Enzymatic synthesis of UDP-(3-deoxy-3-fluoro)-D-galactose and UDP-(2-deoxy-2-fluoro)-D-galactose and substrate activity with UDP-galactopyranose mutase

Article abstract of DOI:10.1016/S0008-6215(00)00135-X

The novel UDP-sugar uridine 5'-(3-deoxy-3-fluoro-D-galactopyranosyl diphosphate) (1) and UDP-(2-deoxy-2-fluoro)-D-galactose (2) have been prepared enzymatically and tested as substrate analogues for the enzyme UDP-galactopyranose mutase (UDP-Galp mutase EC 5.4.99.9). Turnover of both 1 and 2 by UDP-Galp mutase was observed by HPLC and ¹⁹F NMR. The HPLC elution profile and ¹⁹F chemical shift of the products are consistent with the formation of the predicted furanose forms of 1 and 2. The K(m) values for compounds 1 and 2 were similar to those of the natural substrate UDP-Galp (0.26 mM for 1, 0.2 mM for 2, and 0.6 mM for UDP-Galp), but the values for k(cat) were substantially different (1.6/min for 1, 0.02/min for 2, and 1364/min for UDP-Galp). A correlation was also observed between the equilibrium yield of product formed during turnover of UDP-sugar by UDP-Galp mutase (UDP-Galp, compound 1 or compound 2), and the amount of furanose present for the free sugar at thermal equilibrium in aqueous solution, using ¹H and ¹⁹F NMR spectroscopy. The implications of these results to the mechanism of the unusual enzymatic reaction are discussed. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(00)00135-X

www.elsevier.nl/locate/carres
Carbohydrate Research 328 (2000) 473–480
Enzymatic synthesis of
UDP-(3-deoxy-3-fluoro)-
D-galactose and
UDP-(2-deoxy-2-fluoro)-
D-galactose and substrate
activity with UDP-galactopyranose mutase
John N. Barlow, John S. Blanchard *
Department of Biochemistry, Albert Einstein College of Medicine, 1300 Morris Park A6enue, Bronx,
NY 10461, USA
Received 27 December 1999; received in revised form 7 March 2000; accepted 20 March 2000
Abstract
The novel UDP-sugar uridine 5%-(3-deoxy-3-fluoro-
D
-galactopyranosyl diphosphate) (1) and UDP-(2-deoxy-2-
fluoro)- -galactose (2) have been prepared enzymatically and tested as substrate analogues for the enzyme
D
UDP-galactopyranose mutase (UDP-Galp mutase EC 5.4.99.9). Turnover of both 1 and 2 by UDP-Galp mutase was
observed by HPLC and 19F NMR. The HPLC elution profile and 19F chemical shift of the products are consistent
with the formation of the predicted furanose forms of 1 and 2. The Km values for compounds 1 and 2 were similar
to those of the natural substrate UDP-Galp (0.26 mM for 1, 0.2 mM for 2, and 0.6 mM for UDP-Galp), but the
values for kcat were substantially different (1.6/min for 1, 0.02/min for 2, and 1364/min for UDP-Galp). A correlation
was also observed between the equilibrium yield of product formed during turnover of UDP-sugar by UDP-Galp
mutase (UDP-Galp, compound 1 or compound 2), and the amount of furanose present for the free sugar at thermal
1
equilibrium in aqueous solution, using H and 19F NMR spectroscopy. The implications of these results to the
mechanism of the unusual enzymatic reaction are discussed. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: UDP-sugars; Fluoro-sugars; Tautomerization; UDP-galactopyranose mutase; Uridine 5%-(3-deoxy-3-fluoro-D-galac-
topyranosyl diphosphate); UDP-(2-deoxy-2-fluoro)-galactose
1. Introduction
UDP-galactopyranose mutase (UDP-Galp
mutase) catalyzes the interconversion of the
a-pyranose and a-furanose forms of UDP-
galactose (Scheme 1) [1–3]. In vivo, UDP-
galactofuranose (UDP-Galf ) functions as the
activated precursor of Galf in the biosynthesis
of a number of macromolecular structures
found in prokaryotes and lower eukaryotes.
These include the arabinogalactan layer of
mycobacterial cell walls, and the outer leaflet
lipopolysaccharide (LPS) of gram negative
Abbre6iations: UDP-Galp, uridine 5%-(a-D-galactopyranosyl
diphosphate); HPLC, high-performance liquid chromatogra-
phy; NMR, nuclear magnetic resonance spectroscopy;
NADP+, nicotinamide adenine dinucleotide phosphate; ES-
IMS, electrospray ionization mass spectrometry; ATP,
adenosine 5%-triphosphate; DTT, dithiothreitol; FAD,
flavin adenine dinucleotide; AMPSO, (3-[(1,1-dimethyl-2-
hydroxyethyl)amino]-2-hydroxypropanesulfonic acid; 3F-Gal,
3-deoxy-3-fluoro-
galactose.
D-galactose; 2F-Gal, 2-deoxy-2-fluoro-D-
* Corresponding author. Tel.: +1-718-4303096; fax: +1-
718-4308565.
E-mail address: blanchar@aecom.yu.edu (J.S. Blanchard).
0008-6215/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(00)00135-X

474
J.N. Barlow, J.S. Blanchard / Carbohydrate Research 328 (2000) 473–480
bacteria [4,5], both of which play important
roles in the virulence of these organisms. This
occurrence of Galf, together with the absence
of Galf in higher eukaryotes, make UDP-Galp
mutase a potential drug target.
back to the anomeric hydroxyl, since this se-
quence would not account for C-1¦–O-1 bond
cleavage. Unlike other enzymes which catalyze
glycosyl transferase-type chemistry, UDP-
Galp mutase contains bound FAD, making
mechanisms invoking cryptic redox chemistry
a possibility.
The interconversion of UDP-Galp and
UDP-Galf is unprecedented. It has recently
been demonstrated that the anomeric C-1¦–O-
1 bond is cleaved during conversion of UDP-
Galp to UDP-Galf [6]. This observation is
consistent with glycosyltransferase-type chem-
istry occurring at the active site of UDP-Galp
mutase, in this case, intramolecular transfer of
the anomeric carbon of galactose between the
C-5¦ and C-4¦ oxygens [7]. These results are
inconsistent with a mechanism involving UDP
transfer to the C-2¦ hydroxyl, followed by
anomeric equilibration and transfer of UDP
We report here the enzymatic synthesis of
the novel compound UDP-(3-deoxy-3-fluoro-
D-galactopyranose (1) using a variation of the
method used previously to synthesize UDP-(2-
deoxy-2-fluoro)-D-galactopyranose (2) [9]. De-
oxyfluoro sugars have been widely used as
probes of glycosidase mechanisms [8]. The
much slower reactions of 2-deoxy-2-fluoro
sugar analogues in glycosidase-catalyzed
chemistry has been proposed to be due to a
large inductive electronic effect of the 2-fluoro
substituent that destabilizes oxocarbenium ion
intermediates and transition states [8]. Com-
pounds 1 and 2 were synthesized as mechanis-
tic probes of the reaction catalyzed by
UDP-Galp mutase. We also measured the
amounts of furanose present at thermal equi-
librium in aqueous solutions of 3-deoxy-3-
fluoro-
tose and
D
-galactose, 2-deoxy-2-fluoro-
D
-galac-
D
-galactose. The results support the
previously proposed mechanism for UDP-
Galp mutase [6] that invokes anomeric CꢀO
bond cleavage during turnover of UDP-Galp.
Scheme 1.
2. Results and discussion
The synthesis of 1 and 2 was performed in
two steps using a variation of the protocol
published by Hayashi et al. [9] (Scheme 2). In
the first step, phosphorylation of the sugar
was achieved using galactokinase [10]. The
second step, uridylation to give the sugar
nucleotide, was performed using galactose-1-
phosphate uridyl transferase and UDP-glu-
cose. We chose to couple this second step
(Keq =1.1, pH 8.7, 37 °C [11]) to the phospho-
glucomutase-catalyzed isomerization of
D
-
glucose 1-phosphate to -glucose 6-phosphate
D
to drive the reaction forward, instead of
the UDP-glucose pyrophosphorylase-cata-
lyzed regeneration of UDP-glucose as re-
ported previously [9]. This allowed complete
removal of added UDP-glucose and thus
Scheme 2.

J.N. Barlow, J.S. Blanchard / Carbohydrate Research 328 (2000) 473–480
475
(2.2%) was significantly less than that ob-
served with UDP-Galp (7.7%). One possible
explanation for this result was the formation
of a stable, covalent adduct between these
fluoro-sugar nucleotides and the enzyme, since
similar sugars have been shown to be mecha-
nism-based inhibitors of other glycosidases [7].
Addition of UDP-Galp to the equilibrium
mixture resulting from incubation of 1 with
UDP-Galp mutase resulted in 7% conversion
of UDP-Galp to UDP-Galf, demonstrating
that the UDP-Galp mutase was still active.
The activity of UDP-Galp mutase after incu-
bation with 2 could not be determined in this
way because of co-elution of 2 with UDP-Galf
under our HPLC conditions. The steady-state
kinetic parameters, kcat and Km, for 1 and 2
are shown for comparison with those of UDP-
Galp in Table 1.
Fig. 1. Time course of product formation as detected by
HPLC during turnover of 1 (closed circles) and 2 (open
circles) by UDP-Galp mutase.
Table 1
Steady-state kinetic parameters for UDP-Galp mutase
Further support for the formation of the
predicted furanoses was obtained by 19F
NMR spectroscopy. Upon incubation of 1
with UDP-Galp mutase, a new resonance at
l= −202.9 ppm was observed, with an inte-
grated intensity ꢀ1% that of the resonance
from 1 at l= −203.77 ppm, corresponding
to the amount of furanose product determined
by HPLC (1.3%).
The low amounts of furanose product pro-
duced from 1 and 2 by UDP-Galp mutase
prompted us to investigate whether similarly
low levels of furanose forms are present in
aqueous solutions of the corresponding free
sugars at thermal equilibrium. The distribu-
tion of the pyranose and furanose anomers of
Compound
kcat (per min)
Km (mM)
UDP-Galp
1
2
1400
1.6
0.02
645
261
199
avoided problems in separating 1 and 2 from
UDP-glucose. However, a sub-stoichiometric
quantity of UDP-glucose must necessarily be
used with resulting lower yields.
Conversion of 1 and 2 into the predicted
furanose products, uridine 5%-(3-deoxy-3-
fluoro-
D
-galactopyranosyl diphosphate) and
uridine
5%-(2-deoxy-2-fluoro- -galactopyra-
D
nosyl diphosphate), respectively, was per-
formed enzymatically using UDP-Galp
mutase under optimized assay conditions
(Scheme 1). The substrate pyranose and
product furanose forms of the natural sub-
strate of UDP-Galp mutase, UDP-galactose,
are well separated by analytical PA-100 anion-
exchange HPLC. We expected a similar elu-
tion profile for the pyranose and furanose
forms of 1 and 2. Indeed, analysis of the
reaction of 1 and 2 with UDP-Galp mutase by
PA-100 HPLC revealed the time-dependent
formation of peaks eluting at positions ex-
pected for the furanose forms. Integration of
the peak height permitted the time courses of
the reaction to be determined (Fig. 1). Surpris-
ingly, the final amount of furanose product at
enzymatic equilibrium from 1 (1.3%) or 2
D
-galactose at equilibrium has previously been
determined by gas–liquid chromatography
1
and H NMR [12,13]. At 35 °C, 7.0% of
aqueous
D-galactose exists in the a and b
furanose forms at equilibrium [12]. We used a
1
combination of H (1D and 2D COSY) and
19F NMR spectroscopies to quantitate the rel-
ative amounts of furanose at 37 °C in aqueous
solutions of
galactose and 2-deoxy-2-fluoro-
D
-galactose, 3-deoxy-3-fluoro-
D
-
D
-galactose
(Table 2). Identification of the NMR reso-
nances was made using UDP-galactose, 1, and
2 as references for the a-pyranose tautomers,
and using criteria discussed in Ref. [12] (for
example, the b anomers of galactopyranose
and galactofuranose are more stable than the

476
J.N. Barlow, J.S. Blanchard / Carbohydrate Research 328 (2000) 473–480
of furanose form at equilibrium for 3-deoxy-3-
fluoro- -galactose and 2-deoxy-2-fluoro-
galactose are indeed significantly less than the
corresponding a anomers at 37 °C). The rela-
tive amounts of each tautomer are shown in
Table 3. It is evident that the total percentage
D
D-
Table 2
NMR assignments for 3-deoxy-3-fluoro-
D
-galactose, 2-deoxy-2-fluoro-
D
-galactose, and D-galactose
a-Pyranose
b-Pyranose
a-Furanose
b-Furanose
a
3-Deoxy-3-fluoro-D-galactose
H-1
H-2
l=5.45 (t b, J 4.58 Hz)
4.75 (d, J 7.9 Hz)
3.91 (m)
5.50 (d, J 4.8 Hz)
5.49 (s)
4.23 (m)
4.55 (dt b, J 21.2, 5.5 4.44 (d, J 16.7 Hz)
Hz)
H-3
4.91 (ddd, J 49.1, 10.0,
3.4 Hz)
4.71 (ddd, J 48.1, 9.5, 3.7
Hz)
4.35, (dd, J 6.2, 3.7 Hz)
3.84 (m)
5.21 (dt b, J 56.0, 5.2 5.12 (ddd, J 53.0, 4.0, 2.2
Hz)
Hz)
H-4
H-5
4.41 (dd, J 7.5, 3.5 Hz)
4.23 (m)
4.21 (m)
4.53 (dt b, J 24.3, 4.5 Hz)
4.02 (dd, J 24.1, 9.0 Hz)
3.8 (m)
c
c
H-6a+H-6b 3.89 (m)
3.90 (m)
2-Deoxy-2-fluoro- -galactose
D
H-1
H-2
5.60 (d, J 4.2 Hz)
4.81 (ddd, J 50.1, 10.1,
4.0 Hz)
4.97 (dd, J 7.8, 3.5 Hz)
4.47 (ddd, J 51.9, 9.4, 7.9
Hz)
5.55 (s)
5.64 (d, J 11.4 Hz)
5.08 (dt b, J 52.4, 5.3 5.02 (dd, J 51.1, 2 Hz)
Hz)
H-3
4.24 (m)
4.07 (ddd, J 14.3, 9.5, 3.6
Hz)
4.58 (ddt b, J 19.6, 6.0 4.48 (d, J 22 Hz)
Hz)
c
H-4
H-5
4.19 (t b, J 3.7 Hz)
4.25 (m)
4.13 (t b, J 3.3 Hz)
3.88 (m)
3.92 (s)
c
c
c
c
H-6a+H-6b 3.86 (m)
3.90 (m)
D
-Galactose
H-1
H-2
H-3
H-4
H-5
5.40 (d, J 3.9 Hz)
4.71 (d, J 8.0 Hz)
3.63 (dd, J 9.8, 7.9 Hz)
3.78 (dd, J 9.8, 3.5 Hz)
4.06 (d, J 3.4 Hz)
3.84 (m)
5.42 (d, J 4.7 Hz)
4.21 (m)
5.36 (d, J 3.1 Hz)
3.95 (dd, J 10.3, 3.8 Hz)
3.98 (dd, J 10.3, 3.2 Hz)
4.12 (d, J 6.0 Hz)
4.14 (m)
c
3.94 (m)
c
c
c
c
4.21 (t b, J 6.0 Hz)
c
c
H-6a+H-6b 3.86 (m)
3.88 (m)
a Assignments made from H, H{1H} COSY and 19F NMR.
b Apparent triplet.
1
1
c Not assigned because peaks overlapping.
Table 3
Anomeric distributions (%) at thermal equilibrium in aqueous solution a
a-Pyranose
b-Pyranose
a-Furanose
b-Furanose
% Furanose
3F-Gal b
2F-Gal
40 (40) c
41.0 (97) d
31.8
58 (58)
55.7
60.5
0.7 (0.8)
1.0 (0.8)
3.1
1.6 (1.8)
2.2 (2.3)
4.6
2.4 (2.6)
3.2 (3.1)
7.7
D
-Galactose
a Anomeric distribution calculated as follows: (2F-Gal)% a-pyranose and b-pyranose calculated from H-1¦ resonance areas, %
b-furanose and a-furanose calculated from mean of H-1¦ and H-2¦ resonance areas. Area of H-1¦ and H-2¦ of a-furanose are less
accurate because they were not fully resolved (H-2¦ overlaps with isotopic sideband of H-1¦ of b-pyranose) and calculated value
represents upper estimate; (3F-Gal)% a-pyranose and b-pyranose calculated from H-4¦ resonance areas, % b-furanose and
a-furanose calculated from mean of H-1¦ and H-3¦ resonance areas. The H-3¦ resonances were not fully resolved and the
calculated values represent upper estimates; (Galactose)% a-pyranose and b-pyranose calculated from H-4¦ resonance areas, %
b-furanose and a-furanose calculated from H-1¦ resonance areas.
b 3F-Gal, 3-deoxy-3-fluoro-
D-galactose; 2F-Gal, 2-deoxy-2-fluoro-D-galactose.
1
c Figures in parentheses determined from 19F NMR spectra (without H decoupling).
d Total% area of pyranose determined by 19F NMR (a-pyranose and b-pyranose resonances were partially overlapping).

J.N. Barlow, J.S. Blanchard / Carbohydrate Research 328 (2000) 473–480
477
Scheme 3.
reduction in catalytic turnover has been well
documented in the glycosidase-mediated hy-
drolysis of pyranosides containing fluorine at
C-2 or C-5 [7]. The presence of the electroneg-
ative fluorine atom adjacent to the anomeric
carbon has been proposed to destabilize oxo-
carbenium type intermediates or transition
states. In, addition, strong stabilizing interac-
tions between the C-2 hydroxyl group of sug-
ars and enzyme residues have been observed
for retaining glycosidases [15].
A number of mechanisms for the unprece-
dented reaction of UDP-Galp mutase can be
ruled out from these studies and our previ-
ously reported studies [6]. Cleavage of the
C-1¦–O-1 bond must occur during catalysis,
ruling out mechanisms involving enzyme-
UDP intermediates or UDP transfer from C-
1¦ to C-2¦. Mechanisms involving dehydration
or oxidation at C-2¦ or C-3¦ are also ruled out
by the present studies, since HF elimination
was not observed for either 1 or 2, and both 1
and 2 are shown here to be substrates for the
mutase. Two mechanisms that are consistent
with the data presented here are shown in
Scheme 3. Both invoke anomeric CꢀO bond
cleavage in the first step, as required from
previous studies [6]. In pathway (a), CꢀO
bond cleavage takes place during nucleophilic
attack by an enzyme-based nucleophile, remi-
niscent of a retaining-type glycosidase. In the
second step, ring opening takes place to give
an acyclic intermediate which closes in a third
step to give a galactofuranosyl–enzyme ad-
duct. In the final step, nucleophilic substitu-
tion of the enzyme-based nucleophile at the
percentage of furanose form observed for
galactose. In addition, the percentage of fura-
nose forms observed for 3-deoxy-3-fluoro-
D
-
D
-galactose and 2-deoxy-2-fluoro- -galactose
D
is similar to the percentage of furanose
product observed after enzymatic turnover of
1 and 2 by UDP-Galp mutase.
The substitution of an alcohol group with
fluorine in a substrate has been shown to alter
the nature of an enzyme-catalyzed reaction in
several ways, for example disruption of hydro-
gen bonds, suppression of aldol chemistry and
re-routing of the reaction to give net HF
elimination [8]. In the case of UDP-Galp mu-
tase, no fluoride release was observed by 19F
NMR for either 1 or 2. The similar Km values
for UDP-Galp and compounds 1 and 2 also
suggests that interactions in the Michaelis
complexes are not significantly affected. The
reason for the ꢀ1000-fold decrease in the
observed kcat exhibited by 1 is not apparent.
Interestingly, a \100-fold decrease in kcat has
been reported for the b-(1“4)-galactosyl-
transferase with 2-acetamido-2,3-dideoxy-3-
fluoro-glucopyranose as acceptor, although no
explanation for this was reported [14]. Any
possible mechanisms involving redox or aldol
chemistry at the C-3¦ position of UDP-Galp
during turnover by UDP-Galp mutase are
ruled out by the demonstration of turnover of
1 at a significant rate.
Compound 2 was a much poorer substrate,
exhibiting a kcat reduced ꢀ71,000-fold com-
pared to UDP-Galp. We propose that this
decrease in kcat for 2 results from the suppres-
sion of anomeric CꢀO bond cleavage. This

478
J.N. Barlow, J.S. Blanchard / Carbohydrate Research 328 (2000) 473–480
anomeric carbon by UDP gives the product
a-UDP-Galf. In pathway (b), anomeric CꢀO
bond cleavage takes place during nucleophilic
attack by the C-4¦ oxygen to give 1,4-anhy-
drogalactose and UDP. This intermediate sub-
sequently breaks down to give an enzyme-
stabilized oxocarbenium intermediate, which
reacts with UDP to give product.
tarate (20 mmol, 3.8 mg),
D
-glucose 1,6-
biphosphate (0.04 mmol, 0.03 mg), 10 units
galactose-1-phosphate uridyltransferase, 1.6
units phosphoglucomutase, 3 units glucose-6-
phosphate dehydrogenase and 1.4 units -glu-
tamic dehydrogenase. Uridylation was
monitored by ion exchange HPLC and by H
L
1
NMR spectroscopy (described below). Con-
version (100%) of UDP-glucose into 1 was
observed after 48 h at 37 °C. The sample was
then filtered through a YM-10 membrane (10
kDa cut-off, Millipore-Amicon), diluted in
several volumes of water and applied to a
Mono-Q 16/10 anion-exchange column (Phar-
macia), washed previously with 1 M KCl and
then pre-equilibrated in water. A linear 0–300
mM KCl gradient over 1 h was run with
elution of 1 at ꢀ75 mM KCl, with optical
detection at 262 nm. Fractions containing 1
were pooled, lyophilized, resuspended in a
small volume of water and desalted on a
Biogel P-2 gel (Bio-Rad) filtration column.
Fractions containing 1 were then pooled,
3. Experimental
General methods.—All chemicals and en-
zymes were purchased from Sigma–Aldrich,
unless otherwise specified. 3-Deoxy-3-fluoro-
D
-galactose was purchased from Omicron Bio-
chemicals, Inc. (IN, USA). 2-Deoxy-2-fluoro-
-galactose was purchased from Toronto Re-
D
search Chemicals Inc. NMR spectra were run
on a Bruker DRX300 spectrometer and
Bruker DRX600 in 5-mm symmetrical micro-
tubes (D2O matched, Shigemi Co., Ltd.). Re-
combinant UDP-Galp mutase from Klebsiella
pneumoniae was purified from E. coli. Details
of the cloning, expression, and purification of
the enzyme will be published separately.
1
lyophilized and stored at −20 °C. H NMR
(D2O, 600 MHz): l 8.07 (d, 1 H, J 8.1 Hz,
H-6), 6.12 (d, 1 H, J 3.7 Hz, H-1%), 6.11 (d, 1
H, J 8.1 Hz, H-5), 5.83 (dt, 1 H, J 6.8, 4.2 Hz,
H-1¦), 4.95 (ddd, 1 H, J 49.0, 9.7, 3.1 Hz,
H-3¦), 4.51 (d, 2 H, J 3.5 Hz, H-2% and H-3%),
4.45 (dd, 1 H, J 7.3, 3.1 Hz, H-4¦), 4.42 (m, 1
H, H-4%), 4.40–4.30 (m, 3 H, H-5a%, H-5b% and
H-5¦), 3.91 (m, H-6a¦ and H-6b¦); 19F (D2O,
300 MHz): l −203.95 (m). ESIMS: calcu-
lated for C15H23FN2O16P2 (M−H+) 567.1,
observed=567.2 m/z.
Uridine 5%-(3-deoxy-3-fluoro-
nosyl diphosphate) (1).—To a solution (1 mL)
containing 3-deoxy-3-fluoro- -galactose (100
D-galactopyra-
D
mmol, 18 mg), ATP (5 mmol, 2.7 mg), phos-
phoenolpyruvate (20 mmol, 4.1 mg), MgCl2
(10 mmol, 2.0 mg), triethanolamine (100 mM,
pH 7.5) and 9 units pyruvate kinase was
added 4 units of galactokinase. The mixture
was incubated at 37 °C under nitrogen, and
conversion into 3-deoxy-3-fluoro- -galactose
D
Uridine 5%-(2-deoxy-2-fluoro- -galactopyra-
D
1-phosphate monitored by the appearance of
nosyl diphosphate) (2).—The synthesis of 2
was carried out using the same procedure as
that used for the synthesis of 1. To a solution
1
a resonance in the H NMR (D2O) spectrum
at l=5.56 (dt, J 4.5, 6.8 Hz), which was
assigned to the anomeric hydrogen of 3-de-
(1 mL) containing 2-deoxy-2-fluoro- -galac-
D
oxy-3-fluoro- -galactose 1-phosphate. The
D
tose (30 mmol, 5.5 mg), ATP (1.8 mmol, 1 mg),
phosphoenolpyruvate (53 mmol, 11 mg),
MgCl2 (10 mmol, 2 mg), Tris (100 mmol, 12.1
mg), pH adjusted to 8.0 with NaOH, was
added galactokinase (1 unit) and pyruvate ki-
nase (50 units). After 2 h at 37 °C, the compo-
nents of the uridylation reaction were added:
UDP-glucose (23 mmol, 14.2 mg), NADP+ (1
mmol, 0.8 mg), DTT (2 mmol, 0.3 mg), a-ke-
area of this resonance was calibrated against
that of the anomeric hydrogen of the a
anomer of 3-deoxy-3-fluoro- -galactose (l=
D
5.45 ppm). After 72 h, 72% conversion was
achieved.
For uridylation, the mixture was diluted to
2 mL with triethanolamine (100 mM, pH 7.5)
and the following components added: UDP- -
D
glucose (6.6 mmol, 4 mg), NADP+ (2 mmol,
1.5 mg), DTT (2 mmol, 0.3 mg), a-ketoglu-
toglutarate (40 mmol, 7.6 mg),
D
-glucose 1,6-

J.N. Barlow, J.S. Blanchard / Carbohydrate Research 328 (2000) 473–480
479
with optical monitoring at 262 nm. Kinetic
parameters, kcat and kcat/Km, were determined
as described above, with 10-min incubation
times at various concentrations of 1 and 2.
Both compounds exhibited saturation kinetics,
and the kinetic parameters were determined
from fits of the data to the Michaelis–Menten
equation.
biphosphate (0.07 mmol, 0.05 mg), galactose-1-
phosphate uridyltransferase (2 units), phos-
phoglucomutase (1 unit), glucose-6-phosphate
dehydrogenase (1 unit) and -glutamic dehy-
L
drogenase (1 unit, from an ammonium sulfate
suspension). The reaction was left at 37 °C for
48 h during which time greater than 95%
conversion of UDP-glucose into 1 was ob-
served. Purification and storage of 2 was car-
ried out as already described for compound 1.
1H NMR (D2O, 300 MHz): l 8.08 (d, 1 H, J
8.1 Hz, H-6), 6.11 (d, 1 H, J 3.8 Hz, H-1%),
6.09 (d, 1 H, J 8.1 Hz, H-5), 5.93 (dd, 1 H, J
7.4, 3.5 Hz, H-1¦), 4.82 (ddt, 1 H, J 50.1, 9.9,
3.0 Hz, H-2¦), 4.49 (d, 2 H, J 3.5 Hz, H-2% and
H-3%), 4.40 (m, 1 H, H-4%), 4.37–4.25 (m, 4H,
H-5a%, H-5b%, H-3¦ and H-5¦), 4.21 (t, 1 H, J
3.8 Hz, H-4¦), 3.85 (m, 2H, H-6a¦ and H-6b¦);
19F NMR (D2O, 300 MHz): l −208.575
(ddd, F-2, J 49.6, 12.1, 3.9 Hz). ESIMS: calcu-
lated for C15H23FN2O16P2 (M−H+) 567.1,
observed=567.2 m/z.
Incubation of 1 and 2 with UDP-Galp mu-
tase.—Compounds 1 and 2 were tested as
substrates of UDP-Galp mutase using assay
conditions described next. UDP-sugar (0.5
mM) was incubated anaerobically under N2 at
37 °C in 5-mL glass Thunberg tubes (with side
arm) sealed with rubber septa in a volume of
0.5 mL. The assay mixture contained FAD
(0.1 mM), DTT (1 mM) and 50 mM AMPSO,
pH 8.3. UDP-Galp mutase was contained in
the side arm (40 mL of 0.3 mM for 1, 37.5 mM
for 2). The sample was made anaerobic by
passing nitrogen gas over the surface of the
assay mixture for 5 min, followed by bubbling
with N2 for 5 min. The assay mixture was then
mixed with the UDP-Galp mutase in the side
arm and the sample incubated at 37 °C. Small
aliquots were removed at various intervals,
quenched by boiling for 1 min, and the high
molecular weight material removed by cen-
trifugal filtration (Ultrafree-MC tubes, 10,000
nominal molecular weight limit, Millipore).
The sample was then diluted in several vol-
umes of 75 mM KH2PO4, pH 4.5, and injected
onto a 4×250 mm Carbopac PA-100 anion-
exchange HPLC column (Dionex), pre-equili-
brated in 75 mM KH2PO4, pH 4.5. The
sample was eluted isocratically at 1 mL/min,
Turnover of 1 for analysis by 19F NMR was
performed aerobically at 37 °C. To 300 mL of
1 in D2O was added (to final concentrations)
FAD (0.2 mM), DTT (5 mM), potassium
phosphate (20 mM, pH 7.5) and UDP-Galp
mutase (2.1 mM). Samples were incubated for
12 h at 37 °C prior to NMR spectroscopic
analysis.
Measuring pyranose/furanose distribution of
D
-galactose, 3-deoxy-3-fluoro-
2-deoxy-2-fluoro- -galactose.—Solutions (100
mM) of -galactose, 3-deoxy-3-fluoro-
galactose and 2-deoxy-2-fluoro- -galactose
D
-galactose, and
D
D
D-
D
were made up in D2O and incubated for
greater than 12 h at 37 °C, prior to spectral
1
analysis at 37 °C by a combination of H
1
NMR (1D, 2D H{1H} COSY) at 600 MHz,
1
and 19F NMR (with and without H decou-
pling) at 300 MHz.
Acknowledgements
We wish to thank Dr Sean Cahill for expert
technical assistance and helpful discussions
concerning NMR spectroscopy, and Drs
W.W. Cleland and C. Walsh for helpful mech-
anistic discussions. This work was supported
by NIH grant GM33449.
References
[1] P.M. Nassau, S.L. Martin, R.E. Brown, A. Weston, D.
Monsey, M.R. McNeil, K.J. Duncan, J. Bacteriol., 178
(1996) 1047–1052.
[2] R. Ko¨plin, J.-R. Brisson, C. Whitfield, J. Biol. Chem.,
272 (1997) 4121–4128.
[3] R. Lee, D. Monsey, A. Weston, K. Duncan, C. Rithner,
M. McNeil, Anal. Biochem., 242 (1996) 1–7.
[4] C. Whitfield, Trends Microbiol., 3 (1995) 178–185.
[5] P.J. Brennan, H. Nikaido, Annu. Re6. Biochem., 64
(1995) 29–63.

480
J.N. Barlow, J.S. Blanchard / Carbohydrate Research 328 (2000) 473–480
[6] J.N. Barlow, M.E. Girvin, J.S. Blanchard, J. Am. Chem.
Soc., 121 (1999) 6968–6969.
[7] G. Davies, M.L. Sinnott, S.G. Withers, in M. Sinnott
(Ed.), Comprehensi6e Biological Catalysis, Vol. 1, Aca-
demic, New York, 1998, pp. 119–209.
[8] D.L. Zechel, S.G. Withers, Acc. Chem. Res., 33 (2000)
11–18.
[9] T. Hayashi, B.T. Murray, R. Wang, C.-H. Wong,
Bioorg. Med. Chem., 5 (1997) 497–500.
[11] T.E. Barman, Enzyme Handbook, Springer, New York,
1969.
[12] S.J. Angyal, V.A. Pickles, Aust. J. Chem., 25 (1972)
1695–1710.
[13] P.W. Wertz, J.C. Garver, L. Anderson, J. Am. Chem.
Soc., 103 (1981) 3916–3922.
[14] Y. Kajihara, H. Kodama, T. Endo, H. Hashimoto,
Carbohydr. Res., 306 (1998) 361–378.
[15] V. Notenboom, C. Birsan, M. Nitz, D.R. Rose, R.A.J.
Warren, S.G. Withers, Nat. Struct. Biol., 5 (1998) 812–
818.
[10] P. Thomas, E.M. Bessell, J.H. Westwood, Biochem., J.,
139 (1974) 661–664.
.