Development of a chemical strategy to produce rare aldohexoses from ketohexoses using 2-aminopyridine

Article abstract of DOI:10.1016/j.carres.2011.09.021

Rare sugars are monosaccharides that are found in relatively low abundance in nature. Herein, we describe a strategy for producing rare aldohexoses from ketohexoses using the classical Lobry de Bruyn-Alberda van Ekenstein transformation. Upon Schiff-base formation of keto sugars, a fluorescence-labeling reagent, 2-aminopyridine (2-AP), was used. While acting as a base catalyst, 2-AP efficiently promoted the ketose-to-aldose transformation, and acting as a Schiff-base reagent, it effectively froze the ketose-aldose equilibrium. We could also separate a mixture of Sor, Gul, and Ido in their Schiff-base forms using a normal-phase HPLC separation system. Although Gul and Ido represent the most unstable aldohexoses, our method provides a practical way to rapidly obtain these rare aldohexoses as needed.

Full text of DOI:10.1016/j.carres.2011.09.021

Carbohydrate Research 346 (2011) 2693–2698
Contents lists available at SciVerse ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Development of a chemical strategy to produce rare aldohexoses from
ketohexoses using 2-aminopyridine
Kayo Hasehira a, Nobumitsu Miyanishi b, Wataru Sumiyoshi b, Jun Hirabayashi b, Shin-ichi Nakakita a,
⇑
a Department of Functional Glycomics, Life Science Research Center, Kagawa University, 1750-1 Ikenobe, Miki-cho, Kita-gun, Kagawa 761-0793, Japan
b Department of Glyco-Bioindustry, Life Science Research Center, Kagawa University, 1750-1 Ikenobe, Miki-cho, Kita-gun, Kagawa 761-0793, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 June 2011
Received in revised form 15 September 2011
Accepted 21 September 2011
Available online 29 September 2011
Rare sugars are monosaccharides that are found in relatively low abundance in nature. Herein, we
describe a strategy for producing rare aldohexoses from ketohexoses using the classical Lobry de Bru-
yn–Alberda van Ekenstein transformation. Upon Schiff-base formation of keto sugars, a fluorescence-
labeling reagent, 2-aminopyridine (2-AP), was used. While acting as a base catalyst, 2-AP efficiently pro-
moted the ketose-to-aldose transformation, and acting as a Schiff-base reagent, it effectively froze the
ketose–aldose equilibrium. We could also separate a mixture of Sor, Gul, and Ido in their Schiff-base
forms using a normal-phase HPLC separation system. Although Gul and Ido represent the most unstable
aldohexoses, our method provides a practical way to rapidly obtain these rare aldohexoses as needed.
Ó 2011 Elsevier Ltd. All rights reserved.
Keywords:
Rare sugar
2-Aminopyridine
Aldohexose
Lobry de Bruyn–Alberda van Ekenstein
transformation
1. Introduction
in Pseudomonas cichorii ST-24.7,8 Because the enzyme also catalyzes
the 3-epimerization of
the mass production of
D
-Fru, a common ketohexose, it is used for
Rare sugars are monosaccharides that exist in nature in extre-
D
-Psi.9,10
L-Rhamnose isomerase from Esch-
mely small amounts. They include the aldohexoses,
(All), -/ -altrose (Alt), -galactose (Gal), -glucose (Glc),
(Gul), -/ -idose (Ido), -mannose (Man), and -/ -talose (Tal), and
the ketohexoses, -fructose (Fru), -/ -psicose (Psi), -/ -sorbose
(Sor), and -/ -tagatose (Tag). Recently, rare sugars have attracted
D
-/L
-allose
erichia coli11 and many other bacteria12 are also used to produce
D
L
L
L
D
-/L
-gulose
rare sugars.
L-Rhamnose isomerase from Pseudomonas stutzeri has
D
L
L
D
L
been used to produce
D
-All, an aldose, from -Psi, a ketose, proba-
D
L
D
L
D
L
bly via the Lobry de Bruyn–Alberda van Ekenstein rearrange-
ment,13 another classic chemistry that is base catalyzed.14 In
general, enzymatic reactions are preferable to chemical ones as
the former are stereospecific and less dangerous. From a practical
standpoint, however, a complete set of ‘specific’ enzymes for the
production of all aldoses does not currently (and may not) exist.
D
L
attention because some can be used as low calorie sweeteners,1,2
can act as inhibitors of cancer cell growth3 and microbial prolifer-
ation,4 and may also be memory enhancers.5 For the preparation of
rare aldohexoses, the cyanohydrin (Kiliani–Fischer) synthesis has
long been the classic chemistry enabling a systematic synthesis
of aldoses starting from the smallest glyceraldehyde.6 However,
this procedure uses the highly toxic reagent hydrogen cyanide,
and most of the precursor aldopentoses are not available in nature.
During the last 25 years, enzyme catalysis has become a viable,
alternative approach for the production of rare sugars in many
cases. Izumori and his co-workers have extensively studied the
To date,
Arthrobacter globiformis ST48 cells.15
allitol by Gluconobacter frateurii IFO 3254 cells.16 There are also re-
ports of enzymatic production of -Sor from -Tag and -Gal from
-Sor.17,18 Nevertheless, a practical and general strategy for the
D-Tag and
L
-Tag have been produced from galactitol by
L-Psi was also produced from
D
D
L
L
preparation of aldohexoses, many of which are not found in [large
quantities in] nature, does not exist.
converting enzyme,
D-tagatose 3-epimerase, originally identified
Recently, we developed a general analytical method for the syn-
thesis and purification of
D-hexoses using the monoamine-cou-
pling reagent 2-aminobenzamide (2-AB) and subsequent HPLC
purification.19 Although we attempted to use the fluorescence tag
2-aminopyridine (2-AP) in the aforementioned study, the attempt
was unsuccessful, because 2-AP seemed to specifically catalyze
ketose–aldose transformation via the Lobry de Bruyn–Alberda
van Ekenstein rearrangement (hereafter simply denoted the Lobry
Abbreviations: All, allose; Alt, altrose; 2-AP, 2-aminopyridine; Fru, fructose; Gal,
galactose; Glc, glucose; HPLC, high performance liquid chromatography; Man,
mannose; PA-, pyridylaminated; Psi, psicose; RI, refractive index; Tag, tagtose; Tal,
talose; TFA, trifluoroacetic acid.
⇑
Corresponding author. Tel.: +81 87 891 2409; fax: +81 87 891 2410.
E-mail address: nakakita@med.kagawa-u.ac.jp (S. Nakakita).
0008-6215/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2011.09.021

2694
K. Hasehira et al. / Carbohydrate Research 346 (2011) 2693–2698
rearrangement;14), while other labeling reagents, including 2-ami-
nobenzamide, had no such effect. Since it was first introduced by
Hase et al.,20 2-AP has been widely used as a fluorescence-labeling
reagent for glycoproteins,21,22 glycolipids,23,24 and free oligosac-
charides.25,26 The reaction uses a monoamine-coupling mechanism
to produce a Schiff-base derivative, which can be converted into a
stable, reduced pyridylaminated (PA) saccharide by dimethyl-
amine Á borane treatment. As far as we know, however, 2-AP has
never been shown to catalyze a Lobry rearrangement. Therefore,
we first investigated the reaction conditions for the transformation
of a ketose to the two aldose products to establish a general proce-
dure for the chemical synthesis of rare aldoses (Scheme 1).
and 10 Â 250 mm) were obtained from Nacalai tesque Inc. (Kyoto,
Japan). -Psi, -Tag, and -Sor were obtained from Fushimi Pharma-
ceutical Co., Ltd (Kagawa, Japan). PA-Glc was purchased from Takara
Bio Inc. (Otsu, Japan). A TSKgel sugar AXI column (4.6 Â 150 mm)
and TSKgel HW-40F resin were purchased from Tosoh (Tokyo,
Japan).
D
D
D
2.2. Schiff-base formation between 2-AP and monosaccharides
Each ketohexose (10
type glass tube, and it was dissolved in the coupling reagent
(20 L, prepared by mixing 552 mg of 2-AP and 200 of
AcOH20,27), and the reaction mixture was sealed and heated at
90 °C for 60 min. Next, 20 L of MeOH and 40 L of toluene were
lmol) was put into a conical square-cap
l
lL
Another critical issue is the efficient generation of free mono-
saccharides from Schiff-base derivatives. Scheme 2 shows the ba-
sic chemistry that we used to prepare aldohexoses with the
l
l
added into each mixture, over which a stream of N2 was blown
at 60 °C for 10 min in vacuo in a Palstation model 4000 apparatus
(Takara Biomedicals, Kyoto, Japan)20,28 to remove the 2-AP. By
repeating this procedure three times, most of the unreacted 2-AP
was removed. Residual amounts of unwanted materials including
2-AP were removed by gel filtration through a TSK gel HW-40F col-
umn (7 Â 55 mm) equilibrated with 10 mM ammonium acetate
(pH 6.0).21 The fluorescence of each chromatographic fraction
was measured using excitation and emission wavelengths of
320 nm and 400 nm, respectively. Fractions that fluoresced were
pooled and concentrated.
transformation of
was treated with 2-AP, one Schiff-base derivative product was
that of -Fru itself, and the two others were those of Glc and
D-Fru as an example. In brief, when D-Fru
D
Man. The pyridylamino groups were subsequently removed by
trifluoroacetic acid (TFA) treatment, which generated the free
monosaccharides (Fru, Glc, and Man). After removing TFA by
evaporation, the monosaccharides were separated by normal-
phase HPLC with refractive index (RI) detection. For this report,
we show that this scheme is applicable to the other ketohexoses,
D-Psi,
D-Tag, and D-Sor, and that rare aldohexoses can be system-
atically produced.
Notably, certain rare aldohexoses, especially Gul and Ido, are
not very stable, and therefore, easily undergo undesirable transfor-
mation reaction(s) when stored. To circumvent the need to store
2.3. Calculation of Schiff-base formation efficiency
To measure the yields of the Schiff-base derivatized monosac-
charides, a sample of each purified product was reduced to gener-
ate a stable PA-saccharide as follows. To an aliquot of each
free Gul and Ido, D-Sor, D-Gul, and D-Ido were separated by HPLC
as Schiff-base derivatives. These aldohexoses can be stored as sta-
ble Schiff-base derivatives with the pyridylamino group removed
only when a free aldohexose is needed, which makes the method
valuable from an industrial standpoint.
chromatographic fraction containing
70 L of freshly prepared reducing reagent (100 mg of dimethyl-
amine Á borane, 40 L of AcOH, and 25 L of water) was added.
a Schiff-base derivative,
l
l
l
Each reaction mixture was heated at 80 °C for 35 min and then
was applied to a TSKgel Sugar AX-I column (4.6 Â 150 mm) for an-
ion-exchange HPLC.29 The elution buffer was 0.8 M boric acid–
KOH, pH 9.0, 10% (v/v) acetonitrile. The flow rate was 0.3 mL/
min, and the column temperature was 72 °C. The fluorescence of
the PA-monosaccharides was detected using an excitation wave-
length of 310 nm and an emission wavelength of 380 nm.
Each PA-monosaccharide was quantified by normalizing the
chromatographic peak areas of its R/S C2 isomers19 to that of the
PA-Glc standard for which a chromatogram had been obtained un-
der the same conditions.
2. Materials and methods
2.1. Materials
D
-All and
(Tokyo, Japan).
shi (Tokyo, Japan).
from Wako Pure Chemical Industries (Osaka, Japan).
and COSMOSIL Sugar-D columns (4.6 Â 150 mm, 4.6 Â 250 mm,
D
-Tal were purchased from Tokyo Chemical Industry
-Alt, -Gul, and -Ido were purchased from Funako-
-Fru, 2-AP, and dimethylamine Á borane were
-Gal, -Man,
D
D
D
D
D
D
2.4. Generation of free monosaccharides by TFA hydrolysis
D-Ido
D-Gul
To each mixture containing the Schiff-base derivatives Glc/Man/
Fru, Alt/All/Psi, Tal/Gal/Tag, and Sor/Gul/Ido) was added 50 lL of
D-Gul/Schiff
D-Ido/Schiff
2 M TFA. After each reaction mixture had been heated at its opti-
mized temperatures (see below) for 1 h, the TFA was evaporated.
To confirm that the PA groups had been completely removed from
the monosaccharides, an aliquot of each reaction mixture was sub-
jected to size-fractionation HPLC over a Shodex Asahipak NH2P-50
column (4.6 Â 50 mm, Showa Denko, Tokyo) at 25 °C and at a flow
rate of 0.6 mL/min. The fluorescence of the eluate was monitored
using an excitation wavelength of 310 nm and an emission wave-
length of 380 nm. The gradient system consisted of eluents A and
B. Eluent A was 970:70:3 acetonitrile–water–AcOH titrated to pH
7.0 with 7 M aqueous ammonia, and eluent B was 200:800:3 ace-
tonitrile–water–AcOH titrated to pH 7.0 with 7 M aqueous ammo-
nia. The column was equilibrated with 97% eluent A/3% eluent B.
After loading a sample onto the column, eluent B was [immedi-
ately] changed linearly from 3% to 33% (v) in 3 min and then to
71% (v) in 35 min.
D-Glc
D-Gal
D-Tal
D-Sor
D-Psi
D-Glc/Schiff
D-Gal/Schiff
D-Tal/Schiff
D-Tag
D-Fru
D-Man
D-Man/Schiff
D-Alt/Schiff
D-All/Schiff
D-All
D-Alt
Scheme 1. Preparation of
D
-aldohexoses from D-ketohexoses via the Lobry
rearrangement and Schiff-base formation using 2-AP (solid arrows) followed by
TFA treatment (dotted arrows). The Schiff-base derivative of each monosaccharide
is designated with /’Schiff’.

K. Hasehira et al. / Carbohydrate Research 346 (2011) 2693–2698
2695
CH2OH
HOH2C
CH2OH
C
C
C
C
O
N
C
C
C
C
O
C
C
C
C
N
HO
H
H
HO
H
H
HO
H
H
OH
OH
OH
OH
OH
OH
H
H
H
CH2OH
CH2OH
CH2OH
D-Fru
D-Fru/Schiff
D-Fru
H2N
N
2-AP
(Base catalysis)
H2N
O
H
H
C
C
C
C
C
N
O
H
H
C
C
C
C
C
N
H
H
C
C
C
C
C
OH
H
N
OH
H
2-AP
CH2OH
HCOH
OH
H
HO
(Monoamine coupling)
HO
OH
H
C
C
C
C
O
C
C
C
C
HO
OH
OH
H
H
OH
OH
H
H
HO
H
H
HO
H
OH
OH
H
H
OH
OH
OH
OH
Schiff-base
formation
CH2OH
CH2OH
H
H
TFA treatment
CH2OH
CH2OH
CH2OH
D-Glc
D-Glc
D-Glc/Schiff
D-Fru
Enediol intermediate
O
H
HO
HO
C
C
C
C
C
H
O
H
HO
HO
C
C
C
C
C
N
H
HO
HO
C
C
C
C
C
N
H
H
H
OH
OH
H
H
H
H
OH
OH
OH
OH
H
H
H
H
CH2OH
CH2OH
CH2OH
D-Man
D-Man
D-Man/Schiff
Scheme 2. The reaction scheme for the production of
D-Glc and
D-Man from D-Fru using 2-AP. The transformation precedes via 2-AP base-catalyzed Lobry rearrangement and
2-AP monoamine coupling. The Schiff-base derivatives are converted to the monosaccharides by TFA hydrolysis.
2.5. Separation of free monosaccharides
ketose and the two aldose derivatives (see Scheme 2 with D-Fru as
the reactant). After excess reagents were removed, the mixtures of
Schiff-base derivatives were treated with TFA to generate the free
monosaccharides. The TFA was then removed by evaporation. The
three hexoses in each mixture were separated by normal-phase
HPLC with RI detection.
The Glc/Man/Fru, Alt/All/Psi, Tal/Gal/Tag, and Sor/Gul/Ido mix-
tures were each separated through a COSMOSIL Sugar-D column
(10 Â 250 mm) using an 80% acetonitrile isocratic eluent with a
flow rate of 1.0 mL/min at 25 °C. Elution of the monosaccharides
was detected using a Shimadzu RID-6A RI detector. Fractions con-
taining a purified monosaccharide were pooled. To assess the de-
gree of purification, an aliquot of each fraction was
chromatographed through a smaller COSMOSIL Sugar-D column
(4.6 Â 150 mm) under the aforementioned conditions. Each mono-
saccharide was quantified on the basis of its peak area in compar-
To obtain good yields of the aldoses, the experimental condi-
tions used for Schiff-base formation were first investigated for
D-
Fru. -Fru (10 mol) was dissolved in 20 L of the coupling re-
D
l
l
agent, and the mixture was then heated at 90 °C for 60 min.20,27
Next, excess reagent was removed first by solvent evaporation
and then by TSKgel HW-40F chromatography,21 which completely
separated the Schiff-base derivatives from residual 2-AP and other
lower molecular weight compounds (Fig. 1). Fractions containing
the Schiff-base derivatives were pooled (shown in Fig. 1 as the hor-
izontal bar), but before doing so, an aliquot of each sample was re-
duced to generate the PA-monosaccharides. Next, the reaction
mixture was subjected to anion-exchange HPLC TSKgel Sugar AX-
I chromatography (chromatogram not shown). The yields of the
Schiff-base derivatives were 35% (Fru), 12% (Man), and 8% (Glc). To-
tal product recovery was 53% and the keto–aldose conversion ratio
was 36%.
ison with that of standard D-Glc, which had been chromatographed
under the same HPLC conditions.
2.6. Separation of the Schiff-base-derivatized monosaccharides
The Schiff-base derivatives of the D-Sor-series of hexoses, that
is, Sor, Gul, and Ido, were separated by isocratic elution (95% (v/
v) acetonitrile, 2 mM ammonium acetate, pH 4.0) through a COS-
MOSIL Sugar-D column (4.6 Â 250 mm) with
a flow rate of
1.0 mL/min at 25 °C. The fluorescence of the Schiff-base-deriva-
tized monosaccharides was detected using an excitation wave-
length of 320 nm and an emission wavelength of 400 nm.
Fractions that fluoresced were pooled and then treated with TFA
to generate free monosaccharides.
Next, the reaction temperature and time period were varied.
After the coupling reagent was added to samples of D-Fru (10 lmol
each), the following conditions were assessed: 90 °C, 60 min; 90 °C,
120 min; 100 °C, 60 min; and 100 °C, 120 min. But, yield did not in-
crease even if the temperature and time were changed. The yield
was also not substantially improved by increasing the reaction
time. Therefore, 90 °C and 60 min (standard conditions) were used
for the other three ketose–aldose transformations.
3. Results
3.1. Ketose–aldose transformation catalyzed by 2-AP
Using the standard conditions
mol) were individually reacted with 2-AP. Yields for the
-Psi sample were 10% (Psi), 11% (Alt), and 32% (All), for a total
yield of 53%. Yields for the -Tag sample were 28% (Tag), 16%
D-Psi, D-Tag, and D-Sor (each
10
l
The basic strategy for the preparation of rare aldohexoses is
summarized in Scheme 1. The starting materials, the ketohexoses,
were treated with 2-AP, to produce the Schiff-base derivative of the
D
D

2696
K. Hasehira et al. / Carbohydrate Research 346 (2011) 2693–2698
yield, which substantially decreased the higher temperatures (data
not shown). Therefore the reaction temperature for TFA treatment
of the -Psi derivatives was set to 65 °C. Figure 3A shows a typical
D
chromatogram for the products of the reaction at 65 °C.
The optimum temperatures for TFA treatment of the D-Tag and
D-Sor mixtures were also assessed and found to be 75 °C and 70 °C,
respectively (Fig. 3). Table 1 summarizes the yields of the free
monosaccharides for the four ketoses and the ketose–aldose con-
version ratios. The monosaccharides, including the five rare aldo-
hexoses, D-All, D-Alt, D-Tal, D-Gul, and D-Ido, were each separated
to apparent homogeneity by COSMOSIL Sugar-D chromatography
(Fig. 4).
3.3. Separation of the D-Sor/D-Gul/D-Ido Schiff-base derivatives
0
10
20
40
30
We also explored the possibility that the Schiff-base derivatives
-Sor, -Gul, -Ido could be separated chromatographically be-
Fraction Number (1 mL/tube)
of
D
D
D
cause, although the Schiff-base derivatives are stable, the corre-
sponding free monosaccharides are not. A mixture of the Schiff-
Figure 1. Size-fractionation chromatography for the mixture produced by reaction
of 2-AP and
and developed with 10 mM ammonium acetate (pH 6.0) at a flow rate of 0.1 mL/
min with fluorescence detection, and 1-mL fractions. The fractions marked by the
bar contain the Schiff-base products.
D
-Fru. Separation conditions: 7 Â 500 mm HW-40F column equilibrated
base
derivatives
was
chromatographed
through
the
4.6 Â 250 mm COSMOSIL Sugar-D column with fluorescence detec-
tion. The three Schiff-base derivatives had retention times of
21.6 min (D-Sor), 23.7 min (D-Glu), and 11.1 min (D-Ido), and were
well separated (Fig. 5). To confirm that the separated Schiff-base
derivatives could each be converted into the corresponding free
monosaccharide, after pooling the chromatography fractions of
each Schiff-base derivative, the PA groups were removed using
1 M TFA at 70 °C for 1 h. After removing the TFA by evaporation,
a sample of the reaction mixture was subjected to size-fraction-
ation HPLC to confirm that the PA groups had been removed. Final-
ly, the mixture was chromatographed through the 4.6 Â 150 mm
(Tal), and 11% (Gal), for a total yield of 55%, and those for the
sample were 29% (Sor), 24% (Gul), and 10% (Ido), for a total yield of
63%. The ketose–aldose conversion ratio was the greatest for the
Psi sample, (81%) followed by those for the -Sor sample (54%) and
the -Tag sample (49%).
D-Sor
D
-
D
D
3.2. Liberation of the PA moieties to generate free
monosaccharides by TFA treatment
Next, we searched for the best conditions to use for the conver-
sion of the Schiff-base derivatives to free monosaccharides. To each
of the Schiff-base-derivative mixture was added 50 lL of 2 M TFA
(final concentration, 1 M). The mixtures were heated at 100 °C
for 3 h, after which TFA was removed by evaporation. An aliquot
of each reaction mixture was subjected to size-fractionation HPLC
with fluorescence detection to confirm that no Schiff-base deriva-
tives remained. The fractions corresponding to the retention times
of the Schiff-base derivatives did not fluoresce (data not shown).
The remainder of each mixture containing the free monosaccha-
rides was chromatographed through the 10 Â 250 mm COSMOSIL
Sugar-D column with RI detection. Contrary to our expectation,
1
2
3
4
5
we could identify peaks for only
other monosaccharides were absent. Moreover, the yields for
Glc and -Man were low, 8% and 12%, respectively.
Considering that the molar amounts of -Glc and D-Man were
D-Glc and D-Man. Peaks for the
D
-
D
D
almost the same as those of their Schiff-bases, it seemed that the
acid treatment (1 M TFA, 100 °C, 3 h) destroyed the other Schiff-
base derivatives. Therefore, to produce the rare aldohexoses effi-
ciently, for each Schiff-base derivative mixture, the TFA-treatment
conditions needed to be optimized.
First, the conditions for the
D-Psi derivatives were examined. To
four mixtures of the -Psi Schiff-base derivatives was added 50 lL
D
of 2 M TFA (final concentration, 1 M). Each sample was incubated
at 55, 60, 65, or 70 °C for 1 h. An aliquot of each mixture was chro-
matographed by size-fractionation HPLC with fluorescence detec-
tion (Fig. 2). For the reactions run at 55 and 60 °C fluorescent
peaks were observed. For the reactions performed at the higher
temperatures, essentially no fluorescence materials were chromo-
graphically retarded. Then, portions of the reaction mixture were
subjected to TFA hydrolysis at 65, 70, 75, or 80 °C and then
chromatographed through the 10 Â 250 mm COSMOSIL Sugar-D
column with RI detection. TFA treatment at 65 °C gave the largest
0
10
20
Elution time (min)
Figure 2. Size-fractionation HPLC chromatography of the
derivatives after trifluoroacetic acid treatment. To samples of the Schiff-base
Glc/Man mixture, 50 L of 2 M TFA (final concentration, 1 M) was added. The
samples were held for 1 h at (2) 55 °C, (3) 60 °C, (4) 65 °C, or (5) 70 °C before
chromatography. Chromatogram 1 shows the elution profile of a sample that was
held at room temperature for 1 h. The arrowhead marks the elution position of the
Schiff-base derivatives. The size of the peaks indicates the extent to which the PA
group had been removed.
D
-Fru-derived Schiff-base
D-Fru/
l

K. Hasehira et al. / Carbohydrate Research 346 (2011) 2693–2698
2697
25
20
15
10
5
0
100
50
60
70
80
Temperature (°C)
90
Figure 3. Effect of temperature on the yield of free monosaccharides. The residual fluorescence correlates with the amounts of Schiff-base derivatives that remained after TFA
treatment for 1 h at the indicated temperatures. D-Tag (squares), D-Sor (triangles), and D-Psi (diamonds).
Table 1
Yields of monosaccharides derived from the Lobry de Bruyn–Alberda van Ekenstein rearrangement that used 2-APa
Sor
Fru
Sor
Gul
Ido
Total
63
Conversion ratio
54
Fru
Man
Glc
8 (93)
Tag
Total
53
Conversion ratio
36
Yieldb (%)
Yield (%)
29 (93)
24 (97)
10 (95)
Psi
35 (0)c
12 (95)
Psi
Alt
All
Total
53
Conversion ratio
82
Tag
Tal
Gal
Total
55
Conversion ratio
49
10 (91)
11 (100)
32 (92)
28 (97)
16 (94)
11 (80)
a
Yields were normalized to the amounts of the ketose reactants (10
aldohexoses, for example, -Alt + -All, relative to the total yield, for example,
Numbers in parentheses denote the yield for each monosaccharide after TFA hydrolysis.
l
mol = 100%). The value (%) for the ratio of each keto–aldose conversion is the yield of the two
D
D
D
-Psi + -Alt + -All.
D
D
b
c
D-Fru had not recovered probably because its Schiff base was destroyed during TFA hydrolysis (see text).
D-All
D-Alt
D-Psi
D-Sor/Schiff
A
B
D-Gul/Schiff
D
-Tag
D-Ido/Schiff
D
-Tal
D-Gal
0
30
ElutionTime (min)
60
D-Sor
D-Gul
Figure 5. Separation of the
formation and removal of excess reagents, the Schiff-base derivatives of
Gul/
tography with fluorescence detection. The fractions indicated by the bars were
pooled and treated with TFA.
D
-Sor-series Schiff-base derivatives. After Schiff-base
D-Ido
D-Sor/D-
D
-Ido were separated by COSMOSIL Sugar-D column (4.6 Â 250 mm) chroma-
C
0
30
Elution Time (min)
60
COSMOSIL Sugar-D column. The free monosaccharides had been
purified to apparent homogeneity. We therefore developed a prac-
tical procedure for the preparation of D-Gul and D-Ido, which are
the most unstable of the aldohexoses, by Schiff-base derivatization
with 2-AP and subsequent generation by acid treatment that can
be used only when needed.
Figure 4. Separation of monosaccharides after TFA hydrolysis by a COSMOSIL
Sugar-D HPLC column. Schiff-base derivatives of monosaccharides produced from
(A) D-Psi, (B) D-Tag, and (C) D-Sor were treated with TFA under their optimized
conditions, and the free monosaccharide products were separated by COSMOSIL
Sugar-D column (10 Â 250 mm) chromatography with RI detection. The arrow-
heads indicate the elution positions of authentic samples of the monosaccharides.

2698
K. Hasehira et al. / Carbohydrate Research 346 (2011) 2693–2698
concurrent 3-epimerization, because no significant peak corre-
sponding to -Psi was detectable when 2-AP catalyzed the Lobry
4. Discussion
D
rearrangement. Reaction mechanisms in the presence of 2-AP
and pyridine remain to be clarified.
From a practical viewpoint, it is advantageous that the Schiff
base stabilizes the monosaccharide, especially thermodynamically
For the study reported herein, we developed a chemical proce-
dure enabling systematic preparation of aldohexoses from keto-
hexoses, which uses 2-AP as the base catalyst for the Lobry
rearrangement14 and for Schiff-base formation with the reactant
ketohexose and product aldohexoses. The conditions used for the
ketose–aldose transformation and monoamine coupling were
unstable D-Gul and D-Ido. In a similar attempt, an O-isopropylidene
group has previously been shown to be effective to stabilize these
aldoses.36,37 When we use a larger column (e.g., 20 Â 250 mm), a g-
order of preparation of rare sugars is possible in theory by repeat-
ing the simple separation procedure. Now that these rare sugars
are readily available, their potential industrial and biological appli-
cations, as well as their bioactivities, can be characterized more
readily. These monosaccharides can now also be used as lead com-
pounds with their multiple chiral centers being targeted for the
first time for drug development.
20
acid) and 10
the monosaccharide Schiff-base derivatives of
and -Sor were 53%, 53%, 55%, and 63%, respectively (Table 1).
The ketose–aldose conversion ratio was much greater for -Psi
(82%) than for -Fru (36%), -Tag (49%), and -Sor (54%). Usually,
l
L of coupling reagent (552 mg of 2-AP and 200
mol ketohexose, at 90 °C for 60 min. Total yields of
-Fru, -Psi, -Tag,
lL of acetic
l
D
D
D
D
D
D
D
D
it has been found that the transformation precedes easily via base
catalysis when starting with ketoses and 2-AP. However, when
starting with aldoses, the transformation does not occur readily,
because aldoses are pyridylaminated much more readily than are
ketoses.30,31 Thus, the conversion ratio should depend, at least in
part, on the reaction rate of Schiff-base formation in comparison
with the forward isomerization rate. The detailed mechanism for
the 2-AP-catalyzed Lobry rearrangement is largely unknown, and
its elucidation is an important subject for future studies.
References
1. Matsuo, T.; Suzuki, H.; Hashiguchi, M.; Izumori, K. J. Nutr. Sci. Vitaminol. 2002,
48, 77–80.
2. Livesey, G.; Brown, J. C. J. Nutr. 1996, 126, 1601–1609.
3. Sui, L.; Dong, Y.; Watanabe, Y.; Yamaguchi, F.; Hatano, N.; Tsukamoto, I.;
Izumori, K.; Tokuda, M. Int. J. Oncol. 2005, 27, 907–912.
4. Bautista, D. A.; Pegg, R. B.; Shang, P. J. J. Food Prot. 2000, 63, 71–77.
5. Lawson, C. J.; Homewood, J.; Taylor, A. J. Neurobiol. Learn. Mem. 2002, 77, 17–28.
6. Varma, R.; French, D. Carbohydr. Res. 1972, 25, 71–79.
7. Izumori, K. Naturwissenschaften 2002, 89, 120–124.
8. Granström, T. B.; Takata, G.; Tokuda, M.; Izumori, K. J. Biosci. Bioeng. 2004, 97,
89–94.
9. Itoh, H.; Sato, T.; Izumori, K. J. Ferment. Bioeng. 1995, 80, 101–103.
10. Takeshita, K.; Suga, A.; Takada, G.; Izumori, K. J. Biosci. Bioeng. 2000, 90, 453–
455.
11. Wilson, D. M.; Ajl, S. J. Bacteriol. 1975, 73, 410–414.
12. Leang, K.; Maekawa, K.; Menavuvu, B. T.; Morimoto, K.; Granström, T. B.;
Takada, G.; Izumori, K. J. Biosci. Bioeng. 2004, 97, 383–388.
13. Yoshida, H.; Yamada, M.; Ohyama, Y.; Takada, G.; Izumori, K.; Kamitori, S. J.
Mol. Biol. 2007, 365, 1505–1516.
14. Speck, J. C. Adv. Carbohydr. Chem. 1958, 13, 63–103.
15. Izumori, K.; Miyoshi, K.; Tokuda, S.; Yamada, K. Appl. Environ. Microbiol. 1984,
48, 1055–1057.
16. Takeshita, K.; Shimonishi, T.; Izumori, K. J. Ferment. Bioeng. 1996, 81, 212–215.
17. Itoh, H.; Sato, T.; Takeuchi, T.; Anisur, R. K.; Izumori, K. J. Ferment. Bioeng. 1995,
79, 184–185.
18. Leang, . K.; Maekawa, K.; Menavuvu, B. T.; Morimoto, K.; Granstrom, T. B.;
Takada, G.; Izumori, K. J. Biosci. Bioeng. 2004, 97, 383–388.
19. Hasehira, K.; Nakakita, S.; Miyanishi, N.; Sumiyoshi, W.; Hayashi, S.; Takegawa,
K.; Hirabayashi, J. J. Biochem. 2009, 147, 501–509.
20. Hase, S.; Ikenaka, T.; Matsushima, Y. Biochem. Biophys. Res. Commun. 1978, 85,
257–263.
21. Kuraya, N.; Hase, S. J. Biochem. 1992, 112, 122–126.
22. Klausen, N. K.; Bayne, S.; Palm, L. Mol. Biotechnol. 1998, 9, 195–204.
23. Ohara, K.; Sano, M.; Kondo, A.; Kato, I. J. Chromatogr. 1991, 586, 35–41.
24. Korekane, H.; Tsuji, S.; Noura, S.; Ohue, M.; Sasaki, Y.; Imaoka, S.; Miyamoto, Y.
Anal. Biochem. 2007, 364, 37–50.
25. Suzuki, T.; Matsuo, I.; Totani, K.; Funayama, S.; Seino, J.; Taniguchi, N.; Ito, Y.;
Hase, S. Anal. Biochem. 2008, 381, 224–232.
26. Urashima, T.; Odaka, G.; Asakuma, S.; Uemura, Y.; Goto, K.; Senda, A.; Saito, T.;
Fukuda, K.; Messer, M.; Oftedal, O. T. Glycobiology 2009, 19, 499–508.
27. Hase, S. Methods Enzymol. 1994, 230, 225–237.
28. Nakakita, S.; Natsuka, S.; Okamoto, J.; Ikenaka, K.; Hase, S. J. Biochem. 2005, 138,
277–283.
29. Hase, S.; Hatanaka, K.; Ochiai, K.; Shimizu, H. Agric. Biol. Chem. 1992, 55, 283–
284.
30. Grill, E.; Huber, C.; Oefner, P.; Vorndran, A.; Bonn, G. Electrophoresis 1993, 14,
1004–1010.
31. Schwaiger, H.; Oefner, P. J.; Huber, C.; Grill, E.; Bonn, G. Electrophoresis 1994, 15,
941–952.
The optimum temperatures and reaction times for the TFA
treatment, which liberates the PA groups of the Schiff-base deriv-
atives are: 65 °C, 1 h (
D
-Psi); 75 °C, 1 h (D-Tag), and 70 °C, 1 h (D-
Sor). At less than optimum temperatures, the PA group was not
completely removed, whereas at higher temperatures side reac-
tions occurred destroying the monosaccharides. After HPLC separa-
tion, 1.1
obtained from 10
0.9 mol of -Gal (11%) were obtained from 10
and 2.3 mol of -Gul (16%) and 1.0 mol of -Ido (10%) were ob-
tained from 10 mol of -Sor.
l
mol of
D
-Alt (11%) and 2.9
l
l
mol of
mol of
D
-All (32%) were
-Tal (16%) and
mol of -Tag;
l
mol of -Psi; 1.5
D
D
l
D
l
D
l
D
l
D
l
D
The Lobry rearrangement has often been used to prepare ke-
toses from aldoses, but rarely aldoses from ketoses.14,32,33 H.O.L.
Fischer et al. reported that boiling pyridine was useful as both
the solvent and the base catalyst for ketose preparation34 because
it has a lesser tendency to induce side reactions than do aqueous
alkaline solutions, thereby allowing the aldose–ketose transforma-
tion to dominate with epimerization occurring mainly at the C2
position.34–37 Conversely, ketoses have rarely been used to produce
aldoses with boiling pyridine as the solvent. Moreover, as an exten-
sion of the classic Lobry rearrangement, 3-epimerization of the
ketoses occurs.35–37 Landis reported the transformation of
to
-Psi in boiling pyridine (48 h reaction period35). However, for
that reaction, the ketose–aldose conversion ratio was only 23%,
and the reaction was not very specific because -Glc (13%) and
-Man (10%) were present in addition to the expected -Psi (25%)
-Fru (52%). More recently, Ekeberg and co-workers reported
-Fru in boiling pyridine with aluminum
oxide as a catalyst to improve the reaction yield.37 They obtained
-Glc (22%), -Man (13%), -Psi (10%), and -Fru (55%) after a 9-h
reaction period. Unlike the two aforementioned reports, we found
that under our conditions 3-epimerization of -Fru to -Psi was
D-Fru
D
D
D
D
and
D
the transformation of
D
D
D
D
D
D
D
undetectable. Although classic Lobry rearrangement is considered
to proceed to equilibrium that depends on the relative thermody-
namic stabilities of the monosaccharide reactants and products, for
32. El Khadem, H. S.; Ennifar, S.; Isbell, H. S. Carbohydr. Res. 1987, 169, 13–21.
33. El Khadem, H. S.; Ennifar, S.; Isbell, H. S. Carbohydr. Res. 1989, 185, 51–59.
34. Fischer, H. O. L.; Taube, G.; Baer, E. Ber. Dtsch. Chem. Ges. 1927, 60, 479–485.
35. Landis, W. D. Carbohydr. Res. 1979, 70, 209–216.
example, Glc:Fru:Man (5:4:1),38 our method produced more
D-
Fru (35%) and less D-Glc (8%) than expected. A major difference be-
tween the transformations catalyzed by 2-AP and pyridine is that
2-AP exclusively catalyzes the ketose–aldose conversion with no
36. Ekeberg, D.; Morgenlie, S.; Stenstrøm, Y. Carbohydr. Res. 2005, 340, 373–377.
37. Ekeberg, D.; Morgenlie, S.; Stenstrøm, Y. Carbohydr. Res. 2007, 342, 1992–1997.
38. Angyal, S. J. Top. Curr. Chem. 2001, 215, 1–14.