Mode of action on deacetylation of acetylated methyl glycoside by cellulose acetate esterase from Neisseria sicca SB

Article abstract of DOI:10.1271/bbb.69.1292

The regioselective deacetylation of purified cellulose acetate esterase from Neisseria sicca SB was investigated on methyl 2,3,4,6-tetra-O-acetyl- β-D-glucopyranoside and 2,3,4,6-tetra-O-acetyl-β-D-galactopyranoside. The substrates were used as model comp

Full text of DOI:10.1271/bbb.69.1292

Biosci. Biotechnol. Biochem., 69 (7), 1292–1299, 2005
Mode of Action on Deacetylation of Acetylated Methyl Glycoside
by Cellulose Acetate Esterase from Neisseria sicca SB
Kunihiko MORIYOSHI,y Hayato YAMANAKA, Takashi OHMOTO, Tatsuhiko OHE, and Kiyofumi SAKAI
Department of Environmental Technology, Osaka Municipal Technical Research Institute,
1-6-50 Morinomiya, Joto-ku, Osaka 536-8553, Japan
Received January 26, 2005; Accepted April 11, 2005
The regioselective deacetylation of purified cellulose
acetate esterase from Neisseria sicca SB was investigated
on methyl 2,3,4,6-tetra-O-acetyl-ꢀ-D-glucopyranoside
and 2,3,4,6-tetra-O-acetyl-ꢀ-D-galactopyranoside. The
substrates were used as model compounds of cellulose
acetate in order to estimate the mechanism for deace-
tylation of cellulose acetate by the enzyme. The enzyme
rapidly deacetylated at position C-3 of methyl 2,3,4,6-
tetra-O-acetyl-ꢀ-D-glucopyranoside to accumulate 2,4,6-
triacetate as the main initial reaction product in about
70% yield. Deacetylation was followed at position C-2,
and generated 4,6-diacetate in 50% yield. The enzyme
deacetylated the product at positions C-4 and C-6 at
slower rates, and generated 4- and 6-monoacetates
at a later reaction stage. Finally, it gave a completely
deacetylated product. For 2,3,4,6-tetra-O-acetyl-ꢀ-D-
galactopyranoside, CA esterase deacetylated at positions
C-3 and C-6 to give 2,4,6- and 2,3,4-triacetate. Deace-
tylation proceeded sequentially at positions C-3 and C-6
to accumulate 2,4-diacetate in 55% yield. The enzyme
exhibited regioselectivity for the deacetylation of the
acetylglycoside.
polymers calls for a set of enzymes with different
activities acting in concert, because the polysaccharides
in the plant cell-wall contain many side-chain substitu-
ents. In xylan degradation, enzymes act synergistically
on the 1,4-ꢀ-D-xylan backbone (endo-1,4-ꢀ-xylanases)
and side-chain-cleaving enzymes (ꢁ-L-arabinofuranosi-
dase, acetylxylan esterase, and ꢁ-glucuronidase).5–8) The
synergistic action between endo-1,4-ꢀ-xylanases and
acetylxylan esterase (EC 3.1.1.72) results in efficient
degradation of acetylated xylan. The deacetylation
reaction by acetylxylan esterase increases the accessi-
bility of endo-1,4-ꢀ-xylanase to the polysaccharide
backbone.9)
A bacterium capable of assimilating CA, Neisseria
sicca SB, was isolated from soil.10) We purified and
characterized a CA esterase from N. sicca SB that
catalyzes the deacetylation of CA,11) and endo-1,4-ꢀ-
glucanases I (EG I)12) and II (EG II),13) which hydrolyze
the ꢀ-1,4 linkages in CA molecules. We have shown
synergistic degradation of CA particles by CA esterase
with EG I,12) and by CA esterase with EG II.13) CA
esterase facilitates hydrolysis of the CA backbone by the
glucanases. CA esterase is important for complete
degradation of CA particles. The degradation of CA
by a mixture of CA esterase and endo-1,4-ꢀ-glucanases
from N. sicca SB resembled heteroxylan degradation
in its synergistic action. The N-terminal amino acid
sequence of CA esterase was found to have 45%
similarity with that of an acetylxylan esterase from
Aspergillus niger.11) CA esterase catalyzed the hydrol-
ysis of p-nitrophenyl esters of short-chain fatty acids.11)
The enzyme showed only slight activity toward aliphatic
and aromatic acetyl esters. It showed significant activity
on acetyl saccharides, though the compounds are low
solubility. To know the mode of action of CA esterase in
detail is important for elucidating the mechanism of CA
degradation by the enzymes. This paper describes
regioselective deacetylation by CA esterase for methyl
2,3,4,6-tetra-O-acetyl-ꢀ-D-glucopyranoside (2,3,4,6-Ac-
Key words: cellulose acetate; esterase; regioselective
deacetylation; Neisseria sicca
Cellulose acetate (CA) is one of the most important
synthetic organic esters because of its broad applica-
tions, such as in fibers, plastics, films, and membranes,1)
and because it is made from cellulose, the most abundant
biopolymer on earth. CA is biodegradable by certain
unidentified bacteria,2) Pseudomonas paucimobilis,3) and
Bacillus sp. strain S2055,4) but the mechanism of
biodegradation of CA on the enzymatic level has not
been elucidated. Research on the mechanisms of
enzymatic degradation of biodegradable polymers,
including CA, is needed to understand the assimilability
and safety of degradation products in the environment.
In nature, the complete degradation of plant cell-wall
y
To whom correspondence should be addressed. Tel: +81-6-6963-8065; Fax: +81-6-6963-8079; E-mail: moriyosi@omtri.city.osaka.jp
Abbreviations: CA, cellulose acetate; EG I, endo-1,4-ꢀ-glucanase I; EG II, endo-1,4-ꢀ-glucanase II; 2,3,4,6-Ac-MeꢀGlc, methyl 2,3,4,6-tetra-O-
acetyl-ꢀ-D-glucopyranoside; 2,3,4,6-Ac-MeꢀGal, methyl 2,3,4,6-tetra-O-acetyl-ꢀ-D-galactopyranoside; MeꢀGlc, methyl-ꢀ-D-glucopyranoside;
MeꢀGal, methyl-ꢀ-D-galactopyranoside; TLC, thin-layer chromatography; TMS, trimethylsilyl; GLC, gas-liquid chromatography; GLC–MS, gas-
liquid chromatography mass spectrometer; COSY, correlation spectroscopy

Deacetylation by Cellulose Acetate Esterase
1293
MeꢀGlc) and methyl 2,3,4,6-tetra-O-acetyl-ꢀ-D-galacto-
pyranoside (2,3,4,6-Ac-MeꢀGal) as model compounds
for CA. Especially, 2,3,4,6-Ac-MeꢀGlc resembles tri-
acetyl 1,4-linked glucopyranosyl units in the CA
molecule, and this provided insight concerning its
function in CA degradation.
tion mixtures were subjected to trimethylsilylation using
Silblender-HTP (Nacalai Tesque). After incubation at
25 ꢀC for 1 h, samples were analyzed by gas-liquid
chromatography (GLC) with a Shimadzu GC-1700
equipped with a flame ionization detector, and a Supelco
SPB-170l fused-silica capillary column (30 m ꢁ
0:32 mm, 0.25 mm film). The peak areas were integrated
with a Shimadzu C-R8A Chromatopack integrator.
Helium was used as a carrier gas at a flow rate 1.4 ml/
min. The derivatives of MeꢀGlc and MeꢀGal were
chromatographed at an initial temperature of 120 ꢀC for
6 min, which was then increased at a rate of 20 ꢀC/min
to 180 ꢀC, then increased at 2 ꢀC/min to 240 ꢀC. The
injector and detector temperatures were both 260 ꢀC.
Concentrations of the reaction components were deter-
mined by electronic integration. Differences in detector
responses to derivatives having different contents of
acetyl and TMS groups were taken into consideration.
Acetylated and trimethylsilylated derivatives were ana-
lyzed with a gas-liquid chromatography mass spectrom-
eter (GLC–MS) for estimation of number of acetyl
groups. The analysis was done with a Varian Saturn
2000 mass spectrometer at an ionization voltage of
70 eV. The mass spectra were obtained by electron-
impact ionization at 70 eV, using a Supelco MDN-5S
capillary column (30 m ꢁ 0:25 mm, 0.25 mm film).
Materials and Methods
Materials. Lipase from Candida rugosa L-8525
(EC 3.1.1.3) was purchased from Sigma Chemical (St.
Louis, MO), and was dialyzed before use. CA esterase
was produced by N. sicca SB in CA medium, and was
purified as described previously.11) CA esterase activity
was assayed in a reaction mixture containing 50 mM
Tris–HCl buffer, pH 8.0, 1.0% CA (degree of substitu-
tion, 0.88) as the substrate, at 30 ꢀC, and the acetic acid
released was measured as described previously.11) One
unit of the esterase was defined as the amount of enzyme
releasing 1.0 mmol of acetic acid per min.
2,3,4,6-Ac-MeꢀGlc and 2,3,4,6-Ac-MeꢀGal were
obtained by acetylation with acetic anhydride in pyr-
idine from commercially available methyl-ꢀ-D-gluco-
pyranoside (MeꢀGlc, Nacalai Tesque, Kyoto, Japan) and
methyl-ꢀ-D-galactopyranoside (MeꢀGal, Nacalai Tes-
que) respectively. A solution of 2 g of MeꢀGlc or
MeꢀGal in 12 ml of pyridine was treated at 0 ꢀC with
7 ml of acetic anhydride. After 16 h, the products were
crystallized from water and recrystallized from ethanol.
The structures of the products were confirmed by NMR
spectroscopy by way of comparison of the obtained and
reported NMR spectra.14) Methyl 2,3,4-tri-O-acetyl-ꢀ-D-
glucopyranoside (2,3,4-Ac-MeꢀGlc) and methyl 2,3,4-
tri-O-acetyl-ꢀ-D-galactopyranoside (2,3,4-Ac-MeꢀGal)
were enzymatically prepared from 2,3,4,6-Ac-MeꢀGlc
and 2,3,4,6-Ac-MeꢀGal via C-6 selective hydrolysis
catalyzed by Candida rugosa lipase respectively.15)
TLC and NMR spectroscopy. Enzymatic deacetylation
products of 2,3,4,6-Ac-MeꢀGlc and 2,3,4,6-Ac-MeꢀGal
were analyzed on TLC plates of silica gel 60 (Merck,
Darmstadt, Germany). The solvent system was ethyl
acetate/benzene/2-propanol (2:1:0.1, v/v). The sepa-
rated saccharides were detected by spraying H2SO4–
methanol (1:1, w/w), and by heating the plate at 150 ꢀC
for a few minutes. Large-scale isolation of partially
deacetylated compounds for NMR spectroscopy was
accomplished by TLC. The compounds were eluted
from the silica gel with ethyl acetate (tri- and di-
acetates) or methanol (mono-acetates).
Deacetylation of 2,3,4,6-Ac-MeꢀGlc and 2,3,4,6-Ac-
MeꢀGal by CA esterase. A saturated solution of 2,3,4,6-
Ac-MeꢀGlc or 2,3,4,6-Ac-MeꢀGal in 50 mM potassium
phosphate buffer, pH 7.0, at 20 ꢀC was filtered and
mixed with a solution of CA esterase to give a final CA
esterase concentration of 0.15 U/ml in a total volume of
30 ml. The reaction conditions were neutral pH and
relatively low temperature, because the products were
not very stable. Control mixtures without the enzyme
were run in parallel. Initial substrate concentrations were
6.4 mM for 2,3,4,6-Ac-MeꢀGlc and 6.4 mM for 2,3,4,6-
Ac-MeꢀGal, measured by the phenol-sulfuric acid
method.16) After incubation with CA esterase, aliquots
of the mixture were taken periodically and subjected to
thin-layer chromatography (TLC), or were frozen
immediately in liquid nitrogen and lyophilized.
1H- and 13C-NMR spectra of partially acetylated
MeꢀGlc and MeꢀGal derivatives were recorded with a
JEOL JNM-AL300 spectrometer at 25 ꢀC in CDC13 or
1
CD3OD, operated at 300 MHz for H and 75 MHz for
13C. Chemical shifts were given as the ꢂ-value (ppm)
against tetramethylsilane as an internal reference. The
1
sites of the acetyl group were confirmed by 2D H–1H
and 1H–13C correlation spectroscopy (COSY) NMR
spectroscopy.
Results
Enzymatic deacetylation of 2,3,4,6-Ac-MeꢀGlc and
2,3,4,6-Ac-MeꢀGal
2,3,4,6-Ac-MeꢀGlc, 2,3,4-Ac-MeꢀGlc, and MeꢀGlc
were available as standard compounds for GLC. 2,3,4,6-
Ac-MeꢀGlc was deacetylated by CA esterase, and the
products were analyzed by GLC (Fig. 1). CA esterase
hydrolyzed 2,3,4,6-Ac-MeꢀGlc rapidly and generated
GLC and GLC mass spectrometry of trimethylsilyl
(TMS) derivatives of partially acetylated methyl ꢀ-
glycosides. Standard compounds and freeze-dried reac-

1294
K. MORIYOSHI et al.
Table 1. Estimation of Number of Acetyl Groups for Various
Acetylated MeꢀGlc by GLC–MS
Fragment ion Molecular
peak (m=z)
Number of
acetyl group
TMS derivatives
weight
331
(Mþ-CH3O)
377
2,3,4,6-Ac-MeꢀGlcꢂ
2,3,4-Ac-MeꢀGlcꢂ
Product A
362
4
3
3
2
1
1
0
392
(Mþ-CH3)
377
(Mþ-CH3)
407
(Mþ-CH3)
437
(Mþ-CH3)
437
(Mþ-CH3)
467
(Mþ-CH3)
392
Product B
422
Product C
452
Fig. 1. Time Course of 2,3,4,6-Ac-MeꢀGlc Deacetylation by CA
Esterase from N. sicca SB.
Product D
452
MeꢀGlcꢂ
482
2,3,4,6-Ac-MeꢀGlc was incubated with CA esterase. The prod-
ucts were trimethylsilylated and analyzed by gas-liquid chromatog-
ꢂStandard compounds
raphy. ( ) 2,3,4,6-Ac-MeꢀGlc; ( ) product A; ( ) product B; (
product C; ( ) product D; ( ) MeꢀGlc.
)
At a later reaction stage, product H and MeꢀGal were
generated at a much slower rate.
Identification of deacetylated products from 2,3,4,6-
Ac-MeꢀGlc and 2,3,4,6-Ac-MeꢀGal by GLC–MS
The various molecular weights of the enzymatic
products A, B, C, and D were determined by the spectra
of GLC–MS, and the number of acetyl groups of the
products was estimated (Table 1). The enzymatic prod-
ucts were trimethylsilylated and analyzed by GLC–MS.
If the substrate is deacetylated by the enzyme, hydroxyl
groups will arise at the deacetylated positions in the
products. When the hydroxyl groups of the products are
trimethylsilylated, the total molecular weights of the
products increase. The mass spectra of trimethylsilyl
ethers of MeꢀGlc by electron ionization mass spectrom-
etry at 70 eV have been reported.17,18) A more intense
Fig. 2. Time Course of 2,3,4,6-Ac-MeꢀGal Deacetylation by CA
Esterase from N. sicca SB.
2,3,4,6-Ac-MeꢀGal was incubated with CA esterase. The prod-
ucts were trimethylsilylated and analyzed by gas-liquid chromatog-
.
peak is present for the loss of CH3 from one of the
raphy. ( ) 2,3,4,6-Ac-MeꢀGal; ( ) product E; ( ) product F; (
product G; ( ) product H; ( ) MeꢀGal.
)
trimethylsilyl groups. The molecular weight can there-
fore be determined from this if no molecular-ion peak is
present. That was confirmed by GLC–MS spectra of
trimethylsilylated MeꢀGlc and 2,3,4-Ac-MeꢀGlc as
standard compounds. Trimethylsilylated MeꢀGlc and
2,3,4-Ac-MeꢀGlc gave the Mþ-CH3 ion at m=z 467 and
at m=z 377 respectively. 2,3,4,6-Ac-MeꢀGlc gave the
Mþ-CH3O ion at m=z 331. Product A gave the Mþ-CH3
ion at m=z 377. The fragment ion suggested that the
number of the acetyl group for product A was three.
Product B gave the Mþ-CH3 ion at m=z 407, and
products C and D gave the Mþ-CH3 ion at m=z 437. The
fragment ions suggested that the numbers of the acetyl
group were two for product B and one for products C
and D. The reaction products E, F, G, and H were
estimated from the number of acetyl groups in the same
way. Products E, F, G, and H gave the Mþ-CH3 ion at
m=z 377, 377, 407, and 437 respectively. These frag-
ment ion peaks suggest that the numbers of the acetyl
group were three for products E and F, two for G, and
product A as the major initial product in about 70%
yield at 20 min. The initial decreasing rate of the
substrate was 0.56 mM/min when the initial substrate
and enzyme concentrations were 6.4 mM and 0.15 U/ml
respectively. The apparent increasing rate of product A
was almost the same. As the reaction progressed,
product B was accumulated as an intermediate product
in about 50% yield at 3 h. A mixture of products C and D
was generated at a later reaction stage. The formation of
MeꢀGlc proceeded at a much slower rate. In the case of
deacetylation for 2,3,4,6-Ac-MeꢀGal, CA esterase rap-
idly generated products E and F in about 25% and 35%
yield respectively at 1.5 h (Fig. 2). The initial decreasing
rate of the substrate was 0.22 mM/min. The apparent
increasing rates of products E and F were 0.077 and
0.14 mM/min respectively. As the reaction progressed,
product G was accumulated in about 55% yield at 6 h.

Deacetylation by Cellulose Acetate Esterase
1295
Table 2. 1H-NMR Data for Deacetylated Products of 2,3,4,6-Ac-
MeꢀGlc by CA Esterase from N. sicca SB
Table 3. 1H-NMR Data for Deacetylated Products of 2,3,4,6-Ac-
MeꢀGal by CA Esterase from N. sicca SB
Chemical shifts (ppm)
Sample
Chemical shifts (ppm)
Sample
H1
2,3,4,6-Ac-MeꢀGlc
4.43 4.99 5.21 5.09 3.70 4.28 4.15
Product A (2,4,6-Ac)
4.37 4.84 3.72 4.95 3.63 4.29 4.17
Product B (4,6-Ac)
4.23 3.46 3.68 4.93 3.64 4.29 4.14
Product C (4-Ac)
4.20 3.40 3.60 4.88 3.52 3.92 3.71
Product D (6-Ac)
H2
H3
H4
H5
H6
H60 OCH3
H1
2,3,4,6-Ac-MeꢀGal
4.40 5.20 5.02 5.39 3.91 4.18 4.16
Product E (2,3,4-Ac)
4.43 5.22 5.05 5.38 3.77 3.77 3.58
Product F (2,4.6-Ac)
4.36 4.96 3.84 5.33 3.85 4.18 4.16
Product G (2,4-Ac)
H2
H3
H4
H5
H6
H60 OCH3
3.51
3.51
3.57
3.56
3.55
3.53
3.53
3.53
3.52
3.51
3.52
4.38 4.98 3.83 5.25 3.66 3.76 3.76
MeꢀGal
4.21 3.39 3.55 3.55 3.45 4.45 4.32
4.12 3.49 3.47 3.83 3.50 3.75 3.73
MeꢀGlc
4.17 3.16 3.35 3.28 3.26 3.87 3.67
In order to identify the position of deacetylation, the
enzymatic products were measured by 13C-NMR also
(data not shown). In the chemical shifts of 13C-NMR
spectra, the corresponding positions of glucoside deriv-
atives were confirmed by 1H–13C COSY. For acetylated
saccharides, deacetylation of acetylated hydroxyl groups
at positions C-2, C-3, C-4, and C-6 results in a downfield
shift (1–3 ppm) of the peak(s) corresponding to the
neighboring carbon(s).20) When the chemical shifts of
product A were compared with those of 2,3,4,6-Ac-
MeꢀGlc, chemical shifts at positions C-2 and C-4 for
product A were downfield shifts of about 3 ppm each,
and the values at the others positions were almost equal.
This result suggests that product A was a deacetylated
compound of 2,3,4,6-Ac-MeꢀGlc at position C-3,
namely 2,4,6-Ac-MeꢀGlc. When product B was com-
pared with product A, chemical shifts at positions C-1
and C-3 were downfield shifts. This suggests that
product B was 4,6-Ac-MeꢀGlc. When product C was
compared with product B, the chemical shift at position
C-5 was a downfield shift, suggesting that product C
was 4-Ac-MeꢀGlc. When product D was compared with
product B, chemical shifts at positions C-3 and C-5 were
downfield shifts, suggesting that product D was 6-Ac-
MeꢀGlc. These results were confirmed by comparing
products C and D with MeꢀGlc. These results from 13C-
one for H. The numbers of acetyl groups of the
enzymatic products were also confirmed by comparing
relative mobility on TLC (data not shown).
Structure determinations of partially deacetylated
products by NMR
The deacetylated positions in the enzymatic products
1
1
were established by H- and 13C-NMR. The H-NMR
spectra of products A, B, C, and D were measured after
purification by TLC (Table 2). In the chemical shifts of
1
the H-NMR spectra, the corresponding positions of the
1
products were confirmed by H–1H COSY. Deacetyla-
tion at positions C-2, C-3, C-4, and C-6 of acetylated
methyl glucosides results in an upfield shift (1–3 ppm) of
the peak corresponding to the proton of O-acetylated
carbon.19) The shift values are virtually independent of
the nature of the acyl moiety and the solvent. The NMR
spectrum of the triacetate was compared to that of
authentic tetraacetate, and the spectrum of the diacetate
was compared to that of the previously identified
triacetate. The spectra of the monoacetates were com-
pared to that of the previously identified diacetate and
that of unmodified, authentic methyl glucoside. When
the chemical shifts of product A were compared with
that of 2,3,4,6-Ac-MeꢀGlc, the chemical shift at position
C-3 for product A was an upfield shift of about 1.5 ppm,
and the values at the other positions were almost equal.
This result suggests that product A was a deacetylated
compound of 2,3,4,6-Ac-MeꢀGlc at position C-3,
namely 2,4,6-Ac-MeꢀGlc. When product B was com-
pared with product A, the chemical shift at position C-2
was an upfield shift of about 1.4 ppm. This result
suggests that product B was 4,6-Ac-MeꢀGlc. When
product C was compared with product B, the chemical
shift at position C-6 was an upfield shift, suggesting that
product C was 4-Ac-MeꢀGlc. When product D was
compared with product B, the chemical shift at position
C-4 was an upfield shift, suggesting that product D was
6-Ac-MeꢀGlc. These results were confirmed by com-
parison products C and D with MeꢀGlc.
1
NMR corresponded with the results from H-NMR.
In order to identify the position of deacetylation, H-
1
NMR spectra of products E, F, and G were measured
(Table 3). The results suggest that products E, F, and G
were 2,3,4-Ac-MeꢀGal, 2,4,6-Ac-MeꢀGal, and 2,4-Ac-
MeꢀGal respectively. These results were confirmed by
the 13C-NMR spectra (data not shown). We could not
analyze product H due to an insufficient amount of the
pure product.
Discussion
CA esterase from N. sicca SB had been reported on as
to purification and characterization.11) The enzyme
showed significant activity on acetyl saccharides. Inves-
tigation of the mode of action of CA esterase is

1296
K. MORIYOSHI et al.
Fig. 3. Main Deacetylation Reactions Catalyzed by CA Esterase from N. sicca SB for 2,3,4,6-Ac-MeꢀGlc and for 2,3,4,6-Ac-MeꢀGal.
necessary in order to understand the deacetylation
mechanism of CA degradation by CA esterase in detail.
But it is difficult to analyze deacetylation products of
CA polymer directly because partially deacetylated CA
is a heterogeneous molecule with a high molecular
weight and changed solubility. Acetylated methyl gluco-
side was chosen as a model compound of CA. The
compound mimics acetylated glucoside residues in the
CA molecule. Since CA has the ꢀ-1,4 glycosidic
linkage, we used a methylated saccharide at anomeric
carbon by ꢀ-linkage. In this study, we investigated the
substrate specificity of CA esterase as to whether the
enzyme hydrolyzes 2,3,4,6-Ac-MeꢀGlc and 2,3,4,6-Ac-
MeꢀGal regioselectively or randomly.
When CA esterase hydrolyzed 2,3,4,6-Ac-MeGlc, we
observed several partially deacetylated products by
GLC. The reaction products were mainly four com-
pounds, products A, B, C, and D. Each product was
estimated as to the number of acetyl residues by GLC–
MS. The positions of deacetylation were identified by
NMR spectroscopy. The deacetylation reactions of
2,3,4,6-Ac-MeꢀGlc catalyzed by CA esterase from
N. sicca SB are summarized in Fig. 3. There are four
positions of the acetyl group in the 2,3,4,6-Ac-MeꢀGlc
molecule that CA esterase can attack initially. 2,3,4,6-
Ac-MeꢀGlc was most rapidly deacetylated at position
C-3 to give 2,4,6-Ac-MeꢀGlc as the main initial
product. Deacetylation was followed at position C-2,
and generated 4,6-Ac-MeꢀGlc. Deacetylation continued
at position C-4 or C-6 to produce a mixture of 6-Ac-
MeꢀGlc and 4-Ac-MeꢀGlc transiently. 4-Ac-MeꢀGlc
and 6-Ac-MeꢀGlc were subjected to further slow
deacetylation and were converted to MeꢀGlc. CA
esterase deacetylated 2,3,4,6-Ac-MeꢀGlc in preferential
order. These results suggest that CA esterase most
rapidly deacetylated at position C-3 of 2,3,4,6-Ac-
MeꢀGlc, and that the deacetylation rate at position C-
2 was slower than at position C-3, and further that
deacetylation at positions C-4 and C-6 was yet slower.
Regioselective deacetylation for acetylsaccharides has
been reported as to a few acetylxylan esterases.
Acetylxylan esterases from Streptomyces lividans,21)
Trichoderma reesei,22) and Schizophyllum commune23)
were investigated as to deacetylation of 2,3,4,6-Ac-
MeꢀGlc. These enzymes deacetylated 2,3,4,6-Ac-
MeꢀGlc at positions C-2 and C-3, yielding 4,6-Ac-
MeꢀGlc as an final product or an intermediate. But there
were some differences in the action of the acetylxylan
esterases from that of CA esterase. Acetylxylan esterase
from S. lividans removed two acetyl groups simulta-
neously from positions C-2 and C-3, to give 4,6-Ac-
MeꢀGlc as an final product in 90–95% yield.21)
Acetylxylan esterase from T. reesei22) exhibited similar
regioselectivity to acetylxylan esterase from S. lividans.
The enzyme initially removed two acetyl groups at
positions C-2 and C-3, producing 4,6-Ac-MeꢀGlc in
30% yield. The intermediates of this conversion, 2,4,6-
and 3,4,6-Ac-MeꢀGlc, were not detected in the reaction

Deacetylation by Cellulose Acetate Esterase
1297
mixture. 4,6-Ac-MeꢀGlc was further converted, mainly
molecules.
to 4-Ac-MeꢀGlc as an final product in 60% yield.
Acetylxylan esterase from S. commune most rapidly
deacetylated at position C-3 of 2,3,4,6-Ac-MeꢀGlc, and
generated 2,4,6-Ac-MeꢀGlc as the major initial product
in 50–60% yield.23) 2,4,6-Ac-MeꢀGlc was further
deacetylated, mainly at position C-2, to give 4,6-Ac-
MeꢀGlc in 55% yield. 4,6-Ac-MeꢀGlc was not further
converted to 4-Ac-MeꢀGlc or 6-Ac-MeꢀGlc. CA ester-
ase from N. sicca SB produced 2,4,6-Ac-MeꢀGlc in
about 70% yield, the highest yield among these
enzymes. These facts suggest that the deacetylation rate
by CA esterase at position C-3 was faster and/or that
those at other positions were slower than those enzymes.
CA esterase from N. sicca SB gave 4,6-Ac-MeꢀGlc, 4-
Ac-MeꢀGlc, and 6-Ac-MeꢀGlc in 50%, 30%, and 15%
yield respectively as the maximum yield from 2,3,4,6-
Ac-MeꢀGlc. This indicates that CA esterase deacety-
lated at positions C-4 and C-6, though the reaction rates
are very slow, as well as at positions C-2 and C-3.
In nature, deacetylation of acetylxylan residues in
heteroxylans by acetylxylan esterase leads to an efficient
degradation of heteroxylans by xylanases.9) The prop-
erty of rapid deacetylation at positions C-2 and C-3 of
acetylxylose21–23) is important for deacetylation of
acetylxylan by acetylxylan esterase, but the deacetyla-
tion at positions C-4 and C-6 is not important because
there are no acetyl groups at position C-4 and no C-6
carbon in the acetylxylan molecule. On the other hand,
the property of deacetylation of 2,3,4,6-Ac-MeꢀGlc at
position C-6 as well as positions C-2 and C-3 is
important for complete degradation of CA. Kinetic
studies of the hydrolysis of carboxymethyl cellulose by
crude cellulase from Trichoderma viride have been
reported.24) It was concluded that hydrolysis occurred
easily when three adjacent unsubstituted glucosyl
residues were present, and very slowly when only two
adjacent unsubstituted residues were present. Endo-1,4-
ꢀ-glucanases from N. sicca SB did not hydrolyze the ꢀ-
1,4 linkage backbone in the CA molecule with a degree
of substitution of 1.77.12,13) When the degree of
substitution for CA was low, for example less than
1.0, endo-1,4-ꢀ-glucanases hydrolyzed the CA back-
bone. When CA with a degree of substitution of 0.88
was further deacetylated by CA esterase, the glucanases
hydrolyzed the products at a faster rate.12,13) This
suggests that the substituents in the cellulose molecule
influenced hydrolysis by cellulase and that removing the
substituents led to effective hydrolysis of the cellulose
molecule by cellulase. CA esterase deacetylated 2,3,4,6-
Ac-MeꢀGlc at position C-6 as well as at positions C-2
and C-3. This suggests that CA esterase can deacetylate
the CA molecule at position C-6, and that the properties
of CA esterase lead to effective degradation of CA by
synergistic action with endo-1,4-ꢀ-glucanases, and
further that CA esterase is superior for degradation of
CA to acetylxylan esterase, which cannot deacetylate at
position C-6 of acetylated glucosyl residues in CA
There have been several reports of regioselective
deacetylation of acetylated glycosides by lipases in
order to search for preparation of useful intermediates
for synthesis of saccharide derivatives. Lipase from
Candida rugosa hydrolyzed the acetyl group esterifing
at position C-6 in glucoside, and did not hydrolyze at
any other position of the acetyl group.15) Lipase from
Pseudomonas cepacia hydrolyzed 2,3,4-tri-O-acetyl-ꢀ-
D-xylopyranoside at position C-4.25) Lipase from Asper-
gillus niger hydrolyzed 2,3,4,6-Ac-MeꢀGlc at position
C-3 to give 2,4,6-Ac-MeꢀGlc only.26) Crude porcine
pancreatic lipase hydrolyzed ꢀ-D-glucose pentaacetate at
position C-1.27) These lipases displayed high regiose-
lectivity for deacetylation of acetylsaccharides. Deace-
tylation of 2,3,4,6-Ac-MeꢀGlc by CA esterase from
N. sicca SB proceeded easily at position C-3, slowly at
position C-2, and at positions C-4 and C-6 at a much
slower rate. CA esterase had a different regioselectivity
from these lipases.
Deacetylation of cellulose triacetate in dimethylsulf-
oxide was achieved with dimethylamine.28) The contents
of acetyl groups in the products at positions C-2, C-3,
and C-6 were 17%, 25%, and 58% respectively when the
average of degree of substitution for the mixture of the
products was 1.45. The rate of deacetylation was fast in
the order of positions C-2, C-3, and C-6. This suggests
that deacetylation of acetylated saccharides by the basic
catalysts showed much lower regioselectivity than that
by CA esterase, and that the way of deacetylation by CA
esterase was different from that by the basic catalysts.
Regioselective deacetylation by CA esterase is probably
derived from the affinity of acetylated saccharides and
CA esterase. The affinity of CA esterase to acetylated
saccharides probably leads to an effective deacetylation
of CA in CA degradation under neutral conditions.
We also studied the regioselectivity of CA esterase
for 2,3,4,6-Ac-MeꢀGal in order to estimate the effect of
stereoisomers. Deacetylation reactions are summarized
in Fig. 3. For 2,3,4,6-Ac-MeꢀGal, CA esterase rapidly
deacetylated at position C-3 or position C-6 to give
2,4,6-Ac-MeꢀGal or 2,3,4-Ac-MeꢀGal as the intermedi-
ate. Deacetylation was followed at position C-6 for
2,4,6-Ac-MeꢀGal and at position C-3 for 2,3,4-Ac-
MeꢀGal, and generated 2,4-Ac-MeꢀGal in 55% yield.
2,4-Ac-MeꢀGal was subjected to further deacetylation
to generate product H and MeꢀGal at a much slower
rate. The structure of product H is assumed to be 2-Ac-
MeꢀGal or 4-Ac-MeꢀGal generated by hydrolysis of
2,4-Ac-MeꢀGal. The amount of product H was much
smaller, and we could not analyze it further. These
results suggest that CA esterase rapidly deacetylated at
positions C-3 and C-6 of 2,3,4,6-Ac-MeꢀGal, and that
the deacetylation rate at position C-3 was slightly faster
than at position C-6, and further that deacetylation at
positions C-2 and C-4 proceeded at a much slower rate.
A different regioselectivity for 2,3,4,6-Ac-MeꢀGal from
2,3,4,6-Ac-MeꢀGlc was observed on deacetylation by

1298
K. MORIYOSHI et al.
7) Tuohy, M. G., Puls, J., Claeyssens, M., Vrsanska, M.,
and Coughlan, M. P., The xylan-degrading enzyme
system of Talaromyces emersonii: novel enzymes with
activity against aryl ꢀ-D-xylosides and unsubstituted
xylans. Biochem. J., 290, 515–523 (1993).
8) Shao, W., Obi, S. K. C., Puls, J., and Wiegel, J.,
Purification and characterization of the ꢁ-glucuronidase
from Thermoanaerobacterium sp. strain JW/SL-YS485,
an important enzyme for the utilization of substituted
xylans. Appl. Environ. Microbiol., 61, 1077–1081
(1995).
CA esterase. Deacetylation by CA esterase at position
C-3 at a much faster rate for acetylgalactoside was
similar to that for acetylglucoside. The rate of deacety-
lation at position C-6 in acetylgalactoside was much
faster than that in acetylglucoside. On the other hand,
the rate of deacetylation at position C-2 in acetylgalacto-
side was much slower than that in acetylglucoside. The
acetyl group at position C-4 in 2,3,4,6-Ac-MeꢀGlc is
equatorial, whereas that in 2,3,4,6-Ac-MeꢀGal is axial.
These results suggest that the configuration at position
C-4 in the acetylglycoside molecule affected the rate of
deacetylation at positions C-2 and C-6, and that CA
esterase recognizes not only the acetyl moiety in the
acetylglycoside molecule but also part of the saccharide
structure in the acetylglycoside molecule. The config-
uration of acetylglycoside is important for the deacety-
lation reaction by CA esterase. There are few detailed
reports on the regioselectivity of lipases and acetylxylan
esterases for 2,3,4,6-Ac-MeꢀGal. Lipase from C. rugosa
has been shown to deacetylate 2,3,4,6-Ac-MeꢀGal only
at position C-6, not at any other position.15) When
we deacetylated 2,3,4,6-Ac-MeꢀGlc and 2,3,4,6-Ac-
MeꢀGal by lipase from C. rugosa, the rate of deacety-
lation at position C-6 was very slow. This indicates that
regioselective deacetylation by lipase from C. rugosa
was not affected by the configuration of the acetylsac-
charides, and that the enzyme did not bind to the whole
molecule of acetylsaccharides but recognized only the
acetyl group at the primary hydroxy group. Other
acetylated carbohydrates must be examined in order to
understand the recognition of substrate by CA esterase
in detail.
9) Biely, P., MacKenzie, C. R., and Schneider, H.,
Production of acetyl xylan esterase by Trichoderma
reesei and Schizophyllum commune. Can. J. Microbiol.,
32, 767–772 (1988).
10) Sakai, K., Yamauchi, T., Nakasu, F., and Ohe, T.,
Biodegradation of cellulose acetate by Neisseria sicca.
Biosci. Biotechnol. Biochem., 60, 1617–1622 (1996).
11) Moriyoshi, K., Ohmoto, T., Ohe, T., and Sakai, K.,
Purification and characterization of an esterase involved
in cellulose acetate degradation by Neisseria sicca SB.
Biosci. Biotechnol. Biochem., 63, 1708–1713 (1999).
12) Moriyoshi, K., Ohmoto, T., Ohe, T., and Sakai, K.,
Purification and characterization of an endo-1,4-ꢀ-
glucanase from Neisseria sicca SB that hydrolyzes ꢀ-
1,4 linkages in cellulose acetate. Biosci. Biotechnol.
Biochem., 66, 508–515 (2002).
13) Moriyoshi, K., Ohmoto, T., Ohe, T., and Sakai, K., Role
of endo-1,4-ꢀ-glucanases from Neisseria sicca SB in
synergistic degradation of cellulose acetate. Biosci.
Biotechnol. Biochem., 67, 250–257 (2003).
14) Bock, K., and Pedersen, C., Carbon-13 nuclear magnetic
resonance spectroscopy of monosaccharides. Adv. Car-
bohydr. Chem. Biochem., 41, 27–66 (1983).
15) Sweers, H. M., and Wong, C.-H., Enzyme catalyzed
regioselective deacylation of protected sugars in carbo-
hydrate synthesis. J. Am. Chem. Soc., 108, 6421–6422
(1986).
References
1) Bogan, R. T., and Brewer, R. J., ‘‘Encyclopedia of
Polymer Science and Engineering’’ 2nd ed. Vol. 3, eds.
Mark, H. F., Bikales, N. M., Overberger, C. G., and
Menges, G., John Wiley, New York, pp. 174–178
(1985).
2) Cantor, P. A., and Mechalas, B. J., Biological degrada-
tion of cellulose acetate reverse-osmosis membranes.
J. Polym. Sci. C, 28, 225–241 (1969).
16) Dubois, M., Gilles, K. A., Hamilton, J. K., Rebers, P. A.,
and Smith, F., Colorimetric method for determination of
sugars and related substances. Anal. Chem., 28, 350–356
(1956).
17) DeJongh, D. C., Radford, T., Hribar, J. D., Hanessian, S.,
Bieber, M., Dawson, G., and Sweeley, C. C., Analysis of
trimethylsilyl derivatives of carbohydrates by gas chro-
matography and mass spectrometry. J. Am. Chem. Soc.,
91, 1728–1740 (1969).
18) Biemann, K., DeJongh, D. C., and Schnoes, H. K.,
Application of mass spectrometry to structure problems.
XIII. Acetates of pentoses and hexoses. J. Am. Chem.
Soc., 85, 1763–1771 (1963).
19) Iwashita, T., In ‘‘Jikken Kagaku Koza Vol. 5, NMR’’ 4th
ed. (in Japanese), Maruzen, Tokyo, pp. 99–178 (1991).
20) Yoshimoto, K., Itatani, Y., Shibata, K., and Tsuda, Y.,
Utilization of sugars in organic synthesis. II. Syntheses
and proton- and carbon-13-nuclear magnetic resonance
spectra of all positional isomers of methyl mono-O-
tetradecanoyl-ꢁ- and ꢀ-D-glucopyranosides. Chem.
Pharm. Bull., 28, 208–219 (1980).
3) Nelson, M., McCarthy, S. P., and Gross, R. A., Isolation
of
a
Pseudomonas paucimobilis capable of using
insoluble cellulose acetate as a sole carbon source. Proc.
ACS Div. Polym. Mater. Sci. Eng., 67, 139–140 (1992).
4) Ishigaki, T., Sugano, W., Ike, M., and Fujita, M.,
Enzymatic degradation of cellulose acetate plastic by
novel degrading bacterium Bacillus sp. S2055. J. Biosci.
Bioeng., 90, 400–405 (2000).
5) Christov, L. P., and Prior, B. A., Esterases of xylan-
degrading microorganisms: production, properties, and
significance. Enzyme Microb. Technol., 15, 460–475
(1993).
6) Nanmori, T., Watanabe, T., Shinke, R., Kohno, A., and
Kawamura, Y., Purification and properties of thermo-
stable xylanase and ꢀ-xylosidase produced by a newly
isolated Bacillus stearothermophilus strain. J. Bacteriol.,
172, 6669–6672 (1990).
21) Biely, P., Cote, G. L., Kremnicky, L., Greene, R. V.,
Dupont, C., and Kluepfel, D., Substrate specificity and
mode of action of acetylxylan esterase from Streptomy-
ces lividans. FEBS Lett., 396, 257–260 (1996).

Deacetylation by Cellulose Acetate Esterase
1299
22) Biely, P., Cote, G. L., Kremnicky, L., Greene, R. V., and
Tenkanen, M., Action of acetylxylan esterase from
Trichoderma reesei on acetylated methyl glycosides.
FEBS Lett., 420, 121–124 (1997).
23) Biely, P., Cote, G. L., Kremnicky, L., Weisleder, D., and
Greene, R. V., Substrate specificity of acetylxylan
esterase from Schizophyllum commune: mode of action
on acetylated carbohydrates. Biochim. Biophys. Acta,
1298, 209–222 (1996).
24) Eriksson, K.-E., and Hollmark, B. H., Kinetic studies of
the action of cellulase upon sodium carboxymethyl
cellulose. Arch. Biochem. Biophys., 133, 233–237
(1969).
25) Lopez, R., Perez, C., Fernandez-Mayoralas, A., and
Conde, S., Enzymic transesterification of alkyl 2,3,4-tri-
O-acetyl-ꢀ-D-xylopyranoside. J. Carbohydr. Chem., 12,
165–171 (1993).
26) Chen, H.-P., Hsiao, K.-F., Wu, S.-H., and Wang, K.-T.,
Regioselectivity enhancement by partial purification of
lipase from Aspergillus niger. Biotechnol. Lett., 17, 305–
308 (1995).
27) Hennen, W. J., Sweers, H. M., Wang, Y.-F., and Wong,
C.-H., Enzymes in carbohydrate synthesis: lipase-cata-
lyzed selective acylation and deacylation of furanose and
pyranose derivatives. J. Org. Chem., 53, 4939–4945
(1988).
28) Heinze, T., and Liebert, T., Chemical characteristics of
cellulose acetate. In ‘‘Cellulose Acetates: Properties and
Applications’’, ed. Rustemeyer, P., Wiley-VCH, Frei-
burg, pp. 167–237 (2004).