H, H-8), 7.30
(m, 5H, Ar-H), 5.80 (d, 1H, J1,2 5.5 Hz, H-10), 5.49 (d,
1H, J 6.0 Hz, OH), 5.18 (d, 1H, J 5.0 Hz, OH), 5.05
(t, 1H, J 6.0 Hz, OH), 4.41 (q, 1H, H-20), 4.11 (q, 1H,
H-30), 3.89 (dt, 1H, H-40), 3.79 (s, 2H, CH2), 3.63 (m,
1H, J5 a;5 b 10.0, J4 ;5 a 2.5 Hz, H-50a), and 3.53 (m, 1H,
0
0
0
0
J4 ;5 b 3.0 Hz, H-50b); 13C NMR (100.7 MHz, MeOD–
CDCl3): d 173.3 (CO), 155.2 (C-6), 148.2 (C-2), 147.1
(C-4), 137.7 (C-8), 132.8, 128.2, 127.5, 126.2 (Ar-C),
119.3 (C-5), 87.5 (C-10), 84.6 (C-40), 74.2 (C-20), 69.5
(C-30), 60.4 (C-50), and 42.0 (CH2); ESIMS (positive
mode): m/z 402 [M+H]+, 424 [M+Na]+, 440 [M+K]+;
HRESIMS m/z: [M+Na]+ calcd for C18H19N5O6Na,
424.1228; found, 424.1232.
0
0
3.8. 50-O-[[(2-Acetamido-2-deoxy-a-D-glucopyranosyl-
oxy)carbonyl]sulfonyl]-uridine (6)
A suspension of compound 511 (36 mg, 0.05 mmol) in
saturated methanolic ammonia (4 mL) was kept at rt
for 4 h. Then solvents were evaporated and the residue
was purified on a column of Sephadex LH-20 in MeOH
to afford compound 10 as a white solid (26 mg, 90%);
3.10. 9-(20,30-O-Benzylidene-b-D-ribofuranosyl)-2-N-
(phenylacetyl)-guanine (18)
29
1
½aꢁD +49 (c 1, H2O); H NMR (400 MHz, MeOD): d
0
0
7.93 (d, 1H, J5,6 8 Hz, H-6), 5.95 (d, 1H, J1 ;2 4 Hz,
A mixture of compound 17 (500 mg, 1.24 mmol), freshly
distilled benzaldehyde (3.65 mL), and ZnCl2 (576 mg,
4.23 mmol) was stirred until complete dissolution, and
heated at 50 ꢁC for 1 h. Excess of benzaldehyde was
removed by co-evaporation with toluene; the crude
product was precipitated with ether and then purified
by flash chromatography (19:1 CH2Cl2–MeOH) to give
the benzylidene acetal 18 as a white solid (mixture of
H-10), 5.90 (d, 1H, J1,2 3.5 Hz, H-100), 5.79 (d, 1H,
H-5), 4.33 (d, 1H, J5 a;5 b 11.5 Hz, H-50a), 4.27 (d,
1H, H-50b), 4.20 (m, 3H, H-20, H-30, H-40), 4.00 (dd,
1H, J2,3 10.5 Hz, H-200), 3.76 (dd, 1H, J3,4 9.0 Hz,
H-300), 3.70 (m, 3H, H-600a, H-600b, H-500), 3.40 (t, 1H,
J4,5 9.0 Hz, H-400), and 1.98 (s, 3H, NHAc); 13C NMR
(100.7 MHz, MeOD): d 173.9 (CO), 166.2 (C-4), 156.9
(CO), 152.6 (C-2), 142.6 (C-6), 103.2 (C-5), 93.2
(C-100), 90.0 (C-10), 83.9 (C-40), 75.5, 75.4, 72.6, 71.8,
71.6, (C-20, C-30, C-300, C-400, C-500), 69.1 (C-50), 62.4
(C-600), 54.5 (C-200), and 22.8 (CH3); ESIMS (positive
mode): m/z 593.1 [M+Na]+. 615.0 [MꢀH+2Na]+.
HRESIMS (positive mode): [M+Na]+ calcd for
C18H26N4O15NaS, 593.1040; found, 593.1013. Anal.
Calcd for C18H26O15N4S, H2O: C, 36.74; H, 4.80.
Found: C, 36.44; H, 4.92.
0
0
1
two diastereomers, 528 mg, 88%). H NMR (360 MHz,
CDCl3): d 12.05 (NH), 9.84, 9.75 (NH), 7.70, 7.68 (2s,
1H, H-8), 7.53–7.28 (m, 10H, Ar-H), 6.13, 5.99 (2s,
0
0
1H, CHaPh, CHbPh), 5.90, 5.89 (2d, 1H, J1 ;2 3.5 Hz,
H-10), 5.25, 5.26 (2dd, 1H, J2 ;3 6.5 Hz, H-20), 5.14,
0
0
5.09 (2dd, 1H, J3 ;4 3.5 Hz, J3 ;4 2.5 Hz, H-30), 4.51,
0
0
0
0
4.36 (2m, 1H, H-40), 3.96, 3.90 (2dd, 1H, J5 a;5 b
0
0
12.5 Hz, J4;5 a 2.5 Hz, H-50a), 3.80, 3.77 (2dd, 1H, J4,5 b
2.5 Hz, H-50b), and 3.79 (s, 2H, CH2Ph); 13C NMR
(62.9 MHz, CDCl3): d 173.7 (CO), 155.5 (C-6), 147.7
(C-2), 139.0 (C-4), 135.7, 135.5 (C-8), 132.9, 129.3,
128.7, 128.3, 127.4, 126.6, 126.4 (Ar-C), 120.9, 120.8
(C-5), 107.3, 104.0 (CHaPh, CHbPh), 90.7, 89.9 (C-10),
86.6, 85.3 (C-40), 84.5, 83.8, 82.4, 80.6 (C-20, C-30),
62.0 (C-50), and 42.0 (CH2); ESIMS (positive mode):
m/z 490 [M+H]+, 512 [M+Na]+, 528 [M+K]+; HRE-
SIMS m/z: [M+H]+ calcd for C25H24O6N5, 490.1721;
found, 490.1718. Anal. Calcd for C25H23O6N5: C,
61.20; H, 4.93. Found: C, 61.09; H, 4.93.
0
0
3.9. 2-N-(Phenylacetyl)-9-(b-D-ribofuranosyl)-guanine
(17)
To a suspension of guanosine 16 (566 mg, 2 mmol) in
anhyd pyridine (10 mL), was added trimethylsilyl
chloride (1.9 mL, 15 mmol) and the mixture was stirred
at rt for 8 h. Then a soln of phenylacetyl chloride
(0.4 mL, 3 mmol) and 1-hydroxybenzotriazole (432 mg,
3.2 mmol) in CH3CN (1 mL) and pyridine (1 mL) was
added dropwise at 0 ꢁC and the mixture was stirred at
rt overnight. Then the reaction was cooled to 0 ꢁC and
stopped by the addition of water (2.5 mL). After
5 min, 25% aqueous ammonium hydroxide (6 mL) was
added. The aminolysis of the silyl ethers was completed
in 30 min. The solvents were evaporated and the residue
was dissolved in water and extracted with ethyl acetate.
The aqueous phases were concentrated until the product
3.11. 9-(20,30-O-Benzylidene-50-O-[(200,300,400-tri-O-benzyl-
a,b-L-fucopyranosyl)oxycarbonylaminosulfonyl]-b-D-
ribofuranosyl)-2-N-(phenylacetyl)-guanine (21), and 2,3,
4-tri-O-benzyl-1-O-carbamoyl-a,b-L-fucopyranose (20)
Benzyl-protected fucose35 19 (412 mg, 0.95 mmol) was
˚
dissolved in dry CH2Cl2 (3 mL) and stirred with 4 A

176
A. Khaled et al. / Carbohydrate Research 343 (2008) 167–178
molecular sieves (1.3 g) for 20 min under argon. Chloro-
sulfonyl isocyanate (72 ll, 0.95 mmol) was then added to
this soln cooled at ꢀ35 ꢁC and the mixture was stirred at
ꢀ35 ꢁC for 2 h. Temperature was allowed to rise to
ꢀ10 ꢁC and the mixture was further stirred until all
fucose derivative reacted to give a more polar intermedi-
ate. A soln of guanosine derivative 18 (465 mg, 0.95
mmol) in CH2Cl2 (4 mL) containing pyridine (0.2 mL,
2.5 mmol) was added and the mixture kept under argon
was stirred for 20 h at room temperature. TLC (10:1
AcOEt–MeOH) monitoring of the reaction revealed a
major spot (Rf 0.25). After filtration and evaporation
of solvents, purification of the crude product by flash
chromatography on a silica gel column (10:1 AcOEt–
MeOH), afforded first the carbamoyl fucose derivative
20 and then the expected compound still contaminated
by byproducts. This mixture was purified by gel perme-
ation on Sephadex LH-20 (7:3 CH2Cl2–MeOH) afford-
ing two fractions of compound 21 as a white solid both
as an R,S diastereomer mixture: the first one correspond-
ing to the coupling product exclusively as a fucose
(210 mg, 21.5%), the second one as 4:1 a/b anomeric
mixture (180 mg, 18.5%). The overall yield was 40% with
H-1a), 5.48 (d, 0.2H, J1,2 8 Hz, H-1b), 5.0–4.58 (m, 6H,
3CH2Ph), 3.98 (q, 0.8H, H-5a) (dd, 0.8H, J2,3 10.0, J3,4
2.5 Hz, H-3a), 3.86–3.70 (m, 1H, H-5b, H-2a), 3.71 (d,
0.8H, H-4a), 3.67–3.56 (m, 0.6H, H-4b, H-3b, H-2b),
1.19 (d, 0.6H, J5,6 6.5 Hz, H-6b), and 1.12 (d, 2.4H,
J5,6 6.5 Hz, H-6a); 13C NMR (50.3 MHz, CDCl3): d
153.5 (CO), 138.7, 138.1, 137.5, 128.1, 127.6, 127.1
(Ar-C), 95.8 (C-1b), 92.8 (C-1a), 78.8, 75.1, 74.0, 73.8,
73.1, 72.7, 72.2, 71.6, 69.2 (C-2, C-3, C-4, C-5, CH2Ph),
and 16.4 (C-6); ESIMS (positive mode): m/z
500.0 [M+Na]+; 977 [2M+Na]+. HRESIMS m/z:
[M+Na]+ calcd for C28H31O6N1Na, 500.2044; found,
500.2052.
3.12. 9-(50-O-[(a,b-L-Fucopyranosyl)oxycarbonyl-
aminosulfonyl]-b-D-ribofuranosyl)-2-N-(phenylacetyl)-
guanine (22)
To a soln of compound 21 (27 mg, 26 lmol, 4:1 a/b) in
5:3:2 THF–MeOH–water (2.5 mL), 10% Pd on charcoal
(27 mg) was added and the mixture was stirred under
hydrogen pressure (40 psi) for 4 d. Catalyst was filtered,
rinsed with MeOH and water; filtrate and washing were
concentrated under diminished pressure and then freeze-
dried. The residue taken up in water was purified on
Sep-Pak C18. Elution with water afforded methyl L-fuco-
pyranoside 23, predominantly b (2 mg, 45%); then elu-
tion with 4:1 and 7:3 water–MeOH afforded pure
compound 22 as a white powder (6 mg, 35%).1H
NMR (400 MHz, D2O,): d 8.20 (s, 1H, H-8), 7.35 (m,
1
a/b ratio of 91:9; H NMR of 21a (250 MHz, MeOD–
CDCl3) (2 diastereomers 1:1 ratio) d 7.99 (s, 0.5H, H-
8a), 7.91 (s, 0.5H, H-8b), 7.50–7.00 (25H, 5Ar-H), 6.12
(d, 0.5H, J1 ;2 3.5 Hz, H-100a), 6.09 (s, 0.5H, CHbPh),
00 00
6.07 (d, 0.5H, J1 ;2 3.5 Hz, H-100b), 6.06 (d, 0.5H, J1 ;2
00 00
0
0
2.5 Hz, H-10a), 6.03 (d, 0.5H, J1 2 1.5 Hz, H-10b), 5.85
0
0
(s, 0.5H, CHaPh), 5.30 (dd, 0.5H, J2 ;3 6.5 Hz, H-20b),
0
0
5.19 (dd, 0.5H, H-20a), 5.07 (dd, 0.5H, J2 ;3 2.5 Hz, H-
30a), 4.99 (dd, 0.5H, H-30b), 4.89 (d, 0.5H, J 11.5 Hz,
CH2Ph), 4.87 (d, 0.5H, J 11.5 Hz, CH2Ph), 4.80–4.45
(m, 6H, H-40, CH2Ph), 4.30–3.50 (m, 8H, H-200, H-300,
5H, Ar-H), 6.12 (d, 1H, J1 ;2 5.5 Hz, H-10), 5.98 (d,
0
0
0
0
0.8H, J1a00;2 3.0 Hz, H-100a), 5.30 (d, 0.2H, J1b ;2
00
00 00
8.0 Hz, H-100b), 4.78 (t, 1H, superimposed with HOD,
H-20), 4.57 (dd, 1H, J2 ;3 2.5 Hz, H-30), 4.40 (m, 3H,
0
0
H-400, H-500, 2H-50, CH2), 1.06 (d, 1.5H, J5 ;6 6 Hz, H-
H-40, H-50a, H-50b), 4.02 (q, 1H, J5 ,6 6.0 Hz, H-500a),
00 00
00 00
1
600a), and 1.03 (d, 1.5H, J5 ;6 6 Hz, H-600b); H NMR
3.88 (s, 2H, CH2), 3.84 (m, 1.6H, H-200a, H-300a), 3.79
00 00
(m, 0.2H, H-5b), 3.74 (d, 0.8H, J3 ;4 2.5 Hz, H-400a),
00 00
of 21b in the 4:1 a/b mixture (250 MHz, MeOD–CDCl3)
d 7.98 (s, 0.1H, H-8ab), 7.95 (s, 0.1H, H-8bb), 6.09 (s,
0.1H, CHbPh), 5.91 (s, 0.1H, CHaPh), 5.45 (d, 0.1H,
3.72–3.52 (m, 0.6H, H-200b, H-300b, H-400b), 1.15 (d,
0.6H, H-600b), and 1.07 (d, 2.4H, H-600a); 13C NMR
(100.7 MHz, D2O): d 175.0 (CO), 158.7, (CO, C-6),
151.8 (C-2), 149.5 (C-4), 139.6 (C-8), 133.5, 129.5,
128.9, 127.5 (Ar-C) 120.3 (C-5), 95.6 (C-100b), 93.2 (C-
100a), 87.9 (C-10), 82.4 (C-40), 73.4 (C-20), 72.6, 71.5 (C-
400), 70.2 (C-30), 69.4 (C-300), 68.4, 68.3 (C-50, C-500),
67.0 (C-200), 42.8 (CH2), and 15.3 (C-600); ESIMS (posi-
tive mode): m/z 715 [MꢀH+2Na]+; MALDI-TOFMS
m/z: (MꢀH+2Na)+ calcd for C25H29N6O14Na2S,
715.1258; found, 715.1284.
J1 ;2 8 Hz, H-100b), 5.43 (d, 0.1H, J1 ;2 8 Hz, H-100b);
13C NMR (62.9 MHz, MeOD–CDCl3) (4:1 21a/21b ra-
tio) d 174.0 (CO), 158.5, 158.3 (CO), 155.5 (C-6), 147.3,
147.1 (C-2), 139.0 (C-4), 137.8, 137.6, 137.4 (Ar-C),
135.4, 135.1 (C-8), 132.9, 128.8, 127.5, 126.5, 125.9
(Ar-C), 120.6, 120.4 (C-5), 106.7 (CHa-Ph) 103.5 (CHb-
Ph), 95.3 (C-100b), 91.0 (C-100a), 90.4 (C-10), 83.7, 83.6,
82.9, 82.3, 81.6, 80.9 (C-20, C-30, C-40), 78.7, 76.8, 75.9,
74.3, 74.0, 72.1, 71.9, 70.5 (C-200, C-300, C-400, CH2Ph),
67.8, 66.3 (C-50, C-500), 42.3, 42.2 (CH2), and 15.6 (C-
600); ESIMS (positive mode): m/z 1029.5 [M+H]+,
1051.5 [M+Na]+, 1067.5 [M+K]+. HRMS (MALDI-
TOF) m/z: [M+Na]+ calcd for C53H52O14N6S1Na,
1051.3154; found, 1051.3117.
00 00
00 00
1
Methyl b-L-fucopyranoside 23; H NMR (250 MHz,
D2O): d 4.30 (d, 1H, J1,2 7.8 Hz, H-1), 3.81 (q, 1H,
J5,6 6 Hz, H-5), 3.74 (d, 1H, J4,3 3.0 Hz, H-4), 3.64
(dd, 1H, J2,3 10 Hz, H-3), 3.55 (s, 3H, OCH3), 3.47
(dd, 1H, H-2), and 1.27 (d, 3H, H-6); 13C NMR
(90.5 MHz, D2O): d 103.7 (C-1), 72.9 (C-3), 71.3, 70.9,
70.4 (C-2, C-4, C-5), 51.1 (CH3), and 15.3 (C-6); ESIMS
(positive mode): m/z 201.1 [M+Na]+.
Compound 20 first eluted from silica gel as a white so-
1
lid (77 mg, 17%); H NMR (250 MHz, CDCl3–MeOD):
d 7.45–7.24 (m, 15H, Ar-H), 6.12 (d, 0.8H, J1,2 3.5 Hz,

A. Khaled et al. / Carbohydrate Research 343 (2008) 167–178
177
3.13. 9-(50-O-[(a,b-L-Fucopyranosyl)oxycarbonyl-
aminosulfonyl]-b-D-ribofuranosyl)-guanine (7)
GlcNAc, 0.5 mU LgtA and compound 1 or 2 or 3 or 4
or 6 (at concentrations of 1, 2, 5, 10, and 20 mM) in
50 lL total volume. Assays were incubated at 37 ꢁC
for 11 min. Incubation was stopped by immersion in a
boiling water bath for 2 min; the soln was diluted with
water (100 lL) and filtered. The samples were analyzed
by RP-HPLC using fluorescence detection. The percent-
age of conversion was quantified from the fluorescence
intensity of the peaks corresponding, respectively, to
acceptor substrate and product. Assays were duplicated
and a control reaction without the inhibitor was run
simultaneously. In each assay, the amount of the trisac-
charide formed was less than 25% of the total amount of
UDP-GlcNAc.
To a soln of compound 22 (6 mg, 9 lmol) in 40 mM
phosphate buffer pH 7.8 (2.2 mL) Eupergit-immobi-
lized penicillin amidase (75 mg) was added and the
mixture was gently shaken at 25 ꢁC overnight. Enzyme
was filtered, rinsed with water, and the aqueous soln
was freeze-dried. The residue taken up in water was
purified on Sep-Pak C18. Elution with 9:1 water–
MeOH afforded pure compound 7 as a white foam
1
(4.5 mg, 91%). H NMR (400 MHz, D2O): d 8.05 (s,
1H, H-8), 5.95 (d, 1H, J1 ;2 6.0 Hz, H-10), 5.86 (d,
0
0
0.8H, J1a00;2 3.5 Hz, H-100a), 5.28 (d, 0.2H, J1b ;2
00
00 00
8.0 Hz, H-100b), 4.75 (t, 1H, superimposed with HOD,
H-20), 4.48 (dd, 1H, J2 ;3 3.0 Hz, H-30), 4.38 (m, 3H,
0
0
3.16. Assay of compound 7 as a fucosyl donor toward
FucT-III
H-40, 2H-50), 3.99 (q, 0.8H, J5 ;6 6.0 Hz, H-500a), 3.88
00 00
(dd, 0.8H, J2 ;3 10.0, J3 ,4 3.0 Hz, H-300a), 3.83 (dd,
00 00
00 00
0.8H, H-200a), 3.80 (q, 0.2H, H-500b), 3.74 (d, 0.8H,
Assay conditions with soluble enzyme: 400 lM b-D-
Galp-(1!3)-b-D-GlcNAcp-O-(CH2)6-NH-dansyl, 4 mM
compound 7 (4:1 a/b ratio) were incubated at 37 ꢁC for
30 h in a total volume of 50 lL with FucT-III (20 lL,
1 mU) either in 25 mM sodium cacodylate buffer pH
6.0, or in 25 mM sodium cacodylate buffer pH 6.0 con-
taining 10 mM MnCl2, or 25 mM sodium cacodylate
buffer pH 6.0 containing 10 mM Yb(OTf)3. Assay condi-
tions with immobilized enzyme3: 200 lM b-D-Galp-
(1!3)-b-D-GlcNAcp-O-(CH2)6-NH-dansyl, 35 mM com-
pound 7 (6:1 a/b ratio) were stirred at 30 ꢁC for 30 h in
a total volume of 50 lL with FucT-III immobilized on
Ni2+–agarose (1 mU) in 25 mM sodium cacodylate
buffer pH 6.0 containing 20 mM MnCl2 or 20 mM
Yb(OTf)3. Reaction mixtures were analyzed by TLC
(3:1:1 AcOH–AcOEt–water). A blank test was run in
parallel, for which compound 7 was replaced by
100 lM GDP-Fuc. In these conditions 50% substrate
was converted into the product in 10 min.
J3 ;4 2.5 Hz, H-400a), 3.69 (d, 0.2H, J3 ;4 3.0 Hz, H-
400b), 3.64 (dd, 0.2H, H-300b), 3.60 (dd, 0.2H, H-200b),
1.14 (d, 0.6H, H-600b), and 1.09 (d, 2.4H, H-600a); 13C
NMR (100.7 MHz, D2O): d 160.4 (CO), 158.6 (C-6),
155.0 (C-2), 151.4 (C-4), 137.2 (C-8), 125.1 (C-5),
95.3 (C-100b), 93.2 (C-100a), 86.8 (C-10), 82.4 (C-40),
73.4 (C-20), 71.6, 71.5 (C-400), 70.3 (C-30), 69.3 (C-300),
68.4, 68.2 (C-50, C-500), 66.9 (C-200), and 15.3 (C-600);
ESIMS (positive mode): m/z 553.3 [M+H]+, 575.3
[M+Na]+, 597.2 [MꢀH+2Na]+; MALDI-TOFMS
m/z: (M+Na)+ calcd for C17H24N6O13NaS, 575.1014;
found, 575.1010.
00 00
00 00
3.14. Assay of compounds 1, 2, 3, 4, 6 as N-acetylglucos-
amine donors toward LgtA
Assay conditions: 400 lM b-D-Galp-(1!4)-b-D-Glcp-O-
(CH2)6-NH-dansyl, 60 mM compound 1 or 2 or 3 or 4
or 6 and 15 mM MnCl were incubated at 37 ꢁC for
15 h in a total volume of 200 lL with LgtA (40 lL,
2.5 mU) in 50 mM sodium cacodylate buffer pH 7.2.
Two supplementary assays were run with compound 3
at concentrations of 60 and 100 mM in the presence of
uridine used at the same concentrations. Reaction mix-
tures were analyzed by TLC (3:1:1 AcOH–AcOEt–
water). A blank test was run in parallel, for which
compounds to be tested were replaced by 500 lM
UDP-GlcNAc. In these conditions 50% substrate was
converted into the product in 5 min.
3.17. Inhibitory activity of compound 7 toward FucT-III
Enzyme assay conditions: 10 mM MnCl2, 25 mM
cacodylate buffer pH 6.0, 200 lM b-D-Galp-(1!3)-b-
D-GlcNAcp-O-(CH2)6-NH-dansyl, 50 lM GDP-Fuc,
0.05 mU soluble FucT-III and compound 7 (4:1 a/b
ratio, 1, 2, 4, 7.5, 15, and 30 mM) in 50 lL total volume.
Assays were incubated at 37 ꢁC for 9 min. Incubation
was stopped by immersion in a boiling water bath for
2 min; after filtration, samples were analyzed by RP-
HPLC and the percentage of conversion of each assay
was quantified as above. Assays were conducted in
duplicate and a control reaction without the inhibitor
was run simultaneously. In each assay, the amount of
the trisaccharide formed was less than 25% of the total
amount of GDP-Fuc.
3.15. Inhibitory activity of compounds 1, 2, 3, 4, 6 toward
LgtA
Enzyme assay conditions: 25 mM cacodylate buffer pH
7.2 containing 15 mM MnCl2, 800 lM b-D-Galp-
(1!4)-b-D-Glcp-O-(CH2)6-NH-dansyl, 225 lM UDP-

178
A. Khaled et al. / Carbohydrate Research 343 (2008) 167–178
13. Elloumi, N.; Moreau, B.; Aguiar, L.; Jaziri, N.; Sauvage,
Acknowledgments
M.; Hulen, C.; Capmau, M. L. Eur. J. Med. Chem. 1992,
27, 149–154.
14. Kappes, T.; Waldmann, H. Carbohydr. Res. 1998, 305,
341–349.
15. Prata, C.; Mora, N.; Lacombe, J.-M.; Maurizis, J.-C.;
Pucci, B. Tetrahedron Lett. 1997, 38, 8859–8862.
16. Hinklin, R. J.; Kiessling, L. L. J. Am. Chem. 2001, 123,
3379–3380.
17. Palma, A. S.; Morais, V. A.; Coelho, A. V.; Costa, J.
BioMetals 2004, 17, 35–43.
´
The present study was funded partly by the Region Ile
de France within the interregional network (MIIAT)
project ‘glycosylation mutants’. Dr Lo¨ıc Faye (CNRS
UMR 6037) is gratefully acknowledged as the coordina-
tor of this project. A.K. was the recipient of a fellowship
´
from Region Ile de France. We also thank CNRS and
the University of Paris-Sud for financial support,
Professor Andre Lubineau for helpful discussions, and
´
18. Obata, T.; Mukaiyama, T. J. Org. Chem. 1967, 32, 1063–
1065.
´
Professor David Bonnaffe for encouragement.
19. Zhao, Y.; Thorson, J. S. J. Org. Chem. 1998, 63, 7568–
7572.
20. Zatorski, A.; Goldstein, B. M.; Colby, T. D.; Jones, J. P.;
Pankiewicz, K. W. J. Med. Chem. 1995, 38, 1098–
1105.
References
21. Kadokura, M.; Wada, T.; Urashima, C.; Sekine, M.
Tetrahedron Lett. 1997, 38, 8359–8362.
22. Mitsunobu, O. Synthesis 1981, 1–28.
23. Ivannikova, T.; Bintein, F.; Malleron, A.; Juliant, S.;
1. Gijsen, H. J. M.; Qiao, L.; Fitz, W.; Wong, C.-H. Chem.
Rev. 1996, 96, 443–473.
´
2. Bintein, F.; Auge, C.; Lubineau, A. Carbohydr. Res. 2003,
´
Cerutti, M.; Harduin-Lepers, A.; Delannoy, P.; Auge, C.;
338, 1163–1173.
Lubineau, A. Carbohydr. Res. 2003, 338, 1153–1161.
24. Waldmann, H.; Heuser, A.; Reidel, A. Synlett 1994, 65–
67.
25. Dineva, M. A.; Galunsky, B.; Kasche, V.; Petkov, D. D.
Bioorg. Med. Chem. Lett. 1993, 3, 2781–2784.
26. Vincent, S.; Mioskowski, C.; Lebeau, L. Tetrahedron Lett.
1998, 39, 2321–2324.
´
3. Auge, C.; Malleron, A.; Tahrat, H.; Marc, A.; Goergen,
J.-L.; Cerutti, M.; Steelant, W. F. A.; Delannoy, P.;
Lubineau, A. J. Chem. Soc., Chem. Commun. 2000, 2017–
2018.
4. Nahalka, J.; Liu, Z.; Chen, X.; Wang, P. G. Chem. Eur. J.
2003, 9, 373–377.
5. Wang, R.; Steensma, D. H.; Takaoka, Y.; Yun, J. W.;
Kajimoto, T.; Wong, C.-H. Bioorg. Med. Chem. 1997, 5,
661–672.
27. Klein, E.; Mons, S.; Valleix, A.; Mioskowski, C.; Lebeau,
L. J. Org. Chem. 2002, 67, 146–153.
28. Rasmussen, J. K.; Hassner, A. Chem. Rev. 1976, 76, 389–
408.
6. Lairson, L. L.; Wakarchuk, W. W.; Withers, S. G. J.
Chem. Soc., Chem. Commun. 2007, 365–367.
29. Fiandor, J.; Garcia-Lopez, M.-T.; De Las Heras, F. G.;
´
7. Khaled, A.; Ivannikova, T.; Auge, C. Carbohydr. Res.
´
Mendez-Castrillon, P. P.; Felix, A. S.; Alarcon, B.;
2004, 339, 2641–2649.
Carrasco, L. Eur. J. Med. Chem. 1987, 22, 59–65.
30. Johnson, D. C., II; Widlanski, T. S. Org. Lett. 2004, 6,
4643–4646.
8. Wakarchuk, W. W.; Martin, A.; Jennings, M. P.; Moxon,
E. R.; Richards, J. C. J. Biol. Chem. 1996, 27, 19166–
19173.
´
31. Auge, C. Unpublished results.
9. Priem, B.; Gilbert, M.; Wakarchuk, W. W.; Heyraud, A.;
Samain, E. Glycobiology 2002, 12, 235–240.
´
32. Carchon, G.; Chretien, F.; Delannoy, P.; Verbert, A.;
Chapleur, Y. Tetrahedron Lett. 2001, 42, 8821–8824.
33. Sun, H.-Y.; Lin, S.-W.; Ko, T.-P.; Pan, J.-F.; Liu, C.-L.;
Lin, C.-N.; Wang, A. H.-J.; Lin, C.-H. J. Biol. Chem.
2007, 282, 9973–9982.
34. Heidlas, J. E.; Lees, W. J.; Pale, P.; Whitesides, G. M. J.
Org. Chem. 1992, 57, 146–151.
10. Breton, C.; Heissigerova, H.; Jeanneau, C.; Moravcova, J.;
Imberty, A. Biochem. Soc. Symp. 2002, 69, 23–32.
11. Camarasa, M.-J.; Fernandez-Resa, P.; Garcia-Lopez,
M.-T.; De Las Heras, F. G.; Mendez-Castrillon, P. P.;
Alarcon, B.; Carrasco, L. J. Med. Chem. 1985, 28, 40–46.
12. Radominska, A.; Paul, P.; Treat, S.; Towbin, H.; Pratt, C.;
Little, J.; Magdalou, J.; Lester, R.; Drake, R. Biochim.
Biophys. Acta 1994, 1205, 336–345.
35. Dejter-Juszynski, M.; Flowers, H. M. Carbohydr. Res.
1971, 18, 219–226.