The preparation of deoxy derivatives of mannose-1-phosphate and their substrate specificity towards recombinant GDP-mannose pyrophosphorylase from Salmonella enterica, group B

Article abstract of DOI:10.1016/S0957-4166(99)00556-X

2-Deoxy-α-D-glucose-1-phosphate, 3-deoxy-α-D-arabino-hexose-1-phosphate, 4-deoxy-α-D-lyxo-hexose-1-phosphate, and α-D-lyxose-1-phosphate were synthesised chemically, and evaluated as substrates for a recombinant GDP-mannose pyrophosphorylase (Salmonella enterica, group B, cloned in Escherichia coli). The deoxy derivatives were all substrates for the enzyme, with slightly reduced V(max) values but significantly higher K(m) values than those recorded for the native substrate, mannose-1-phosphate. The pyrophosphorylase was used for the synthesis of GDP-mannose analogues GDP-2-deoxy-glucose and GDP-lyxose on a milligram scale. Copyright (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0957-4166(99)00556-X

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 11 (2000) 621–628
The preparation of deoxy derivatives of mannose-1-phosphate
and their substrate specificity towards recombinant GDP-
mannose pyrophosphorylase from Salmonella enterica, group B
Gregory M. Watt,a Sabine L. Flitsch,a,∗ Sven Fey,b Lothar Elling c and Udo Kragl b,†
aThe Edinburgh Centre for Protein Technology, Department of Chemistry, University of Edinburgh, King’s Buildings,
West Mains Road, Edinburgh EH9 3JJ, UK
bForschungzentrum Jülich, Institut für Biotechnologie 2, D-52425 Jülich, Germany
cHeinrich-Heine-Universität Düsseldorf, Institut für Enzymtechnologie, D-52426 Jülich, Germany
Received 5 November 1999; accepted 2 December 1999
Abstract
2-Deoxy-α-D-glucose-1-phosphate, 3-deoxy-α-D-arabino-hexose-1-phosphate, 4-deoxy-α-D-lyxo-hexose-1-
phosphate, and α-D-lyxose-1-phosphate were synthesised chemically, and evaluated as substrates for a recom-
binant GDP-mannose pyrophosphorylase (Salmonella enterica, group B, cloned in Escherichia coli). The deoxy
derivatives were all substrates for the enzyme, with slightly reduced Vmax values but significantly higher Km
values than those recorded for the native substrate, mannose-1-phosphate. The pyrophosphorylase was used for
the synthesis of GDP-mannose analogues GDP-2-deoxy-glucose and GDP-lyxose on a milligram scale. © 2000
Elsevier Science Ltd. All rights reserved.
1. Introduction
Oligosaccharides are ubiquitous in Nature and can display many biological activities, which makes
them important synthetic targets.1 However, chemical synthesis of oligosaccharides is difficult due to
multi-step protection/deprotection strategies, and in some cases low stereoselectivity of chemical glyco-
sidation reactions. Glycosyltransferases catalyse the formation of glycosidic linkages in Nature, and have
become major tools for the synthesis of complex oligosaccharides.2 Catalysis by glycosyltransferases is
both regio- and stereoselective, allowing relatively straightforward purifications of glycosylated products.
We are particularly interested in the application of mannosyltransferases, which we have used for the
selective formation of the core β-mannosidic linkage of N-glycans.3,4
∗
†
Corresponding author.
New address: Universität Rostock, Fachbereich Chemie, Buchbinderstraße 9, D-18055 Rostock, Germany.
0957-4166/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(99)00556-X
tetasy 3191

622
G. M. Watt et al. / Tetrahedron: Asymmetry 11 (2000) 621–628
The requirement of expensive sugar nucleotides as glycosyl donor substrates for glycosyltransferases
is a disadvantage, which has largely been overcome with the development of in situ sugar nucleotide
recycling systems.5,6 This has allowed the large-scale production of carbohydrate-based drugs for the
pharmaceutical industry;7,8 two important examples are the selectin inhibitor sialyl Lewis X and the
microbial adhesion inhibitor 30-sialyllactose.
An efficient enzymatic synthesis of GDP-mannose 2 from mannose-1-phosphate 1 and GTP using
GDP-mannose pyrophosphorylase from Salmonella enterica, group B (Scheme 1) has been developed9,10
with the aim of producing this sugar nucleotide on a large scale at low cost. Here we report the substrate
specificity of this GDP-mannose pyrophosphorylase for chemically synthesised deoxy-derivatives of
mannose-1-phosphate. In addition, we show that the enzyme can be used for the preparation of GDP-
2-deoxy-α-D-glucose 18 and GDP-α-D-lyxose 19 on a milligram scale.
Scheme 1.
2. Results and discussion
Glycosyl-1-phosphates are biologically important molecules, and much research has been carried out
on possible synthetic routes to these compounds. We chose the MacDonald reaction,11 which involves
treatment of peracetylated sugars with phosphoric acid, followed by addition of lithium hydroxide, thus
providing a straightforward route to glycosyl-1-phosphates. Mannose-1-phosphate 1 was synthesised
from peracetylated mannose 3 on a multi-gram scale using this reaction (Scheme 2). Due the absence
of the 2-hydroxyl group, 2-deoxy-α-D-glucose-1-phosphate 5 is more labile to hydrolytic cleavage
of the anomeric phosphate, which makes its chemical synthesis difficult.12–14 However, we found
that under carefully controlled reaction conditions both 2-deoxy-glucose-1-phosphate 515 and 3-deoxy-
arabino-hexose-1-phosphate 7 could be prepared from 4 and peracetylated 3-deoxy-arabino-hexose 6,16
respectively, using the MacDonald reaction (Scheme 2).11
Scheme 2. Reagents: (a) (i) H3PO4; (ii) LiOH
The preparation of 4-deoxy-lyxo-hexose-1-phosphate 13, shown in Scheme 3, involved selective
benzoylation of mannose17 affording tetra-benzoate 8, followed by deoxydation of the free 4-position
by treatment with thiocarbonyl-dipyridone then tributyltin hydride and AIBN18 affording perbenzoylated
4-deoxy-lyxo-hexose 9 in 54% yield. This was deprotected using sodium methoxide in methanol to give
4-deoxy-lyxo-hexose, which was acetylated using acetic anhydride and pyridine to give peracetylated
4-deoxy-lyxo-hexose 10 in quantitative yield.
Peracetylated lyxose 14 was subjected to MacDonald reaction conditions11 but an inseparable 4:1 α:β
mixture of lyxose-1-phosphates was obtained. This was possibly due to the high temperatures required

G. M. Watt et al. / Tetrahedron: Asymmetry 11 (2000) 621–628
623
Scheme 3. Reagents: (a) (i) N,N0-thiocarbonyl-dipyridone, toluene; (ii) AIBN, tributyltin hydride; (b) (i) NaOMe, MeOH;
(ii) Ac2O, pyridine; (c) benzylamine, THF; (d) (i) LDA, THF; (ii) tetrabenzyl pyrophosphate; (e) (i) H2(g), Pd/C, EtOH, 10%
NaHCO3 solution; (ii) NaOMe, MeOH
for this reaction. Thus, an alternative approach to lyxose-1-phosphate 17, whereby phosphorylation was
carried out at low temperature, was used.
The preparation of 4-deoxy-lyxo-hexose-1-phosphate 13 and lyxose-1-phosphate 17 (Scheme 3)
started from the peracetylated deoxy sugars 10 and 14, respectively, which were firstly deprotected at
the anomeric centre using benzylamine.19 The resulting hemiacetals 11 and 15 were each treated with
lithium diisopropylamide at low temperature to ensure exclusive formation of the α-anomeric oxides;
tetrabenzyl pyrophosphate was added to give the protected sugar phosphates 12 and 16. Removal of the
protecting groups afforded the sugar phosphates 13 and 17 in 54 and 49% yield, respectively.
2.1. Incubation of deoxy analogues with GDP-mannose pyrophosphorylase
Previous studies have focused on deoxy analogues of GDP-mannose 2, which were tested as substrates
for a GDP-mannose:dolichyl phosphate mannosyltransferase in chick embryo microsomes.20 The results
of this study revealed that GDP-mannose 2 has an identical apparent Km value to the apparent Ki
value measured for GDP-6-deoxymannose, indicating that the 6-hydroxyl group of mannose does not
participate in enzyme–substrate recognition. The 3-deoxy, 2-deoxy, and 4-deoxy analogues were worse
inhibitors by factors of 2, 2.5, and 6, respectively, indicating that the hydroxyl groups at these positions, in
particular position 4, may have some influence in enzyme–substrate recognition. The substrate specificity
of GDP-mannose 2 has also been tested on a mannosyltransferase from Salmonella anatum,21 whereby
the 6-deoxy, 3-deoxy, and 2-deoxy analogues were shown to be transferred at relative velocities of 0.29,
0.12, and 0.11, respectively, to the native compound.
The present study investigates the role hydroxyl groups play in the biosynthesis of GDP-mannose
itself. Thus, mannose-1-phosphate 1, 2-deoxy-glucose-1-phosphate 5, 3-deoxy-arabino-hexose-1-
phosphate 7, 4-deoxy-lyxo-hexose-1-phosphate 13, and lyxose-1-phosphate 17 were studied in detail
with fully purified recombinant GDP-mannose pyrophosphorylase. The apparent Km and Vmax values
are listed in Table 1. The Vmax values were least affected by the removal of the 3-hydroxyl group and
were similarly affected by removal of the 2-, 4-hydroxyl groups and the hydroxymethylene group in the
5 position. The highest reduction in Vmax was seen for 17 which had a value of about 10% of that for the
natural substrate 2.
The effect on the Km values was much more significant, with the 2-deoxy-glucose-1-phosphate 5 being
the worst substrate, showing a 300 fold increase in Km. This would suggest that all four hydroxy groups
in mannose-1-phosphate contribute to the binding to the pyrophosphorylase.
It has been shown previously that the present GDP-mannose pyrophosphorylase can be used to catalyse

624
G. M. Watt et al. / Tetrahedron: Asymmetry 11 (2000) 621–628
Table 1
the synthesis of GDP-mannose itself on a large scale.10 It was, therefore, of interest to see if the
enzyme could also be applied to the preparation of deoxy analogues of GDP-mannose, despite their
less favourable kinetic constants. We found that even with the poorest substrate in our studies, the 2-
deoxy analogue 5, synthesis on a preparative scale was possible. Thus, GDP-2-deoxy-glucose 18 and
GDP-lyxose 19 were synthesised, purified and characterised by NMR spectroscopy. This shows that the
GDP-mannose pyrophosphorylase from S. enterica can be used for the synthesis of GDP-mannose deoxy
derivatives. These are useful probes for monitoring the substrate specificity of mannosyltransferases and
will be used by us in further studies of the N-glycan biosynthetic pathway in yeast.22
3. Experimental
3.1. General methods
Proton NMR spectra were recorded at 200 MHz on Varian and Bruker spectrometers, at 400 MHz
on a Bruker spectrometer, and at 600 MHz on a Varian spectrometer. HPLC23 was carried out using
an Hypersil ODS-5 µm (RP-18) column of dimensions 125×4.8 mm; the eluents were for A: 100 mM
KH2PO4, 8 mM tetrabutylammonium sulfate (pH 5.3); and for B: MeOH; the gradient employed was for
0 min: 0% B, 2.5 min: 0% B, 16.5 min: 12% B, 17.5 min: 30% B, 29 min: 30% B, 30 min: 0% B, and
35 min 0% B; nucleotide derivatives were detected by UV absorption (253 nm); the injection volume
was 20 µl. Enzymes used were: inorganic phosphatase (PPase) from yeast (Boehringer M.) and alkaline
phosphatase molecular biology grade 713023 from calf intestine (Boehringer M.). All other chemicals
were purchased from Aldrich/Sigma/Fluka or Merck.
3.2. 2-Deoxy-α-D-glucose-1-phosphate (dilithium salt) 5
Peracetylated 2-deoxy-α-D-glucose 4 (0.20 g, 0.60 mmol) was dissolved in a 1:10 w/v solution of
dry phosphoric acid in THF (5 ml, 6.00 mmol) and the solvent was removed under reduced pressure.
The resulting syrup was stirred under reduced pressure (0.01 kPa) at room temperature for 40 min,
whereafter the vacuum was removed and a 1 M solution of lithium hydroxide (25 ml) was added.
The resulting milky suspension was stirred for 3 h and then filtered through a Whatman® GF/A glass
microfibre filter. The filtrate was neutralised with AG® 50W-X2 (H+ form) ion exchange resin and the
aqueous solvent was removed under reduced pressure. The resulting residue was treated with methanol
(10 ml) and the resulting suspension was centifuged at 4000 rpm for 5 min, whereafter the supernatant

G. M. Watt et al. / Tetrahedron: Asymmetry 11 (2000) 621–628
625
was removed. After this was repeated three times, the pellet was dried in vacuo to give 2-deoxy-α-D-
glucose-1-phosphate 5 (72 mg, 41%) as a white powder; [α]D +73 (c 0.4 in H2O) [lit.,15 +72.4 (c 0.5 in
H2O)]; NMR (D2O) (200 MHz) δH 1.59 (ddd, 1H, J1,2ax 2 Hz, J2ax,2eq 13 Hz, J2ax,3 9 Hz, H-2ax), 2.12
(ddd, 1H, J1,2eq <1 Hz, J2ax,2eq 13 Hz, J2eq,3 5 Hz, H-2eq), 3.25 (dd, 1H, J3,4 9 Hz, J4,5 9 Hz, H-4), 3.66
(dd, 1H, J5,6a 6 Hz, J6a,6b 12 Hz, H-6a), 3.83 (m, 2H, H-5 and H-6b), 3.95 (ddd, 1H, J2eq,3 5 Hz, J2ax,3
9
Hz, J3,4 9 Hz, H-3), and 5.45 (br-d 1H, J1,P 8 Hz, H-1); ESMS (−) 243.0.
3.3. 3-Deoxy-α-D-arabino-hexose-1-phosphate (dilithium salt) 7
Peracetylated 3-deoxy-α-D-arabino-hexose 6 (0.46 g, 1.37 mmol) was dissolved in anhydrous THF
containing 1:10 w/v phosphoric acid (11.5 ml, 11.73 mmol), and the solvent was removed under reduced
pressure. The resulting syrup was stirred at room temperature under reduced pressure (0.01 kPa) for 90
min, the temperature was increased to 60°C, and stirring was continued at the same pressure for 50 min.
After the reaction had cooled to room temperature, the vacuum was removed and 1 M lithium hydroxide
(45 ml) was added. The resulting suspension was stirred for 3 h, filtered through a Whatman® GF/A glass
microfibre filter, and the filtrate was neutralised with AG® 50W-X2 (H+ form) ion exchange resin, then
concentrated. The residue was washed with methanol as described for 2-deoxy-glucose-1-phosphate 5,
and dried in vacuo to give 3-deoxy-α-D-arabino-hexose-1-phosphate 7 (0.13 g, 38%) as a white powder;
[α]D +53 (c 0.4 in H2O) [lit.,24 +53.3 (c 0.13 in H2O)]; NMR (D2O) (200 MHz) δH 1.85 (m, 2H, H-3ax,
H-3eq), 3.51 (dd, 1H, J5,6a 7 Hz, J6a,6b 12 Hz, H-6a), 3.67 (ddd, 1H, J3eq,4 6 Hz, J3ax,4 10 Hz, J4,5 10 Hz,
H-4), 3.63–3.81 (m, 3H, H-2, H-5, and H-6b), and 5.02 (dd, 1H, J1,2 <2 Hz, J1,P 9 Hz, H-1); ESMS (−),
243.0.
3.4. 2,3,6-Tri-O-acetyl-4-deoxy-D-lyxo-hexose 11
A solution of the peracetylated compound 10 (860 mg, 2.60 mmol) in THF (8 ml) was treated with
benzylamine (0.43 ml, 3.90 mmol) for 20 h. The reaction was diluted with chloroform (100 ml) and
washed with cold 0.2 M hydrochloric acid (200 ml), saturated solution of sodium hydrogencarbonate
(200 ml) and water (200 ml). The organic layer was dried (Na2SO4), filtered and concentrated under
reduced pressure. Flash-column chromatography (eluent: 50 to 60% gradient of ethyl acetate in hexane)
of the residue afforded the hemiacetal 11 as a syrup (487 mg, 65%); NMR (CDCl3) (200 MHz) δH
1.70–1.90 (m, 2H, H-4ax and H-4eq), 2.07, 2.12, and 2.13 (3×s, 9H, 3×COCH3), 4.10–4.40 (m, 3H,
H-5, H-6a, and H-6b), and 5.25–5.44 (m, 3H, H-1, H-2, and H-3).
3.5. Dibenzyl 2,3,6-tri-O-acetyl-4-deoxy-α-D-lyxo-hexose-1-phosphate 12
A cooled (–60°C) and stirred solution of the hemiacetal 11 (487 mg, 1.68 mmol) in freshly dis-
tilled THF (12 ml) was treated with a 2 M solution of lithium diisopropylamide (2.18 mmol) in
1:1:1 heptane:THF:ethylbenzene (1.10 ml) for 45 min at −60°C, whereafter a solution of tetrabenzyl
pyrophosphate (1.27 g, 2.35 mmol) in THF (3 ml) was added. Stirring was continued at −60°C for 4
h, after which time the reaction mixture was slowly allowed to attain room temperature. The solvent
and diisopropylamine formed in the reaction were removed under reduced pressure. Flash-column
chromatography (eluent: 40 to 50% gradient of ethyl acetate in hexane) of the residue afforded the title
compound 12 as a white solid (600 mg, 65%); NMR25 (CDCl3) (200 MHz) δH 1.82–1.90 (m, 2H, H-4ax
and H-4eq), 2.01, 2.07, and 2.17 (3×s, 9H, 3×COCH3), 4.05–4.18 (m, 3H, H-5, H-6a, and H-6b), 5.11

626
G. M. Watt et al. / Tetrahedron: Asymmetry 11 (2000) 621–628
(m, 1H, H-2), 5.11 and 5.15 (2×s, 4H, 2×CH2Ph), 5.29 (m, 1H, H-3), 5.71 (dd, 1H, J1,2 2 Hz, J1,P 6 Hz,
H-1), and 7.39 (m, 10H, 2×Ph).
3.6. 4-Deoxy-α-D-lyxo-hexose-1-phosphate (disodium salt) 13
Ethanol (6 ml) and 10% sodium hydrogen carbonate (4 ml) were added to a flask containing the
protected deoxy-α-D-lyxo-hexose-1-phosphate 12 (193 mg, 0.35 mmol) and 10% palladium on carbon
(40 mg). The resulting suspension was stirred vigorously under a slight positive pressure of hydrogen for
15 h, whereafter TLC (eluent: 40% ethyl acetate in hexane) showed that no starting material remained.
The mixture was filtered through a Celite pad and washed in turn with 2:1 (20 ml) and 1:1 (20 ml)
mixtures of methanol:water. The filtrate and washing were concentrated under reduced pressure and
the resulting residue was taken up in a 10:10:1 mixture of methanol:water:triethylamine and left to
stand at room temperature for 3 h, whereafter TLC (eluent: 40% 0.1 M ammonium chloride solution
in acetonitrile) indicated the formation of a single product. The reaction mixture was concentrated under
reduced pressure to give a white solid, which was purified by ion exchange chromatography DOWEX®
50W-X8 (sodium form, Serva Feinbiochemica) to afford 4-deoxy-α-D-lyxo-hexose-1-phosphate 13 (62
mg, 54%) as a white solid; NMR25 (D2O) (200 MHz) δH 1.40–1.59 (m, 2H, H-4ax and H-4eq), 3.48–4.10
(m, 5H, H-2, H-3, H-5, H-6a and H-6b), and 5.11 (dd, 1H, J1,2 2 Hz, J1,P 9 Hz, H-1); ESMS (−): 243.
3.7. 2,3,4-Tri-O-acetyl-D-lyxose 15
Peracetylated α-D-lyxose 16 (0.73 g, 2.31 mmol) in THF (6 ml) was treated with benzylamine (0.38
ml, 3.47 mmol) as described for 2,3,6-tri-O-acetyl-4-deoxy-D-lyxo-hexose 11 to afford the hemiacetal 15
(0.46 g, 72%) as a syrup; [α]D −12 (c 0.7 in CHCl3); λmax(MeOH)/nm 213 and 265 (ε/dm3 mol−1 cm−1
91 and 33); νmax/cm−1 3450m (OH) and 1750s (CO); NMR (CHCl3) δH 2.04, 2.06, and 2.11 (3×s, 9H,
3×COCH3), 3.88 (m, 2H, H-5ax and H-5eq) 5.08 (d, 1H, J1,2 3 Hz, H-1), 5.08 (m, 1H, H-4), 5.17 (dd,
1H, J1,2 3 Hz, J2,3 3 Hz, H-2), and 5.38 (dd, 1H, J2,3 3 Hz, J3,4 8 Hz, H-3); δC 20.72, 20.80, and 20.83
(3C, 3×COCH3), 60.59 (C-5), 67.26 (C-4), 68.27 (C-3), 70.03 (C-2), 92.37 (C-1), 169.86 (COCH3), and
170.38 (2C, 2×COCH3); m/z (EI) 170 (34), 157 (43), 128 (100), 115 (56), 103 (12), 86 (28), and 73
(16).
3.8. Dibenzyl 2,3,4-tri-O-acetyl-α-D-lyxose-1-phosphate 16
Hemiacetal 15 (0.30 g, 1.08 mmol) was subjected to analogous conditions used for the preparation of
dibenzyl 2,3,6-tri-O-acetyl-4-deoxy-α-D-mannose-1-phosphate 12 to afford the title compound 16 (0.30
g, 52%) as a syrup; [α]D +22 (c 0.3 in CHCl3); λmax(MeOH)/nm 213 and 265 (ε/dm3 mol−1 cm−1 4390
and 523); νmax/cm−1 1750s (CO), 1280s (PO), 740 and 700 (CH (Ar)); NMR (CHCl3) (200 MHz) δH
2.01, 2.03, and 2.07 (3×s, 9H, 3×COCH3), 3.67 (dd, 1H, J4,5ax 9 Hz, J5ax,5eq 11 Hz, H-5ax), 3.90 (dd,
1H, J4,5eq 5 Hz, J5ax,5eq 11 Hz, H-5eq), 5.02 and 5.09 (2×s, 4H, 2×OCH2Ph), 5.13 (m, 1H, H-4), 5.21
(dd, 1H, J1,2 3 Hz, J2,3 3 Hz, H-2), and 5.29 (dd, 1H, J2,3 3 Hz, J3,4 9 Hz, H-3), 7.33 and 7.34 (2×s,
10H, 2×OCH2Ph); m/z (EI) 259.0 (65, M−277.1 OPO(OCH2Ph)2), 186.9 (43), 157.0 (37), 107.0 (13,
PhCHOH), and 91.0 (100, PhCH2).

G. M. Watt et al. / Tetrahedron: Asymmetry 11 (2000) 621–628
627
3.9. α-D-Lyxose-1-phosphate (disodium salt) 17
Dibenzyl 2,3,4-tri-O-acetyl-α-D-lyxose-1-phosphate 16 was subjected to analogous hydrogenation
conditions used for the preparation of 4-deoxy-α-D-lyxo-hexose-1-phosphate 13. The debenzylated
material was taken up in methanol (5 ml), filtered to remove any sodium hydrogen carbonate present, and
washed with methanol (7 ml). The combined filtrate and washings were treated with a 0.5 M solution
of sodium methoxide in methanol (0.3 ml) at room temperature for 2 h, whereafter the reaction mixture
was neutralised with AG® 50W-X2 (H+ form) ion exchange resin and the solvent removed under reduced
pressure. The residue was taken up in a minimum amount of water and methanol was added, whereupon
a precipitate formed, which was collected by filtration to give the title compound 17 (76 mg, 49%) as
a white solid; [α]D +11 (c 0.3 in H2O); NMR (D2O) (200 MHz) δH 3.57–3.90 (m, 5H, H-2, H-3, H-4,
H-5ax, and H-5eq) and 5.15 (dd, 1H, J1,2 2 Hz, J1,P 8 Hz, H-1); ESMS (−) 229.1.
3.10. Kinetic measurements of GDP-mannose pyrophosphorylase with glycosyl-1-phosphates as subs-
trates
To Eppendorf tubes containing GTP (2 mM), MgCl2 (2 mM), Tris® buffer (50 mM, pH 8), and the
appropriate concentration of glycosyl-1-phosphate was added GDP-mannose pyrophosphorylase (100
µl). After 5 min at room temperature the tubes were heated at 100°C for 5 min to stop the reaction by
denaturation of the enzyme. The tubes were centrifuged at 14 000 rpm for 5 min and the supernatant
subjected to HPLC analysis,23 whereby the concentation of GDP-sugar was calculated by integration of
the peak areas using commercially available GDP-mannose as a standard.
3.11. Synthesis of GDP-sugars using GDP-mannose pyrophosphorylase
A solution of DTT (5 mM), BSA (1 mg/ml), MgCl2 (4 mM), GTP (4 mM), and the respective
glycosyl phosphate (4 mM) in water (19.5 ml) was adjusted to pH 8.0 using 4 M sodium hydroxide.
GDP-Man pyrophosphorylase (250 µl of 1.65 U/ml) and inorganic phosphatase (250 µl of 200 U/ml)
were added (final volume 20 ml) and the mixture was passed through a sterilizing filter. The pH of
8.0 was maintained by the addition of 4 M sodium hydroxide using Metrohm pH-Stat. apparatus,
and the progress of the reaction was monitored by subjecting aliquotes of the reaction mixture (10
µl) diluted with water (490 µl) to HPLC analysis. After 60 h the reaction mixture was treated with
alkaline phosphatase for 24 h, whereafter any protein was removed by filtration through Centriplus-10™
ultrafiltration concentraters (centrifugation at 4000 g for 1 h). The filtrate was concentrated to 2 ml under
reduced pressure and purified by semi-preparative HPLC. Fractions containing the GDP-sugar derivative
were pooled, neutralised (2 M NaOH), washed three times with 1:1 chloroform:acetonitrile (10 ml), and
concentrated under reduced pressure to give the GDP-deoxy sugar (contaminated with phosphate salt).
3.11.1. GDP-2-deoxy-α-D-glucose 18
00,2ax00
00,2eq00
NMR (D2O) (600 MHz) δH 1.63 (ddd, 1H, J1
3 Hz, J2ax
13 Hz, J2ax
9 Hz, H-2ax00),
00,300
2.12 (ddd, 1H, J1
<1 Hz, J2ax
13 Hz, J2eq
5 Hz, J6a
5 Hz, H-2eq00), 3.34 (dd, 1H, J3
00,400
10 Hz,
2 Hz,
00,2eq00
00,2eq00
00,300
00,6b00
J4
J3
10 Hz, H-400), 3.71 (dd, 1H, J5
13 Hz, H-6a00), 3.77 (dd, 1H, J5
00,500
00,6b00
00,300
00,6b00
13 Hz, H-6b00), 3.79 (m, 1H, H-5), 3.91 (ddd, 1H, J2eq
10 Hz,
00,300
00,300
5 Hz, J2ax 9 Hz, J3
00,400
H-300), 4.16 (m, 2H, H-5a0 and H-5b0), 4.31 (m, H1, H-40), 4.46 (dd, 1H, J2 ,3 5 Hz, J3 ,4 4 Hz, H-30),
0
0
0
0
4.71 (dd, 1H, J1 ,2 6 Hz, J2 ,3 5 Hz, H-20), 5.63 (br-d 1H, J1 7 Hz, H-1 ), 5.90 (d, 1H, J1 ,2 6 Hz,
0
0
00,P
0
0
0
0
00
H-10), and 8.07 (s, 1H, H-8).

628
G. M. Watt et al. / Tetrahedron: Asymmetry 11 (2000) 621–628
3.11.2. GDP-α-D-lyxose 19
NMR (D2O) (400 MHz) δH 3.62–3.95 (m, 5H, H-200, H-300, H-400, H-5ax00, and H-5eq00), 4.15 (m,
0
2H, H-5a0 and H-5b0), 4.31 (m, 1H, H-40), 4.45 (dd, 1H, J2 ,3 5 Hz, J3 ,4 4 Hz, H-3 ), 4.71 (dd, 1H, J1 ,2
0
0
0
0
0
0
6 Hz, J2 ,3 5 Hz, H-20), 5.38 (dd, 1H, J1
2 Hz, J1
8 Hz, H-100), 5.92 (d, 1H, J1 ,2 6 Hz, H-10),
0 0
0
0
00,200
00,P00
and 8.22 (s, 1H, H-8).
Acknowledgements
The authors are grateful to the BBSRC and the British Council who supported the work carried out by
G.M.W. in Edinburgh and Jülich, respectively. We also thank Deutschen Akademischen Austauschdienst
for giving S.F. the opportunity to study in Edinburgh. Finally, we thank the technical staff at the University
of Edinburgh and Forschungzentrum Jülich for their help regarding the analysis of our products.
References
1. Varki, A. Glycobiology 1993, 3, 97–130.
2. Palcic, M. M. Methods Enzymol. 1994, 230, 300–360.
3. Watt, G. M.; Revers, L.; Webberley, M. C.; Wilson, I. B. H.; Flitsch, S. L. Carbohydr. Res. 1998, 305, 533–541.
4. Watt, G. M.; Revers, L.; Webberley, M. C.; Wilson, I. B. H.; Flitsch, S. L. Angew. Chem., Int. Ed. Engl. 1997, 36, 2354–2356.
5. Ichikawa, Y.; Wang, R.; Wong, C.-H. Methods Enzymol. 1994, 247, 107–127.
6. Wong, C.-H.; Haynie, S. L.; Whitesides, G. M. J. Org. Chem. 1982, 47, 5416–5418.
7. McAuliffe, J. C.; Hindsgaul, O. Chem. Ind. 1997, 3, 170–174.
8. Simon, P. M. Drug Discovery Today 1996, 1, 522–528.
9. Elling, L.; Ritter, J. E.; Verseck, S. Glycobiology 1996, 6, 591–597.
10. Fey, S.; Elling, L.; Kragl, U. Carbohydr. Res. 1998, 305, 475–481.
11. MacDonald, D. L. J. Org. Chem. 1962, 27, 1107.
12. Percival, M. D.; Withers, S. G. Can. J. Chem. 1988, 66, 1970–1972.
13. Niggemann, J.; Lindhorst, T. K.; Walfort, M.; Laupichler, L.; Sajus, H.; Thiem, J. Carbohydr. Res. 1993, 246, 173–183.
14. Shibaev, V. N.; Kusov, Y. Y.; Kucár, S.; Kochetkov, N. K. Izv. Akad. Nauk. SSSR 1973, 4, 922–926.
15. Kucár, S.; Zámocky, J.; Bauer, S. Chem. Zvesti 1974, 28, 115–119.
16. Christian, R.; Schreiner, E.; Zbiral, E.; Schulz, G. Carbohydr. Res. 1989, 194, 49–61.
17. Williams, J. M.; Richardson, A. C. Tetrahedron 1967, 23, 1369.
18. Kim, S.; Yi, K. Y. J. Org. Chem. 1986, 51, 2613–2615.
19. Sim, M.; Kondo, H.; Wong, C. J. Am. Chem. Soc. 1993, 115, 2260–2267.
20. McDowell, W.; Schwartz, R. T. FEBS Lett. 1989, 243, 413–416.
21. Shibaev, V. N. Pure Appl. Chem. 1978, 50, 1421–1436.
22. Wilson, I. B. H.; Webberley, M. C.; Revers, L.; Flitsch, S. L. Biochem. J. 1995, 310, 909–916.
23. Ryll, T.; Wagner, R. J. Chromatography 1991, 570, 77–88.
24. Kucár, S.; Zámocky, J.; Zemek, J.; Bauer, S. Chem. Zvesti 1976, 32, 414–419.
25. Niggemann, J.; Thiem, J. Liebigs Ann. Chem. 1992, 535–358.