Báo cáo khoa học: Kinetic studies on endo-b-galactosidase by a novel colorimetric assay and synthesis of N -acetyllactosamine-repeating oligosaccharide b-glycosides using its transglycosylation activity pptx

Kinetic studies on endo-b-galactosidase by a novel colorimetric assay
and synthesis of
N
-acetyllactosamine-repeating oligosaccharide
b-glycosides using its transglycosylation activity
Takeomi Murata, Takeshi Hattori, Satoshi Amarume, Akiko Koichi and Taichi Usui
Department of Applied Biological Chemistry, Shizuoka University, Japan
Novel chromogenic substrates for endo-b-galactosidase
were designed on the basis of the structural features of
keratan sulfate. Galb1-4GlcNAcb1-3Galb1-4GlcNAcb-
pNP (2), which consists of two repeating units of N-acetyl-
lactosamine, was synthesized enzymatically by consecutive
additions of GlcNAc and Gal residues to p-nitrophenyl
b-N-acetyllactosaminide. In a similar manner, Glc-
NAcb1-3Galb1-4GlcNAcb-pNP (1), GlcNAcb1-3Galb1-
4Glcb-pNP (3), Galb1-4GlcNAcb1-3Galb1-4Glcb-pNP (4),
Galb1-3GlcNAcb1-3Galb1-4Glcb-pNP (5), and Galb1-
6GlcNAcb1-3Galb1-4Glcb-pNP (6) were synthesized as
analogues of 2. Endo-b-galactosidases released GlcNAcb-
pNP or Glcb-pNP in an endo-manner from each substrate.
A colorimetric assay for endo-b-galactosidase was devel-
oped using the synthetic substrates on the basis of the
determination of p-nitrophenol liberated from GlcNAcb-
pNP or Glcb-pNP formed by the enzyme through a coupled
reaction involving b-N-acetylhexosaminidase (b-NAHase)
or b-
D
-glucosidase. Kinetic analysis by this method showed
that the value of V
max
/K

m
of 2 for Escherichia freundii endo-
b-galactosidase was 1.7-times higher than that for keratan
sulfate, indicating that 2 is very suitable as a sensitive sub-
strate for analytical use in an endo-b-galactosidase assay.
Compound 1 still acts as a fairly good substrate despite the
absence of a Gal group in the terminal position. In addition,
the hydrolytic action of the enzyme toward 2 was shown to
be remarkably promoted compared to that of 4 by the
presence of a 2-acetamide group adjacent to the p-nitro-
phenyl group. This was the same in the case of a comparison
of 1 and 3. Furthermore, the enzyme also catalysed a
transglycosylation on 1 and converted it into GlcNAc
b1-3Galb1-4GlcNAcb1-3Galb1-4GlcNAcb-pNP (9)and
GlcNAcb1-3Galb1-4GlcNAcb1-3Galb1-4GlcNAcb1-3Galb1-
4GlcNAcb-pNP (10) as the major products, which have
N-acetyllactosamine repeating units.
Keywords: endo-b-galactosidase; enzyme assay; kinetics;
poly (N-acetyllactosamine); transglycosylation.
Endo-b-galactosidases were discovered as keratan sulfate-
degrading enzymes, so-called keratanases, in culture filtrates
of Escherichia freundii (glycoside hydrolase family 16) [1],
Coccobacillus sp. [2], Pseudomonas sp. [3], Flavobacterium
keratolyticus (glycoside hydrolase family 16) [4,5], and
Bacteroides fragilis [6]. E. freundii keratanase was found to
have hydrolysing activity for a wide range of nonsulfated
oligosaccharides isolated from human milk and carbohy-
drate moieties of glycoproteins and glycolipids [7–10]. The
use of endo-b-galactosidase has been expanded to detection
of poly (N-acetyllactosamine) chains in a variety of complex

glycoconjugates in addition to keratan sulfate. Bacteroides
fragilis endo-b-galactosidase has properties similar to those
of E. freundii endo-b-galactosidase [11–13]. Therefore, the
endo-b-galactosidases from E. freundii and B. fragilis have
been widely used as tools for structural and functional
analyses of glycans involved in glycoconjugates.
An assay using keratan sulfate as a substrate has been
widely used for estimation of endo-b-galactosidase activity.
However, this method is not always reproducible because
of lack of uniformity of the polymer. Methods using
low molecular mass substrates have been preferred and
recommended for accurate determination of activities of
endoglycosidases such as a-amylase [14], lysozyme [15],
and endo-b-N-acetylglucosaminidase [16], because the
purity of the substrate and the reaction pattern can be
determined exactly. This led us to develop a substrate
suitable for use in the analysis of endo-b-galactosidase. A
series of chromogenic substances having a partially substi-
tuted unit of poly (N-acetyllactosamine) were designed as
substrate analogues for this enzyme because systematic
kinetic studies on the structural modification of substrates
would be helpful in revealing the requirements for binding
and catalytic specificity. In general, glycosidase can cause
transglycosylation as well as hydrolysis as a reverse
reaction [17–19]. Transglycosylation of endo-type glycosi-
dase is now used for the synthesis of N-acetyllactosamine-
repeating oligosaccharide.
Correspondence to T. Murata, Department of Applied Biological
Chemistry, Shizuoka University, 836 Ohya, Shizuoka, 422-8529,
Japan. Fax and Tel.: +81 54 238 4872,

E-mail:
Abbreviations: pNP, p-nitrophenyl; b-NAHase, b-N-acetylhexos-
aminidase; b4GalT, b-1,4-galactosyltransferase; b3GnT, b-1,3-N-
acetylglucosaminyltransferase; HPAEC-PAD, high-performance
anion exchange chromatography-pulsed amperometric detection.
Enzymes: endo-b-galactosidase (EC 3.2.1.103); b-
D
-galactosidase
(EC 3.2.1.23); b-1,4-galactosyltransferase (EC 2.4.1.22); b-1,3-N-
acetylglucosaminyltransferase (EC 2.4.1.149); b-D-glucosidase
(EC 3.2.1.21); b-N-acetylhexosaminidase (EC 3.2.1.52).
(Received 20 December 2002, revised 11 July 2003,
accepted 17 July 2003)
Eur. J. Biochem. 270, 3709–3719 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03757.x
In this paper, we describe the enzymatic synthesis of a
novel substrate 2 and its analogues for use a colorimetric
assay of endo-b-galactosidase activity and the usefulness of
the resulting chromogenic substrates for kinetic studies on
the enzyme. In the latter part of this paper, synthesis of
N-acetyllactosamine-repeating oligosaccharide b-glycoside
utilizing endo-b-galactosidase-mediated transglycosylation
is described.
Materials and methods
Materials
Endo-b-galactosidases from E. freundii and B. fragilis were
from Seikagaku Corporation (Tokyo, Japan) and Wako
Pure Chemical Industries, Ltd. (Osaka, Japan), respectively.
b-
D
-Galactosidase from Bacillus circulans ATCC31382 [20]

was a kind gift from Meiji Milk Products Co. Ltd. (Tokyo,
Japan). b-
D
-Glucosidase from almonds was from Sigma
Chemicals. b-N-acetylhexosaminidase (b-NAHase) from
Amycolatopsis orientalis IFO 12806T was purified by 80%
saturated ammonium sulfate precipitation followed by
GlcNAc-cellulofine affinity chromatography. Bovine milk
b-1,4-galactosyltransferase (b4GalT) was from Calbiochem
(CA, USA). Crude b-1,3-N-acetylglucosaminyltransferase
(b3GnT) from bovine serum was prepared as follows.
Bovine serum was brought to 80% saturation with solid
ammonium sulfate and left standing overnight at 4 °C. The
precipitate was collected by centrifugation at 5000 g for
30 min, dissolved in 50 m
M
Tris/HCl buffer (pH 8.0), and
dialysed against distilled water overnight at 4 °C. The
enzyme solution was lyophilized and then used for synthesis
of oligosaccharides without further purification. The crude
enzyme preparation catalysed the transfer of a GlcNAc
residue from UDP-GlcNAc to the OH-3¢ positions of
Galb1-4Glc, Galb1-4GlcNAc, Galb1-4Glcb-pNP and
Galb1-4GlcNAcb-pNP. The specific activity for Galb1-
4GlcNAcb-pNP as an acceptor substrate was 69 lUÆmg
)1
.
Galb1-4GlcNAc, GlcNAcb1-3Galb1-4GlcNAc, Galb1-
4Glcb-pNP, Galb1-4GlcNAcb-pNP, and GlcNAcb1-
6Galb1-4Glcb-pNP were synthesized by our previously

described methods [21–24]. UDP-GlcNAc and UDP-Gal
were kind gifts from Yamasa Corporation (Choshi,
Japan). All other chemicals were obtained from commercial
sources.
Enzyme assay
b-
D
-Galactosidase, b-
D
-glucosidase and b-NAHase activit-
ies were assayed as follows. A mixture containing 2 m
M
substrate solution (Galb-pNP, Glcb-pNP, and GlcNAcb-
pNP) in 0.4 mL 50 m
M
sodium phosphate buffer pH 6.0
and an appropriate amount of enzyme in a total volume of
0.1 mL was incubated for 10 min at 40 °C. One hundred
microlitres of the reaction mixture were then added to
0.1 mL 1.0
M
Na
2
CO
3
on a microplate at 2-min intervals
during to stop the reaction, and the amount of liberated
p-nitrophenol was determined by measuring absorbance at
405 nm using a microplate reader (Biolumin 960, Amer-
sham Pharmacia). One unit of enzyme was defined as the

amount releasing 1 lmol p-nitrophenolÆmin
)1
.Theb3GnT
assay was carried out as follows. Galb1-4GlcNAcb-pNP
(5 mg) and UDP-GlcNAc (3.6 mg) were dissolved in
0.5mL50m
M
Tris/HCl buffer pH 8.0 containing 1.6 mg
MnCl
2
and 0.5 mg ATP, followed by addition of an
appropriate amount of b3GnT preparation. The reaction
mixture was incubated at 37 °C for 96 h. Fifty microlitres of
the reaction mixture was taken out at 24-h intervals during
the reaction and boiled for 5 min. Resulting GlcNAcb1-
3Galb1-4GlcNAcb-pNP was measured by HPLC as des-
cribed in the Analytical methods.
Analytical methods
HPLC was carried out using a Mightysil RP18ODS column
(4.6 · 150 mm, Kanto Chemical Co. Ltd, Tokyo, Japan) in
a Hitachi 6000-series liquid chromatograph with an L-4000
ultraviolet detector (absorbance at 300 nm). Elution of the
column was performed with H
2
O/CH
3
OH (95 : 5, v/v). The
flow rate was 1.0 mLÆmin
)1
at 40 °C. HPAEC-PAD

analysis was conducted on a DX-300 Bio-LC system
equipped with a pulsed amperometric detector (Dionex,
Sunnyvale, USA). Oligosaccharides were analysed by a
CarboPac P-1 column (Dionex, 4 · 250 mm) at a flow rate
of 1 mLÆmin
)1
at room temperature. The elution was
effected with 100 m
M
NaOH for 40 min. For NMR
analysis, an appropriate oligosaccharide sample was dis-
solved in 200 lLofD
2
O and filtrated with a Millipore filter
(0.22 lm) and then put into a sample tube (i.d. 3 mm).
1
H- and
13
C-NMR spectra were recorded on a JEOL
JNM-LA 500 spectrometer at 25 °C. Chemical shifts are
expressed in d relative to sodium 3-(trimethylsilyl) propio-
nate as an external standard. ESI-MS analysis was carried
out in the positive-ion mode on a JEOL MS-700 (JEOL
Ltd, Akishima, Japan) using 2,5-dihydroxy benzoic acid as
the matrix. Determination of the amount of protein was
carried out using a Bio-Rad protein assay kit. Determin-
ation of total carbohydrate was carried out as follows. One
hundred microliters of sample was put into a test tube
(1 · 10 cm), and 100 lL of 5% (w/v) phenol and 0.5 mL
concentrated sulfuric acid were immediately added. The

sample mixture was vortexed and then kept for 20 min at
room temperature, and absorbance was read at 490 nm.
Preparation of GlcNAcb1-3Galb1-4GlcNAcb-
p
NP (1)
and GlcNAcb1-3Galb1-4GlcNAcb-
p
NP (3)
Galb1-4GlcNAcb-pNP (504 mg, 1 mmol) and UDP-Glc-
NAc (651 mg, 1 mmol) were dissolved in 25 mL 50 m
M
Tris/HCl buffer pH 8.0 containing 99 mg MnCl
2
, followed
by addition of 850 mU of crude b3GnT preparation
from bovine serum. The mixture was incubated for 172 h
at 37 °C, and the reaction was terminated by boiling for
5 min. The precipitate was removed by centrifugation
(8 000 g, 20 min), and the supernatant was loaded onto a
Toyopearl HW-40S column (5 · 100 cm) equilibrated with
25% methanol at a flow rate of 2 mLÆmin
)1
.Afterthe
chromatography, the eluate was monitored by measuring
the absorbance at 300 nm (p-nitrophenyl group) and at
490 nm (phenol-sulfuric acid method) by a spectrometer.
Both chromatograms showed two peaks (F-1, tubes 24–33;
F-2, tubes 49–60). F-1 was combined, concentrated, and
lyophilized to produce 1 (24.1 mg) in a 3.4% total
yield based on the acceptor added. F-2 was recovered as

3710 T. Murata et al. (Eur. J. Biochem. 270) Ó FEBS 2003
Galb1-4GlcNAcb-pNP (0.35 g). In the same way, com-
pound 3 was prepared from Galb1-4Glcb-pNP (0.5 g) and
UDP-GlcNAc (450 mg) by use of bovine serum b3GnT in
a 4% total yield based on the acceptor added.
1
Hand
13
C-NMR data of 1 and 3 were almost identical to data
reported previously [24].
Preparation of GlcNAcb1-3Galb1-4GlcNAcb-
p
NP (2)
and GlcNAcb1-3Galb1-4GlcNAcb-
p
NP (4)
Compound 1 (24.1 mg, 34.1 lmol) and UDP-Gal (42.6 mg,
68.2 lmol) were dissolved in 3.4 mL 100 m
M
sodium
cacodylate buffer (pH 6.8) containing 67.3 mg MnCl
2
followed by addition of 0.2 U b4GalT from bovine milk.
The mixture was incubated for 6 h at 37 °C and separated
by a Toyopearl HW-40S column (2.5 · 80 cm) as described
above. Compound 2 was obtained in an 82% total yield
(24.3 mg) based on the acceptor added. In the same
way, compound 4 was obtained in a 71% total yield
(13.2 mg) based on the acceptor from 3 and UDP-Gal.
1

Hand
13
C-NMR data of 2 and 4 are summarized in
Table 1.
Preparation of GlcNAcb1-3Galb1-4GlcNAcb-
p
NP (5) and
its positional isomer GlcNAcb1-3Galb1-4GlcNAcb-
p
NP (6)
Compound 3 (85 mg, 128 lmol) and o-nitrophenyl b-
D
-
galactopyranoside (Galb-oNP, 238 mg, 790 lmol) were
dissolved in 5.8 mL 40 m
M
sodium acetate buffer pH 5.5
followed by addition of 98 mU of b-
D
-galactosidase from
Bacillus circulans ATCC31382. The mixture was incubated
for 10 h at 50 °C and was loaded onto an ODS DM1020T
column (5 · 100 cm) equilibrated with 5% methanol at
a flow rate of 5 mLÆmin
)1
in order to remove the
o-nitrophenol liberated during the reaction. The fractions
showing absorbance at 300 nm were concentrated and
loaded onto a Toyopearl HW-40S column as above. The
chromatogram showed two peaks (F-1, 1340–1520 mL;

F-2, 1640–1880 mL). F-2 contained 3 (43 mg) used as the
acceptor substrate. F-1 was further separated by a Shodex
Asahipak NH2P-50 column (21.5 · 300 mm) equilibrated
with 80% acetonitrile at a flow rate of 5 mLÆmin
)1
at 40 °C.
Eluate was monitored on-line by measuring the absorbance
at 300 nm. The chromatogram showed two peaks (F-1a,
327–390 mL; F-1b, 396–450 mL). These peaks were
Table 1.
1
H- and
13
C-chemical shifts of transfer products in D
2
O solution. J
1,2
, coupling constants are given in Hz.
Compounds
Chemical Shifts (d)
C-1 C-2 C-3 C-4 C-5 C-6 NHCOCOCH
3
o-ph m-ph p-ph C-O H-1 J
1,2
CH
3
2 pNP 119.3 128.9 145.5 164.4
GlcNAc 101.3 57.6 74.8 80.9 77.9 62.6 177.7 24.8 5.35 8.5 2.02
Gal 105.7 72.7 84.9 71.1 77.7 63.81 4.51 8.3
GlcNAc 105.63 58.0 75.0 80.8 77.3 62.5 177.7 24.9 4.72 8.3 2.05

Gal 105.58 73.7 75.3 71.3 78.1 63.77 4.49 8.2
4 pNP 119.3 128.9 145.5 164.5
Glc 102.0 75.2 77.9 80.7 76.9 62.7 5.31 7.7
Gal 105.8 72.8 84.9 71.2 77.7 63.7 4.486 7.7
GlcNAc 105.7 58.0 77.9 81.0 77.4 62.6 177.7 25.0 4.72 8.2 2.05
Gal 105.6 73.8 76.9 71.4 78.2 63.8 4.490 7.7
5 pNP 119.2 128.9 145.4 164.4
Glc 102.0 75.2 77.9 80.6 76.8 62.6 5.30 8.0
Gal 105.4 72.8 84.79 71.1 77.7 63.77 4.47 7.9
GlcNAc 105.7 57.5 84.83 71.2 78.0 63.3 177.8 25.0 4.73 8.5 2.02
Gal 106.3 73.5 75.3 71.3 78.1 63.82 4.44 7.7
6 pNP 119.3 128.9 145.4 164.4
Glc 102.0 75.2 77.9 80.6 76.8 62.6 5.30 8.0
Gal 105.74 72.8 84.8 71.1 77.7 63.78 4.47 7.7
GlcNAc 105.71 58.94 76.3 72.3 77.4 71.3 177.8 25.0 4.69 8.6 2.02
Gal 106.36 73.5 75.5 71.4 78.0 63.81 4.44 8.0
7 pNP 118.2 127.3 143.3 164.0
Gal 101.8 72.1 83.4 68.6 77.1 61.7 5.20 6.5
GlcNAc 103.8 57.9 75.4 70.6 78.5 62.6 171.8 24.7 4.78 8.1 2.06
8 pNP 118.2 127.5 143.2 164.2
Gal 102.3 72.2 75.1 70.4 78.4 70.1 5.22 7.6
GlcNAc 103.2 57.1 74.5 71.5 78.7 62.8 170.7 24.8 4.55 8.4 1.78
9 pNP 119.3 128.9 145.5 164.5
GlcNAc 101.3 57.6 74.8 81.0 78.0 62.7 177.7 24.9 5.34 8.3 2.01
Gal 105.8 72.8 84.9 71.2 77.7 63.8 4.49 8.0
GlcNAc 105.67 58.0 75.0 80.9 77.4 62.6 177.7 25.0 4.70 8.2 2.03
Gal 105.58 72.8 84.8 71.2 77.7 63.8 4.46 8.0
GlcNAc 105.71 58.5 76.4 72.5 78.5 63.3 177.8 25.0 4.67 8.6 2.03
Ó FEBS 2003 Kinetic studies on endo-b-galactosidase (Eur. J. Biochem. 270) 3711
combined, concentrated, and lyophilized to produce 5

(5.5 mg) and 6 (7.2 mg) in 5.2 and 6.8% yields based on the
acceptor added, respectively.
1
H- and
13
C-NMR data of 5
and 6 are summarized in Table 1.
Preparation of GlcNAcb1-3Galb1-4GlcNAcb-
p
NP (7)
and GlcNAcb1-3Galb1-4GlcNAcb-
p
NP (8)
Galb-pNP (390 mg, 1.29 mmol) and N, N¢-diacetylchito-
biose (GlcANc
2
, 531 mg, 1.25 mmol) were dissolved in
7.5mL20m
M
sodium acetate buffer pH 5.0 followed by
addition of 8.7 U of A. orientalis b-N-acetylhexosamini-
dase. The mixture was incubated for 100 h at 40 °C, and the
reaction was terminated by boiling for 5 min. The precipi-
tate was removed by centrifugation (8 000 g, 15 min), and
the supernatant was loaded onto a Toyopearl HW-40S
column (5 · 100 cm) as above. Eluate was monitored by
measuring the absorbance at 300 nm (p-nitrophenyl group)
and at 490 nm (phenol-sulfuric acid method). The chroma-
togram showed four peaks (F-1, 135–150 mL; F-2, 190–
225 mL; F-3, 275–295 mL; F-4, 420–470 mL). F-2 and F-3

were combined, concentrated, and lyophilized to produce 8
(33.8 mg) and 7 (8.6 mg), respectively, in a 6.5% total yield
based on the donor added.
1
H- and
13
C-NMR data of these
disaccharides are summarized in Table 1.
Hydrolytic actions of endo-b-galactosidase
on
p
-nitrophenyl b-glycosides
The hydrolytic actions of endo-b-galactosidase on p-nitro-
phenyl oligosaccharide b-glycosides and a reducing oligo-
saccharide listed in Table 2 were investigated by incubating
a mixture (50 lL) containing 1 m
M
of substrates in 50 m
M
sodium acetate buffer pH 5.8 with 1 mU of the enzymes at
37 °C for 20 min. The enzyme hydrolysates were analysed
by HPLC or HPAEC-PAD as described in the Analytical
method section. The reaction was linear from 5 to 15 min.
Therateofattackon2 was arbitrarily set at 100.
Colorimetric assay of endo-b-galactosidase activity
A mixture containing 0.5 m
M
of each p-nitrophenyl oligo-
saccharide b-glycoside and 50 mU of b-
D

-glucosidase or
25 mU of b-N-acetyllactosaminide in 500 mL 50 m
M
sodium acetate buffer pH 5.8 and an appropriate amount
of the enzyme was incubated at 37 °C. Samples (each
50 lL) were taken at intervals (0, 5, 10, 15 and 20 min)
during the incubation and inactivated by adding 50 lL
1.0
M
Na
2
CO
3
. The amount of liberated p-nitrophenol was
determined by measuring absorbance at 405 nm using a
microplate reader. One unit of the enzyme was defined as
the amount hydrolysing 1 lmol of 2 per min. The initial
rates of the enzymatic reaction were evaluated from kinetic
curves of product accumulation as described above. The
parameters of Michaelis–Menten-type kinetics were evalu-
ated by 1/v–1/[S] plots and the least-squares method. The
substrate concentration ranges used for compounds 1, 2, 3,
4 and 5 were 0.05–0.4, 0.02–0.75, 0.1–0.8, 0.25–2.0 and 0.25–
1.5 m
M
, respectively.
Assay of endo-b-galactosidase activity by HPLC
The standard assay was carried out as follows. A reaction
mixture (500 lL) containing an appropriate substrate and
endo-b-galactosidase in 10 m

M
sodium acetate buffer
(pH 5.8) was incubated at 37 °C, and samples (each
50 lL) were taken at 3-min intervals during incubation.
After inactivation of each sample by adding 150 mL of 1
M
acetic acid, the amount of liberated GlcNAcb-pNP was
determined by HPLC as described in Analytical methods.
Transglycosylation reaction of endo-b-galactosidase
from
E. freundii
Compound 1 (16 mg, 23 lmol) was dissolved in 1.9 mL
20 m
M
sodium acetate buffer pH 5.8 followed by addition
of 2.3 mU endo-b-galactosidase from E. freundii.The
mixture was incubated for 30 days at 37 °C and was loaded
onto a Sep-pak accel QMA column (2 · 4 cm) equilibrated
with H
2
O at a flow rate of 1 mLÆmin
)1
. Eluate was collected
in 2-mL fractions and monitored by measuring the absorb-
ance at 300 nm using a spectrometer. The fractions showing
absorbance at 300 nm were combined, concentrated and
loaded onto a Shodex Asahipak GS-220FP column
(7.6 · 250 mm) equilibrated with H
2
O at a flow rate of

0.6 mLÆmin
)1
at 40 °C. Eluate was monitored on-line by
measuring the absorbance at 300 nm. The chromatogram
showed four peaks (Fig. 1A). Peak B, which was presumed
to be a transglycosylation product, was concentrated and
lyophilized to produce 9 (0.8 mg) in 3.3% yield based on the
substrate added.
1
H- and
13
C-NMR data of 9 are summar-
ized in Table 1.
Results
Preparation of colorimetric substances
A series of chromogenic substances were designed as
substrates of endo-b-galactosidase based on the structural
features of keratan sulfate, which is an alternating polymer
of N-acetyllactosamine units jointed to each other by a
Table 2. Relative hydrolytic rates of endo-b-galactosidases on p-nitro-
phenyl oligosaccharide b-glycosides. The hydrolytic actions of endo-
b-galactosidase on different substrates were investigated as described in
Materials and methods. The vertical arrow indicates the point of
cleavage. One mM of each substrate was used for the determination of
relative hydrolytic rates. –, not hydrolyzed even in the presence of 10
mU of the enzyme.
3712 T. Murata et al. (Eur. J. Biochem. 270) Ó FEBS 2003
b-(1-3) linkage. Tetrasaccharide 2 containing two N-acetyll-
actosamine repeats and its analogues were synthesized by
the alternative addition of b-(1-3) linked GlcNAc and b-(1-

4) linked Gal to Galb1-4GlcNAcb-pNP and Galb1-4Glcb-
pNP, respectively, using two kinds of glycosyltransferases.
Thus, compounds 1 and 3 were first prepared by the
regioselective transfer of GlcNAc residue from UDP-
GlcNAc to Galb1-4GlcNAcb-pNP and Galb1-4Glcb-pNP
by b3GnT from bovine serum. They were further converted
into 2 and 4 utilizing b4GalT from bovine milk (Fig. 2A,B).
The enzyme efficiently catalysed the transfer of a Gal moiety
to the OH-4¢ position of the acceptors in high yields (82 and
71%) depending on the acceptor. The positional isomers 5
and 6 were prepared simultaneously by Gal transfer from
Galb-oNP to the OH-3¢ and OH-6¢ positions of 3 using
B. circulans b-
D
-galactosidase-mediated transglycosylation
(Fig. 2B). The resulting products were obtained in a molar
ratioof1:1.3andina12%overallyieldbasedonthe
acceptor added. Compound 7 and its isomer 8 were
prepared from Galb-pNP using b-NAHase-mediated trans-
glycosylation (Fig. 2C).
Hydrolytic actions of endo-b-galactosidases
The hydrolytic actions of endo-b-galactosidases on synthetic
chromogenic substances (each 1 m
M
) were investigated by
using enzyme preparations from E. freundii and B. fragilis.
Each enzyme splits compounds 1–6 into the corresponding
reducing oligosaccharides and chromogenic substances,
GlcNAcb-pNP/Glcb-pNP. For example, compound 2 was
completely hydrolysed in an endo-manner into Galb1-

4GlcNAcb1-3Galb and GlcNAcb-pNP. The relative hydro-
lytic rates of 1 and 4 compared with that of 2 (set at 100)
were 47 and 10, i.e. 2- and 10-fold differences, respectively.
The hydrolytic activities toward 3, 5 and 6 were not detected
under the experimental conditions as described in the
Materials and methods. Furthermore, the hydrolytic rate of
reducing trisaccharide GlcNAcb1-3Galb1-4GlcNAc was
compared with that of its glycoside 1 in order to examine
how the p-nitrophenyl group participates in the hydrolytic
action. There was little progression of hydrolysis of the
reducing trisaccharide under the experimental conditions,
Fig. 1. HPLC analysis of the reaction mixture obtained by endo-b-
galactosidase-mediated transglycosylation and time courses of the pro-
duction of transglycosylation products 9 and 10 from 1 and degradation
of 1. (A) HPLC analysis was performed as described in Materials and
methods. (B) A reaction mixture (50 lL) containing 11.5 m
M
com-
pound 1, 0.1% BSA and 1.5 mU E. freundii endo-b-galactosidase in
20 m
M
sodium acetate buffer, pH 5.8, was incubated at 37 °C. The
amount of each product formed from the initial substrate was deter-
mined by HPLC. s, Peak B (compound 9); d, peak A (compound 10);
h, GlcNAcb-pNP; j, compound 1.
Fig. 2. Summary of enzymatic synthesis of p-nitrophenyl oligosaccha-
ride b-glycosides used in this work. (A) Consecutive additions of Glc-
NAc and Gal to Galb1-4GlcNAcb-pNP by b3GnT and b4GalT. (B)
Consecutive additions of GlcNAc and Gal to Galb1-4Glcb-pNP by
b3GnT and b-

D
-galactosidase or b4GalT. (C) N-acetylglucosaminy-
lation of Galb-pNP by b-N-acetylhexosaminidase-mediated transgly-
cosylation.
Ó FEBS 2003 Kinetic studies on endo-b-galactosidase (Eur. J. Biochem. 270) 3713
although the reducing trisaccharide was hydrolysed very
slowly to form GlcNAcb1-3Gal and GlcNAc when a
10-fold amount of enzyme was added (Table 2). This
indicates that the p-nitrophenyl group is critical for the
hydrolytic action of endo-b-galactosidase. p-Nitrophenyl
disaccharide b-glycosides 7, 8, Galb1-4GlcNAcb-pNP,
and GlcNAcb1-6Galb1-4GlcNAcb-pNP and GlcNAcb1-
6Galb1-4Glcb-pNP did not act as substrates for the enzyme.
Colorimetric assay of endo-b-galactosidase activity
As shown in Fig. 3, a novel method for endo-b-galactosi-
dase assay using 2 was designed on the basis of the results
described above. This assay involves the colorimetric
determination of p-nitrophenol liberated from the substrate
by the action of the enzyme through a coupled reaction
involving b-NAHase. Thus, the enzyme exclusively produ-
ces GlcNAcb-pNP from 2 and then b-NAHase hydrolyses
GlcNAcb-pNP to free p-nitrophenol. In this case, com-
pound 1 is not suitable for this assay because the terminal
GlcNAc residue is subjected to hydrolysis by b-NAHase.
Compound 2 was incubated with E. freundii endo-b-galac-
tosidase in the presence and absence of b-NAHase
(Fig. 4A). The rate of hydrolysis was first-order with respect
to the enzyme throughout the course of the determination.
The reaction proceeded linearly for at least 20 min under
the experimental conditions used (Fig. 4A). In this case,

only 10 ng of the endo-b-galactosidase could be determined
by this assay method. The dose–response plot of the coupled
enzyme vs. the colour intensity of the enzyme was found to
be a plateau in the range of 20–40 mU for 15 min. The
addition of 20 mU of coupled enzyme to the assay system
was sufficient to obtain the maximal activity of endo-b-
galactosidase (Fig. 4B). In a similar manner, compound 4
was applied to determination of endo-b-galactosidase
activity with b-
D
-glucosidase as a coupled enzyme (Fig. 4C).
When 4 was used as a substrate, 100 ng of the enzyme was
required for determination of the activity by this assay
method. The addition of 50 mU of coupled enzyme to the
assay system was sufficient to obtain the maximal activity of
endo-b-galactosidase (Fig. 4D).
Kinetics of endo-b-galactosidase
In order to elucidate in more detail the substrate specificity
of endo-b-galactosidase, parameters of Michaelis–Menten-
type kinetics for 1–5 were evaluated by a 1/v–1/[S] plot
(Fig. 5). Compound 2 was assayed with A. orientalis
b-NAHase as a coupled enzyme and 3–5 with almond
b-
D
-glucosidase using the newly developed colorimetric
assay described above. Compound 1 was assayed by HPLC.
The kinetic parameters are summarized in Table 3. Com-
pound 2 was the best among the synthetic substrates with a
k
cat

/K
m
value of 911Æm
M
)1
Æs
)1
for E. freundii endo-b-galac-
tosidase. Compound 1 still acts as a fairly good substrate
despite the absence of a Gal group at the terminal position,
because the k
cat
/K
m
value of 1 was 31% of that of 2.
However, the k
cat
/K
m
value of 4, in which Glc had been
replaced with GlcNAc, was only 0.9% of that of 2.The
same tendency was seen in a comparison of 1 and 3.This
was also the case for B. fragilis endo-b-galactosidase as
showninTable3.
Transglycosylation reaction of endo-b-galactosidase
from
E. freundii
In order to examine the transglycosylation activity of
endo-b-galactosidase, E. freundii endo-b-galactosidase was
Fig. 3. Principle of the colorimetric assay method for endo-b-galac-

tosidase using compounds 2 and 4 through a coupled reaction involving
b-NAHase or b-
D
-glucosidase.
Fig. 4. Effects of coupled enzymes on the release of p-nitrophenol from
synthetic substrates. (A) Effect of b-NAHase on the production of
p-nitrophenol from 2. Compound 2 (125 nmol) was incubated with
E. freundii endo-b-galactosidase (9.3 ng) in 0.5 mL 10 m
M
sodium
acetate buffer, pH 5.8, in the presence or absence of b-NAHase
(25 mU). The reaction was stopped by the addition of 1.0
M
Na
2
CO
3
,
and then liberated p-nitrophenol was determined spectrophotometri-
cally at 405 nm. d, Endo-b-galactosidase and b-NAHase; s, endo-
b-galactosidase; m, b-NAHase. (B) Time course of the production of
p-nitrophenol formed from 2 through a coupled reaction involving
b-NAHase. (C) Effect of b-
D
-glucosidase on the production of
p-nitrophenol from 4. Compound 4 (250 nmol) was incubated with
E. freundii endo-b-galactosidase (93 ng) in the presence or absence of
b-
D
-glucosidase (50 mU) as a coupled enzyme. d, Endo-b-galactosi-

dase and b-
D
-glucosidase; s, endo-b-galactosidase; m, b-
D
-glucosi-
dase. (D) Time course of the production of p-nitrophenol from
4 through a coupled reaction involving b-
D
-glucosidase.
3714 T. Murata et al. (Eur. J. Biochem. 270) Ó FEBS 2003
incubated with a high substrate concentration of 1 (8.4%).
After the incubation, the reaction mixture was analysed by
HPLC. As shown in Fig. 1A, two peaks, A and B, which
were presumed to be transglycosylation products, appeared
at retention times of 20 and 25 min, respectively. Based on
the results of ESI-MS analysis, the molecular mass of peak
B was estimated to be 1072, which coincides with the value
of GlcNAc-Gal-GlcNAc-Gal-GlcNAc-pNP. Peak B was
collected and lyophilized to produce 9 (3.3% actual yield
base on the donor substrate 1). The
1
H- and
13
C-NMR
signals of 9 were easily assigned by correlation with the
spectra of 1 and 2 (Table 1). The most direct evidence that
the GlcNAcb1-3Gal unit is bound to the C-4¢ position of
GlcNAc was obtained from an 8-p.p.m. downfield shift of
C-4¢ and the HMBC spectrum (Fig. 6). The results of NMR
and ESI-MS analyses revealed that 9 is a pentasaccharide

b-glycoside, GlcNAcb1-3Galb1-4GlcNAcb1-3Galb1-4Glc-
NAcb-pNP. Peak A was also subjected to ESI-MS analysis.
The molecular mass of peak A was estimated to be 1437.5,
which coincides with the theoretical value of GlcNAc-Gal-
GlcNAc-Gal-GlcNAc-Gal-GlcNAc-pNP. This result shows
that the GlcNAcb1-3Gal residue from 1 was also trans-
ferred to 9 by the enzyme. Taking into account the
regioselective synthesis of 9 from 1, peak A was estimated
to be a heptasaccharide b-glycoside, GlcNAcb1-3Galb1-
4GlcNAcb1-3Galb1-4GlcNAcb1-3Galb1-4GlcNAcb-pNP
(10). The time course of the production of transfer products
and the degradation of 1 was observed by HPLC analysis as
shown in Fig. 1B. The maximum production of 9 and 10
was reached at 4 h, and their amounts were gradually
decreased during the subsequent reaction, finally causing
complete hydrolysis.
Discussion
Endo-b-galactosidases have been used widely as analytical
tools in the field of glycobiology and glycotechnology.
However, there have been few studies on kinetics of the
enzyme because a simple and sensitive assay method has not
been developed. The conventional assay for endo-b-galac-
tosidase activity is based on the method of Park and
Johnson using keratan sulfate as a substrate. However, this
method is not always reproducible because of lack of
uniformity of the polysaccharide. Therefore, we synthesized
p-nitrophenyl di, tri, and tetrasaccharide b-glycosides 1–8 as
substrates for endo-b-galactosidase by combining a glycos-
idase-catalysed transglycosylation and a glycosyltrans-
ferase-catalysed reaction. Endo-b-galactosidases from

E. freundii and B. fragilis hydrolysed p-nitrophenyl oligo-
saccharide b-glycosides 1–6 as shown in Table 2. Based on
the hydrolytic action toward the enzyme, 2 wasfoundtobe
the best substrate among the synthetic substances. Com-
pound 1, in which terminal Gal was trimmed from 2, still
acts as a fairly good substrate. The present assay system has
better reproducibility and is simpler than the method of
Park and Johnson. From a practical viewpoint, 2, a well-
defined substrate, was shown to be very useful for a routine
submicrogram assay of endo-b-galactosidase in biological
materials.
Further kinetic studies on endo-b-galactosidases were
carried out using a series of modified substrates mentioned
above. Kinetic parameters of activities of E. freundii and
B. fragilis endo-b-galactosidases toward 1–5 are summar-
ized in Table 3. The catalytic efficiencies of 2 for both
enzymes were the highest among the synthetic substrates.
The value of V
max
/K
m
of 2 for the E. freundii enzyme was
1.7-times higher than that for keratan sulfate as shown in
Table 3, indicating that it is very useful as a substrate
instead of keratan sulfate for analytical use in the endo-
b-galactosidase assay. In addition, this similarity in the
values of V
max
/K
m

suggests that the sulfate group on the
6-position on GlcNAc of keratan sulfate is not always
essential for the hydrolytic action of the enzyme. Replace-
ment of Glc by internal GlcNAc of 2 resulted in a
remarkable reduction in the catalytic efficiency on 4.This
was also true for a comparison of 1 and 3. The increase in
Fig. 5. Effects of concentrations of substrate on endo-b-galactosidase
activity using compounds 2 (A) and 4 (B). The results shown in (A) were
obtained from an experiment in which the substrate concentration was
varied over a range of 0.02–0.75 m
M
, and the results shown in (B) were
obtained from an experiment in which the substrate concentration was
varied over a range of 0.25–2.0 m
M
. Insets show the Linweaver–Burk
plots.
Ó FEBS 2003 Kinetic studies on endo-b-galactosidase (Eur. J. Biochem. 270) 3715
catalytic efficiency was clearly due to the N-acetyl group on
C-2 of GlcNAc linked to the aglycon moiety. However,
compound 4 still acts as a fairly good substrate: the
V
max
/K
m
value of 4 is14.1%ofthatofGalb1-4GlcNAcb1-
3Galb1-4Glcb-Cer. It may be used as a substitute for the
detection of glycosphingolipid-degrading endo-b-galactosi-
dase. This concept for the enzyme assay could be applied to
other types of endo-b-galactosidases from Diplococcus

pneumonia [25], Clostridium perfringens [26,27], and mollusk
Painopecten sp. [28], which hydrolyse internal b-galactosidic
linkages of blood group A and B antigens, Gala1-4Galb1-
4GlcNAc and GlcNAca1-4Galb1-4GalNAc, and GlcAb1-
3Galb1-3Gal structures, respectively.
The structure of the site of cleavage by the enzyme, which
was deduced from results of kinetic studies using well-
defined synthetic oligosaccharides, is shown in Fig. 7A. We
propose a binding structure so that Galb1-4GlcNAcb1-
3Galb1-4GlcNAcb-pNP has a matching shape, which can
accommodate a chain of five residues (A, B, C, D, and E) in
the active site. Synthetic tri- and tetrasaccharide glycosides
had only one cleavage site for the endo-b-galactosidase,
which splits the glycoside bond between C and D. On the
other hand, Fukuda and Matsumura reported that endo-
b-galactosidases from E. freundii hydrolysed corneal
Fig. 6. HMBC spectrum of compound 9 with
1
Hand
13
C spectra on the
sides of the two-dimensional spectrum. Only the expanded region of the
HMBC spectrum showing connectivity of GlcNAc C-1¢¢ with Gal C-1¢,
newly formed by endo-b-galactosidase-mediated transglycosylation, is
presented.
Fig. 7. Proposed structure of the cleavage site of endo-b-galactosidase.
TheenzymehydrolysesbothN-acetyllactosamine-repeating tetrasac-
charide b-glycoside (A) and keratan sulfate (B). Arrows show the
cleavage site of the glycosidic linkages of each substrate.
Table 3. Kinetic parameters of endo-b-galactosidases from Escherichia freundii and Bacteroides fragilis. The parameters of Michaelis–Menten-type

kinetics were evaluated by 1/v-1/[S] plots and the least-squares method. This summary was compiled from results reported here and from data in the
literature. The substrate concentration ranges used for compounds 1, 2, 3, 4 and 5 were 0.05–0.4, 0.02–0.75, 0.1–0.8, 0.25–2.0 and 0.25–1.5 mM,
respectively. K
m
, Mean ± SEM (mM); V
max
, Mean ± SEM (lmolÆmin
)1
Æmg
)1
); k
cat
, Mean ± SEM (sec
)1
); k
cat
/K
m
, sec
)1
ÆmM
)1
.–,not
determined.
3716 T. Murata et al. (Eur. J. Biochem. 270) Ó FEBS 2003
keratan sulfate, releasing GlcNAc6SO
3
–
b1-3Gal and
GlcNAc6SO

3
–
b1-3Gal6SO
3
–
b1-4GlcNAc6SO
3
–
b1-3Gal as
major products [8]. This finding suggests that the enzyme
tolerates C-6 sulfation of the sugar residues A and B, which
have to be partially O-sulfated in keratan sulfate, but not
C-6 sulfation of the sugar residue C. The hydrolysates
GlcNAc6SO
3
–
b1-3Gal and GlcNAc6SO
3
–
b1-3Gal6-
SO
3
–
b1-4GlcNAc6SO
3
–
b1-3Gal may occupy corresponding
sites B-C and -A-B-C, respectively, as shown in Fig. 7B. The
sugar residue D at the cleavage site influences the sensitivity
of oligosaccharides to the enzyme, because hydrolytic action

was promoted by the presence of an N-acetyl group on C-2
of GlcNAc corresponding to sugar residue D (Table 3).
Disaccharide b-glycoside 7 did not act as a substrate.
Reduction of the reducing-end residue of Galb1-4Glc-
NAcb1-3Galb1-4Glc inhibited the action of endo-b-gal-
actosidases from E. freundii [7,29]. These results indicate
that a sugar pyranose structure such as GlcpNAc or Glcp on
leaving site D is required for the hydrolytic action of endo-
b-galactosidases. Furthermore, the mode of linkage of sugar
residues A to B was strict for the binding locus in the active
site, because conversion of the (1–4) into (1–3)-linkages of
terminus Gal to GlcNAc residues remarkably decreased the
enzyme action (Table 2). These observations indicate that a
tetrasaccharide sequence consisting of two LacNAc repeat-
ing units such as 2 is preferable to the binding locus in the
active site. A series of chromogenic substrates were shown
to be advantageous as probes for substrate recognition at
theactivesiteintheenzyme.
Generally, glycosidase has transglycosylation activity as
a reverse reaction of hydrolysis. Taking into account the
hydrolytic action of endo-b-galactosidase, 1 wasusedasa
substrate for transglycosylation reaction. The enzyme
produced 9 and 10 as major products through consecutive
transfer of the GlcNAcb1-3Gal unit (Fig. 8). Thus, the
enzyme catalysed regioselective transfer to the OH-4¢
position of 1 of the GlcNAcb1-3Gal unit from the same
substrate to produce 9.Inthiscase,1 serves as both
donor and acceptor substrates in the transglycosylation.
Furthermore, the disaccharide elongation reaction pro-
gresses in order and produces heptasaccharide glycoside

10 from 9. However, once 9 and 10 had reached
maximum concentrations, they gradually decreased in a
subsequent reaction and finally caused complete degrada-
tion (Fig. 1B,C) indicating that 9 and 10 are fairly good
substrates for endo-b-galactosidase and reflect an N-
acetyllactosamine-repeating structure. The analytical yield
of 9 was estimated by HPLC analysis (Fig. 1) to be 7.0%
based on the donor substrate. The large difference
between yields in actual and analytical data is though to
be caused by a loss of 9 through the process of
chromatographic separation procedures.
Poly (N-acetyllactosamine) has been shown to be present
on membrane glycoconjugates [30,31] and has been identi-
fied as a precursor of Lewis X, sialyl Lewis X, and blood
group antigens. The amounts of poly (N-acetyllactosamine)
chains have been shown to be changed during cellular
differentiation [32] and malignant transformation of cells
[33,34]. Furthermore, poly (N-acetyllactosamine)s are
recognized with high affinity by galectins [35] and are
involved in apoptosis [36]. Recently, Ando et al. reported
that sialylated poly (N-acetyllactosamine) on the cell surface
facilitated apoptotic cell uptake by macrophages [37]. On
the other hand, it has been reported that Helicobacter pylori
selectively interacted with poly (N-acetyllactosamine) of
human erythrocytes [38,39] and with Neu5Aca2-3Galb1-
4GlcNAcb1-3Galb1-4GlcNAcb1-3Galb1-4Glcb-Cer and
NeuAca2-3Galb1-4GlcNAcb1-3Galb1-4GlcNAcb1-
3Galb1-4GlcNAcb1-3Galb1-4Glcb-Cer isolated from
human gastric adenocarcinoma [40]. These findings suggest
that poly (N-acetyllactosamine) chains involved in glyco-

conjugates play important roles in biological events such as
metastasis, apoptosis, phagocytosis, and infection. Accord-
ingly, compounds 9 and 10 should be useful as model
compounds for studying biological functions of N-acetyl-
lactosamine repeating domains involved in glycoproteins
and glycolipids.
In this study, the first enzymatic synthesis of poly
(N-acetyllactosamine) has been achieved by endo-b-galac-
tosidase-mediated transglycosylation. This enzyme would
be an excellent tool for producing a poly (N-acetyllactosa-
mine) chain as well as for the detection of poly (N-acetyl-
lactosamine)s involved in glycoconjugates.
Acknowledgements
We thank Yamasa Corporation and Meiji Milk Products for the gift of
UDP-GlcNAc and UDP-Gal and the gift of b-
D
-galactosidase from
B. circulans ATCC31382, respectively. We also thank JEOL Hightech
Ltd. (Akishima, Japan) for ESI-MS analysis of the transfer products.
This work was supported by a Grant-in-Aid for Scientific Research
(No. 14760047) from the Ministry of Education, Science, Sports,
Culture, and Technology of Japan.
References
1. Kitamikado, M., Ueno, R. & Nakamura, T. (1970) Enzymic
degradation of whale cartilage keratosulfate. II. Identification of a
keratosulfate-degrading bacterium. Bull.Jap.Soc.Sci.Fish.36,
1174–1176.
2. Hirano, S. & Meyer, K. (1971) Enzymatic degradation of corneal
and cartilaginous keratosulfates. Biochem. Biophys. Res. Commun.
44, 1371–1375.

3. Nakazawa, K. & Suzuki, S. (1975) Purification of keratan sulfate-
endogalactosidase and its action on keratan sulfates of different
origin. J. Biol. Chem. 250, 912–917.
Fig. 8. Proposed mechanism of E. freundii endo-b-galactosidase-medi-
ated transglycosylation reaction. Compound 1 as both a donor and
acceptor substrate was incubated with E. freundii endo-b-galactosi-
dase. The enzyme catalysed the transfer of the GlcNAcb1-3Gal unit
from 1 to nonreducing-terminal GlcNAc residues of 1 and 9.
Ó FEBS 2003 Kinetic studies on endo-b-galactosidase (Eur. J. Biochem. 270) 3717
4. Kitamikado, M., Ito, M. & Li, Y.T. (1981) Isolation and char-
acterization of a keratan sulfate-degrading endo-b-galactosidase
from Flavobacterium keratolyticus. J. Biol. Chem. 256, 3906–3909.
5. Ito, M., Hirabayashi, Y. & Yamagata, T. (1986) Substrate speci-
ficity of endo-b-galactosidases from Flavobacterium keratolyticus
and Escherichia freundii is different from that of Pseudomonas sp.
J. Biochem. (Tokyo). 100, 773–780.
6. Scudder, P., Uemura, K., Dolby, J., Fukuda, M.N. & Feizi, T.
(1983) Isolation and characterization of an endo-b-galactosidase
from Bacteroides fragilis. Biochem. J. 213, 485–494.
7. Fukuda, M. & Matsumura, G. (1975) Endo-b-galactosidase of
Escherichia freundii. Hydrolysis of pig colonic mucin and milk
oligosaccharides by endoglycosidic action. Biochem. Biophys. Res.
Commun. 64, 465–471.
8. Fukuda, M.N. & Matsumura, G. (1976) Endo-b-galactosidase of
Escherichia freundii. Purification and endoglycosidic action on
keratan sulfates, oligosaccharides, and blood group active glyco-
protein. J. Biol. Chem. 251, 6218–6225.
9. Fukuda, M.N., Watanabe, K. & Hakomori, S.I. (1978) Release
of oligosaccharides from various glycosphingolipids by endo-
b-galactosidase. J. Biol. Chem. 253, 6814–6819.

10. Fukuda, M.N. (1981) Purification and characterization of endo-
b-galactosidase from Escherichia freundii induced by hog gastric
mucin. J. Biol. Chem. 256, 3900–3905.
11. Scudder, P., Hanfland, P., Uemura, K. & Feizi, T. (1984)
Endo-b-
D
-galactosidases of Bacteroides fragilis and Escherichia
freundii hydrolyze linear but not branched oligosaccharide
domains of glycolipids of the neolacto series. J. Biol. Chem. 259,
6586–6592.
12. Scudder, P., Tang, P.W., Hounsell, E.F., Lawson, A.M.,
Mehmet, H. & Feizi, T. (1986) Isolation and characterization of
sulphated oligosaccharides released from bovine corneal keratan
sulphate by the action of endo-b-galactosidase. Eur. J. Biochem.
157, 365–373.
13. Scudder, P., Lawson, A.M., Hounsell, E.F., Carruthers, R.A.,
Childs, R.A. & Feizi, T. (1987) Characterization of oligosacchar-
ides released from human-blood-group O erythrocyte glycopep-
tides by the endo-b-galactosidase of Bacteroides fragilis.Astudyof
the enzyme susceptibility of branched poly (N-acetyllactosamine)
structures. Eur. J. Biochem. 168, 585–593.
14. Usui, T., Ogawa, K., Nagai, H. & Matsui, H. (1992) Enzymatic
synthesis of p-nitrophenyl 4
5
-O-b-
D
-galactosyl-a-maltopentaoside
as a substrate for human a-amylase. Anal. Biochem. 202, 61–67.
15. Nanjo, F., Sakai, K. & Usui, T. (1988) p-Nitrophenyl penta-
N-acetyl-b-chitopentaoside as a novel synthetic substrate for the

colorimetric assay of lysozyme. J. Biochem. (Tokyo) 104, 255–
258.
16. Takegawa, K., Fujita, K., Fan, J.Q., Tabuchi, M., Tanaka, N.,
Kondo, A., Iwamoto, H., Kato, I., Lee, Y.C. & Iwahara, S. (1998)
Enzymatic synthesis of a neoglycoconjugate by transglycosylation
with Arthrobacter endo-b-N-acetylglucosaminidase: a substrate
for colorimetric detection of endo-b-N-acetylglucosaminidase
activity. Anal. Biochem. 257, 218–223.
17. Ichikawa, Y., Look, G.C. & Wong, C H. (1992) Enzyme-
catalyzed oligosaccharide synthesis. Anal. Biochem. 20, 215–238.
18. Shoda, S., Fujita, M. & Kobayashi, S. (1998) Glycanase-catalyzed
synthesis of non-natural oligosaccharides. Trends Glycosci. Gly-
cotech. 10, 279–289.
19. Murata, T. & Usui, T. (2000) Enzymatic synthesis of important
oligosaccharide units involved in N-andO-linked glycans. Trends
Glycosci. Glycotech. 12, 161–174.
20. Fujimoto, H., Miyasato, M., Ito, Y., Sasaki, T. & Ajisaka, K.
(1998) Purification and properties of recombinant b-galactosidase
from Bacillus circulans. Glycoconj. J. 15, 155–160.
21. Sakai, K., Kastumi, R., Ohi, H., Usui, T. & Ishido, Y. (1992)
Enzymatic synthesis of N-acetyllactosamine and N-acetylallo-
lactosamine by the use of b-
D
-galactosidase. J. Carbohydr. Chem.
11, 553–565.
22. Usui, T., Kubota, S. & Ohi, H. (1993) A convenient synthesis of
b-
D
-galactosyl disaccharide derivatives using the b-
D

-galactosidase
from Bacillus circulans. Carbohydr. Res. 244, 315–323.
23. Matahira, Y., Tashiro, A., Sato, T., Kawagishi, H. & Usui, T.
(1995) Enzymic synthesis of lacto-N-triose ll and its positional
analogues. Glycoconj. J. 12, 664–671.
24. Murata, T., Tashiro, A., Itoh, T. & Usui, T. (1997) Enzymic
synthesis of 3¢-O-and6¢-O-N-acetylglucosaminyl-N-acetyllactos-
aminide glycosides catalyzed by b-N-acetyl-
D
-hexosaminidase
from Nocardia orientalis. Biochim. Biophys. Acta 1335, 326–334.
25. Takasaki, S. & Kobata, A. (1976) Purification and characteriza-
tion of an endo-b-galactosidase produced by Diplococcus pneu-
moniae. J. Biol. Chem. 251, 3603–3609.
26. Fushuku, N., Muramatsu, H., Uezono, M.M. & Muramatsu, T.
(1987) A new endo-b-galactosidase releasing Gala1–3Gal from
carbohydrate moieties of glycoproteins and from a glycolipid.
J. Biol. Chem. 262, 10086–10092.
27. Ashida, H., Anderson, K., Nakayama, J., Maskos, K., Chou,
C W.,Cole,R.B.,Li,S C.&Li,Y T.(2001)Anovelendo-
b-galactosidase from Clostridium perfringens that liberates the
disaccharide GlcNAca1–4Gal from glycans specifically expressed
in the gastric gland mucous cell-type mucin. J. Biol. Chem. 276,
28226–28232.
28. Takagaki, K., Nakamura, T., Takeda, Y., Daidouji, K. & Endo,
M. (1992) A new endo-b-galactosidase acting on the Galb1–3Gal
linkage of the proteoglycan linkage region. J. Biol. Chem. 267,
18558–18563.
29. Nakagawa, H., Yamada, T., Chien, J.L., Gardas, A., Kitamikado,
M., Li, S.C. & Li, Y.T. (1980) Isolation and characterization of an

endo-b-galactosidase from a new strain of Escherichia freundii.
J. Biol. Chem. 255, 5955–5959.
30. Fukuda, M., Fukuda, M.N. & Hakomori, S. (1979) Develop-
mental change and genetic defect in the carbohydrate structure of
band 3 glycoprotein of human erythrocyte membrane. J. Biol.
Chem. 254, 3700–3703.
31. Spooncer, E., Fukuda, M., Klock, J.C., Oates, J.E. & Dell, A.
(1984) Isolation and characterization of polyfucosylated lactosa-
minoglycan from human granurocytes. J. Biol. Chem. 259, 4792–
4801.
32. Muramatsu, T., Gachelin, G., Nicolas, J.F., Condamine, H.,
Jakob, H. & Jacob, F. (1978) Carbohydrate structure and cell
differentitation: unique properties of fucosylglycopeptides isolated
from embryonal carcinoma cells. Proc. Natl Acad. Sci. USA 75,
2315–2319.
33. Yamashita, K., Ohkura, T., Tachibana, Y., Takasaki, S. &
Kobata, A. (1984) Comparative study of the oligosaccharides
released from baby hamster kidney cells and their polyoma
transformant by hydrazinolysis. J. Biol. Chem. 259, 10834–10840.
34. Pierce, M. & Arango, J. (1986) Rous sarcoma virus-transformed
baby hamster kidney cells express higher levels of aspar-
agine-linked tri- and tetraantennary glycopeptides containing
[GlcNAcb (1,6) Man-a (1,6) Man] and poly–N–acetyllactosamine
sequences than baby hamster kidney cells. J. Biol. Chem. 261,
10772–10777.
35. Barondes, S.H., Cooper, D.N., Gitt, M.A. & Leffler, H. (1994)
Galectins. Structure and function of a large family of animal lec-
tins. J. Biol. Chem. 269, 20807–20810.
36. Perillo, N.L., Pace, K.E., Seilhamer, J.J. & Baum, L.G. (1995)
Apoptosis of T cells mediated by galectin-1. Nature 378, 736–739.

37. Ando, K., Hagiwara, T., Beppu, M. & Kikugawa, K. (2000)
Naturally occurring anti-band 3 antibody binds to apoptotic
human T-lymphoid cell line Jurkat through sialylated poly-
N-acetyllactosaminyl saccharide chains on the cell surface.
Biochem. Biophys. Res. Commun. 275, 412–417.
3718 T. Murata et al. (Eur. J. Biochem. 270) Ó FEBS 2003
38. Miller-Podraza, H., Milh, M.A., Bergstrom, J. & Karlsson, K.A.
(1996) Recognition of glycoconjugates by Helicobacter pylori:
an apparently high-affinity binding of human polyglycosyl-
ceramides, a second sialic acid-based specificity. Glycoconj. J. 13,
453–460.
39. Karlsson, H., Larsson, T., Karlsson, K. & Miller-Podraza, H.
(2000) Polyglycosylceramides recognized by Helicobacter pylori:
analysis by matrix-assisted laser desorption/ionization mass
spectrometry after degradation with endo-b-galactosidase and by
fast atom bombardment mass spectrometry of permethylated
undegraded material. Glycobiology 10, 1291–1309.
40. Roche, N., Larsson, T., Angstrom, J. & Teneberg, S. (2001)
Helicobacter pylori-binding gangliosides of human gastric adeno-
carcinoma. Glycobiology 11, 935–944.
Ó FEBS 2003 Kinetic studies on endo-b-galactosidase (Eur. J. Biochem. 270) 3719