Thermal decomposition of β-d-galactopyranosyl-(1→3)-2-acetamido- 2-deoxy-d-hexopyranoses under neutral conditions

Article abstract of DOI:10.1016/j.carres.2010.06.003

β-d-Galactopyranosyl-(1→3)-2-acetamido-2-deoxy-d-glucose (LNB) and β-d-galactopyranosyl-(1→3)-2-acetamido-2-deoxy-d-galactose (GNB) decompose rapidly upon heating into d-galactose and mono-dehydrated derivatives of the corresponding 2-acetamido-2-deoxy-d-

Full text of DOI:10.1016/j.carres.2010.06.003

Carbohydrate Research 345 (2010) 1901–1908
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Thermal decomposition of b-D-galactopyranosyl-(1?3)-2-acetamido-
2-deoxy-D-hexopyranoses under neutral conditions
*
Kazuhiro Chiku, Mamoru Nishimoto, Motomitsu Kitaoka
National Food Research Institute, National Agriculture and Food Research Organization, 2-1-12 Kannondai, Tsukuba, Ibaraki 305-8642, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
b-
acetamido-2-deoxy-
drated derivatives of the corresponding 2-acetamido-2-deoxy-
dideoxy-hex-2-enofuranoses and bicyclic 2-acetamido-3,6-anhydro-2-deoxy-hexofuranoses. The decom-
position is conducted under neutral conditions where glycosyl linkages are generally believed to be sta-
ble. The half-lives of LNB and GNB were 8.1 min and 20 min, respectively, at 90 °C and pH 7.5. The pH
dependency of decomposition rates suggests that the instabilities are an extension of the conditions
for the peeling reaction, often observed with glycans of O-linked glycoproteins under alkaline conditions.
Such decomposition under the neutral conditions is commonly observed with 3-O-linked reducing
aldoses.
D
-Galactopyranosyl-(1?3)-2-acetamido-2-deoxy-
D-glucose (LNB) and b-D-galactopyranosyl-(1?3)-2-
Received 16 March 2010
Received in revised form 3 June 2010
Accepted 9 June 2010
D-galactose (GNB) decompose rapidly upon heating into
D-galactose and mono-dehy-
D-hexoses, including 2-acetamido-2,3-
Available online 15 June 2010
Keywords:
b-
D-Galactopyranosyl-(1?3)-2-acetamido-
2-deoxy-D-hexopyranose
2-Acetamido-2,3-dideoxy-hex-2-
enofuranose
Ó 2010 Elsevier Ltd. All rights reserved.
2-Acetamido-3,6-anhydro-2-deoxy-
hexofuranose, bicyclic
Peeling reaction
Decomposition under neutral conditions
Reducing aldose, 3-O-linked
1. Introduction
rides containing 3-O-linked HexNAc as the reducing end show fur-
ther decomposition, and unsaturated sugars are formed.16–19 This
b-
(1?3)HexNAc] are common building units of biologically func-
tional sugar chains. b- -Galactopyranosyl-(1?3)-2-acetamido-2-
deoxy-
D
-Galactopyranosyl-(1?3)-2-acetamido-2-deoxy-hexoses [Galb
undesirable reaction is sometimes called ‘peeling’, and it proceeds
until the alkali-stable glycosidic linkage appears at the reducing
end. The products of the peeling reaction of HexNAc can undergo a
color change with the addition of p-dimethylaminobenzaldehyde
in strong hydrochloric acid (Ehrlich reagent); the coloring reaction
is known as the Morgan–Elson reaction.20 The reaction has been
used as a method of detecting 3-O-linked HexNAc residues in carbo-
hydrates and in their decomposition products. Based on these meth-
ods, Kuhn et al.21,22 demonstrated that LNB is easily degraded under
alkaline conditions. In addition, some 3-methylated derivatives of
GlcNAc have also shown facile decomposition under alkaline condi-
tions.23,24 Therefore, it is commonly thought that 3-O-linked Hex-
NAc possesses alkaline instability.
D
D
-glucose [lacto-N-biose I (LNB)] is often found in sugar
lipids, blood-type antigens, and human milk oligosaccharides.1–5
The core 1 structure of O-linked glycans of glycoproteins contains
a
-linked b-D-galactopyranosyl-(1?3)-2-acetamido-2-deoxy-D-gal-
actose [galacto-N-biose (GNB)] to serine/threonine (Ser/Thr).6–9
We reported one-pot enzymatic process to produce LNB10 and
GNB11 from sucrose and the corresponding 2-acetamido-2-deoxy-
hexoses [HexNAc; 2-acetamido-2-deoxy-D-glucose (GlcNAc) and
2-acetamido-2-deoxy- -galactose (GalNAc), respectively]. During
D
the development of this method, we observed a strange phenome-
non wherein LNB and GNB completely decomposed when the solu-
tion was heated at neutral pH for a short time, despite the general
belief that glycosyl linkages are stable under neutral conditions.
It is known that b-elimination under alkaline conditions to liber-
ate glycans from Ser/Thr often causes the removal of GalNAc from
glycan by another b-elimination.12–15 The reductive alkaline cleav-
age reaction to form oligosaccharide alditols is required to liberate
the glycans without decomposition. In addition, some oligosaccha-
In this study we found that the thermal instabilities of LNB and
GNB under neutral conditions are an extension of their alkaline
instabilities, and that the phenomena commonly occur with free
3-O-linked reducing aldoses. Herein, we show a kinetic analysis
of the decomposition of LNB and GNB by heat treatment.
2. Results
2.1. Isolation of decomposition products of LNB and GNB
LNB and GNB mostly decomposed within 1 h at 90 °C under
neutral conditions when reacted with 100 mM sodium phosphate
* Corresponding author. Tel.: +81 29 838 8071; fax: +81 29 838 7321.
E-mail address: mkitaoka@affrc.go.jp (M. Kitaoka).
0008-6215/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2010.06.003

1902
K. Chiku et al. / Carbohydrate Research 345 (2010) 1901–1908
amido-2,3-dideoxy-D-erythro-2-enofuranoses. These compounds
a
b
were previously reported by Ogata et al.25 and Beau et al.26
Bicyclic 3,6-anhydrohexofuranoses were suggested for the
structures of 4 and 5 based on the long-range C–H coupling corre-
lations of C-3–H-6, C-4–H-1, and C-6–H-3 that appeared on HMBC
spectra like 1 and 2. The relative stereoconfiguration of the com-
pounds was deduced by determining cross peaks of the protons
at C-1–C-3 using nuclear Overhauser effect spectroscopy (NOESY)
(Table 1). Finally 4 and 5 were identified to be C-2 epimers of 2-
acetamido-3,6-anhydrohexofuranoses with the
D-gulo and D-ido
configurations, respectively. The inversion of the conformation at
C-3 from GalNAc should be noted.
The anomeric proton (H-1) of 6 with a 2-enol structure gave a
doublet of doublets (dd, H-1
a, J1,3 1.0 Hz, J1,4 4.0 Hz) and an appar-
ent triplet (H-1b, J1,3 1.0 Hz, J1,4 1.0 Hz,
a:b = 1:0.6). In addition, the
heteronuclear multiple-bond correlation (HMBC) spectrum exhib-
ited a long-range C–H coupling correlation of C-4–H-1, correspond-
ing
to
the
adapted
2-acetamido-2,3-dideoxy-D-threo-2-
enofuranose.
Figure 1. Heat treatment of (a) LNB and (b) GNB with sodium phosphate buffer (pH
7.5) at 90 °C for 0–120 min.
High performance liquid chromatography (HPLC) analysis
showed that the decomposition of LNB/GNB proceeds concomi-
tantly with the release of an equimolar amount of Gal and
mono-dehydrated derivatives of the corresponding HexNAc. The
time-course of the LNB decomposition (Fig. 2a) showed that 1–3
were generated in yields of about 18%, 32%, and 46%, respectively,
after heat treatment for 60 min. In contrast, GNB decomposition
(Fig. 2b) gave two major bicyclic 3,6-anhydrohexofuranoses, 4
(37%) and 5 (47%), after heat treatment for 100 min. In these reac-
tions, levels of the hex-2-enofuranosides 3 or 6 increased at an
early stage (reaching about 48% or 33%, respectively), and then re-
mained steady or decreased gradually. These results suggest that
the bicyclic 3,6-anhydrohexofuranoses 1–2 and 4–5 could also be
generated from the hex-2-enofuranoses 3 and 6.
buffer (pH 7.5). In both cases, two spots were detected on thin-
layer chromatography (TLC) (Fig. 1). One spot showed the same
Rf value as D-galactose (Gal), while the other product showed a
higher Rf value than the corresponding HexNAc in each case. The
reaction products showing the higher Rf values were separated
by two-column chromatography using a Shodex Asahipak ODP-
50 and NH2P50-4E column to give three LNB-decomposition
products 1–3, and three GNB-decomposition products 4–6. Mass
spectrometry (MS) showed that all products 1–6 could be repre-
sented by the molecular formula C8H13NO5 that corresponds to
mono-dehydrated HexNAc. Nuclear magnetic resonance (NMR)
spectra suggested that the products are equilibrium mixtures of
a
and b anomers.
2.2. Temperature- and pH-dependence on decomposition of
LNB and GNB
H
H
HO
HO
O
O
OH
OH
We investigated the stabilities of LNB and GNB under several pH
and temperature conditions by determining the decomposition
rate using the Gal 1-P assay.27 The rate constant (k) of LNB decom-
position was 0.085 0.002 minꢀ1 with 100 mM sodium phosphate
buffer (pH 7.5) at 90 °C, which is about 2.5 times higher than the
rate constant of GNB decomposition (k = 0.034 0.001 minꢀ1).
The half-lives of LNB and GNB under the above-mentioned condi-
tions were 8.1 min and 20 min, respectively. LNB and GNB decom-
positions were undetectable at temperatures below 55 °C and
65 °C, respectively, in 100 mM sodium phosphate buffer (pH 7.5)
for 1 h. The activation energy (Ea) values calculated for the decom-
position of LNB and GNB were 125 5 kJ molꢀ1 and
132 2 kJ molꢀ1, respectively (Fig. 3). The concentration of phos-
phate buffer between 20 mM and 250 mM did not significantly af-
fect the reaction rate of LNB decomposition.
O
O
NHAc
NHAc
H
H
1
2
OH
H
O
HO
OH
NHAc
H
3
H
HO
HO
O
O
OH
OH
The pH dependence of LNB decomposition was studied by treat-
ing LNB at 90 °C for 0–60 min with the various above-mentioned
buffers at pH values in the range 4.0–10.5. LNB and GNB were
mostly stable at pH conditions below 5.5 at 90 °C for 1 h (remain-
ing amount >95%). The rate constants of the decomposition of LNB
and GNB increased exponentially with pH. Linear relationships be-
tween log10 k and pH observed for both LNB and GNB (Fig. 4) indi-
cated that the decomposition was a function of the hydroxyl ion
concentration. The reaction was not significantly affected by the
variation in buffer, and depended on the pH value (concentration
of hydroxide ions) and temperature. The identical compounds
were detected under various conditions on HPLC analysis at the
early stage of the reaction. Simple acid hydrolysis was not ob-
served under the conditions examined.
O
O
NHAc
NHAc
H
H
4
5
OH
H
O
HO
OH
NHAc
6
The compounds 1 and 2 were identified to be C-2 epimers of bicy-
clic 2-acetamido-3,6-anhydrohexofuranoses with the -manno and
-gluco configurations, respectively. The compound 3 was 2-acet-
D
D

K. Chiku et al. / Carbohydrate Research 345 (2010) 1901–1908
1903
Table 1
NOESY spectral data of 3,6-anhydro-HexNAc furanoses
Cross peak on NOESY spectruma
Identification of 3,6-anhydro-HexNAc furanose
H-1–H-2
H-2–H-3
Compound 1
1a
1b
ꢀ
+
+
N-Acetyl-
N-Acetyl-b-
a
-
D
-mannosamine form
+
D-mannosamine form
Compound 2
2a
2b
+
ꢀ
ꢀ
ꢀ
N-Acetyl-
N-Acetyl-b-
a
-
D
-glucosamine form
D-glucosamine form
Compound 4
4a
4b
+
ꢀ
+
+
N-Acetyl-
N-Acetyl-b-
a
-
D
-gulosamine form
D-gulosamine form
Compound 5
5a
5b
ꢀ
ꢀ
ꢀ
N-Acetyl-
N-Acetyl-b-
a
-
D
-idosamine form
+
D-idosamine form
a
Cross peaks of NOESY spectrum were clearly (+) or undetectable (ꢀ).
a
b
100
80
60
40
20
0
2.3. Heat treatment of various LNB- and GNB-related sugars
The heat stabilities of 17 sugars were examined under neutral
conditions. Only the free 3-O-linked hexoses showed decomposi-
tion. The rate constants and the half-lives are summarized in
Table 2. All compounds whose structures can be described as
3-O-linked reducing hexoses were decomposed to remove the
reducing hexoses under these conditions. No decomposition of lac-
to-N-tetraose was detected under these conditions, suggesting that
the instability required the reducing end of LNB to be free. N-Acet-
yllactosamine, lactose, and cellobiose were also stable under these
conditions, suggesting that the 4-O-linked reducing hexoses were
0
10
20
30
40
50
60
Time (min)
stable. The rate constant of b-D-galactopyranosyl-(1?3)-2-deoxy-
100
80
60
40
20
0
D-glucose [Galb(1?3)2dGlc] was comparable to that of LNB,
whereas b-D-galactopyranosyl-(1?3)-D-glucose [Galb(1?3)Glc]
and laminaribiose decomposed much more slowly. Lewisa trisaccha-
ride28 decomposed with the release of Gal slightly faster than LNB,
indicating that the additional 4-O-linkage did not stabilize LNB.
Nigerose decomposed at a similar rate to laminaribiose, indicating
that cleavage was not affected whether the linkage was
a or b. It
should be noted that laminaribiose and nigerose produced an equi-
molar amount of glucose (not double molar), probably from their
non-reducing ends. N-Acetylmuramic acid was found to decompose,
indicating that the decomposition did not require a glycosyl linkage.
0
20
40
60
80
100
120
Time (min)
3. Discussion
Figure 2. Time course of the decomposition reaction of (a) LNB and (b) GNB in
sodium phosphate buffer (pH 7.5). The symbols shown are as follows: s, LNB; d,
GNB; 4, Gal; h, 1; j, 2; O, 3; }, 4; ꢀ, 5; ., 6. GlcNAc and GalNAc were not detected.
GNB and LNB are often found in mammalian biologically func-
tional carbohydrates. We report herein that the free forms of
GNB and LNB are actually easily decomposed by heating under
neutral conditions, where glycosyl linkages are generally consid-
ered to be stable. The pH dependencies of the decomposition rate
clearly indicate that the instabilities are an extension of the condi-
tions for the peeling reaction, which is often observed under alka-
line conditions. We also demonstrated that instability under
neutral conditions commonly occurs with 3-O-linked reducing
hexoses.
-2
-4
k
The products of the peeling reaction have been predicted but
not determined earlier. This is probably due to two reasons; the
low availability of LNB and GNB, and the fact that the alkaline con-
ditions of the peeling reaction cause further decomposition of the
products. The large-scale production of the compounds and the
-6
0.0027
0.0028
1/T (K-1
0.0029
)
Figure 3. The Arrhenius plot of the decompositions of LNB and GNB. s, LNB; d,
GNB.
IUPAC name is b-D-galactopyranosyl-(1?3)-2-deoxy-D-arabino-hexose.

1904
K. Chiku et al. / Carbohydrate Research 345 (2010) 1901–1908
Judging from the compounds formed, the reaction is initiated
0
-1
-2
-3
-4
a
b
from the linear form of HexNAc. Scheme 1 illustrates the cleavage
of LNB. Initially, the enol form of the aldehyde is generated as the
intermediate, and the C–O linkage at position 3 is cleaved to form
the 2-deoxy-2,3-unsaturated sugar (the linear form of compound
3) with the release of galactose through b-elimination. The hydro-
xy group at the C-6 position attacks the double bond at the C-3 po-
sition to form
a furan ring of the enol compound. The
stereochemical configuration at the C-2 position is lost through
the formation of an aldehyde.29,30 It should be noted that the re-
verse reaction, the formation of the bicyclic compound, is identical
to the peeling reaction where compounds 1 and 2 are intramolec-
ular 3-O-substituted 2-acetamido-2-deoxy-hexoses.
This mechanism explains the difference in the rate of decompo-
sition of LNB, Galb(1?3)2dGlc, and Galb(1?3)Glc. Only
Galb(1?3)Glc can quench the enol intermediate by the formation
of a carbonyl group at the C-2 position (ketose). Kainuma et al.31
reported that glucose was transformed into fructose in 200 mM so-
dium phosphate buffer (pH 6.5–7.5) at 97 °C. We confirmed that
the treatment of 50 mM glucose in 100 mM sodium phosphate buf-
fer (pH 7.5) at 90 °C for 1 h caused transformation into fructose at a
yield of 11%. This is also evidence for the formation of the enol
structure under neutral conditions.
N-Acetylmuramic acid decomposed much slower than LNB did.
This can be explained by the reaction mechanism in which the C–O
linkage is stabilized from cleavage by the acidity of the lactic acid.
It should be noted that N-acetylmuramic acid 6-phosphate hydro-
lase (or etherase), an enzyme that hydrolyzes N-acetylmuramic
acid 6-phosphate into GlcNAc 6-phosphate and lactic acid, is con-
sidered to catalyze the reaction via a similar mechanism in which
the elimination of lactate gives a 2,3-unsaturated aldehyde and is
followed by the addition of water to give the saturated product.32
The reactions which transformed LNB and GNB into 1–3 and 4–
6 in total yields of approximately 100% in the reaction mixture are
economical and simple methods to give libraries of 2,3-unsatu-
rated furanose and bicyclic furanoses. These compounds have
effectively not been isolated from the peeling reaction under alka-
line conditions because the intermediates are easily decomposed
under the severe conditions.
5
6
7
8
9
10
pH
0
-1
-2
-3
-4
5
6
7
8
9
10
pH
Figure 4. Relationship between rate constant (k) and pH on LNB and GNB
decompositions. The following relationships on (a) LNB and (b) GNB decomposition
were derived from linear plot with a correlation coefficient greater than 0.91 and
0.92, respectively: (a) y = 0.57x ꢀ 5.7 (x = [pH value]; y = log10 [k]); (b)
y = 0.60x ꢀ 6.3 (x = [pH value]; y = log10 [k]). Buffers used are as follows: s,
100 mM sodium phosphate buffer (pH 5.7, 6.3, 6.9, 7.4, and 7.9); d, 100 mM
citrate–NaOH buffer (pH 5.3, 5.7, and 6.3); }, 100 mM TAPS–NaOH buffer (pH 6.2,
6.8, and 7.4); ꢀ, 100 mM HEPES–NaOH buffer (pH 6.8, 7.4, and 7.8); 4, 100 mM
CHES–NaOH buffer (pH 8.5, 8.8, and 9.5); N, 100 mM glycine–NaOH buffer (pH 7.2,
7.8, 8.4, 8.9, and 9.4); and j, 100 mM tricine–NaOH buffer (pH 6.6, 7.1, 7.7, and 8.2).
much milder conditions for the degradation used in this study
made detailed analyses of the reaction easier.
The HexNAc parts of LNB and GNB are transferred to the fura-
noses of 3-deoxy-2,3-unsaturated compounds (3 and 6) and bicy-
clic 3,6-anhydro compounds (1–2 and 4–5), losing the
configuration at the C-2 position of GlcNAc and GalNAc,
respectively.
The bicyclic furanoses 1 and 2 have been reported as intermedi-
ates in the synthesis of the furanodictines A and B, which show a
neuronal differentiation activity in rat PC12 cells.33 Ogata et al.25
reported the transformation of GlcNAc into 1–3 by treating
Table 2
Thermal decomposition of various sugars under neutral conditions
Substrate
Structure
Rate constant (k) (minꢀ1
)
Half-life (t1/2) (min)
Lacto-N-biose I (LNB)
Lacto-N-tetraose (LNT)
Lewisa trisaccharide
Galacto-N-biose (GNB)
Galb(1?3)GlcNAc
Galb(1?3)GlcNAcb(1?3)Galb(1?4)Glc
0.085
N.D.a
0.13
0.034
0.0056
0.069b
0.0059
0.0081
N.D.a
8.1
—
Galb(1?3)[Fuca(1?4)]GlcNAc
5.3
20
124
10
117
86
—
133
—
—
—
—
72
—
—
Galb(1?3)GalNAc
Galb(1?3)Glc
Galb(1?3)2dGlc
Glcb(1?3)Glc
Glcb(1?3)Glcb(1?3)Glc
Glcb(1?3)Glcb-O-Me
b-
b-
D
-Galactopyranosyl-(1?3)-
D-glucose
D-Galactopyranosyl-(1?3)-2-deoxy-
D
-glucose
Laminaribiose
Laminaritriose
b-Methyl laminaribioside
Nigerose
Glc
a(1?3)Glc
0.0052
Cellobiose
Glcb(1?4)Glc
Glcb(1?4)2dGlc
Galb(1?4)Glc
Galb(1?4)GlcNAc
GlcNAc 3-O-lactyl ether
GlcNAc
N.D.a
b-
D-Glucopyranosyl-(1?4)-2-deoxy-
D-glucose
N.D.a,b
N.D.a
Lactose
N-Acetyllactosamine
N-Acetylmuramic acid
N-Acetylglucosamine (GlcNAc)
N-Acetylgalactosamine (GalNAc)
N.D.a
0.010c
N.D.a
GalNAc
N.D.a
a
b
c
N.D., not detected under the conditions.
The rate constant calculated by quantifying Gal or Glc produced.
The rate constant calculated by quantifying the total yield of compounds 1–3.

K. Chiku et al. / Carbohydrate Research 345 (2010) 1901–1908
1905
OH
OH
OH
OH
OH
OH
OH
H
OH
OH
OH
O
O
O
OH
O
HO
OH
H
HO
HO
O
HO
O
HO
O
HO
O
OH
OH
NHAc
OH
NHAc
OH
H
NHAc
- Gal
OH
H
HO
H
OH
OH
OH
O
HO
linear form
H
H
O
NHAc
H
NHAc
OH
H
H
H
O
HO
OH
HO
HO
O
O
OH
H
OH
furanose form
H
H
NHAc
O
O
NHAc
NHAc
H
3
H
O
H
HO
HO
O
OH
OH
furanose form
O
O
NHAc
NHAc
H
H
1
2
Scheme 1. Proposed mechanism of thermal decomposition of LNB into compounds 1–3.
100 mM GlcNAc in 400 mM borate buffer (pH 7.0) at 100 °C for 2 h
at an approximately 80% conversion yield. The thermal decomposi-
tion of LNB presented here may be another practical method of
synthesizing 1–3 by improving the procedures for the isolation of
each compound, since it gives a high conversion yield and does
not require excess amounts of ionic compounds. It should also be
noted that LNB can be easily prepared in a one-pot enzymatic sys-
tem from sucrose and GlcNAc.11
a laminaribiose phosphorylase.34,35 b-
-glucose [Galb-(1?3)-Glc] and b- -galactopyranosyl-(1?3)-2-
deoxy- -glucose [Galb-(1?3)-2dGlc] were prepared by the meth-
od given by Nakajima et al.36 b-
-Glucopyranosyl-(1?4)-2-
deoxy- -glucose [Glcb-(1?4)-2dGlc] was prepared as described
D-Galactopyranosyl-(1?3)-
D
D
D
D
D
previously.37 Lacto-N-tetraose and N-acetylmuramic acid were
purchased from Sigma–Aldrich Chemical Co. (St. Louis, MO). Nige-
rose and cellobiose were purchased from Wako Pure Chemicals
(Osaka, Japan). N-Acetyllactosamine was purchased from Sei-
kagaku Kogyo (Tokyo, Japan). Lewisa trisaccharide {Galb-
We conclude that 3-O-substituted reducing aldoses are gener-
ally decomposed under neutral conditions by heating.
(1?3)[Fuca-(1?4)]GlcNAc} was purchased from Dextra Laborato-
ries (Reading, UK).
4. Experimental
4.1. Sugars used
4.2. Decomposition of LNB and GNB
LNB or GNB was dissolved to approximately 50 mM with the fol-
lowing buffers: 100 mM citrate–NaOH buffer (pH 4.0, 4.5, 5.0, 5.5,
and 6.0), 100 mM sodium phosphate buffer (pH 6.0, 6.5, 7.0, 7.5,
LNB and GNB were synthesized from sucrose using GNB/LNB
phosphorylase as described previously.11,12 Laminaribiose, lami-
naritriose, and methyl b-laminaribioside were synthesized using

1906
K. Chiku et al. / Carbohydrate Research 345 (2010) 1901–1908
8.0, and 8.5), 100 mM HEPES–NaOH buffer (pH 7.0, 7.5, 8.0, and 8.5),
100 mM Tricine–NaOH buffer (pH 7.5, 8.0, 8.5, and 9.0), 100 mM
TAPS–NaOH buffer (pH 8.0, 8.5, and 9.0), 100 mM glycine–NaOH
buffer(pH8.5, 9.0, 9.5, 10.0, and10.5), and100 mMCHES–NaOHbuf-
4.7. Determination of kinetic parameters
Rate constant (k) of the decomposition was calculated by
regressing data with Eq. 1 or 2 using Grafit version 4 (Erithacus
Software, Middlesex, United Kingdom):
fer (pH 9.7, 10.0, and 10.5). The solutions (20
200-
lL) were placed into a
l
L thermal tube and subjected to heat treatment at 90 °C for 0–
½Substrateꢂt ¼ ½Substrateꢂ0 ꢃ expðꢀk ꢃ tÞ
½Productꢂt ¼ ½Substrateꢂ0 ꢃ ½1 ꢀ expðꢀk ꢃ tÞꢂ
ð1Þ
ð2Þ
60 min using a GeneAmpÒ PCRSystem 9700 (Applied Biosystems,
CA, USA). The products were periodically analyzed by TLC, HPLC,
and colorimetric quantification employing carbohydrate enzymes
(each of the measurement methods are described below).
The apparent energy of activation (Ea) was obtained by regress-
ing data with the Arrhenius Eq. 3 using Grafit version 4:
ln k ¼ ln A ꢀ Ea=RT ðA : frequency factorÞ
ð3Þ
4.3. Thin-layer chromatography (TLC)
TLC was performed on Silica Gel-60 F254 (E. Merck) using 80%
MeCN in distilled water as a solvent. The products were visualized
by heating the plate after dipping it in 20% H2SO4 in MeOH.
4.8. Influence of temperature, pH, and treatment times on
decomposition of LNB and GNB
LNB and GNB were treated under various conditions of pH (the
previously mentioned buffers at pH 4.0–10.5 were used), at tem-
peratures of 25–95 °C and the treatment time was between 0
and 60 min, and the products were detected by TLC. In addition,
the concentration of the LNB or GNB reduced with the heat treat-
ment was sequentially measured using Gal 1-P assay.27
4.4. HPLC analysis
Both the substrate and the products of heat treatment were
analyzed by HPLC systems (Shimadzu, Kyoto, Japan) with a Corona
Charged Aerosol Detector (ESA Inc., Chelmsford, MA) using a Sho-
dex Asahipak NH2P50-4E column (4.6 mm i.d. ꢁ 250 mm) with
75% MeCN, 90% MeCN (for separation of 3,6-anhydro compounds),
and 70% MeCN (for LNT and Lewisa trisaccharide) in distilled water
as the solvent at a flow rate of 1.0 mL/min.
4.9. Heat treatment of various sugars
The sugars were dissolved to approximately 50 mM with
100 mM sodium phosphate buffer (pH 7.5). The pH value in the solu-
tion of MurNAc was readjusted to 7.5 with NaOH solution. The mix-
ture was incubated at 90 °C for 0–180 min, and the concentrations of
the sugars and the products were sequentially analyzed by HPLC.
4.5. Spectrum analysis
Electrospray-ionization mass spectrometry (ESIMS) spectra of
the purified decomposition products were recorded in the posi-
tive-ion mode on a Bruker APEX II 70e Fourier-transform ion-cyclo-
tron resonance mass spectrometer (Bruker Daltonics, Billerica, MA,
USA). One-dimensional (1H and 13C) and two-dimensional [dou-
ble-quantum-filtered correlation spectroscopy (DQF-COSY), NOESY,
heteronuclear single quantum coherence (HSQC), and HMBC] NMR
spectra of the purified products were recorded with a Bruker Avance
800 or Avance 500 spectrometer (Bruker Biospin, Rheinstetten, Ger-
many) in D2O with t-BuOH as the internal standard.
4.10. Isolation of LNB decomposition products (1–3)
LNB (768 mg, 2.00 mmol) was dissolved in 40 mL of 100 mM so-
dium phosphate buffer (pH 7.5) and incubated at 90 °C for 2 h. The
product was desalted with Amberlite MB-3 (Organo) and concen-
trated. The solution was purified using a Shodex Asahipak ODP-50
column (10 mm i.d. ꢁ 250 mm, distilled water), and the products
were lyophilized. A significant amount of a new compound that
was not detected in the reaction mixture and which showed a lower
Rf value than 1–3 on TLC with exact mass corresponding to [2Hex-
NAcꢀ4H2O] {ESIMS: m/z 393.12 [M+Na]+ (calcd for C16H22N2NaO8,
393.13)} was formed during this stage causing the drastic decreases
in the isolation yields of 1–3. The residue was dissolved in ace-
tone–water, and the solution was further purified by Shodex Asahi-
pak NH2P50-4E column chromatography (4.6 mm i.d.ꢁ 250 mm;
37:3 acetone–water) to give the LNB decomposition products 1
(9.6 mg) and 2 (13 mg), and product 3 (24 mg).
4.6. Enzymatic quantification of LNB, GNB, and Gal
The amounts of LNB, GNB, and Gal were monitored by quantify-
ing the produced Gal 1-P27 from Gal by galactokinase (GalK) and
from LNB and GNB by GNB/LNB phosphorylase.38 Working reagent
A for the Gal 1-P assay was prepared from the following mixture:
0.5 IU/mL of UDP-glucose hexose-1-phosphate uridylyltransferase,
2.5 IU/mL of phosphoglucomutase, 2.0 IU/mL of glucose 6-phos-
phate dehydrogenase, 20
2.0 mM thio-NAD+, and 2.0 mM UDP-
HCl buffer (pH 7.5) containing 10 mM MgCl2. The determination
of LNB and GNB was performed by the preparation of a 100-
sample in 96-well microtiter plate. The solution was added to
50 L of reagent A and reagent B containing 2 IU/mL of GNB/LNB
l
M
a
-
D
-glucose 1,6-bisphosphate,
4.11. Isolation of GNB decomposition products (4–6)
D-glucose in 50 mM Tris–
GNB (768 mg, 2.00 mmol) was dissolved in 40 mL of 100 mM
sodium phosphate buffer (pH 7.5) and incubated at 90 °C for 3 h.
The purification of the decomposition products was performed
using the same operation as for the purification of the LNB decom-
position products, to give the GNB decomposition products 4
(26 mg) and 5 (37 mg), and product 6 (11 mg). A significant
amount of a new compound showing lower Rf value than 4–6 on
TLC formed during the first stage of the purification, causing the
drastic decreases in the isolation yields of 4–6.
lL
l
phosphorylase in 100 mM sodium phosphate buffer (pH 7.5) and
incubated at 37 °C for 1 h. It is important that the reaction solution
containing reagent A and B be prepared daily because magnesium
phosphate precipitates upon storage. The concentration of Gal 1-P
produced was quantified by measuring absorbance at 400 nm de-
rived from thio-NADH (e
398 = 11,700 cmꢀ1 Mꢀ1). The determina-
tion of Gal was performed via the same operation by replacing
the reagent B with reagent C containing 2 IU/mL mutarotase,
4 IU/mL GalK, and 2 mM ATP in 50 mM Tris–HCl buffer (pH 7.5)
containing 10 mM MgCl2. The LNB/GNB or Gal calibration curve
was linear in the concentration range from 0 to 0.25 mM.
4.12. 2-Acetamido-3,6-anhydro-2-deoxy-D-mannofuranose (1)
Compound 1 was obtained in a yield of 2.4% as a colorless pow-
der, and included two anomers (
:b = 1:1.5) in the mixture; 1H
a

K. Chiku et al. / Carbohydrate Research 345 (2010) 1901–1908
1907
(D2O, 500 MHz)
a
anomer: d 5.51 (d, 1H, J1,2 5.5 Hz, H-1), 4.68 (t, 1H,
(CH3CONH–); ESIMS: m/z 226.07 [M+Na]+ (calcd for C8H13NNaO5,
J3,4 4.8 Hz, J4,5 4.8 Hz, H-4), 4.63 (dd, 1H, J2,3 5.5 Hz, H-3), 4.37–4.40
226.07).
0
(m, H-5), 4.34 (t, 1H, H-2), 3.93 (dd, 1H, J5,6 6.8 Hz, J6,6 8.4 Hz, H-6),
3.89 (t, 1H, J5,6 8.4 Hz, H-60), 2.06 (s, 3H, CH3CONH–); b anomer: d
4.16. 2-Acetamido-3,6-anhydro-2-deoxy-D-idofuranose (5)
0
5.28 (d, 1H, J1,2 6.3 Hz, H-1), 4.79 (H-4), 4.61 (dd, 1H, J2,3 5.6 Hz,
J3,4 4.8 Hz, H-3), 4.37–4.40 (m, H-5), 4.23 (dd, 1H, H-2), 4.01 (dd,
Compound 5 was obtained in a yield of 9.1% as a colorless oil and
included two anomers
:b = 1:2) in the mixture; 1H (D2O,
800 MHz) anomer: d 5.31 (d, 1H, J1,2 1.8 Hz, H-1), 4.64–4.66
1H, J5,6 6.5 Hz, J6,6 8.9 Hz, H-6), 3.53 (dd, 1H, J5,6 8.3 Hz, H-60),
(a
0
0
2.04 (s, 3H, CH3CONH–); 13C NMR (D2O, 125 MHz)
a
anomer: d
a
175.9 (CH3CONH–), 97.2 (C-1), 83.6 (C-4), 81.9 (C-3), 72.9 (C-5),
72.4 (C-6), 56.4 (C-2), 23.3 (CH3CONH–); b anomer: d 176.2
(CH3CONH–), 102.5 (C-1), 82.4 (C-4), 81.6 (C-3), 73.2 (C-5), 72.8
(C-6), 60.7 (C-2), 23.4 (CH3CONH–); ESIMS: m/z 226.07 [M+Na]+
(calcd for C8H13NNaO5, 226.07).
(m, H-4), 4.62 (dd, 1H, J2,3 1.0 Hz, J3,4 4.6 Hz, H-3), 4.42 (m, 1H, H-
0
5), 4.27 (m, 1H, J5,6 3.6 Hz, J6,6 10.2 Hz, H-6), 4.13 (br s, 1H, H-2),
3.83–3.84 (m, 1H, H-60), 2.00 (s, 3H, CH3CONH–); b anomer: d
5.50 (d, 1H, J1,2 4.4 Hz, H-1), 4.75–4.76 (H-3), 4.64–4.66 (m, H-4),
4.32 (dd, 1H, J 1.5 Hz, 0.7 Hz, H-5), 4.17 (t, 1H, J2,3 4.4 Hz, H-2),
3.9 (m, 2H, H-6 and H-60), 2.03 (s, 3H, CH3CONH–); 13C NMR (D2O,
4.13. 2-Acetamido-3,6-anhydro-2-deoxy-
D
-glucofuranose (2)
200 MHz) a anomer: d 175.7 (CH3CONH–), 103.8 (C-1), 89.8 (C-4),
87.0 (C-3), 76.8 (C-5), 75.5 (C-6), 63.1 (C-2), 23.5 (CH3CONH–); b
anomer: d 176.1 (CH3CONH–), 98.3 (C-1), 87.1 (C-3), 86.6 (C-4),
76.0 (C-5), 74.1 (C-6), 59.6 (C-2), 23.4 (CH3CONH–); ESIMS: m/z
226.07 [M+Na]+ (calcd for C8H13NNaO5, 226.07).
Compound 2 was obtained in a yield of 3.2% as a colorless pow-
der, and included two anomers (
:b = 1:0.4) in the mixture; 1H
(D2O, 800 MHz) anomer: d 5.59 (d, 1H, J1,2 4.4 Hz, H-1), 4.73 (t,
a
a
1H, J4,5 5.4 Hz, H-4), 4.63 (dd, 1H, J2,3 4.4 Hz, J3,4 5.7 Hz, H-3),
0
4.28 (ddd, 1H, J5,6 5.9 Hz, J5,6 7.5 Hz, H-5), 4.24 (t, 1H, H-2), 3.97
4.17. 2-Acetamido-2,3-dideoxy-D-threo-hex-2-enofuranose (6)
(dd, 1H, J6,6 9.3 Hz, H-6), 3.63 (dd, 1H, H-60), 2.03 (s, 3H,
0
CH3CONH–); b anomer: d 5.41 (d, 1H, J1,2 1.5 Hz, H-1), 4.76 (H-4),
4.49 (dd, 1H, J2,3 1.3 Hz, J3,4 4.7 Hz, H-3), 4.33–4.39 (m, H-5), 4.15
Compound 6 was obtained in a yield of 2.8% as a colorless pow-
der, and included two anomers (
:b = 1:0.6) in the mixture; 1H NMR
(D2O, 800 MHz): anomer: d 6.12 (dd, 1H, J3,4 1.8 Hz, H-3), 5.96 (t,
1H, J1,3 1.0 Hz, J1,4 1.0 Hz, H-1), 4.83 (H-4), 3.68–3.73 (m, 2H, H-5 and
H-6), 3.59–3.62 (m, 1H, H-60), 2.10 (s, 3H CH3CONH–); b anomer: d
6.11 (dd, 1H, J3,4 1.6 Hz, H-3), 6.01 (dd, 1H, J1,3 1.0 Hz, J1,4 4.0 Hz,
H-1), 5.04 (ddd, 1H, J4,5 3.2 Hz, H-4), 3.71 (ddd, 1H, J5,6 4.3 Hz,
a
0
(br s, 1H, H-2), 3.93 (dd, 1H, J5,6 6.5 Hz, J6,6 7.5 Hz, H-6), 3.86 (t,
a
1H, J5,6 7.5 Hz, H-60), 1.99 (s, 3H, CH3CONH–); 13C NMR (D2O,
0
200 MHz)
a anomer: d 176.0 (CH3CONH–), 99.3 (C-1), 87.4 (C-3),
80.6 (C-4), 71.9 (C-5, C-6), 60.4 (C-2), 23.4 (CH3CONH–); b anomer:
d 175.7 (CH3CONH–), 104.0 (C-1), 87.4 (C-3), 84.5 (C-4), 72.7 (C-6),
72.4 (C-5), 63.5 (C-2), 23.4 (CH3CONH–); ESIMS: m/z 226.07
[M+Na]+ (calcd for C8H13NNaO5, 226.07).
0
0
H-5), 3.68 (dd, 1H, J6,6 11.6 Hz, H-6), 3.62 (dd, 1H, J5,6 7.5 Hz,
H-60), 2.10 (s, 3H, CH3CONH–); 13C NMR (D2O, 200 MHz);
anomer:
175.2 (CH3CONH–), 135.5 (C-2), 111.6 (C-3), 100.9 (C-1), 86.1 (C-4),
75.1 (C-5), 64.3 (C-6), 24.2 (CH3CONH–); anomer: 175.1
a
4.14. 2-Acetamido-2,3-dideoxy-
(3)
D-erythro-hex-2-enofuranose
b
(CH3CONH–), 135.5 (C-2), 111.9 (C-3), 101.4 (C-1), 86.4 (C-4), 74.4
(C-5), 65.4 (C-6), 24.2 (CH3CONH–); ESIMS: m/z 226.07 [M+Na]+
(calcd for C8H13NNaO5, 226.07).
Compound 3, called Chromogen I, was obtained in a yield of
6.0% as a colorless oil, and included two anomers ( :b = 1:0.6) in
the mixture; 1H NMR (D2O, 500 MHz)
anomer: d 6.14 (br s, 1H,
a
a
Acknowledgments
H-3), 6.01 (dd, 1H, J1,3 0.9 Hz, J1,4 4.0 Hz, H-1), 5.03 (dt, 1H, J3,4
1.6 Hz, J4,5 4.0 Hz, H-4), 3.80 (dt, 1H, J5,6 4.0 Hz, H-5), 3.69 (dd,
We thank the staffs of Instrumental Analysis Center for Food
Chemistry of National Food Research Institute for recording NMR
and MS spectra. This work was supported in part by a grant from
the Promotion of Basic Research Activities for Innovative Biosci-
ences (PROBRAIN) of Japan.
1H, J6,6 11.9 Hz, H-6), 3.55 (dd, 1H, J5,6 7.2 Hz, H-60), 2.10 (s, 3H,
CH3CONH–); b anomer: d 6.20 (dd, 1H, J3,4 = 1.6 Hz, H-3), 5.97 (d,
1H, J1,3 = 1.0 Hz, H-1b), 4.81 (H-4), 3.75–3.71 (m, 2H, H-5 and H-
6), 3.59–3.64 (m, 1H, H-60), 2.10 (s, 3H CH3CONH–); 13C NMR
0
0
(D2O, 125 MHz):
a anomer: d 176.5 (CH3CONH–), 137.3 (C-2),
112.2 (C-3), 102.4 (C-1), 87.9 (C-4), 76.4 (C-5), 65.4 (C-6), 25.6
(CH3CONH–); b anomer: d 176.5 (CH3CONH–), 136.8 (C-2), 113.1
(C-3), 102.3 (C-1), 87.5 (C-4), 76.8 (C-5), 65.5 (C-6), 25.6
(CH3CONH–); ESIMS: m/z 226.07 [M+Na]+ (calcd for C8H13NNaO5,
226.07).
References
1. Haeuw-Fievre, S.; Wieruszeski, J. M.; Plancke, Y.; Michalski, J. C.; Montreuil, J.;
Strecker, G. Eur. J. Biochem. 1993, 215, 361–371.
2. Urashima, T.; Saito, T.; Nakamura, T.; Messer, M. Glycoconjugate J. 2001, 18,
357–371.
3. Kuhn, R.; Gauhe, A.; Baer, H. H. Chem. Ber. 1954, 87, 289–300.
4. Cote, R. H.; Morgan, W. T. J. Nature 1956, 178, 1171–1172.
5. Schiffman, G.; Kabat, E. A.; Leskowitz, S. J. Am. Chem. Soc. 1962, 84, 73–77.
6. Springer, G. F. Science 1984, 224, 1198–1206.
7. Bhavanandan, V. P.; Zhu, Q.; Yamakami, K.; Dilulio, N. A.; Nair, S.; Capon, C.;
Lemoine, J.; Fournet, B. Glycoconjugate J. 1998, 15, 37–49.
8. Berger, E. G. Biochim. Biophys. Acta 1999, 1455, 255–268.
9. Hanisch, F. A. Biol. Chem. 2001, 382, 143–149.
10. Nishimoto, M.; Kitaoka, M. Biosci. Biotechnol. Biochem. 2007, 71, 2101–2104.
11. Nishimoto, M.; Kitaoka, M. Carbohydr. Res. 2009, 344, 2573–2576.
12. Jorpes, E. Biochem. Zool. 1929, 204, 354–360.
13. Anderson, B.; Meyer, K.; Hoffman, P. Biochim. Biophys. Acta 1963, 74, 309–311.
14. Carlson, D. M. J. Biol. Chem. 1968, 243, 616–626.
15. Lloyd, K. O.; Kabat, E. A.; Licerio, E. Biochemistry 1968, 7, 2976–2990.
16. Kabat, E. A.; Bassett, E. W.; Pryzwans, K.; Lloyd, K. O.; Kaplan, M. E.; Layug, E. J.
Biochemistry 1965, 4, 1632–1638.
17. Lloyd, K. O.; Kabat, E. A. Carbohydr. Res. 1967, 4, 165–177.
18. Lloyd, K. O.; Kabat, E. A. Proc. Natl. Acad. Sci. U.S.A. 1968, 61, 1470–1477.
19. Lloyd, K. O.; Kabat, E. A. Carbohydr. Res. 1969, 9, 41–48.
20. Morgan, W. T. J.; Elson, L. A. Biochem. J. 1934, 28, 988–995.
4.15. 2-Acetamido-3,6-anhydro-2-deoxy-D-gulofuranose (4)
Compound 4 was obtained in a yield of 6.3% as a colorless pow-
der, and included two anomers (
:b = 1:0.9) in the mixture; 1H
(D2O, 500 MHz) anomer: d 5.22 (d, 1H, J1,2 5.6 Hz, H-1), 4.73–
a
a
4.76 (H-3, H-4), 4.37–4.40 (m, 1H, H-5), 4.23 (t, 1H, J2,3 5.6 Hz, H-
0
2), 3.90–3.94 (m, 1H, J5,6 6.8 Hz, H-6), 3.89 (dd, 1H, J5,6 3.0 Hz,
J6,6 10.5 Hz, H-60), 2.06 (s, 3H, CH3CONH–); b anomer: d 5.43 (d,
0
1H, J1,2 5.6 Hz, H-1), 4.73–4.76 (H-3), 4.59 (d, 1H, J3,4 4.0 Hz), 4.40
0
(d, 1H, J5,6 3.6 Hz, H-5), 4.34 (dd, 1H, J6,6 10.1 Hz, H-6), 4.31 (t,
1H, J2,3 5.6 Hz, H-2), 3.83 (d, 1H, H-60), 2.04 (s, 3H, CH3CONH–);
13C NMR (D2O, 125 MHz)
a anomer: d 176.1 (CH3CONH–), 102.2
(C-1), 87.5 (C-4), 81.7 (C-3), 76.7 (C-5), 76.1 (C-6), 60.7 (C-2),
23.4 (CH3CONH–); b anomer: d 175.9 (CH3CONH–), 97.2 (C-1),
89.0 (C-4), 81.6 (C-3), 77.6 (C-5), 75.6 (C-6), 56.3 (C-2), 23.3

1908
K. Chiku et al. / Carbohydrate Research 345 (2010) 1901–1908
21. Kuhn, R.; Kruger, G. Chem. Ber. 1957, 90, 264–277.
32. Hadi, T.; Dahl, U.; Mayer, C.; Tanner, M. E. Biochemistry 2008, 47,
11547–11558.
33. Kikuchi, H.; Saito, Y.; Komiya, J.; Takaya, Y.; Honma, S.; Nakahata, N.; Ito, A.;
Oshima, Y. J. Org. Chem. 2001, 66, 6982–6987.
34. Kitaoka, M.; Sasaki, T.; Taniguchi, H. Denpun Kagaku 1993, 40, 311–314.
22. Kuhn, R.; Gauhe, A.; Baer, H. H. Chem. Ber. 1956, 89, 1027–1033.
23. Aminoff, D.; Morgan, W. T. J.; Watkins, W. M. Biochem. J. 1952, 51, 379–389.
24. Jeanloz, R. W.; Tremege, M. Fed. Proc. 1956, 15, 282.
25. Ogata, M.; Hattori, T.; Takeuchi, R.; Usui, T. Carbohydr. Res. 2010, 345, 230–234.
26. Beau, J. M.; Rollin, P.; Sinaÿ, P. Carbohydr. Res. 1977, 53, 187–195.
27. Nihira, T.; Nakajima, M.; Inoue, K.; Nishimoto, M.; Kitaoka, M. Anal. Biochem.
2007, 371, 259–261.
35. Kitaoka, M.; Sasaki, T.; Taniguchi, H. Arch. Biochem. Biophys. 1993, 304, 508–
514.
36. Nakajima, M.; Nishimoto, M.; Kitaoka, M. J. Biol. Chem. 2009, 284, 19220–
28. Rege, V. P.; Morgan, W. T. J.; Painter, T. J.; Watkins, W. M. Nature 1964, 204,
740–742.
19227.
37. Tariq, M. A.; Hayashi, K.; Tokuyasu, K.; Nagata, T. Carbohydr. Res. 1995, 275, 67–
29. Coxon, B.; Hough, L. J. Chem. Soc. 1961, 1577–1579.
30. Angyal, S. J. Top. Curr. Chem. 2001, 215, 1–14.
72.
38. Kitaoka, M.; Tian, J.; Nishimoto, M. Appl. Environ. Microbiol. 2005, 71, 3158–
31. Kainuma, K.; Tadokoro, K.; Suzuki, S. Nougeikagakukaishi 1968, 42, 173–177 (in
Japanese).
3162.