Regioselective C-6 hydrolysis of methyl O-benzoyl-pyranosides catalysed by candida rugosa lipase

Article abstract of DOI:10.1002/ejoc.200801268

Hydrolysis of six methyl O-benzoyl-pyranosides has been investigated using Candida rugosa lipase in dioxane/buffer mixtures. The lipase catalysed the hydrolysis of all substrates in a regiospecific manner at C-6, The rate of reaction was dependent on pyra

Full text of DOI:10.1002/ejoc.200801268

FULL PAPER
DOI: 10.1002/ejoc.200801268
Regioselective C-6 Hydrolysis of Methyl O-Benzoyl-pyranosides Catalysed by
Candida Rugosa Lipase
Aslan Esmurziev,[a,b] Eirik Sundby,[b] and Bård Helge Hoff*[a]
Keywords: Carbohydrates / Glycosides / Biocatalysis / Hydrolysis / Enzyme catalysis / Candida rugosa lipase
Hydrolysis of six methyl O-benzoyl-pyranosides has been in-
vestigated using Candida rugosa lipase in dioxane/buffer
mixtures. The lipase catalysed the hydrolysis of all substrates
in a regiospecific manner at C-6. The rate of reaction was
dependent on pyranoside structure, reaction temperature
and scale, dioxane concentration and agitation speed. Start-
ing from their C-6 O-benzoyl precursors, the methyl 2,3,4-
methyl 2,3,4,6-tetra-O-benzoyl-β-D-galactopyranoside sub-
strate inhibition were observed, which in part could be over-
come by increasing the reaction volume. Methyl 2,3,4,6-
tetra-O-benzoyl-β-
D
-glucopyranoside and methyl 2,3,4,6-
-mannopyranoside were poor substrates
tetra-O-benzoyl-α-
D
for Candida rugosa lipase and low degree of conversion
towards products were obtained under all conditions. No acyl
migration was detected in any of the products.
tri-O-benzoyl-pyranosides of α-
-glucose, and methyl 2,3-di-O-benzoyl-α-
oside could be isolated in 85–96% yield. In hydrolysis of
D-galactose, β-
D-galactose, α-
D
D-galactopyran-
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2009)
methyl 2,3,4,6-tetra-O-benzoyl-α-D-glucopyranoside and
pH 7 a mixture of the C-6 and C-4 deacetylated products
were obtained. Also the hydrolysis of methyl and butyl
2,3,4,6-tetra-O-acetyl α-- and β--glucopyranosides pro-
ceeded regioselectively at C-6 at pH 4 using 25% acetoni-
trile as a co-solvent.[20] Good regiocontrol was also ob-
served in hydrolysis of methyl 2,3,4,6-tetra-O-acetyl-α--
mannopyranoside.[28] Hydrolysis of acetylated hexa-
pyranosides using a PEG-modified variant of CRL in 1,1,1-
trichloroethane/water lead to mixtures of methyl 2,3-diace-
tyl hexapyranosides and methyl 2,3,4-triacetyl hexa-
pyranosides.[29] As an alternative, ethanolysis reactions can
be utilised. Using CRL, methyl 2,3,4-triacetyl-α--gluco-,
-β--gluco- and -α--mannopyranosides have been prepared
by this method.[28] Besides lipases, various esterases could
be envisioned as suitable catalysts for regioselective hydroly-
sis.[21,30]
A disadvantage with pyranoside acetates is their ten-
dency to undergo acyl migrations.[20,21,31] Benzoyl mi-
gration in carbohydrate chemistry is not unknown.[32,33]
However, due to their higher stability towards base,[34] ben-
zoate pyranosides are expected to be less problematic in this
respect compared to their acetylated analogues. The benzo-
ates also offer the possibility of analysing products by UV
detectors,[35] or by circular dichroism spectroscopy,[36]
which is often troublesome for acetate esters. In contrast to
the methyl 2,3,4-tri-O-acetyl-pyranosides, enzymatic ap-
proaches have not been developed or investigated in the
benzoyl series. This is probably due to expected challenges
with solubility, and the increase in size of the substrates
possibly limiting the number of useful catalysts for this
transformation. On this background, the present work de-
Introduction
Partially acylated pyranosides have widespread use in
carbohydrate chemistry. Derivatives containing a free C-6
hydroxy, such as methyl 2,3,4-tri-O-acyl-pyranosides, are
useful building blocks for uronic acid derivatives,[1] C-6
modified hexoses,[2,3] disaccharides,[4–8] and oligosaccha-
rides.[9–12] However, regioselective deprotection of carbo-
hydrates is a considerable challenge due to the presence of
multiple hydroxy groups of similar reactivity. Therefore,
non-enzymatic methods for synthesising 2,3,4-tri-O-ben-
zoyl-pyranosides includes a 3-step sequence of C-6 protec-
tion, acylation, and cleavage of the protection group. Com-
monly used protection groups includes trityl,[13,14] bulky
silane ethers,[15–17] bromoacetyl,[18] and formyl.[19]
As opposed to standard methods, regioselective hydroly-
sis catalysed by enzymes offers a two step route.[20–24] Can-
dida rugosa lipase (CRL) has proven to be the catalyst of
choice for deacetylation at C-6.[20,24] The lipase has a rather
wide active site and the ability to tolerate high concentra-
tions of organic co-solvents.[25–27] The hydrolysis of glucose
pentaacetate has been investigated at different pH by sev-
eral authors.[20,23,24] Reactions at pH 4–5 using CRL led to
high regioselectivity favouring hydrolysis at C-6, whereas at
[a] Department of Chemistry, Norwegian University of Science
and Technology,
Høgskoleringen 5, 7491 Trondheim, Norway
Fax: +47-72593973
E-mail: bhoff@chem.ntnu.no
[b] Department of Chemistry and Material Science, Sør-Trøndelag
University College,
E. C. Dahls gate 2, 7004 Trondheim, Norway
1592
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1592–1597

Regioselective C-6 Hydrolysis of Methyl O-Benzoyl-pyranosides
Scheme 1. Synthesis of starting materials: a: α--glucopyranoside, b: β--glucopyranoside, c: α--galactopyranoside, d: β--galactopyr-
anoside, e: α--mannopyranoside.
scribes regioselective enzymatic hydrolysis of methyl O-ben- served in acetone and 2-propanol, whereas reactions in di-
zoyl-pyranosides as a preparative route to derivatives hav- oxane and THF gave acceptable reaction rate. Dioxane (20
ing a free C-6 hydroxy group.
vol.-%) was selected as a co-solvent and hydrolysis of 1a–e
and 3c were investigated at different temperatures on 25-mg
scale (Scheme 2). Hydrolysis of all substrates proceeded in
a regiospecific manner at C-6.
Results and Discussion
Preparation of Substrates
The methyl 2,3,4,6-tetra-O-benzoyl-pyranosides 1a–e of
α--glucose, β--glucose, α--galactose, β--galactose, and
α--mannose, respectively, were synthesised from the corre-
sponding methyl -pyranosides using benzoyl chloride in
pyridine (Scheme 1).
Benzoylation of methyl α--galactopyranoside to 1c was
sluggish, and methyl 2,3,6-tri-O-benzoyl-α--galactopyr-
Scheme 2. CRL-catalysed hydrolysis of 1a–e and 3c. 1a–e: R = Bz,
anoside (3c) could easily be isolated.[37] Compound 3c was 3c: R = H.
included in the study to investigate the effect of steric crowd
on the enzymatic hydrolysis, and possibly to provide a new
The results of the screening experiments are summarised
method for the isolation of methyl 2,3-di-O-benzoyl-α-- in Table 1. The degree of conversion depended on the py-
galactopyranoside (4c).
ranoside structure and the reaction temperature.
Table 1. Effect of pyranoside structure and reaction temperature
on conversion in hydrolysis of 1a–e and 3c using CRL.
Lipase-Catalysed Regioselective Hydrolysis
Conversion (%)[b]
Initially, a screening using several lipases was performed.
Compound 3c was selected as a test substance because it Substrate Enzyme 20 °C[c] 30 °C[c] 40 °C[c]
30 °C[d]
Product
loading[a]
was expected to have the highest reactivity due to size ef-
fects and solubility. The screening revealed that only Can-
dida rugosa lipase (CRL) had an acceptable reaction rate.
No useful activity was observed using porcine pancreas li-
pase type 2, several formulations of Candida antarctica li-
pase A and B, Humicola insolens lipase (Lipozyme TL
100 L), Rhizomucor miehei lipase (Lipozyme RM IM) and
Rhizopus delemar lipase. This parallels the activity observed
1a
1b
1c
1d
1e
3c
0.5
1
0.5
0.5
1
35
0
66
42
0
71
6
67
70
4
47
4
88
91
4
85
6
75
86
6
2a
2b
2c
2d
2e
4c
0.5
98
99
97
99
[a] Weigh amount relative to substrate. [b] By HPLC. [c] Reaction
time: 20 h agitating at 140 rpm. [d] Reaction time 40 h agitating at
for various lipases towards other acetylated carbo- 140 rpm.
hydrates.[23,24]
A challenge in hydrolysis of benzoylated pyranosides is
The highest conversion rates were observed in hydrolysis
their low solubility in water, causing mass-transfer limita- of 3c, possibly due to a more accessible site of reaction.
tions and rate retardation. Therefore the hydrolytic activity Hydrolysis of the α--glucopyranoside 1a, proceeded with
of CRL towards 3c was examined in the presence of ace- the highest rate at 30 °C, whereas for the 1c–d the best con-
tone, THF, dioxane and 2-propanol. A low activity was ob- versions were experienced at 40 °C. Methyl 2,3,4,6-tetra-O-
Eur. J. Org. Chem. 2009, 1592–1597
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1593

A. Esmurziev, E. Sundby, B. H. Hoff
FULL PAPER
Table 2. Hydrolysis of selected substrates in acetate buffer containing 20% dioxane.[a]
Substrate
Agitation
(rpm)
Enzyme loading
(w/w)[b]
Reaction time [h]
Conversion
(%)[c]
Isolated yield
(%)
Product
1a
1b
1c
1e
3c
140
140
140
140
140
0.5
1
0.5
1
60
20
60
20
20
40
9
67
7
33
7
53
5
2a
2b
2c
2e
4c
0.5
82
76
[a] All reactions were performed in a 100-mg scale at 30 °C. [b] Weight amount relative to substrate. [c] By HPLC.
benzoyl-β--glucopyranoside (1b) and the α--mannopyr- after half of the reaction time (30 h). No change in conver-
anoside 1e, were hydrolysed at a low rate at all tempera- sion was the result. Therefore, deactivation of the enzyme
tures, and were poor substrates for CRL.
by dioxane is a not a likely explanation.
When the reaction scale was increased to 100 mg, a lower
Further, the effect of varying the volume and strength of
degree of conversion, and thus a lower yield were the result the buffer medium was investigated in hydrolysis of 1a, see
in hydrolysis of all substrates, see Table 2. The highest con- Table 4. By increasing the volume of the buffer solution
version was obtained in hydrolysis of 3c, reaching a moder- from 12 to 24 mL (Entry 2), the degree of conversion could
ate 82% conversion after 20 h reaction time.
be increased to 87%. The use of even higher volumes (Entry
Phase-transfer limitation was expected to be the main 3) did not improve the degree of conversion. Applying a
reason for the drop in conversion. Therefore, the effect of stronger buffer medium (Entry 4) led to a slight rate retar-
enzyme loading and agitation on conversion was investi- dation, possibly due to a salting out effect.
gated in 100-mg scale using 3c as substrate. By increasing
Table 4. Effect of buffer reaction volume and buffer strength in
agitation from 140 to 200 rpm and doubling catalyst load-
ing, full conversion could be obtained within 20 h when re-
hydrolysis of 1a.[a]
Entry
Buffer concentration
Volume
[mL]
Conversion
(%)
acting at 30 °C. Also, hydrolysis of 1c gave a high conver-
sion (96%) when applying the same agitation and catalyst
amount and reacting at 40 °C for 40 h.
[]
1
2
3
4
0.2
0.2
0.2
0.5
12
24
30
12
65[b]
87
83
Hydrolysis of the α--glucopyranoside 1a proved more
difficult to optimise. Experiments were undertaken varying
the dioxane concentration and enzyme loading, see Table 3.
The highest conversion obtained was 72%, applying 30%
dioxane as co-solvent. Moderate activity was observed at
lower (Entry 1) and higher (Entry 7) dioxane concentration.
Surprisingly, increasing the enzyme loading did not affect
degree of conversion to a significant extent (Entry 4 and 6).
The reaction was also performed in the presence of 1 mol
equivalent of benzoic acid (Entry 3). However, no signifi-
cant effect on conversion was detected.
60
[a] Conditions: 100 mg of 1a, temperature 30 °C, agitation: 200 rpm
for 60 h, enzyme loading: 1. [b] Table 3 Entry 2.
Then the effect of the reaction components 1a and 2a on
the performance of CRL was studied, using hydrolysis of
3c as model system (Table 5). In the presence of 2a (Entry
2), the degree of conversion of 3c dropped from 99 to 80%.
However, the effect of introducing 1a (Entry 3) upon the
equivalent reaction was much more profound and the de-
gree of conversion dropped to 32%. By increasing the en-
zyme loading higher conversions could be obtained, but still
the inhibiting action of 1a was evident.
Table 3. Effect of dioxane concentration, enzyme loading and ben-
zoic acid on conversion in hydrolysis of 1a.[a]
Entry
Dioxane
(%)
Enzyme loading
(w/w)
Conversion
(%)
Table 5. Effect of 1a and 2a on hydrolysis of 3c (140 rpm, 30 °C,
20 h).
1
2
3
4
5
6
7
10
20
20
20
30
30
40
1
1
1
2
1
2
1
42
65
Entry Added substance Enzyme loading
Conversion of 3c
63[b]
65
(mol-equiv.)
(w/w)
(%)
1
2
3
4
5
none (0)
2a (1.2)
1a (1.0)
1a (1.0)
1a (1.0)
0.5
0.5
0.5
1
99
81
32
55
71
70
72
49
[a] Scale: 100 mg of 1a, 200 rpm agitation at 30 °C for 60 h. [b] One
mol equivalent of benzoic acid was added.
2
The moderate conversions being almost independent of
During investigation of the inhibition phenomenon, it
enzyme loading was interpreted as either a deactivation of was discovered that when a 1:1 mixture of 1a and 2a were
the enzyme by the co-solvent, or enzyme inhibition caused subjected to fresh buffer and CRL, above 90% overall con-
by some of the reaction components. The hypothesis of a version could be obtained. Although inconvenient, the
time dependant inactivation of the lipase affected by diox- double batch protocol allowed for the isolation of 2a and
ane was tested by adding a second portion of fresh enzyme 2d in 87 and 85% yield from 1a and 1d, respectively. More-
1594
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1592–1597

Regioselective C-6 Hydrolysis of Methyl O-Benzoyl-pyranosides
β--glucopyranoside, methyl α--galactopyranoside and β--galac-
topyranoside were purchased from Sigma–Aldrich. Benzoyl chlo-
ride and pyridine were from Fluka. Column chromatography was
performed using silica gel 60A from Fluka (pore size 40–63 µm).
Candida rugosa lipase was purchased from Sigma–Aldrich (type
VII, Ն 700 units/mg solid).
over, these experiments also confirm that the rate retar-
dation compared to hydrolysis of 1c is related to special
effects caused by 1a and 1d.
Hydrolysis of 1b and 1d were also attempted by this
method, but as expected this did not improve the conver-
sions towards 2b and 2e noteworthy.
Analysis: The HPLC system consisted of an Agilent 1100 series
quaternary pump, Agilent 1100 series variable wavelength UV-de-
tector (200–315 nm) and a thermostatted column compartment.
Conversion was analysed on a Supelcosil C18 column (5 µm par-
ticle size, 25 cm), UV detection (254 nm). All HPLC analyses were
performed using a mixture of deionised water containing TFA
(0.01%) and MeOH (30:70, vol.-%) with increase in flow rate after
elution of the tribenzoate. 1a/2a: Flow rate 1 mL/min. for 12 min,
then 2 mL/min. for 33 min. Retention times 2a: 11.9 min. and 1a:
31.4 min. 1b/2b: Flow rate 1 mL/min. for 9 min, then 2 mL/min. for
16 min. Retention time: 2b: 8.5 min. and 1b: 14.1 min. 1c/2c: Flow
rate 1 mL/min. for 12 min, then 2 mL/min. for 26 min. Retention
time: 2c: 11.4 min, 1c: 24.7 min. 1d/2d: Flow rate 1 mL/min. for
8 min, then 2 mL/min for 15 min. Retention time: 2d: 7.6 min, 1d:
12. 9 min. 1e/2e: Flow rate 1 mL/min. 10 min, then 2 mL/min. for
22 min. Retention time: 2e: 7.6 min, 1e: 20.7 min. 3c/4c: Flow rate
1 mL/min. 15 min. Retention time: 4c: 5.0 min, 3c: 14.4 min.
The rate retardation in hydrolysis of 1a and 1d is not
fully understood. However, all the results taken together
indicate that the main cause is substrate inhibition. The ef-
fect of the inhibition could be reduced by increasing the
reaction volume, indicating that solubility is important. -
Glucose has been shown to be a competitive inhibitor in
CRL-catalysed esterifications forming a dead-end com-
plex.[38] Probably also 1a and 1d could bind to the active
site of the enzyme in a non-productive manner blocking
further catalysis. Also, larger molecules as saponins, fla-
vonoids, alkaloids and antimicrobial agents are known in-
hibitors of CRL without their detailed mechanism of action
been clear.[39,40] A more thorough kinetic investigation of
this inhibition is underway.
Benzoyl migration has been observed in structurally re-
lated compounds when exposed to bases.[32,33] No benzoyl
migration was observed for 2a–e and 4c upon one month
storage in CDCl3 at 25 °C, as detected by NMR and HPLC.
The presented methodology complements the established
enzymatic procedures toward methyl 2,3,4-tri-O-acetyl-pyr-
anosides, by providing building blocks which are less prone
to acyl migration and that allow for tougher conditions in
subsequent chemical steps.
NMR spectra were recorded with Bruker Avance DPX 400 op-
1
erating at 400 MHz for H, and 100 MHz for 13C. Chemical shifts
are in ppm relative to TMS. Coupling constants are in Hertz. All
spectral assignments were confirmed by COSY, HMBC and HSQC
experiments. Melting points were measured with a Mettler FP 5
melting point apparatus and are uncorrected. Optical rotations
were measured using sodium D line at 589 nm with a Perkin–Elmer
243 B polarimeter.
Starting Materials: The compounds 1a–e were prepared as de-
scribed previously.[42] Compound 3c was isolated after benzoylation
of methyl α--galactopyranoside by column chromatography (sil-
ica, toluene/ethyl acetate, 9:1). Recrystallisation from ethanol gave
methyl 2,3,6-tri-O-benzoyl-α--galactopyranoside as a white solid,
m.p. 135.0–135.5 °C, lit.[43] 135.5–137 °C, [α]2D3 = +118.0 (c = 1.00,
CHCl3), lit.[44] +119.8 (c = 1.01, CHCl3). 1H NMR (400 MHz,
CDCl3, 25 °C): δ = 8.08–7.94 (m, 6 H, Ar), 7.61–7.31 (m, 9 H, Ar),
5.75 (dd, 3J = 2.9, 10.6 Hz, 1 H, 3-H), 5.70 (dd, 3J = 10.6, 3.01 Hz,
1 H, 2-H), 5.22 (d, 3J = 3.0 Hz, 1 H, 1-H), 4.66 (dd, 3J = 6.0,
Conclusions
The enzymatic hydrolysis of six methyl O-benzoyl-pyr-
anosides has been investigated using Candida rugosa lipase.
All substrates were hydrolysed in a regioselective manner at
C-6. The highest conversions were observed in hydrolysis
of methyl 2,3,6-tri-O-benzoyl-α--galactopyranoside (3c).
Also methyl 2,3,4,6-tetra-O-benzoyl-α--galactopyranoside
(1c) could be hydrolysed at a reasonable time scale. This
3
2
11.3 Hz, 1 H, 6a-H), 4.58 (dd, J = 6.8, J = 11.3 Hz, 1 H, 6b-H),
3
allowed for the isolation of methyl 2,3-di-O-benzoyl-α-- 4.42 (m, 1 H, 4-H), 4.35 (t, J = 6.0 Hz, 1 H, 5-H), 3.45 (s, 3 H,
OCH3) ppm. 13C NMR spectrum was identical, but assignment
galactopyranoside (4c) and methyl 2,3,4-tri-O-benzoyl-α--
galactopyranoside (2c) in high yield. Hydrolysis of methyl
2,3,4,6-tetra-O-benzoyl-α--glucopyranoside (1a) was com-
plicated by what appeared as substrate inhibition. However,
increasing the reaction volume, or by using a two-batch
protocol, the product could be obtained in up to 87% iso-
lated yield. The same method provided methyl 2,3,4-tri-O-
differed from that reported previously.[44] 13C NMR (100 MHz,
CDCl3, 25 °C): δ = 166.5 (C=O), 166.0 (C=O), 165.8 (C=O), 133.4–
128.4 (18 C, Ar), 97.5 (C-1), δ = 70.8 (C-3), 68.9 (C-2), 68.2 (C-4),
67.7 (C-5), 63.4 (C-6), 55.5 (CH3) ppm.
General Procedure Hydrolysis (25-mg Scale): The hydrolysis was
performed by suspending compound 1a–e or 3c (25 mg,
0.041 mmol) in acetate buffer (2.4 mL, 0.2 , pH: 4.8) and dioxane
benzoyl-β--galactopyranoside (2d) in 85% yield. Methyl
(0.6 mL) and adding Candida rugosa lipase (12.5 or 25 mg). The
2,3,4,6-tetra-O-benzoyl-β--glucopyranoside
(1b)
and
reaction mixture was shaken using an incubator at 140–200 rpm at
the specified temperature and time. The reaction mixture was fil-
tered through Celite, followed by washing of the Celite using ethyl
acetate (15 mL). The combined organic fraction was washed with
aqueous NaHCO3 (20 mL). After drying with Na2SO4, and evapo-
ration of the solvents, the residue was analysed by HPLC to deter-
mine conversion. Reported conversions are the results of averaging
two runs.
methyl 2,3,4,6-tetra-O-benzoyl-α--mannopyranoside (1e)
were poor substrates for CRL. Acyl migration to the free
hydroxy at C-6 was not observed for any of the product
studied.
Experimental Section
General Information: Methyl α--mannopyranoside was prepared
by a known method,[41] while methyl α--glucopyranoside, methyl
General Procedure Hydrolysis (100-mg Scale): The hydrolysis was
performed by suspending compound 1a–e or 3c (100 mg,
Eur. J. Org. Chem. 2009, 1592–1597
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1595

A. Esmurziev, E. Sundby, B. H. Hoff
FULL PAPER
0.163 mmol) in acetate buffer (9.6 mL, 0.2 , pH: 4.8) and dioxane
(2.4 mL) and adding Candida rugosa lipase (50–200 mg). The reac-
tion mixture was shaken using an incubator agitating at 140 or
200 rpm at the specified temperature and time. The reaction mix-
ture was filtered through Celite, followed by washing of the Celite
using ethyl acetate (60 mL). The combined organic fraction was
washed with aqueous NaHCO3 (80 mL). After drying with
Na2SO4, and evaporation of the solvents, the residue was analysed
by HPLC to determine conversion. Selected reactions were purified
by column chromatography (silica, toluene/ethyl acetate, 1:1). For
preparative reactions, yields are reported for single runs.
cedure as for substance 1a to give methyl 2,3,4-tri-O-benzoyl-α--
mannopyranoside (2e) (5 mg, 6%) as a colourless syrup, [α]2D1
=
–126.0 (c = 0.5, CHCl3), lit.[18] –154 (c = 1.1, CHCl3). 1H and
13C NMR spectra for a [D6]DMSO solution were slightly shifted
compared to those reported previously.[47] Assignments were veri-
fied by HSQC and HMBC experiments. 1H NMR (400 MHz, [D6]-
3
DMSO, 25 °C): δ = 8.02–7.33 (m, 15 H, Ar), 5.80 (t, J = 10.0 Hz,
1 H, 4-H), 5.64 (dd, 3J = 10.0, 3.4 Hz, 1 H, 3-H), 5.58 (dd, 3J =
3
3.4, 1.7 Hz, 1 H, 2-H), 5.04 (d, J = 1.7 Hz, 1 H, 1-H), 4.06 (m, 1
H, 5-H), 3.63 (m, 2 H, 6a-H, 6b-H), 3.48 (s, 3 H, OCH3) ppm.
13C NMR (100 MHz, [D6]DMSO, 25 °C): δ = 164.7 (C=O), 164.6
(C=O), 164.6 (C=O), 133.9–128.6 (18 C, Ar), 97.6 (C-1), 70.8 (C-
5), 70.6 (C-3), 69.8 (C-2), 66.4 (C-4), 60.0 (C-6), 54.7 (OCH3) ppm.
Products
Methyl 2,3,4-Tri-O-benzoyl-α-D
-glucopyranoside (2a):[12,45] Methyl
Methyl 2,3-Di-O-benzoyl-α-D
-galactopyranoside (4c):[43,48] Methyl
2,3,4,6-tetra-O-benzoyl-α--glucopyranoside (1a) (100 mg, 0.163
mmol) was suspended in 20% dioxane/acetate buffer at pH 4.8
(12 mL) using lipase from Candida rugosa (100 mg) at 30 °C agitat-
ing at 200 rpm for 40 h. Extraction, drying and concentration
yielded a 58:42 mixture of 2a and 1a, which was treated once more
under identical conditions. This gave a 92:8 mixture of 2a and 1a.
Flash chromatography yielded 70 mg (87%) of 2a as a white
amorphous solid, m.p. 133.5–134.5 °C, [α]2D1 = +51.3 (c = 1.50,
CHCl3), lit.[19] +53.5 (c = 1.5, CHCl3). 1H NMR corresponded
with data published by Verduyn et al.[12] 13C NMR (CDCl3) of the
sugar backbone corresponded with data reported by Kovac et
al.,[45] additional shifts for benzoyl C atoms, δ = 166.4 (C=O), 165.8
(C=O, 2 C), 133.5–128.2 (18 C, Ar) ppm.
2,3,6-tri-O-benzoyl-α--galactopyranoside (3c) (100 mg, 0.197
mmol) was hydrolysed by CRL at 30 °C agitating at 200 rpm for
20 h, to yield a 1:99 mixture of 3c and 4c. Purification by flash
chromatography gave 77 mg (96%) of 4c as a white solid, m.p.
95.0–96.0 °C, lit.[48] amorphous solid, [α]2D1 = +198 (c = 1.3,
1
CHCl3), lit.[48] +200 (c = 1.2, CHCl3). H NMR (400 MHz, [D5]-
3
pyridine, 25 °C): δ = 8.40–7.16 (m, 10 H, Ar) 6.44 (dd, J = 10.7,
3
3.6 Hz, 1 H, 2-H), 6.21 (dd, J = 10.7, 3.2 Hz, 1 H, 3-H), 5.50 (d,
3
3J = 3.6 Hz, 1 H, 1-H), 5.00 (d, J = 3.2 Hz, 1 H, 4-H), 4.45 (m, 1
H, 5-H), 4.40 (m, 2 H, 6a-H, 6b-H), 3.39 (s, 3 H, OCH3) ppm. 13C
NMR data for a [D5]pyridine solution were slightly shifted com-
pared to those reported previously (25 MHz).[49] 13C NMR
(100 MHz, [D5]pyridine, 25 °C): δ = 166.6 (C=O), 166.5 (C=O),
133.6–128.7 (12 C, Ar), 98.3 (C-1), 73.0 (C-3), 72.6 (C-5), 70.5 (C-
2), 68.2 (C-4), 61.9 (C-6), 55.1 (CH3) ppm.
Methyl 2,3,4-Tri-O-benzoyl-β-D
-glucopyranoside (2b):[18] Methyl
2,3,4,6-tetra-O-benzoyl-β--glucopyranoside (1b) (100 mg, 0.163
mmol) was hydrolysed and purified using the exactly the same pro-
cedure as for substance 1a to give methyl 2,3,4-tri-O-benzoyl-β--
glucopyranoside (2b) (6 mg, 8%) as a white solid, m.p. 82–83 °C.
Optical rotation differed in sign of that reported previously; [α]2D1
= +6.8 (c = 1.1, CHCl3), lit.[18] –6.6 (c = 1.1, CHCl3). 1H NMR
(CDCl3) and 13C NMR (CDCl3) of the sugar backbond corre-
sponded with data reported by Ziegler et al.,[18] additional shifts
for benzoyl C atoms, δ = 166.2 (C=O), 165.9 (C=O), 166.3 (C=O),
133.0–128.4 (18 C, Ar) ppm.
Acknowledgments
Sør-Trøndelag University College is gratefully acknowledged for
financial support. NTNU is acknowledged for providing NMR
facilities.
[1] B. Yu, X. Zhu, Y. Hui, Org. Lett. 2000, 2, 2539–2541.
[2] T. Lintunen, J. T. Yli-Kauhaluoma, Bioorg. Med. Chem. Lett.
2000, 10, 1749–1750.
[3] W. T. Haskins, R. M. Hann, C. S. Hudson, J. Am. Chem. Soc.
1946, 68, 628–632.
[14,18]
Methyl 2,3,4-Tri-O-benzoyl-α-D-galactopyranoside (2c):
Methyl 2,3,4,6-tetra-O-benzoyl-α--galactopyranoside (1c) (100
mg, 0.163 mmol) was hydrolysed in 20% dioxane/acetate buffer
(12 mL) using lipase from Candida rugosa (100 mg) at 40 °C agitat-
ing at 200 rpm for 40 h, yielding a 96:4 mixture of 2c and 1c. Purifi-
cation by flash chromatography yielded 70 mg (86%) of 2c as a
[4] T. Ziegler, E. Eckhardt, G. Pantkowski, J. Carbohydr. Chem.
1994, 13, 81–109.
[5] Y. Sone, A. Misaki, S. Shibata, Carbohydr. Res. 1989, 191, 79–
white solid, m.p. 57.0–57.5 °C, lit.[14,46] amorphous solid, [α]2D1
=
89.
1
+246.6 (c = 1.2, CHCl3), lit.[18] +248.4 (c = 1.2, CHCl3). H and
[6] F. Dasgupta, P. J. Garegg, Carbohydr. Res. 1990, 202, 225–238.
[7] S. J. Danishefsky, M. P. DeNinno, S. H. Chen, J. Am. Chem.
Soc. 1988, 110, 3929–3940.
[8] C. De Meo, M. Farris, N. Ginder, B. Gulley, Eur. J. Org. Chem.
2008, 3673–3677.
[9] Y. Morii, H. Matsuda, K. Ohara, M. Hashimoto, K. Miyairi,
T. Okuno, Bioorg. Med. Chem. 2005, 13, 5113–5144.
[10] T. Ziegler, H. Sutoris, C. P. J. Glaudemans, Carbohydr. Res.
1992, 229, 271–291.
[11] T. Hashihayata, K. Ikegai, K. Takeuchi, H. Jona, T. Mukai-
yama, Bull. Chem. Soc. Jpn. 2003, 76, 1829–1848.
[12] R. Verduyn, M. Douwes, P. A. M. Van der Klein, E. M. Moes-
inger, G. A. Van der Marel, J. H. Van Boom, Tetrahedron 1993,
49, 7301–7316.
[13] M. C. Yan, Y. N. Chen, H. T. Wu, C. C. Lin, C. T. Chen, C. C.
Lin, J. Org. Chem. 2007, 72, 299–302.
[14] N. S. Utkina, A. V. Nikolaev, V. N. Shibaev, Bioorg. Khim.
1991, 17, 531–539.
[15] G. D. K. Kumar, S. Baskaran, J. Org. Chem. 2005, 70, 4520–
4523.
13C NMR data (CDCl3) corresponded with those reported.[14,18]
Methyl 2,3,4-Tri-O-benzoyl-β-D
-galactopyranoside (2d):[47] Methyl
2,3,4,6-tetra-O-benzoyl-β--galactopyranoside (1d) (100 mg, 0.163
mmol) was hydrolysed in 20% dioxane/acetate buffer (12 mL) using
lipase from Candida rugosa (100 mg) at 40 °C for 40 h agitating
at 200 rpm. Extraction, drying and concentration yielded a 55:45
mixture of 2d and 1d, which was subjected to enzymatic hydrolysis
once more under identical conditions. This gave a 88:12 mixture of
2d and 1d, which upon flash chromatography yielded 69 mg (85%)
of 2d as a white amorphous solid, m.p. 89.0–90.0 °C, lit.[18] 90–
91 °C, [α]2D1 = +124.0 (c = 1.0, CHCl3). 1H and 13C NMR data
corresponded with those reported for solutions in [D6]DMSO[47]
and CDCl3.[16]
Methyl 2,3,4-Tri-O-benzoyl-α-D
-mannopyranoside (2e):[15,47] Methyl
2,3,4,6-tetra-O-benzoyl-α--mannopyranoside (1e) (100 mg, 0.163
mmol) was hydrolysed and purified using exactly the same pro-
1596
www.eurjoc.org
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2009, 1592–1597

Regioselective C-6 Hydrolysis of Methyl O-Benzoyl-pyranosides
[16]
[33] G. J. F. Chittenden, J. G. Buchanan, Carbohydr. Res. 1969, 11,
379–385.
E. M. Nashed, C. P. J. Glaudemans, J. Org. Chem. 1987, 52,
5255–5260.
[34] T. W. Greene, P. G. M. Wuts, Protective Groups in Organic Syn-
thesis, 2nd ed., John Wiley & Sons, New York, 1991.
[35] K. Yoshida, K. Kotsubo, H. Shigematsu, J. Chromatogr. 1981,
208, 104–108.
[17]
[18]
V. Bou, J. Vilarrasa, Tetrahedron Lett. 1990, 31, 567–568.
T. Ziegler, P. Kovac, C. P. J. Glaudemans, Carbohydr. Res. 1989,
194, 185–198.
[19]
[20]
L. X. Gan, R. L. Whistler, Carbohydr. Res. 1990, 206, 65–69.
[36] H. W. Liu, K. Nakanishi, J. Am. Chem. Soc. 1982, 104, 1178–
M. Terreni, R. Salvetti, L. Linati, R. Fernandez-Lafuente, G.
Fernandez-Lorente, A. Bastida, J. M. Guisan, Carbohydr. Res.
2002, 337, 1615–1621.
1185.
[37] J. M. Williams, A. C. Richardson, Tetrahedron 1967, 23, 1369–
1378.
[38] B. R. Somashekar, K. Lohith, B. Manohar, S. Divakar, J. Bio-
sci. Bioeng. 2007, 103, 122–128.
[39] C. Ruiz, S. Falcocchio, E. Xoxi, L. Villo, G. Nicolosi, F. I. J.
Pastor, P. Diaz, L. Saso, J. Mol. Catal. B: Enzym. 2006, 40,
138–143.
[40] E. Grippa, R. Valla, L. Battinelli, G. Mazzanti, L. Saso, B.
Silvestrini, Biosci. Biotechnol. Biochem. 1999, 63, 1557–1562.
[41] A. Nandi, P. Chattopadhyay, Tetrahedron Lett. 2002, 43, 5977–
5980.
[42] A. Esmurziev, N. Simic, E. Sundby, B. H. Hoff, Magnetic Res.
Chem. 2009, manuscript accepted for publication.
[43] E. J. Reist, R. R. Spencer, D. F. Calkins, B. R. Baker, L. Good-
man, J. Org. Chem. 1965, 30, 2312–2317.
[44] G. Vic, J. J. Hastings, H. G. Crout, Tetrahedron: Asymmetry
1996, 7, 1973–1984.
[45] P. Kovac, V. Sklenar, C. P. J. Glaudemans, Carbohydr. Res.
1988, 175, 201–213.
[46] A. Rashid, W. Mackie, I. J. Colquhoun, D. Lamba, Can. J.
Chem. 1990, 68, 1122–1127.
[47] K. Agoston, L. Kroeger, G. Dekany, J. Thiem, J. Carbohydr.
Chem. 2007, 26, 513–525.
[48] E. Seymour, J. M. J. Frechet, Tetrahedron Lett. 1976, 1149–
1152.
[21]
[22]
[23]
T. Horrobin, C. H. Tran, D. Crout, J. Chem. Soc. Perkin Trans.
1 1998, 1069–1080.
W. J. Hennen, H. M. Sweers, Y. F. Wang, C. H. Wong, J. Org.
Chem. 1988, 53, 4939–4945.
G. Fernandez-Lorente, J. M. Palomo, J. Cocca, C. Mateo, P.
Moro, M. Terreni, R. Fernandez-Lafuente, J. M. Guisan, Tet-
rahedron 2003, 59, 5705–5711.
[24] T. Rodriguez-Perez, I. Lavandera, S. Fernandez, Y. S. Sanghvi,
M. Ferrero, V. Gotor, Eur. J. Org. Chem. 2007, 2769–2778.
[25] K. Watanabe, S. I. Ueji, J. Chem. Soc. Perkin Trans. 1 2001,
1386–1390.
[26] C. Torres, C. Otero, J. Mol. Catal. A: Chem. 1995, 97, 119–
134.
[27] F. Yang, T. W. Weber, J. L. Gainer, G. Carta, Biotechnol. Bio-
eng. 1997, 56, 671–680.
[28] K. F. Hsiao, F. L. Yang, S. H. Wu, K. T. Wang, Biotechnol.
Lett. 1995, 17, 963–968.
[29] Y. Kodera, K. Sakurai, Y. Satoh, T. Uemura, Y. Kaneda, H.
Nishimura, M. Hiroto, A. Matsushima, Y. Inada, Biotechnol.
Lett. 1998, 20, 177–180.
[30] K. Moriyoshi, H. Yamanaka, T. Ohmoto, T. Ohe, K. Sakai,
Biosci. Biotechnol. Biochem. 2005, 69, 1292–1299.
[31] S. Tomic, D. Ljevakovic, J. Tomasic, Tetrahedron 1993, 49,
10977–10986.
[49] Y. Tsuda, M. E. Haque, K. Yoshimoto, Chem. Pharm. Bull.
1983, 31, 1612–1624.
[32] A. E. Salinas, Carbohydr. Res. 1999, 316, 34–46.
Received: December 19, 2008
Published Online: February 18, 2009
Eur. J. Org. Chem. 2009, 1592–1597
© 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1597