Enzymatic synthesis of 4-hydroxyphenyl β-D-oligoxylosides and their notable tyrosinase inhibitory activity

Article abstract of DOI:10.1271/bbb.80885

We have purified and characterized an oligoxylosyl transfer enzyme (OxtA) from Bacillus sp. strain KT12. In the present study, a N-terminally His-tagged recombinant form of the enzyme, OxtA(H)^E, was overproduced in Escherichia coli and applied

Full text of DOI:10.1271/bbb.80885

Biosci. Biotechnol. Biochem., 73 (5), 1123–1128, 2009
Enzymatic Synthesis of 4-Hydroxyphenyl ꢀ-D-Oligoxylosides
and Their Notable Tyrosinase Inhibitory Activity
y
Kazuhiro CHIKU,1; Hirofumi DOHI,2 Akihiro SAITO,3 Hiroki EBISE,4 Yusuke KOUZAI,3
3
Hirofumi SHINOYAMA,5 Yoshihiro NISHIDA,2 and Akikazu ANDO
1Department of Advanced Bioresource Science, Graduate School of Science and Technology,
648 Matsudo, Matsudo, Chiba 271-8510, Japan
2Department of Bioorganic Chemistry, Graduate School of Horticulture,
648 Matsudo, Matsudo, Chiba 271-8510, Japan
3Department of Nanobiology, Graduate School of Advanced Integration Science,
648 Matsudo, Matsudo, Chiba 271-8510, Japan
4Department of Bioresource Chemistry, Faculty of Horticulture, Chiba University,
648 Matsudo, Matsudo, Chiba 271-8510, Japan
5School of Art and Design, Meisei University, Nagabuchi, Ome, Tokyo 198-8655, Japan
Received December 15, 2008; Accepted February 4, 2009; Online Publication, May 7, 2009
[doi:10.1271/bbb.80885]
We have purified and characterized an oligoxylosyl
transfer enzyme (OxtA) from Bacillus sp. strain KT12.
In the present study, a N-terminally His-tagged re-
combinant form of the enzyme, OxtA(H)E, was over-
produced in Escherichia coli and applied to the reaction
with xylan and hydroquinone to produce 4-hydroxy-
phenyl ꢀ-D-oligoxylosides, ꢀ-(Xyl)n-HQ (n ¼ 1{4), by
one step reaction. The obtained ꢀ-(Xyl)n-HQ inhibited
mushroom tyrosinase, which catalyzes the oxidation of
L-DOPA to L-DOPA quinine, and the IC50 values of
ꢀ-Xyl-HQ, ꢀ-(Xyl)2-HQ, ꢀ-(Xyl)3-HQ, and ꢀ-(Xyl)4-HQ
were 3.0, 0.74, 0.48, and 0.18 mM respectively. ꢀ-(Xyl)4-
HQ showed 35-fold more potent inhibitory activity than
ꢀ-arbutin (4-hydroxyphenyl ꢀ-D-glucopyranoside), of
which the IC50 value was measured to be 6.3 mM.
Kinetic analysis revealed that ꢀ-(Xyl)2-HQ, ꢀ-(Xyl)3-
HQ, and ꢀ-(Xyl)4-HQ competitively inhibited the en-
zyme, and the corresponding Ki values were calculated
to be 0.20, 0.29, and 0.057 mM respectively.
brain. The excessive generation of L-DOPA quinone and
dopamine, which are enzymatically produced from
L-DOPA, can irreversibly alter protein functions, induc-
ing dopaminergic neurodegeneration in the mammalian
brain, and contribute to the neurodegeneration associated
with Parkinson’s disease.4,5) In addition, the browning
reaction on foods is undesired in the food industry.
Therefore, tyrosinase inhibitors are of extensive interest
in medicine, and the cosmetics and food industries.
As popular tyrosinase inhibitors, hydroquinone6,7) and
ꢀ-arbutin (4-hydroxyphenyl ꢀ-D-glucopyranoside)8,9) are
used to prevent deposition of melanin pigments. How-
ever, the hydroquinone is an irritating agent inducing
ochronosis10) and can cause permanent leucoderma in
long-term use.11) Animal studies and in vitro cell studies
have demonstrated that the hydroquinone is a possible
carcinogenic compound.12–14) On the other hand, ꢀ-
arbutin is known as a harmless, water-soluble com-
pound.15) Tyrosinase inhibitors with more effective, less
irritating, and water-soluble properties are still to be
desired.
Recently we isolated a novel enzyme which can
catalyze a ꢀ-oligoxylosyl transfer reaction to polyhydric
phenol.16) This enzyme showed transxylosylation activ-
ity toward catechol and hydroquinone with xylan as
xylosyl donor. We named it OxtA (oligoxylosyl transfer
enzyme A). The N-terminal amino acid sequences of
OxtA showed similarities with endo-1,4-ꢀ-xylanhydro-
lases that belong to the family 11 of the glycosyl
hydrolase classification.17)
Key words: transxylosylation; tyrosinase inhibitor; 4-
hydroxyphenyl ꢀ-oligoxylosides
Tyrosinase (EC 1.14.18.1) is a rate-limiting enzyme in
melanin synthesis that catalyzes the hydroxylation of
L-tyrosine to 3,4-dihydroxyphenyl L-alanine (L-DOPA),
followed by oxidation of L-DOPA to L-DOPA qui-
none.1,2) The product is finally transformed to a black
pigment, melanin, with the aid of both enzymatic and
non-enzymatic reactions. Melanin formation can protect
the skin from ultraviolet-induced skin injury, but
excessive melanin production has hyperpigmenting
effects such as sunburn, freckles, lentigo, and melasma.3)
Melanin pigments are also found in the mammalian
In this study, we constructed a system to overproduce
the N-terminally His-tagged recombinant OxtA protein,
OxtA(H)E, in E. coli and applied the protein to the
production of oligoxylosyl hydroquinones, which were
y
To whom correspondence should be addressed. Tel/Fax: +81-47-308-8871; E-mail: chiku06UD7204@graduate.chiba-u.jp
Abbreviations: DOPA, 3,4-dihydroxyphenylalanine; DP, degree of polymerization; D2O, deuterium oxide; FABMS, fast atom bombardment mass
spectrometry; HPLC, high performance liquid chromatography; HRESIMS, high resolution electrospray ionization mass spectrometry; IC50, 50%
inhibitory concentration; NMR, nuclear magnetic resonance; N-terminal, amino-terminal; OxtA, oligoxylosyl transfer enzyme; OxtA(H)E, His-tagged
recombinant OxtA protein produced in Escherichia coli; TLC, thin-layer chromatography; ꢀ-(Xyl)n-Cat, 2-hydroxyphenyl ꢀ-oligoxyloside; ꢀ-(Xyl)n-
HQ, 4-hydroxyphenyl ꢀ-oligoxyloside; ꢀ-Xyl-HQ, 4-hydroxyphenyl ꢀ-xyloside; ꢀ-(Xyl)2-HQ, 4-hydroxyphenyl ꢀ-xylobioside; ꢀ-(Xyl)3-HQ,
4-hydroxyphenyl ꢀ-xylotrioside; ꢀ-(Xyl)4-HQ, 4-hydroxyphenyl ꢀ-xylotetraoside

1124
K. CHIKU et al.
hydroquinone (Sigma) and 5.0 g of xylan (beach wood, Sigma) and
incubated at 40 ꢀC for 24 h, and the products were periodically
analyzed by HPLC and TLC. For convenience of purification of the
reaction products, the buffer concentration was lowered over that used
for the enzymatic characteristics, which was originally set to 50 mM.
The reaction mixture, incubated for 24 h, was centrifuged at 8;000 ꢁ g
for 20 min, and 400 ml of EtOAc was added to the supernatant. The
water layer was concentrated and lyophilized. The residue containing
xylosylated products was purified by active carbon column chroma-
tography (Wako, Tokyo) and each product was subsequently eluted
with H2O, 10, 20% EtOH, and 40% acetone. Two fractions, eluted
with 20% EtOH and 40% acetone, were collected and concentrated
under diminished pressure. Each residue was dissolved in a small
amount of MeOH and purified by silica gel column chromatography
(gradually eluted by EtOAc:MeOH:H2O = 50:2:1 ! 34:2:1 !
17:2:1 ! 15:2:1 ! 10:2:1) to afford four fractions containing trans-
xylosylated products. Crystallization of two fractions, obtained by
elution with EtOAc:MeOH:H2O = 34:2:1 and 17:2:1, afforded com-
pounds 2 and 3 respectively. The other two fractions, eluted with
EtOAc:MeOH:H2O = 50:2:1 and 15:2:1, were further purified on an
Shodex Asahipak NH2P-50 4E column (4.6 i.d. ꢁ 250 mm, Showa
Denko, Tokyo) using a linear gradient (80–70%) of MeCN in water to
give compounds 1 and 4. The structures of compounds 1–4 were
identified by MS data and 1H-NMR spectra.
designed as candidates for novel tyrosinase inhibitors.
Inhibition assay of the derived 4-hydroxyphenyl ꢀ-
oligoxylosides with tyrosinase revealed that all the
compounds designed possessed a notable inhibitory
effect against a mushroom tyrosinase.
Materials and Methods
Overproduction and purification of the recombinant OxtA protein.
To construct an overproduction system for the recombinant oligox-
ylosyl transfer enzyme OxtA, the coding region of the gene was first
amplified, based on the high identity level (91%) of the N-terminal
amino acid sequence of OxtA with the Bacillus subtilis xylanase
(NC000964).16) A set of primers (50-ATGTTTAAGTTTAAAAAG-
AATTTCTTAGTT-30 and 50-TTACCACACTGTTACGTTAGAAC-
TTCCACT-30) were designed according to the nucleotide sequence of
the Bacillus subtilis xylanase gene. PCR was done using the genomic
DNA of Bacillus sp. strain KT12 as the template. The amplified DNA
fragment was inserted into the plasmid vector pGEMꢀ-T Easy
(Promega, Medison, WI) to obtain plasmid pGEM/oxtA, which was
subjected to analysis of the nucleotide sequence. An expression
plasmid that carries a gene encoding the OxtA protein without the
signal sequence, and tagged with six histidine residues (His-tag) at the
N-terminus, was thereafter constructed as follows: Primers 50-
CTCGAGGCAGCTGGCACAGATTAC-30 and 50-GGATCCTTACC-
ACACTGTTACGTTAG-30, in which the BamHI and XhoI sites
respectively are underlined were designed to amplify the part of the
oxtA gene that encodes the part of the OxtA protein without the
putative signal peptide. The part of the oxtA gene was amplified with
the primer set using pGEM/oxtA as a template. The PCR product was
cloned in plasmid vector pGEM-T Easy to obtain plasmid pGEM/
oxtAR. The nucleotide sequence of the cloned fragment was
ascertained, and the BamHI-XhoI fragment of pGEM/oxtAR was then
ligated to the corresponding sites of pET15b (Novagen, Darmstadt,
Germany). The resulting plasmid, pET15b/OxtA, was introduced into
E. coli BL21(DE3) (Novagen). The recombinant OxtA protein was
overproduced in soluble form and purified in accordance with
Novagen’s instructions. The His-tagged recombinant OxtA protein
obtained was used in this study. The His-tagged protein was designated
OxtA(H)E to distinguish it from the native OxtA protein.
Mass and 1H-NMR analysis of the products. A FABMS spectrum
was obtained on JEOL JMS-AX50 (Jeol, Tokyo). HRESIMS spectra
were recorded in positive ion mode on a Bruker micro TOF (Bruker
Daltonics, Bremen, Germany). 1H-NMR spectra were recorded with a
model JNM-ECA500 (500 MHz, JEOL, Tokyo) at 25 ꢀC, for solution
in D2O with tert-BuOH (1.23 ppm), or (3-trimethylsilyl)-propane
sulfonic acid sodium salt (0 ppm), as the internal standard. The
splitting patterns are reported as d (doublet), t (triplet), m (multiplet),
dd (double of doublets), and dt (double of triplets).
4-Hydroxyphenyl ꢀ-D-xylopyranoside (1). 1H-NMR (500 MHz,
D2O): ꢁ 6.85 and 7.01 (d ꢁ 2, 4H, J ¼ 8:9 Hz, ArH), 4.91 (d, 1H,
J1;2 ¼ 7:3 Hz, H-1), 3.97 (dd, 1H, J4;5e ¼ 5:5 Hz, J5a;5e ¼ 11:5 Hz,
H-5e), 3.65–3.71 (m, 1H, H-4), 3.48–3.54 (m, 2H, H-2 and H-3),
3.38 (t, 1H, J4;5a ¼ 10:8 Hz, H-5a); FABMS: m=z: 242 (Mþ), 281
(Mþ þ K).
4-Hydroxyphenyl ꢀ-D-xylopyranosyl-(1-4)-ꢀ-D-xylopyranoside (2).
1H-NMR (500 MHz, D2O): ꢁ 6.85 and 7.01 (d ꢁ 2, 4H, J ¼ 9:0 Hz,
Enzyme assays. ꢀ-xylanase activity of OxtA toward xylan from
beech wood (Sigma, St. Louis, MO) was determined at 40 ꢀC, as
reported previously.16) The reaction was terminated by heating at
100 ꢀC for 10 min. The amount of xylooligosaccharides released into
the solution was measured by the Somogyi–Nelson method.18) One unit
of hydrolysis activity corresponded to 1 mmol of reducing sugar
(expressed as D-xylose) released per min. The transxylosylation
activity of OxtA toward catechol was determined at 40 ꢀC. The
amount of 2-hydroxyphenyl ꢀ-oligoxylosides was analyzed by
HPLC.16) One unit of transxylosylation activity corresponded to
1 mmol of 2-hydroxyphenyl ꢀ-oligoxylosides (expressed as the total
amount of ꢀ-(Xyl)2{4-Cat) release per min. The buffer systems for
measuring xylanolytic and transxylosylation activities were as follows:
50 mM glycine-HCl buffer (pH 1.3, 2.0, and 3.0), 50 mM sodium citrate
buffer (pH 3.0, 4.0, and 5.0), 100 mM sodium acetate buffer (pH 5.0),
50 mM potassium phosphate buffer (pH 6.0, 7.0, and 8.0), 50 mM Tris–
HCl buffer (pH 7.0, 8.0, 9.0, and 10.0), or 50 mM glycine-NaOH buffer
(pH 10.0, 11.0, 12.0, and 13.5).
0
0
ArH), 4.92 (d, 1H, J1;2 ¼ 7:7 Hz, H-1), 4.45 (d, 1H, J1 ;2 ¼ 7:9 Hz,
H-10), 4.10 (dd, 1H, J4;5e ¼ 5:3 Hz, J5a;5e ¼ 11:9 Hz, H-5e), 3.97 (dd,
1H, J4 ;5 e ¼ 5:4 Hz, J5 a;5 e ¼ 11:9 Hz, H-50e), 3.83 (dt, 1H, J3;4
¼
0
0
0
0
9:0 Hz, J4;5eq ¼ 5:3 Hz, H-4), 3.60–3.66 (m, 2H, H-3 and H-40), 3.53
(t, 1H, J1;2 ¼ J2;3 ¼ 7:7 Hz, H-2), 3.40–3.48 (m, 2H, H-5a and H-20),
3.24–3.32 (m, 2H, H-30 and H-50a); HRESIMS: calcd for C16H22NaO10
(Mþ þ Na): 397.1111, found: m=z 397.1096.
4-Hydroxyphenyl ꢀ-D-xylopyranosyl-(1-4)-ꢀ-D-xylopyranosyl-(1-4)-
ꢀ-D-xylopyranoside (3). 1H-NMR (500 MHz, D2O): ꢁ 6.85 and 7.01
(d ꢁ 2, 4H, J ¼ 8:9 Hz, ArH), 4.92 (d, 1H, J1;2 ¼ 7:7 Hz, H-1), 4.43–
4.48 (d ꢁ 2, 2H, J ¼ 7:6 and 7.7 Hz, H-10 and H-100), 4.09–4.12
(m, 2H, H-5e and H-50e), 3.94–3.98 (dd, 1H, J4 ;5 e ¼ 5:4 Hz,
00 00
J5 a;5 e ¼ 11:6 Hz, H-5’’e), 3.75–3.85 (m, 2H, H-4 and H-40), 3.23–
3.66 (m, 10H, sugar-H); HRESIMS: calcd for C21H30NaO14
(Mþ þ Na): 529.1533, found: m=z 529.1539.
00
00
TLC analysis. TLC was performed on Silica gel-60 plates
(Whatman, Maidstone, UK) and Silica gel-60 F254 (Merck, Darmstadt,
Germany) using EtOAc:AcOH:H2O (3:1:1) as a solvent. Spots were
visualized at 254 nm using a SLUV-6 UV-Handy Lamp (As One,
Osaka) and spraying with 20% sulfuric acid in MeOH, followed by
heating at 120 ꢀC for 7 min.
4-Hydroxyphenyl ꢀ-D-xylopyranosyl-(1-4)-ꢀ-D-xylopyranosyl-(1-4)-
ꢀ-D-xylopyranosyl-(1-4)-ꢀ-D-xylopyranoside (4). 1H-NMR (500 MHz,
D2O): ꢁ 6.85 and 7.02 (d ꢁ 2, 4H, J ¼ 8:9 Hz, ArH), 4.93 (d, 1H,
J1;2 ¼ 7:6 Hz, H-1), 4.43–4.49 (m, 3H, H-10, H-100, and H-1000), 3.23–
4.11 (m, 20H, sugar-H); HRESIMS: calcd for C26H38NaO18
(Mþ þ Na): 661.1956, found: m=z 661.1960.
Preparation of 4-hydroxyphenyl ꢀ-D-oligoxylosides. 4-Hydroxy-
phenyl ꢀ-oligoxylosides, ꢀ-(Xyl)n-HQ, were prepared as follows.
OxtA(H)E (100 U of xylanase activity) was added to 100 ml of 20 mM
potassium phosphate buffer (pH 8.0) containing 5.0 g (4.5 mmol) of
Quantitative analysis of 4-hydroxyphenyl ꢀ-D-oligoxylosides. The
concentration of 4-hydroxyphenyl ꢀ-D-oligoxyloside was determinated
by HPLC. Samples were centrifuged at 8;000 ꢁ g for 10 min at 4 ꢀC,
passed through a 0.20-mm membrane filter, and subjected to an HPLC

Enzymatic Synthesis of 4-Hydroxyphenyl ꢀ-D-Oligoxylosides
system (Hitachi, Tokyo) with an L-7400 type UV detector (Hitachi)
1125
1
2
and an Shodex Asahipak NH2P-50 4E column (4.6 i.d. ꢁ 250 mm,
Showa Denko). The HPLC system was used under the following
conditions: mobile phase, MeCN:H2O (5:1); flow rate, 0.6 ml/min;
absorbance, 280 nm; sample volume, 5 ml; and temperature, 40 ꢀC.
94
66
Tyrosinase inhibition assay. The effects of the purified ꢀ-(Xyl)n-HQ
on a reaction catalyzed by a mushroom tyrosinase were evaluated
in vitro. A series of various concentrations (0.025–10 mM) of ꢀ-(Xyl)n-
HQ was set in a reaction premixture containing L-DOPA (0.25–
1.5 mM) as substrate in 100 mM sodium phosphate buffer (pH 6.8). The
substrate was freshly prepared before the experiment. The reaction was
initiated by adding 30 ml of the enzyme (30 U/reaction) at 25 ꢀC, and
the amounts of L-DOPA quinone produced in the reaction mixture
were determined spectrophotometrically at 475 nm (OD475) at 10 s
intervals. IC50 is the concentration of a drug that inhibits a standard
reaction at the 50% level. Ki values were calculated from shared non-
linear regression fits of substrate-velocity curves taken at different
concentrations of the inhibitors. The data was analyzed using Enzyme
Kinetics Module software (Hulinks, Tokyo).
43
30
20
(kDa)
Fig. 1. SDS–PAGE of the Recombinant OxtA Protein OxtA(H)E.
Lane 1, Protein size markers (phosphorylase b (94 kDa), bovine
serum albumin (66 kDa), ovalbumin (43 kDa), carbonic anhydrase
(20 kDa)); lane 2, the purified OxtA(H)E protein. The gel was
stained with Commassie Brilliant Blue R-250.
Results
Production and purification of recombinant OxtA
The N-terminal amino acid sequence of the OxtA
protein exhibited a striking identity (91%) with the
Bacillus subtilis xylanases belonging to family 11 of the
glycosyl hydrolase classification.16) Based on the high
identity level with the B. subtilis xylanase, we success-
fully amplified the gene for the OxtA protein (see
‘‘Materials and Methods’’ for details). The deduced
amino acid sequence of OxtA without the putative signal
sequence was identical to that of the endo-xylanase
(AAD10834), a family 11 of the glycosyl hydrolase
classification, from Bacillus sp.19) Two glutamic acid
residues that are active centers in family 11 xylanses20)
were conserved in OxtA. The recombinant OxtA protein
OxtA(H)E, which was N-terminally His-tagged, was
overproduced in a soluble form in E. coli and purified
(Fig. 1). The purified OxtA(H)E protein was subjected to
enzymatic characterization, without removing the His-
tag. OxtA(H)E catalyzed the transxylosylation reaction
to catechol (33 U/mg) in 100 mM sodium acetate buffer
(pH 5.0), and hydrolyzed xylan (0.46 U/mg) in the same
buffer. OxtA(H)E was thus found to exhibit transxylo-
sylation and xylanolytic activities, as the native OxtA
did.16) The pH dependence on OxtA(H)E, however,
differed from that on the native OxtA. OxtA(H)E
exhibited the highest transxylosylation activity (79.2
U/mg) in 50 mM potassium phosphate buffer (pH 8.0)
(Fig. 2), whereas the native OxtA exhibited constant
transxylosylation activity in a pH range, 4–11.16) The
N-terminally attached His-tag in the OxtA(H)E might
affect the mode of pH dependence.
100
80
60
40
20
0
1
2
3
4
5
6
7
pH
8
9 10 11 12 13 14
Fig. 2. Effects of pH on the Enzymatic Activity of the OxtA(H)E
Protein.
Transxylosylation (filled symbols) and hydrolytic activities (open
symbols) were measured under standard assay conditions in the
following buffers: and , 50 mM glycine-HCl buffer (pH 1.3, 2.0,
and 3.0); and , 50 mM sodium citrate buffer (pH 3.0, 4.0, and
5.0); and , 50 mM potassium phosphate buffer (pH 6.0, 7.0, and
8.0);
and
and 13.5).
and , 50 mM tris-HCl buffer (pH 7.0, 8.0, 9.0, and 10.0);
, 50 mM glycine-NaOH buffer (pH 10.0, 11.0, 12.0,
1
Compounds 1–4 were identified by MS and H-NMR
spectroscopy. Compound 1 gave a ½M þ Kꢂþ signal at
Synthesis and identification of oligoxylosyl hydro-
quinones by recombinant OxtA
m=z 281 by FABMS. A doublet signals at ꢁ4.9 (J1:2
¼
7:3 Hz, ꢀ-linkage) was observed by 1H-NMR spec-
troscopy. Compound 1 was determined to be 4-hydroxy-
phenyl ꢀ-xyloside21) from the molecular ion peak at m=z
281 ½M þ Kꢂþ obtained by FABMS and a doublet signal
at ꢁ4.9 (J1:2 ¼ 7:3 Hz, ꢀ-linkage) observed by
1H-NMR spectroscopy. The other compounds 2–4
showed no molecular-related ion peak on FABMS,
while their HRESIMS spectra gave ½M þ Naꢂþ signals
at m=z 397.1096, 529.1539, and 661.1960, respectively.
The structures of compounds 2–4 were analyzed by
1H-NMR. In a previous study, 2-hydroxyphenyl ꢀ-
The enzymatic reaction was performed as follows: A
mixture of 5 g of xylan, 5 g of hydroquinone, 100 ml of
20 mM potassium phosphate buffer (pH 8.0), and the
purified OxtA(H)E protein containing 500 U of xylanase
activity was incubated in a shake flask with stirring at
40 ꢀC for 24 h. The xylosylated hydroquinones produced
were purified by two or three chromatographic steps, as
described in ‘‘Materials and Methods,’’ to afford differ-
ent compounds 1 (25 mg), 2 (626 mg), 3 (210 mg), and 4
(39.9 mg) (Fig. 3).

1126
K. CHIKU et al.
O
O
HO
HO
HO
1
OH
HO
O
HO
O
HO
O
HO
O
HO
OH
2
HO
O
O
HO
HO
O
HO
O
O
HO
O
HO
HO
OH
3
HO
HO
O
O
O
HO
O
O
HO
HO
O
O
HO
HO
O
HO
HO
OH
4
Fig. 3. Chemical Structures of 4-Hydroxyphenyl ꢀ-D-Oligoxylosides (Compounds 1–4).
oligoxylosides were fully characterized by spectroscopic
data (1H- and 13C-NMR, DQF-COSY, HMBC, HSQC,
DPFGSE-TOCSY, and DPFGSE-ROESY).22) The data
were applied to the structural determination of com-
pounds 2–4. The 1H-NMR spectra gave peaks in a range
of ꢁ3.2–4.5, characteristic of hexoside ring protons, very
similar to the signals of the 2-hydroxyphenyl ꢀ-
oligoxylosides16) and p-nitrophenyl ꢀ-oligoxylosides.23)
Two doublet signals at ꢁ6.9 and ꢁ7.0 (J ¼ 8:9 Hz) were
assigned to the aromatic protons. A doublet signal at
ꢁ4.9 (J1:2 ¼ 7:3{7:8 Hz, ꢀ-linkage) was assigned to the
anomeric proton of the xylopyranosyl residue at the
reducing end attached to the hydroquinone. The other
doublet signals at ꢁ4.4–4.5 (J1:2 ¼ 7:6{7:9 Hz, ꢀ-link-
age) were assigned to the anomeric protons of ꢀ-
xylosides at the internal and the non-reducing end.
These NMR and MS data confirmed that the structures
of compounds 1–4 were 4-hydroxyphenyl ꢀ-D-xylopy-
ranoside (1, ꢀ-Xyl-HQ), 4-hydroxyphenyl ꢀ-D-xylopy-
ranosyl-(1-4)-ꢀ-D-xylopyranoside (2, ꢀ-(Xyl)2-HQ), 4-
50
40
30
20
10
0
a b c a b c a b c a b c a b c a b c a b c a b c a b c a b c
0.33 12 24 48 72
0
1
3
6
0.17
Reaction time (h)
Fig. 4. Time Course of Transxylosylation to Hydroquinone by the
OxtA(H)E Protein.
Xylosylation of hydroquinone was monitored in 20 mM potassium
phosphate buffer (pH 8.0) containing 5% (w/v) of xylan, 5% (w/v)
of hydroquinone, and the purified OxtA(H)E protein corresponding
to 0.1 (a), 1.0 (b), and 5.0 U/ml (c) of xylanase activity. Reaction
mixtures were incubated at 40 ꢀC for 0–72 h and the products were
periodically analyzed by HPLC. The compositions of the reaction
products are shown as follows: , ꢀ-(Xyl)2-HQ; , ꢀ-(Xyl)3-HQ;
, ꢀ-(Xyl)4-HQ.
hydroxyphenyl
ꢀ-D-xylopyranosyl-(1-4)-ꢀ-D-xylopy-
ranosyl-(1-4)-ꢀ-D-xylopyranoside (3, ꢀ-(Xyl)3-HQ), and
4-hydroxyphenyl ꢀ-D-xylopyranosyl-(1-4)-ꢀ-D-xylopyr-
anosyl-(1-4)-ꢀ-D-xylopyranosyl-(1-4)-ꢀ-D-xylopyrano-
side (4, ꢀ-(Xyl)4-HQ).
Effects of concentration of recombinant OxtA
Inhibition of mushroom tyrosinase by 4-hydroxy-
phenyl ꢀ-D-oligoxylosides
Transxylosylation of hydroquinone with the
OxtA(H)E protein containing 0.1, 1.0, and 5.0 U/ml of
xylanase activity occurred in 5.0 ml of 20 mM potassium
phosphate buffer (pH 8.0). These reaction mixtures were
incubated at 40 ꢀC for 0–72 h, and the enzyme reaction,
yielding ꢀ-(Xyl)2{4-HQ, was periodically analyzed by
HPLC (Fig. 4). The concentration of ꢀ-(Xyl)2{4-HQ
increased concurrently after the OxtA(H)E protein was
added. Production of ꢀ-(Xyl)2-HQ increased sharply
with increasing OxtA(H)E protein concentrations,
although ꢀ-(Xyl)4-HQ decreased slightly with the
OxtA(H)E protein, which showed 5 U/ml of xylanase
activity. These data indicate that ꢀ-(Xyl)2-HQ is hardly
hydrolyzed by the OxtA(H)E protein. After 24 h of
incubation, the yields of ꢀ-(Xyl)2-HQ, ꢀ-(Xyl)3-HQ,
and ꢀ-(Xyl)4-HQ were 33, 13, and 1.4 mM respectively,
and conversion reached 10% based on the amount of
hydroquinone supplied (the transfer rate). Production of
the ꢀ-(Xyl)2{4-HQ by the OxtA(H)E protein was more
than that of ꢀ-(Xyl)2{4-Cat by the native OxtA protein
from Bacillus sp. KT12 (4.2% of the transfer rate).16)
To investigate the biological significance of the
prepared ꢀ-(Xyl)n-HQ, the IC50 and Ki values against
mushroom tyrosinase were measured. The IC50 values of
ꢀ-Xyl-HQ, ꢀ-(Xyl)2-HQ, ꢀ-(Xyl)3-HQ, ꢀ-(Xyl)4-HQ,
and the ꢀ-arbutin were 3.7, 0.74, 0.48, 0.18, and 6.3 mM
respectively (Fig. 5). ꢀ-(Xyl)4-HQ showed 35-fold more
active inhibition than ꢀ-arbutin. According to Line-
weaver-Burk plots, the Ki values of ꢀ-(Xyl)2-HQ and
ꢀ-(Xyl)3-HQ were calculated to be 0.20 and 0.29 mM
respectively, and ꢀ-(Xyl)4-HQ showed the highest
inhibitory activity (Ki ¼ 0:057 mM, Fig. 6, Table 1).
The inhibition types of compounds 1–4 to the enzyme
were competitive.
Discussion
We constructed a system to overproduce the His-
tagged recombinant OxtA, OxtA(H)E, in E. coli, and
applied the protein to the production of oligoxylosyl
hydroquinones. The purified OxtA(H)E showed high

Enzymatic Synthesis of 4-Hydroxyphenyl ꢀ-D-Oligoxylosides
1127
transxylosylation activity. The deduced amino acid
sequence of OxtA(H)E showed a complete similarity to
xylanases belonging to family 11 of glycosyl hydrolase.
The family 11 enzymes catalyze a hydrolysis reaction
by a double displacement mechanism, in which two
glutamic acid residues act at the active center.20) The
catalytic mechanism of the glycosynthases has been
explained in terms of a double displacement mechanism
as well.24–26) The OxtA protein might also catalyze the
transxylosylation reaction in the same manner.
The OxtA(H)E obtained produced ꢀ-(Xyl)n-HQ by the
transxylosylation reaction in one step. OxtA(H)E pro-
duced ꢀ-(Xyl)2-HQ as the major product, together with
ꢀ-(Xyl)3-HQ and ꢀ-(Xyl)4-HQ. The transxylosylation
reaction by OxtA(H)E could be an endo-type reaction
in view of the characteristics of family 11 enzymes,20)
which have a syn-protonation mechanism at the active
site. OxtA(H)E can not hydrolyze ꢀ-(Xyl)2-HQ. This
accords with the fact that family 11 enzymes are
inactive on aryl cellobiosides.20)
100
80
60
40
20
0
0.010
0.10
1.0
10
100
Inhibitor (mM)
Fig. 5. Inhibitory Effects of 4-Hydroxyphenyl ꢀ-D-Oligoxylosides on
Mushroom Tyrosinase.
Tyrosinase activity was measured using 1.0 mM L-DOPA as the
substrate. The results are expressed as the percentages of inhibi-
tion by , ꢀ-Xyl-HQ; , ꢀ-(Xyl)2-HQ; , ꢀ-(Xyl)3-HQ; and
, ꢀ-(Xyl)4-HQ; and , ꢀ-Glc-HQ (ꢀ-arbutin) with respect to the
untreated control. Fifty percent inhibition is indicated by a dotted
line. Inhibitor concentrations are logarithmically plotted.
In the effort to develop water-soluble and more effec-
tive tyrosinase inhibitors, glycosylated hydroquinones
such as ꢀ-arbutin are lead compounds.27–29) For example,
Moon et al.28) synthesized a series of arbutin ꢀ-glucosides
and showed their notable inhibitory activity to a mush-
room tyrosinase. They reported that 4-hydroxyphenyl
A
0.30
0.25
0.20
0.15
0.10
0.05
0
ꢀ-D-glucopyranosyl-(1-4)-ꢀ-D-glucopyranoside (IC50
¼
10 mM) showed lower inhibitory activity than 4-hydrox-
yphenyl ꢀ-D-glucopyranoside (IC50 ¼ 6 mM).27) Our
present data showed the opposite tendency: 4-hydrox-
yphenyl ꢀ-D-oligoxylosides showed stronger inhibitory
activity than 4-hydroxyphenyl ꢀ-D-xyloside. Obviously,
the inhibitory activity increased with the molecular
weight in our case. The ꢀ-(Xyl)4-HQ was found to have a
35-fold more inhibitory effect than arbutin. The strong
activity originates in its high affinity with the enzyme
(a low Ki value). The xylotetraoside moiety of ꢀ-(Xyl)4-
HQ interact with this enzyme to enhance the affinity.
In addition, aryl ꢀ-xylosides are known to penetrate
through cell membranes and to activate a biosynthesis of
chondroitin and dermatan sulfates.30–32) These glycosa-
minoglycans are recognized by various extracellular
matrix proteins, cytokines, growth factors, enzymes, and
inhibitors, and are responsible for diverse biological
functions. Hence, we think that the ꢀ-oligoxylosides
prepared in the present study have much potential for
biomedical applications.
0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
L-DOPA (mM)
B
80
60
40
20
Acknowledgments
-1
0
1
2
3
4
5
1/L-DOPA (mM)
The authors thank Professor Makoto Kiso, Professor
Hideharu Ishida, Dr. Hiromune Ando, Dr. Akihiro
Imamura, and Mr. Kohki Fujikawa of Gifu University
for their kind help in determination of HRESIMS. We
also thank Mr. Takashi Kamijo and Mr. Tomofumi
Yokoyama of Chiba University for technical assistance
in cloning and expression of the oxtA gene.
Fig. 6. Kinetics of Mushroom Tyrosinase Inhibition by ꢀ-(Xyl)4-HQ.
A series of substrate-velocity plots were taken at increasing
concentrations of L-DOPA as substrate in the presence of ꢀ-(Xyl)4-
HQ. ꢀ-(Xyl)4-HQ were added as follows: , 0 mM; , 0.05 mM;
, 0.18 mM; and , 0.36 mM in reaction mixture. Michaelis-Menten
(A) and Lineweaver-Burk (B) plots for the inhibition of mushroom
tyrosinase by ꢀ-(Xyl)4-HQ are shown.
Table 1. Inhibitory Effects of ꢀ-Oligoxylosyl Hydroquinones and ꢀ-Arbutin on a Mushroom Tyrosinase
ꢀ-Xyl-HQ
3.7
—
ꢀ-(Xyl)2-HQ
0.74
0.20
ꢀ-(Xyl)3-HQ
0.48
0.29
ꢀ-(Xyl)4-HQ
0.18
0.057
ꢀ-Glc-HQ
IC50 (mM)
6.3
—
Ki (mM)

1128
K. CHIKU et al.
17) Henrissat B and Bairoch A, Biochem. J., 293, 781–788 (1993).
References
18) Somogyi M, J. Biol. Chem., 195, 19–23 (1952).
19) Jeong K, Lee P, Park I, Kim M, and Kim S, Enzyme Microb.
Technol., 22, 599–605 (1998).
1) Hearing VJ and Jimenez M, Int. J. Biochem., 19, 1141–1147
(1987).
2) Garcia-Borron JC and Solano F, Pigment Cell Res., 15, 162–173
(2002).
20) Collins T, Gerday C, and Feller G, FEMS Microbiol. Rev., 29,
3–23 (2005).
3) Briganti S, Camera E, and Picardo M, Pigment Cell Res., 16,
101–110 (2003).
21) Sulistyo J, Kamiyama Y, Ito H, and Yasui T, Biosci. Biotechnol.
Biochem., 58, 1311–1313 (1994).
4) Asanuma M, Miyazaki I, and Ogawa N, Neurotox. Res., 5, 165–
176 (2003).
22) Chiku K, Shinoyama H, Uzawa J, and Seki H, 24th AOmA
TCMMQ Joint Symposium (in Japanese), 60–63 (2007).
23) Eneyskaya EV, Brumer H3rd, Backinowsky LV, Ivanen DR,
Kulminskaya AA, Shabalin KA, and Neustroev KN, Carbohydr.
Res., 338, 313–325 (2003).
5) Xu Y, Stokes AH, Freeman WM, Kumer SC, Vogt BA, and
Vrana KE, Mol. Brain Res., 45, 159–162 (1997).
6) Arndt K and Fitzpatrick T, J. Am. Med. Assoc., 194, 965–967
(1965).
24) Armand S, Andrews SR, Charnock SJ, and Gilbert HJ,
Biochemistry, 40, 7404–7409 (2001).
7) Palumbo A, d’Ischia M, Misurata G, and Prota G, Biochim.
Biophys. Acta, 1073, 85–90 (1991).
25) Kurakake M, Fujii T, Yata M, Okazaki T, and Komaki T,
Biochim. Biophys. Acta, 1726, 272–279 (2005).
8) Sugai T, Hifu (in Japanese), 34, 522–529 (1992).
9) Tomita K, Fukuda M, and Kawasaki K, Fragrance J., 6, 72–77
(1990).
26) Sinnott ML, Chem. Rev., 90, 1171–1202 (1990).
27) Jun SY, Park KM, Choi KW, Jang MK, Kang HY, Lee SH, Park
KH, and Cha J, Biotechnol. Lett., 30, 743–748 (2008).
28) Moon YH, Nam SH, Kang J, Kim YM, Lee JH, Kang HK,
Breton V, Jun WJ, Park KD, Kimura A, and Kim D, Appl.
Microbiol. Biotechnol., 77, 559–567 (2007).
10) Levin CY and Maibach H, Am. J. Clin. Dermatol., 2, 213–217
(2001).
11) Fernandes D, S. Afr. Med. J., 90, 829 (2000).
12) Whysner J, Verna L, English JC, and Williams GM, Regul.
Toxicol. Pharmacol., 21, 158–176 (1995).
13) Hard G, Whysner J, English JC, Zang E, and Williams G,
Toxicol. Pathol., 25, 132–143 (1997).
29) Funayama M, Arakawa H, Yamamoto R, Nishino T, Shin T, and
Murao S, Biosci. Biotechnol. Biochem., 59, 143–144 (1995).
30) Cheng F, Havsmark B, Sakurai K, Habuchi H, Suzuki S,
Yoshida K, and Fransson LA, Glycoconj. J., 14, 297–305
(1997).
14) Joseph P, Klein-Szanto AJ, and Jaiswal AK, Br. J. Cancer, 78,
312–320 (1998).
15) Maeda K and Fukuda M, J. Pharmacol. Exp. Ther., 276, 765–
769 (1996).
31) Yamaguchi T and Iki M, Agric. Biol. Chem., 50, 2983–2988
(1986).
16) Chiku K, Uzawa J, Seki H, Amachi S, Fujii T, and Shinoyama
H, Biosci. Biotechnol. Biochem., 72, 2285–2293 (2008).
32) Robinson J and Gospodarowicz D, J. Biol. Chem., 259, 3818–
3824 (1984).