Enzymatic synthesis of 4-methylumbelliferyl (1→3)-β-D-glucooligosaccharides - New substrates for β-1,3-1,4-D-glucanase

Article abstract of DOI:10.1016/S0008-6215(03)00199-X

The transglycosylation reactions catalyzed by β-1,3-D-glucanases (laminaranases) were used to synthesize a number of 4-methylumbelliferyl (MeUmb) (1→3)-β-D-gluco-oligosaccharides having the common structure [β-D-Glcp-(1→3)]_n-β-D-Glcp-MeUmb, where n=1-5. The β-1,3-D-glucanases used were purified from the culture liquid of Oerskovia sp. and from a homogenate of the marine mollusc Spisula sachalinensis. Laminaran and curdlan were used as (1→3)-β-D-glucan donor substrates, while MeUmb-β-D-glucoside (MeUmbGlcp) was employed as a transglycosylation acceptor. Modification of [β-D-Glcp-(1→3)]₂-β-D-Glcp-MeUmb (MeUmbG₃) gives 4,6-O-benzylidene-D-glucopyranosyl or 4,6-O-ethylidene-D-glucopyranosyl groups at the non-reducing end of artificial oligosaccharides. The structures of all oligosaccharides obtained were solved by ¹H and ¹³C NMR spectroscopy and electrospray tandem mass spectrometry. The synthetic oligosaccharides were shown to be substrates for a β-1,3-1,4-D-glucanase from Rhodothermus marinus, which releases MeUmb from β-di- and β-triglucosides and from acetal-protected β-triglucosides. When acting upon substrates with d.p.>3, the enzyme exhibits an endolytic activity, primarily cleaving off MeUmbGlcp and MeUmbG₂.

Full text of DOI:10.1016/S0008-6215(03)00199-X

Carbohydrate Research 338 (2003) 1455ꢁ1467
/
www.elsevier.com/locate/carres
Enzymatic synthesis of 4-methylumbelliferyl (10
/
3)-b-
D
-
-glucanase
glucooligosaccharides*new substrates for b-1,3-1,4-
/
D
Rainer Borriss,a Martin Krah,a Harry Brumer, III,b Maxim A. Kerzhner,c
Dina R. Ivanen,c Elena V. Eneyskaya,c Lyudmila A. Elyakova,d
Sergei M. Shishlyannikov,c Konstantin A. Shabalin,c Kirill N. Neustroevc,*
a AG Bakteriengenetik, Institut fur Biologie, Humboldt Universitat Berlin Chausseestrasse 117, D-10115 Berlin, Germany
¨
b Department of Biotechnology, Royal Institute of Technology (KTH), Stockholm Center for Physics, Astronomy, & Biotechnology, Roslagsvagen 30 B,
S-106 91 Stockholm, Sweden
¨
c Petersburg Nuclear Physics Institute, Russian Academy of Science, Molecular and Radiation Biology Division, Gatchina, St. Petersburg 188300,
Russian Federation
d Pacific Insititute of Bioorganic Chemistry, Far East Division Russian Academy of Sciences, pr. 100-letiya Vladivostoka 159, Vladivostok 690022,
Russian Federation
Received 2 January 2003; accepted 20 April 2003
Abstract
The transglycosylation reactions catalyzed by b-1,3-
methylumbelliferyl (MeUmb) (103)-b- -gluco-oligosaccharides having the common structure [b-
MeUmb, where nꢀ1ꢁ5. The b-1,3-D-glucanases used were purified from the culture liquid of Oerskovia sp. and from a
homogenate of the marine mollusc Spisula sachalinensis. Laminaran and curdlan were used as (103)-b- -glucan donor substrates,
while MeUmb-b- -glucoside (MeUmbGlcp) was employed as a transglycosylation acceptor. Modification of [b- -Glcp-(103)]2-b-
-Glcp-MeUmb (MeUmbG3) gives 4,6-O-benzylidene- -glucopyranosyl or 4,6-O-ethylidene- -glucopyranosyl groups at the non-
D
-glucanases (laminaranases) were used to synthesize a number of 4-
/
D
D
-Glcp-(103)]n-b- -Glcp-
/
D
/
/
/
D
D
D
/
D
D
D
reducing end of artificial oligosaccharides. The structures of all oligosaccharides obtained were solved by 1H and 13C NMR
spectroscopy and electrospray tandem mass spectrometry. The synthetic oligosaccharides were shown to be substrates for a b-1,3-
1,4-
D
-glucanase from Rhodothermus marinus, which releases MeUmb from b-di- and b-triglucosides and from acetal-protected b-
3, the enzyme exhibits an endolytic activity, primarily cleaving off
MeUmbGlcp and MeUmbG2.
triglucosides. When acting upon substrates with d.p.ꢀ
/
# 2003 Elsevier Science Ltd. All rights reserved.
Keywords: 4-Methylumbelliferyl (10
/
3)-b-D
-glucoside; b-1,3-1,4-
D
-Glucanase; Rhodothermus marinus; Transglycosylation; Laminaranase
1. Introduction
kinetic assays which are both more sensitive and more
simple to analyse when compared with assays where
mono- or oligosaccharides are detected as products of
hydrolysis.1,2 Synthetic fluorogenic mono- and oligosac-
charides have also been used as modified ligands to
study lectins.3 Commercially available 4-methylumbelli-
feryl-, o-nitrophenyl- and p-nitrophenyl (PNP) glyco-
sides used as substrates for exo-glycosidases are, as a
general rule, products of relatively high-yielding organic
synthesis.4,5 Obtaining fluorogenic and chomogenic
oligosaccharides, substrates for endo-glycosidases,6 is
Fluorogenic and chromogenic saccharides are conveni-
ent and widely used substrates to study the kinetic
properties of glycosyl hydrolases and transglycosylases,
since many exo- and endo-glycosidases are able to
release chromophoric or fluorophoric aglycon groups
from these substrates. This allows the development of
* Corresponding author. Tel.: ꢂ
/
7-812-7132014; fax: ꢂ
/
7-
essentially a more difficult task for organic chemistry.
Classical synthetic approaches suffer the drawback,
however, that they require numerous protection and
812-7132303.
E-mail
Neustroev).
address:
neustk@omrb.pnpi.spb.ru (K.N.
0008-6215/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0008-6215(03)00199-X

1456
R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ1467
/
deprotection steps in addition to those in which
glycosidic bonds are formed.7,8 Considerable effort has
gone into studying the application of the transglycosy-
lating ability of retaining glycosidases as an alternative
approach for the synthesis of chromo- and fluorogenic
oligosaccharide substrates.9ꢁ11 One potential limitation
with this approach, however, is that the products of
transglycosylation may themselves be hydrolysed by the
enzyme during the course of the reaction. The introduc-
tion of glycosynthases, mutant glycosidases which lack a
catalytic nucleophile and are thus incapable of carrying
out substrate hydrolysis but which catalyse transglyco-
sylation in the presence of glycosyl fluoride donors, has
overcome this problem for the synthesis of a number of
b-12,13 and a-aryl14 oligosaccharides.
ability of the enzyme was studied by analysing MeUmb-
containing oligosaccharides with thin-layer chromato-
graphy (TLC) and analytical reverse-phase high perfor-
mance liquid chromatography (HPLC). Separated
oligosaccharides containing the fluorophoric group
were isolated from the reaction mixture by gel-permea-
tion chromatography followed by a second purification
step using reverse-phase HPLC. Individual fractions
obtained by HPLC were more than 95% pure and were
further characterized with electrospray ionisation mass
1
spectrometry, H and 13C NMR.
ESIꢁ
/
TOF MS data recorded for the (Mꢂ
Na)ꢂ ion,
/
which was formed to the exclusion of any other
molecular adducts under the experimental conditions,
indicate that the observed masses correlate well with the
calculated monoisotopic masses of the expected pro-
ducts (Table 1). In order to provide further evidence for
the proposed structures, tandem mass spectra (MS/MS)
were recorded for each compound in Table 1. The
benzylidene-protected oligosaccharides gave excellent
fragmentation patterns, which allowed the confirmation
of the stoichiometry of glucosyl, methylumbelliferyl and
benzylidene residues (Fig. 1). However, none of the
unprotected MeUmb b-laminara-oligosaccharides, with
the exception of the laminaratriose substrate, produced
a complete Y- or B-type fragment ion series under a
variety of CID conditions studied. Instead, the only
predominant fragment ions observed in MS/MS spectra
of these compounds were those resulting from the
neutral loss of methylumbelliferone (7-hydroxy-4-
methylcoumarin) to yield a B-type ion and of a
MeUmb-dehydroglucosyl residue to yield a C-type ion
(Fig. 2). Generation of the observed C1 ions may occur
In this work, we used the inherent transglycosylating
activity of two endo-b-1,3-D-glucanases (laminara-
nases), one from the bacterium Oerskovia sp. and one
from the marine mollusc S. sachalinensis, to synthesize
b-
of [b-
as novel substrates for analysing the b-1,3-1,4-
nase (b-1,3-1,4- -glucan glucohydrolase; E.C. 3.2.1.73,
b-1,3-1,4-glucanase) from Rhodothermus marinus. This
-glucanase belongs to glycosyl
D
-gluco-oligosaccharides with the common structure
-Glc-(103)]n-b- -Glc-MeUmb, where nꢀ1ꢁ5,
-gluca-
D
/
D
/
/
D
D
thermophilic b-1,3-1,4-
D
hydrolase family 16, which is comprised mainly of endo-
acting b-glucanases. The range of substrates acted upon
by enzymes in this family includes (10
/
3)-b-
D
-glucans,
such as laminaran, or mixed linkage (10
/
3),(10
/
4)-b-
D-
glucans like lichenan and barley glucan.15 Although the
enzymatic properties, DNA sequence and transglycosy-
lation activity of this enzyme have been described
recently,15,16 elucidating the structureꢁ
function aspects
/
of the transglycosylation activity requires more detailed
study, which is made possible only with non-natural
substrates. The novel MeUmb substrates presented in
this study should prove generally useful for future
comparative studies of wild-type and mutant b-1,3-1,4-
sequentially by b elimination from the Bn ion (nꢀ
/
Table 1
Calculated and observed monoisotopic molecular masses of
sodium adducts using externally calibrated ESIꢁ
/TOF MS
D-glucanases from a variety of sources.
Product
Calculated
Observed Error
(ppm)
20
2. Results and discussion
MeUmbG2
C22H28NaO13:
523.1428
523.1535
685.2074
847.2820
1009.3408
1171.4065
773.2299
935.2888
1097.3329
An extracellular b-1,3-D-glucanase (laminaranase) pos-
MeUmbG3
C28H38NaO18:
685.1956
17
sessing transglycosylation activity was isolated from
culture filtrates of Oerskovia sp. The partially purified
enzyme contained less than 0.1% each of b-glucosidase,
b-1,6-glucanase and cellulase activities, thus permitting
use of the preparation for the enzymatic synthesis of
MeUmbG4
C34H48NaO23:
847.2484
40
MeUmbG5
C40H58NaO28:
1009.3012
39
MeUmbG6
C46H68NaO33:
1171.3541
45
(10
/
3)-b-
D
-glucosides. The enzyme preparation was
7 (20
found to be stable at 37 8C in the pH range of 4ꢁ
/
BnzMeUmbG3
BnzMeUmbG4
BnzMeUmbG5
C35H42NaO18:
773.2269
3.8
9.7
0.36
mM sodium acetate or 20 mM sodium phosphate
buffers) for at least 72 h.
The enzyme exhibits transglycosylation activity when
C41H52NaO23:
935.2797
curdlan [(10
/
3)-b-
D-glucan] was used as a donor and
MeUmbGlcp as an acceptor. The transglycosylation
C47H62NaO28:
1097.3325

R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ
/
1467
1457
Fig. 1. CID MS/MS spectra of benzylidene-protected MeUmb-(10
/
3)-glucooligosaccharides. (A) Bzl-Glc3-MeUmb; (B) Bzl-Glc4-
MeUmb; (C) Bzl-Glc5-MeUmb. Labels in bold italics marked with an asterix indicate peaks used for ‘lock mass’ correction.
number of glucose units in the parent ion) and provides
additional evidence for a (103)-b linkage between the
first and second glucosyl residues. By comparison,
b(10
/
4)-linked MeUmb gluco-oligosaccharides give
/
complete B-ion series when subjected to CID under
similar conditions [Brumer, Neustroev, and coworkers,

1458
R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ1467
/
unpublished]. To further characterize these fragment
ions, ‘virtual’ MS3 experiments were performed on the
shown in Fig. 3, the fragment ions observed in spectrum
B at m/z 833.3 and m/z 689.2 yield complete Y- and B-
type ion series in addition to ions resulting from cross-
ring fragmentation. The observation of these latter ions
is most certainly a consequence of the considerably
higher energy imparted on the parent ion in the source
region of the mass spectrometer, since the collision
energy is similar (approx 50 V) to that used to obtain
MeUmb-Glc5 compound by subjecting the (Mꢂ
Na)ꢂ
/
adduct to higher cone voltages sufficient to cause in-
source fragmentation. Fragment ions with identical
masses to those observed previously in the CID spectra
were produced, and these were individually selected with
the quadrupole to generate CID MS/MS spectra. As
Fig. 2. CID MS/MS spectra of MeUmb-(10
/
3)-glucooligosaccharides. (A) Glc2-MeUmb; (B) Glc3-MeUmb; (C) Glc4-MeUmb; (D)
Glc5-MeUmb; (E) Glc6-MeUmb. Labels in bold italics marked with an asterix indicate peaks used for ‘lock mass’ correction.

R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ
/
1467
1459
Fig. 2 (Continued)
CID of the (Mꢂ
/
Na)ꢂ ion under lower cone voltage
conditions. Taken together, the monoisotopic mass
value obtained in spectrum A and the tandem MS
determined by analysing the positions of the low-field
signals for C-3 in the 13C NMR spectra. Data from 2D
1
NMR spectroscopy (COSY, Hꢁ13
/
C-heteronuclear cor-
relations, and NOESY) establish the structures of the
data in spectra Bꢁ
/
D (Fig. 3) provide conclusive evidence
of the stoichiometry of glucosyl and methylumbelliferyl
residues in the molecule. Similar exhaustive fragment-
ion analysis was not carried out for other molecules in
the glycan series, but the correctness of the ion assign-
ments in Fig. 2 may be inferred by analogy.
oligosacharides unequivocally. After enzymatic synth-
esis with Oerskovia b-1,3-
GlcnMeUmb oligosaccharides with nꢀ
tained from curdlan. The yield for each of these
oligosaccharides is shown in Table 4. Thus, only (10
3)-b-linked MeUmb b-laminara-oligosaccharides with
d.p. 2ꢁ4 were produced as a result of the regio- and
D
-glucanase, a set of (10
/
3)-D-
/
2ꢁ4 was ob-
/
/
The exact numbers of glucosyl residues were also
1
derived from H and 13C NMR spectra, which contain
/
from two to four characteristic, low-field signals for
anomeric protons and carbon atoms (Table 2). The 13C
NMR spectra contain also the same numbers of high-
field signals for C-6, which can be easily identified by the
DEPT technique. Analysis of vicinal coupling constants
stereoselective synthesis performed with the aid of the
transglycosylation ability of the endo-b-glucanase from
Oerskovia sp. When laminaran from Laminaria digitata
was also used as a glycosyl donor of b-laminara-
oligosaccharides for the transglycosylation reaction
1
of H spectra shows that all glucose units are hexopyr-
catalyzed by the b-1,3-
D-glucanase from Oerskovia sp.,
anoses with equatorial hydroxy groups, namely b-
glucopyranose. Anomeric configurations of all the
monosaccharide residues were confirmed by the low-
field chemical shift values for the anomeric carbon
TLC and HPLC analyses of these products showed
similar sets of products ranging from d.p. 2 to 4.
However, the use of laminaran as a source of b-
oligosaccharides for the enzymatic synthesis is less
atoms (Table 3). In addition, the (10
/
3)-type of the
preferable due to presence of about 10% b-(10
/
6)-linked
glycosidic linkages in the oligosaccharides obtained was
glucan branches on the main chain. As a result, the

1460
R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ1467
/
Fig. 3. CID MS/MS spectra of in-source fragment ions of the Glc5-MeUmb sodium adduct. (A) TOF MS spectrum of Glc5-
MeUmb with cone voltage set at 80 V; (B) TOF MS spectrum of Glc5-MeUmb with cone voltage set at 160 V; (C) MS/MS spectrum
of in-source fragment 833.25 (cone 160 V, collision energy 60 V); (D) MS/MS spectrum of in-source fragment 689.21 (cone 160 V,
collision energy 60 V). Labels in bold italics marked with an asterix indicate peaks used for ‘lock mass’ correction.

R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ
/
1467
1461
Table 2
Chemical shifts obtained from H NMR spectra of MeUmb b-laminaraoligosaccharides
1
Product
H-1
H-2
H-3
H-4
H-5
H-6a
H-6b
MeUmbG2
b-
b-
D
-Glcp-(10
/
3)
B
5.130
4.803
3.805
3.398
3.881
3.551
3.636
3.429
3.695
3.517
3.977
3.946
3.817
3.742
D-Glcp
A
MeUmbG3
b-
b-
b-
D
D
D
-Glcp-(10
-Glcp-(10
-Glcp
/
3)
3)
C
B
A
5.196
4.851
4.770
3.826
3.595
3.374
3.908
3.801
3.538
3.647
3.538
3.416
3.722
3.537
3.494
3.986
3.953
3.927
3.819
3.764
3.727
/
MeUmbG4
b-
b-
b-
b-
D
D
D
D
-Glcp-(10
-Glcp-(10
-Glcp-(10
-Glcp
/
3)
3)
3)
D
C
B
5.220
4.848
4.811
4.759
3.829
3.593
3.565
3.363
3.909
3.808
3.784
3.530
3.647
3.530
3.530
3.409
3.728
3.525
3.531
3.486
3.982
3.950
3.929
3.921
3.815
3.761
3.756
3.721
/
/
A
Table 3
Chemical shifts obtained from 13C NMR spectra of MeUmb b-laminaraoligosaccharides
Product
C-1
C-2
C-3
C-4
C-5
C-6
MeUmbG2
b-
b-
D
-Glcp-(10
/
3)
B
101.95
105.32
75.03
75.96
86.51
78.04
70.34
72.07
78.31
78.51
62.90
63.20
D-Glcp
A
MeUmbG3
b-
b-
b-
D
D
D
-Glcp-(10
-Glcp-(10
-Glcp
/
3)
3)
C
B
A
101.96
105.04
105.32
75.09
71.76
75.94
86.35
86.77
76.05
70.31
70.63
72.06
78.32
78.13
78.49
62.92
63.19
63.19
/
MeUmbG4
b-
b-
b-
b-
D
D
D
D
-Glcp-(10
-Glcp-(10
-Glcp-(10
-Glcp
/
3)
3)
3)
D
C
B
101.95
105.03
105.00
105.28
75.09
75.80
75.72
75.92
86.28
86.52
86.71
78.02
70.29
70.58
70.58
72.04
78.31
78.12
78.09
78.47
62.88
63.15
63.15
63.15
/
/
A
products of the reaction had the same length but
differed in configuration of b-glucosidic bonds. These
products were only poorly separated by HPLC with a
relatively low yield of approx 40%, which limited the
based upon its recent successful application in the
synthesis of PNP b-laminara-oligosaccharides.18 Trans-
Table 4
amount of MeUmb (103)-b-glucosides which could be
/
Yields of products from enzymatic synthesis catalyzed by b-
1,3-D-glucanases from Oerskovia sp. and mollusc S. sachali-
nensis using laminaran and curdlan as donors
obtained (Table 4). The pH optimum of the transglyco-
sylation reaction lies between 5 and 6.5 (Fig. 4),
however, the ratio of the individual (103)-b-gluco-
/
Product
Oerskovia b-1,3-
canase
D
-glu- Mollusc b-1,3-
D-gluca-
oligosaccharides formed in the course of the reaction is
pH independent.
nase
With the Oerskovia sp. enzyme, the lowest yield
obtained was for MeUmbG4 (Table 4) and MeUmb-
Donor, yield (%)
Donor, yield (%)
Curdlan Laminaran Curdlan Laminaran
containing b-D-glucosides of higher polymerization were
detected in trace quantities. Since higher oligosacchar-
ides with a d.p. of 4 or greater would serve as useful
MeUmbG2 30
MeUmbG3 25
21
17
12
8
30
20
15
8
substrates for kinetic studies of the b-1,3-1,4-
nases,17 enzymatic synthesis was explored with the endo-
1,3-b- -glucanase (laminaranase) from the marine mol-
lusc S. sachalinensis. The selection of this enzyme was
D-gluca-
MeUmbG4
8
4
Traces
traces
traces
MeUmbG5 traces
MeUmbG6 traces
traces
traces
D
6

1462
R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ1467
/
Fig. 4. The pH-optima of the Oerskovia endo-b-glucanase-catalyzed reactions: k, product formation during the hydrolysis of
laminaran; m and j, product formation during the transglycosylation reactions with laminaran and curdlan as donors, respectively.
The analyses were done on a reversed-phase Waters Hypersil ODS column using a linear gradient (0ꢁ/90%) of MeCN in water. The
amounts of products formed were calculated by integrating the corresponding chromatographic peaks.
matic synthesis reported herein has certain advantages
in comparison to the traditional synthetic route which
has been presented for oligosaccharides possessing the
Table 5
Kinetic parameters of hydrolysis of MeUmb-containing b-
glucooligosaccharides by b-1,3-1,4-
inus
D-
D-glucanase from R. mar-
common structure [b-
D
-Glcp-(10
/
4)]n-b-
D
3).8 The enzymatic ap-
-Glcp-(10
/
3)-b- -Glcp-MeUmb (nꢀ
D
/
1ꢁ
/
Substrate
Km (mmol) 105ꢃ
Vmax (mmol/(min mg))
/
proach does not require as many individual reaction
steps and, as a result, has a higher total yield of the final
products. Most notably, the enzymatic synthesis permits
one to obtain glucooligosaccharides with higher d.p.,
such as MeUmbG5 and MeUmbG6. Additionally, the
pure organic synthetic route to produce MeUmb-b-
laminara-oligosaccharides requires sizable quantities of
individually purified oligosaccharides derived from
laminaran as starting materials. Such oligosaccharides
can only be produced from laminaran or curdlan by
enzymatic digestion or acid hydrolysis followed by
chromatographic purification, which is both low yield-
ing and tedious.21,22
MeUmbG2
MeUmbG3
103.1
28.6
32.5
94.3
20.2
25.4
15.8
5.4
BnzMeUmbG3
EtMeUmbG3
glycosylation with the b-1,3-D-glucanase from S. sacha-
linensis lead to the formation of products with a d.p. of
5 and 6 in quantities suitable for enzymatic studies
(Table 5). Minor quantities of MeUmb-b-laminara-
oligosaccharides with d.p. more than 6 were detected
in the reaction mixtures by TLC and HPLC, however,
these were not pursued due to their low yields (B
analytical HPLC analysis).
/2% by
All the synthesized MeUmb oligosaccharides were
D-glucanase from
tested as substrates for the b-1,3-1,4-
It should be noted that microbial glycosyl hydrolases
are traditionally considered to be good tools for the
enzymatic synthesis of glycosides due to their high pH-
and thermal-stability, and because they are readily
produced in large scale.19 In comparison, eukaryotic
glycosyl hydrolases are used less frequently.20 In this
study, the combined application of the microbial
R. marinus. As determined by HPLC and TLC analyses,
the enzyme releases MeUmb from MeUmbG2 and
MeUmbG3 (kinetic parameters for these reactions are
shown in Table 5), while the reaction of the enzyme with
MeUmb oligosaccharides of higher d.p. (4ꢁ6) gives a
/
number of products. Table 6 shows the ratio of
MeUmb-containing products released during enzymatic
hydrolysis. The data indicate that MeUmbG2 and
MeUmbG3 are ideal substrates for precise kinetic
studies of the enzyme due to a higher sensitivity of
detection when compared with the methods of detecting
Oerskovia and mollusc endo-b-1,3-
D-glucanases proved
D-
effective in the production of a set of MeUmb-b-
glucooligosaccharides due to their complimentary sub-
strate specificities and product distributions. The enzy-

R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ
/
1467
1463
Table 6
Yields a of products from hydrolysis of MeUmb b-laminaraoligosaccharides by b-1,3-1,4-
D
-glucanase from R. marinus
Substrate
Products (mol/mol of MeUmb products) b
MeUmb
MeUmbG
MeUmbG2
0.76
MeUmbG3
MeUmbG4
MeUmbG5
MeUmbG2
MeUmbG3
MeUmbG4
MeUmbG5
MeUmbG6
0.24
0.34
0.22
0.23
0.345
0.65
0.11
0.14
0.09
0.083
0.03
0.005
0.005
0.5
0.06
0.012
0.009
0.007
0.6
0.44
a
The yields given are an average of at least 3 determinations.
b
Experimental error is 95% of values given.
/
reducing sugars released during hydrolysis23,24 or the use
of alternative substrates such as Congo Red dye-labeled
3. Experimental
D
-glucans.25 The ability to liberate MeUmb from
3.1. Materials
b-
MeUmbG2 and MeUmbG3 is most likely a common
feature in mode of action of b-1,3-1,4- -glucanases and
b-1,3- -glucanases. For example, MeUmb is the only
chromophoric product in hydrolysis of these substrates
by a b-1,3- -glucanase from Trichoderma sp. (data not
shown). A similar mode of action was also reported for
a b-1,3-1,4-
-glucanase from Bacillus licheniformis.26
Laminaran from L. digitata, barley (10
glucan, cellulose, p-nitrophenyl b-
/
3),(10
/
4)-b-
-glucosides were
D
D
D
from Sigma Chemical Co. (St. Louis, MO, USA).
Curdlan from Alcaligenes faecalis was kindly donated
by Professor Irwin J. Goldstein, University of Michigan,
Ann Arbor, USA. All reagents were of analytical or
research grade. 4-Methylumbelliferyl b-D-glucopyrano-
side was synthesized according to Ref. 5.
D
D
Consequently, we may suppose that the synthesized
substrates can be used broadly as effective tools in the
investigation of glycosidase activities in different mem-
bers of glycosyl hydrolase family 16 which act upon
3.2. Analytical methods
(10
glucans.
Protected MeUmb oligosaccharides with 4,6-O-ben-
zylidene- -glucopyranosyl or 4,6-O-ethylidene- -gluco-
/
3)-b-glucans and mixed linked (10
/
3),(10
/
4)-b-
D-
Protein concentrations were measured in accordance
with the Lowry procedure using BSA as a standard.29
The protein concentration of b-1,3-1,4-
D-glucanase
D
D
from R. marinus was determined using an absorption
coefficient at 282 nm of A1% 32.38 mL/(mg cm).16
Optical rotations were determined with a JASCO DIP-
360 polarimeter. Qualitative TLC analysis of oligosac-
charide substrates and products of enzymatic hydrolysis
were performed using Kieselgel 60 plates from Merck
(Darmstadt, Germany) with a mobile phase of 2:1:1
pyranosyl groups (BzlG3MeUmb and EtG3MeUmb)
were also shown to serve as substrates for the R.
marinus b-1,3-1,4-
D-glucanase. As with the unprotected
analogues, the enzyme releases MeUmb yielding BzlG3
and EtG3 as hydrolysis products. Interestingly, the KM
and Vmax values for these substrates (Table 5) are
essentially the same as those obtained for unprotected
EtOAcꢁ
/
AcOHꢁ
water. All 1H and 13C NMR spectra
/
MeUmbG2 and MeUmbG3 oligosaccharides and b-
glucans, laminaran and curdlan, that was reported
D-
were recorded with an AMX-500 Bruker spectrometer
(1H at 500.13 MHz, 13C at 125.13 MHz) in D2O at
ambient temperature with acetone as an internal refer-
ence standard (dH 2.225; dC 31.45). One-dimensional
1H, NOE and 13C spectra and phase-sensitive two-
earlier.15 The utility of these substrates for the screening
of bacterial strains for the presence of endo-b-1,3-1,4-
D-
glucanase activity is directly indicated, since protection
of the non-reducing end of the substrate discriminates
against b-glucosidase activity. It has previously been
demonstrated that fluorescent and chromogenic sub-
strates significantly improve the sensitivity of microtiter
plate-based screening assays of glycosyl hydrolases.27,28
The favorable kinetic parameters observed for the
hydrolysis of protected MeUmb-b-laminarabioside and
dimensional spectra (COSY-DQF, NOESY and 1Hꢁ13
C
/
heteronuclear correlations) were recorded using stan-
dard pulse programs at 20 and 50 8C. Data were
analysed using the XWINNMR software package (Bruker,
Germany). The assignment of individual sugar residues
was based on two-dimensional phase-sensitive COSY
data. Monosaccharide sequence assignments were based
on NOE cross-peaks between the anomeric proton and
proton at the substitution position in NOESY experi-
MeUmb-b-laminaratrioside by R. marinus b-1,3-1,4-
D-
glucanase supposes a successful application of these
substrates in such studies including detection of b-
glucanase in the presence of b-glucosidase.
ments using a mixing time of 180 ms. Assignment of 13
signals was based on the protonꢁcarbon correlations
C
/

1464
R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ1467
/
observed by 1Hꢁ13
/
C correlation spectroscopy. Determi-
All steps were carried out at 4 8C. Cells were separated
by centrifugation (3000g, 30 min), washed twice with 10
nation of the anomeric configuration of each mono-
saccharide residue was based on the observed chemical
shifts of the anomeric (C-1) protons in one-dimensional
1H spectra and the J1,2 coupling constants.
Positive-ion mass spectra were recorded on a Micro-
mass Q-TOF2 orthogonal acceleration quadrupole/
time-of-flight mass spectrometer (Micromass, Manche-
ster, UK). With the TOF analyser operating in single-
reflectron ‘V’ mode, typical peak resolution on singly-
mM TrisꢁHCl, pH 8.0, 5 mM EDTA, 0.02% NaN3.
/
Cells were re-suspended in a minimal volume of the
same buffer containing 1 mM DTT, 0.2 M KCl, and 1
mg/mL of lysozyme (Sigma Chemical Co., USA). The
mixture was then left at ꢂ
(6ꢃ20 s) and centrifuged (30,000g, 20 min). The
supernatant was adjusted to a pH value of 5.5 by
addition of 2 N HCl, heated for 20 min at ꢂ60 8C and
/
4 8C for 45 min, sonicated
/
/
charged sodiated carbohydrate adducts was 10,000ꢁ
12,000 FWHM in TOF MS mode. TOF MS mass
calibration was obtained over the m/z range 130ꢁ1980
using a solution of NaI (2 g/L) and CsI (0.05 g/L) in 1:1
PriOHꢁ
water. Solutions of MeUmb glucooligosacchar-
/
centrifuged (30,000g, 20 min). The resulting supernatant
was dialysed against 20 mM AcONa buffer, pH 4.5,
containing 5 mM CaCl2 (buffer A). The crude b-1,3-1,4-
glucanase preparation was loaded on a Toypearl SP-
/
/
650M column (10ꢃ100 mm) equilibrated with buffer A.
/
ides were infused into the mass spectrometer through a
nanoflow ion source at a concentration of 0.03 mg/mL
Bound protein was eluted with buffer A containing 0.25
M NaCl. Fractions containing laminaranase activity
were pooled, concentrated to 15 mL on an Amicon PM-
30 membrane, dialysed against 50 mM potassium
phosphate buffer, pH 6.0 and loaded onto a Sephadex
in 1:1 MeOHꢁwater containing 0.5 mM NaCl by a
/
syringe pump at 500 nL/min. During the analysis of
benzylidene-protected MeUmb glucooligosaccharides
(0.002ꢁ
/
0.005 mg/mL, in 1:1 MeOHꢁwater containing
/
G-50 (Fine) column (22ꢃ960 mm) equilibrated with the
/
0.5 mM NaCl, flow rate 5 mL/min) an electrospray ion
source (desolvation temp. 150 8C) was mounted on the
instrument. The nanospray source capillary voltage
same buffer. Fractions containing b-1,3-1,4-glucanase
activity were pooled, dialyzed against deionized water
and lyophilized.
Endo-b-1,3-glucanase (laminaranase) was isolated
from the culture filtrate of Oerskovia sp. Following
cultivation under conditions previously described,30 cells
were removed by centrifugation (3000g, 20 min). The
supernatant was then simultaneously concentrated 20-
optimized at 3.0ꢁ3.5 kV and the electrospray source
/
voltage was maintained at 3.0 kV. The choice of source
had no significant effect on the ionization of the analytes
studied. The cone voltage was varied during TOF MS
acquisition to optimise the intensity of the (Mꢂ
ion signal (60ꢁ90 V) or to increase the extent of in-
source fragmentation of the (Mꢂ 170
Na)ꢂ ion (150ꢁ
3.6ꢃ
/
Na)ꢂ
/
fold and transferred to 20 mM TrisꢁHCl, pH 7.5 (buffer
/
/
/
B) using of hollow fibers with a nominal molecular
weight limit of 25 kDa (‘Kirishi’, Kirishi, Russia). The
resulting protein solution was loaded onto a DEAE
V). Argon was present in the collision cell (3.4ꢁ
/
/
10ꢄ5 mbar analyser Penning gauge reading) during both
MS and MS/MS experiments. For MS/MS experiments,
Sephadex A-50 column (100ꢃ50 mm) equilibrated with
/
the collision energy was varied (40ꢁ
the resilience of the (Mꢂ
optimal balance between parent and fragment ion
intensities. A scan time of 2.4 or 2.5 s with an interscan
delay of 0.1 s was used in all MS modes. Data were
collected until an acceptable signal-to-noise ratio was
achieved after the combination of individual spectra,
/
60 V), depending on
Na)ꢂ ion, to achieve an
the same buffer. Bound protein was eluted with 1 M
NaCl in buffer B. Fractions containing b-1,3-glucanase
activity were pooled and the activity was precipitated by
adding (NH4)2SO4 to 1.7 M. The precipitate was
separated by centrifugation (3000g, 20 min) and the
supernatant was dialyzed sequentially against 20 mM
AcONa, pH 4.0, and H2O prior to lyophilization. The
enzyme was obtained in 50% yield and had a specific b-
/
typically 5ꢁ20. To provide additional mass accuracy for
/
the analysis of MS/MS spectra, the calculated mono-
isotopic mass for the parent ion was used as a ‘lock
mass’ to recalibrate these spectra after the combination
of individual spectra.
1,3-D-glucanase activity of 120 U/mg.
Marine mollusc S. sachalinensis was kindly provided
by Professor V.G. Tarasov (Institute of Bioorganic
Chemistry, Far East Division of Russian Academy of
Sciences) and then endo-b-1,3-glucanase was partially
purified. Mollusc material was homogenised for 20 min
at 4 8C in 3 vol of 20 mM AcONa buffer (pH 5.2),
soluble fraction was removed by centrifugation (3000g
for 20 min) and the supernatant was desalted on a
Sephadex G-25 (fine) column equilibrated in the same
buffer. Fractions with the laminaranase active enzyme
(approx 100 mL) were pooled, applied onto a SP-
Sephadex column, and eluted with 1 M NaCl in 20
3.3. Purification of enzymes
Recombinant plasmids pMKM1 derived from the wild-
type strain of R. marinus16 were used in this study.
Culture of Escherechia coli DH5a used as a host strain
for the plasmids was grown in LuriaꢁBertani (LB)
/
medium with addition of 150 mg/mL of ampicillin at
37 8C for 7 h with shaking.15 Recombinant laminar-
anase was purified according to the following scheme.

R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ
/
1467
1465
mM AcONa buffer, pH 5.2. The active fraction was
then used in further experiments.
Enzyme activity during the purification process was
mL). The Michaelis constant was determined for each
substrate from the MichaelisꢁMenten equation by non-
linear regression analysis.32
/
measured according to the Somogyiꢁ
/
Nelson method23
using laminaran as a substrate.
3.7. Investigation of transglycosylation activity of b-1,3-
-glucanases from Oerskovia and S. sachalinensis
D
3.4. Enzyme assays
The transglycosylation activity of b-1,3-D-glucanases
toward laminaran and curdlan was investigated at
concentrations from 5 to 100 mg/mL in 20 mM AcONa
b-1,3-1,4-
anase) activities were determined by measuring the
amount of reducing sugar released from laminaran.
D
-Glucanase and b-1,3-
D
-glucanase (laminar-
or sodium phosphate buffer over the pH range 3.5ꢁ8.5.
/
Standard assays for b-1,3-1,4-D-glucanase activity from
Reaction products were analysed on a reverse-phase
Waters Hypersil ODS column using UV detection at 254
nm. Product yields were calculated by integration of the
corresponding chromatographic peaks.
R. marinus were carried out in 50 mM AcONa buffer,
pH 6.0, (0.1 mL) at 75 8C for 10 min. One unit of the
enzyme activity was defined as the amount of the
enzyme required to produce 1 mmol of reducing sugar
per minute from laminaran in the above conditions. b-
Glucosidase activity was measured using p-nitrophenyl
3.8. Synthesis of 4-methylumbelliferyl b-
laminaraoligosaccharides with d.p. 2
/4
ꢁ
D-glucanase from Oerskovia were incubated in
b-
phosphate buffer, pH 6.0, at 37 8C.31 Other endo-
glycosidase activities were analysed by the Somogyiꢁ
D-glucopyranoside as a substrate in 50 mM sodium
Curdlan (500 mg), MeUmbGlcp (100 mg) and 20ꢁ
of b-1,3-
/
30 U
/
Nelson method using appropriate polysaccharide sub-
strates.
20 mM AcONa buffer, pH 4.8, (5 mL) at 37 8C for 48 h.
The extent of the reaction was tracked by TLC and
reverse-phase HPLC. Products of the transglycosylation
reaction were sequentially isolated on a Bio-Gel P2 extra
3.5. Mode of action of b-1,3-1,4-
D-glucanase
fine (Bio Rad, USA) column (10ꢃ
mL/h) in H2O followed by an INERTSIL PREP-ODS
column (20ꢃ250 mm) using a linear gradient (0ꢁ100%)
/
1200 mm, flow rate 9
The mode of action of the b-1,3-1,4-
D
-glucanase was
investigated during the hydrolysis of MeUmbGnꢀ2ꢁ5 in
20 mM AcONa buffer, pH 4.0, at 37 8C, using substrate
concentrations in the range from 0.8 to 1.2 mM over
different time intervals and following the reaction by
TLC. The reaction was terminated by freezing and
lyophilisation. Analysis of the hydrolysis products was
/
/
of MeCN in water.
3.9. 4-Methylumbelliferyl b-laminarabioside
[a]D
3.10. 4-Methylumbelliferyl b-laminaratrioside
[a]D 44.28 (c 0.335, water).
3.11. 4-Methylumbelliferyl b-laminaratetraoside
[a]D 26.38 (c 0.335, water).
ꢄ
/
60.38 (c 0.335, water).
performed on a Waters Hypersil ODS column (200ꢃ
mm) using linear gradient (0ꢁ90%) of MeCN in water
with spectrophotometric detection at 254 nm.
/
4.8
/
3.6. Determination of kinetic characteristics of b-1,3-1,4-
-glucanase hydrolysis of MeUmb-b-
laminaraoligosaccharides
ꢄ
/
D
The activity of the b-1,3-1,4-D-glucanase in the hydro-
ꢄ
/
lysis of MeUmbG2, MeUmbG3, and acetal-protected
MeUmbG3 was evaluated spectrofluorometrically after
incubation of the reaction mixture in 50 mM AcONa
buffer, pH 6.0, by measuring released MeUmb in
accordance with Ref. 29. To reduce the rate of non-
enzymatic hydrolysis of the substrate, the reaction was
carried out at 37 8C.
3.12. Synthesis of 4-methylumbelliferyl b-
laminaraoligosaccharides with d.p 5 and 6
Laminaran (200 mg), MeUmbGlcp (100 mg) and
mollusc b-1,3- -glucanase (20 U) were incubated in 50
D
mM AcONa buffer, pH 4.8, (5 mL) for 48 h at 37 8C.
Products of the reaction were isolated as described
above for the di-, tri-, and tetrasaccharide analogues.
The Michaelisꢁ/Menten parameters Km and Vmax were
determined from the Lineweaverꢁ/Burk plots obtained
by measuring the initial rate of hydrolysis of acetal-
protected and non-protected MeUmb-b-laminaraoligo-
saccharides. The purified b-1,3-1,4-
D
-glucanase (0.009ꢁ
/
3.13. 4-Methylumbelliferyl b-laminarapentaoside
0.010 U) was incubated with MeUmb-b-laminara-oli-
gosaccharides in 50 mM AcONa buffer, pH 5.5 (0.1
[a]D
ꢄ21.08 (c 0.335, water).
/

1466
R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ1467
/
3.14. 4-Methylumbelliferyl b-laminarahexaoside
[a]D 18.48 (c 0.335, water).
Acknowledgements
ꢄ
/
The present work was supported in part by a grant (03-
04-48756-a) from Russian Foundation for Basic Re-
search and by a grant of Program for Basic Research
3.15. Synthesis of 4-methylumbelliferyl v-4,6-O-
benzylidene-b-laminaraoligosaccharides with d.p 3
‘PhysicalꢁChemical Biology’ from the Presidium of
/
/
5
ꢁ
Russian Academy of Sciences. The Wallenberg Con-
sortium North for Functional Genomics is thanked for
financial support for the purchase of mass spectrometry
equipment at the Royal Institute of Technology. K.N.
thanks Dr Anna A. Kulminskaya for her skillful
assistance in preparation the material for publication.
Derivatisation of MeUmbG3, MeUmbG4, and
MeUmbG5 was performed as follows. A mixture of 3
mmol MeUmbGn, 10 mmol (1.5 mL) of a,a-dimethox-
ytoluene, and 50 mg of p-toluenesulfonic acid mono-
hydrate in dry N,N-dimethylformamide (10 mL) was
incubated overnight at 37 8C.33 The resulting syrup was
fractionated on an INERTSIL PREP-ODS column
using a linear gradient (0ꢁ100%) of MeCN in water.
/
References
3.16. 4-Methylumbelliferyl v-4,6-O-benzylidene-b-
laminaratrioside
1. Honda, Y.; Tanimori, S.; Kirihata, M.; Kaneko, S.;
Tokuyasu, K.; Hashimoto, M.; Watanabe, T.; Fukamizo,
NMR 1H d: 5.283 (1 H, JH1,H2 7.93 Hz, H-1-1), 4.923 (1
H, JH1,H2 7.86 Hz, H-1-3), 4.868 (1 H, JH1,H2 7.86 Hz,
H-1-2), 4.384 (1 H, JH5,H6a 4.85, JH6a,H6b 10.5 Hz, H-6a-
3), 3.982 (1 H, JH5,H6a 2.36, JH6a,H6b 12.4 Hz, H-6a-1),
3.976 (1 H, JH5,H6a 1.8, JH6a,H6b 11.9 Hz, H-6a-2), 3.934
(1 H, JH2,H3 9.2, JH3,H4 8.8 Hz, H-3-1), 3.892 (1 H,
JH5,H6b 9.8 Hz, H-6b-3), 3.838 (1 H, H-2-1), 3.824 (1 H,
JH2,H3 9.2, JH3,H4 8.9 Hz, H-3-3), 3.818 (1 H, JH5,H6b 5.5
Hz, H-6b-1), 3.814 (1 H, JH3,H4 8.9, JH4,H5 9.5 Hz, H-4-
2), 3.765 (1 H, JH4,H5 9.5 Hz, H-5-1), 3.743 (1 H, JH5,H6b
4.8 Hz, H-6b-2), 3.684 (1 H, JH4,H5 9.2 Hz, H-4-3), 3.659
(1 H, H-4-1), 3.640 (1 H, JH2,H3 9.3 Hz, H-3-2), 3.633 (1
H, H-5-3), 3.632 (1 H, H-2-2), 3.557 (1 H, H-5-2), 3.521
T. FEBS Lett. 2000, 476, 194ꢁ
2. De La Mata, I.; Estrada, P.; Macarron, R.; Dominguez, J.
M. Biochem. J. 1992, 283, 679ꢁ782.
3. Mo, H.; Winter, H. C.; Goldstein, I. J. J. Biol. Chem.
2000, 275, 10623ꢁ10629.
/
197.
/
/
4. Lemieux, R. U. In Methods in Carbohydrate Chemistry;
Whistler, R. L.; Wolfrom, M. L., Eds.; Vol. II; Academic
Press: New York, 1963; pp 22ꢁ
5. De Boeck, H.; Matta, K. L.; Claeyssens, M.; Sharon, N.;
Loontiens, F. G. Eur. J. Biochem. 1983, 131, 453ꢁ460.
6. Kandra, L.; Gye´ma´nt, G.; Farkas, E.; Lipta´k, A. Carbo-
hydr. Res. 1997, 298, 237ꢁ242.
7. Takeo, K.; Ohguchi, Y.; Hasegawa, R.; Kitamura, S.
Carbohydr. Res. 1995, 277, 231ꢁ244.
8. Malet, C.; Validot, J. L.; Ochoa, A.; Ga´llego, B.; Brosa,
C.; Planas, A. Carbohydr. Res. 1995, 274, 285ꢁ301.
9. Zeng, X.; Yoshino, R.; Murata, T.; Ajisaka, K.; Usui, T.
Carbohydr. Res. 2000, 325, 120ꢁ131.
/27.
/
/
/
(1 H, H-2-3). [a]D
ꢄ23.28 (c 0.15, water).
/
/
3.17. 4-Methylumbelliferyl v-4,6-O-benzylidine-b-
laminaratetraoside
/
10. Eneyskaya, E. V.; Brumer III, H.; Backinowsky, L. V.;
Ivanen, D. R.; Kulminskaya, A. A.; Shabalin, K. A.;
[a]D
ꢄ14.28 (c 0.15, water).
/
Neustroev, K. N. Carbohydr. Res., 2002, 338, 313ꢁ325.
/
11. Eneyskaya, E. V.; Golubev, A. M.; Kachurin, A. M.;
Savel’ev, A. N.; Neustroev, K. N. Carbohydr. Res. 1998,
3.18. 4-Methylumbelliferyl v-4,6-O-benzylidene-b-
laminarapentaoside
305, 83ꢁ
12. Williams, S. J.; Withers, S. G. Australian J. Chem. 2002,
55, 3ꢁ12.
13. Jakeman, D. L.; Withers, S. G. Trends in Glycosci.
Glycotechnol. 2002, 14, 13ꢁ25.
14. Okuyama, M.; Mori, H.; Watanabe, K.; Kimura, A.;
Chiba, S. Biosci. Biotechnol. Biochem. 2002, 66, 928ꢁ933.
15. Petersen, B. O.; Krah, M.; Duus, J. Ø; Tomsen, K. K. Eur.
J. Biochem. 2000, 267, 361ꢁ369.
16. Krah, M.; Misselwitz, R.; Politz, O.; Tomsen, K. K.;
Welfle, U.; Borriss, R. Eur. J. Biochem. 1998, 257, 101ꢁ
111.
/
91.
/
[a]D
ꢄ9.58 (c 0.15, water).
/
/
3.19. Synthesis of 4-methylumbelliferyl v-4,6-O-
ethylidine-b-laminarabioside
/
/
Synthesis of 4,6-O-ethylidene-modified MeUmb b-lami-
narabioside was performed as described above for a
benzylidene-protected analogue.
/
17. Malet, C.; Planas, A. Biochemistry 1997, 36, 13838ꢁ
13848.
18. Zvyagintseva, T. N.; Makar’eva, T. N.; Ermakova, S. P.;
/
3.20. 4-Methylumbelliferyl v-4,6-O-ethylidine-b-
laminarabioside
Elyakova, L. A. Russian J. Bioorg. Chem. 1998, 24, 219ꢁ
/
[a]D
ꢄ21.48 (c 0.15, water).
/
223.

R. Borriss et al. / Carbohydrate Research 338 (2003) 1455ꢁ
/
1467
1467
19. Wong, C. H.; Whitesides, G. M.; Baldwin, J. E.; Magnus,
P. D. Enzymes in Synthetic Organic Chemistry, Chapter 5,
Tetrahedron Organic Chemistry Series, Pergamon, Ox-
26. Abel, M.; Planas, A.; Christensen, U. Biochem. J. 2001,
357, 195ꢁ202.
27. Kaneko, S.; Kitaoka, M.; Kuno, A.; Hayashi, K. Biosci.
Biotechnol. Biochem. 2000, 64, 741ꢁ745.
28. Margolles-Clark, E.; Tenkanen, M.; Luonteri, E.; Penttila,
/
ford, 1994, pp. 252ꢁ
20. Stowell, C. P.; Lee, Y. C. Adv. Carbohydr. Chem. Biochem.
1980, 37, 225ꢁ281.
/311.
/
¨
/
M. Eur. J. Biochem. 1996, 240, 104ꢁ
29. Lowry, O. H.; Rosenbrough, N. J.; Farr, A. L.; Randall,
R. J. J. Biol. Chem. 1951, 193, 265ꢁ275.
/111.
21. Kulminskaya, A. A.; Thomsen, K. K.; Shabalin, K. A.;
Sidorenko, I. A.; Eneyskaya, E. V.; Savel’ev, A. N.;
/
Neustroev, K. N. Eur. J. Biochem. 2001, 268, 6123ꢁ
22. Ogawa, K.; Tsuruggi, J.; Watanabe, T. Carbohydr. Res.
1973, 29, 397ꢁ403.
23. Somogyi, M. J. Biol. Chem. 1952, 195, 19ꢁ
24. Høj, P. B.; Rodriguez, E. B.; Stick, R. V.; Stone, B. A. J.
Biol. Chem. 1989, 264, 4939ꢁ4947.
25. Wood, P. J. Carbohydr. Res. 1981, 94, 35ꢁ
/
6131.
30. Bagiyan, F. G.; Eneyskaya, E. V.; Kulminskaya, A. A.;
Savel’ev, A. N.; Shabalin, K. A.; Neustroev, K. N. Eur. J.
/
Biochem. 1997, 249, 286ꢁ
31. Bahl, O. P.; Agrawal, K. M. J. Biol. Chem. 1969, 244,
2970ꢁ2978.
32. Perella, F. Anal. Biochem. 1988, 174, 437ꢁ
33. Takeo, K.; Shinmitsu, K. Carbohydr. Res. 1984, 135ꢁ145.
/292.
/
23.
/
/
/
447.
/
39.
/