A new methodology for the synthesis of fluorinated exo-glycals and their time-dependent inhibition of UDP-galactopyranose mutase

Article abstract of DOI:10.1002/chem.200500991

Fluorinated carbohydrates constitute a very important class of mechanistic probes for glycosyl-processing enzymes. In this study, we describe the first synthesis of fluorinated and phosphonylated exo-glycals and their corresponding nucleotide sugars in the galactofuranose series. The synthetic protocol that we have developed is a Selectfluor-mediated fluorination/ elimination sequence on phosphonylated exo-glycals, and it offers a new entry into fluorinated carbohydrate chemistry. The challenging E/Z stereochemical assignment of the resulting tetra substituted alkenes, which bear an alkoxy, an alkyl, a fluoro, and a phosphonyl group, has been achieved through NMR experiments. The corresponding (E)- and (Z)-nucleotide fluorosugars have been prepared and tested as inhibitors of UDP-galactopyranose mutase (UGM). UGM is a flavoenzyme that catalyzes the isomerization of uridine diphosphate(UDP)-gal actopyranose into UDP-galactofuranose, a key step of the biosynthesis of important mycobacterial cell-wall glycoconjugates. The two diastereomeric molecules were found to display timedependent inactivation of UGM, as expected from preliminary results using non-fluorinated exo-glycal nucleotides. The inhibitory properties of the two fluorinated molecules led us to suggest that the inactivation mechanism proceeds through two-electron processes, despite the presence of the flavin cofactor within the UGM catalytic site.

Full text of DOI:10.1002/chem.200500991

DOI: 10.1002/chem.200500991
A New Methodology for the Synthesis of Fluorinated exo-Glycals and Their
Time-Dependent Inhibition of UDP-Galactopyranose Mutase
Audrey Caravano,[a] Hirofumi Dohi,[a] Pierre Sinaþ,*[a] and StØphane P. Vincent*[a, b]
Abstract: Fluorinated carbohydrates
constitute a very important class of
mechanistic probes for glycosyl-pro-
cessing enzymes. In this study, we de-
scribe the first synthesis of fluorinated
and phosphonylated exo-glycals and
their corresponding nucleotide sugars
in the galactofuranose series. The syn-
thetic protocol that we have developed
is a Selectfluor-mediated fluorination/
elimination sequence on phosphonylat-
ed exo-glycals, and it offers a new entry
into fluorinated carbohydrate chemis-
try. The challenging E/Z stereochemi-
cal assignment of the resulting tetra-
substituted alkenes, which bear an
alkoxy, an alkyl, a fluoro, and a phos-
phonyl group, has been achieved
through NMR experiments. The corre-
sponding (E)- and (Z)-nucleotide fluo-
rosugars have been prepared and
tested as inhibitors of UDP-galactopyr-
anose mutase (UGM). UGM is a fla-
voenzyme that catalyzes the isomeriza-
tion of uridine diphosphate(UDP)-gal-
actopyranose into UDP-galactofura-
nose, a key step of the biosynthesis of
important mycobacterial cell-wall gly-
coconjugates. The two diastereomeric
molecules were found to display time-
dependent inactivation of UGM, as ex-
pected from preliminary results using
non-fluorinated exo-glycal nucleotides.
The inhibitory properties of the two
fluorinated molecules led us to suggest
that the inactivation mechanism pro-
ceeds through two-electron processes,
despite the presence of the flavin co-
factor within the UGM catalytic site.
Keywords: glycals
·
inhibitors
·
mechanistic probes · nucleotides ·
Selectfluor
Introduction
Several mechanisms have been proposed to describe this in-
triguing reaction: the involvement of a 1,4-anhydro-galac-
tose as a low-energy intermediate,[11] the formation of an
anomeric radical after a single-electron transfer from the re-
duced flavin to an intermediate oxycarbenium species,[8] and
the formation of a covalent adduct between the enzyme[12]
(or its cofactor)[9,13] and the substrate. In a previous study,
we described the synthesis and binding properties of a con-
strained analogue of UDP-galactose locked in a 1,4B boat
conformation, to probe the conformational itinerary of the
galactose residue during its enzymatic interconversion.[14]
Recently, a mechanistic study using radiolabeled UDP-gal-
actose and aimed at trapping a covalent intermediate led to
the characterization of an adduct between the flavin cofac-
tor and the galactose residue.[13] Whereas a stepwise mecha-
nism involving both an SET[8] and the formation a covalent
adduct[9,12,13] may be considered, it was concluded from the
recently resolved three-dimensional structure of reduced
UGM that, given the mode of bending of the flavin cofactor
upon reduction, the formation of a covalent intermediate
and the galactose anomeric position seemed unlikely.[15]
Nevertheless, time-dependent inactivation of UGM has
been observed by Liu et al. with the fluorinated UDP-Galf
analogues 1 and 2[16] and by ourselves using UDP-exo-glycal
The biosynthesis of poly-galactofuranose (Galf) glycoconju-
gates has recently received particular and increasing atten-
tion due to their essential role in maintaining the integrity
of mycobacteria.[1–7] From the biocatalytic viewpoint, the
quest to unravel the biogenesis of galactofuranose (Galf)
led to the discovery of a unique enzymatic ring-contraction/
expansion, namely the interconversion between uridine
diphosphate(UDP)-galactopyranose and the corresponding
furanose, which is catalyzed by UDP-galactopyranose
mutase (UGM or Glf). UGM is a flavoprotein but, unex-
pectedly, no redox process has been evidenced to date.[8–10]
[a]Dr. A. Caravano, Dr. H. Dohi, Prof. P. Sinaþ, Prof. S. P. Vincent
École Normale SupØrieure, DØpartement de Chimie
Institut de Chimie MolØculaire (FR 2769)
UMR 8642: CNRS-ENS-UPMC Paris 6, 24 rue Lhomond, 75231
Paris Cedex 05 (France)
Fax : (+33)1-44-32-33-97
E-mail: pierre.sinay@ens.fr
stephane.vincent@fundp.ac.be
[b]Prof. S. P. Vincent
University of Namur, DØpartement de Chimie, Laboratoire de
Chimie Bio-Organique, rue de Bruxelles
5000 Namur (Belgium)
3114
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3114 – 3123

FULL PAPER
3.[17] Although these results were only observed when UGM
was used in its native (oxidized) state, they strongly suggest-
ed that a covalent bond was transiently formed between the
substrate and a catalytic nucleophile such as the flavin itself.
phosphonyl, an alkoxy, and an alkyl group. Moreover, to the
best of our knowledge, this kind of functionality has not yet
been explored in enzyme inhibition or inactivation studies.
On the other hand, the past few years have seen a great de-
velopment in the field of car-
bohydrate fluorinations, thanks
to the development of a new
generation of fluorinating re-
agents. For instance, the fluori-
nation of endo-glycals by Se-
lectfluor is noteworthy.[21,22]
This reagent appeared to be
the best fluorinating molecule
for the preparation of 2-fluoro-
glycosides functionalized at
the anomeric position with a
broad range of substituents
In view of the interesting properties of exo-glycal 3, we pur-
sued our investigation by designing a new generation of
UDP-exo-glycals that would behave as inactivators of UGM
and would clarify the inactivation mechanism of UGM by
exo-glycals. Thus, the fluorinated exo-glycals 4 and 5 were
naturally envisaged, considering the unique stabilizing prop-
erties of the fluorine atom. In
(see Scheme 1).[21] The mechanism of this electrophilic addi-
tion/nucleophilic substitution is now well understood and in-
volves a 2-fluoro-1-trialkylammonium intermediate adduct
6. Given that this chemistry was developed with endo-gly-
cals, we reasoned that Selectfluor should also be a very
good fluorinating agent for exo-glycals, even when phospho-
mechanistic studies of retain-
ing glycosidases, 2-fluoro-gly-
cosides now constitute key
tools to evidence a transient
covalent process as well as to
identify the catalytic nucleo-
philes.[18] Unfortunately, this
approach (molecules 1 and 2)
did not allow the characteriza-
tion of a covalent adduct with
UGM by means of mass spec-
trometry.[16] From a more gen-
eral perspective, since fluori-
nated substrate analogues con-
stitute a very important class
Scheme 1. NuH=ROH, RNH2, RSH, (RO)2P(O) OH; X=BF4ꢀ, TfOꢀ.
ꢀ
of probes for the elucidation of enzyme mechanisms,[19] and
exo-glycal nucleotides derived from neuraminic acid display
excellent inhibition properties against sialyl transferases,[20]
we decided to explore the chemistry of exo-glycal fluorina-
tion. In addition, developing new techniques for the fluori-
nation of exo-glycals might lead to applications in inhibi-
tion/inactivation studies of numerous glycosyl-processing en-
zymes. Herein, we describe a new methodology leading to
the synthesis of fluorinated nucleotide sugars 4 and 5, their
stereochemical assignment, and their inhibition properties
against UGM.
nylated. In fact, the only reported case in which an endo-
glycal was transformed into a fluoro-endo-glycal was a side
reaction giving poor yields of fluoroalkene 7 (Scheme 1).[22]
As a first approach, we reasoned that if the fluorination
were to be performed on exo-glycal 8 with a base instead of
a nucleophile, an addition–elimination should occur, directly
yielding the desired fluorinated exo-glycal 9 (Scheme 2).
The starting tetrasilylated exo-glycal 8 could be easily pre-
pared in two steps from tetrasilylated g-galactonolactone.[17]
We thus screened the reaction conditions (solvent, tempera-
ture, several non-nucleophilic bases, etc.) to evaluate this
new reaction, but the expected fluorinated exo-glycals 9
were never isolated when the reaction was run under basic
conditions. This result might be attributed to the incompati-
bility of Selectfluor with the different bases that we used.
Adding the base (1,8-diazabicyclo[5.4.0]undec-7-ene (DBU),
trialkylamines, tBuOK, Py, 2,6-di-tert-butylphenol (2,6-
DTBP), etc.) after enol ether 8 had completely reacted with
Results and Discussion
A survey of the literature describing the synthesis of fluori-
nated enol ethers did not lead us to find a general method
for preparing tetrasubstituted alkenes bearing a fluoro, a
Chem. Eur. J. 2006, 12, 3114 – 3123
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3115

S. P. Vincent, P. Sinaþ et al.
transformation has been de-
scribed by Wong et al. for sev-
eral endo-glycals but, surpris-
ingly, applying the published
procedures did not allow the
generation of the desired
fluoro-ketol 10 (Table 1).[26,27]
The typical reaction conditions
(DMF/H2O, CH3CN/H2O or
CH3NO2/H2O; excess Select-
fluor) led only to a complex
mixture of side products
Scheme 2. TBS=tert-butylsilyl, DMAP=4-dimethylaminopyridine.
(Table 1, entries 1–5). Knowing
Selectfluor did not give the expected molecules, but rather a
complex mixture of inseparable products. Somewhat surpris-
ingly, we finally observed the formation of the addition/
elimination products 9, in 30% overall yield, when no base
was added to promote the desired elimination (reaction per-
formed in anhydrous CH3CN/
that Selectfluor reactivity can be strongly modulated by all
of the reaction parameters, especially the solvent system
and the temperature,[22] we carried out an extensive screen-
ing of the reaction conditions (see Table 1). We eventually
found the appropriate conditions, which gave a satisfactory
CH3NO2, 1:1, at 408C). This
Table 1. Fluoro-hydroxylation of tetrasilylated exo-glycal 8.
result can be explained by
the fact that the leaving group
in intermediate
6
is
a
1,4-diazabicyclo[2.2.2]octane
(DABCO) derivative that can
itself play the role of a base.
However, the yield was only
30% and the fluorinated enol
ethers were very difficult to
separate from side products:
under these conditions, some
40% of ketols 10 could be iso-
lated, demonstrating that the
limiting step was indeed the
Entry
MS[a]
Solvent
SF[b]
[equiv][
T
Time
Conv.[c]
[%][%]product
10[d]
9[d]
Side
[f]
8C][h] [%]
[%]
1
2
3
4
5
6
7
8
–
–
–
–
–
4 Š
4 Š
4 Š
4 Š
4 Š
4 Š
4 Š
–
DMF/H2O
DMF/H2O
CH3NO2/H2O
CH3CN/H2O
CH3CN/H2O
CH3NO2
CH3NO2
CH3NO2
CH3NO2
CH3CN
4
4
4
1.3
1.3
3
3
6
10
3
3
RT
50
RT
RT
50
RT
40
40
40
30
30
40
36
12
48
48
12
72
12
12
12
48
12
24
24
100[e]
100[e]
100[e]
100[e]
100[e]
50
80
85
88
93
0
0
0
0
0
0
0
0
0
0
100
100
100
100
100
30
12
25
38
7
20[f]
48[f]
50[f]
43[f]
82
85
62
0
0
10
10
7
4
5
elimination
path A).
(Scheme 2,
9
10
11
12
13
[g]
[g]
[g]
CH3CN/CH3NO2
CH3CN/CH3NO2
CH3CN/CH3NO2
90
0
7
100
Fluoro-hydroxylation reaction:
In view of the above result, we
had to envisage the prepara-
tion of the target molecules 9
through a two-step sequence.
In our hands, the best proce-
dure to produce phosphonylat-
3
3
93
–
16
0
[d]
40
[a]Molecular sieves. [b]Selectfluor (TfO form). [c]Determined by 31P NMR of the crude reaction mixture.
[d]Yield of isolated product. [e]The starting material was totally decomposed. [f]Ketols 10 were inseparable.
[g]1:1 (v/v). [h]Detected by 31P and 19F NMR spectroscopy.
ꢀ
ed exo-glycals proved to be the method developed by Lin
et al., which involves activating an anomeric ketol by tri-
fluoroacetylation under basic conditions, giving exo-glycals
directly in good yields with a Z stereospecificity.[23–25] For
this reaction, the nature of the activating reagent, generating
an anomeric leaving group, is of extreme importance since
other common leaving groups, such as sulfonates, do not
cleanly produce exo-glycals. The synthesis of 9 therefore im-
plied the two-step sequence depicted in Scheme 2 (path B):
i) synthesis of a fluorinated and phosphonylated intermedi-
ate ketol 10; ii) application of the Lin elimination procedure
to generate product 9.[23,24] Thus, we chose to perform a
fluoro-hydroxylation of tetrasilylated exo-glycal 8. Such a
85% yield (Table 1, entry 11) after purification. Unexpect-
edly, the key parameter that ensured a successful and repro-
ducible fluoro-hydroxylation was that this reaction had to
be performed under strictly anhydrous conditions. Acetoni-
trile was found to be the solvent of choice, either pure or
mixed with anhydrous nitromethane (Table 1, entries 6–12).
This is in marked contrast to literature reports, which de-
scribe the same reaction with water as cosolvent. Moreover,
nitromethane had been described as the ideal solvent for
electrophilic fluorination/nucleophilic substitution of endo-
glycals.[22] For the fluoro-hydroxylation of phosphonylated
exo-glycal 8, running the reaction in pure nitromethane led
only to moderate yields of fluoro-ketol 10, which was conta-
3116
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3114 – 3123

Synthesis and Inhibitory Effect on UGM of Fluorinated exo-Glycals
FULL PAPER
minated with inseparable side products (Table 1, entries 6–
9). When the reaction was monitored by 19F NMR spectrom-
etry, we did not observe the characteristic signals of the in-
termediate adduct 6, even though such an adduct has been
purified and characterized with other carbohydrates.[22] This
probably signifies that this intermediate is rather unstable in
this case. Surprisingly, after filtering off the required molec-
ular sieves, the best yields were not obtained by addition of
aqueous solutions, but rather by diluting the crude reaction
mixture with dichloromethane and treating it with silica gel.
Fluoro-ketol 10 was isolated as a mixture of two diastereo-
mers in a ratio of 1:1.3.
characterized, probably because they are too labile. This
shows that when the elimination is performed on fluoro-
ketol 10, the rate-determining step is the elimination of the
trifluoroacetate, whereas when applied to non-fluorinated
ketols,[17,23,24,28] the acylation determines the elimination ki-
netics. All of these observations are consistent with an E1-
type mechanism, but do not rule out an E1cb mechanism.
The desired exo-glycals 9 were obtained in 65% yield in a
Z/E ratio of 64:36.
E/Z structural assignment: Because of the lack of 1H–1H
coupling constants, the stereochemistry of tetrasubstituted
alkenes is generally difficult to establish. Determining the
configuration of exo-glycals 9 proved to be even more diffi-
cult since the exocyclic double bond bears three heteroele-
ments. The stereochemistry of
Standard conditions[14,28] used for the preparation of exo-
glycal 6 were applied to generate the fluorinated molecules
9, but the reaction proved very sluggish (Table 2, entries 1–
the starting (Z)-exo-glycal
8
Table 2. Conversion of the fluoro-ketol 10 to the desired fluorinated exo-glycal 9.
was conclusively assigned on
the basis of 1H NMR and
NOE experiments. Further-
more, given the novelty of this
kind of functionality, we did
not know if the coupling con-
stants between 19F, 31P, and the
two olefinic 13C were charac-
teristic of the stereochemistry.
Moreover, we had to face an-
other difficulty: the non-fluori-
nated (E)-exo-glycal (E)-8 had
never been synthesized since
the methods for producing it
are Z stereospecific. We rea-
soned that with (E)-exo-glycal
Entry
Solvent
(CF3CO)2O
[equiv][eq[euqi[vu]iv]
Py
DMAP
T
Time
Yield
Z/E
8C][h] [%] ratio
1
2
3
4
5
6
THF
THF
THF
10
10
10
10
10
10
20
40
40
excess
excess
excess
–
–
–
20
20
20
RT
RT
35
35
35
24
48
52
24
24
10
12.5
18
30
63
72
n.d.[a]
4:6
4:6
7:3
6:4
THF/Py (6:4)
CH2Cl2/Py (6:4)
CH2Cl2/Py (6:4)
35
80
7:3
[a]Not determined.
3). Although pyridine and triethylamine are the usual bases
for this elimination, an excess of DMAP gave by far the
best results (Table 2, entries 4–6). The fact that the reaction
is much slower when applied to fluoro-ketol 10 as compared
to its non-fluorinated homologue[17] is interesting from a
mechanistic point of view. Indeed, it strongly suggests that
this elimination involves an oxycarbenium species 12 as an
intermediate (Scheme 3): the electron-withdrawing charac-
ter of the fluorine atom disfavors the formation of a cationic
intermediate and thus reduces the reaction rate. A competi-
tion between an E1-type and a concerted E2-type mecha-
nism has been previously discussed by Lin et al., but was
not fully demonstrated.[28] A second feature reinforces the
suggestion of an E1-type stepwise mechanism: we were able
to isolate intermediate 11, whereas, to date, such quaternary
anomeric trifluoroacetates have never been observed or
(E)-8 in hand, we would strengthen our assignments based
on NMR methods such as NOE experiments. In a related
project aimed at producing C-glycosidic Galf-1-phosphate
analogues bearing a leaving group at the anomeric position
to generate potential suicide inhibitors of UGM, we fortui-
tously prepared and fully characterized the (E)-exo-glycal
(E)-8 (Scheme 4). The reaction of ketol 13 with (diethylami-
no)sulfurtrifluoride (DAST) under classical conditions gave
the 1-fluoro-phosphonate 14 in 41% yield. The stereochem-
istry at the anomeric position could be easily assigned by
3
virtue of the characteristic J coupling constant between H-2
and the anomeric fluorine atom.[29] Under these conditions,
an elimination also occurred, leading to a mixture of diaste-
reomeric exo-glycals 8 in a Z/E ratio of 2:1. The yield of the
E diastereomer was poor, but sufficient to produce enough
material for a full NMR characterization. Interestingly, the
E/Z diastereomers were very
easily separated since they dis-
played very different Rf values
(DRf >0.3) upon silica gel
chromatography. The (Z)-exo-
glycals 8 (with different protec-
tive groups) were always more
Scheme 3.
polar than their E isomers. We
Chem. Eur. J. 2006, 12, 3114 – 3123
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3117

S. P. Vincent, P. Sinaþ et al.
showed poor inhibition percen-
tages (< 10% at [I] = [UDP-
Galf] = 1 mm). Under non-re-
ducing conditions, a time-de-
pendent inactivation was ob-
served (Figure 1), as in the
case of exo-glycal 3[17] or Liuꢁs
fluorinated UDP-Galf ana-
logues 1 and 2.[16] Unfortunate-
ly, we were unable to find evi-
dence of a covalent adduct be-
tween the inactivators and the
protein, or the cofactor, by
mass spectrometry. This result
indicates that the still putative
Scheme 4.
tentatively assigned the Z configuration to the slower-mi-
grating fluorinated species 9. This assignment was eventually
confirmed by NOE experiments. Indeed, distinct NOEs
were observed between the benzylic protons and H-4 of
(Z)-8 and (Z)-9 (Scheme 4). These effects were not seen
with (E)-8 and (E)-9, but instead NOEs between the benzyl-
ic protons and H-3 were observed for both E diastereomers.
The clear similarities in migratory aptitudes and NOEs con-
stitute convincing evidence for the assignment of the E/Z
stereochemistry of molecules 9.
Synthesis of target nucleotide sugars: From the UGM inhib-
ition perspective, we were satisfied by the fact that the elim-
ination leading to exo-glycals 9 was not stereoselective. As a
matter of fact, no efficient method has been described for
the preparation of (E)-phosphonylated exo-glycals. There-
fore, when we described the synthesis and the inactivation
properties of (Z)-exo-glycal 3, we were unable to study the
inhibition of its E isomer. We now have the two fluorinated
diastereomers at our disposal. As in the synthesis of UDP-
(Z)-exo-glycal-Galf 3,[17] (E)-9 and (Z)-9 could be efficiently
deprotected if the hydrogenolysis step was performed first,
followed by desilylation (compounds (E)-15 and (Z)-15,
Scheme 5). The corresponding nucleotide sugars were then
obtained in yields of 64% and 70% for molecules 4 and 5,
respectively, using the Bogachev procedure.[14,30]
~
*
Figure 1. Time-dependent inactivation of 4 ( )and 5 ( ).
adduct is not sufficiently stable under mass spectrometry
analysis conditions, despite the presence of a fluorine atom
in the molecule. Both the (Z)- and (E)-fluorinated exo-gly-
cals 4 and 5 displayed much slower inactivation kinetics
compared to the non-fluorinated analogue 3. Indeed, a com-
plete loss of UGM residual activity was only observed after
50–70 h of incubation with the fluorinated molecules 4 and 5
(Figure 1), whereas under the same conditions UGM was in-
activated within 15 h by non-fluorinated exo-glycal 3.[17]
Since molecules 3 and 4 are structurally very similar, they
should display similar binding properties towards UGM.
Therefore, the significant difference in their inhibition kinet-
ics must originate from the inactivation mechanism itself. It
is reasonable to conclude that the electron-withdrawing
character of the fluorine atom present in molecule 4 is the
cause of the dramatic decrease
UGM inactivation study: Fluorinated exo-glycals 4 and 5
were found to display the same type of inhibition properties
as the non-fluorinated analogue 3:[17] under reducing condi-
tions (obtained by dithionite addition), the two molecules
in the inactivation rate. Since
captodative (or push–pull) ef-
fects stabilize radical inter-
mediates, the presence of the
fluorine atom should not de-
crease the rate of a radical pro-
cess. On the contrary, if the in-
activation mechanism is a two-
electron process, such as a pro-
Scheme 5. Reagents and conditions: a) H2, Pd/C, Et3N CH2Cl2, RT; b) nNu4NF·3H2O, THF, 08C; c) UMP-N-
methylimidazolium, CH3CN, RT.
3118
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3114 – 3123

Synthesis and Inhibitory Effect on UGM of Fluorinated exo-Glycals
FULL PAPER
silica gel (Kieselgel Si 60; 40–63 mm). Size-exclusion chromatography was
performed on a Sephadex G15 column (2.5”60 cm) using a Pharmacia
Biotech ¾kta FPLC apparatus; fractions containing uridine derivatives
were detected with UV light. When required, nucleotide sugars were pu-
rified by semipreparative HPLC using a Waters Delta prep 4000 chroma-
tonation (Scheme 6), the presence of the fluorine atom
should strongly alter the inactivation kinetics. It is well-es-
tablished that enol ethers are readily activated by acids:
their intrinsic basicity can indeed be strongly decreased if an
electron-withdrawing group, such as a fluorine atom, is di-
rectly linked to the double bond (their stabilities towards
acids are then improved). The decrease in basicity of the
enol ether functionality in molecules 4 and 5, compared to
3, is likely to be the origin of the observed difference in in-
hibition kinetics. Therefore, the inactivation mechanism of
UGM with fluorinated molecules 4 and 5 is unlikely to in-
tography system equipped with
a NovaPak C18 (1”10 cm) column
(eluent: 50 mm triethylammonium acetate, pH 6.8). For enzymatic assays,
a Waters 600E analytical apparatus equipped with a Zorbax C18-SB
column (25”0.46 cm, 5 mm) was used (eluent: 50 mm triethylammonium
acetate, pH 6.8). UDP-Galf was prepared according to known proce-
dures.[31, 32]
Enzyme kinetics and inhibition assays: The conditions described by Liu
et al. were followed.[16] All assays were performed using a potassium
phosphate buffer (100 mm, pH 6.8) at T = 218C. To obtain reductive
conditions, freshly prepared sodium dithionite solutions were used to
allow a final concentration of 20 mm. When native mutase was used
(without sodium dithionite), reactions were conducted in the dark under
aerobic conditions. All inhibition studies involved measuring the conver-
sion of pure starting UDP-Galf into UDP-Galp (compared with a com-
mercially available sample) by analytical HPLC (C18 column, elution
with 50 mm triethylammonium acetate, pH 6.8; detection at 262 nm).
Substrate and inhibitor concentrations were adjusted to 1 mm from titrat-
ed mother solutions. As a control, the relative concentrations of any spe-
cies in the reaction mixtures could be determined titrimetrically using
UMP as an internal standard. Enzyme concentrations were adjusted to
produce a conversion of UDP-Galp between 5 and 15%. Reactions were
stopped by freezing the solution in liquid nitrogen. Residual enzyme ac-
tivities were then measured in the presence of inhibitors at three differ-
ent times, compared with the same experiment conducted without inhibi-
tors, and the values were averaged.
volve
a single-electron transfer or radical chemistry
(Scheme 6, pathways B and C). Instead, a protonation
(Scheme 6, pathway A) can reasonably be invoked as the
first step of the inactivation mechanism of UGM by exo-gly-
cals.
In conclusion, we have described an efficient methodolo-
gy for the synthesis of phosphonylated and fluorinated exo-
glycals, using Selectfluor as fluorinating agent. This new re-
action sequence has been exploited for the synthesis of both
diastereomers of UDP-fluoro-exo-galactal, 4 and 5. As ex-
pected, these two molecules display time-dependent inacti-
vation of UDP-galactopyranose mutase. Comparison with
the inhibition properties of the non-fluorinated analogue 3
has led to a clarification of the inactivation mechanism of
UGM by UDP-exo-glycals: despite the presence of the FAD
cofactor close to the galactose binding site, the mechanism
of inactivation most probably proceeds through a sequence
of two-electron processes. We are now investigating applica-
tion of this new methodology to other carbohydrates, with a
view to generating inhibitors or mechanistic probes of other
glycosyl-processing enzymes.
Inactivation assays: Several reactors containing the enzyme (5.4 mm) and
the inactivator 4 or 5 (3.56 mm) were incubated for different times (0, 3,
6, 9, 14, 24, 36, 48, and 72 h). UDP-a-d-Galf (1 mm) was then added to
initiate the reaction and the sample was freeze-dried in liquid nitrogen to
stop the enzymatic reaction. The residual activity of the enzyme was de-
termined as described previously for the inhibition assays.[17]
Test for reversibility of the inactivation by UDP-1’-fluoro-exo-glycal-d-
galactofuranose: Two samples containing UDP-1(1’Z)-[1’F]-exo-glycal-
Galf 4 (3.2 mm) and UDP-1(1’E)-[1’F]-exo-glycalGalf 5 (3.2 mm) were in-
dividually incubated with the mutase (5.4 mm) in phosphate buffer (total
volume 60 mL, 100 mm, pH 7.4) for 72 h, cooled in ice and in the dark
under aerobic conditions. Similar experiments were performed without
the inactivator. The resulting solutions were applied to filters of a Viva-
spin 0.5 mL concentrator equipped with a 10000 MWCO PES mem-
brane; they were diluted with 300 mL of phosphate buffer and centri-
fuged (2000 g for 2 min) to obtain a final volume of 60 mL. This proce-
dure was performed four times to remove all small molecules. The resid-
ual enzyme activity was measured before and after filtration (with and
without dithionite).[16,17]
Experimental Section
Materials and procedures: All chemicals were purchased from Sigma, Al-
drich or Fluka and were used without further purification. Tetrahydrofur-
an, diethyl ether, and toluene were freshly distilled from sodium benzo-
phenone, dichloromethane from P2O5, and acetonitrile from CaH2. 1H,
13C, 19F, and 31P NMR spectra were recorded with Bruker AC-250 and
AMX-400 spectrometers. All compounds were characterized by 1H, 13C,
19F, and 31P NMR as well as by 1H–1H and 1H–13C correlation experi-
ments. Specific optical rotations were measured on a Perkin-Elmer 241
polarimeter with solutions in a 1 dm cell. Melting points were determined
with a Büchi 535 apparatus. Column chromatography was performed on
Atom and position numberings: We have systematically numbered the
phosphonate methylene group as 1’ and adopted the usual numbering for
carbohydrates from 1 to 6, with 1 for the anomeric position. For nucleo-
tide sugars, we used the conventional ribose and pyrimidine numberings:
Scheme 6.
Chem. Eur. J. 2006, 12, 3114 – 3123
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3119

S. P. Vincent, P. Sinaþ et al.
1’ for the anomeric position, 1’’ for the nitrogen atom linked to the
J(6a,6b) = 10.2 Hz, 1H; H-6aA), 3.62 (ABX, J(5,6b) = 5.4 Hz, J-
ribose.
(6a,6b) = 10.2 Hz, 1.8H; H-6bA,B), 0.97 (s, 7.2H; Si-tBu), 0.96 (s, 7.2H;
Si-tBu), 0.95 (s, 9H; Si-tBu), 0.94 (s, 9H; Si-tBu), 0.93 (s, 9H; Si-tBu),
0.92 (s, 9H; Si-tBu), 0.92 (s, 7.2H; Si-tBu), 0.87 (s, 7.2H; Si-tBu), 0.25 (s;
Si-Me), 0.21 (s; Si-Me), 0.19 (s; 2Si-Me), 0.18 (2s; 2Si-Me), 0.17 (2s;
2Si-Me), 0.16 (s; Si-Me), 0.14 (s; Si-Me), 0.13 (s; Si-Me), 0.09 (s; 2Si-
Me), 0.08 (s; Si-Me), 0.04 (s; Si-Me), 0.02 ppm (s; Si-Me); 13C NMR
(100 MHz, CDCl3): d = 136.4 (d, J(C,P) = 6.4 Hz; Cq arom.), 136.15 (d,
J(C,P) = 6.2 Hz; 2Cq arom.), 136.12 (d, J(C,P) = 6.5 Hz; Cq arom.),
(1(1’)Z)-2,3,5,6-Tetra-O-tert-butyldimethylsilyl-1-(dibenzyloxyphosphor-
yl)methylidene-d-galactofuranose (8): A cooled (ꢀ708C) solution of di-
benzyl methylphosphonate (4.96 g, 18.0 mmol) in THF (36 mL) under
argon atmosphere was treated first with nBuLi (6.9 mL, 2.5m solution in
hexane, 17.3 mmol) and, after 20 min, with a solution of 2,3,5,6-tetra-O-
tert-butyldimethylsilyl-d-galactono-1,4-lactone (4.56 g, 7.19 mmol) in
THF (7 mL). The temperature was maintained at ꢀ708C for 10 min, and
then the mixture was allowed to warm to ꢀ408C over a period of 1 h.
The solution was then diluted with 1m phosphate buffer (pH 7; 170 mL)
and extracted with CH2Cl2 (2”350 mL). The combined organic phases
were dried over MgSO4, filtered, concentrated, and dried overnight in
vacuo. The crude material was then dissolved in THF (50 mL) at 08C,
and pyridine (5.81 mL, 71.9 mmol) and trifluoroacetic anhydride (5.0 mL,
36.0 mmol) were added. After 4 h at 08C, the reaction was stopped by
adding saturated aqueous NaHCO3, and the mixture was extracted with
EtOAc (350 mL). The organic phase was dried over MgSO4, filtered, and
concentrated under reduced pressure. Purification by chromatography on
silica gel (cyclohexane/EtOAc, 5:1) afforded 8 (4.55 g, 71% yield) as a
colorless syrup.
128.42–127.70 (20CH arom.), 105.59 (dd, J(1,P)
=
2.2 Hz, J(1,F)
=
24.3 Hz; C-1A), 102.13 (d, J(1,P) 1.0 Hz, J(1,F)
=
= 20.5 Hz; C-1B),
88.49 (C-4A), 88.19, 86.28 (2d, J(1’,P) = 162.8 Hz, J(1’,F) = 191.9 Hz; C-
1’B), 87.80, 85.97 (2d, J(1’,P) = 162.7 Hz, J(1’,F) = 184.0 Hz; C-1’A),
84.62 (C-4B), 82.82 (d, J(2,P) = 7.5 Hz; C-2A), 79.61 (d, J(2,P) = 3.5 Hz,
J(2,F) = 5.3 Hz, C-2B), 78.90 (C-3A), 78.05 (C-3B), 73.61 (C-5B), 73.23 (C-
5A), 68.61 (d, J(C,P)
= 5.9 Hz; CH2Ph), 68.50 (d, J(C,P) = 5.8 Hz,
CH2Ph), 68.38 (d, J(C,P) = 6.3 Hz; CH2Ph), 67.67 (d, J(C,P) = 6.3 Hz;
CH2Ph), 65.92 (C-6A), 64.40 (C-6B), 26.07 (Si-C(CH3)3), 25.98 (Si-C-
(CH3)3), 25.90 (2Si-C(CH3)3), 25.86 (Si-C(CH3)3), 25.77 (2Si-C(CH3)3),
25.68 (2Si-C(CH3)3), 18.48 (Si-C(CH3)3), 18.27 (Si-C(CH3)3), 18.25 (Si-C-
(CH3)3), 18.14 (Si-C(CH3)3), 18.03 (Si-C(CH3)3), 17.94 (Si-C(CH3)3), 17.81
(Si-C(CH3)3), 17.79 (Si-C(CH3)3), ꢀ3.97 (Si-Me), ꢀ4.18 (2Si-Me), ꢀ4.22
(2Si-Me), ꢀ4.26 (Si-Me), ꢀ4.29 (3Si-Me), ꢀ4.49 (Si-Me), ꢀ4.54 (Si-Me),
ꢀ4.65 (Si-Me), ꢀ4.66 (Si-Me), ꢀ5.19 (Si-Me), ꢀ5.22 (Si-Me), ꢀ5.35 (Si-
Me), ꢀ5.39 ppm (2Si-Me); 31P NMR (101 MHz, CDCl3): d = 17.76 (d,
J(P,F) = 75.9 Hz; major diastereomer), 16.00 ppm (d, J(P,F) = 72.9 Hz;
minor diastereomer); 19F NMR (235 MHz, CDCl3): d = ꢀ210.33 (dd,
J(F,P) = 75.3 Hz, J(F,H1) = 44.7 Hz; major diastereomer), ꢀ211.12 ppm
(dd, JF-P = 72.9 Hz, JF-H1’ = 44.7 Hz; minor diastereomer); MS (DCI-
NH3): m/z (%): 929 (100) [M+H]+, 946 (95) [M+NH4]+; elemental anal-
ysis calcd (%) for C45H82FO9PSi4: C 58.15, H 8.89; found: C 57.75, H
9.22.
[a]2D1
=
+6.4 (c = 1.1 in CHCl3); 1H NMR (400 MHz, CDCl3): d =
7.38–7.30 (m, 10H; H arom.), 5.09 (ABX, J(H,P) = 7.4 Hz, J(H,H) =
12.0 Hz, 2H; CH2Ph), 5.02 (ABX, J(H,P) = 7.9 Hz, J(H,H) = 12.0 Hz,
2H; CH2Ph), 4.61 (dd, J(1’,2) = 1.1 Hz, J(1’,P) = 10.5 Hz, 1H; H-1’),
4.47 (dt, J(1’,2) = 1.1 Hz, J(2,P) = J(2,3) = 3.9 Hz, 1H; H-2), 4.41 (t,
J(3,4) = J(4,5) = 3.9 Hz, 1H; H-4), 4.33 (dt, J(2,3) = J(3,4) = 3.9 Hz,
J(3,P) = 0.6 Hz, 1H; H-3), 3.92 (dt, J(5,6a) = 7.0 Hz, J(5,6b) = 4.6 Hz,
1H; H-5), 3.77 (ABX, J(5,6a) = 7.0 Hz, J(6a,6b) = 10.2 Hz, 1H; H-
6a), 3.65 (ABX, J(5,6b) = 4.6 Hz, J(6a,6b) = 10.2 Hz, 1H; H-6b), 0.94
(s, 9H; Si-tBu), 0.92 (s, 9H; Si-tBu), 0.90 (s, 9H; Si-tBu), 0.88 (s, 9H; Si-
tBu), 0.16 (s, 6H; 2Si-Me), 0.13 (s, 3H; Si-Me), 0.12 (s, 3H; Si-Me), 0.11
(s, 3H; Si-Me), 0.09 (s, 3H; Si-Me), 0.08 (s, 3H; Si-Me), 0.02 ppm (s, 3H;
Si-Me); 13C NMR (100 MHz, CDCl3): d = 173.03 (d, J(1,P) = 2.6 Hz; C-
1), 136.93 (d, J(C,P) = 7.3 Hz; Cq arom.), 136.91 (d, J(C,P) = 7.5H; Cq
arom.), 128.23–127.57 (10CH arom.), 90.27 (C-4), 81.78 (d, J(1’,P)
195.7 Hz; C-1’), 81.05 (d, J(2,P) = 13.9 Hz; C-2), 76.39 (C-3), 72.57 (C-
5), 66.63 (d, J(C,P) 4.5 Hz; CH2Ph), 66.60 (d, J(C,P) 5.0 Hz;
CH2Ph), 64.39 (C-6), 25.93 (Si-C(CH3)3), 25.82 (Si-C(CH3)3), 25.66 (2Si-
C(CH3)3), 18.28 (Si-C(CH3)3), 18.05 (Si-C(CH3)3), 17.83 (Si-C(CH3)3),
17.72 (Si-C(CH3)3), ꢀ3.82 (Si-Me), ꢀ4.02 (Si-Me), ꢀ4.34 (Si-Me), ꢀ4.38
(Si-Me), ꢀ4.41 (Si-Me), ꢀ4.81 (Si-Me), ꢀ5.40 (Si-Me), ꢀ5.42 ppm (Si-
Me); MS (DCI-NH3): m/z: 893 [M+H]+; elemental analysis calcd (%)
for C45H81O8PSi4: C 60.49, H 9.14; found: C 60.36, H 9.27.
1-(Dibenzyloxyphosphoryl)fluoromethyl-1-O-trifluoroacetyl-2,3,5,6-tetra-
O-tert-butyldimethylsilyl-d-galactofuranose (11): The mixture of ketols
10 (100 mg, 0.11 mmol) was dissolved in anhydrous THF (2.5 mL), and
anhydrous pyridine (349 mL, 4.40 mmol) and trifluoroacetic anhydride
(150 mL, 1.10 mmol) were successively added at 08C. The solution was
stirred for 3 h at 08C and then allowed to warm to room temperature.
After concentration under vacuum, the crude reaction mixture was puri-
fied by flash chromatography on silica gel (cyclohexane/EtOAc, 11:1) to
give compound 11 (68 mg, 62% yield) as a colorless oil.
=
=
=
1H NMR (400 MHz, CDCl3): d = 7.37–7.30 (m, 10H; H arom.), 6.02 (dd,
J(1’,P)
= 14.6 Hz, J(1’,F) = 45.7 Hz, 1H; H-1’), 5.76 (dd, J(2,3) =
3.0 Hz, J = 9.8 Hz, 1H; H-2), 5.17–4.96 (m, 4H; 2CH2Ph), 5.07 (d, J(3,4)
= 5.7 Hz, 1H; H-4), 4.29–4.26 (m, 2H; H-3 and H-5), 4.00 (ABX, J-
(5,6a) = 1.7 Hz, J(6a,6b) = 11.4 Hz, 1H; H-6a), 3.73 (ABX, J(5,6b) =
1.8 Hz, J(6a,6b) = 11.4 Hz, 1H; H-6b), 0.94 (s, 18H; 2Si-tBu), 0.92 (s,
9H; Si-tBu), 0.86 (s, 9H; Si-tBu), 0.23 (s, 3H; Si-Me), 0.19 (s, 3H; Si-
Me), 0.13 (s, 3H; Si-Me), 0.12 (s, 3H; Si-Me), 0.11 (s, 3H; Si-Me), 0.09
(s, 3H; Si-Me), 0.08 (s, 3H; Si-Me), 0.07 (s, 3H; Si-Me); 13C NMR
(100 MHz, CDCl3): d = 197.97 (d, J(C,F) = 12.5 Hz; C-1), 157.04 (q,
1-(Dibenzyloxyphosphoryl)fluoromethyl-2,3,5,6-tetra-O-tert-butyldime-
thylsilyl-d-galactofuranose (10): exo-Glycal 8 (2.5 g, 2.80 mmol) was dis-
solved in a mixture of anhydrous MeCN (100 mL) and nitromethane
(100 mL) under argon atmosphere in the presence of 4 Š molecular
sieves (6.6 g). The suspension was stirred for 3 h at room temperature,
and then Selectfluor (bis(triflate), 4.0 g, 8.41 mmol) was added at 08C.
The resulting solution was stirred for 12 h at 308C, then diluted with
EtOAc and filtered through a pad of Celite. The filtrate was concentrat-
ed under reduced pressure, then diluted with CH2Cl2 (2 mL), and silica
gel (ca. 2 g) was added. The residue was purified by chromatography on
silica gel (cyclohexane/EtOAc, 15:1 ! 12.5:1) to give colorless 10 (2.08 g,
80% yield) as a mixture of two diastereomers A and B in a ratio of
1:1.25.
1H NMR (400 MHz, CDCl3): d = 7.40–7.30 (m, 20H; H arom.), 5.28–
5.11 (m, 8H; 4CH2Ph), 5.11 and 5.01 (2d, J(1’,P) = 3.8 Hz, J(1’,F) =
44.2 Hz, 1H; H-1’A), 4.98 and 4.87 (2d, J(1’,P)
44.6 Hz, 0.8H; H-1’B), 4.98 (brs, 1H; OH-1A), 4.57 (d, J(OH,2) = 2.0 Hz,
0.8H; OH-1B), 4.48 (dd, J(2,3) = 4.3 Hz, J(2,OH) = 2.0 Hz, 0.8H; H-
2B), 4.33 (dd, J(2,3) = 4.3 Hz, J(3,4) = 3.1 Hz, 0.8H; H-3B), 4.31 (dd,
J(3,4) = 1.0 Hz, J(4,5) = 5.4 Hz, 1H; H-4A), 4.20 (m, 1H; H-3A), 4.17 (s,
1H; H-2A), 4.14 (t, J(3,4) = J(4,5) = 3.1 Hz, 0.8H; H-4B), 3.91 (q, J(4,5)
= J(5,6a) = J(5,6b) = 5.4 Hz, 1H; H-5A), 3.84 (ddd, J(4,5) = 3.1 Hz,
J(5,6a) = 6.8 Hz, J(5,6b) = 5.4 Hz, 0.8H, H-5B), 3.77 (ABX, J(5,6a) =
6.8 Hz, J(6a,6b) = 10.2 Hz, 0.8H; H-6aB), 3.73 (ABX, J(5,6a) = 5.4 Hz,
J(C,F)
135.08 (d, J(C,P)
=
41.9 Hz; CO-CF3), 135.16 (d, J(C,P)
6.1 Hz; Cq arom.), 128.71–127.72 (10CH arom.),
=
6.0 Hz; Cq arom.),
=
114.54 (q, J(C,F) = 284.1 Hz; CO-CF3), 90.88 (dd, J(1’,P) = 145.5 Hz,
J(1’,F) = 199.2 Hz; C-1’), 79.37 (C-2), 78.58 (C-4), 73.25 (C-3), 69.49 (d,
J(C,P) = 7.5 Hz; CH2Ph), 69.10 (d, J(C,P) = 6.8 Hz; CH2Ph), 68.74 (C-
5), 65.86 (C-6), 26.01 (Si-C(CH3)3), 25.79 (Si-C(CH3)3), 25.71 (Si-C-
(CH3)3), 25.67 (Si-C(CH3)3), 18.47 (Si-C(CH3)3), 18.20 (Si-C(CH3)3), 17.82
(Si-C(CH3)3), 17.61 (Si-C(CH3)3), ꢀ5.02 (Si-Me), ꢀ5.04 (Si-Me), ꢀ5.12
(2Si-Me), ꢀ5.20 (Si-Me), ꢀ5.34 (Si-Me), ꢀ5.47 (Si-Me), ꢀ5.75 (Si-Me);
31P NMR (101 MHz, CDCl3): d = 9.77 ppm (d, J(P,F1’) = 64.8 Hz); 19F
NMR (235 MHz, CDCl3): d = ꢀ74.55 (d, J(F,F1’) = 14.8 Hz; OCOCF3),
ꢀ210.17 ppm (ddd, J(F,F1’) = 14.8 Hz, J(F1’,H1’) = 45.8 Hz, J(F1’,P) =
64.6 Hz; F-1’); MS (DCI-NH3): m/z (%): 1025 (55) [M+H]+, 1042 (100)
[M+NH4]+; HRMS: m/z: calcd for C47H82O10F4Si4P: 1025.4659; found
1025.4646.
= 6.4 Hz, J(1’,F) =
(1(1’)Z)-1-Deoxy-1-(dibenzyloxyphosphoryl)fluoromethylidene-2,3,5,6-
tetra-O-tert-butyldimethylsilyl-d-galactofuranose ((Z)-9) and (1(1’)E)-1-
3120
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3114 – 3123

Synthesis and Inhibitory Effect on UGM of Fluorinated exo-Glycals
FULL PAPER
deoxy-1-(dibenzyloxyphosphoryl)fluoromethylidene-2,3,5,6-tetra-O-tert-
butyldimethylsilyl-d-galactofuranose ((E)-9): Compound 10 (826 mg,
0.89 mmol) was dissolved in a mixture of anhydrous CH2Cl2 and pyridine
(6:4, 25 mL) under argon atmosphere. The solution was cooled to 08C,
and then DMAP (2.18 g, 17.8 mmol) and trifluoroacetic anhydride
(1.2 mL, 8.9 mmol) were successively added. The reaction mixture was
stirred for 24 h at 358C, and then the solvents were evaporated under re-
duced pressure. The residue was diluted with CH2Cl2 (60 mL) and the or-
ganic layer was washed with water, dried over MgSO4, filtered, and con-
centrated. The residue was then purified by chromatography on silica gel
(cyclohexane/EtOAc, 9:1) to give (Z)-9 (381 mg, 46% yield) and (E)-9
(146 mg, 26% yield) in a ratio of 64:36 (72% overall yield).
to furnish monofluoride 14 (98 mg, 41%), (Z)-exo-glycal (Z)-8 (49 mg,
22%), and (E)-exo-glycal (E)-8 (24 mg, 11%), respectively.
1H NMR (400 MHz, CDCl3): d = 7.36 (m; 10H, ArH), 4.97–5.15 (m;
ArCH2), 4.56 (d, J(2,3)
= 5.5 Hz, J(2,F) = 13.7 Hz; H-2a), 4.39 (dt,
J(2,3) = 5.5 Hz, J(3,F) = 1.0 Hz; H-3a), 4.35 (d, J(2,F) = 5.4 Hz; H-2b),
4.09–4.14 (m; H-4a and H-4b), 4.07 (brs; H-3b), 3.86–3.90 (m; H-5a), 3.81–
3.90 (m; H-4b and H-5b), 3.59–3.74 (m; H-6a and H-6b), 2.51–2.74 (m; H-
1’a and H-1’b), 0.88–0.94 (m; MeSi), 0.03–0.20 ppm (m; tBuSi); 13C NMR
(100 MHz, CDCl3): d = 136.6”3, 136.5, 136.4”2, 136.3”2, 128.5, 128.4,
128.3, 128.2”3, 128.1, 128.0”3, 127.8, 120.4”2, 118.2”2, 115.4, 113.1,
82.3”2, 81.9”2, 81.0”2, 80.8”2, 78.5, 76.4, 74.2, 67.7, 67.6”3, 66.9,
66.8”3, 65.8, 64.5, 34.7, 34.3, 33.3, 32.9, 31.8, 31.6, 30.4, 30.1, 26.0”2,
25.9”3, 25.8”2, 25.6, 18.4, 18.3, 18.0, 17.9”2, 17.8, ꢀ3.7”2, ꢀ4.0”2,
ꢀ4.1”2, ꢀ4.2, ꢀ4.3”2, ꢀ4.6, ꢀ4.8, ꢀ5.2, ꢀ5.3, ꢀ5.4”2 ppm; 19F NMR
(235 MHz, CDCl3): d = ꢀ95.1 (dd, J(2,F) = 9.4 Hz, J(1’,F) = 23.5 Hz;
(Z)-9: [a]1D8 = ꢀ7.3 (c = 1.0 in CHCl3); 1H NMR (400 MHz, CDCl3): d
= 7.38–7.35 (m, 10H; H arom.), 5.17 (ABX, J(H,P) = 6.5 Hz, J(H,H) =
11.7 Hz, 2H; 2CHPh), 5.12 (ABX, J(H,P) = 7.7 Hz, J(H,H) = 11.7 Hz,
Fa), ꢀ99.0 ppm (td, J(2,F)
= 4.7 Hz, J(P,F) = 14.1 Hz, J(1’,F) =
1H; CHPh), 5.05 (ABX, J(H,P)
= 7.0 Hz, J(H,H) = 11.7 Hz, 1H;
23.5 Hz; Fb); 31P NMR (101 MHz, CDCl3): d = 24.95 (s, Pb), 23.96 ppm
(d, J(P,F) = 4.7 Hz; Pa).
CHPh), 4.78 (d, 1H; H-3), 4.41 (dd, J(3,4) = 1.6 Hz, J(4,5) = 9.1 Hz,
1H; H-4), 4.27 (d, J(2,3) = 0.7 Hz, 1H; H-2), 4.01 (td, J(4,5) = 9.1 Hz,
J(5,6a,b) = 4.0 Hz, 1H; H-5), 3.74 (AX, J(5,6a,b) = 4.0 Hz, 2H; H-6a
and H-6b), 0.94 (s, 9H; Si-tBu), 0.92 (s, 9H; Si-tBu), 0.89 (s, 9H; Si-tBu),
0.87 (s, 9H; Si-tBu), 0.20 (s, 3H; Si-Me), 0.15 (2s, 6H; 2Si-Me), 0.14 (s,
3H; Si-Me), 0.11 (s, 3H; Si-Me), 0.10 (s, 6H; 2Si-Me), 0.04 ppm (s, 3H;
Si-Me); 13C NMR (100 MHz, CDCl3): d = 160.85 (dd, J1-P = 21.6 Hz,
J1-F = 37.3 Hz; C-1), 136.03 (d, JC-P = 8.3 Hz; 2Cq arom.), 133.57 (dd,
J1’-P = 232.4 Hz, J1’-F = 241.4 Hz; C-1’), 128.53–127.61 (10C arom.), 95.91
(1(1’)E)-2,3,5,6-Tetra-O-TBDMS-1-(dibenzyloxyphosphoryl)methylidene-
d-galactofuranose ((E)-8): [a]D23 = ꢀ21.7 (c = 0.5 in CHCl3); 1H NMR
(400 MHz, CDCl3): d = 7.36 (m, 10H; ArH), 4.94–5.07 (m; ArCH2), 4.90
(d, J(1’,P) = 7.7 Hz, 1H; H-1’), 4.31 (d, J(4,5) = 9.8 Hz, 1H; H-4), 4.13
(brs, 1H; H-2), 4.10 (td, J(4,5) = 9.8 Hz, J(5,6) = 3.6 Hz, 1H; H-5), 3.75
(d, J(5,6) = 3.6 Hz, 2H; H-6), 0.89–0.94 (m; MeSi), 0.06–0.29 ppm (m;
tBuSi); 13C NMR (100 MHz, CDCl3): d = 176.1, 175.8, 128.8, 128.4”2,
128.1, 128.0, 93.4, 84.7, 82.7, 76.4, 75.8, 72.5, 66.6”2, 66.5, 66.3, 26.1,
25.9”2, 25.6, ꢀ4.5, ꢀ4.6”2, ꢀ4.7”2, ꢀ5.3 ppm; 31P NMR (101 MHz,
CDCl3): d = 21.51 ppm (s); MS (DCI-NH3): m/z: 893 [M+H]; HRMS
(DCI): m/z: calcd for C49H50O8P (MH+): 893.4834; found: 893.4824.
(C-4), 76.75 (C-3), 76.10 (C-2), 72.35 (C-5), 67.80 (d, JC-P
= 4.5 Hz;
CH2Ph), 67.71 (d, JC-P = 5.1 Hz; CH2Ph), 66.50 (C-6), 26.11 (Si-C(CH3)3),
25.89 (Si-C(CH3)3), 25.61 (Si-C(CH3)3), 25.55 (Si-C(CH3)3), 18.56 (Si-C-
(CH3)3), 18.07 (Si-C(CH3)3), 17.93 (Si-C(CH3)3), 17.75 (Si-C(CH3)3),
ꢀ4.55 (Si-Me), ꢀ4.66 (Si-Me), ꢀ4.70 (Si-Me), ꢀ4.88 (Si-Me), ꢀ5.14 (Si-
Me), ꢀ5.17 (Si-Me), ꢀ5.22 (Si-Me), ꢀ5.27 ppm (Si-Me); 31P NMR
((1(1’)Z)-1-Deoxy-1-methylidene-1’-fluoro-d-galactofuranosyl)phosphon-
ic acid ((Z)-15): A solution of compound (Z)-9 (297 mg, 0.33 mmol) in
anhydrous CH2Cl2 (22 mL) was stirred in the presence of Et3N (91 mL,
0.66 mmol, 2 equiv) and Pd/C (10%, 90 mg) for 1 h at room temperature
under H2 atmosphere (1 bar). The suspension was then filtered through a
pad of Celite and the filtrate was concentrated to quantitatively afford a
viscous oil (304 mg). Tetrabutylammonium fluoride (433 mg, 1.39 mmol,
4.2 equiv) was added to a solution of this compound in THF (15 mL) at
08C. After stirring for 1 h at room temperature, the reaction mixture was
concentrated under reduced pressure, and the residue was redissolved in
the minimum amount of water. This solution was filtered through paper,
and the aqueous filtrate was freeze-dried. The residue, redissolved in the
minimum amount of water, was applied to a short column of Dowex
50WX8-200 (Na+ form). The appropriate fractions (detected by TLC,
eluting with EtOH/NH4OH/H2O, 5:3:1) were pooled and freeze-dried.
The compound was then purified on Sephadex G15 (eluent: H2O) and
(101 MHz, CDCl3): d
= 6.81 ppm (d, J(P,F) =
94.2 Hz); 19F NMR
(235 MHz, CDCl3): d = ꢀ174.10 ppm (d, J(P,F) = 94.1 Hz); MS (DCI-
NH3): m/z (%): 911 (20) [M+H]+, 928 (100) [M+NH4]+; HRMS: m/z:
calcd for C45H81O8FSi4P: 911.4730; found 911.4722.
(E)-9: [a]1D8 = ꢀ40.4 (c = 1.0 in CHCl3); 1H NMR (400 MHz, CDCl3): d
= 7.41–7.38 (m, 10H; H arom.), 5.25 (s, 1H; H-3), 5.19 (ABX, J(H,P) =
6.8 Hz, J(H,H) = 11.9 Hz, 1H; CHPh), 5.15 (ABX, J(H,P) = 6.2 Hz,
J(H,H) = 11.7 Hz, 2H; 2CHPh), 5.04 (ABX, J(H,P) = 6.8 Hz, J(H,H)
= 11.7 Hz, 1H; CHPh), 4.47 (d, J(4,5) = 9.9 Hz, 1H; H-4), 4.23 (d,
J(2,3) = 3.3 Hz, 1H; H-2), 4.17 (td, J(4,5) = 9.9 Hz, J(5,6a,b) = 3.5 Hz,
1H; H-5), 3.81 (AX, J(5,6a,b) = 3.5 Hz, 2H; H-6a and H-6b), 0.98 (s,
9H; Si-tBu), 0.97 (s, 9H; Si-tBu), 0.94 (s, 9H; Si-tBu), 0.91 (s, 9H; Si-
tBu), 0.30 (s, 3H; Si-Me), 0.29 (s, 3H; Si-Me), 0.20 (s, 3H; Si-Me), 0.17
(s, 3H; Si-Me), 0.16 (2s, 6H; 2Si-Me), 0.15 (s, 3H; Si-Me), 0.10 ppm (s,
3H; Si-Me); 13C NMR (100 MHz, CDCl3): d = 159.15 (dd, J(1,P) =
4.4 Hz, J(1,F) = 48.4 Hz; C-1), 135.82 (d, J(C,P) = 7.4 Hz; Cq arom.),
135.79 (d, J(C,P) = 8.8 Hz; Cq arom.), 131.30 (dd, J(1’,P) = 247.6 Hz,
J(1’,F) = 254.9 Hz; C-1’), 128.45–127.67 (10C arom.), 94.52 (C-4), 77.27
(d, J(3,P) = 2.6 Hz; C-3), 76.73 (C-2), 72.30 (C-5), 67.90 (d, J(C,P) =
5.2 Hz; CH2Ph), 67.64 (d, J(C,P) = 4.2 Hz; CH2Ph), 64.00 (C-6), 26.10
(Si-C(CH3)3), 25.87 (Si-C(CH3)3), 25.80 (Si-C(CH3)3), 25.55 (Si-C(CH3)3),
18.56 (Si-C(CH3)3), 18.24 (Si-C(CH3)3), 17.95 (Si-C(CH3)3), 17.79 (Si-C-
(CH3)3), ꢀ4.65 (Si-Me), ꢀ4.70 (Si-Me), ꢀ4.73 (3Si-Me), ꢀ4.75 (Si-Me),
ꢀ5.27 ppm (2Si-Me); 31P NMR (101 MHz, CDCl3): d = 7.85 ppm (d,
J(P,F) = 78.0 Hz); 19F NMR (235 MHz, CDCl3): d = ꢀ163.69 ppm (d,
J(P,F) = 77.7 Hz); MS (DCI-NH3): m/z (%): 911 (100) [M+H]+, 928
(20) [M+NH4]+; HRMS: m/z: calcd for C45H81O8FSi4P: 911.4730; found
911.4732.
subsequently applied to
a
column of Dowex 50WX8-200 (Bu3NH+
form). The desired fractions were combined and freeze-dried to give (Z)-
15 (170 mg, 88% yield, 1.7 equivalents tributylammonium salt) as a white
hygroscopic solid.
[a]2D2 = ꢀ14.1 (c = 0.8 in H2O, 1.7 equiv Bu3N); 1H NMR (400 MHz,
D2O): d = 4.92 (m, 1H; H-2), 4.22 (t, J(2,3) = J(3,4) = 4.6 Hz, 1H; H-
3), 4.17 (t, J(3,4) = J(4,5) = 4.6 Hz, 1H; H-4), 3.86 (td, J(4,5) = J(5,6a)
= 4.6 Hz, J(5,6b) = 6.5 Hz, 1H; H-5), 3.73 (ABX, J(5,6a) = 4.6 Hz,
J(6a,6b) = 11.8 Hz, 1H; H-6a), 3.68 (ABX, J(5,6b) = 6.6 Hz, J(6a,6b)
= 11.8 Hz, 1H; H-6a), 3.11 (m; CH2(d), Bu3N), 1.65 (m; CH2(c), Bu3N),
1.35 (sext., J(H,H) = 7.5 Hz; CH2(b), Bu3N), 0.91 ppm (t, J(H,H) =
7.5 Hz; CH3(a), Bu3N); 13C NMR (100 MHz, D2O): d
= 153.10 (dd,
J(1,P) = 18.9 Hz, J(1,F) = 37.7 Hz; C-1), 142.93 (dd, J(1’,P) = 236.8 Hz,
J(1’,F) = 247.9 Hz; C-1’), 86.67 (C-4), 76.08 (C-3), 74.74 (d, J(2,P) =
5.5 Hz; C-2), 70.92 (C-5), 62.46 (C-6), 53.02 (CH2(d), Bu3N), 25.54
(CH2(c), Bu3N), 19.63 (CH2(b), Bu3N), 13.15 ppm (CH3(a), Bu3N); 31P
NMR (101 MHz, D2O): d = 0.16 ppm (d, J(P,F) = 85.1 Hz); 19F NMR
(235 MHz, D2O): d = ꢀ169.80 ppm (dd, J(P,F) = 84.7 Hz, J(F,H2) =
2.4 Hz); MS (FABꢀ): m/z: 273 [MꢀH]ꢀ; HRMS: m/z: calcd for
C7H11O8FP: 273.0176; found 273.0172.
Dibenzyl (1-deoxy-1-fluoro-2,3,5,6-tetra-O-TBDMS-d-galactofuranosyl)-
methyl phosphonate (14): DAST (41 mL, 0.312 mmol) was gradually
added to a cooled (08C) solution of lactol 10 (238 mg, 0.260 mmol) in
CH2Cl2 (3 mL), and the mixture was stirred for 1 h at 08C. After washing
with saturated aqueous NaHCO3 solution, the aqueous phase was ex-
tracted with CH2Cl2. The combined organic layers were washed with
brine, dried over MgSO4, and concentrated in vacuo. The residue was pu-
rified by column chromatography on silica gel (cyclohexane/EtOAc, 5:1)
UDP-(1(1’)Z)-1’-fluoro-exo-glycal-d-galactofuranose (4): A suspension of
5’-UMP (triethylammonium salt, 65.1 mg, 0.153 mmol) in a mixture of
freshly distilled MeCN (800 mL), N,N-dimethylaniline (77 mL, 0.61 mmol),
Chem. Eur. J. 2006, 12, 3114 – 3123
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3121

S. P. Vincent, P. Sinaþ et al.
and Et3N (43 mL, 0.31 mmol) was stirred under argon atmosphere at 08C.
Trifluoroacetic anhydride (128 mL, 0.92 mmol) was then slowly added
dropwise. The reaction mixture was stirred for a few minutes at room
temperature, after which time a red-brown coloration was observed.
Excess trifluoroacetic anhydride and trifluoroacetic acid were removed
from the reaction mixture under vacuum (using an oil pump and a liquid
nitrogen trap). In a separate flask, a mixture of N-methylimidazole
(61 mL, 0.76 mmol) and Et3N (128 mL, 0.92 mmol) in anhydrous MeCN
(150 mL) was prepared, cooled to 08C, and then added to the flask con-
taining the mixed phosphoryl anhydride. The reaction mixture was stirred
for 10 min at 08C, after which time a bright-yellow solution was obtained.
In the meantime, in another flask a mixture of (Z)-15 (tributylammonium
salt, 71.8 mg, 0.118 mmol) and preactivated 4 Š molecular sieves in
MeCN (800 mL) was stirred for 30 min at 08C. The solution of the UMP-
N-methylimidazolium species was then added dropwise to the solution
containing (Z)-15. The resulting mixture was stirred at 08C under argon
for 2 h, and then allowed to warm to room temperature. The reaction
was complete after 20 h, whereupon it was quenched with cold aqueous
ammonium formate (3 mL, 250 mm, pH 7). After filtration through a
Celite pad, the amines were extracted from the aqueous phase with
CH2Cl2 (3 mL). The organic phase was washed with cold aqueous ammo-
nium formate (2 mL, 250 mm, pH 7) and the combined aqueous phases
were pooled and freeze-dried. The residue was purified by size-exclusion
chromatography (Sephadex G15) eluting with a 50 mm triethylammoni-
um acetate buffer (pH 6.8). The appropriate fractions were pooled and
freeze-dried. Compound 4 was further purified by HPLC on a C18
column, using 50 mm triethylammonium acetate buffer (pH 6.8) contain-
ing 1% MeCN as eluent at a flow rate of 1 mLminꢀ1. This protocol af-
forded compound 4 with the (Z)-configuration as a white solid (triethyl-
ammonium salt) in 64% yield.
149.26 (dd, J(1,P) = 7.2 Hz, J(1,F) = 37.1 Hz; C-1), 141.38 (dd, J(1’,P)
= 216.7 Hz, J(1’,F) = 254.9 Hz; C-1’), 84.00 (C-4), 75.65 (C-3), 74.75 (C-
2), 70.43 (C-5), 62.76 (C-6), 52.96 (CH2(d), Bu3N), 25.52 (CH2(c), Bu3N),
19.65 (CH2(b), Bu3N), 13.00 ppm (CH3(a), Bu3N); 31P NMR (101 MHz,
D2O): d = 0.95 ppm (d, J(P,F) = 82.0 Hz); 19F NMR (235 MHz, D2O): d
= ꢀ155.43 ppm (d, J(P,F) = 82.4 Hz); MS (FABꢀ): m/z: 273 [MꢀH]ꢀ;
HRMS: m/z: calcd for C7H11O8FP: 273.0176; found 273.0186.
UDP-(1(1’)E)-1’-fluoro-exo-glycal-d-galactofuranose (5): The coupling
procedure was essentially identical to that described for compound 4.
The reaction time for the coupling of exo-glycal (E)-15 (62.7 mg,
0.12 mmol) with activated UMP (0.15 mmol) was 24 h at room tempera-
ture. The crude mixture was purified by size-exclusion chromatography
(Sephadex G15, eluent H2O). Compound 5 was further purified by
HPLC on a C18 column, eluting with 50 mm triethylammonium acetate
buffer (pH 6.8) containing 1% MeCN at a flow rate of 1 mLminꢀ1. This
protocol afforded compound 5 as a white solid (triethylammonium salt)
in 70% yield.
1H NMR (400 MHz, D2O): d = 7.97 (d, J(5’’,6’’) = 8.1 Hz, 1H; H-6’’),
6.01 (d, J(1’,2’) = 4.7 Hz, 1H; H-1’), 5.97 (d, J(5’’,6’’) = 8.1 Hz, 1H; H-
5’’), 5.01 (dd, J(2,3) = 1.7 Hz, J(2,P) = 2.9 Hz, 1H; H-2), 4.38–4.34 (m,
2H; H-2’ and H-3’), 4.32 (dd, J(2,3) = 1.7 Hz, J(3,4) = 1.7 Hz, 1H; H-
3), 4.30–4.27 (m, 2H; H-4 and H-4’), 4.23 (ABXX’, J(4’,5’a) = 4.3 Hz,
J(5’a,5’b)
= 11.9 Hz, J(5’a,P) = 2.5 Hz, 1H; H-5’a), 4.17 (ABXX’,
J(4’,5’b) = 5.2 Hz, J(5’a,5’b) = 11.9 Hz, J(5’b,P) = 2.5 Hz, 1H; H-5’b),
3.92 (td, J(4,5) = J(5,6a) = 4.4 Hz, J(5,6b) = 6.5 Hz, 1H; H-5), 3.76
(ABX, J(5,6a) = 4.4 Hz, J(6a,6b) = 11.9 Hz, 1H; H-6a), 3.70 (ABX,
J(5,6b) = 6.5 Hz, J(6a,6b) = 11.9 Hz, 1H; H-6b), 3.16 (q, J = 7.3 Hz;
CH2, Et3N), 1.23 ppm (t, J = 7.3 Hz, CH3, Et3N); 13C NMR (100 MHz,
D2O): d = 166.57 (C-4’’), 153.04 (dd, J(1,P) = 6.3 Hz, J(1,F) = 43.4 Hz;
C-1), 152.22 (C-2’’), 141.98 (C-6’’), 137.88 (dd, J(CF,P)
= 239.5 Hz,
1H NMR (400 MHz, D2O): d = 7.97 (d, J(5’’,6’’) = 8.2 Hz, 1H; H-6’’),
6.00 (d, J(1’,2’) = 4.6 Hz, 1H; H-1’), 5.98 (d, J(5’’,6’’) = 8.2 Hz, 1H; H-
5’’), 4.95 (dd, J(2,3) = 3.4 Hz, J(2,P) = 7.1 Hz, 1H; H-2), 4.39–4.36 (m,
2H; H-2’ and H-3’), 4.29–4.25 (m, 2H; H-3 and H-4’), 4.22 (t, J(3,4) =
J(4,5) = 4.8 Hz, 1H; H-4), 4.18 (ABXX’, J(4’,5’) = 5.6 Hz, J(5’a,5’b) =
J(CF,F) = 255.1 Hz; C-F), 103.07 (C-5’’), 88.57 (C-1’), 87.84 (C-4), 83.79
(d, J(4’,P) = 9.2 Hz; C-4’), 75.79 (C-3), 75.38 (C-2), 74.27 (C-3’), 70.93
(C-5), 70.27 (C-2’), 65.30 (d, J(5’,P) = 5.4 Hz; C-5’), 62.72 (C-6), 47.04
(CH2, Et3N), 8.61 ppm (CH3, Et3N); 31P NMR (101 MHz, D2O): d =
ꢀ7.84 (dd, J(Pa,Pb) = 23.3 Hz, J(Pa,F) = 84.0 Hz; Pa), ꢀ11.63 ppm (d,
J(Pa,Pb) = 22.3 Hz; Pb); 19F NMR (235 MHz, D2O): d = ꢀ156.70 ppm
(d, J(F,Pa) = 84.7 Hz); MS (ESI+): m/z: 603 [M+Na]+; HRMS: m/z:
calcd for C16H23O16FN2P2Na: 603.0405; found 603.0385.
9.0 Hz, J(5’,P)
4.8 Hz, J(5,6b)
=
2.6 Hz, 2H; H-5’a,b), 3.88 (td, J(4,5)
6.7 Hz, 1H; H-5), 3.77 (ABX, J(5,6a)
=
J(5,6a)
= 4.8 Hz,
=
=
J(6a,6b) = 11.8 Hz, 1H; H-6a), 3.71 (ABX, J(5,6b) = 6.7 Hz, J(6a,6b)
= 11.8 Hz, 1H; H-6b), 3.21 (q, J = 7.3 Hz; CH2, Et3N), 1.29 ppm (t, J =
7.3 Hz, CH3, Et3N); 13C NMR (100 MHz, D2O): d
=
166.61 (C-4’’),
36.9 Hz; C-1), 152.22 (C-2’’),
232.9 Hz, J(CF,F) 239.7 Hz;
154.96 (dd, J(1,P)
= 20.6 Hz, J(1,F) =
142.01 (C-6’’), 140.60 (dd, J(CF,P)
=
=
C-F), 103.05 (C-5’’), 88.61 (C-1’), 86.88 (C-4), 83.73 (d, J(4’,P) = 9.1 Hz;
C-4’), 76.11 (d, J(2,P) = 7.1 Hz, C-2), 75.99 (C-3), 74.17 (C-3’), 70.97
(C-5), 70.10 (C-2’), 65.18 (d, J(5’,P) = 5.5 Hz, C-5’), 62.46 (C-6), 47.04
(CH2, Et3N), 8.60 ppm (CH3, Et3N); 31P NMR (101 MHz, D2O): d =
ꢀ8.87 (dd, J(Pa,Pb) = 24.3 Hz, J(Pa,F) = 90.1 Hz; Pa), ꢀ11.69 ppm (d,
J(Pa,Pb) = 25.3 Hz; Pb); 19F NMR (235 MHz, D2O): d = ꢀ172.06 ppm
Acknowledgements
This work was supported by the Centre National de la Recherche Scienti-
fique (CNRS), by the Minist›re DØlØguØ à la Recherche et aux Nouvelles
Technologies (Ph.D. grant to A.C.), and by the Japan Society for the Pro-
motion of Science (JSPS, Research Fellowships for Young Scientists
(PD) to H.D.). We warmly thank Dr. Didier Blanot (IBBMC, UMR-8619
du CNRS, UniversitØ Paris-Sud) for the mass spectrometric analyses.
(dd, J(F,Pa)
= 91.8 Hz, J(F,H2) =
2.35 Hz); MS (ESI+): m/z: 603
[M+Na]+; HRMS: m/z: calcd for C16H23O16FN2P2Na: 603.0405; found
603.0432.
((1(1’)E)-1-Deoxy-1-methylidene-1’-fluoro-exo-glycal-d-galactofurano-
syl)phosphonic acid ((E)-15): Compound (E)-9 (294 mg, 0.32 mmol) was
hydrogenated (reaction time: 3 h) and desilylated (reaction time: 1.5 h)
following the same procedures as used for the preparation of compound
(Z)-15. After purification of the crude reaction mixture, first by size-ex-
clusion chromatography (Sephadex G15, eluent H2O) and then by ion-ex-
change chromatography (Dowex Na+ and Bu3NH+ forms), compound
(Z)-15 (166 mg, 80% yield, bis(tributylammonium) salt) was obtained as
a hygroscopic white solid.
[1]A. Weston, R. J. Stern, R. E. Lee, P. M. Nassau, D. Monsey, S. L.
Martin, M. S. Scherman, G. S. Besra, K. Duncan, M. R. McNeil,
Tuber. Lung Dis. 1998, 78, 123–131.
[2]J. B. Houseknecht, T. Lowary, Curr. Opin. Chem. Biol. 2001, 5, 677–
682.
[3]L. Kremer, G. S. Besra, Expert Opin. Invest. Drugs 2002, 11, 1033–
1049.
[a]2D2 = ꢀ19.5 (c = 0.77 in H2O, 2 equiv Bu3N); 1H NMR (400 MHz,
[4]L. Kremer, L. G. Dover, C. Morehouse, P. Hitchin, M. Everett,
H. R. Morris, A. Dell, P. J. Brennan, M. R. McNeil, C. Flaherty, K.
Duncan, G. S. Besra, J. Biol. Chem. 2001, 276, 26430–26440.
[5]L. L. Pedersen, S. J. Turco, Cell. Mol. Life Sci. 2003, 60, 259–266.
[6]K. Mikusova, T. Yagi, R. Stern, M. R. McNeil, G. S. Besra, D. C.
Crick, P. J. Brennan, J. Biol. Chem. 2000, 275, 33890–33897.
[7]M. S. Scherman, K. A. Winans, R. J. Stern, V. Jones, C. R. Bertozzi,
M. R. McNeil, Antimicrob. Agents Chemother. 2003, 47, 378–382.
D2O): d = 4.81 (m, 1H; H-2), 4.25 (t, J(2,3) = J(3,4) = 6.6 Hz, 1H; H-
3), 4.08 (dd, J(3,4)
= 6.6 Hz, J(4,5) = 4.0 Hz, 1H; H-4), 3.85 (ddd,
J(5,6a) = 4.9 Hz, J(5,6b) = 7.1 Hz, 1H; H-5), 3.72 (ABX, J(5,6a) =
4.9 Hz, J(6a,6b) = 11.8 Hz, 1H; H-6a), 3.67 (ABX, J(5,6b) = 7.1 Hz,
J(6a,6b)
= 11.8 Hz, 1H; H-6a), 3.09 (m; CH2(d), Bu3N), 1.64 (m;
CH2(c), Bu3N), 1.34 (sext, J(H,H) = 7.3 Hz; CH2(b), Bu3N), 0.90 ppm (t,
J(H,H)
= =
7.3 Hz; CH3(a), Bu3N); 13C NMR (100 MHz, D2O): d
3122
www.chemeurj.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2006, 12, 3114 – 3123

Synthesis and Inhibitory Effect on UGM of Fluorinated exo-Glycals
FULL PAPER
[8]S. W. B. Fullerton, S. Daff, D. A. R. Sanders, W. J. Ingledew, C. Whit-
field, S. K. Chapman, J. H. Naismith, Biochemistry 2003, 42, 2104–
2109.
[21]P. T. Nyffeler, S. G. Duron, M. D. Burkart, S. P. Vincent, C.-H.
Wong, Angew. Chem. 2005, 117, 196–217; Angew. Chem. Int. Ed.
2005, 44, 192–212.
[9]Z. Huang, Q. Zhang, H.-w. Liu, Bioorg. Chem. 2003, 31, 494–502.
[10]S. Bornemann, Nat. Prod. Rep. 2002, 19, 761–772.
[11]J. N. Barlow, M. E. Girvin, J. S. Blanchard, J. Am. Chem. Soc. 1999,
121, 6968–6969.
[22]S. P. Vincent, M. D. Burkart, C.-Y. Tsai, Z. Zhang, C.-H. Wong, J.
Org. Chem. 1999, 64, 5264–5279.
[23]W.-B. Yang, C.-F. Chang, S.-H. Wang, C.-F. Teo, C.-H. Lin, Tetrahe-
dron Lett. 2001, 42, 4657–4660.
[12]J. N. Barlow, J. S. Blanchard, Carbohydr. Res. 2000, 328, 473–480.
[13]M. Soltero-Higgin, E. E. Carlson, T. D. Gruber, L. L. Kiessling, Nat.
Struct. Mol. Biol. 2004, 11, 539–543.
[14]A. Caravano, D. Mengin-Lecreulx, J.-M. Brondello, S. P. Vincent, P.
Sinaþ, Chem. Eur. J. 2003, 9, 5888–5898.
[24]W.-B. Yang, C.-Y. Wu, C.-C. Chang, S.-H. Wang, C.-F. Teo, C.-H.
Lin, Tetrahedron Lett. 2001, 42, 6907–6910.
[25]C. Taillefumier, Y. Chapleur, Chem. Rev. 2004, 104, 263–292.
[26]M. D. Burkart, S. P. Vincent, C.-H. Wong, Chem. Commun. 1999,
1525–1526.
[15]K. Beis, V. Srikannathasan, H. Liu, S. W. B. Fullerton, V. A. Bam-
ford, D. A. R. Sanders, C. Whitfield, M. R. McNeil, J. H. Naismith,
J. Mol. Biol. 2005, 348, 971–982.
[16]Q. Zhang, H.-W. Liu, J. Am. Chem. Soc. 2001, 123, 6756–6766.
[17]A. Caravano, S. P. Vincent, P. Sinaþ, Chem. Commun. 2004, 1216–
1217.
[27]M. D. Burkart, Z. Zhang, S.-C. Hung, C.-H. Wong, J. Am. Chem.
Soc. 1997, 119, 11743–11746.
[28]W.-B. Yang, Y.-Y. Yang, Y.-F. Gu, S.-H. Wang, C.-F. Chang, C.-H.
Lin, J. Org. Chem. 2002, 67, 3773–3782.
[29]M. Michalik, M. Hein, M. Frank, Carbohydr. Res. 2000, 327, 185–
218.
[18]S. G. Withers, Carbohydr. Polym. 2001, 44, 325–337.
[19]For a recent review on the elucidation of enzyme mechanisms using
fluorinated substrate analogues, see R. Pongdee, H.-W. Liu, Bioorg.
Chem. 2004, 32, 393–437.
[30]V. S. Bogachev, Russ. J. Bioorg. Chem. 1996, 22, 599–604.
[31]A. L. Marlow, L. L. Kiessling, Org. Lett. 2001, 3, 2517–2519.
[32]Q. Zhang, H.-w. Liu, J. Am. Chem. Soc. 2000, 122, 9065–9070.
[20]B. Müller, C. Schaub, R. R. Schmidt, Angew. Chem. 1998, 110,
3021–3024; Angew. Chem. Int. Ed. 1998, 37, 2893–2896.
Received: August 15, 2005
Published online: January 23, 2006
Chem. Eur. J. 2006, 12, 3114 – 3123
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
3123