Novel and efficient chemoenzymatic synthesis of D-glucose 6-phosphate and molecular modeling studies on the selective biocatalysis

Article abstract of DOI:10.1002/ejoc.200700017

A concise chemoenzymatic synthesis of glucose 6-phosphate is described. Candida rugosa lipase was found to be an efficient catalyst for both regio- and stereoselective deacetylation of the primary hydroxy group in the peracetylated D-glucose. In addition, we report an improved synthesis of 1,2,3,4,6-penta-O- acetyl-α-D-glucopyranose providing a large-scale procedure for the acetylation of α-D-glucose without isomerization at the anomeric center. The high overall yield and the easy scalability makes this chemoenzymatic strategy attractive for industrial application. Furthermore, molecular modeling of phosphonate transition-state analog for the enzymatic hydrolysis step supports the substrate selectivity observed with Candida rugosa lipase. Wiley-VCH Verlag GmbH & Co. KGaA, 2007.

Full text of DOI:10.1002/ejoc.200700017

FULL PAPER
DOI: 10.1002/ejoc.200700017
Novel and Efficient Chemoenzymatic Synthesis of
D
-Glucose 6-Phosphate and
Molecular Modeling Studies on the Selective Biocatalysis
Tatiana Rodríguez-Pérez,[a] Iván Lavandera,[a] Susana Fernández,[a] Yogesh S. Sanghvi,[b]
Miguel Ferrero,*[a] and Vicente Gotor*[a]
Keywords: Carbohydrates / Enzyme catalysis / Molecular modeling
A concise chemoenzymatic synthesis of glucose 6-phosphate
is described. Candida rugosa lipase was found to be an ef-
yield and the easy scalability makes this chemoenzymatic
strategy attractive for industrial application. Furthermore,
ficient catalyst for both regio- and stereoselective deacety- molecular modeling of phosphonate transition-state analog
lation of the primary hydroxy group in the peracetylated
D-
for the enzymatic hydrolysis step supports the substrate
glucose. In addition, we report an improved synthesis of
selectivity observed with Candida rugosa lipase.
1,2,3,4,6-penta-O-acetyl-α-
D-glucopyranose providing
a
large-scale procedure for the acetylation of α-
out isomerization at the anomeric center. The high overall
D
-glucose with-
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
Robison and King[6] were first to isolate G-6-P from a
crude mixture of hexose monophosphates obtained by yeast
fermentation. Phosphorylation of glucose in the presence of
hexokinase and acetate kinase with ATP also furnished a
route to G-6-P.[7] The phosphorolysis of cellobiose to G-1-
P which converts to G-6-P by phosphoglucomutase pro-
vides yet another procedure.[8] Treatment of fructose 1,6-
diphosphate with mild acid furnished fructose 6-phosphate
which is transformed into G-6-P in the presence of glucose
phosphate isomerase.[9] Recently, bacterial acid phospha-
tases have been described as an alternative for production
of G-6-P.[10] The foregoing protocols are of limited value
for scale-up because of the expensive enzymes used and the
degradation of enzymes during the reaction and tedious iso-
lation of the G-6-P from the reaction mixture.
Lardy and Fischer[11] reported a 3-step protocol for the
synthesis of G-6-P starting with 1,2,3,4-tetra-O-acetyl-β--
glucopyranose (TAG). However, synthesis of the key start-
ing material TAG is not straightforward and requires sev-
eral protection-deprotection steps.[12] Other chemical meth-
ods for the preparation of G-6-P describes the phosphoryla-
tion of glucose by phosphorus oxychloride,[13] phosphorus
pentoxide, or polyphosphoric acid.[14] Unfortunately, direct
phosphorylation of a polyhydroxy-sugar leads to unsatis-
factory low yields and requires extensive purification.[15]
The central role of G-6-P in biology and the increase in
demand for industrial applications makes it a desirable tar-
get for synthesis. Most of the enzymatic approaches need
very expensive and highly specific enzymes or cofactors, or
proceed with low yields. Chemical routes involve extra pro-
tecting and deprotection steps to phosphitylate selectively
the C-6-position. These limitations in the access to G-6-P
Introduction
-Glucose 6-phosphate (G-6-P) is vital to the glucide me-
tabolism for higher organisms. It appears as an intermediate
in several relevant routes such as glycolysis, gluconeogene-
sis, and in the synthesis and degradation of glycogen. More-
over, it has been widely used in pharmaceutical and medici-
nal chemistry applications. G-6-P is a normal constituent of
the resting muscle and used in cases of low blood pressure,
anesthesiology, asthenia, cardiology, and alcoholism.[1] In-
teraction of aluminum(III), which accumulates in the or-
ganism is associated with some human pathologies, where
G-6-P plays an important role in the intake and circulation
of this ion in biological media.[2] Conjugation of food pro-
teins with G-6-P was useful in improving its functional
properties such as water solubility, emulsifying activity,
foaming and gel-forming properties.[3] On the other hand,
this metabolite is a key intermediate in the biosynthesis of
clinically important 2-deoxy-streptamine-containing ami-
nocyclitol antibiotics[4] and has proven to also be the me-
tabolite involved in the stimulation of lipogenic enzyme
gene expression.[5]
[a] Departamento de Química Orgánica e Inorgánica and Instituto
Universitario de Biotecnología de Asturias, Universidad de
Oviedo,
33006 Oviedo (Asturias), Spain
Fax: +34-98-510-3448
E-mail: MFerrero@fq.uniovi.es
VGS@fq.uniovi.es
[b] Rasayan Inc.,
2802 Crystal Ridge Road, Encinitas, CA 92024-6615, USA
Fax: +1-760-944-1543
E-mail: Rasayan@sbcglobal.net
Supporting information for this article is available on the
WWW under http://www.eurjoc.org or from the author.
Eur. J. Org. Chem. 2007, 2769–2778
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2769

M. Ferrero, V. Gotor
FULL PAPER
motivated us to develop an efficient route for its synthesis material. Following a literature protocol, acylation of α--
on large-scale starting from TAG. glucose (1) with acetic anhydride in the presence of pyridine
A major challenge faced by carbohydrate chemists is to furnished 2 in 95% isolated yield (Scheme 1). The structure
1
orchestrate selective transformation of hydroxy groups with of 2 was confirmed by H NMR, which shows a doublet
similar reactivity in a single molecule. Enzyme-catalyzed re- for the α-anomer at δ = 6.39 ppm. However, the spectrum
actions have been successfully utilized in selective deacyla- also showed a small doublet at δ = 5.69 ppm corresponding
tion or acylation of sugars and hence have provided an ef- to the β-hydrogen on the anomeric carbon of 2.
fective method of manipulating protecting-group strategies
Integration of the 1H NMR signals of compound 2 indi-
in carbohydrate synthesis.[16] For example, Candida rugosa cated that it is a mixture of α/β-anomers in a ratio of 92:8.
1
(CRL) and Pseudomonas fluorescens (PFL) lipases afford The observed H NMR values are in accordance with the
regioselective hydrolysis of acetylated glycopyranoses and literature report.[19] A commercial sample of 2 also exhib-
glycopyranosides only in the C-6- and C-1-positions, ited a similar H NMR pattern for the α/β-anomers.[20]
1
respectively.[17] These results obtained by Guisán and co-
workers depend strictly on the use of enzyme preparations
of PFL and CRL obtained by adsorption on an octyl aga-
rose support. However, this reaction is limited to analytical
scale because enzymes require purification and immobiliza-
Enzymatic Hydrolysis Studies of 2
tion and these lipase forms are not commercially available.
The outcome of the synthesis led us to study the enzy-
In another report, the esterase from Rhodosporidium toru-
matic hydrolysis of 2 as an anomeric mixture. Regioselective
loides has been used to catalyze the hydrolysis of a series of
hydrolysis of 2 to 3 was carried out with commercial lipases
peracetylated α--hexopyranoses and α--hexopyranosides.
only because of their easy access in bulk quantity and reli-
For example, per-acetylated glucose has been selectively
able activity. Six enzymes [Pseudomonas cepacia lipase
cleaved at the C-6-position, but the yield of the isolated 6-
(PSL-C), Candida rugosa lipase (CRL), Candida antarctica
OH derivative was a modest 54%.[18]
lipase A (CAL-A), Candida antarctica lipase B (Novozym
Herein, we present a novel and efficient chemoenzymatic
435, CAL-B), Chirazyme L-2, and porcine pancreatic lipase
route to obtain G-6-P on a preparative scale, where the key
(PPL)] were evaluated to determine which enzyme gives the
step is the lipase-catalyzed regioselective hydrolysis of
best regioselectivity for transformation of 2 to 3. Interest-
1,2,3,4,6-penta-O-acetyl-α--glucopyranose (2) to furnish
ingly, hydrolysis of the α/β-mixture of 2 resulted in the non-
TAG. Different commercial lipases have been screened and
reactive nature of the β-anomer and selective hydrolysis of
various reaction parameters, such as temperature, co-sol-
the α-anomer. Among the six enzymes tested, CRL afforded
vent, and buffer pH are studied for yield optimization of
the best results where hydrolysis of the C-6 or C-4 acetyl
TAG. We also propose the rationale for our finding of the
group (≈1:1) of α-2 occurred at 30 °C in 25 m phosphate
enzyme selectivity during hydrolysis of 2 using computer-
buffer pH 4 with acetonitrile as co-solvent (20%).[17] In our
aided molecular modeling.
studies, the use of PSL-C triggered the hydrolysis at the C-
1 and C-2 positions of 2 furnishing 3,4,6-tri-O-acetyl-α--
glucopyranose. Various attempts to hydrolyze 2 with CAL-
A, CAL-B, PPL, and Chirazyme L-2 indicated that these
Results and Discussion
enzymes were completely inert for an extended period of
time (5–6 d) with full recovery of the unchanged starting
materials α/β-anomers (92:8).
Synthesis of 1,2,3,4,6-Penta-O-acetyl-α-D-glucopyranose (2)
In order to develop a direct and concise synthesis of
TAG, we needed good quality pentaacetate 2 as the starting
Process Optimization of CRL-Mediated Hydrolysis of 2
The success with CRL prompted us to further investigate
the influence of temperature and buffer pH on the rate and
regioselectivity of enzymatic hydrolysis of 2. The increase
in temperature from 30 °C to 60 °C has a dramatic effect
on the conversion rate where enzyme activity was totally
abolished at higher temperatures (Entries 1–4, Table 1). We
found that at 40 °C the CRL is able to maintain its activity
and offer best selectivity toward the hydrolysis of the C-6
acetyl group in 2. Modulation of the buffer pH clearly indi-
cates that maximum activity is exhibited around pH 4–5.
The activity of CRL is compromised at a higher pH re-
sulting in lower conversion rates (Entries 5–7, Table 1).
Scheme 1.
2770
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2769–2778

Chemoenzymatic Synthesis of -Glucose 6-Phosphate
FULL PAPER
Table 1. Hydrolysis catalyzed by CRL of 1,2,3,4,6-penta-O-acetyl- n-hexane (4:1) was possible leaving behind unreacted β-an-
α--glucopyranose (2).[a,b]
omer of 2 and isomeric 4 in the mother liquor.
Entry pH 20% co-solvent T/°C
t/h Conv. (%)[c] Ratio 3/4[c]
An Improved Process for the Synthesis of α-Anomeric 2 and
Its Hydrolysis with CRL
1
2
3
4
5
6
7
8
4
4
4
4
5
6
7
4
4
4
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
CH3CN
THF
30
40
50
60
40
40
40
40
40
40
2
43
48:52
66:34
n.d.
s.m.
64:36
n.d.
s.m.
s.m.
88:12
91:9
48 55
48 Ͻ10
48
48 58
Our results indicate that CRL is highly regioselective in
cleaving the 6-O-acetyl group only from the α-anomer in a
mixture of α/β-2. We were intrigued by the selective nature
of CRL and questioned the possible use of pure α-anomeric
2 to further improve the yield and isolation of the hy-
drolyzed 3. Our current route for the synthesis of 2 resulted
in a mixture of two anomeric products (92:8). The use of
commercial sample exhibited a similar pattern. This fact
encouraged us to explore an improved synthesis of 2 where
α-anomer is the only product. A literature search for per-
O-acetylation of carbohydrates[21] revealed use of both ex-
otic as well as simple reagents such as Er(OTf)3,[22] Ce-
(OTf)3,[23] HClO4/SiO2,[24] or iodine.[19,25] However, cost,
availability, toxicity, and difficulty in handling can limit the
use of these catalysts on large-scale. In addition, epimeri-
zation at the anomeric carbon has been of serious concern.
Because of these limitations, we decided to carefully investi-
gate the current protocol with acetic anhydride in pyridine.
We observed that lowering the reaction temperature im-
proved the α-anomeric ratio. Thus, using 8 equiv. of freshly
distilled acetic anhydride and lowering the reaction tem-
perature to –10 °C, 1,2,3,4,6-penta-O-acetyl-α--glucopyr-
anose (α-2) was obtained as sole product in 96% yield
(Scheme 2). Work-up and isolation of the product was eas-
ily done by simple precipitation from the reaction mixture
without a need for further purification. With this simple
improvement in place, we had pure α-anomer of 2 in hand
for further studies.
0
48 Ͻ10
48
95
0
0
9
10
1,4-dioxane
acetone
27 Ͼ97[d]
48 63
[a] Starting material contains 8% of 1,2,3,4,6-penta-O-acetyl-β--
glucopyranose which was recovered unchanged. [b] Enzymatic hy-
drolysis was carried out in 25 mM phosphate buffer using 20% of
1
co-solvent. [c] Calculated by H NMR integration. [d] Conversion
indicated as Ͼ97% (Ϯ3% by NMR analysis) means that no starting
material was detected even at high spectrum amplitude. n.d.: not
determined, s.m.: starting material recovered.
It is well established that the presence of a co-solvent
during lipase-mediated hydrolysis could influence the out-
come of selectivity in a favorable manner. To further in-
crease the selectivity of the transformation of 2 to 3, we
examined the use of THF, 1,4-dioxane and acetone as co-
solvents. Improved results were obtained with 1,4-dioxane
or acetone, whereas THF showed no hydrolysis (Entries 8–
10, Table 1). Use of 1,4-dioxane as a co-solvent offered ex-
cellent conversion (Ͼ97%) of the two regioisomers 3 and 4
in a 88:12 ratio, respectively, in 27 h. Although the use of
acetone gave slightly better selectivity, the conversion was
far too low to be in par with 1,4-dioxane.
In anticipation of improving the product ratio, additional
experiments were carried out in 1,4-dioxane as a co-solvent
at variable temperature and pH. The results are summa-
rized in Table 2. Again, lower reaction temperature (30 °C)
and lower buffer pH 4 drove the hydrolysis catalyzed by
CRL to completion furnishing isomers 3 and 4 in 97:3 ratio,
respectively, in 48 h (Entry 3, Table 2). Compounds 3 and
4 have identical Rf values and their separation by
chromatography was not practical. Gratifyingly, isolation of
pure 3 via crystallization from a mixture of diethyl ether/
Scheme 2.
Next, the enzymatic hydrolysis with CRL was set-up with
5 g of pure α--peracetylated glucose (α-2) under fully opti-
mized reaction conditions (Scheme 2). After 48 h, we ob-
tained a mixture of 3 and 4 (95:5 ratio, respectively) in 95%
yield. Clearly, this is an improvement over previous condi-
tions where we had a mixture of three products (Scheme 1).
However, we were puzzled with the formation of 4 in minor
amounts despite the selectivity demonstrated by CRL. We
argued that formation of 4 is due to the acyl group mi-
gration and not the CRL promiscuity.
Table 2. Hydrolysis catalyzed by CRL of 1,2,3,4,6-penta-O-acetyl-
α--glucopyranose (2)[a] in 1,4-dioxane.[b]
Entry pH
T/°C
t/h
Conv. (%)[c] Ratio 3/4[c]
1
2
3
4
5
6
7
4
4
4
4
5
6
3
40
50
30
25
30
30
30
27
88
48
72
48
48
48
Ͼ97[d]
47
88:12
61:39
97:3
Ͼ97[d]
Ͼ97[d]
Ͼ97[d]
Ͼ97[d]
62
92:8
93:7
90:10
87:13
[a] Starting material contains 8% of 1,2,3,4,6-penta-O-acetyl-β--
glucopyranose which was recovered unchanged. [b] Enzymatic hy-
drolysis was carried out in 25 m phosphate buffer using 20% of
co-solvent. [c] Calculated by H NMR integration. [d] Conversion
indicated as Ͼ97% (Ϯ3% by NMR analysis) means that no starting
Studies Related to the Confirmation of 4-O to 6-O Acyl
Migration in 3
1
It is well documented that acyl groups have the propen-
sity to migrate under both acidic or basic conditions.[17b,18]
material was detected even at high spectrum amplitude.
Eur. J. Org. Chem. 2007, 2769–2778
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2771

M. Ferrero, V. Gotor
FULL PAPER
Based on this fact, we decided to generate proof for the derivatives 3 and 4 in a 97:3 ratio, respectively. This experi-
formation of 4 presumably through 4-O to 6-O acyl mi- ment further confirmed that CRL is able to recognize the
gration in the compound 3. First, we carried out experi- 6-position in 5 to afford 3, despite of a bulky acyl group.
ments in the absence of CRL to assure that the enzyme was Should CRL be promiscuous in its activity we would have
not responsible for the acyl migration.
seen the formation of 1,2,3-tri-O-acetyl-6-O-benzoyl-α--
As expected, the results in Table 3 indicated that the acyl glucopyranose (6). The later compound was not detected in
group undergoes migration from 4-O to 6-O in all co-sol- the reaction mixture.
vents tested. The rearrangement was slower in 1,4-dioxane
and THF compared to the acetonitrile and acetone (Entries
1–4, Table 3). These experiments were carried out with
25 m phosphate buffer in the presence of the co-solvent.
Next, we wanted to better understand the role of phosphate
buffer in the CRL mediated hydrolysis process. Therefore,
we designed another set of experiments only with organic
solvents.
Table 3. Studies on transformation of 3 to 4 through acyl migration
in the absence of enzyme.[a]
Entry
Co-solvent
3/4 final ratio[b]
1
2
3
4
1,4-dioxane
THF
CH3CN
acetone
86:14
86:14
82:18
80:20
[a] Reaction conditions: initial ratio 3:4 is 91:9, 25 m phosphate
buffer using 20% of co-solvent and at 30 °C during 48 h. [b] Calcu-
lated by H NMR spectroscopy.
Scheme 3.
1
Further evidence for the selectivity of CRL was demon-
strated by enzyme-catalyzed hydrolysis of the 13C-labeled
(C-6 position in α-2*, Scheme 4) peracetylated α--glucose.
The α-2* was conveniently synthesized by acylation of 3
with 13C-labeled acetic anhydride in pyridine at room tem-
perature. The CRL-mediated hydrolysis of α-2* furnished a
mixture of unlabeled compounds 3 and 4 in a 96:4 ratio,
respectively. Assuming that the CRL was responsible for the
hydrolysis at the C-4 position, the product in the foregoing
reaction should have been the 13C labeled derivative 4*
(Scheme 4). The 13C NMR experiment of the reaction mix-
ture confirmed that the label was totally absent in the prod-
ucts. Therefore, we propose that the CRL selectively trans-
forms 2 or 5 into 3, which undergoes 4-O to 6-O acetyl
migration furnishing the regioisomer 4 as a minor by-prod-
uct.
The evaluation of acyl group migration was performed
in four organic solvents that were used for the prior studies.
No evidence of the 4-O to 6-O acyl migration was found in
1,4-dioxane or THF (Entries 1 and 2, Table 4) determined
1
by H NMR spectroscopic data. Interestingly, the acyl mi-
gration was observed in acetonitrile or acetone as a solvent
(Entries 3 and 4, Table 4). These experiments suggest that
acyl migration could take place either in organic solvents
alone or mixed with the phosphate buffer.
Table 4. Studies on transformation of 3 to 4 via acyl migration in
the organic solvents.[a]
Entry
Co-solvent
3/4 final ratio[b]
1
2
3
4
1,4-dioxane
THF
CH3CN
acetone
91:9
91:9
71:29
87:13
In order to decrease the possibility of acetyl migration,
we increased the amount of CRL during hydrolysis reac-
tion. Indeed, the overall reaction time was reduced from 48
to 24 h, but the ratio of hydrolyzed products 3 and 4 remain
unchanged. This experiment provided additional support to
our hypothesis that CRL did not trigger the acyl migration
during hydrolysis step.
[a] Reaction conditions: initial ratio 3/4 is 91:9, 30 °C during 48 h.
1
[b] Calculated by H NMR spectroscopy.
Studies Related to the Confirmation of the Selectivity
Demonstrated by CRL
Recently, Otera and co-workers[26] described a chemical
Next, we wished to generate clear chemical evidence that deprotection strategy using [tBu2SnOH(Cl)]2 for the synthe-
the CRL is not responsible for the formation of 4. For this, sis of 3 from 2. Despite of good yield (88%), this protocol
we synthesized 1,2,3,4-tetra-O-acetyl-6-O-benzoyl-α--glu- has severe limitations due to the toxicity of tin and the lack
copyranose (5) from 1,2,3,4-tetra-O-acetyl-α--glucopyr- of a commercial source for the organotin catalyst. In com-
anose (3) by benzoylation with benzoyl chloride in pyridine parison, our enzymatic protocol allows the synthesis of 3
(Scheme 3). Installation of a 6-O-benzoyl group in 5 al- in high yield through a much simpler procedure using a
lowed us to test the selectivity of CRL under a modified commercial catalyst that is biodegradable and has many en-
chemical environment. Treatment of 5 with CRL under hy- vironmental advantages over other traditional chemical
drolytic conditions provided a mixture of 6-OH and 4-OH processes.
2772
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2769–2778

Chemoenzymatic Synthesis of -Glucose 6-Phosphate
FULL PAPER
The acetyl groups were removed by saponification with
K2CO3 in MeOH, and subsequent treatment with Dowex
50W afforded the desired G-6-P 10 in 99% yield.
We repeated the synthesis of G-6-P on 10-gram scale fol-
lowing our five-step protocol starting from α--glucose in
77% overall yield. We find this protocol to be very efficient
1
because it utilized only one chromatographic step. The H
NMR spectrum of the G-6-P synthesized herein showed the
distinctive anomeric signal of the α- and β-anomers at δ =
5.10 and 4.53 ppm, respectively, in a 60:40 ratio. This data
is in accordance with the literature report and correlated
with the reference sample purchased from a commercial
supplier.
Although the chemical purity of the G-6-P was con-
firmed by NMR and optical rotation, it was important to
assure that this product possesses its biochemical activity
measured by an enzymatic assay.[27] Therefore, our sample
of G-6-P was converted by glucose 6-phosphate dehydroge-
nase (EC 1.1.1.49) to 6-phosphogluconolactone following a
literature protocol.[28] In this experiment, we also compared
the activity of our product with three commercial G-6-P
samples purchased from Sigma, Fluka and Roche (Fig-
ure 1). This data provided convincing evidence that the G-
6-P prepared by our protocol is biologically active.
Scheme 4.
Phosphorylation Studies of 3 Leading to Synthesis of G-6-P
With a convenient synthesis of 3 in hand, the stage was
set for phosphorylation at the 6-OH group. We decided to
use a mixture of 3 and 4 (97:3 ratio) assuming that phos-
phorylation of the C4 secondary hydroxy group of the
minor impurity will be significantly slower and may not
compete with the reaction of the primary hydroxy group
in 3. Thus, treatment of the sugars 3 and 4 with diphenyl
chlorophosphate in pyridine and DMAP gave the phos-
phate 7 in a moderate yield of 58% (Scheme 5). The reac-
tion was further optimized by performing phosphorylation
in dichloromethane as a solvent and Et3N as a base instead
of pyridine. Phosphate 7 was formed in 85% isolated yield
under these conditions. A trace amount of the undesired
phosphate 8 was also observed. This by-product was sepa-
rated easily from 7 by filtration through silica gel followed
by precipitation in CHCl3/n-hexane.
Figure 1. Activity of different commercial G-6-P sources compared
with G-6-P synthesized in this study (10).
Molecular Modeling Studies
The exceptionally high regio- and stereoselectivity ob-
served with the CRL-catalyzed hydrolysis of the anomeric
mixture of peracetylated -glucose motivated us to under-
take a modeling study of this reaction to better understand
the results. Thus, we obtained the X-ray structure of the
open form of CRL (1LPM)[29] from the Protein Data Bank.
This structure presents an inhibitor covalently linked to the
active site. The catalytic triad is formed by residues Glu341–
Scheme 5.
Reductive cleavage of the phenyl groups in the phosphate His449–Ser209 and the oxyanion hole is formed by the
was carried out by hydrogenolysis with platinum(IV) oxide main chain NH groups of residues Gly123, Gly124, and
to give the deprotected phosphate 9 in quantitative yield. Ala210.[29,30]
Eur. J. Org. Chem. 2007, 2769–2778
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2773

M. Ferrero, V. Gotor
FULL PAPER
To qualitatively explain the regioselectivity of CRL, we clashes within the phosphonate. After several arrangements
modeled phosphonate analogs of the key intermediates for of the α-anomer sugar within the active site of the lipase, a
hydrolysis of both anomers of peracetylated glucose 2. Both model (Figure 3, A) in which this substrate perfectly fitted
phosphonates were able to mimic key features of the inter-
mediates in hydrolysis reactions and allowed the computa-
tionally simpler molecular mechanics approach to be used.
This approach focuses on how the substrate fits in the en-
zyme, but might omit subtle details about the transition
state. We started with a simplified phosphonate that mim-
icked the tetrahedral intermediate for the hydrolysis of ethyl
acetate (Figure 2). Geometry optimization yielded a struc-
ture containing all six catalytically essential hydrogen bonds
in the catalytic site. To model the sugar substrates, we re-
placed the ethoxy moiety with the α- or β--glucose group
linked to the phosphonate at 6-position and then added the
other four acetate groups. We carried out a systematic se-
arch in order to identify the catalytically productive confor-
modeling. Key hydrogen bonds between CRL and the tetrahedral
Figure 2. Tetrahedral intermediate phosphonate transition-state
analog used to mimic the hydrolysis of ethyl acetate in computer
mations (see Exp. Sect.), which we defined as those that: a)
contained all six catalytically essential hydrogen bonds (a–
e, in addition of key hydrogen bond between Nδ of His449
and the carboxylate of Glu341) shown in Figure 2, b) avo-
ided steric hindrances between the peracetylated glucose
intermediate analog are: two from Nε of His449 to the oxygen
atom of Ser209 (a) and the OEt group of the tetrahedral intermedi-
ate (b), and three from the oxyanion to Ala210 (c), to Gly124 (d),
and to Gly123 (e). A sixth key hydrogen bond is from Nδ of His449
to the carboxylate of Glu341 (not shown). Further, the OEt group
is replaced by α- or β-glucose peracetylated in a stepwise fashion
phosphonate and the lipase, and c) avoided internal steric as described in the text.
Figure 3. Best models of phosphonate analogs for CRL-catalyzed hydrolysis of peracetylated glucose. Residues surrounding the active
site (Glu208, Gly122, Phe133, Phe296, Ile297, Phe344, Phe345, Ser450, and Ile453) are colored dark blue. The oxyanion hole residues
Gly123, Gly124, and Ala210 are colored red, green, and yellow, respectively. A) Peracetylated α-glucose conformation of the phosphonate
mimics hydrolysis of the 6-acetate group, it places the anomeric acetate carbonyl group close to Ser450 forming an additional hydrogen
bond (bond f, in a detailed view in panel B), in addition of the key hydrogen bonds (a–e in panel B, a sixth key hydrogen bond between
Glu341 and His449 is not shown). C) Peracetylated β-glucose conformation of the phosphonate does not form the same additional
hydrogen bond f because of the bad angle [f = 2.82 Å, angle (O–H–O) 84°], and also loses three key hydrogen bonds (bonds b, c, and e;
panel D; see text). To allow a better view of the active site, panel A (Phe344 and Leu445) and panel C (Phe344) display some amino
acids in a line representation.
2774
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2769–2778

Chemoenzymatic Synthesis of -Glucose 6-Phosphate
FULL PAPER
was derived. This model was devoid of any steric hin- ity of an enzyme.[33] The key difference between the inter-
drances and maintained all the six key hydrogen-bond inter- mediates obtained for both peracetylated anomers of -glu-
actions (Figure 3, B).[31] The sugar was partly shared be- cose is the superior binding of the α-intermediate caused at
tween the acyl pocket and the alcohol pocket of the lipase, a site remote from the reaction center. Such remote binding
thus an additional H bond was observed between the an- can lower the free energy of the corresponding transition
omeric carbonyl oxygen and the OH of residue Ser450 (see state by favoring a catalytically productive orientation, and
part B of Figure 3, bond f, 2.84 Å).[32] These interactions at the same time disfavoring the intermediate for the ob-
can fix the sugar substrate in a desirable manner to react served β-anomer.
regioselectively towards the 6-O position.
When the β-anomer was built and minimized, no single
Experimental Section
productive conformation was obtained. In all structures
some of the key H-bond interactions were lost. Finally, we
tried to place the β-anomer structure in a similar way as
that for the α-anomer (Figure 3, C); the substrate moved
away from the catalytic His449 (b = 3.63 Å, Figure 3, D)
site and two of the oxyanion key hydrogen bonds [c =
2.88 Å, angle (N–H–O) 117°; e = 3.59 Å, Figure 3, D] were
lost.[31]
General: Candida rugosa lipase (CRL, Type VII, 1410 U/mg) was
purchased from Sigma. [α]D values are given in 10–1 degcm2 g–1.
Molecular Modeling: The program Insight II, version 2000.1, was
used for viewing the structures. The geometric optimizations and
molecular dynamics were performed using Discover, version 2.9.7
(Accelrys, San Diego CA, USA), using the AMBER[34] force field.
The distance dependent dielectric constant was set to 4.0 to mimic
the electrostatic shielding of the solvent and the 1–4 van der Waals
interactions were scaled to 50%. The crystal structure for the CRL
(1LPM[29]
)
was obtained from the Protein Data Bank
Conclusions
(www.pdb.org), and includes a phosphonate. The methodology fol-
lowed for localization of energy minima structures by means of
a systematically search has been described elsewhere (see also the
electronic Supporting Information).[35] Protein structures in Fig-
ure 3 were generated using PyMOL 0.99.[36]
In conclusion, the route detailed here has proven to be a
highly efficient method for the large-scale synthesis of -
glucose 6-phosphate. A versatile acetylating method, very
useful in carbohydrate chemistry has been developed based
on lower reaction temperature (–10 °C) and use of freshly
distilled acetic anhydride. Under these conditions, exclu-
sively 1,2,3,4,6-penta-O-acetyl-α--glucopyranose was ob-
tained. In addition, the enzymatic hydrolysis of peracety-
lated α--glucose catalyzed by the commercial Candida
rugosa lipase allowed the regioselective deprotection of the
C-6-acetyl group in excellent yield. An acetyl migration
1,2,3,4,6-Penta-O-acetyl-D-glucopyranose (2): Acetic anhydride
(12 mL, 0.13 mol) was added dropwise to a stirred solution of α-
-glucose (3.0 g, 17 mmol) in anhydrous pyridine (30 mL) at 0 °C.
After stirring at room temperature for 17 h, the solvent was evapo-
rated under vacuum, and the residue was purified by flash
chromatography (EtOAc/n-hexane, 1:2) to afford 6.2 g of 2 (ratio
α/β: 92:8) as a white solid (95% yield). 1H NMR (CDCl3,
300 MHz): δ = 2.01–2.18 (several s, 30 H, 10 CH3 α anomer and β
from 4-O to 6-O explains the formation of the 4-OH depro- anomer), 3.81–3.86 (m, 1 H, 5-H β anomer), 4.05–4.15 (m, 3 H, 5-
2
3
H, 6-H α anomer, 6-H β anomer), 4.25 (dd, JHH = 12.4, JHH
=
=
tected derivative as a minor byproduct. It is worth emphas-
izing that the preparation of G-6-P is accomplished in 77%
overall yield from readily available starting material and re-
quires only one chromatographic step. This work provides
further evidence that the combined use of enzymatic and
conventional chemical methods offers a convenient alterna-
tive to the synthesis of biological active molecule with a
host of pharmaceutical applications.
Results obtained from the molecular modeling study on
the enzymatic step are in agreement with the observed ex-
perimental data. Thus, review of our results indicated that
the location of the acetyl group in the anomeric position
has a strong influence in the reactivity of the enzyme. More
3
4.1 Hz, 2 H, 6-H α anomer and 6-H β anomer), 5.10 (dd, JHH
10.2, JHH = 3.5 Hz, 2 H, 2-H α anomer and 2-H β anomer), 5.14
3
3
(t, JHH = 9.8 Hz, 2 H, 4-H α anomer and 4-H β anomer), 5.24 (t,
3
3JHH = 9.2 Hz, 1 H, 3-H β anomer), 5.47 (t, JHH = 9.9 Hz, 1 H,
3
3-H α anomer), 5.69 (d, JHH = 8.1 Hz, 1 H, 1-H β anomer), 6.39
3
(d, JHH = 3.8 Hz, 1 H, 1-H α anomer) ppm.
1,2,3,4,6-Penta-O-acetyl-α-D
-glucopyranose (α-2):[17a] Freshly dis-
tilled acetic anhydride (12 mL, 0.13 mol) was added dropwise to a
stirred solution of α--glucose (3.0 g, 17 mmol) in anhydrous pyri-
dine (15 mL) at –10 °C, and the stirring continued at the same tem-
perature for 17 h. After addition of ice into the reaction mixture a
powdery white solid precipitated. Filtration of the solid affords
6.3 g of α-2 (96% yield). Rf (EtOAc) = 0.64; m.p. 110–112 °C.
importantly, the existence of this interaction between the 1- [α]2D0 = +101 (c = 1.0, in CHCl ). IR (KBr): ν = 2940, 2875,
˜
3
1753 cm–1. H NMR (CDCl3, 300 MHz): δ = 2.03 (s, 3 H, CH3),
1
O carbonyl group and Ser450 could explain the data ob-
tained when another type of moiety is present in the anom-
eric position. In this sense, it has been described by Terreni
2.05 (s, 3 H, CH3), 2.07 (s, 3 H, CH3), 2.10 (s, 3 H, CH3), 2.19 (s,
2
3 H, CH3), 4.08–4.15 (m, 2 H, 5-H, 6-H), 4.28 (dd, JHH = 12.6,
3JHH = 4.2 Hz, 1 H, 6-H), 5.11 (dd, JHH = 10.1, JHH = 3.7 Hz, 1
3
3
et al.[17b] that when the acetyl group in the β-anomer is
changed by a methoxy ether (smaller group) or a butoxy
ether (bigger moiety), the CRL-catalyzed hydrolysis in the
6-position is accomplished, pointing out again the relevance
of the nature of the group present at the anomeric carbon.
This study constitutes another fine example of how re-
3
3
H, 2-H), 5.14 (t, JHH = 9.6 Hz, 1 H, 4-H), 5.48 (t, JHH = 9.6 Hz,
1 H, 3-H), 6.33 (d, JHH = 3.7 Hz, 1 H, 1-H) ppm. 13C NMR
3
(CDCl3, 75.5 MHz): δ = 20.3 (CH3), 20.4 (CH3), 20.5 (CH3), 20.6
(CH3), 20.7 (CH3), 61.3 (CH2), 67.8 (C-4), 69.1 (C-2), 69.7 (C-3,
C-5), 88.9 (C-1), 168.6 (C=O), 169.3 (C=O), 169.5 (C=O), 170.1
(C=O), 170.5 (C=O) ppm. MS (ESI+ ): m/z (%) = 413 (100) [M +
mote interactions can affect the stereo- and/or regioselectiv- Na]+.
Eur. J. Org. Chem. 2007, 2769–2778
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2775

M. Ferrero, V. Gotor
FULL PAPER
Enzymatic Hydrolysis of α-2: A mixture of α-2 (5.0 g, 12.81 mmol),
Candida rugosa lipase (2.5 g), and 25 m phosphate buffer pH 4
which contain 20% of 1,4-dioxane (600 mL) was stirred (250 rpm)
at 30 °C during 48 h. The progress of the reaction was monitored
by TLC (EtOAc). The suspension was filtered through Celite, the
enzyme washed with 1,4-dioxane and EtOAc, and the residue ex-
tracted with EtOAc. Solvents were evaporated under vacuum, and
the residue analyzed by 1H NMR, which indicated a 3/4 ratio of
95:5. No further purification was necessary. Yield: 95% (4.2 g).
Pure regioisomer 3 can be isolated by crystallization in diethyl
ether/n-hexane (4:1).
orated under vacuum, and the residue subjected to flash
chromatography (EtOAc/n-hexane, 2:1), affording 5 as a white solid
(116 mg, 88% yield). Rf (50% n-hexane/EtOAc) = 0.43; m.p. 125–
127 °C. [α]2D0 = +99 (c = 1.1, in CHCl ). IR (KBr): ν 3063, 2940,
˜
3
1757, 1724 cm–1; 1H NMR (CDCl3, 600 MHz): δ = 2.03 (s, 3 H,
CH3), 2.04 (s, 3 H, CH3), 2.05 (s, 3 H, CH3), 2.19 (s, 3 H, CH3),
2
3
4.23–4.28 (m, 1 H, 5-H), 4.39 (dd, JHH = 12.2, JHH = 3.9 Hz, 1
2
3
H, 6-H), 4.49 (dd, JHH = 12.4, JHH = 2.4 Hz, 1 H, 6-H), 5.12
(dd, 3JHH = 10.1, 3JHH = 3.8 Hz, 1 H, 2-H), 5.25 (t, 3JHH = 9.9 Hz,
3
3
1 H, 4-H), 5.52 (t, JHH = 9.8 Hz, 1 H, 3-H), 6.36 (d, JHH
=
3.8 Hz, 1 H, 1-H), 7.44 (t, 3JHH = 7.3 Hz, 2 H, H-m), 7.56 (t, JHH
3
= 7.5 Hz, 1 H, H-p), 8.04 (d, JHH = 7.8 Hz, 2 H, H-o) ppm. 13C
3
1,2,3,4,6-Penta-O-acetyl-D
-glucopyranose [6-O-acetyl-1Ј-13C] (α-
NMR (CDCl3, 150.5 MHz): δ = 20.4 (CH3), 20.5 (CH3), 20.6
(CH3), 20.8 (CH3), 61.9 (CH2), 68.2 (C-4), 69.2 (C-2), 69.8 (C-3,
C-5), 89.0 (C-1), 128.4 (C-m), 129.4 (C-i), 129.7 (C-o), 133.2 (C-p),
166.1 (Ph-C=O), 168.7 (C=O), 169.3 (C=O), 169.6 (C=O), 170.2
(C=O) ppm. MS (ESI+ ): m/z (%) = 475 (100) [M + Na]+.
2*): Acetic anhydride-1,1Ј-13C2 (122 µL, 1.29 mmol) was added
dropwise to a stirred solution of 1,2,3,4-tetra-O-acetyl-α--gluco-
pyranose (300 mg, 0.86 mmol) in anhydrous pyridine (1 mL) at
0 °C. After stirring at room temperature for 16 h, the solvent was
evaporated under vacuum, and the residue was purified by flash
chromatography (EtOAc/n-hexane, 1:2) to afford 335 mg of α-2* as
a white solid (99% yield). Rf (EtOAc) = 0.64. 1H NMR (CDCl3,
600 MHz): δ = 2.00 (s, 3 H, CH3), 2.02 (s, 3 H, CH3), 2.03 (s, 3 H,
Enzymatic Hydrolysis of 5: A mixture of 5 (48.3 mg, 0.107 mmol),
Candida rugosa lipase (24 mg), and 25 m phosphate buffer pH 4
which contain 20% of 1,4-dioxane (6 mL) was stirred (250 rpm) at
30 °C during 49 h. The suspension was filtered through Celite, the
enzyme washed with 1,4-dioxane and EtOAc, and the residue ex-
tracted with EtOAc. The combined organic layers were washed
with aqueous NaHCO3. Solvents were evaporated under vacuum,
2
CH3), 2.08 (d, JHC = 7.0 Hz, 3 H, CH3), 2.17 (s, 3 H, CH3), 4.08–
2
3
4.13 (m, 2 H, 5-H, 6-H), 4.25 (dt, JHH = 12.1, JHH = 4.2 Hz, 1
3
3
H, 6-H), 5.09 (dd, JHH = 10.3, JHH = 3.7 Hz, 1 H, 2-H), 5.13 (t,
3
3JHH = 9.8 Hz, 1 H, 4-H), 5.46 (t, JHH = 9.8 Hz, 1 H, 3-H), 6.32
1
(d, 3JHH = 3.7 Hz, 1 H, 1-H) ppm. 13C NMR (CDCl3, 150.5 MHz):
and the residue analyzed by H NMR, which indicated a 3/4 ratio
2
of 97:3. Yield: 91% (34 mg).
δ = 20.3 (CH3), 20.4 (CH3), 20.5 (CH3), 20.6 (CH3), 20.7 (d, JCC
= 59 Hz, CH3), 20.8 (CH3), 61.3 (CH2), 67.8 (C-4), 69.1 (C-2), 69.7
(C-3, C-5), 88.9 (C-1), 168.6 (C=O), 169.3 (C=O), 169.6 (C=O),
170.1 (C=O), 170.5 (13C=O) ppm. MS (ESI+ ): m/z (%) = 414 (100)
[M + Na]+.
1,2,3,4-Tetra-O-acetyl-6-O-diphenyloxyphosphoryl-α-D-gluco-
pyranose (7): Diphenyl chlorophosphate (1.7 mL, 8.31 mmol), Et3N
(1.2 mL, 8.31 mmol), and a catalytic amount of DMAP were added
to a solution of 3/4 (ratio α/β: 95:5), (2.41 g, 6.92 mmol) in anhy-
drous CH2Cl2 (43 mL). The mixture was stirred at room tempera-
ture for 6 h. MeOH (3 mL) was added and after 30 min of ad-
ditional stirring, solvents were evaporated under vacuum. The yel-
low-orange residue was filtered through a short column of silica
gel (EtOAc/n-hexane, 2:3), and the resulting white solid (mixture
of 7/8) was dissolved in CHCl3 and precipitated in n-hexane to
afford 3.5 g of 7 (85% yield). 7: Rf (50% n-hexane/EtOAc): 0.26;
1,2,3,4-Tetra-O-acetyl-α-
D-glucopyranose and 1,2,3,6-Tetra-O-ace-
tyl-α-
D
-glucopyranose (3/4): Hygroscopic white solid. 1H NMR
(CDCl3, 300 MHz): δ = 2.01 (s, 3 H, CH3 of 3), 2.04 (s, 3 H, CH3
of 3), 2.07 (s, 3 H, CH3 of 3), 2.11 (s, 3 H, CH3 of 4), 2.13 (s, 3 H,
2
3
CH3 of 4), 2.17 (s, 3 H, CH3 of 3), 3.58 (dd, JHH = 12.7, JHH
=
2
3
4.1 Hz, 2 H, 6-H of 3 and 4-H of 4), 3.72 (dd, JHH = 12.7, JHH
= 2.3 Hz, 1 H, 6-H of 3), 3.92 (ddd, 3JHH = 10.2; 3JHH = 4.1, 3JHH
= 2.2 Hz, 2 H, 5-H of 3, and 5-H of 4), 4.25 (dd, 2JHH = 12.6, 3JHH
m.p. 97–100 °C. [α]2D0 = +74 (c = 1.1, in CHCl ). IR (KBr): ν 3054,
˜
3
2965, 1738, 1590 cm–1. H NMR (CDCl3, 400 MHz): δ = 1.99 (s,
1
2
3
= 2.3 Hz, 1 H, 6-H of 4), 4.51 (dd, JHH = 12.6, JHH = 3.7 Hz, 1
H, 6-H of 4), 5.07 (dd, 3JHH = 10.2, JHH = 3.7 Hz, 2 H, 2-H of 3,
3
3 H, CH3), 2.01 (s, 3 H, CH3), 2.02 (s, 3 H, CH3), 2.11 (s, 3 H,
CH3), 4.12–4.38 (m, 3 H, 5-H, 2ϫ6-H), 4.97 (dd, 3JHH = 10.2, 3JHH
and 2-H of 4), 5.10 (t, 3JHH = 9.9 Hz, 1 H, 4-H of 3), 5.33 (t, JHH
3
3
3
= 3.7 Hz, 1 H, 2-H), 5.08 (t, JHH = 10.1 Hz, 1 H, 4-H), 5.44 (t,
= 10.1 Hz, 1 H, 3-H of 4), 5.52 (t, JHH = 9.9 Hz, 1 H, 3-H of 3),
6.29 (d, JHH = 3.7 Hz, 1 H, 1-H of 4), 6.34 (d, JHH = 3.7 Hz, 1
H, 1-H of 3) ppm.
3JHH = 9.8 Hz, 1 H, 3-H), 6.25 (d, 3JHH = 3.8 Hz, 1 H, 1-H), 7.19–
3
3
7.34 (m, 10 H aromatics) ppm. 13C NMR (CDCl3, 100.6 MHz): δ
2
= 20.3 (CH3), 20.4 (CH3), 20.5 (CH3), 20.7 (CH3), 66.4 (d, JCP
=
=
1,2,3,4-Tetra-O-acetyl-α-D
-glucopyranose (3):[17a] White solid. Rf
3
5.2 Hz, CH2), 67.8 (C-4), 68.9 (C-2), 69.7 (C-3), 70.2 (d, JCP
(EtOAc) = 0.25; m.p. 102–105 °C. [α]2D0 = +117 (c = 1.1, in CHCl3).
3
6.9 Hz, C-5), 88.7 (C-1), 119.9 (apparent t, JCP = 4.7 Hz, C-o),
1
IR (KBr): ν 3501, 2962, 1752 cm–1. H NMR (CDCl , 300 MHz):
˜
3
2
125.4 (C-p) ppm. 129.7 (C-m), 150.2 (d, JCP = 4.6 Hz, C-i), 168.5
δ = 2.00 (s, 3 H, CH3), 2.02 (s, 3 H, CH3), 2.06 (s, 3 H, CH3), 2.16
(C=O), 169.2 (C=O), 169.4 (C=O), 170.1 (C=O) ppm. 31P NMR
(CDCl3, 162.0 MHz): δ = –11.99 ppm. HRMS (EI): calcd for
C26H29O13P (M+): 580.13458. found 580.13459. 8: Rf (50% n-hex-
2
3
(s, 3 H, CH3), 3.57 (dd, JHH = 12.6, JHH = 4.3 Hz, 1 H, 6-H),
2
3
3
3.71 (dd, JHH = 12.6, JHH = 2.4 Hz, 1 H, 6-H), 3.91 (ddd, JHH
= 10.3, JHH = 4.1, JHH = 2.2 Hz, 1 H, 5-H), 5.06 (dd, JHH
10.3, JHH = 3.7 Hz, 1 H, 2-H), 5.10 (t, JHH = 9.5 Hz, 1 H, 4-H),
5.50 (t, JHH = 9.9 Hz, 1 H, 3-H), 6.33 (d, JHH = 3.7 Hz, 1 H, 1-
H) ppm. 13C NMR (CDCl3, 75.5 MHz): δ = 20.3 (CH3), 20.4
(CH3), 20.5 (CH3), 20.7 (CH3), 60.7 (CH2), 68.1 (C-4), 69.2 (C-2),
69.5 (C-3), 72.0 (C-5), 89.0 (C-1), 169.0 (C=O), 169.6 (C=O), 170.0
(C=O), 170.1 (C=O) ppm. MS MS (ESI+ ): m/z (%) = 471 (100)
[M + Na]+.
3
3
3
=
ane/EtOAc) = 0.39. [α]2D0 = +40 (c = 1.7, in CHCl ). IR (NaCl): ν
˜
3
3
3
2960, 1756 cm–1. 1H NMR (CDCl3, 400 MHz): δ = 1.87 (s, 3 H,
3
3
CH3), 1.96 (s, 6 H, 2CH3), 2.14 (s, 3 H, CH3), 4.07–4.14 (m, 2 H,
2
5-H, 6-H), 4.25 (apparent d, JHH = 11.4 Hz, 1 H, 6-H), 4.81 (c,
3
3
3JHH = 9.4 Hz, 1 H, 4-H), 5.04 (dd, JHH = 10.2, JHH = 3.8 Hz, 1
3
3
H, 2-H), 5.61 (t, JHH = 9.8 Hz, 1 H, 3-H), 6.28 (d, JHH = 3.6 Hz,
1 H, 1-H), 7.12–7.31 (m, 10 H aromatics) ppm. 13C NMR (CDCl3,
100.6 MHz): δ = 20.1 (CH3), 20.3 (CH3), 20.5 (CH3), 20.6 (CH3),
61.2 (CH2), 69.1 (C-2), 69.4 (C-3), 69.9 (d, JCP = 6.2 Hz, C-5),
73.3 (d, JCP = 4.3 Hz, C-4), 88.6 (C-1), 119.6 (apparent t, JCP =
3
1,2,3,4-Tetra-O-acetyl-6-O-benzoyl-α-
D-glucopyranose (5): Benzoyl
2
3
chloride (34 µL, 0.293 mmol) was added dropwise to a solution of
3 (102 mg, 0.293 mmol) in anhydrous pyridine (0.5 mL) at 0 °C.
After stirring at room temperature overnight, the solvent was evap-
2
5.5 Hz, C-o), 125.4 (C-p), 129.6 (C-m), 149.9 (d, JCP = 5.3 Hz, C-
i), 150.0 (d, JCP = 5.6 Hz, C-i), 168.3 (C=O), 169.3 (C=O), 169.8
2
2776
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2769–2778

Chemoenzymatic Synthesis of -Glucose 6-Phosphate
FULL PAPER
(C=O), 170. (C=O) ppm. 31P NMR (CDCl3, 162.0 MHz): δ =
–12.61 ppm. MS (ESI+ ): m/z (%) = 581 (19) [M + H]+, 603 (100)
[M + Na]+ 619 (55) [M + K]+.
[4]
[5]
a) T. Eguchi, S. Sasaki, Z. Huang, K. Kakinuma, J. Org. Chem.
2002, 67, 3979–3984; b) N. Yamauchi, K. Kakinuma, J. Org.
Chem. 1995, 60, 5614–5619; c) Aminocyclitol antibiotics (Ed.:
K. L. Rinehart Jr, T. Suami), ACS Symposium Series No.125;
American Chemical Society, Washington, DC, 1980; d) K. L.
Rinehart Jr, R. M. Stroshane, J. Antibiotic. 1976, 29, 319–353.
F. Foufelle, B. Gouhot, J. P. Pegorier, D. Perdereau, J. Girard,
P. Ferre, J. Biol. Chem. 1992, 267, 20543–20546.
R. Robison, E. J. King, Biochem. J. 1931, 25, 323–328.
a) D. C. Crans, G. M. Whitesides, J. Org. Chem. 1983, 48,
3130–3132; b) A. Pollak, R. L. Baughn, G. M. Whitesides, J.
Am. Chem. Soc. 1977, 99, 2366–2367.
B. P. Colowick, E. W. Sutherland, J. Biol. Chem. 1942, 44, 423–
437.
C. Neuberg, H. Lustig, M. A. Rathenberg, Arch. Biochem.
1943, 3, 33–44.
T. van Herk, A. F. Hartog, A. M. van der Burg, R. Wever, Adv.
Synth. Catal. 2005, 347, 1155–1162.
a) H. A. Lardy, H. O. L. Fischer, Biochem. Prep. 1952, 2, 39–
44; b) H. A. Lardy, H. O. L. Fischer, J. Biol. Chem. 1946, 164,
513–519.
R. L. Whistler, L. W. Doner, K. Martin, Methods Carbohydr.
Chem. 1972, 6, 411–412.
a) R. Hua, S. Zhang, Y. Song, Huaxue Tongbao 1992, 33–34;
b) K. Schutzner, Cesk. Farm. 1974, 23, 34–36; Chem. Abstr.
1974, 81, 13715; c) P. A. Levene, A. L. Raymond, J. Biol. Chem.
1931, 92, 757–763.
a) F. R. Zuleski, E. T. McGuinness, J. Labelled Compd. 1969,
5, 371–376; b) M. Viscontini, C. Oliver, Helv. Chim. Acta 1953,
36, 466–470; c) J. E. Seegmiller, B. L. Horecker, J. Biol. Chem.
1951, 192, 175–180.
1,2,3,4-Tetra-O-acetyl-α-D-glucopyranosyl-6-phosphoric Acid (9): A
solution of 7 (2.29 g, 3.95 mmol) in anhydrous MeOH (23 mL) was
added to a flask containing PtO2 (445 mg, 1.96 mmol) and stirred
under a positive pressure of hydrogen (balloon). The suspension
was stirred for 16 h at room temperature. Then, the mixture was
filtered through Celite and concentrated to give 9 as a white solid
in quantitative yield (1.69 g). The crude was sufficiently pure for
[6]
[7]
the next step. [α]2D0 = +91 (c = 0.8, in MeOH). IR (KBr): ν 3423,
˜
[8]
1638 cm–1. 1H NMR ([D4]MeOH, 300 MHz): δ = 2.15 (s, 3 H,
CH3), 2.17 (s, 6 H, CH3), 2.27 (s, 3 H, CH3), 2.37 (s, 3 H, CH3),
[9]
3
3
4.16–440 (m, 3 H, 5-H, 26-H), 5.25 (dd, JHH = 10.2, JHH
=
3
3
3.4 Hz, 1 H, 2-H), 5.34 (t, JHH = 9.3 Hz, 1 H, 4-H), 5.63 (t, JHH
[10]
[11]
= 9.8 Hz, 1 H, 3-H), 6.49 (d, JHH = 3.4 Hz, 1 H, 1-H) ppm. 13C
3
NMR ([D4]MeOH, 75.5 MHz): δ ppm = 20.3 (CH3), 20.6 (2CH3),
20.7 (CH3), 65.6 (CH2), 69.4 (C-4), 70.7 (C-2), 71.3 (C-3), 71.9 (C-
5), 90.1 (C-1), 170.6 (C=O), 171.1 (C=O), 171.3 (C=O), 171.7
(C=O) ppm. 31P NMR ([D4]MeOH, 121.5 MHz): δ = 0.22 ppm. [12]
MS (ESI+ ): m/z (%) = 429 (5) [M + H]+, 451 (38) [M + Na]+, 619
(23) [M + K]+.
[13]
D-Glucose 6-Phosphate (10): Potassium carbonate (387 mg,
2.80 mmol) was added to a solution of 9 (1 g, 2.34 mmol) in anhy-
drous MeOH (40 mL) at 0 °C. After being stirred for 28 h at 0 °C
a precipitate was observed, which was isolated by filtration. The
white solid was dissolved in water and Dowex 50Wx4–400 (H+
form) was added. The mixture was stirred for 30 min, the resin was
filtered, and the filtrate evaporated to afford 602 mg of 10 (99%
[14]
[15]
[16]
a) C.-H. Wong, G. M. Whitesides, J. Am. Chem. Soc. 1981, 103,
4890–4899; b) C.-H. Wong, J. Gordon, C. L. Cooney, G. M.
Whitesides, J. Org. Chem. 1981, 46, 4676–4679.
yield) (ratio α/β: 40:60). [α]2D0 = +24 (c = 1.1, in D2O). H NMR
1
3
(D2O, 400 MHz): δ = 3.10 (t, JHH = 8.1 Hz, 1 H, 2-H β anomer),
a) F. Ganske, U. T. Bornscheuer, Org. Lett. 2005, 7, 3097–3098;
b) T.-C. Chien, J.-W. Chern, Carbohydr. Res. 2004, 339, 1215–
1217; c) M.-J. Kim, M. Y. Choi, J. K. Lee, Y. Ahn, J. Mol.
Catal. B 2003, 26, 115–118; d) S. Park, R. J. Kazlauskas, J. Org.
Chem. 2001, 66, 8395–8401; e) R. Khan, L. Groppen, P. A.
Konpwicz, M. Matulová, S. Paoletti, Tetrahedron Lett. 1993,
34, 7767–7770; f) D. G. Drueckhamer, J. W. Hennen, R. L. Ped-
erson, C. F. Barbas, C. M. Gautheron, T. Krach, C.-H. Wong,
Synthesis 1991, 7, 499–525.
a) G. Fernández-Lorente, J. M. Palomo, J. Cocca, C. Mateo, P.
Moro, M. Terreni, R. Fernández-Lafuente, J. M. Guisán, Tet-
rahedron 2003, 59, 5705–5711; b) M. Terreni, R. Salvetti, L.
Linati, R. Fernández-Lafuente, G. Fernández-Lorente, A. Bas-
tida, J. M. Guisán, Carbohydr. Res. 2002, 337, 1615–1621; c) A.
Bastida, R. Fernández-Lafuente, G. Fernández-Lorente, J. M.
Guisán, Bioorg. Med. Chem. Lett. 1999, 9, 633–636.
T. Horrobin, C. H. Tran, D. Crout, J. Chem. Soc., Perkin Trans.
1 1998, 1069–1080.
3
3.31–4.11 (m, 11 H), 4.53 (d, JHH = 7.9 Hz, 1 H, 1-H β anomer),
5.10 (d, 3JHH = 3.5 Hz, 1 H, 1-H α anomer) ppm. 13C NMR (D2O,
100.6 MHz): δ = 66.8 (C-6 α anomer and β anomer), 70.2, 72.3 (d,
JCP = 8.1 Hz), 73.6, 74.8, 76.2, 76.6 (d, JCP = 8.1 Hz), 77.7, 94.4
(C-1 α anomer), 98.2 (C-1 β anomer) ppm. 31P NMR (D2O,
162.0 MHz): δ = 0.79 ppm. MS (ESI–): m/z (%) = 259 (100) [(M –
H]–.
Supporting Information (see also the footnote on the first page of
this article): General spectroscopic and experimental data, and mo-
lecular modeling calculations description are shown. Furthermore,
copies of 1H, 13C, DEPT, and 31P NMR spectra in addition of
some 2D NMR experiments are included.
[17]
[18]
Acknowledgments
[19]
[20]
P. F. Schatz, J. Chem. Edu. 2001, 78, 1378.
The authors are grateful to Dr. P. Herrero for the support in the
enzymatic assays of G-6-P. Financial support of this work by the
European Community, Sixth Framework Program (EU-04-LSHB-
2003-503017) is gratefully acknowledged. S. F. thanks the Minis-
terio de Educación y Ciencia (MEC) for a personal grant (Ramón
y Cajal Program).
α--1,2,3,4,6-pentaacetylglucose purchased from Innovassynth
Technologies (www.Innovassynth.com) had 6% of β--
1,2,3,4,6-pentaacetylglucose.
P. M. Collins, R. J. Ferrier In Monosaccharides. Their Chemis-
try and Their Roles in Natural Products, John Wiley & Sons,
New York, 1995, pp. 360.
[21]
[22]
[23]
A. Procopio, R. Dalpozzo, A. De Nino, L. Maiuolo, B. Russo,
G. Sindona, Adv. Synth. Catal. 2004, 346, 1465–1470.
a) G. Bartoli, R. Dalpozzo, A. De Nino, L. Maiuolo, M.
Nardi, A. Procopio, A. Tagarelli, Green Chem. 2004, 6, 191–
192; b) R. Dalpozzo, A. De Nino, L. Maiuolo, A. Procopio,
M. Nardi, G. Bartoli, R. Romeo, Tetrahedron Lett. 2003, 44,
5621–5624.
a) A. K. Misra, P. Tiwari, S. K. Madhusudan, Carbohydr. Res.
2005, 340, 325–329; b) A. K. Chakraborti, R. Gulhane, Chem.
Commun. 2003, 1896–1897.
[1] Fr Pat. 5782 19680318, 1968; CAN 71:3626. Hypertensive glu-
cose 6-(sodium hydrogen phosphate).
[2] D. Champmartin, P. Rubini, A. Lakatos, T. Kiss, J. Inorg. Bio-
chem. 2001, 84, 13–21.
[3] a) T. Aoki, K. Kitahata, T. Fukumoto, Y. Sugimoto, H. R.
Ibrahim, T. Kimura, Y. Kato, T. Matsuda, Food Res. Int. 1997,
30, 401–406; b) Y. Kato, T. Aoki, N. Kato, N. Nakamura, T.
Matsuda, J. Agric. Food Chem. 1995, 43, 301–305; c) G. Ma-
theis, Food Chem. 1991, 39, 13–26.
[24]
Eur. J. Org. Chem. 2007, 2769–2778
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
2777

M. Ferrero, V. Gotor
FULL PAPER
[25] K. P. R. Kartha, R. A. Field, Tetrahedron 1997, 53, 11753–
11766.
[26] A. Orita, Y. Hamada, T. Nakano, S. Toyoshima, J. Otera,
Chem. Eur. J. 2001, 7, 3321–3327.
[27] E. Postma, C. Verduyn, W. A. Scheffers, J. P. van Dijken, Appl.
Environ. Microbiol. 1989, 55, 468–477.
[28] Spectrometric assays were carried out at 340 nm and 25 °C.
The reaction rates were proportional to the amount of G-6-P
added. The assay mixture contained Tris hydrochloride buffer
(pH 8, 50 m), MgCl2 (5 m), and NADP+ (0.4 m). The re-
action was started with 5 m glucose 6-phosphate.
[29] P. Grochulski, P. Bouthillier, R. J. Kazlauskas, A. N. Serreqi,
J. D. Schrag, E. Ziomek, M. Cygler, Biochemistry 1994, 33,
3494–3500.
[32] It has been described previously another example where this
amino acid seemed to have a remarkable effect in the hydrolysis
of 1-pyridylethanol esters. See, for instance: F. Bellezza, A. Cip-
iciani, G. Cruciani, F. Fringuelli, J. Chem. Soc., Perkin Trans.
1 2000, 4439–4444.
[33] For other examples of influenced selectivity in enzymes by re-
mote interactions, see: a) I. Lavandera, S. Fernández, J. Mag-
dalena, M. Ferrero, H. Grewal, C. K. Savile, R. J. Kazlauskas,
V. Gotor, ChemBioChem 2006, 7, 693–698; b) B. Botta, G. Zap-
pia, A. Tafi, M. Botta, F. Manetti, E. Cernia, G. Milana, C.
Palocci, S. Soro, G. D. Monache, J. Mol. Catal. B 2002, 16,
241–247; c) D. G. Gascoyne, H. L. Finkbeiner, K. P. Chan, J. L.
Gordon, K. R. Stewart, R. J. Kazlauskas, J. Org. Chem. 2001,
66, 3041–3048.
[30] Other examples of molecular modeling or structural studies on
CRL: a) J. Schmitt, S. Brocca, R. D. Schmid, J. Pleiss, Protein
Eng. 2002, 15, 595–601; b) T. Schulz, R. D. Schmid, J. Pleiss,
J. Mol. Model. 2001, 7, 265–270; c) M. Cygler, J. D. Schrag,
Biochim. Biophys. Acta 1999, 1441, 205–214; d) M. Botta, E.
Cernia, F. Corelli, F. Manetti, S. Soro, Biochim. Biophys. Acta
1997, 1337, 302–310; e) M. Holmquist, F. Hæffner, T. Norin,
K. Hult, Protein Sci. 1996, 5, 83–88.
[34] a) W. D. Cornell, P. Cieplak, C. I. Bayly, P. A. Kollman, J. Am.
Chem. Soc. 1993, 115, 9620–9631; b) S. J. Weiner, P. A. Koll-
man, D. A. Case, U. C. Singh, C. Ghio, G. Alagona, S. Profeta,
P. Weiner, J. Am. Chem. Soc. 1984, 106, 765–784.
[35] I. Lavandera, S. Fernández, J. Magdalena, M. Ferrero, R. J.
Kazlauskas, V. Gotor, ChemBioChem 2005, 6, 1381–1390.
[36] DeLano, W. L. “The PyMOL Molecular Graphics System”;
DeLano Scientific LLC, San Carlos, CA, USA; http://www.py-
mol.org.
[31] To identify hydrogen bonds, a donor-acceptor separation of
less than 3.20 Å and a donor-hydrogen-acceptor angle of 120°
or greater are required.
Received: January 9, 2007
Published Online: April 12, 2007
2778
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 2769–2778