Appl Microbiol Biotechnol (2013) 97:2961–2970

DOI 10.1007/s00253-012-4184-z

BIOTECHNOLOGICALLY RELEVANT ENZYMES AND PROTEINS

Enzymatic production of 3,6-anhydro-L-galactose

from agarose and its purification and in vitro
skin whitening and anti-inflammatory activities
Eun Ju Yun & Saeyoung Lee & Ji Hye Kim & Bo Bae Kim &
Hee Taek Kim & Sun Hee Lee & Jeffrey G. Pelton &
Nam Joo Kang & In-Geol Choi & Kyoung Heon Kim

Received: 2 February 2012 / Revised: 19 March 2012 / Accepted: 15 May 2012 / Published online: 8 June 2012
# Springer-Verlag 2012

Abstract 3,6-Anhydro- L-galactose (L-AHG) constitutes
50 % of agarose, which is the main component of red macro-
algae. No information is currently available on the mass pro-
duction, metabolic fate, or physiological effects of L-AHG.
Here, agarose was converted to L-AHG in the following three
steps: pre-hydrolysis of agarose into agaro-oligosaccharides by
using acetic acid, hydrolysis of the agaro-oligosaccharides into
neoagarobiose by an exo-agarase, and hydrolysis of neoagar-
obiose into L-AHG and galactose by a neoagarobiose hydro-
lase. After these three steps, L-AHG was purified by
adsorption and gel permeation chromatographies. The final
product obtained was 95.6 % pure L-AHG at a final yield of
4.0 % based on the initial agarose. In a cell proliferation assay,
E. J. Yun : H. T. Kim : S. H. Lee : K. H. Kim (*)
School of Life Sciences and Biotechnology, Korea University,
Seoul 136-713, Korea
e-mail: khekim@korea.ac.kr
S. Lee : I.-G. Choi (*)
Division of Biotechnology, College of Life Sciences and
Biotechnology, Korea University,
Seoul 136-713, Korea
e-mail: igchoi@korea.ac.kr
J. H. Kim : B. B. Kim : N. J. Kang (*)
School of Applied Biosciences, Kyungpook National University,
Daegu 702-701, Korea
e-mail: njkang@knu.ac.kr
J. G. Pelton
Physical Biosciences Division, Lawrence
Berkeley National Laboratory,
Berkeley, CA 94720, USA
J. H. Kim : B. B. Kim : N. J. Kang
School of Food Science and Biotechnology,
Kyungpook National University,
Daegu 702-701, Korea
L-AHG at a concentration of 100 or 200 μg/mL did not exhibit
any significant cytotoxicity. In a skin whitening assay, 100 μg/
mL of L-AHG showed significantly lower melanin production
compared to arbutin. L-AHG at 100 and 200 μg/mL showed
strong anti-inflammatory activity, indicating the significant
suppression of nitrite production. This is the first report on
the production of high-purity L-AHG and its physiological
activities.
Keywords 3,6-Anhydro-L-galactose . Agar . Red
macroalgae . Skin whitening . Anti-inflammation
Introduction
An anhydrosugar is synthesized from a monosaccharide
derivative, such as a glucopyranoside, through the forma-
tion of an intramolecular ether bond by releasing a water
molecule. Anhydrosugars are considered rare, and their
synthesis, metabolism, and biological functions are not
well studied (Kitamura et al. 1991; Kuhn et al. 2006;
Rees 1961; Yamaji et al. 2002). 3,6-Anhydro-L-galactose
(L-AHG) naturally occurs as one the monomers of aga-
rose by alternating galactose and L-AHG units; agarose
is the main component of red macroalgae (Rhodophyta)
(Rees 1961; Sekkal et al. 1993; Su and Hassid 1962). To
date, no information is available on the biological func-
tions and metabolic pathways of L-AHG, as is the case
with other anhydrosugars; therefore, L-AHG is not com-
mercially available even as a reagent. Many marine bac-
teria are capable of degrading and metabolizing agarose,
but the metabolic pathway of L-AHG is almost complete-
ly unknown (Ekborg et al. 2006; Shin et al. 2010; Yun et
al. 2011).
2962
Although the biological activity of L-AHG itself remains
unknown, many biological functions of the oligomers of L-
AHG and D-galactose, namely, agaro-oligosaccharides, have
been reported. These functions include apoptosis induction,
carcinostatic activity, antioxidation, hepatoprotection, im-
munoregulation, anti-allergy, anti-inflammation, and α-
glucosidase inhibition (Chen et al. 2005; Chen et al. 2006;
Tomono et al. 2009). Neoagarobiose (3-O-(3,6-anhydro-L-
galactopyranosyl)-D-galactose), the dimeric form of L-AHG
and galactose, can be used as a novel moisturizer with
whitening effect (Kobayashi et al. 1997). The bioactivities
of the agaro-oligosaccharides are presumed to be primarily
due to the presence of L-AHG (Tomono et al. 2009). There-
fore, it is reasonable to examine the possible skin whitening
and anti-inflammatory effects of L-AHG and thereby to test
its application in the treatment of hyperpigmentation, which
is commonly caused by inflammatory skin disorders (Urabe
et al. 1998). Although a number of whitening agents includ-
ing arbutin are currently used for the treatment of acquired
skin hyperpigmentation, they are only modestly effective,
and some are known to produce adverse side effects (Lee et
al. 2003; Penney et al. 1984).
For the industrial production of agaro-oligosaccharides, the
most common method used is acid hydrolysis at high tempera-
ture because this method is simple, cheap, and rapid and affords
a high yield (Chen et al. 2005; Yang et al. 2009). Acid hydrolysis
leads mainly to the cleavage of α-1,3 glycosidic linkages of
agarose to produce agaro-oligosaccharides. To further degrade
agaro-oligosaccharides into monomeric sugars (i.e., L-AHG and
D-galactose), hydrolysis by a strong acid is required (Jol et al.
1999; Kim et al. 2010a, b). However, such a strong acid can
overdegrade L-AHG and galactose to generate toxic byproducts,
such as 5-hydroxy-methyl-furfural (HMF). In particular, L-
AHG is known to be readily converted into HMF (Kim et al.
2010a, b; Stevenson and Furneaux 1991; Yang et al. 2009).
Therefore, when producing L-AHG from agarose, agar, or red
macroalgae, it would be preferable to rely on enzymatic degra-
dation to prevent the further degradation of L-AHG.
In the enzymatic hydrolysis of agarose, the β-agarase
system, which contains endo-type β-agarases and exo-type
β-agarases to produce neoagaro-oligosaccharides and neo-
agarobiose, respectively, is the major pathway in marine
bacteria, such as Pseudoalteromonas atlantica and Saccha-
rophagus degradans 2-40 (Ekborg et al. 2006; Kim et al.
2010a, b; Vera et al. 1998). Neoagarobiose is finally hydro-
lyzed into L-AHG and D-galactose by neoagarobiose hydro-
lase (NABH), which is known to be rare but was recently
studied by our group (Ha et al. 2011; Lee et al. 2009). Thus,
enzymatic hydrolysis may be a suitable method for the
industrial production of commercially unavailable L-AHG.
To date, crude enzymes of cell-free lysate from S. degradans
2-40 have been applied to hydrolyzing agarose into L-AHG
as the only enzymatic conversion method (Yun et al. 2011).
Appl Microbiol Biotechnol (2013) 97:2961–2970
In this study, we produced L-AHG from agarose by a series
of chemical and enzymatic steps comprising mild pre-
hydrolysis using acetic acid and enzymatic saccharification
by using an exo-type β-agarase (Aga50D) (Kim et al. 2010a,
b) and an NABH extracted from S. degradans 2-40
(SdNABH, formerly AgaJ) (Ha et al. 2011; Lee et al. 2009).
To obtain highly pure L-AHG, the reaction product of agarose
was purified through adsorption and gel permeation chroma-
tographies using silica gel and Bio-Gel P-2 columns, respec-
tively. The structure of purified L-AHG was verified by gas
chromatography–mass spectrometry (GC–MS). We used this
purified L-AHG to perform the first biological study of its
whitening and anti-inflammatory effects. This is the first re-
port on the enzymatic conversion of agarose into L-AHG by
using necessary recombinant enzymes and the purification of
L-AHG for the elucidation of its possible bioactivity.
Materials and methods
Pre-hydrolysis of agarose into agaro-oligosaccharides
To prepare agaro-oligosaccharides from agarose (Intron
Biotechnology, Seongnam, Korea), 5 g of agarose was pre-
hydrolyzed in 100 mL of 3 M acetic acid at 80 °C for 70 min
with agitation at 150 rpm in a water bath. The reaction
product was dried by using a rotary vacuum evaporator at
65 °C for 2 h to remove water and acetic acid. The dried
residues, including acetic acid, were washed with 1.5 L of
94.0 % (v/v) ethanol (Daejung, Siheung, Korea) and filtered
through a membrane filter (pore size, 0.45 μm; Whatman,
Dassel, Germany). The filtered solids were collected as
hydrolyzed agaro-oligosaccharides, and they were dried in
a vacuum drying oven at 60 °C for 24 h.
Enzymatic hydrolysis of agaro-oligosaccharides
Agaro-oligosaccharides obtained from acetic acid hydrolysis
were further hydrolyzed sequentially by a neoagarobiose-
producing β-agarase (i.e., Aga50D) and an NABH (i.e.,
SdNABH). Aga50D and SdNABH both originate from S.
degradans 2-40 (ATCC 43961), and their modes of action have
been published previously (Kim et al. 2010a, b; Lee et al.
2009). The expression of those recombinant enzymes in
Escherichia coli BL21 (DE3) and their purification by affinity
chromatography with a HisTrap HP (GE Healthcare, Piscat-
away, NJ, USA) were performed according to a previously
described method (Kim et al. 2010a, b; Lee et al. 2009). For
enzymatic hydrolysis, agaro-oligosaccharides at a concentra-
tion of 5 % (w/v) were first hydrolyzed by 10 mg of Aga50D in
100 mL of 50 mM Tris–HCl buffer (pH 7.4) at 30 °C for 72 h in
a shaking incubator at 150 rpm. After the enzymatic reaction by
Aga50D, the reaction mixture was centrifuged, and the
Appl Microbiol Biotechnol (2013) 97:2961–2970
supernatant was then hydrolyzed by 2.5 mg of NABH under
the same reaction conditions used for the Aga50D reaction.
One hundred milliliters of the reaction product was concentrat-
ed to 20 mL using a rotary vacuum evaporator at 40 °C for 2 h
and dried with 10 g of Celite 545 (Yakuri, Tokyo, Japan) in a
drying oven at 40 °C for 24 h.
Chromatographic purification of L-AHG
To purify L-AHG from the dried enzymatic reaction
product on Celite, we performed silica gel chromatog-
raphy. Before loading the sample, a column (I.D. 4×
100 cm) packed with silica gel 60 (70-230 mesh
ASTM; Merck, Darmstadt, Germany) was equilibrated
with chloroform. After the dried enzymatic reaction
product was directly applied on top of the column,
elution was performed with 3 L of the mobile phase
(chloroform/methanol/H2O078:20:2, v/v/v), and 20 mL
of each fraction was collected for thin-layer chromato-
graphic (TLC) analysis. Only the fractions which were
verified to contain only AHG without galactose by TLC
were collected (Fig. 1) and dried using a rotary vacuum
evaporator at 35 °C for 2 h to remove the organic
solvent. These fractions were then dissolved in water
at a concentration of 123.0 mg/mL. For further purifi-
cation of L-AHG by size-exclusion chromatography, the
L-AHG solution obtained from the silica gel chromatog-
raphy was loaded onto a Bio-Gel P-2 column (I.D. 2.1×
35.8 cm, Amersham Biosciences, Piscataway, NJ, USA)
equilibrated with water. Each 2 ml fraction was obtained
by elution with water as the mobile phase, and fractions
containing AHG only, as confirmed by TLC, were
collected.
Fig. 1 TLC of the products obtained from each step of agarose
hydrolysis and purification: processing step 1, agarose; processing step
2, acetic acid pre-hydrolysis and washing and drying; processing step
3, enzymatic hydrolysis by Aga50D; processing step 4, enzymatic
hydrolysis by SdNABH; processing step 5, silica gel chromatography;
and processing step 6, Bio-Gel P-2 chromatography. Neoagarobiose
(NAB), D-AHG, and galactose were used as the standards (first and
second lanes)
2963
Thin-layer chromatography of the reaction products
TLC analysis of the enzymatic reaction product was con-
ducted on a silica gel 60 (Merck, Darmstadt, Germany)
plate, and the plate was developed with n-butanol/ethanol/
water (3:1:1, v/v/v) for 1 h. Thereafter, the plate was dried
and visualized using a solution composed of 10 % (v/v)
sulfuric acid and 0.2 % (w/v) naphthoresorcinol (Tokyo
Chemical Industry, Tokyo, Japan) in ethanol at 90 °C for
1 min.
Preparation and purification of neoagarobiose
To obtain neoagarobiose as the standard for TLC and GC–
MS, agarose at a concentration of 10 % (w/v) was hydro-
lyzed into neoagarobiose at 30 °C for 72 h in a shaking
incubator at 150 rpm after adding 50 mg of Aga50D to
100 mL of 50 mM Tris–HCl buffer (pH 7.4). After the
enzymatic reaction by Aga50D, the reaction mixture was
centrifuged, and 50 mL of the supernatant was concentrated
to 2 mL using a rotary vacuum evaporator at 35 °C for 2 h.
The concentrated reaction product was loaded onto a Bio-
Gel P-2 column equilibrated with water for the purification
of neoagarobiose by gel permeation chromatography. Each
fraction of 2 mL was obtained by elution with water as the
mobile phase, and the fractions containing neoagarobiose
only were collected.
Quantification of L-AHG and neoagarobiose by GC–MS
To quantify L-AHG, the previously developed analytical
method using GC–MS was used (Yun et al. 2011). Briefly,
in the range of 50–1000 μg, 3,6-anhydro-D-galactose (D-
AHG; Dextra Laboratories, Berkshire, UK) was derivatized
prior to GC–MS analysis. The aldehyde group of D-AHG
was protected by methoximation by treatment with 50 μL of
2 % (w/v) of methoxyamine hydrochloride in pyridine (Sig-
ma, St. Louis, MO, USA) at 75 °C for 30 min. To increase
the volatility of D-AHG, 80 μL of N-methyl-N-trimethyl-
silyl-trifluoroacetamide (MSTFA; Fluka, St. Louis, MO,
USA) was added to the methoxymized sample. GC–MS
analysis was performed using an Agilent 7890A GC/
5975 C MSD system (Agilent Technologies, Wilmington,
DE, USA) equipped with a DB5-MS column (I.D. 30 m×
0.25 mm, 0.25-μm film thickness; Agilent Technologies).
The GC–MS operating conditions were as follows: The
column temperature was initially set at 100 °C for 3.5 min
and increased to 160 °C at 15 °C/min, at which it was
maintained for 20 min. It was then increased to 200 °C at
20 °C/min and held at 200 °C for 15 min. Finally, the
temperature was increased to 280 °C at 20 °C/min and
maintained at the same temperature for 5 min. The mass
spectra were acquired in a scan range of 50–700 m/z, and
2964
the electron impact and the temperature of the ion source
were 70 eV and 230 °C, respectively. The amount of AHG
was calibrated on the basis of the intensity of its peaks in the
total ion chromatogram. Neoagarobiose was also quantified
by GC–MS by using the earlier method for quantifying L-
AHG. To determine the L-AHG purity of each sample
obtained from the production and purification steps, the
amount of L-AHG in each sample was quantified by GC–
MS and divided by the total dry weight of each sample.
Nuclear magnetic resonance spectroscopy of L-AHG
Two milligrams of L-AHG, which was produced from aga-
rose and purified in this study, was dried in a centrifugal
vacuum concentrator at 30 °C for 6 h, and the dried L-AHG
was dissolved in D2O. 1 H–13 C nuclear magnetic resonance
(NMR) spectroscopy was performed using a Bruker Avance
II 900 MHz NMR spectrometer equipped with a cryoprobe
(Bruker, Billerica, MA, USA). Chemical shifts were
referenced to 3-(trimethylsilyl)-propionic-2,2,3,3-d4 acid.
The chemical structure of L-AHG was confirmed by com-
paring its 1 H and 13 C chemical shifts measured in this study
with its literature values (Sugano et al. 1994) and agarobiose
(4-O-(D-galactosyl)-3,6-anhydro-L-galactopyranose) (Miller
et al. 1982), respectively.
Polarimetry of L-AHG
To verify that the AHG obtained in this study is L-AHG, we
have measured the specific rotation of L-AHG using a polar-
imeter, and it was compared with that of D-AHG and the
reported value of L-AHG. The purified L-AHG (4 mg) was
dissolved in 2 mL of dH2O and left to stand overnight. The
polarimetry of L-AHG was measured with a JASCO P-2000
(JASCO, Easton, MD, USA). The specific rotation, ½a25 D , was
calculated using the following equation: ½a25 D ¼ a = lc, where α
is the observed optical rotation (°), l is the length of a sample
tube (dm), and c is the solution concentration (g/mL). The
specific rotation of L-AHG was compared with that of D-AHG
at a concentration of 0.3 % (w/v) in dH2O as well as the
previously published value of L-AHG as the reference (Araki
and Hirase 1953).
Cell proliferation assay
To estimate the possible cytotoxicity of L-AHG and other
substances tested in this study, cell proliferation assay was
performed. B16F10 cells and RAW264.7 cells were seeded
1×104 cells/well in 96-well plates with 10 % (v/v) FBS/
DMEM at 37 °C in a 5 % CO2 (v/v) incubator. After
culturing for 24 h, 20 μL of 3-(4,5-dimethylthiazol-2-yl)-
5-(3-carboxymethoxyphenyl)-2-(4-sulfonyl)-2 H-
Appl Microbiol Biotechnol (2013) 97:2961–2970
tetrazolium (MTT) solution was added to each well. The
cells were then incubated for 2 h at 37 °C in the CO2
incubator, and the absorbance of the cell culture was mea-
sured at 570 nm. Experiments were conducted in triplicate,
and data were expressed as mean ± standard deviation.
Student'st-test was performed for single statistical compar-
isons. A probability value of p<0.05 was used as the crite-
rion for statistical significance.
Skin whitening activity assay
A murine melanoma cell line, B16F10, was obtained from
Korean Cell Line Bank (KCLB, Seoul, Korea). The cells
were cultured in a Dulbecco’s modified Eagle’s medium
(DMEM; Invitrogen, Auckland, New Zealand) supple-
mented with 10 % (w/v) of heat-inactivated fetal bovine
serum (FBS; Sigma) and 1 % (w/v) of penicillin–streptomy-
cin at 37 °C in a humidified atmosphere with 5 % CO2.
RAW264.7, a mouse macrophage cell line, was obtained
from KCLB and cultured in DMEM supplemented with
2 mM of glutamine, antibiotics such as penicillin A and
streptomycin at concentrations of 100 units/mL each, and
10 % (v/v) of heat-inactivated FBS. RAW264.7 cells were
grown at 37 °C in a fully humidified atmosphere containing
5 % CO2.
The melanin content was measured using a previously
reported method (Tsuboi et al. 1998). Briefly, confluent
monolayers of B16F10 melanoma cells were trypsinized,
and 8×103 viable cells suspended in 3 mL of 10 % FBS
DMEM were added to each well of a 96-well plate. The
plate was incubated at 37 °C in a humidified atmosphere
containing 5 % CO2. When the cells reached 80–90 %
confluence, they were starved by culturing them in serum-
free DMEM for further 12 h. Before being exposed to
100 nM of α-melanocyte-stimulating hormone (α-MSH;
Sigma), the cells were treated with arbutin, neoagarobiose,
galactose, D-AHG, and L-AHG each for 1 h at concentra-
tions of 1, 10, and 100 μg/mL. After 4 days of cultivation
with α-MSH, the culture medium was collected to deter-
mine the relative melanin contents by measuring the absor-
bance at 475 nm with an enzyme-linked immunosorbent
assay (ELISA) reader (Sunrise Basic Tecan, Gröding, Aus-
tria). These experiments were conducted in triplicate, and
data were expressed as mean ± standard deviation. For
single statistical comparisons, Student's t-test was per-
formed, and a probability value of p<0.05 was used as the
criterion for statistical significance.
Anti-inflammatory activity assay
As an indicator of anti-inflammatory activity, nitrite concen-
tration in the RAW264.7 cell culture medium was measured
using Griess reaction (Ding et al. 1988). Cells were plated at
Appl Microbiol Biotechnol (2013) 97:2961–2970 2965

Table 1 Product recovery yield and L-AHG purity at the stages of Results
production from agarose and purification

Processing step Product Recovery yield Purity of Production of L-AHG from agarose
of product based L-AHG
on initial agarose (%, w/w) To obtain L-AHG from agarose, the substrate agarose was
(%, w/w)
sequentially hydrolyzed by acetic acid and enzymes
Initial substrate Agarose 100.0 NA (processing step 1 on the thin layer chromatogram shown
Acetic acid Agaro- NA NA in Fig. 1). The products obtained from each reaction and
pre-hydrolysis oligosacchar- purification step were analyzed by TLC, as shown in Fig. 1.
ides Processing step 2 on the chromatogram in Fig. 1 also
Ethanol Agaro- 76.0 NA
indicates agaro-oligosaccharides with various degrees of
washing and oligosacchar-
drying ides polymerization (DPs), which were obtained by acetic acid
Enzymatic Neoagarobiose 37.6 NA pre-hydrolysis performed at 80 °C for 70 min. After pre-
hydrolysis by hydrolysis using acetic acid, the filtered solids were washed
Aga50D
with ethanol and dried. As seen in Table 1, the total recovery
Enzymatic L-AHG 15.2 18.0
hydrolysis by yield of the filtered, ethanol-washed, and dried solids, which
SdNABH were presumably agaro-oligosaccharides, was as high as
Silica gel L-AHG 7.7 57.5 76.0 % (w/w) of the initial input agarose used in the pre-
chromatogra-
hydrolysis.
phy
Bio-Gel P-2 L-AHG 4.0 95.6 The agaro-oligosaccharides obtained as solids from the
chromatogra- acetic acid hydrolysis product were then hydrolyzed by the
phy exo-type agarase, Aga50D. The enzymatic reaction product
obtained by Aga50D showed a strong band that was pre-
NA not applicable
sumed to represent neoagarobiose in processing step 3 on
the chromatogram in Fig. 1. These results were in good
agreement with our previous finding that Aga50D was an
a density of 1×106 cells/well in a 96-well plate containing exo-type agarase, producing neoagarobiose predominantly
phenol red-free DMEM and incubated overnight. The cells from agarose (Kim et al. 2010a, b). After the enzyme reac-
were treated with 2 μg/mL of lipopolysaccharide (LPS; E. tion using Aga50D, the enzymatic conversion yield was
coli 0127:B8, Sigma) to induce nitrite production in the 49.4 %, and the final recovery yield of neoagarobiose from
absence or presence of testing samples, such as neoagaro- the initial agarose (including the pre-hydrolysis process)
biose, galactose, D-AHG, and L-AHG, at concentrations of was 37.6 % (Table 1).
50, 100, and 200 μg/mL and incubated for 24 h. After The reaction product (mainly neoagarobiose) from the
incubation, the plate was centrifuged at 3,000 rpm for enzymatic hydrolysis by Aga50D was further hydrolyzed
15 min at room temperature to obtain the supernatant of by sdNABH and analyzed by TLC, as shown in Fig. 1. The
the cell culture. The supernatant was analyzed using an monomeric sugars, AHG and galactose, showed strong
ELISA reader for measurement of nitrite concentration in bands in processing step 4 of the TLC (Fig. 1). Thus, we
terms of absorbance at 540 nm. Data were expressed as confirmed that SdNABH cleaved the glycosidic linkage of
mean ± standard deviation of three independent experi- neoagarobiose as we had previously reported (Ha et al.
ments, and Student'st-test was used for single statistical 2011; Lee et al. 2009). As shown in Table 1, the conversion
comparisons. Probability values of p<0.05 were considered yield of the reaction product of agaro-oligosaccharides by
as statistically significant. Aga50D into AHG was 40.5 %, and the purity of L-AHG in

Fig. 2 TLC of each fraction

obtained by silica gel
chromatography, where the
accumulated elution volume of
fractions from the silica gel
column was indicated. Only the
fractions containing L-AHG
were collected (i.e., 0.77–
1.77 L of elution volume) for
further chromatographic
purification
2966 Appl Microbiol Biotechnol (2013) 97:2961–2970

the reaction product by SdNABH was 18.0 %. Overall, it is

expected that, when 100 g of agarose is subjected to acid
pre-hydrolysis and enzymatic hydrolysis, 15.2 g of L-AHG
can be produced.

Purification of L-AHG by chromatography

The purity of L-AHG was only 18.0 % after pre-hydrolysis

and enzymatic hydrolysis; therefore, the obtained L-AHG
needed to be purified before being subjected to structural
analysis by GC–MS and the biological activity tests. First,
we performed silica gel chromatography, and only the frac-
tions that showed L-AHG without galactose on the TLC
(elution volume of 0.77–1.77 mL) in Fig. 2 were collected.
The original purity of L-AHG in the final enzymatic reac-
tion mixture with SdNABH, 18.0 %, increased to 57.5 % by Fig. 3 Total ion chromatograms from the GC–MS analyses of a D-
silica gel chromatography, showing a purification fold of 3.2 AHG (authentic standard) and b L-AHG produced from agarose and
(Table 1). The production and recovery yield of L-AHG purified by silica gel and Bio-Gel P2 size-exclusion chromatographies
after hydrolyses and silica gel chromatography was 7.7 %
on the basis of the initial agarose amount, but considering NMR spectra of purified L-AHG
silica gel chromatography only, the recovery yield of L-
AHG was 50.5 %, as calculated on the basis of the initial The 1 H–13 C NMR spectrum of L-AHG, which was produced
amount of L-AHG in the reaction product produced by from agarose and then purified in this study, is shown in
SdNABH. The partially purified L-AHG sample was sepa- Fig. 4. As shown in Table 2, the 1 H and 13 C chemical shifts
rated into three spots on TLC in processing step 5 of the were compared with those of L-AHG and agarobiose in the
chromatograph in Fig. 1. literature (Sugano et al. 1994; Miller et al. 1982), respectively.
The partially purified L-AHG was further purified by Although the same solvent and reference, D2O and 3-(trime-
size-exclusion chromatography by using Bio-Gel P-2 with thylsilyl)-propionic-2,2,3,3-d4 acid, respectively, were used,
a molecular cutoff size of 1,800 Da. When the fractions our measured 1 H chemical shifts differed −0.1 ppm in com-
showing only the spot of L-AHG were pooled, its purity parison with those of literature values. Since all the proton
of L-AHG was 95.6 %, with a purification fold of 1.7. The peaks exhibited the same shifts in our spectrum, and the 1 H
recovery yield of L-AHG in Bio-Gel P-2 chromatography chemical shifts matched well with the published values, the 1
was 51.8 %. As shown in processing step 6 on the TLC in H chemical shifts of −0.1 ppm were considered only as
Fig. 1, the final purified product showed only the spot of L-
AHG. Consequently, through the hydrolysis and purifica-
tion processes, 4.0 g of L-AHG with a purity of 95.6 %
could be obtained from 100 g of agarose.

GC-mass spectrum of purified L-AHG

To quantify and identify the purified L-AHG, the total ion

chromatograms (TICs) of the purified L-AHG was com-
pared with that of D-AHG, as shown in Fig. 3. Both the
mass spectra of L- and D-AHG matched exactly as they
appeared as one major peak at 20.67 min and two minor
peaks at 24.15 and 27.35 min in the TIC. The fragmentation
of the mass spectra into the major and minor peaks may be
atribtued to the structural instability of anhydrosugars (Chen
et al. 2005; Yang et al. 2009). Since AHG can also be
converted into its hydrate form, L-AHG can be detected in Fig. 4 1H-13C NMR spectrum of the purified L-AHG at 900 MHz.
hydrate form as well as in aldehyde form (Ducatti et al. The purified L-AHG was dissolved in D2O, and 3-(trimethylsilyl)
2011). propionic-2,2,3,3-d4 acid (sodium salt) was used as the standard
Appl Microbiol Biotechnol (2013) 97:2961–2970 2967

Table 2 1H- and 13 C chemical 1 13

shifts from the NMR spectros- H chemical shift (PPM) C chemical shift (PPM)
copy of the L-AHG produced
and purified in this study Measured Published Difference Measured Published Difference

H-1 5.00 5.10 −0.10 92.1 91.4 0.7

H-2 3.61 3.70 −0.10 75.0 74.2 0.8
H-3 3.94 4.02 −0.08 86.0 84.3 1.7
The 1 H and 13 C chemical shifts
of the L-AHG purified in this H-4 4.21 4.30 −0.09 78.8 87.1 −8.3
study were compared with those H-5 4.25 4.34 −0.09 80.2 76.7 3.5
of L-AHG (Sugano et al. 1994) H-6 3.84 3.93 −0.09 74.9 74.5 0.4
and agarobise (Miller et al. H-6′ 4.01 4.11 −0.10
1982) in the literature

referencing issue. Moreover, the 13 C chemical shifts of the
purified L-AHG showed good agreement with the previously
reported values except for C-4 assignment (Miller et al. 1982).
Polarimetry of purified L-AHG
The specific rotations (½a25
D ) of the L-AHG produced and
purified in this study were measured by using a polar-
imeter. As shown in Table 3, the ½a25 D value of L-AHG
was measured as −27.96 mL/dm/g and was distinct from
that of D-AHG, +27.56 mL/dm/g. The ½a25 D value of the
purified L-AHG in this study also agreed well with the
previously reported value (Araki and Hirase 1953).
Cytotoxicity of L-AHG and other saccharides
We evaluated the cytotoxic effects of neoagarobiose, galac-
tose, D-AHG, and L-AHG through proliferation assays us-
ing B16F10 and RAW264.7 cells. Figure 5 shows that the
24-h treatment with neoagarobiose, galactose, D-AHG, and
L-AHG at concentrations up to 100 and 200 μg/mL did not
significantly inhibit the growth of B16F10 cells and
RAW264.7 cells, respectively.
Skin whitening activity of L-AHG
To evaluate the possible skin whitening activity of the L-
AHG obtained and purified in this study, α-MSH-treated
Table 3 Specific rotations of L-AHG and D-AHG
Sample ½a25
D
a
L-AHG −27.96
D-AHG +27.56
NA not applicable
a
½a25
D was expressed in mL/dm/g
b
The specific rotation of L-AHG was compared with that of L-AHG in
the literature (Araki et al. 1953)
B16F10 melanoma cells induced for melanin production
were cultured in the presence of neoagarobiose, galac-
tose, D-AHG (commercial), or L-AHG. For this experi-
ment, arbutin, a well-known skin whitening agent, was
used as the positive control (Maeda and Fukuda 1996).
As shown in Fig. 6, the extracellular melanin concen-
tration of cells treated with 100 μg/ mL L-AHG was
significantly lower than those of the cells treated with
the same concentration of arbutin, neoagarobiose, or D-
AHG. In particular, the extracellular melanin concentra-
tion of the cells treated with 100 μg/mL of L-AHG was
only 23.9 % that of cells treated with 100 nM α-MSH.
Visual examination of the cell culture media revealed
the same results, showing the brightest color in the L-
AHG-treated sample. These results suggested that treat-
ment with L-AHG strongly suppressed melanin produc-
tion in B16F10 melanoma cells.
Anti-inflammatory activity of L-AHG
The cellular nitrite level increases considerably under
inflammatory conditions (Lee et al. 2003). To investi-
gate the possible anti-inflammatory effect of L-AHG,
we measured the nitrite levels in the culture media of
RAW264.7 cells that were stimulated by LPS in the
presence or absence of neoagarobiose, galactose, D-
AHG, or L-AHG (all at 50, 100, and 200 μg/ mL).
Nitrite production was significantly (p<0.05) suppressed
by 100 and 200 μg/ mL L-AHG (Fig. 7). The nitrite
levels in the culture media of the cells treated with
100 and 200 μg/ mL were, respectively, 64.5 and
Reference 38.8 % of those in the LPS-treated controls. D-AHG
showed a nitrite-suppressing effect only at 200 μg/mL,
−25.20b
but its effect was significantly lower than that of L-
NA
AHG. Other saccharides, such as neoagarobiose and
galactose, did not induce any significant reduction in
the nitrite production of RAW264.7 cells. These results
indicate that L-AHG can be used as an anti-
inflammatory agent.
2968
Fig. 5 Cytotoxic effects of neoagarobiose (NAB), galactose, D-AHG,
and L-AHG on a B16F10 cells and b RAW264.7 cells. Cells were
treated with neoagarobiose, galactose, D-AHG, or L-AHG at the indicated
concentrations in 10 % FBS DMEM for 24 h. Proliferation of cells was
Discussion
Strong acids, such as hydrochloric acid, sulfuric acid, and
trifluoroacetic acid, at high temperatures (80–121 °C) are
widely used to prepare agaro-oligosaccharides or AHG from
agarose or agar (Jol et al. 1999; Kim et al. 2010a, b). L-
AHG from agarose is known to be converted to HMF during
hydrolysis by a strong acid due to the instability of its
reducing terminal (Kim et al. 2010a, b; Quemener et al.
1995; Stevenson and Furneaux 1991; Yang et al. 2009). In
the present study, we conducted acid pre-hydrolysis by
using acetic acid at a relatively mild temperature (80 °C)
to minimize the degradation of L-AHG into HMF and other
overreaction products. Aga50D is a unique β-agarase since
most β-agarases release neoagaro-oligosaccharides with
Fig. 6 Effects of arbutin, neoagarobiose (NAB), galactose, D-AHG,
and L-AHG on α-MSH-induced melanin production in B16F10 mel-
anoma cells. Data of three independent experiments were expressed as
mean ± standard deviation. Asterisks indicate a significant inhibition of
melanin content compared to that in the cells treated with α-MSH
alone (p<0.05)
Appl Microbiol Biotechnol (2013) 97:2961–2970
determined by the MTT assay. Results were expressed as cell viability
relative to the untreated control. Data of three independent experiments
were expressed as mean ± standard deviation. Asterisk indicates a signif-
icant difference (p<0.05) compared with the untreated control
various DPs, but Aga50D predominantly forms neoagaro-
biose (Kim et al. 2010a, b). Therefore, it can be considered
an effective enzyme for the production of neoagarobiose or
the generation of L-AHG when combined with an NABH.
The combination of pre-hydrolysis by acetic acid for the
production of agaro-oligosaccharides and enzymatic hydro-
lysis of agaro-oligosaccharides by Aga50D was effective in
producing neoagarobiose from agarose.
Although β-agarases are prevalent in microorganisms, α-
agarases that cleave α-1,3 glycosidic linkages are rare (Fu and
Kim 2010; Hassairi et al. 2001). To produce L-AHG from
agarose or neoagarobiose, the α-1,3 linkages between L-AHG
and galactose should be hydrolyzed by an α-agarase or a
NABH. NABHs that preferentially or solely act on neoagar-
obiose to produce L-AHG have been reported in Pseudomo-
nas atlantica (currently P. atlantica), Vibrio sp., and
Fig. 7 Effects of neoagarobiose (NAB), galactose, D-AHG, and L-
AHG on LPS-induced production of nitrite in RAW 264.7 cells. Data
of three independent experiments were expressed as mean ± standard
deviation. Asterisks indicate significant differences relative to the LPS-
treated group (p<0.05)
Appl Microbiol Biotechnol (2013) 97:2961–2970
Cytophaga flevensis (Day and Yaphe 1975; Sugano et al.
1994; Van Der Meulen and Harder 1976). However, the
industrial application of these NABHs was not possible since
their amino acid sequences were not yet fully discovered.
Overexpresssion of a recombinant NABH was recently per-
formed for the first time, and this NABH originating from S.
degradans 2-40 was characterized to attack the α-1,3 glyco-
sidic linkage from the non-reducing end of a neoagaro-
oligosaccharide and neoagarobiose, resulting in the formation
of predominantly L-AHG (Ha et al. 2011; Lee et al. 2009).
This recombinant NABH is valuable since it can be used as
the key enzyme in the enzymatic saccharification of agarose
into monomeric sugars (L-AHG and galactose) and in the
production of L-AHG from neoagaro-oligosaccharides.
Desirable skin whitening agents should inhibit the secre-
tion of melanin in melanosomes by acting specifically to
reduce the synthesis or activity of tyrosinase; they should also
exhibit low cytotoxicity and no mutagenicity (Dooley 1997).
In particular, the melanin-suppressing effect of L-AHG was
significantly (p < 0.05) higher than that of arbutin. The
melanin-suppressing effect of neoagarobiose, which has been
reported previously (Kobayashi et al. 1997), was also shown
in this study. Our findings of no activity of galactose itself and
the lower activity of neoagarobiose than that of L-AHG on
melanin production provide experimental evidence for the
first time that the bioactivity of neoagarobiose may mainly
be attributable to the L-AHG moiety of neoagarobiose but not
to the galactose moiety.
In conclusion, agarose, which is composed of galactose and
L-AHG, is the main constituent of red macroalgal biomass. In
this study, agrose was pre-hydrolyzed into agaro-
oligosaccharides by acetic acid, and the agaro-oligosaccharides
were hydrolyzed by an exo-type β-agarase and an NABH to
obtain L-AHG. The enzyme reaction product containing L-AHG
was sequentially purified by adsorption and gel permeation
chromatographies. The purified L-AHG exhibited significant
skin whitening and anti-inflammatory activities. This is the first
report on the recombinant enzyme-based production of L-AHG
and its purification and biological activities. These results could
initiate the industrial production of L-AHG from red macroalgae
as an important health-benefiting agent.
Acknowledgments This work was supported by the National Re-
search Foundation (NRF) Grant (2011-0015629) funded by the Korean
Government (MEST) and the Korea University Grant. Facility support
at Korea University Food Safety Hall for the Institute of Biomedical
Science and Food Safety is also acknowledged.
References
Araki C, Hirase S (1953) Studies on the chemical constitution of agar-
agar. XV. Exhaustive mercaptolyses of agar-agar. Bull Chem Soc
Jpn 26:463–467
2969
Chen H-M, Zheng L, Yan X-J (2005) The preparation and bioactivity
research of agaro-oligosaccharides. Food Technol Biotechnol
43:29–36
Chen H, Yan X, Zhu P, Lin J (2006) Antioxidant activity and hepato-
protective potential of agaro-oligosaccharides in vitro and in vivo.
Nutr J 5:1–12
Day DF, Yaphe W (1975) Enzymatic hydrolysis of agar: purification
and charaterization of neoagarobiose hyrolase and ρ-nitrophenyl
α-galactoside hydrolase. Can J Microbiol 21:1512–1518
Ding AH, Nathan CF, Stuehr DJ (1988) Release of reactive nitrogen
intermediates and reactive oxygen intermediates from mouse
peritoneal macrophages: comparison of activating cytokines and
evidence for independent production. J Immunol 141:2407–2412
Dooley TP (1997) Topical skin depigmentation agents: current prod-
ucts and discovery of novel inhibitors of melanogenesis. J Der-
matol Treat 8:275–279
Ducatti DRB, Colodi FG, Goncalves AG, Duarte MER, Noseda MD
(2011) Production of agaro- and carra-oligosaccharides by partial
acid hydrolysis of galactans. Bras J Pharmacogn 21:296–304
Ekborg NA, Taylor LE, Longmire AG, Henrissat B, Weiner RM,
Hutcheson SW (2006) Genomic and proteomic analyses of the
agarolytic system expressed by Saccharophagus degradans 2-40.
Appl Environ Microbiol 72:3396–3405
Fu XT, Kim SM (2010) Agarase: review of major sources, categories,
purification method, enzyme characteristics and applications. Mar
Drugs 8:200–218
Ha SC, Lee S, Lee J, Kim HT, Ko H-J, Kim KH, Choi I-G (2011)
Crystal structure of a key enzyme in the agarolytic pathway, α-
neoagarobiose hydrolase from Saccharophagus degradans 2-40.
Biochem Biophys Res Commun 412:238–244
Hassairi I, Amar RB, Nonus M, Gupta BB (2001) Production and
separation of α-agarase from Altermonas agarlyticus strain
GJ1B. Bioresour Technol 79:47–51
Jol CN, Neiss TG, Penninkhof B, Rudolph B, De Ruiter GA (1999) A
novel high-performance anion-exchange chromatographic meth-
od for the analysis of carrageenans and agars containing 3,6-
anhydrogalactose. Anal Biochem 268:213–222
Kim C, Ryu HJ, Kim SH, Yoon J-J, Kim HS, Kim YJ (2010a) Acidity
tunable ionic liquids as catalysts for conversion of agar into mixed
sugars. Bull Korean Chem Soc 31:511–514
Kim HT, Lee S, Lee D, Kim H-S, Bang W-G, Kim KH, Choi I-G (2010b)
Overexpression and molecular characterization of Aga50D from
Saccharophagus degradans 2-40: an exo-type β-agarase producing
neoagarobiose. Appl Microbiol Biotechnol 86:227–234
Kitamura Y, Abe Y, Yasui T (1991) Metabolism of levoglucosan (1,6-
anhydro-β-D-glucopyranose) in microorganisms. Agric Biol
Chem 55:515–521
Kobayashi R, Takisada M, Suzuki T, Kirimura K, Usami S (1997)
Neoagarobiose as a novel moisturizer with whitening effect. Bio-
sci Biotechnol Biochem 61:162–163
Kuhn A, Yu S, Giffhorn F (2006) Catabolism of 1,5-anhydro-D-fruc-
tose in Sinorhizobium morelense S-30.7.5: discovery, character-
ization, and overexpression of a new 1,5-anhydro-D-fructose
reductase and its application in sugar analysis and rare sugar
synthesis. Appl Environ Microbiol 72:1248–1257
Lee AK, Sung SH, Kim YC, Kim SG (2003) Inhibition of
lipopolysaccharide-inducible nitric oxide synthase, TNF-α and
COX-2 expression by sauchinone effects on I-κBα phosphoryla-
tion, C/EBP and AP-1 activation. Br J Pharmacol 139:11–20
Lee S, Lee JY, Ha SC, Jung J, Shin DH, Kim KH, Choi I-G (2009)
Crystallization and preliminary X-ray analysis of neoagarobiose
hydrolase from Saccharophagus degradans 2-40. Acta Crystal-
logr Sect F-Struct Biolo Cryst Commun 65:1299–1301
Maeda K, Fukuda M (1996) Arbutin: mechanism of its depigmenting
action in human melanocyte culture. J Pharmacol Exp Ther
276:765–769
2970
Miller IJ, Wong H, Newman RH (1982) A 13C N.M.R. study of some
disaccharides from algal polysaccharides. Aust J Chem 35:853–
856
Penney KB, Smith CJ, Allen JC (1984) Depigmenting action of hy-
droquinone depends on disruption of fundamental cell processes.
J Investig Dermatol 82:308–310
Quemener B, Lahaye M, Metro F (1995) Assessment of methanolysis
for the determination of composite sugars of gelling carageenans
and agarose by HPLC. Carbohydr Res 266:53–64
Rees DA (1961) Enzymic synthesis of 3:6-anhydro-L-galactose within
porphyran from L-galactose 6-sulphate units. Biochem J 81:347–
352
Sekkal M, Huvenne J-P, Legrand P, Sombret B, Mollet J-C, Mouradi-
Givernaud A, Verdus M-C (1993) Direct structural identification
of polysaccharides from red algae by FTIR microspectrometry. I:
localization of agar in Gracilaria verrucosa sections. Mikrochim
Acta 112:1–10
Shin MH, Lee DY, Wohlgemuth G, Choi I-G, Fiehn O, Kim KH (2010)
Global metabolite profiling of agarose degradation by Saccharo-
phagus degradans 2-40. New Biotechnol 27:156–168
Stevenson TT, Furneaux RH (1991) Chemical methods for the analysis
of sulfated galactans from red algae. Carbohydr Res 210:277–298
Su J-C, Hassid WZ (1962) Carbohydrates and nucleotides in the red
algal Porphyra perforata. I. Isolation and identification of carbo-
hydrates. Biochem 3:468–474
Sugano Y, Kodama H, Terada I, Yamazaki Y, Noma M (1994) Purification
and characterization of a novel enzyme, α-neoagarooligosaccharide
hydrolase (α-NAOS hydrolase), from a marine bacterium, Vibrio sp.
strain JT0107. J Bacteriol 176:6812–6818
Appl Microbiol Biotechnol (2013) 97:2961–2970
Tomono J, Sagawa H, Kato I (2009) Agarase and gene thereof. US
Patent 7622291 B2
Tsuboi T, Kondoh H, Hiratsuka J, Mishima Y (1998) Enhanced mela-
nogenesis induced by tyrosinase gene-transfer increases boron-
uptake and killing effect of boron neutron capture therapy for
amelanotic melanoma. Pigment Cell Res 11:275–282
Urabe K, Nakayama J, Hori Y (1998) Mixed epidermal and dermal
hypermelanoses. In: Norlund JJ, Boissy RE, Hearing VJ (eds) The
pigmentary system: physiology and pathophysiology. Oxford
University Press, New York, pp 909–911
Van Der Meulen HJ, Harder W (1976) Characterization of neoagarotetra-
ase and neoagarobiase of Cytophaga flevensis. Antonie van Leeu-
wenhoek 42:81–94
Vera J, Alvarez R, Murano E, Slebe JC, Leon O (1998) Identification
of a marine agarolytic Pseudoalteromonas isolate and character-
ization of its extracellular agarase. Appl Environ Microbiol
64:4378–4383
Yamaji K, Sarker KP, Maruyama I, Hizukuri S (2002) Antioxidant
effects of 1,5-anhydro-D-fructose, a new natural sugar, in vitro.
Planta Med 68:16–19
Yang B, Yu G, Zhao X, Jiao G, Ren S, Chai W (2009) Mechanism of
mild acid hydrolysis of galactan polysaccharides with highly
ordered disaccharide repeats leading to a complete series of ex-
clusively odd-numbered oligosaccharides. FEBS J 276:2125–
2137
Yun EJ, Shin MH, Yoon J-J, Kim YJ, Choi I-G, Kim KH (2011)
Production of 3,6-anhydro-L-galactose from agarose by agarolytic
enzymes of Saccharophagus degradans 2-40. Process Biochem
46:88–93