An efficient chemoenzymatic route to dihydroxyacetone phosphate from glycidol for the in situ aldolase-mediated synthesis of monosaccharides

Article abstract of DOI:10.1016/j.tetlet.2006.03.036

We report a new two-step procedure that uses inexpensive rac-glycidol to obtain valuable dihydroxyacetone phosphate (DHAP), a building block for the synthesis of monosaccharide analogues.

Full text of DOI:10.1016/j.tetlet.2006.03.036

Tetrahedron Letters 47 (2006) 3261–3263
An efficient chemoenzymatic route to dihydroxyacetone
phosphate from glycidol for the in situ aldolase-mediated
synthesis of monosaccharides
a
b
c
a,
*
Franck Charmantray, Phillipe Dellis, Soth Samreth and Laurence Hecquet
a
Laboratoire SEESIB, UMR 6504 CNRS, Universit e´ Blaise Pascal, 24 Avenue des Landais, 63177 Aubi e` re cedex, France
b
Synkem S.A.S, 47, rue de Longvic, F-21300 Chen oˆ ve, France
c
FournierPharma, 50, rue de Dijon, F-21121 Daix, France
Received 10 February 2006; revised 2 March 2006; accepted 7 March 2006
Available online 23 March 2006
Abstract—We report a new two-step procedure that uses inexpensive rac-glycidol to obtain valuable dihydroxyacetone phosphate
DHAP), a building block for the synthesis of monosaccharide analogues.
Ó 2006 Elsevier Ltd. All rights reserved.
(
1
. Introduction
multistep purification procedures and the need to pro-
tect functional groups. A promising alternative would
be a short and inexpensive DHAP synthesis followed
by an in situ aldolisation reaction catalysed by aldolase.
Asymmetric framework construction by enzymatic
carbon–carbon bond formation is an attractive
alternative to conventional chemical methods. It offers
stereochemical control, mild conditions and needs no
protecting group. In the stereoselective synthesis of
carbohydrates and carbohydrate mimetics, aldolases
and in particular DHAP-utilising aldolases such as
D-fructose-1,6-bisphosphate aldolase, D-tagatose-1,6-
bisphosphate aldolase, L-fuculose-1-phosphate aldolase
and L-rhamnulose-1-phosphate aldolase have proved
to be useful catalysts. Enzymes of this family catalyse
the aldol addition of dihydroxyacetone phosphate
By this approach, DHAP was recently obtained by dihy-
droxyacetone (DHA) phosphorylation catalysed by
1
,2
6
DHA kinase with ATP. One drawback of this proce-
dure was the need for a system of ATP regeneration
7
,8
from acetyl phosphate. Sheldon and co-workers also
describe a coupled multienzymatic system to prepare
carbohydrates from glycerol. The key step in this stra-
tegy was the efficient DHAP synthesis from L-glycerol-
3-phosphate (L-G-3-P) catalysed by L-glycerophosphate
9
(
DHAP) with a wide range of aldehydes to form a
oxidase (L-GPO) in the presence of catalase. The
new C–C bond, creating two hydroxylated stereogenic
centres. D-Fructose-1,6-bisphosphate aldolase (FruA,
EC 4.1.2.13), the aldolase most widely used for the
synthesis of ketoses and analogues, produces an aldol
authors elegantly prepared L-G-3-P by phosphorylation
of glycerol with inexpensive pyrophosphate catalysed by
phytase. However, the quantitative conversion of pyro-
phosphate into D,L-G-3-P required a glycerol concentra-
tion of 95% (v/v), restricting the synthesis to the
hydrophobic carbohydrates.
3
adduct with D-threo configuration. However, wider
practical application of aldolases requires cheap and
4
ready access to DHAP. The literature and our own
5
earlier work has recently described efficient labora-
Here we describe a short, practical and efficient chemo-
enzymatic synthesis of DHAP from inexpensive
rac-glycidol (Scheme 1). Our two-step procedure was
performed in one pot and successfully applied to the
synthesis of 5-halo-D-xylulose mediated in situ by FruA.
Our route was in four steps: (i) regioselective opening of
the rac-glycidol epoxide ring with phosphate to generate
D,L-G-3-P; (ii) oxidation of L-G-3-P in DHAP under
tory-scale chemical syntheses of DHAP. These strategies
are made complicated by expensive or toxic reagents,
Keywords: DHAP; Multienzymatic synthesis; Aldolase; C–C Bond
formation.
*
Corresponding author. Tel.: +33 4 73 407871; fax: +33 4 73
07717; e-mail: Laurence.HECQUET@univ-bpclermont.fr
4
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.03.036

3262
F. Charmantray et al. / Tetrahedron Letters 47 (2006) 3261–3263
O
X 1. FruA, pH 6.8
H
OH O
O
2. Pase, pH 4.7
Na HPO4
OH
OH
L
L-GPO, pH 6.8
2
X
OH
HO
OPO H
3 2
O
HO
HO
OPO H2
3
OH
DHAP
1
/2 O2
H O
2
2
5-halo-D-xylulose
X = Cl, Br
OH
D
OPO H2
3
H O
2
Scheme 1. Multienzyme system for one-pot 5-halo-D-xylulose synthesis.
oxygen mediated by L-GPO, coupled with hydrogen
peroxide decomposition by catalase; (iii) FruA catalysed
aldol addition of DHAP on 2-haloacetaldehyde, and (iv)
used in the previous step. In the latter case, whatever
the quantity of L-GPO, the oxidation reaction stopped
after a few minutes. By contrast, when Na HPO was
2
4
5
-halo-D-1-phosphate hydolysis catalysed by acid phos-
used, L-G-3-P was fully converted into DHAP after oxi-
dation by L-GPO. This two-step procedure from rac-
glycidol gave DHAP in 28% overall yield (maximum
theoretical yield was 50%). This compound was thus
readily available for in situ coupled aldol reaction.
phatase (Pase).
First step: The controlled opening of the rac-glycidol
epoxide ring with various phosphate sources in refluxing
water gave D,L-G-3-P in moderate to good yields (Table
1
). The reaction was monitored by L-G-3-P detection
Third step: FruA from rabbit muscle (RAMA) was added
when the oxidation reaction with L-GPO was complete.
1
2
with L-GPO and subsequent assay for quantification
1
0
of the hydrogen peroxide released (equal amounts of
D-isomer were assumed to be formed). We showed that
the opening of rac-glycidol was pH-dependent with an
optimum pH above 10 obtained with Na (K )HPO
4
RAMA was used to catalyse the aldol addition of DHAP
onto 2-chloroacetaldehyde (or 2-bromoacetaldehyde)
leading to aldol adduct 5-chloro-D-xylulose-1-phosphate
(or 5-bromo-D-xylulose-1-phosphate). The reaction was
2
2
1
3
31
and Na PO . In these cases, L-G-3-P was obtained in
monitored by in situ C and P NMR (see Supplemen-
tary data). We observed total DHAP consumption
and appearance of a single product. This compound
was identified as 5-chloro-D-xylulose-1-phosphate (or
3
4
5
0–60% yield (entries 1, 2, 4 and 7). We also studied
the influence of the counterion effect and the glycidol/
phosphate ratio. Similar results were observed when
K was replaced by Na (entries 3 and 4). A threefold
+
+
5-bromo-D-xylulose-1-phosphate) by its NMR signal at
2
excess of rac-glycidol gave only a 5% increase in yield
d = 68 ppm, J = 3.8 Hz for C carbon coupled with
5
(
entries 4 and 5). Surprisingly, the yield of D,L-G-3-P
the phosphorus. The reaction yield was quantitative as
decreased when the concentrations of the two reagents
increased (entries 4 and 6). For these reasons, we
decided to use stoichiometric amounts of rac-glycidol
and Na HPO or Na PO as phosphate sources (entries
determined by NMR.
Fourth step: The dephosphorylation of 5-chloro-D-xylu-
lose-1-phosphate (or 5-bromo-D-xylulose-1-phosphate)
was carried out by the addition of acid phosphatase
2
4
3
4
4
and 7).
(Pase) after adjusting the pH to 4.7. The final products
Second step: After conversion of glycidol into D,L-G-3-P
5-chloro-D-xylulose (or 5-bromo-D-xylulose) were char-
acterised as the major compounds in the reaction
mixture, while glycerol was the sole by-product. This
latter came from the hydrolysis of the phosphate group
of D-G-3-P (obtained from D,L-G-3-P resolution cata-
lysed by L-GPO in the second step) by Pase. After
purification by column chromatography on silica
gel, 5-chloro-D-xylulose and 5-bromo-D-xylulose were
recovered as pure compounds in 47% and 12% yield,
respectively, from L-G-3-P.
in the presence of Na HPO or Na PO , the pH was
2
4
3
4
adjusted to 6.8 to optimise GPO activity. Catalase was
added first (1200 units/mmol of L-G-3-P). The reaction
was initiated by addition of L-GPO (30 units/mmol of
L-G-3-P). DHAP was assayed with NADH-consuming
1
1
a-glycerophosphate dehydrogenase. The behaviour of
L-GPO was different when Na HPO or Na PO was
2
4
3
4
Table 1. Reaction conditions for D,L-glycerol-3-phosphate (D,L-G-3-
P) synthesis from rac-glycidol
a
In conclusion, we report an attractive two-step synthesis
of DHAP from rac-glycidol, a cheap commercially
available starting material. The controlled opening of
Entry rac-Glycidol Phosphate Phosphate pH D,L-G-3-P
b
(M)
source
source (M)
Yield (%)
1
2
3
4
5
6
7
0.5
0.5
0.5
0.5
1.5
2.0
0.5
H
KH
K
Na
Na
Na
Na
3
PO
4
0.5
0.5
0.5
0.5
0.5
2.0
0.5
2
4
16
26
the rac-glycidol epoxide ring with Na HPO in water
2
PO
4
2
4
2
HPO
4
10 50
10 55
10 60
10 42
12 55
gave D,L-G-3-P in 55% yield. L-G-3-P conversion to
DHAP by means of L-GPO and catalase was found to
be quantitative. We note that L-GPO and catalase could
2
2
2
3
HPO
HPO
HPO
4
4
4
1
3
be co-immobilised as described earlier. We show that
DHAP can be used in situ as a donor substrate of FruA
in the presence of either 2-chloro- or 2-bromoacetalde-
hydes as acceptor substrates for the synthesis of either
PO
4
a
Reaction carried out in refluxing water for 3 h.
Yields were determined enzymatically.
b
10

F. Charmantray et al. / Tetrahedron Letters 47 (2006) 3261–3263
3263
5
-chloro- or 5-bromo-D-xylulose, suitable intermediates
Supplementary data
for the synthesis of substituted xylulose analogues in
3
1
13
the C position. As DHAP aldolases display a broad
Supplementary data for P and C NMR spectra
associated with this article can be found, in the online
version, at doi:10.1016/j.tetlet.2006.03.036.
5
specificity towards acceptor substrates, this four-step
strategy can be generally applied to the synthesis of var-
ious analogues of monosaccharides.
References and notes
2
. General procedure
1
2
. (a) Gijsen, H. J. M.; Qiao, L.; Fitz, W.; Wong, C.-H.
Chem. Rev. 1996, 96, 443–474; (b) Wymer, N.; Toone, E.
J. Curr. Opin. Chem. Biol. 2000, 4, 110–119.
. (a) Andr e´ , C.; Gu e´ rard, C.; Hecquet, L.; Demuynck, C.;
Bolte, J. J. Mol. Catal. B, Enzym. 1998, 5, 459–466; (b)
Crestia, D.; Gu e´ rard, C.; Veschambre, H.; Hecquet, L.;
Demuynck, C.; Bolte, J. Tetrahedron: Asymmetry 2001,
To a 10 mL solution of rac-2,3-epoxypropanol (0.38 g,
mmol) in distilled water was added solid disodium
hydrogen phosphate Na HPO (0.74 g, 5 mmol). The
mixture was heated at 100 °C for 3 h, and assayed for
L-G-3-P content. The yield was 55% from (S)-2,3-epoxy-
propanol (and 28% from rac-2,3-epoxypropanol). After
cooling to room temperature, the pH was adjusted to 6.8
with 3 N HCl and 0.1 mL GPO/catalase mixture
5
2
4
1
2, 869–876.
3
. Brockamp, H.; Kula, M. Appl. Microbiol. Biotechnol.
1990, 34, 287–291.
(
(
45 units/1800 units), 70 lL FruA from rabbit muscle
20 units) and 2-haloacetaldehyde (2 mmol) were succes-
4. Meyer, O.; Rohmer, M.; Grosdemange-Billiard, C. Tetra-
hedron Lett. 2004, 45, 7921–7923.
5
6
7
. Charmantray, F.; El Blidi, L.; Gefflaut, T.; Hecquet, L.;
Bolte, J.; Lemaire, M. J. Org. Chem. 2004, 69, 9310–
sively added. The reaction proceeded with stirring at
room temperature overnight. The pH was adjusted to
4
9
312.
.7 with 1 N HCl and 50 units of acid phosphatase
Pase) was added. The reaction mixture was stirred
. S a´ nchez-Moreno, I.; Garc ´ı a-Garc ´ı a, J. F.; Bastida, A.;
Garc ´ı a-Junceda, E. Chem. Commun. 2004, 14, 1634–
(
for a further 24 h at room temperature. The pH was
adjusted to 7.0 with 1 N NaOH, and MeOH (30 mL)
was added to the solution. The resulting precipitate
was removed by filtration through Celite. The filtrate was
concentrated under vacuum and the residue then under-
went silica gel chromatography with cyclohexane/
AcOEt (2:8) as an eluent. 5-Chloro-D-xylulose was
recovered in 47% yield and 5-bromo-D-xylulose in 12%
yield.
1
635.
. Schoevaart, R.; van Rantwijk, F.; Sheldon, R. A. Chem.
Commun. 1999, 24, 2465–2466.
8. Schoevaart, R.; van Rantwijk, F.; Sheldon, R. A. J. Org.
Chem. 2000, 65, 6940–6943.
9. Fessner, W.-D.; Sinerius, G. Angew. Chem., Int. Ed. Engl.
1
994, 33, 209–212.
1
1
1
1
0. Bergmeyer, H. U. In Methods in Enzymatic Analysis;
Verlag Chemie, 1984; Vol. 4, pp 342–350.
1. Bergmeyer, H. U. In Methods in Enzymatic Analysis;
Verlag Chemie, 1984; Vol. 2, pp 146–147.
2. The concomitant reactions of oxidation and aldolisation
were slightly less yield effective.
Acknowledgements
3. Kr a¨ mer, L.; Steckhan, E. Tetrahedron 1997, 53, 14645–
14650.
We thank Fournier Pharma, for financial support.