A short route to nucleoside diphosphate activated D- and L-hexoses

Article abstract of DOI:10.1016/S0040-4039(01)00352-5

Leloir transferases utilise nucleoside diphosphate sugars, which are notoriously difficult to synthesise and handle. Starting off from D- or L-configurated glycals, a facile synthesis of nucleotide sugars by epoxidation and direct coupling with uridine di

Full text of DOI:10.1016/S0040-4039(01)00352-5

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 2973–2975
A short route to nucleoside diphosphate activated
D- and
L-hexoses
Christiane Ernst and Werner Klaffke*
Organisch-Chemisches Institut der Westf a¨ lischen Wilhelms-Universit a¨ t, Corrensstr. 40, D-48149 M u¨ nster, Germany
Received 13 February 2001; accepted 26 February 2001
Abstract—Leloir transferases utilise nucleoside diphosphate sugars, which are notoriously difficult to synthesise and handle.
Starting off from
D- or L-configurated glycals, a facile synthesis of nucleotide sugars by epoxidation and direct coupling with
uridine diphosphate is described. © 2001 Published by Elsevier Science Ltd.
One, if not the greatest obstacle in using Leloir trans-
ferases for glycosylation reactions is their absolute
requirement for activated sugars. The syntheses of these
are non-trivial and their handling is requiring profound
preparative skills. Contributions, in which the natural
pathway towards activated sugars has been exploited
are nowadays classical examples of the elegant combi-
inspired by reports on the regioselective attack of 1,2-
anhydro sugars using protic acids, such as carboxy-
lates, phosphate mono- and diesters. Such reagents
gave rise to high a/b-selectivities at good overall yields,
depending on the used solvent.
6
6,7
Starting off from easily obtainable
D
-glucal (1),
D-
1
8
nation of biochemistry and carbohydrate synthesis.
galactal (2), and
L
-rhamnal (7) the respective a-
configurated 1,2-anhydro sugars 3, 4 and 8 were
The advantages, however, of chemical syntheses are
embedded in a broader applicability since activated
sugars can be envisaged, which are complementing the
repertoire of hexoses abundant in mammalian systems.
This is paid for by a more cumbersome preparative
procedure, the synthesis of a glycosyl phosphate, which
is subsequently coupled to a nucleotide to yield the
obtained quantitatively according to the standard
9
procedures (Scheme 1).
Special attention had to be paid to the preparation of
the diphosphate, which was converted to the free acid
+
by passing through a column of DOWEX-50 (H ) and
titrated with tetrabutyl ammonium hydroxide solution
2
diphosphate. Yields are generally quite disappointing
(
40% w/v) to pH 6. Contrary to the procedure reported
at this final step, which unfortunately also marks the
end point of the synthetic assembly line. We were
prompted by this dilemma to embark on the direct
coupling of glycosyl donors with a nucleoside diphos-
phate. This idea has a short history in the carbohydrate
community, which is examplified by contributions of
3b
earlier, pH in our examples showed to be the parame-
ter to which utmost attention had to be given. At pH 7
all charges of the diphosphate were screened by tetra-
butyl ammonium counterions, hence the coupling yields
were dramatically decreased. The other corner of the
operating window was marked by pH 5. Here, initially
product formation was observed by MS, but extensive
decomposition into cyclic hexosyl 1,2-phosphates and
UMP was taking place.
3
4
5
Hindsgaul, Hanessian, and Schmidt. These groups
have approached the problem by reacting their respec-
tively favoured glycosyl donors with nucleoside diphos-
phates or nucleotides, as in the latter case.
10
Thus, the lyophilised TBA salt (UDP×1.7 TBA,
titrated to pH 6) was dispersed in anhydrous
dichloromethane (approx. 3 ml/mmol) and added to the
solution of the glycosyl donor. In a stream of argon,
the reaction mixture was concentrated to yield a syrup,
which was stirred for not more than 20 h. After this
period the reaction was complete and worked up by
removing the organic solvent, filtration through Celite
Here, we present a very general layout for the synthesis
of nucleoside diphosphate activation, which was
Keywords: carbohydrates; nucleoside diphosphate sugars; Leloir-
transferase donors; 1,2-anhydride.
*
Corresponding author. Fax: +49 251 8336501; e-mail:
werner.klaffke@uni-muenster.de
0
040-4039/01/$ - see front matter © 2001 Published by Elsevier Science Ltd.
PII: S0040-4039(01)00352-5

2
974
C. Ernst, W. Klaffke / Tetrahedron Letters 42 (2001) 2973–2975
O
R2
OBn
O
R2
OBn
O
R2 OBn
O
NH
i
ii
R1
BnO
R1
BnO
R1
BnO
O
P
OR
P
O
O
N
OH
O
O
O
O
RO
O
1
2
1
2
1
2
1
R = OBn, R = H
3 R = OBn, R = H
5 R = OBn, R = H
OH
OH
1
2
1
2
1
2
2
R = H, R = OBn
4 R = H, R = OBn
6 R = H, R = OBn
O
NH
O
P
OR
P
O
O
O
N
O
O
O
H3C
BnO
ii
O
i
3
H C
O
H C
3
RO
OH
O
H
O
BnO
H
BnO
BnO
BnO
BnO
OH
OH
7
8
9
Scheme 1. Syntheses of nucleotide sugars starting from D- and L-glycal precursors. (i) DMDO, CH Cl , 0°C, 1 h, quant.; (ii) UDP
2 2
TBA salt, CH Cl , rt, 20 h (40–60%).
2
2
with buffered solution (30 mM NH HCO ) and purified
References
4
3
by ion-exchange chromatography using DEAE cellu-
−
3
lose (HCO form, gradient elution 30 mM“400 mM
1. Wong, C. H.; McGarvey, G. J. Liebigs Ann. Chem. 1997,
NH HCO ).
1059–1074.
4
3
2
3
. Moffatt, J. G. Meth. Enzymol. 1966, 136–142.
. (a) Uchiyama, T.; Hindsgaul, O. J. Carbohydr. Chem.
998, 17, 1181–1190; (b) Arlt, M.; Hindsgaul, O. J. Org.
Chem. 1995, 60, 14–15.
After pooling of the appropriate fractions and lyophili-
sation, UDP 3,4,6-tri-O-benzyl- -glucose and -galac-
tose, 5 and 6, were isolated as white powders.
Interestingly, -gluco configurated 9 could not be iso-
1
D
4
5
6
. Hanessian, S.; Lu, P.-P.; Ishida, H. J. Am. Chem. Soc.
L
1
998, 120, 13296–13300.
. Schmidt, R. R.; Braun, H.; Jung, K.-H. Tetrahedron Lett.
992, 33, 1585–1588.
. Timmers, C. M.; van Straten, N. C. R.; van der Marel,
G. A.; van Boom, J. H. J. Carbohydr. Chem. 1998, 17 (3),
471–481.
lated from excess UDP as such, but instead had to be
treated with alkaline phosphatase (pH 7.8, 37°C) prior
to ion-exchange chromatography. All benzyl ethers 5,
1
6, and 9 gave clean analytical and spectroscopic analy-
ses; the isolated yields were typically 40–60%, the a/b
11
ratio varied around 1:1. (typical data of D- and L-sug-
7
8
9
. Plante, O. J.; Andrade, R. B.; Seeberger, P. H. Org. Lett.
13
ars are given in Ref. ). The benzylated UDP hexoses
were stable upon storage and could be deblocked by
1
999, 1, 211–214.
. Shull, B. K.; Wu, Z.; Koreeda, M. J. Carbohydr. Chem.
996, 15 (8), 955–964.
. Halcomb, L.; Danishefski, S. J. J. Am. Chem. Soc. 1989,
11, 6661–6666.
3b
hydrogenation as was described earlier.
1
For the
D-series it has been shown that enzymatic
1
transformations by Leloir transferases are not ham-
1
1
0. UDP trisodium salt (100 mg, 0.2 mmol) was dissolved in
water (1 ml) and the solution was passed through
3
b
pered by the presence of anomeric mixtures. In the
-hexosyl phosphate series, b-configurated glycosyl
donors are substrates for the respective -sugar trans-
ferases, the stereochemistry of which is difficult to
L
+
DOWEX 50W X8 (H ). Fractions containing protonated
L
UDP were detected by UV absorption, pooled and
titrated using aqueous Bu N(OH) (40% w/v) to pH 6.
4
1
2
obtain by conventional methodology.
The resulting solution was lyophilised to yield UDP
Bu N salt as a white powder.
4
In summary, this method comprises of a very fast and
simple reagent-free procedure for the synthesis of stable
a- and b-configurated UDP sugars, which can easily be
1. Benzylated UDP sugars were obtained as a white pow-
ders upon lyophilisation, 5 (ꢀ50%, a/b ratio 1:1), 6
(ꢀ64%, a/b ratio 1:0.75), 9 (47%, a/b ratio 0.8:1).
12. Barber, G. A.; Behrman, E. J. Arch. Biochem. Biophys.
1991, 288, 239–242.
extended to functionalised glycals of either the
D
- or
L
-series. Further investigations are presently under way
to exploit this method for the synthesis of libraries of
nucleoside diphosphates.
13. Selected physical data of compounds 5 and 9: compound
1
5: H NMR (600 MHz, D O): l=7.77 (d, 1H, H-6a),
2

C. Ernst, W. Klaffke / Tetrahedron Letters 42 (2001) 2973–2975
2975
7
(
5
.74 (d, 1-H, H-6b), 7.15–7.30 (m, 26H, arom.), 6.87–6.95
m, 4H, arom.), 5.85 (m, 1H, H-1%a), 5.83 (d, 1H, H-1%b),
.80 (d, 1H, H-5a), 5.72 (d, 1H, H-5b), 5.45 (dd, 1H,
NMR (600 MHz, D O): l=7.81 (d, 1-H, H-6b), 7.77 (d,
2
1H, H-6a), 7.16–7.33 (m, 20H, arom.), 5.83 (d, 1H,
H-1%b), 5.82 (d, 1H, H-5b), 5.78 (d, 1H, H-1%a), 5.76 (d,
1H, H-5a), 5.39 (dd, 1H, H-1%%a), 4.85 (ddꢀt, 1H, H-
H-1%%a), 4.87 (ddꢀt, 1H, H-1%%b), 4.74–4.80 (2*d, 2H,
-
CH , benzyl), 4.58–4.60 (2*d, 2H, -CH , benzyl), 4.42–
1%%b), 4.40–4.80 (8*d, 8H, -CH , benzyl), 4.52–4.60 (m,
2
2
2
4
.50 (m, 2H, H-3%a and b), 4.24–4.46 (4*d, 8H, -CH2,
2H, H-3%a and b), 4.22 (m, 1H, H-2%b), 4.16 (m, 1H,
H-2%a), 4.03–4.21 (m, 6H, H-4%, H-5%a, H-5%b, a and b),
3.89 (m, 1H, H-5%%a), 3.73 (ddꢀt, 1H, H-3%%a), 3.53 (m,
1H, H-2%%a), 3.50 (ddꢀt, 1H, H-3%%b), 3.44 (m, 1H, H-
5%%b), 3.37 (ddꢀt, 1H, H-2%%b), 3.12–3.20 (2*dd, 2H, H-
4%%a and b), 1.18 (d, 3H, -CH b), 1.15 (d, 3H, -CH a)
benzyl), 4.10 (dd, 1H, H-2%a), 4.14–4.20 (m, 2H, H-4%a
and b), 4.11–4.19 (m, 1H, H-2%b), 4.05–4.10 (m, 4H,
H-5%a and H5%b a and b), 3.87 (m, 1H, H-5%%a), 3.77
(
ddꢀt, 1H, H-3%%a), 3.60 (m, 4H, H-6%%a and H-6%%b a and
b), 3.54 (dd, 1H, H-2%%a), 3.53 (m, 1H, H-3%%b), 3.45 (m,
H, H-4%%a and H-4%%b a and b), 3.39 (ddꢀt, 1H, H-2%%b)
ppm; J5,6=8.0, J1%,2%=4.3, J1%%a,2%%a=3.2, J1%%a,P=7.0,
J2%%a,3%%a=8.9, J3%%a,4%%a=9.4, 4%%a,5%%a=10.4, J1%%b,2%%b=7.9,
3
3
2
ppm; J5,6=8.2, J1%,2%=4.2, J1%%a,2%%a=2.8, J1%%a,P=7.8,
J2%%a,3%%a=9.7, J3%%a,4%%a=9.4, J4%%a,5%%a=9.4, J5%%a,CH3=5.7,
J
J1%%b,2%%b=8.0, J1%%b,P=7.9,
J4%%b,5%%b=9.4, J5%%b,CH3=5.7 Hz; C NMR (90 MHz,
J
=9.1, J3%%b,4%%b=9.4,
2%%b,3%%b
13
13
J1%%b,P=7.9, J2%%b,3%%b=8.9 Hz; C NMR (90 MHz, D O):
2
l=168.94, 168.77 (C-4, a/b), 154.63, 154.30 (C-2, a/b),
D O): l=188.60, 188.56 (C-2, a/b), 169.02, 168.90 (C-4,
2
1
44.52, 144.20 (C-6, a/b), 139.5–140.3 (arom.), 131.1–
31.7 (arom.), 105.56, 105.34 (C-5, a/b), 100.81 (C-1%%b,
a/b), 144.54, 144.42 (C-6, a/b), 139.8–140.3 (arom.),
131.2–131.7 (arom.), 105.48, 105.57 (C-5, a/b), 100.57
(C-1%%b), 98.73 (C-1%%a), 91.27, 91.22 (C-1%, a/b), 86.17,
86.02, 85.40, 85.21 (C-3%%, C-4%, a/b), 78.58 (C-2%%b), 77.29
(C-2%%a, JP-O-C-C=7.5 Hz), 78.09, 76.67, 74.35, 72.54,
1
JP-O-C=6.7 Hz), 98.66 (C-1%%a, JP-O-C=6.7 Hz), 91.49,
1.19 (C-1%, a/b), 86.28, 85.81, 85.76, 84.76 (C-3%%, C-4%,
9
a/b), 79.44 (C-2%%b, JP-O-C-C=7.8 Hz), 77.77 (C-2%%a, JP-O-
C-C=7.8 Hz), 78.60, 78.06, 77.11, 76.87, 76.72, 76.69
72.42, 72.24 (C-3%, C-4%%, C-5%%, a/b, -CH , benzyl), 71.12
2
(
7
C-4%%, a/b, -CH , benzyl), 75.79, 75.08 (C-3%, a/b), 73.79,
(C-2%, a/b), 67.83, (C-5%, a/b, JP-O-C=4.1 Hz), 19.90 (C-6%%,
2
−
2.59, 72.27, 72.19 (C-5%%, C-6%%, a/b), 70.41, 70.04 (C-2%,
a/b) ppm; MS (ESI-MS, nanospray) (m/z) 729 [M] , 364
31
2−
a/b), 67.83, 67.61 (C-5%, a/b, JP-O-C=6.7 Hz) ppm;
NMR (242.5 MHz, D O): l=−10.57 (2*d, P1,a/b), -12.43
P
[M] . Upon hydrogenolysis (cf. lit. 3b), compounds 5
and 6 were identical in all spectroscopic aspects to com-
mercial samples of UDP glucose and UDP galactose
(Sigma).
2
(
2*d, P2,a/b
)
ppm;
J
=19.6 Hz; MS (ESI-MS,
P-O-P
−
2−
1
nanospray) (m/z) 835 [M] , 417 [M] . Compound 9: H
.
.