The first C-glycosidic analogue of a novel galactosyltransferase inhibitor

Article abstract of DOI:10.1039/c0ob00630k

Structural analogues and mimics of the natural sugar-nucleotide UDP-galactose (UDP-Gal) are sought after as chemical tools for glycobiology and drug discovery. We have recently developed a novel class of galactosyltransferase (GalT) inhibitors derived from UDP-Gal, bearing an additional substituent at the 5-position of the uracil base. Herein we report the first C-glycosidic derivative of this new class of GalT inhibitors. We describe a practical convergent synthesis of the new UDP-C-Gal derivative, including a systematic study into the use of radical chemistry for the preparation of galactosyl ethylphosphonate, a key synthetic intermediate. The new inhibitor showed activity against a bacterial UDP-Gal 4′-epimerase at micromolar concentrations. This is the first example of a base-modified UDP-sugar as an inhibitor of a UDP-sugar-dependent enzyme which is not a glycosyltransferase, and these results may therefore have implications for the design of inhibitors of these enzymes in the future.

Full text of DOI:10.1039/c0ob00630k

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 1855
www.rsc.org/obc
PAPER
The first C-glycosidic analogue of a novel galactosyltransferase inhibitor†
Karine Descroix and Gerd K. Wagner*‡
Received 26th August 2010, Accepted 29th November 2010
DOI: 10.1039/c0ob00630k
Structural analogues and mimics of the natural sugar-nucleotide UDP-galactose (UDP-Gal) are sought
after as chemical tools for glycobiology and drug discovery. We have recently developed a novel class of
galactosyltransferase (GalT) inhibitors derived from UDP-Gal, bearing an additional substituent at the
5-position of the uracil base. Herein we report the first C-glycosidic derivative of this new class of GalT
inhibitors. We describe a practical convergent synthesis of the new UDP-C-Gal derivative, including a
systematic study into the use of radical chemistry for the preparation of galactosyl ethylphosphonate, a
key synthetic intermediate. The new inhibitor showed activity against a bacterial UDP-Gal
4¢-epimerase at micromolar concentrations. This is the first example of a base-modified UDP-sugar as
an inhibitor of a UDP-sugar-dependent enzyme which is not a glycosyltransferase, and these results
may therefore have implications for the design of inhibitors of these enzymes in the future.
Encouraged by the exciting biological activity of 1, and with a
view towards potential cellular applications, we became interested
Introduction
UDP-Galactose (UDP-Gal, Fig. 1) is the natural donor substrate
in the development of a hydrolytically stable analogue of 1. In
for a range of carbohydrate-active and glycoprocessing enzymes
particular, we were keen to replace the scissile galactosyl phosphate
of biological and biomedical interest, including the galacto-
bond in 1 with a chemically stable surrogate. These considerations
syltransferases (GalTs),1 UDP-Gal 4¢-epimerases,2 and UDP-
informed the design of the new C-glycoside 5-FT UDP-C-Gal 2
galactopyranose mutases.3 These enzymes are critically involved
(Fig. 1), which combines structural features of 5-FT UDP-Gal 1
in the biosynthesis of glycan components of the cell wall in plants,
and UDP-C-Gal 3, a previously reported C-glycosidic UDP-Gal
bacteria and mycobacteria.4 Derivatives and mimics of UDP-Gal
mimetic.10 The replacement of the b-phosphate group in sugar-
are therefore sought after as chemical tools for glycobiology5
nucleotides with a phosphonate group is a common strategy for
and as potential enzyme inhibitors for drug discovery.6 In the
past, the development of UDP-Gal-based enzyme inhibitors has
focused predominantly on the modification of the sugar moiety7
or the pyrophosphate linkage.8 We have recently developed a novel
class of UDP-Gal derivatives modified at position 5 of the uracil
base,9 as exemplified by compound 1 (Fig. 1). 5-FT UDP-Gal 1
inhibits Gal transfer by a range of different GalTs with Ki values in
the low to sub-micromolar range.9 Evidence from structural and
enzymological studies with a representative GalT shows that 1
binds at the UDP-Gal binding site, locking the target enzyme in a
catalytically inactive conformation.9 For a GalT inhibitor, this is a
unique and unprecedented mode of action. These results also raise
the intriguing possibility that as a close structural analogue of the
natural sugar-nucleotide UDP-Gal, 1 and related molecules may
also have inhibitory activity towards other UDP-Gal-dependent
enzymes.
the generation of hydrolytically stable analogues.11 Interestingly,
in the case of UDP-Gal the best surrogate for the glycosyl
phosphate linkage, with regard to bioactivity, is a non-isosteric C2
fragment. UDP-C-Gal 3 inhibited a representative GalT with an
IC50 value of 40 mM, similar to the Km value of the natural donor
UDP-Gal,10 while the corresponding isosteric C-1 phosphonate
analogue of UDP-Gal11g showed no biological activity against a
pig a-1,3-GalT.11f In view of these previous findings we chose a
C2-linker between the galactose and phosphonate moieties in our
design of 5-FT UDP-C-Gal 2, the first C-glycosidic derivative
of 1.
Herein, we report the chemical synthesis of the new 5-FT UDP-
C-Gal 2 and its biological evaluation against a UDP-Gal 4¢-
epimerase. The key step in our synthesis is the installation of the
5-formylthienyl substituent in position 5 under aqueous Suzuki–
Miyaura conditions. This flexible synthetic approach allows the
introduction of the 5-substituent in the final step of the synthesis.
We also describe our findings from a systematic study into the
use of radical chemistry10,12 for the preparation of galactosyl
phosphonate 5a (Scheme 1), a central intermediate in our synthesis
of 2. Finally, we present enzymological data which show that
the new 5-FT UDP-C-Gal 2 has inhibitory activity against a
UDP-Gal 4¢-epimerase (GalE) from Streptococcus thermophilus.
GalE catalyses the interconversion of UDP-Gal and UDP-Glc
School of Pharmacy, University of East Anglia, Norwich, UK NR4 7TJ.
E-mail: gerd.wagner@kcl.ac.uk; Fax: +44 (0)20 7848 4747; Tel: +44 (0)20
7848 4045
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c0ob00630k
‡ Current address: King’s College London, School of Biomedical Sciences,
Institute of Pharmaceutical Science, Franklin-Wilkins Building, 150
Stamford Street, London, UK SE1 9NH.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1855–1863 | 1855
©

View Article Online
Fig. 1 Chemical structures of UDP-Gal, 5-FT UDP-Gal 1 (ref. 9), UDP-C-Gal 3 (ref. 10), and the synthetic target 5-FT UDP-C-Gal 2.
(Fig. 2a) and plays an important role in the biosynthesis of
bacterial cell-surface carbohydrates such as the O-antigen in gram-
negative species.13 In thermophilic bacteria, GalE is involved in
the biosynthesis of exopolysaccharides, and overexpression of
GalE in T. thermophilus resulted in an increased capacity of
biofilm production.14 In addition, the corresponding enzyme in
trypanosoma species is a validated anti-parasitic drug target.15
GalE inhibitors are therefore of interest as potential anti-bacterial
and anti-parasitic agents. To the best of our knowledge, this is the
first example of a base-modified UDP-sugar as an inhibitor of a
UDP-sugar-dependent enzyme which is not a glycosyltransferase,
and these results may therefore have implications for the design of
inhibitors of these enzymes in the future.
single-step procedure affords predominantly the a-anomer, as
required in our case.
Starting from galactose, glycosyl bromide 4 was obtained in two
synthetic steps in 95% overall yield using standard carbohydrate
chemistry (Scheme 1). Previously, the best results for the radical
addition of 4 to diethyl vinylphosphonate had been achieved by
the use of AIBN (azobisisobutyronitrile) as the radical initiator
in combination with photochemical activation.12,17 In order to
have access to sufficient quantities of 5a, and ultimately 7, we
decided to revisit this radical reaction, and to systematically
investigate the experimental conditions for this critical step in
our synthesis (Table 1). We were particularly keen to avoid
photochemical activation and to replace AIBN, which can be
difficult to source commercially in the UK, with an alternative
radical initiator. We therefore explored different combinations of
radical initiators and solvents, parameters that had not previously
been investigated, while adopting other experimental conditions
from Praly and co-workers.12 In particular, to promote efficient C–
C bond formation we carried out all radical reactions of glycosyl
bromide 4 with an excess of diethyl vinylphosphonate, and in the
presence of catalytic amounts of n-Bu3SnCl and 1.5 equivalents of
sodium cyanoborohydride as the reducing agent. These conditions
help minimize the competing hydrogen abstraction from the
intermediate glycosyl radical, a potential side reaction leading to
1-deoxysugar 5b.12
Results and discussion
Chemistry
Our synthetic strategy for the target 5-FT UDP-C-Gal 2 was
centred around the formation of the phosphate–phosphonate
bond. To implement this strategy, galactosyl-1-phosphonate 7 was
required as a key synthetic building block. For the preparation of
7, we adopted the synthetic approach of Vidal and Praly.10 These
authors successfully reacted peracetylated galactopyranos-1-yl
bromide 4 with diethyl vinylphosphonate under radical chemistry
conditions to obtain the fully protected galactosyl-1-phosphonate
5a, a direct synthetic precursor of 7 (Scheme 1). Among the
various methods that have been described in the literature for
the preparation of glycosylphosphonates,16 advantageously, this
We found that, interestingly, the radical reaction is relatively
tolerant to changes in radical initiator and solvent (Table 1). With
both AIBN and ABCN (1,1¢-azobiscyclohexanecarbonitrile),
which is known to be a more efficient radical initiator than
1856 | Org. Biomol. Chem., 2011, 9, 1855–1863
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Fig. 2 (a) GalE-catalysed interconversion of UDP-Gal and UDP-Glc; (b) Dixon plot for GalE-inhibition by 5-FT UDP-C-Gal 2.
Scheme 1 Synthesis of C-glycoside 7. Reagents and conditions: (i) Ac2O, pyridine, rt, overnight; (ii) HBr (33% in AcOH), 0 ◦C, 2 h; (iii) diethyl
vinylphosphonate, n-Bu3SnCl, NaCNBH3, reflux, radical initiator, solvent, 1 to 4 days (see Table 1); (iv) Me3SiBr, pyridine, CH3CN, 0 ◦C, 3 h; (v)
H2O/MeOH/Et3N, rt, 16 h.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1855–1863 | 1857
©

View Article Online
Table 1 Radical chemistry conditions for the preparation of C-glycoside
5a (see Scheme 1 for other conditions)
the intermediate glycosyl radical. Depending on the experimental
conditions, 5b was isolated in 10–40% yield. 5b was formed to a
substantial degree particularly in those cases where the reactants
had not been specially dried prior to the reaction. This suggests
that meticulously dry reagents and solvents are a prerequisite to
favour radical addition over radical reduction under these condi-
tions. On the other hand, we never observed the corresponding 2-
deoxysugar resulting from the rearrangement of the peracetylated
glycopyranosyl radical as described by Praly.12 Following the
successful preparation of C-glycoside 5a, we sequentially removed
the various protecting groups (Scheme 1). First, the phosphonate
ester groups were readily cleaved with bromotrimethylsilane in a
mixture of acetonitrile/pyridine, followed by in situ hydrolysis of
the resulting bis(trimethylsilyl)-phosphonosugar intermediate.19
Finally, deacetylation of 6 and purification of the reaction
product by C18 reverse phase chromatography afforded the fully
deprotected galactosyl phosphonate 7 in high yield.
With sufficient quantities of 7 in hand, we approached the
synthesis of target 5-FT UDP-C-Gal 2 by joining the C-glycoside
and UMP building blocks (Scheme 2). In principle, the installation
of the 5-formylthien-2-yl substituent in position 5 of the uracil
base can be carried out before or after the formation of the
phosphate/phosphonate linkage. In both cases, we planned to
activate the respective 5-substituted UMP derivative in the form
of its phosphoromorpholidate, a well established methodology in
our laboratory.9,20,21 However, the conversion of the cross-coupled
UMP derivative 8 into the corresponding phosphoromorpholidate
9 under standard Mukaiyama conditions22 proved unexpectedly
problematic (Scheme 2). Although TLC indicated the formation
of morpholidate 9, this material was unusually difficult to isolate
Entry Radical initiator
Equiv. Solvent
Yielda
76%b
1
2
3
4
5
6
7
8
9
AIBN/photoactivationb 0.7 tBuOH
AIBN
AIBN
AIBN
ABCN
ABCN
ABCN
Benzoylperoxide
Benzoylperoxide
1
diethylether/tBuOH 62%
0.7
0.7
1
0.7
0.7
0.7
0.7
diethylether/tBuOH 55%
tBuOH
75%
diethylether/tBuOH 44%
diethylether/tBuOH 71%
tBuOH
64%
diethylether/tBuOH 53%
tBuOH
54%
a Isolated yields, corrected for residual diethyl vinylphosphonate (0.1–0.5
molar equivalents by 1H-NMR). b Ref. 12.
AIBN,18 5a was obtained in fair to good isolated yields. In the
case of AIBN the best results were obtained in tBuOH (entry 4),
while for ABCN the addition of diethylether as co-solvent was
beneficial (entry 6). The composition of the solvent had almost no
influence when benzoylperoxide was used as the radical initiator
(entries 8 and 9). Importantly, our best yields are comparable
to the yield previously reported by Praly under UV-activation,12
although longer reaction times were required under these thermal
conditions. The exclusive formation of the desired a-anomer under
1
these conditions was evident from the H NMR spectrum of 5a.
The vicinal coupling constant 3JH3,H4 of 5.3 Hz is indicative of the
a configuration at the anomeric center of C-galactosides and in
agreement with the previously reported value for a-configurated
5a.12 As expected, 1-deoxysugar 5b was observed as a side product
in several experiments, resulting from the in situ reduction of
Scheme 2 Synthesis of target 5-FT UDP-C-Gal 2. Reagents and conditions: (i) morpholine, dipyridyldisulfide, Ph3P, DMSO, rt, 2 h; (ii) Gal-1-phosphonate
7; tetrazole (5 equiv.), DMF, rt, 4 days; (iii) 5-formylthien-2-ylboronic acid, Cs2CO3, TPPTS, Na2Cl4Pd, H2O, 17 h/40 ◦C or 2 h/55 ◦C.
1858 | Org. Biomol. Chem., 2011, 9, 1855–1863
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
with our standard work-up protocol (precipitation with 0.1 M
NaI in acetone). During the work-up, the reaction mixture turned
bright red immediately upon addition of the NaI solution, with
TLC indicating the formation of side products. As a result,
phosphoromorpholidate 9 was isolated in only 7% yield via this
route. In view of this disappointing yield, we decided to explore
the alternative strategy, coupling 7 to the morpholidate of 5-
iodo UMP 1021 followed by Suzuki–Miyaura installation of the
5-substituent in the final step of the synthesis. Advantageously, this
strategy also allows greater flexibility regarding the installation of
different 5-substituents.
No practical problems were encountered upon conversion of 5-
iodo UMP 10 into the corresponding phosphoromorpholidate 11,
which was isolated in quantitative yield. With this phosphoromor-
pholidate in hand, the formation of the phosphate–phosphonate
linkage was carried out under the same conditions previously
developed for the preparation of the parent 5-FT UDP-Gal 1,
using tetrazole as the catalyst.9 However, compared to galactosyl-
1-phosphate during the synthesis of 1, galactosyl-1-phosphonate
7 reacted only relatively sluggishly in this transformation. This
reduced reactivity necessitated the use of 5 equivalents of tetrazole
and longer reaction times (4 days instead of 5 h). Fortuitously,
despite the longer reaction time only relatively limited decompo-
sition was observed, although some loss of material did occur
through hydrolysis of 5-iodo UMP phosphoromorpholidate 11
to the parent nucleoside monophosphate 10. Once formed, the
product 5-iodo UDP-C-Gal 12 proved to be chemically stable
under these conditions, as was expected. Unfortunately, the
removal of the contaminant 10 and excess tetrazole required
the sequential purification of the crude product by C-18 reverse
phase and anion-exchange chromatography. Due to these repeated
chromatographic purification steps, 5-iodo UDP-C-Gal 12 was
obtained in only 28% isolated yield, in addition to 18% 5-iodo
UMP 10. The successful formation of the P–O–P linkage in 12 was
confirmed by 31P NMR spectroscopy, which showed two doublets
at d ~ -9 and ~22 ppm with a coupling constant of 2JP,P = 27 Hz
for this phosphonophosphate derivative.
This protocol provided 5-FT UDP-C-Gal 2 in isolated yields of
35% (17 h/40 ◦C) and 59% (2 h/55 ◦C), respectively.
The relatively modest cross-coupling yields can be attributed
to the formation of UDP-C-Gal 3 as a major side product, due
to hydrodehalogenation in position 5 under the cross-coupling
conditions. Such hydrodehalogenations are a frequently observed
side reaction in heterocyclic couplings.23 They are usually the
result of an inefficient transmetallation step,23 and the nature of
the leaving group, the nature of the Pd ligand and the catalyst
loading can all influence the degree of hydrodehalogenation. We
found that at higher catalyst loadings, UDP-C-Gal 3 became
the main reaction product, with complete conversion of starting
material 12 after 3 h at room temperature, yet only traces
of the cross-coupled product 2 were observed, as indicated by
TLC. After work-up and purification, the 1H NMR spectrum of
the reaction product displayed two doublets resonating at d ~7.53
and 5.67 ppm, with a coupling constant of 3JH5,H6 = 7.9 Hz. These
signals are characteristic for the two heteroaromatic protons at the
uracil base, proving unambiguously the formation of the dehalo-
genated parent compound 3 rather than the desired cross-coupled
product 2.
Enzymology
Against the bacterial a-1,4-GalT LgtC24, the new 5-FT UDP-C-
Gal 2 has shown similar activity as the parent UDP-sugar 1.25 In
the present study, 2 was evaluated as a potential inhibitor of UDP-
Gal 4¢-epimerase (GalE, E.C. 5.1.3.2), an enzyme of the Leloir
pathway of galactose metabolism.26 For these experiments, we used
a commercial GalE from Streptococcus thermophilus as a model
enzyme. First, we developed an assay protocol that allowed us to
follow the GalE-catalysed conversion of UDP-Gal into UDP-Glc
by HPLC. For this protocol, we adapted previously published ion-
pair conditions for the separation of sugar-nucleotides.27 Despite
the minimal structural difference between UDP-Gal and UDP-
Glc, our optimized HPLC conditions allowed us to clearly separate
the two UDP-sugars (retention times: UDP-Gal 7.7 min, UDP-
Glc 8.2 min). With this protocol in hand, we incubated GalE
with the natural substrate UDP-Gal at different concentrations,
monitoring the reaction progress by HPLC. From these experi-
ments, we determined a Km for UDP-Gal of 233 15 mM. Finally,
we co-incubated GalE and UDP-Gal with inhibitor 2. In the
presence of the UDP-C-Gal derivative 2, the conversion of UDP-
Gal into UDP-Glc was significantly reduced, and we determined
a Ki of 426 mM for 2, based on the formation of UDP-Glc
(Fig. 2b).
In principle, two explanations are conceivable for the inhibitory
effect of 2 on GalE-catalysed UDP-Gal epimerization: 2 might act
either as an alternative substrate of the enzyme, leading to the for-
mation of the corresponding 5-substituted UDP-C-Glc derivative,
or as an inhibitor. During the co-incubation experiments, no new
peak was observed in the HPLC chromatograms of the enzymatic
reactions. This suggests that GalE does not use 2 as a substrate
and that 2 may indeed be a true GalE inhibitor. An intriguing
question concerns the contribution of the C-glycoside motif to the
GalE inhibitory activity of 2. To answer this question, a direct
comparison of the biological activity of 1 and 2 towards GalE
would provide an important clue. Unfortunately, all attempts to
separate 5-FT UDP-Gal 1 and the corresponding 5-FT UDP-Glc
In the final step of the synthesis, the 5-formylthienyl substituent
was introduced at the uracil base using Suzuki–Miyaura chemistry
previously developed in our group for the cross-coupling of
UDP-sugars.21 With Na2Cl4Pd as the Pd source and TPPTS
(triphenylphosphine trisulfonate, sodium salt) as the water-soluble
ligand, no conversion was observed at room temperature, and only
starting material was recovered after 3 h. However, running the
◦
reaction overnight at 40 C under otherwise identical conditions
successfully produced the cross-coupled 5-FT UDP-C-Gal 2. In
the case of the corresponding diphosphate analogue, the onset of
decomposition had limited the cross-coupling time to 10 min.9
However, the greater chemical stability of phosphonate derivative
12 permitted a significantly longer reaction time, which helped
drive the cross-coupling reaction to completion at 40 ◦C. Im-
portantly, no decomposition was observed even after 17 h at
this temperature. Due to the pronounced chemical stability of
phosphonate 12, the cross-coupling reaction could also be carried
out successfully at slightly elevated temperature, which allowed a
shorter reaction time. In both cases, the cross-coupling product 2
was purified by ion-pair and/or anion-exchange chromatography
and converted into its disodium salt form by eluting the purified
material through a short column of Dowex-XW20 (Na+ form).
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1855–1863 | 1859
©

View Article Online
derivative under the conditions of our HPLC assay have so far
failed, precluding the determination of Km (GalE) for 1. A direct
comparison of 1 and 2 with regard to GalE inhibition is therefore
not possible at present, preventing any conclusion on the effect of
replacing oxygen by carbon in this particular system. Addressing
this important question will be a priority in future studies on this
new class of GalE inhibitors.
recorded at 25 ◦C on a Varian VXR 400 S spectrometer (400 MHz
1
for H, 100 MHz for 13C, 161.9 MHz for 31P). Chemical shifts
1
(d) are reported in ppm (parts per million). Assignments of H
signals were made by first-order analysis of 1D spectra, as well
as analysis of 2D 1H–1H correlation maps (COSY). The 13C
NMR assignments are supported by 2D 13C–1H correlations maps
(HSQC). Preparative chromatography was performed on Silica
Gel 60 (particle size 0.063–0.200 mm). Ion-pair and ion-exchange
chromatography was performed on a Biologic LP chromatography
system equipped with a peristaltic pump and a 254 nm UV Optics
Module under the following conditions:
Conclusion
Structural analogues and mimics of the natural sugar-nucleotide
UDP-Gal are sought after as chemical tools for glycobiology and
drug discovery. Starting from 5-FT UDP-Gal 1, a novel type of
GalT inhibitor,9 we have now developed the new C-glycoside 5-FT
UDP-C-Gal 2. We have devised a practical convergent synthesis
for 2 which provides access to this new inhibitor, in 6 synthetic steps
from D-galactose, in sufficient quantities for biological studies.
During the preparation of 2 we successfully circumvented several
synthetic pitfalls. Thus, we identified suitable conditions that allow,
for the first time, the preparation of C-galactoside 5a, a key
synthetic building block, under thermal conditions in practically
useful yields.
In initial enzymological experiments, 2 showed inhibitory
activity against the UDP-Gal 4¢-epimerase GalE. This is the first
example that a base-modified UDP-sugar derivative can act as an
inhibitor of a carbohydrate-active enzyme that is not a glycosyl-
transferase. While the GalE inhibitory activity of 2 is relatively
modest, our results suggest that the additional substituent at the
uracil base is tolerated by this glycoprocessing enzyme. Modifica-
tion of the donor at the nucleobase has recently been introduced as
a novel strategy for inhibition of glycosyltransferases.9 Our results
suggest that base-modification may also represent a promising
strategy for the development of inhibitors for carbohydrate-active
enzymes other than glycosyltransferases. While results from our
initial enzymological experiments indicate that 2 is not used as
a substrate by GalE, but rather acts as an inhibitor, further
biological studies will be required to clarify the precise mode
of action of this new sugar-nucleotide analogue. Importantly,
the chemical stability of C-glycoside 2 is significantly enhanced
compared to the parent UDP-sugar 1. This pronounced chemical
stability will greatly facilitate such biological studies and the
application of 2 as a novel chemical tool in enzymology and
crystallography, as well as in cellular studies.
Purification method 1. Ion-pair chromatography was performed
using Lichroprep RP-18 resin and a standard gradient (unless
stated otherwise) of 0–15% MeOH against 0.05 M TEAB (triethy-
lammonium bicarbonate) over a total volume of 480 mL (flow
rate: 5 mL min-1). Product-containing fractions were combined
and reduced to dryness. The residue was co-evaporated repeatedly
with methanol to remove residual TEAB.
Purification method 2. Anion exchange chromatography was
performed on BioscaleTM Mini Macro-Prep High Q cartridges
and a gradient of 0–100% 1 M TEAB (pH 7.3) against H2O
over a total volume of 480 mL (flow rate: 3 mL min-1. Product-
containing fractions were combined and reduced to dryness. The
residue was co-evaporated repeatedly with methanol to remove
residual TEAB.
Diethyl 2-(2,3,4,6-tetra-O-acetyl-a-D-galactopyranosyl)-ethyl-
phosphonate
(5a). Representative
procedure
(ABCN,
diethylether/tBuOH system): under a nitrogen atmosphere, a
solution of 4 (274 mg, 0.67 mmol), n-Bu3SnCl (57 mL, 0.21 mmol,
0.3 equiv.), NaBH3CN (95% grade, 65 mg, 1.04 mmol, 1.5 equiv.),
diethyl vinylphosphonate (1.1 mL, 6.91 mmol, 10 equiv.), ABCN
(110 mg, 0.45 mmol, 0.67 equiv.), and tert-butanol (0.65 mL,
6.91 mmol, 10 equiv.) in diethylether (5 mL) was stirred at reflux
temperature (35 ◦C). After 4 days, TLC (cyclohexane/EtOAc
1 : 1) showed complete consumption of the starting material and
the formation of two new species, the desired diethyl 2-(2,3,4,6-
tetra-O-acetyl-a-D-glucopyranosyl)-ethylphosphonate 5a (Rf 0.1)
and side product 2,3,4,6-tetra-O-acetyl-b-D-glucopyranose 5b (Rf
0.47). The reaction was concentrated in vacuo, and the oily residue
was dissolved in CH2Cl2 (30 mL). The organic solution was
washed with water (3¥ 30 mL), dried over MgSO4, and reduced
to dryness. The residue was purified by chromatography on a
silica gel column, which was eluted first with cyclohexane/EtOAc
(1 : 1) and then with EtOAc/EtOH (20 : 1), to afford 5a as a
syrup (257 mg, 71%): Rf 0.5 (EtOAc/EtOH 20 : 1); dH (400 MHz,
CDCl3) 5.37 (dd, 1H, J5,6 3.0 Hz, J6,7 2.8 Hz, H-6), 5.24 (dd, 1H,
Experimental section
General methods. All reagents were obtained commercially
and used as received, including anhydrous solvents over molec-
ular sieves, unless otherwise stated. Galactose pentaacetate and
galactosyl bromide 4 were prepared as previously reported (see the
ESI†). Anhydrous acetonitrile was obtained after distillation over
CaH2 under nitrogen atmosphere. All moisture-sensitive reactions
were carried out under an atmosphere of nitrogen in oven-dried
glassware. TLC was performed on precoated slides of Silica Gel 60
F254 (Merck). Spots were visualised under UV light (254/280 nm)
and/or by charring in anisaldehyde stain. Reaction products
were characterised by low- and high-resolution mass spectrometry
(LR/HR-MS) as well as 1H, 13C and, in the case of phosphorus-
containing molecules, 31P NMR spectroscopy. NMR spectra were
J
3,4 5.2 Hz, J4,5 9.3 Hz, H-4), 5.14 (dd, 1H, J5,6 3.3 Hz, J4,5 9.5 Hz,
H-5), 4.22–4.00 (m, 7H, H-3, H-8a, H-8b, 2 OCH2CH3), 3.97
(m, 1H, H-7), 2.09, 2.04, 2.01, 1.98 (all s, 12H, 4¥ C(O)CH3),
2.98–1.56 (m, 4H, H-1a, H-1b, H-2a, H-2b), 1.29 (dt, 6H,
4JH,P 2.8 Hz, JH,H 7.0 Hz, 2 OCH2CH3); dC (100 MHz, CDCl3)
170.4, 170.0, 169.9, 169.8 (4¥ C O), 72.9 (d, JC,P 17.0 Hz,
C-3), 68.3 (C-7), 68.2 (C-4), 67.9 (C-5), 67.6 (C-6), 61.8 (d,
2
2JC,P 6.0 Hz, OCH2CH3), 61.7 (d, JC,P 5.0 Hz, OCH2CH3),
61.5 (C-8), 22.1 (C-1), 20.7 (4¥ C(O)CH3), 19.3 (C-2), 16.4 (d,
JC,P 6.0 Hz, 2¥ OCH2CH3); dP (161.9 MHz, CDCl3) 32.0. m/z
(ESI) 497.1772 M+H+, C20H33O12P requires 497.1782. 5b: Rf 0.47
(cyclohexane/EtOAc 1 : 1); dH (400 MHz, CDCl3) 5.35 (dd, 1H,
1860 | Org. Biomol. Chem., 2011, 9, 1855–1863
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
J3,4 3.4 Hz, J6,7 <1.0 Hz, H-4), 5.11 (dd, 1H, J1a,2 5.5 Hz, J1b,2
10.3 Hz, J2,3 10.3 Hz, H-2), 4.95 (dd, 1H, H-3), 4.10 (dd, 1H, J1a,1b
11.1 Hz, H-1a), 4.00 (d, 2H, J5,6 6.4 Hz, H-6a, H-6b), 3.74 (t,
1H, H-5), 3.20 (dd, 1H, H-1b), 2.06, 1.97, 1.96, 1.91 (all s, 12H,
4¥ C(O)CH3); dC (100 MHz, CDCl3) 170.6, 170.4, 170.3, 170.1
(4¥ C O), 75.0 (C-5), 71.6 (C-3), 67.9 (C-4), 67.2 (C-1), 66.5
(C-2), 62.2 (C-6), 20.9, 20.8 (4¥ C(O)CH3). m/z (ESI) 350.1445
7.93 (d, 1H, J 4.2 Hz, Hthienyl), 7.60 (d, 1H, J 4.2 Hz, Hthienyl), 6.02
(d, 1H, J1¢,2¢ 4.9 Hz, H-1¢), 4.40 (1H, t, J 5.1 Hz, H-2¢), 4.32 (1H,
t, J 4.8 Hz, H-3¢), 4.24–4.15 (1H, m, H-4¢), 4.09– 4.04 (2H, m,
H-5¢), 3.47 (m, 4H, 2¥ CH2), 3.04 (q, J = 7.3 Hz, NCH2CH3, 3
equiv.), 2.88 (m, 4H, 2¥ CH2), 1.17 (t, 28.4H, J 7.3 Hz, NCH2CH3,
3 equiv.); dP (161.9 MHz, D2O) 10.6.
5-Iodo UMP phosphoromorpholidate (11). 5-Iodo UMP 10 18
(292 mg, 0.65 mmol) was dissolved in DMSO and co-evaporated
(3¥) with DMF to remove residual water. The residue was
dissolved in DMSO (0.5 mL) and morpholine (400 mL, 4.6 mmol)
was added to the solution. The mixture was stirred at room
temperature for 5 min. Dipyridyl disulfide (500 mg, 2.3 mmol)
and triphenylphosphine (600 mg, 2.3 mmol) were added in 5 min
intervals, and the reaction was stirred for 60 min at room
temperature. The reaction product was precipitated by addition
of NaI in acetone (0.1 M). The supernatant was removed with a
pipette. The colourless residue of crude 11 was filtered off, washed
with cold acetone, and used in the next reaction step without
further purification.
+
M+NH4 , C14H20O9 requires 350.1446.
Bis(triethylammonium) 2-(2,3,4,6-tetra-O-acetyl-a-D-galacto-
pyranosyl)-ethylphosphonate (6). To a solution of 5a (260 mg,
0.53 mmol) in CH3CN (5 mL) at 0◦C, pyridine (450 mL, 5.52 mmol,
10.5 equiv.) was added, followed by Me3SiBr (730 mL, 5.52 mmol,
10.5 equiv.). The solution was stirred at 0 ◦C for 3 h. The reaction
was quenched with H2O/C5H5N (9 : 1, 5 mL) and the solution was
evaporated to dryness. The residue was purified by Purification
Method 1 (0–50% MeOH against 0.05 M TEAB) to afford the
triethylammonium salt of phosphonic acid 6 in quantitative yield
(230 mg, 0.52 mmol): Rf 0.1 (EtOAc/EtOH 20 : 1); dH (400 MHz,
CDCl3) 5.25 (dd, 1H, J5,6 and J6,7 2.4 Hz, H-6), 5.15–5.05 (m, 2H,
H-4, H-5), 4.14–3.95 (m, 4H, H-3, H-7, H-8a, H-8b), 2.98 (q, 6H, J
7.3 Hz, NCH2CH3), 1.85–1.30 (m, 4H, H-1a, H-1b, H-2a, H-2b),
1.06 (t, 7.6H, J 7.3 Hz, NCH2CH3); dC (100 MHz, CDCl3) 72.3
(d, JC,P 14.3 Hz, C-3), 68.4, 68.3, 68.2 (C-4/C-5/C-6/C-7), 62.0
(C-8), 46.7 (NCH2CH3), 22.5 (d, JC,P 136.4 Hz, C-1), 20.2, 20.0
(4¥ C(O)CH3), 19.3 (C-2), 8.3 (NCH2CH3). m/z (ESI) 439.0999
M-H-, C16H24O12P requires 439.1011.
Bis(triethylammonium) 2-(a-D-galactopyranosyl)-ethylphos-
phono-5-iodouridin-5¢-yl phosphate (12). 11 (129 mg, 0.25 mmol)
and 7 (135 mg, 0.50 mmol, 2 equiv.) were repeatedly co-evaporated
with pyridine (3 mL). The residue was dried under high vacuum
and dissolved in anhydrous DMF (2 mL). To this solution,
tetrazole (0.45 M in CH3CN, 2.8 mL, 1.25 mmol, 5 equiv.) was
added under a nitrogen atmosphere. The reaction was stirred
at room temperature for 4 days. The reaction was evaporated
to dryness and the yellow powder was purified sequentially by
Purification Methods 1 and 2 to afford the triethylammonium salt
of 12 in 28% yield (87.5 mg, 0.07 mmol): Rf 0.5 (iPrOH/H2O/aq.
NH4OH 6 : 3 : 1); dH (400 MHz, D2O) 8.24 (s, 1H, H-6), 5.90 (d,
1H, J1¢,2¢ 4.6 Hz, H-1¢), 4.37–4.31 (m, 2H, H-2¢, H-3¢), 4.27–4.12
(m, 3H, H-4¢, H-5¢a, H-5¢b), 4.03–3.88 (m, 3H, H-3¢¢, H-4¢¢,
H-6¢¢), 3.80 (dd, 1H, J5¢¢,6¢¢ 3.3 Hz, J4¢¢,5¢¢ 9.6 Hz, H-5¢¢), 3.77–3.62
(m, 3H, H-7¢¢, H-8¢¢a, H-8¢¢b), 3.16 (q, J 7.4 Hz, NCH2CH3, 1.7
equiv.), 2.00–1.58 (m, 4H, H-1¢¢a, H-1¢¢b, H-2¢¢a, H-2¢¢b), 1.24 (t, J
7.4 Hz, NCH2CH3, 1.7 equiv.); dC (150.9 MHz, D2O) 163.2 (C-4),
151.6 (C-2), 145.9 (C-6), 89.7 (C-1¢), 83.4 (d, JC,P 9.0 Hz, C-4¢),
75.8 (d, JC,P 16.5 Hz, C-3¢¢), 73.8 (C-2¢), 71.5 (C-7¢¢), 69.6 (C-2¢),
69.1, 68.6, 68.3 (C-4¢¢/C-5¢¢/C-6¢¢), 64.7 (C-5¢), 61.2 (C-8¢¢), 58.6
(C-5), 46.6 (NCH2CH3), 23.9 (d, JC,P 138.0 Hz, C-1¢¢), 18.3 (d,
JC,P < 5 Hz, C-2¢¢), 8.2 (NCH2CH3); dP (161.9 MHz, D2O) 22.3
(d, JP,P 26.8 Hz, CPOPO), -8.7 (d, JP,P 26.8 Hz, CPOPO). m/z
(ESI) 702.9808 M-H-, C17H27IN2O16P2 requires 702.9808.
Bis(triethylammonium) 2-(a-D-galactopyranosyl)-ethylphos-
phonate (7).
A solution of 6 (49.3 mg, 0.11 mmol) in
H2O/MeOH/Et3N (7 : 3 : 1, 11 mL) was stirred for 16 h at room
temperature. The reaction was evaporated to dryness and the
residual white powder was purified by Purification Method 1
(100% 0.05 M TEAB) to afford the triethylammonium salt of
7 as a colorless foam (30 mg, 99%): Rf 0.2 (iPrOH/H2O/aq.
NH4OH 6 : 3 : 1); dH (400 MHz, D2O) 3.82–3.70 (m, 3H, H-3,
H-6, H-7), 3.59 (dd, 1H, J1,2 3.4 Hz, J2,3 9.5 Hz, H-4), 3.54–3.43
(m, 3H, H-5, H-8a, H-8b), 2.97 (q, J 7.3 Hz, NCH2CH3, 1.4
equiv.), 1.70–1.20 (m, 4H, H-1a, H-1b, H-2a, H-2b), 1.04 (t, J
7.3 Hz, NCH2CH3, 1.4 equiv.); dC (100 MHz, D2O) 76.2 (d, JC,P
16.3 Hz, C-3), 71.6 (C-7), 69.6, 69.2 (C-5/C-6), 68.3 (C-4), 61.3
(C-8), 46.7 (NCH2CH3), 23.9 (d, JC,P 134.2 Hz, C-1), 18.4 (C-2),
8.3 (NCH2CH3); dP (161.9 MHz, D2O) 24.9. m/z (ESI) 271.0590
M-H-, C8H16O8P requires 271.0588.
5-(5-Formylthien-2-yl) UMP phosphoromorpholidate (9). 5-(5-
Formylthien-2-yl) UMP 8 (134 mg, 0.31 mmol) was dissolved
in DMSO and co-evaporated (3¥) with DMF to remove residual
water. The residue was dissolved in DMSO (1 mL) and morpholine
(100 mL, 1.66 mmol) was added to the solution. The mixture was
stirred at room temperature for 5 min. Dipyridyl disulfide (221 mg,
1.0 mmol) and triphenylphosphine (263 mg, 1.0 mmol) were added
in 5 min intervals, and the reaction was stirred for 60 min at room
temperature. The reaction product was precipitated by addition
of NaI in acetone (0.1 M). The supernatant was removed with a
pipette. The bright red residue was filtered off, washed with cold
acetone, and purified by Purification Method 1 (0–20% MeOH
against 0.05 M TEAB) to afford phosphoromorpholidate 9 in
7% yield (17 mg, 0.02 mmol): Rf 0.78 (iPrOH/H2O/aq. NH4OH
6 : 3 : 1); dH (400 MHz, D2O) 9.72 (s, 1H, CHO), 8.25 (s, 1H, H-6),
5-FT UDP-C-Gal: Bis(sodium) 2-(a-D-galactopyranosyl)-
ethylphosphono-5-(5-formylthien-2-yl)uridin-5¢-yl phosphate (2).
To a 2-necked round bottom flask charged with 12 (56.7 mg,
2.3 equiv. TEA salts, 0.07 mmol), 5-formylthien-2-ylboronic acid
(20 mg, 0.13 mmol, 1.8 equiv.) and Cs2CO3 (39 mg, 0.16 mmol,
2.3 equiv.) was added degassed H2O (5 mL). The flask was
purged with N2. TPPTS (2.5 mg, 0.004 mmol, 0.06 equiv.) and
Na2Cl4Pd (0. 5 mg, 0.002 mmol, 0.025 equiv.) were added, and
◦
the reaction was stirred under N2 for 2 h at 55 C. The reaction
was cooled to room temperature and filtered through a Millipore
syringe filter (0.22 mm, 33 mm). The filtrate was concentrated in
vacuo to give a white powder, which was purified sequentially
by Purification Methods 1 and 2. Side product 3, resulting
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1855–1863 | 1861
©

View Article Online
from dehydrohalognation, eluted first (17.3 mg, 2.0 equiv. TEA,
0.02 mmol, 32% yield), followed by cross-coupling product 2
(38 mg, 2.3 equiv. TEA, 0.04 mmol, 59% yield). Purified 2 was
converted into its sodium salt form by elution from a Dowex-Na+
column: Rf 0.6 (iPrOH/H2O/aq. NH4OH 6 : 3 : 1); dH (400 MHz,
D2O) 9.76 (s, 1H, CHO), 8.44 (s, 1H, H-6), 7.98 (d, 1H, J 4.2 Hz,
Hthienyl), 7.73 (d, 1H, J 4.2 Hz, Hthienyl), 6.01 (d, 1H, J1¢,2¢ 4.8 Hz, H-
1¢), 4.46–4.39 (m, 2H, H-2¢, H-3¢), 4.33–4.23 (m, 3H, H-4¢, H-5¢a,
H-5¢b), 3.95–3.89 (m, 2H, H-3¢¢, H-4¢¢), 3.85 (dd, 1H, J6¢¢,7¢¢ < 1 Hz,
H-6¢¢), 3.71 (dd, 1H, J5¢¢,6¢¢ 3.4 Hz, J4¢¢,5¢¢ 9.4 Hz, H-5¢¢), 3.67–3.55
(m, 3H, H-7¢¢, H-8¢¢a, H-8¢¢b), 1.95–1.53 (m, 4H, H-1¢¢a, H-1¢¢b,
H-2¢¢a, H-2¢¢b); dC (150.9 MHz, D2O) 187.1 (CHO), 163.5 (C-4),
151.2 (C-2), 144.5 (C-6), 141.3, 139.6, 138.3, 125.2, 108.9 (C5, 4¥
Cthienyl), 89.1 (C-1¢), 83.5 (d, JC,P 8.9 Hz, C-4¢), 75.6 (d, JC,P 18.2 Hz,
C-3¢¢), 74.2 (C-2¢), 71.4 (C-7¢¢), 69.5, 69.5, 69.0 (C-3¢/C-5¢/C-6¢¢),
68.3 (C-4¢¢), 64.7 (C-5¢), 61.1 (C-8¢¢), 23.8 (d, JC,P 140.0 Hz, C-
1¢¢), 18.2 (d, JC,P < 5 Hz, C-2¢¢); dP (161.9 MHz, D2O) 22.2 (d,
JP,P 27.4 Hz, CPOPO), -8.4 (d, JP,P 27.4 Hz, CPOPO). m/z (ESI)
687.0651 M-H-, C22H29N2O17P2S1 requires 687.0668.
using GraFit 5.0.10. The Ki value of 2 was determined by linear
regression analysis (Dixon plot) using 0, 10, 25, 50, 100, 250, 5000
and 1000 mM of inhibitor with 232 mM UDP-Gal in Tris/HCl
buffer. All experiments were carried out in triplicate. Control
experiments carried out in the absence of enzyme showed no
significant degree of chemical hydrolysis (<2% after 3 h).
Acknowledgements
We thank the MRC (Discipline Hopping Award G0701861, to
G.K.W.) for financial support, the UEA for a senior demonstra-
torship (to K.D.), the EPSRC National Mass Spectrometry Service
Centre, Swansea, for the recording of mass spectra, and Ms Niina
Go¨o¨s for help with the HPLC assay. We gratefully acknowledge
support of this work by a Royal Society Research Grant.
References
1 T. Hennet, The galactosyltransferase family., Cell. Mol. Life Sci., 2002,
59, 1081–1095.
UDP-C-Gal: Bis(triethylammonium) 2-(a-D-galactopyranosyl)-
ethylphosphonouridin-5¢-yl phosphate (3)10. Rf 0.44 (iPrOH/
H2O/aq. NH4OH 6 : 3 : 1); dH (400 MHz, D2O) 7.95 (d, 1H, J
8.2 Hz, H-5), 5.94 (m, 2H, H-1¢, H-6), 4.36–4.32 (m, 2H, H-2¢,
H-3¢), 4.26–4.12 (m, 3H, H-4¢, H-5¢a, H¢5¢b), 4.02–3.91 (m, 3H,
H-3¢¢, H-4¢¢, H-6¢¢), 3.82–3.77 (m, 1H, H-5¢¢), 3.76–3.53 (m, 3H,
H-7¢¢, H-8¢¢a, H-8¢¢b), 3.16 (q, J 7.4 Hz, NCH2CH3, 3.2 equiv.),
2.00–1.57 (m, 4H, H-1¢¢a, H-1¢¢b, H-2¢¢a, H-2¢¢b), 1.24 (t, J 7.4 Hz,
NCH2CH3, 3.2 equiv.); dP (161.9 MHz, D2O) 22.2 (d, JP,P 26.8 Hz,
CPOPO), -8.4 (d, JP,P 26.8 Hz, CPOPO)
2 S. Bernatchez, C. M. Szymanski, N. Ishiyama, J. Li, H. C. Jarrell, P. C.
Lau, A. M. Berghuis, N. M. Young and W. W. Wakarchuk, A single
bifunctional UDP-GlcNAc/Glc 4-epimerase supports the synthesis of
three cell surface glycoconjugates in Campylobacter jejuni., J. Biol.
Chem., 2005, 280, 4792–4802.
3 S. K. Partha, K. E. van Straaten and D. A. R. Sanders, Structural
Basis of Substrate Binding to UDP-galactopyranose Mutase: Crystal
Structures in the Reduced and Oxidized State Complexed with UDP-
galactopyranose and UDP., J. Mol. Biol., 2009, 394, 864–877.
4 (a) J. Roesti, C. J. Barton, S. Albrecht, P. Dupree, M. Pauly, K. Findlay,
K. Roberts and G. J. Seifert, UDP-Glucose 4-Epimerase Isoforms
UGE2 and UGE4 Cooperate in Providing UDP-Galactose for Cell
Wall Biosynthesis and Growth of Arabidopsis thaliana., Plant Cell,
2007, 19, 1565–1579; (b) H. Guo, W. Yi, J. K. Song and P. G. Wang,
Current understanding on biosynthesis of microbial polysaccharides.,
Curr. Top. Med. Chem., 2008, 8, 141–151; (c) M. Belanova, P.
Dianiskova, P. J. Brennan, G. C. Completo, N. L. Rose, T. L. Lowary
and K. Mikusova, Galactosyl transferases in mycobacterial cell wall
synthesis., J. Bacteriol., 2008, 190, 1141–1145.
5 (a) L. L. Kiessling and R. A. Splain, Chemical approaches to
glycobiology., Annu. Rev. Biochem., 2010, 79, 619–653; (b) G. K.
Wagner, T. Pesnot and R. A. Field, A survey of chemical methods
for sugar-nucleotide synthesis., Nat. Prod. Rep., 2009, 26, 1172–
1194.
6 (a) X. Qian, M. M. Palcic, “Glycosyltransferase Inhibitors”, in
Carbohydrates in Chemistry & Biology (ed.: B. Ernst, G. Hart, P. Sinay¨),
Wiley-VCH, Weinheim, 2000, pp. 293–328; (b) R. Roychoudhury and
N. L. B. Pohl, New structures, chemical functions, and inhibitors
for glycosyltransferases., Curr. Opin. Chem. Biol., 2009, 14, 168–173;
(c) G. K. Wagner and T. Pesnot, Glycosyltransferases and their assays.,
ChemBioChem, 2010, 11, 1939–1949.
7 (a) H. Hashimoto, T. Endo and Y. Kajihara, Synthesis of the First
Tricomponent Bisubstrate Analogue That Exhibits Potent Inhibition
against GlcNAc:a-1,4-Galactosyltransferase., J. Org. Chem., 1997, 62,
1914–1915; (b) K. Takaya, N. Nagahori, M. Kurogochi, T. Furuike,
N. Miura, K. Monde, Y. C. Lee and S.-I. Nishimura, Rational Design,
Synthesis, and Characterization of Novel Inhibitors for Human b-1,4-
Galactosyltransferase., J. Med. Chem., 2005, 48, 6054–6065.
8 (a) S. Niewiadomski, Z. Beebeejaun, H. Denton, T. K. Smith, R. J.
Morris and G. K. Wagner, Rationally designed squaryldiamides - a
novel class of sugar-nucleotide mimics?, Org. Biomol. Chem., 2010, 8,
3488–3499 and references therein; (b) K. K. Yeoh, T. D. Butters, B. L.
Wilkinson and A. J Fairbanks, Probing replacement of pyrophosphate
via click chemistry: synthesis of UDP-sugar analogues as potential
glycosyl transferase inhibitors., Carbohydr. Res., 2009, 344, 586–591;
(c) L. Ballell, R. J. Young and R. A. Field, Synthesis and evaluation
of mimetics of UDP and UDP-a-D-galactose, dTDP and dTDP-a-D-
glucose with monosaccharides replacing the key pyrophosphate unit.,
Org. Biomol. Chem., 2005, 3, 1109–1115; (d) R. Wang, D. H. Steensma,
Y. Takaoka, J. W. Yun, T. Kajimoto and C. H. Wong, A search for
Enzymology. Streptococcus thermophilus UDP-Gal 4¢-
epimerase (GalE) was purchased from Calbiochem. For the
determination of Km UDP-Gal, GalE (15 mL, 2.5 mg L-1),
UDP-Gal (15 mL, final concentrations: 10 mM–1mM) and MnCl2
(15 mL, 1mM) in Tris/HCl buffer (pH 7) were incubated for
15 min at 30 ◦C (total volume 150 mL, all concentrations are
final concentrations). For the determination of Ki [2], GalE
(15 mL, 10 mg L-1), UDP-Gal (15 mL, 232 mM), 2 (15 mL, final
concentrations: 10 mM–1mM) and MnCl2 (15 mL, 1mM) in
Tris/HCl buffer (pH 7) were incubated for 15 min at 30 ◦C (total
volume 150 mL, all concentrations are final concentrations). The
reactions were stopped by cooling in dry ice. Samples (injection
volume 80 mL) were analyzed immediately by HPLC on a
PerkinElmer Series 200 machine equipped with a SupelcosilTM
LC-18-T column (5 mm, 25 cm ¥ 4.6 mm), a column oven (set
to 35 ◦C), and a diode array detector. The following buffers
were used for HPLC analysis25: buffer A—potassium phosphate
(100 mM), tetrabutylammonium bisulfate hydrogen sulfate
(8 mM), pH 6.5; buffer B—buffer A/methanol (70/30), pH 6.5.
All buffers were filtered through 0.2 mm filters prior to use. The
elution gradient was as follows (flow rate: 1.5 mL min-1): 5%
buffer B for 2 min, 5 to 50% buffer B linearly for 15 min, 50%
B for 1 min, 50 to 100% A for 2 min, and 95% A for 5 min.
The depletion of UDP-Gal and the formation of UDP-Glc,
the product of the epimerization reaction, were monitored at
254 nm. The formation of UDP-Glc was quantified based on
peak area, in reference to a UDP-Glc calibration curve, and used
for the calculation of kinetic parameters. Km and vmax values were
determined by fitting data points to a Michaelis–Menten curve
1862 | Org. Biomol. Chem., 2011, 9, 1855–1863
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
pyrophosphate mimics for the development of substrates and inhibitors
of glycosyltransferases., Bioorg. Med. Chem., 1997, 5, 661–672.
9 T. Pesnot, R. Jørgensen, M. M. Palcic and G. K. Wagner, Structural
and mechanistic basis for a new mode of glycosyltransferase inhibition.,
Nat. Chem. Biol., 2010, 6, 321–323.
10 S. Vidal, I. Bruyere, A. Malleron, C. Auge and J.-P. Praly, Non-
isosteric C-glycosyl analogues of natural nucleotide diphosphate sugars
as glycosyltransferase inhibitors., Bioorg. Med. Chem., 2006, 14, 7293–
7301.
abandoned route., Tetrahedron: Asymmetry, 2000, 11, 305–317; (c) J.
Kovensky, A. Fernandez Cirelli and P. Sinay¨, Synthesis of two isosteric
phosphono analogues of D-galactofuranosyl phosphate., Carbohydr.
Lett., 1999, 3, 271–276; (d) H.-D. Junker, N. Phung and W.-D. Fessner,
Diastereoselective free-radical synthesis of a-substituted C-glycosyl
phosphonates, and their use as building blocks in the HWE-reaction.,
Tetrahedron Lett., 1999, 40, 7063–7066; (e) H.-D. Junker and W.-D.
Fessner, Diastereoselective synthesis of C-glycosylphosphonates via
free-radical glycosylation., Tetrahedron Lett., 1998, 39, 269–272; (f) F.
Casero, L. Cipolla, L. Lay, F. Nicotra, L. Panza and G. Russo, Stereos-
elective Synthesis of the Isosteric Phosphono Analogues of N-Acetyl-
a-D-glucosamine 1-Phosphate and N-Acetyl-a-D-mannosamine 1-
Phosphate., J. Org. Chem., 1996, 61, 3428–3432; (g) R. Meuwly and
A. Vasella, Synthesis of isosteric mono-phosphonate analogues of b-
and a-D-fructose 2,6-bisphosphate., Helv. Chim. Acta, 1986, 69, 751–
760; (h) S. Mirza and A. Vasella, Synthesis of methyl shikimate and of
diethyl phosphashikimate from D-ribose., Helv. Chim. Acta, 1984, 67,
1562–1567; (i) M. Chmielewski, J. N. BeMiller and D. Pat Cerretti,
Synthesis of phosphonate analogs of a-D-glucopyranosyl and a-D-
galactopyranosyl phosphate., Carbohydr. Res., 1981, 97, C1–C4.
17 G. R. Chen and J.-P. Praly, Free-radical and ionic routes towards
hydrolytically stable and bioactive C-glycosyl compounds., C. R. Chim.,
2008, 11, 19–28.
18 G. E. Keck and D. A. Burnett, b-Stannyl enones as radical traps – a
very direct route to PGF2-a., J. Org. Chem., 1987, 52, 2958–2960.
19 R. Rabinowitz, The Reactions of Phosphonic Acid Esters with Acid
Chlorides, A Very Mild Hydrolytic Route., J. Org. Chem., 1963, 28,
2975–2978.
20 A. Collier and G. K. Wagner, A fast synthetic route to GDP-sugars
modified at the nucleobase., Chem. Commun., 2008, 178–180.
21 (a) T. Pesnot and G. K. Wagner, Novel derivatives of UDP-glucose:
concise synthesis and fluorescent properties., Org. Biomol. Chem., 2008,
6, 2884–2891; (b) T. Pesnot, M. M. Palcic and G. K. Wagner, A novel
fluorescent probe for retaining galactosyltransferases., ChemBioChem,
2010, 11, 1392–1398.
11 (a) J. Hajduch, G. Nam, E. J. Kim, R. Fro¨hlich, J. A. Hanover and
K. L. Kirk, A convenient synthesis of the C1-phosphonate analogue of
UDP-GlcNAc and its evaluation as an inhibitor of O-linked GlcNAc
transferase (OGT)., Carbohydr. Res., 2008, 343, 189–195; (b) R. Chang,
T.-T. Vo and N. S. Finney, Synthesis of the C1-phosphonate analog of
UDP-GlcNAc., Carbohydr. Res., 2006, 341, 1998–2004; (c) F. Stolz, M.
Reiner, A. Blume, W. Reutter and R. R. Schmidt, Novel UDP-glycal
derivatives as transition state analogue inhibitors of UDP-GlcNAc 2-
epimerase., J. Org. Chem., 2004, 69, 665–679; (d) F. Stolz, A. Blume, S.
Hinderlich, W. Reutter and R. R. Schmidt, C-Glycosidic UDP-GlcNAc
Analogues as Inhibitors of UDP-GlcNAc 2-Epimerase., Eur. J. Org.
Chem., 2004, 15, 3304–3312; (e) A. Caravano, D. Mengin-Lecreulx,
J.-M. Brondello, S. P. Vincent and P. Sinay¨, Synthesis and Inhibition
Properties of Conformational Probes for the Mutase-Catalyzed UDP-
Galactopyranose/Furanose Interconversion., Chem.–Eur. J., 2003, 9,
5888–5898; (f) B. Waldscheck, M. Streiff, W. Notz, W. Kinzy and R. R.
Schmidt, a(1-3)-Galactosyltransferase Inhibition Based on a New Type
of Disubstrate Analogue, Angew. Chem., Int. Ed., 2001, 40, 4007–4011;
(g) A. Scha¨fer and J. Thiem, Synthesis of Novel Donor Mimetics of
UDP-Gal, UDP-GlcNAc, and UDP-GalNAc as Potential Transferase
Inhibitors., J. Org. Chem., 2000, 65, 24–29; (h) J. Kovensky, M. McNeil
and P. Sinay¨, First Synthesis of UDP-C-D-galactofuranose., J. Org.
Chem., 1999, 64, 6202–6205.
12 J.-P. Praly, A. S. Ardakani, I. Bruye`re, C. Marie-Luce and B. B.
Qin, Towards a- or b-D-C-glycosyl compounds by tin-catalyzed
addition of glycosyl radicals to acrylonitrile and vinylphosphonate, and
flexible reduction of tetra-O-acetyl-a-D-glucopyranosylbromide with
cyanoborohydride., Carbohydr. Res., 2002, 337, 1623–1632.
13 H. Guo, W. Yi, L. Li and P. G. Wang, Three UDP-hexose 4-epimerases
with overlapping substrate specificity coexist in E. coli O86:B7.,
Biochem. Biophys. Res. Commun., 2007, 356, 604–609.
22 T. Mukaiyama and M. Hashimoto, Phosphorylation by Oxidation-
Reduction Condensation. Preparation of Active Phosphorylating
Reagents., Bull. Chem. Soc. Jpn., 1971, 44, 2284.
23 I. J. S. Fairlamb, Regioselective (site-selective) functionalisation of
unsaturated halogenated nitrogen, oxygen and sulfur heterocycles by
Pd-catalysed cross-couplings and direct arylation processes., Chem.
Soc. Rev., 2007, 36, 1036–1045.
24 K. Persson, H. D. Ly, M. Dieckelmann, W. W. Wakarchuk, S. G.
Withers and N. C. J. Strynadka, Crystal structure of the retaining
galactosyltransferase LgtC from Neisseria meningitidis in complex
with donor and acceptor sugar analogs., Nat. Struct. Biol., 2001, 8,
166–175.
25 K. Descroix, T. Pesnot, S. Gehrke, M. M. Palcic and G. K. Wagner,
manuscript in preparation.
14 Y.-K. Niou, W.-L. Wu, L.-C. Lin, M.-S. Yu, H.-Y. Shu, H.-H. Yang
and G.-H. Lin, Role of galE on biofilm formation by Thermus spp.,
Biochem. Biophys. Res. Commun., 2009, 390, 313–318.
15 (a) M. D. Urbaniak, J. N. Tabudravu, A. Msaki, K. M. Matera,
R. Brenk, M. Jaspars and M. A. Ferguson, Identification of novel
inhibitors of UDP-Glc 4¢-epimerase, a validated drug target for African
sleeping sickness., Bioorg. Med. Chem. Lett., 2006, 16, 5744–5747;
(b) J. D. Durrant, M. D. Urbaniak, M. A. J. Ferguson and J. A.
McCammon, Computer-Aided Identification of Trypanosoma brucei
Uridine Diphosphate Galactose 4¢-Epimerase Inhibitors: Toward the
Development of Novel Therapies for African Sleeping Sickness.,
J. Med. Chem., 2010, 53, 5025–5032.
26 H. M Holden, I. Rayment and J. B. Thoden, Structure and function
of enzymes of the Leloir pathway for galactose metabolism., J. Biol.
Chem., 2003, 278, 43885–43888.
16 (a) X. Wen and P. G. Hultin, Concise, stereocontrolled syntheses
of phosphonate C-glycoside analogues of b-D-ManNAc and b-D-
GlcNAc 1-O-phosphates., Tetrahedron Lett., 2004, 45, 1773–1775;
(b) A. Dondoni, A. Marra and C. Pasti, Stereoselective synthesis
of C-glycosylphosphonates from their ketols. Reconsideration of an
27 N. Kochanowski, F. Blanchard, R. Cacan, F. Chirat, E. Guedon, A.
Marc and J.-L. Goergen, Intracellular nucleotide and nucleotide sugar
contents of cultured CHO cells determined by a fast, sensitive, and
high-resolution ion-pair RP-HPLC., Anal. Biochem., 2006, 348, 243–
251.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 1855–1863 | 1863
©