Synthesis of non-natural carbohydrates from glycerol and aldehydes in a one-pot four-enzyme cascade reaction

Article abstract of DOI:10.1039/c1gc15429j

A simple procedure has been developed for the synthesis of enantio- and diastereomerically pure carbohydrate analogues from glycerol and a variety of aldehydes in one pot using a four-enzyme cascade reaction. As a proof of concept of the usefulness of this enzymatic catalytic cascade the naturally occurring azasugar d-fagomine was synthesized. This work highlights the potential value of using enzymes in cascade reactions to selectively form complex products that by previous traditional organic chemistry could only be obtained via repeated isolation and purification of intermediates.

Full text of DOI:10.1039/c1gc15429j

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2011, 13, 2895
www.rsc.org/greenchem
PAPER
Synthesis of non-natural carbohydrates from glycerol and aldehydes in a
one-pot four-enzyme cascade reaction
Lara Babich,†a Lieke J. C. van Hemert,†b Aleksandra Bury,a Aloysius F. Hartog,a Pierpaolo Falcicchio,c
John van der Oost,c Teunie van Herk,a Ron Wever*a and Floris P. J. T. Rutjesb
Received 19th April 2011, Accepted 28th June 2011
DOI: 10.1039/c1gc15429j
A simple procedure has been developed for the synthesis of enantio- and diastereomerically pure
carbohydrate analogues from glycerol and a variety of aldehydes in one pot using a four-enzyme
cascade reaction. As a proof of concept of the usefulness of this enzymatic catalytic cascade the
naturally occurring azasugar D-fagomine was synthesized. This work highlights the potential
value of using enzymes in cascade reactions to selectively form complex products that by previous
traditional organic chemistry could only be obtained via repeated isolation and purification of
intermediates.
kinases are needed and that regeneration of ATP is required. The
Introduction
group of Sheldon was among the first to explore the formation
In the search for efficient, sustainable synthetic strategies,
of a heptose sugar using a multi-enzyme system in which a
chemists increasingly pursue approaches that are inspired by
phosphorylated intermediate was continuously generated using
Nature.1 The ultimate goal is to mimic biosynthetic pathways
and construct complex products via ingenious sequences of
multiple enzymatic conversions in one pot, preferably with both
pyrophosphate as a phosphate donor and in which the DHAP
formed was coupled enzymatically to an aldehyde. However, a
pH switch was needed to turn off the activity of one of the
high chemo- and stereoselectivity.2 This concept is demonstrated
enzymes and also a very high concentration of glycerol (85%)
in the synthesis of complex carbohydrate building blocks,3 which
had to be present.13
may serve amongst others as intermediates for iminosugars.4
Besides impressive achievements in the chemical synthesis
of carbohydrate building blocks using asymmetric catalysis5
In our group, we recently reported the efficient phosphoryla-
tion of dihydroxyacetone by the acid phosphatase from Shigella
flexneri (PhoN-Sf) using pyrophosphate (PPi) as the phosphate
and organocatalysis,6 use of aldolases could provide direct
source. In addition, it was shown that this phosphorylation
stereoselective access to carbohydrate structures without going
process could be applied in one pot in combination with
through protecting group manipulations.7 In order to explore
dihydroxyacetone phosphate (DHAP)-dependent rabbit muscle
the full potential of existing aldolases in C–C bond formation,
aldolase (RAMA) to give the desired aldol product.14 Motivated
it is crucial to have facile access to phosphorylated substrates.8
Dihydroxyacetone phosphate (DHAP) is one of the key com-
pounds in aldol condensation using DHAP dependant aldolases
by this outcome, we now report a one-pot synthesis of chiral
carbohydrate fragment 1 by the simultaneous action of four
different enzymes starting from inexpensive glycerol 2 and an
and the research groups of Whitesides, Wong and Fessner have
appropriate aldehyde (Scheme 1).
used several enzymatic procedures to arrive at the formation
of DHAP starting from glycerol,9 glycerol phosphate10 or
dihydroxyacetone.11,12 These methods have the disadvantage that
aUniversity of Amsterdam, J. H. van’t Hoff Institute for Molecular
Scheme 1 One-pot retrosynthesis of carbohydrate fragments 1.
Sciences, Science Park 904, 1098 XH, Amsterdam, The Netherlands.
E-mail: r.wever@uva.nl; Fax: +31 5255604; Tel: +31 5255110
bRadboud University Nijmegen, Institute for Molecules and Materials,
Heyendaalseweg 135, 6525 AJ, Nijmegem, The Netherlands.
E-mail: F.Rutjes@science.ru.nl; Fax: +31 243653393; Tel: +31
243653202
The newly envisaged cascade was based on initial phos-
phorylation of glycerol into glycerol-3-phosphate,9,13 followed
by L-glycerol-3-phosphate oxidase (GPO)-mediated oxidation
to DHAP. Subsequent RAMA-catalyzed C–C bond formation
and dephosphorylation with PhoN-Sf should then give rise to
target molecules 1. Having PhoN-Sf available which operates at
cWageningen University, Laboratory of Microbiology, 6703 CT,
Wageningen, The Netherlands
† Both authors contributed evenly to this manuscript.
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2895–2900 | 2895
©

View Article Online
Table 1 Optimization of the four-enzyme cascade
pH 6, we envisioned that we should be able to carry out the
glycerol-DHAP cascade in a single pot with all enzymes acting
simultaneously.14,15 Furthermore, we show that it is possible to
obtain all the stereocomplementary products by using all four
DHAP-dependent aldolases present in nature.
Results and discussion
PhoN-Sf
GPO
Conv.
Reaction
time (h)
Entry
R
Product (U mL-1) (U mL-1)
(%)a
Scheme 2 illustrates the sequence of events occurring in the
cascade. The PhoN-Sf-mediated phosphorylation of glycerol 2
with the phosphate donor PPi produces glycerol-3-phosphate
3 in a chemoselective manner, due to the known selectivity
of PhoN-Sf for primary hydroxy functions (Scheme 2).15 The
L-enantiomer is then oxidized by GPO in the presence of
oxygen to produce DHAP 4.10 The oxidation takes place with
concomitant formation of hydrogen peroxide, which is converted
into water and oxygen by a third enzyme, catalase. In the
next step of the cascade, DHAP reacts with the aldehyde
catalyzed by RAMA to provide the phosphorylated aldol
product 5. Under the reaction conditions, the aldol product
is ultimately dephosphorylated by PhoN-Sf leading to enantio-
and diasteromerically pure diol 1. The resulting enzyme-bound
phosphate may then be either transferred to water (hydrolysis)
or to glycerol (transphosphorylation).14 Hydrolysis is generally
preferred over transphosphorylation so that Pi is released unless
a suitable substrate is present. This essentially irreversible
step fortuitously shifts the thermodynamic equilibrium of the
cascade to aldol product 1 once PPi becomes exhausted.
1
2
3
4
5
6
7
8
Et
Et
Et
Et
Et
Et
Et
1a
1a
1a
1a
1a
1a
1a
1.5
1.5
1.5
1.5
3.0
3.0
3.0
3.0
2
5
42
50
58
84
24
24
24
24
8
8
8
10
50
50
50
50
50
87
95b
100b,c
48
i-Pr 1b
8
a Conversions were based on the aldehyde and determined using HPLC
on aliquots of the reaction mixture. b Reaction was performed at 20 ◦C.
c Reaction was performed in a closed vessel.
the phosporylated product, the dephosphorylated end product,
and the amount of aldehyde converted. This allowed us to study
in detail the effect of changing the amount of activity of GPO
from 2 U mL-1 to 50 U mL-1 in the cascade. Optimal conversion
was obtained with 50 U mL-1 GPO (entry 4).
Encouraged by these results the effect of changing the activity
of the acid phosphatase was also investigated. When the activity
of PhoN-Sf in the cascade was doubled, within 8 h a conversion
of 87% was reached (entry 5). Changing the activity of GPO or
PhoN-Sf had no effect on the diastereoselectivity, a d.r. of 14 : 1
((3S,4R) vs. (3S,4S)) was observed in all cases. Lowering the
temperature to 20 ◦C resulted in 95% conversion and increased
diastereoselectivity (entry 6). Performing the reaction in a closed
vessel for 8 h without taking samples led to a conversion of 100%
(entry 7). These effects are probably due to evaporation of the
rather volatile propanal.
A branched aldehyde was also successfully used as a substrate,
albeit that the yield was lower compared to propanal (Table
1, entry 8), which is in line with the aldehyde specificity
of RAMA.16 The cascade reaction using propanal was also
performed on a 10 mL scale in the presence of 3 U mL-1 of
PhoN-Sf and 50 U mL-1 of GPO resulting in 87% conversion
and an isolated yield of 97 mg (65%).
Scheme 2 Four-step one-pot catalytic cascade.
Fig. 1 shows the time course of the reaction and a comparison
between the currently described cascade and the previous
cascade starting from DHA instead of glycerol.14 It is obvious
that the glycerol cascade is superior as compared to the DHA
one. This may relate to the value of the Km for glycerol which is
0.7 M (not shown), whereas the Km for DHA is 3.6 M.14
This may represent an important factor for the cascade
reaction resulting in higher conversions compared to the DHA
route. Further glycerol stabilizes most enzymes. The glycerol
cascade has the additional advantage that since DHA is prone
to oxidation and side reactions significantly fewer by-products
are formed.
Supposedly, PhoN-Sf-induced phosphorylation of glycerol
gives rise to a racemic mixture of D- and L-glycerol-3-phosphate.
Since GPO would oxidize only the L-enantiomer10 the D-isomer
might accumulate. However, the acid phosphatase possesses
strong transphosphorylation activity14 and most likely PhoN-Sf
will transfer the phosphate group from D-glycerol-3-phosphate
to glycerol resulting in racemic glycerol-3-phosphate so that
accumulation of the D-isomer will not occur.
The first cascade was conducted at 0.5 mL scale (Table 1,
entry 1) with propanal as acceptor aldehyde.14 A solution of
500 mM glycerol, 250 mM PPi and 100 mM propanal at pH 6
was incubated with the four enzymes at 30 ◦C, resulting in 42%
conversion after 24 h. By HPLC it was possible to monitor at
selected time intervals the ratio of PPi to Pi, the formation of
The effect of the glycerol concentration on the rate of
formation was also investigated (Fig. 2). A concentration of
0.5 M glycerol was optimal. At 1 M glycerol both the rate
2896 | Green Chem., 2011, 13, 2895–2900
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Table 2 Four-enzyme cascade with four different aldolases and 3 different aldehydes
Aldehyde
Conv. (%)a
Aldolase17
d.r.
Configuration17
1a propanal
87
72
15
77
61
25
—
RAMA
RhuA
TagA
14 : 1
2 : 1
8 : 1
(3S,4R) : (3S,4S)
(3R,4S) : (3R,4R)
(3S,4S) : (3S,4R)
(3R,4R) : (3R,4S)
(3S,4R) : (3S,4S)
(3R,4S) : (3R,4R)
FucA
2 : 1
1b isobutanal
RAMA
RhuA
TagA
>20 : 1
13 : 1
—
16
FucA
>20 : 1
11 : 1
2 : 1
(3R,4R) : (3R,4S)
(3S,4R) : (3S,4S)
(3R,4S) : (3R,4R)
(3S,4S) : (3S,4R)
1c Alloc-amino-propanal
61/87b
44
RAMA
RhuA
TagA
8
2 : 1
a Conversions were based on the aldehyde and determined with HPLC on aliquots of the reaction mixture. The diastereoisomeric ratio was determined
based upon the peak areas in the HPLC profile and the assignments of the stereochemistry on the reported specificity of these aldolases in literature.17
b It was possible to increase the yield by addition of an extra portion of PPi.
Fig. 2 Effect of glycerol concentration on the yield. 250 mM PPi,
20 mM acetate pH 6, 100 mM propanal, 6 U ml-1 RAMA, 50 U ml-1
GPO, 10 U ml-1 catalase, 1 or 3 U ml-1 of PhoN-Sf.
Fig. 1 Time course of the cascade reaction starting from DHA or
glycerol. The reactions were carried out with 1.5 U mL-1 PhoN-Sf, 6 U
mL-1 RAMA, 250 mM PPi, 100 mM propanal and 0.5 M glycerol or 0.5
M DHA.
iminosugar D-fagomine (8, Scheme 3). D-Fagomine was isolated
from buckwheat seeds (Fagopyrum esculentum Moench), charac-
terized for the first time in 1974, and shown to possess relevant
biological activity.18 It is a potent inhibitor of isomaltase and
specific a- and b-galactosidases. Furthermore, D-fagomine has a
potent antihyperglycemic effect in streptozocin-induced diabetic
mice. Over the years, several pathways have been developed
and the yields were lower. The lower activity at low glycerol
concentrations probably relates to the Km value of PhoN-Sf for
glycerol of about 0.7 M.
To probe the broader applicability of the glycerol cascade,
the reaction was also carried out with three other aldolases in
combination with propanal (Table 2). Rhamnulose-1-phosphate
aldolase (RhuA, entry 2) and fuculose-1-phosphate aldolase
(FucA, entry 4) gave similar conversions as compared to RAMA
(entry 1), although the d.r. in both cases was significantly lower.
In all cases the enantioselectivity with respect to the C3 center
was complete, while C4 gave a 2 : 1 mixture in favor of the
anticipated diastereoisomer.17
Tagatose-1,6-diphosphate aldolase (TagA) showed higher
diastereoselectivity (entry 3), but in this case the conversion
was decreased, probably due to lower activity of the aldolase
(entry 3).
The efficiency of the newly developed four-enzyme cascade
was evaluated in a two-step synthesis of the naturally occurring
Scheme 3 Two-step synthesis of D-fagomine (8).
Green Chem., 2011, 13, 2895–2900 | 2897
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
for the synthesis of D-fagomine,19 including chemoenzymatic
approaches involving aldolases.20
Experimental
Chemicals and enzymes
We envisioned that the cascade with glycerol 2, RAMA and
aldehyde 6 under the optimized reaction conditions would result
in the formation of the intermediate product 7 (Scheme 3),
which may then be converted into D-fagomine 8. The first
attempt using Cbz-protected aldehyde 6 resulted in 21% isolated
yield of aldol product 7a. Due to the low solubility of the
aldehyde, the experiment was repeated in the presence of 10%
DMSO, which resulted in a significant increase in yield to 49%.
Besides, an improvement of the d.r. was observed from 6 : 1
to 10 : 1. This could be due to a solvent effect on the rate
of formation of one of the diastereoisomers caused by active
site perturbation of the aldolase. With the more soluble Alloc-
protected aldehyde 6 under the optimal conditions (without
cosolvent) a conversion of 75% into 7b was observed and a d.r. of
5 : 1. Since it was noted that the temperature of incubation had
an effect on the diastereoselectivity (cf. Table 1, entry 6), we also
investigated the effect of temperature on the conversion and the
diastereoselectivity of the resulting Alloc-modified carbohydrate
7b. There was a marked increase in the isomeric ratio from
5 : 1 to 30 : 1 when the temperature was lowered from 30 to
10 ◦C though the conversion dropped to approximately 50%.
Optimal results in terms of isolated yield were obtained at
20 ◦C giving product 7b in in 69% yield and a d.r. of 11 : 1.
When instead of RAMA the RhuA aldolase derived from the
thermophilic organism Thermatoga maritima was used and the
temperature lowered to 10 ◦C the conversion dropped consider-
ably (not shown), but unfortunately the isomeric ratio remained
2 : 1.
All chemicals were purchased from commercial suppliers and
used without purification. Recombinant Shigella flexneri acid
phosphatase (PhoN-Sf) was expressed in E.coli and purified
as previously described.23 Catalase from bovine liver and
L-glycerol-3-phosphate oxidase were purchased from Sigma-
Aldrich and Toyobo, respectively.
Recombinant rhamnulose-1-phosphate aldolase from the
thermophilic organisms T. maritima (pBAD/gIII plasmid pro-
vided by DSM) was expressed in E. coli strain TOP10 by
inducing with 0.02% L-arabinose overnight at 28 ◦C. The
resulting culture broth was centrifuged at 6000 rpm for 15 min at
4 ◦C and the cells were resuspended in 100 mM Bis-Tris pH 6.5,
0.1 mM ZnCl2. Cell disruption was performed by sonication
followed by centrifugation at 14 000 rpm for 30 min at 4 ◦C.
The thermophilic enzyme RhuA was isolated from the cell free
extract by heat-shock in a 70 ◦C water bath for 20 min, followed
by centrifugation at 14 000 rpm for 30 min.
Tagatose-1,6-diphosphate aldolase from E. coli (pBAD plas-
mid supplied by DSM) was expressed in E. coli TOP10 cells
and induced by 0.02% L-arabinose in the presence of 0.3 mM
ZnCl2. Cells were harvested by centrifugation at 6000 rpm
for 15 min and resuspended in 20 mM Tris/HCl pH 7.5.
Cell disruption was performed by sonication. From the cell-
free extract, TagA was recovered by 40% ammonium-sulfate
precipitation and centrifugation at 13 000 rpm for 30 min. The
supernatant containing TagA was dialyzed overnight against
20 mM Tris/HCl pH 7.5 and loaded onto HiTrap DEAE FF 1
mL FPLC column for further purification. TagA was eluted
at 200 mM NaCl and dialyzed against 20 mM Tris/acetate
pH 7.5, 0.1 mM ZnCl2.
Subsequent Pd-mediated Alloc-deprotection of 7b (Pd on
C, Et3SiH), immediately followed by intramolecular reductive
amination led to the diastereoselective formation of D-fagomine
8 in quantitative yield, thereby realizing a two-step synthesis
of the natural product in 69% overall yield.21 These yields are
in the same range as previously reported by the Clape´s group
for a two-step approach using DHA and D-fructose-6-phosphate
aldolase (FSA),22 but the four-enzyme cascade has the advantage
that it can also be applied to the stereocomplementary DHAP-
dependent aldolases.
Recombinant fuculose-1-phosphate (FucA) aldolase from the
thermophilic bacterium T. ethanolicus (pBAD/gIII plasmid
provided by DSM) was expressed in E. coli strain TOP10
by inducing with 0.02% L-arabinose overnight at 37 ◦C. The
resulting culture was centrifuged at 7000 rpm for 15 min at
◦
4 C and the cells were resuspended in 100 mM Tris-HCl (pH
7.5), 150 mM NaCl and 0.1 mM ZnCl2. Cell disruption was
performed by French Press followed by centrifugation at 17 000
rpm for 30 min at 4 ◦C. The thermophilic enzyme was isolated
Conclusions
◦
from the cell free extract by heat-shock in a 70 C water bath
A convenient procedure was developed for the production
of carbohydrates in a highly enantio- and diastereoselective
manner using an efficient one-pot four-enzyme catalytic cascade.
This cascade starts from inexpensive reagents without need for
protection-deprotection steps. The primary step is chemoenzy-
matic phosphorylation of glycerol and subsequent oxidation to
the energy rich phosphate ester DHAP at the expense of py-
rophosphate. In the same pot, the formed DHAP is coupled to a
variety of aldehydes by DHAP-dependent aldolases allowing the
synthesis of all four possible stereoisomers. Dephosphorylation
to the final aldol product occurs once pyrophosphate is fully
consumed. The value of the catalytic cascade was demonstrated
in an efficient two-step synthesis of the naturally occurring
azasugar D-fagomine. Thus, this approach may lead to the
simple and inexpensive synthesis of a variety of non-natural
heterocycles that presently can only be produced at high costs.
for 20 min, followed by centrifugation at 17000 rpm for 30 min
at 4 ◦C. The supernatant containing FucA was loaded onto
a Q-Sepharose FF FPLC column (40 mL) and eluted with a
linear gradient of NaCl (0–1 M NaCl in 20 column volumes).
The purified fractions were collected and desalted using 50 mM
Tris-HCl pH 7.5 and 150 mM NaCl buffer.
Enzyme activity assays
Aldolase activity was assayed by measuring the decrease of
DHAP concentration in a reaction in which the aldolase
performs the aldol coupling in the presence of DHAP
(50 mM), rac-glyceraldehyde (100 mM), 0.1 mM ZnCl2, 50 mM
acetate pH 6, 30 ◦C. The decrease in DHAP concentration was
assayed every 5 min with a coupled-enzyme system in presence
of NADH-consuming glycerol-3-phosphate dehydrogenase by
2898 | Green Chem., 2011, 13, 2895–2900
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
measuring NADH concentration by UV spectroscopy. Samples
of the aldolase reaction mixture were diluted and incubated
in 100 mM Tris/acetate pH 7.5, 0.16 mM NADH, and
1 U mL-1 glycerophosphate dehydrogenase. The absorption was
monitored at 340 nm at 20 ◦C (molar absorption coefficient
6.22 mM -1 cm -1).
The Km value of PhoN-Sf for glycerol was determined with a
coupled enzyme system as described by Bergmeyer.24 L-Glycerol-
3-phosphate resulting from the phosphatase reaction was as-
sayed by coupling with glycerol-3-phosphate dehydrogenase in
the presence of hydrazine and NAD+ at pH 9.5 and 25 ◦C. The
reaction was spectrophotometrically monitored at 340 nm.
(3S,4R) - 6 - [(Allyloxycarbonyl)amino] - 5,6 - dideoxy -2-
hexulose (7b, 20 mg, 0.081 mmol) was dissolved in EtOH (10
mL) and trietylsilane (0.16 mL, 0.97 mmol) was added, followed
by the addition of Pd on C (10 mg, 0.094 mmol). The reaction
mixture was stirred overnight at room temperature and filtered
over Celite. The filtrate was concentrated in vacuo and purified
with solid phase extraction, giving rise to D-fagomine (8, 12 mg,
1
0.082 mmol, 100%. H NMR (CD3OD, 400 MHz) d 3.87 (dd,
J = 10.9, 3.1 Hz, 1H), 3.63–3.55 (m, 1H), 3.42–3.38 (m, 1H),
3.11 (t, J = 9.0 Hz, 1H), 3.02 (ddd, J = 12.7, 4.6, 2.4 Hz, 1H),
2.66–2.58 (m, 2H), 2.48–2.41 (m, 1H), 1.97–1.90 (m, 1H), 1.54–
1.43 (m, 1H); 13C NMR (CD3OD, 75 MHz) d 73.3, 73.1, 61.4,
61.2, 42.6, 32.8; [a]2D0 = +6.6 (c = 0.60, MeOH); LRMS 148.10
[M+H]+. These data are in agreement with literature.19
Analytical methods and product characterization
The time-course of the cascade reaction and the conversions
were determined by HPLC using an Alltech OA 1000 organic
acid column (0.65 ¥ 30 cm). 20 ml of the reaction mixture was
diluted 10-fold before injection on the HPLC. Isocratic runs
were performed using 4.5 mM H2SO4, with a flow rate of 0.4 ml
min-1. The effluent was monitored at 210, 215, 254, and 275 nm
and by refractive index detector.
1H NMR and 13C NMR spectra were recorded at 300 (75)
or 400 (100) MHz. Optical rotations were determined with a
Perkin Elmer 241 polarimeter. IR spectra were recorded on a
Thermo Mattson IR300 equipped with a Harrick split pea ATR
unit. Fast Atom Bombardment (FAB) mass spectrometry was
carried out using a JEOL JMS SX/SX 102A four-sector mass
spectrometer coupled to a JEOL MS-MP9021D/UPD system
program.
Spectroscopic data
5,6-Dideoxy-D-threo-2-hexulose (1a). 1H NMR (CD3OD,
300 MHz) d 4.53 (d, 1H, J = 19.2 Hz), 4.42 (d, 1H, J = 19.2
Hz), 4.14 (d, 1H, J = 2.4 Hz), 3.78 (dt, 1H, J = 2.4, 6.0 Hz),
1.65–1.50 (m, 2H), 0.96 (t, 3H, J = 7.4 Hz); 13C NMR (CD3OD,
75 MHz) d 212.0, 77.1, 73.2, 66.0, 25.3, 8.7; [a]2D0 -17.7 (c 0.22,
MeOH); IR (neat) 1723, 3397 cm-1; LRMS 149.08 [M+H]+,
171.09 [M+Na]+. These data are in agreement with literature.10
5,6,7-Trideoxy-D-threo-2-heptulose
(1b). 1H-NMR
(CD3OD, 300 MHz) d 4.47 (q, 2H, J = 19.2 Hz), 4.11 (d,
1H, J = 2.3 Hz), 3.89 (dt, 1H, J = 2.4, 6.7 Hz), 3.30 (td, 1H,
J = 1.6, 3.1 Hz), 1.56–1.40 (m, 4H), 0.95 (t, 3H, J = 7.2 Hz);
13C-NMR (CD3OD, 75 MHz) d 211.9, 77.5, 71.4, 66.0, 34.5,
18.2, 12.4; [a]2D0 -14.1 (c 0.29, MeOH)); IR (neat) 1722, 3367
cm-1; LRMS 163.10 [M+H]+, 185.12 [M+Na]+.
General procedure for the preparative scale one-pot four-enzyme
cascade
To a solution of 500 mM glycerol, 250 mM PPi (3 : 2
Na2PPi/Na4PPi),25 20 mM sodium acetate (pH 6), and 100 mM
aldehyde in water 10 U mL-1 catalase from bovine liver, 6 U
mL-1 RAMA, and 50 U mL-1 GPO were added. The reactions
were started by adding 3 U mL-1 PhoN-Sf and incubated at
20 ◦C under mild shaking until completion. Dephosphorylation
to the end product was completed after 24 h and 4 g of silica
gel was added. The reaction slurry was concentrated under
reduced pressure, and the silica gel was poured on top of a
silica gel column and eluted with EtOAc/MeOH (19 : 1). The
pure fractions were collected and concentrated under reduced
pressure to give the product as a light-yellow oil.
5,6-Dideoxy-5-methyl-D-threo-2-hexulose (1c). 1H NMR
(CD3OD, 300 MHz) d 4.53 (d, 1H, J = 19.2 Hz), 4.44 (d, 1H, J =
19.2 Hz), 4.31 (d, 1H, J = 1.8 Hz), 3.45 (dd, 1H, J = 2.0, 9.1 Hz),
1.90 (ddd, 1H, J = 2.3, 6.7, 13.4 Hz), 1.01 (d, 3H, J = 6.7 Hz), 0.93
(d, 3H, J = 6.7 Hz); 13C NMR (CD3OD, 75 MHz) d 212.6, 77.4,
75.8, 65.9, 29.8, 17.8, 17.6; [a]2D0 -12.2 (c 0.16, MeOH); IR (neat)
1723, 3396 cm-1; LRMS 163.10 [M+H]+, 185.12 [M+Na]+.
(3S,4R)-6-[(Benzyloxycarbonyl)amino]-5,6-dideoxy-2-hexu-
lose (7a). 1H-NMR (CD3OD, 400 MHz) d 7.48–7.21 (m, 5H),
5.19–5.02 (2H, m), 4.48 (q, 2H, J = 16.4 Hz), 4.19–4.09 (m, 1H),
3.97 (t, 1H, J = 7.2 Hz), 3.29–3.19 (m, 2H), 1.76 (q, 2H, J =
8.8 Hz); 13C (CD3OD, 75 MHz) d 213.5, 159.2, 138.6, 129.6,
129.1, 129.0, 79.6, 71.3, 68.0, 67.6, 38.9, 34.6. These data are in
agreement with literature.19
Two step synthesis of D-fagomine (8)
3-Alloc-aminopropanal (157 mg, 1.0 mmol) was dissolved in
H2O (5.47 mL). Subsequently, glycerol (2.00 mL, 2.5 M 5.0
mmol) and PPi (591 mg, 2.5 mmol) were added, followed by
the addition of 1.00 mL GPO (50 U mL-1), 20 mL (10 U mL-1)
catalase from bovine liver and 161 mL (6 U mL-1) RAMA. The
reactions were initiated by adding 600 mL (3U mL-1) PhoN-Sf
and incubated until completion at 20 ◦C under mild shaking.
Silica gel (2 g) was added to the reaction and the resulting
mixture was concentrated under reduced pressure. The product
was purified by flash chromatography (EtOAc/MeOH/19 : 1),
giving product 7b in 69% yield.
(3S,4R)-6-[(Allyloxycarbonyl)amino]-5,6-dideoxy-2-hexulose
(7b). 1H-NMR (CD3OD, 400 MHz) d 7.13–6.60 (1H, m),
6.12–5.67 (1H, m), 5.25 (d, 1H, J = 18 Hz), 5.17 (d, 1H, J = 9.0
Hz), 4.65–4.30 (m, 4H), 4.10 (d, 1H, J = 3.0 Hz), 3.96 (dt, 1H,
J = 7.5, 3.0 Hz), 3.55 (s, 2H), 3.20 (s, 1H), 1.74 (q, 2H, J = 6
Hz); 13C (CD3OD, 75 MHz) d 213.7, 134.8, 117.7, 79.7, 71.4,
68., 66.6, 38.9, 34.7; IR (neat) 1649, 1689, 3338 cm-1; LRMS
248.10 [M+H]+, 270.3 [M+Na]+.
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 2895–2900 | 2899
©

View Article Online
9 D. C. Crans and G. M. Whitesides, J. Am. Chem. Soc., 1985, 107,
7019–7026.
Acknowledgements
10 W.-D. Fessner and G. Sinerius, Angew. Chem., Int. Ed. Engl., 1994,
This research was performed under the ACTS-NWO Integration
of Biosynthesis & Organic Synthesis (IBOS) program. The
companies MSD (Oss, NL), DSM (Geleen, NL) and Syncom
(Groningen, NL) are kindly acknowledged for financial support.
We also wish to thank Dr M. Schu¨rmann (DSM, Geleen) for
providing the plasmids harboring the aldolase genes.
33, 209–212.
11 O. Eyrisch, G. Sinerius and W.-D Fessner, Carbohydr. Res., 1993,
238, 287–306.
12 C. H. Wong and G. M Whitesides, J. Org. Chem., 1983, 48, 3199–
3205.
13 R. Schoevaart, F. van Rantwijk and R. A. Sheldon, J. Org. Chem.,
2000, 65, 6940–6943.
14 T. van Herk, A. F. Hartog, H. E. Schoemaker and R. Wever, J. Org.
Chem., 2006, 71, 6244–6247.
15 T. van Herk, A. F. Hartog, A. M. van der Burg and R. Wever, Adv.
Synth. Catal., 2005, 347, 1155–1162.
16 M. D. Bednarski, E. S. Simon, N. Bischofberger, W.-D. Fessner, M. J.
Kim, W. Lees, T. Saito, H. Waldmann and G. M. Whitesides, J. Am.
Chem. Soc., 1989, 111, 627–635.
17 K. Faber, Biotransformations in Organic Chemistry, 5th ed., Springer-
Verlag, Berlin, 2004, pp. 277–284.
Notes and references
1 (a) A. Bruggink, R. Schoevaart and T. Kieboom, Org. Process Res.
Dev., 2003, 7, 622–640; (b) R. A Sheldon, in Multi-step enzyme
catalysis: Biotransformations and chemoenzymatic synthesis, ed. E.
Garc´ıa-Junceda, Wiley-VCH, Weinheim, 2008, Ch. 6, pp. 109–136.
2 R. A. Sheldon, Chem. Commun., 2008, 3352–3365.
3 S. M. Dean, W. A. Greenberg and C.-H. Wong, Adv. Synth. Catal.,
2007, 349, 1308–1320.
4 (a) Y. Ichikawa, Y. Igarashi, M. Ichikawa and Y. Suhara, J. Am.
Chem. Soc., 1998, 120, 3007–3018; (b) Iminosugars: from synthesis to
therapeutic applications, ed. P. Compain and O. R. Martin, Wiley-
VCH, Weinheim, 2007; (c) N. Asano, Cell. Mol. Life Sci., 2009, 66,
1479–1492.
5 (a) M. Movassaghi and E. N. Jacobsen, Science, 2002, 298, 1904–
1905; (b) S. Y. Ko, A. W. M. Lee, S. Masamune, L. A. Reed, K. B.
Sharpless and F. J. Walker, Science, 1983, 220, 949–951.
6 A. B. Northrup and D. W. C. MacMillan, Science, 2004, 305, 1752–
1755.
7 (a) H. J. M. Gijsen and C.-H. Wong, J. Am. Chem. Soc., 1995, 117,
7585–7591; (b) K. M. Koeller and C.-H. Wong, Chem. Rev., 2000,
100, 4465–4494; (c) P. Clape´s, W.-D. Fessner, G. A. Sprenger and A.
K. Samland, Curr. Opin. Chem. Biol., 2010, 14, 154–167.
8 (a) A. Bolt, A. Berry and A. Nelson, Arch. Biochem. Biophys., 2008,
474, 318–330; (b) M. G. Silvestri, G. Desantis, M. Mitchell and C.-H.
Wong, Top. Stereochem., 2003, 23, 267–342; (c) M. Schumperli, R.
Pellaux and S. Panke, Appl. Microbiol. Biotechnol., 2007, 75, 33–45.
18 A. Kato, N. Asano, H. Kizu and K. Matsui, J. Nat. Prod., 1997, 60,
312–314.
19 (a) N. Kumari, B. G. Reddy and Y. D. Vankar, Eur. J. Org. Chem.,
2009, 160–169; (b) L. Bartali, D. Scarpi, A. Guarna, C. Prandi and
E. G. Occhiato, Synlett, 2009, 913–916.
20 M. Sugiyama, Z. Hong, P.-H. Liang, S. M. Dean, L. J. Whalen,
W. A. Greenberg and C.-H. Wong, J. Am. Chem. Soc., 2007, 129,
14811–14817, and references cited therein.
21 P. K. Mandl and J. S. McMurray, J. Org. Chem., 2007, 72, 6599–6601.
22 J. A. Castillo, J. Calveras, J. Casas, M. Mitjans, M. Pilar Vinardell, T.
Parella, T. Inoue, G. A. Sprenger, J. Joglar and P. Clape´s, Org. Lett.,
2006, 8, 6067–6070.
23 N. Tanaka, Z. Hasan, A. F. Hartog, T. van Herk and R. Wever, Org.
Biomol. Chem., 2003, 1, 2833–2839.
24 H. U. Bergmeyer, K. Gawehn, M. Grassl, Glucose-6-phosphate
dehydrogenase from yeast, in Methods of Enzymatic Analysis, ed. H.
U. Bergmeijer, Verlag Chemie, vol. 1, 1986, pp. 458–459.
25 PPi was added as a mixture of 3.5 equiv. (350 mg) Na2PPi and 2 equiv.
(241 mg) Na4PPi resulting in a buffered solution of pH 6.
2900 | Green Chem., 2011, 13, 2895–2900
This journal is
The Royal Society of Chemistry 2011
©