Identification of rice Os4BGlu13 as a β-glucosidase which hydrolyzes gibberellin A4 1-O-β-d-glucosyl ester, in addition to tuberonic acid glucoside and salicylic acid derivative glucosides

Article abstract of DOI:10.1016/j.abb.2015.07.021

Abstract Gibberellin 1-O-β-d-glucose ester hydrolysis activity has been detected in rice seedling extracts, but no enzyme responsible for this activity has ever been purified and identified. Therefore, gibberellin A4 glucosyl ester (GA₄-GE) β-d-glucosidase activity was purified from ten-day rice seedling stems and leaves. The family 1 glycoside hydrolase Os4BGlu13 was identified in the final purification fraction. The Os4BGlu13 cDNA was amplified from rice seedlings and expressed as an N-terminal thioredoxin-tagged fusion protein in Escherichia coli. The purified recombinant Os4BGlu13 protein (rOs4BGlu13) had an optimum pH of 4.5, for hydrolysis of p-nitrophenyl β-d-glucopyranoside (pNPGlc), which was the best substrate identified, with a k_cat/K_m of 637 mMmM^{-1 -1}. rOs4BGlu13 hydrolyzed helicin best among natural glycosides tested (k_cat/K _m of 74.4 mM^{-1 -1}). Os4BGlu13 was previously designated tuberonic acid glucoside (TAG) β-glucosidase (TAGG), and here the k_cat/K_m of rOsBGlu13 for TAG was 6.68 mM^{-1 -1}, while that for GA₄-GE was 3.63 mM^{-1 -1} and for salicylic acid glucoside (SAG) is 0.88 mM^{-1 -1}. rOs4BGlu13 also hydrolyzed oligosaccharides, with preference for short β-(1 → 3)-linked over β-(1 → 4)-linked glucooligosaccharides. The enzymatic data suggests that Os4BGlu13 may contribute to TAG, SAG, oligosaccharide and GA₄-GE hydrolysis in the rice plant, although helicin or a similar compound may be its primary target.

Full text of DOI:10.1016/j.abb.2015.07.021

Archives of Biochemistry and Biophysics 583 (2015) 36e46
Contents lists available at ScienceDirect
Archives of Biochemistry and Biophysics
journal homepage: www.elsevier.com/ locate/yabbi
Identification of rice Os4BGlu13 as a
gibberellin A4 1-O- -glucosyl ester, in addition to tuberonic acid
glucoside and salicylic acid derivative glucosides
b-glucosidase which hydrolyzes
b-D
Yanling Hua a, b, Watsamon Ekkhara , Sompong Sansenya , Chantragan Srisomsap ,
b
,
c
d
e
f
g
g
g
Sittiruk Roytrakul , Wataru Saburi , Ryosuke Takeda , Hideyuki Matsuura ,
Haruhide Mori , James R. Ketudat Cairns
The Center for Scientific and Technological Equipment, Suranaree University of Technology, Nakhon Ratchasima 30000, Thailand
Center for Biomolecular Structure, Function and Application, Suranaree University of Technology, Nakhon Ratchasima 30000, Thailand
School of Biochemistry, Institute of Science, Suranaree University of Technology, Nakhon Ratchasima 30000, Thailand
Department of Chemistry, Faculty of Science, Rajamangala University of Technology, Thanyaburi, Pathun Thani 12110, Thailand
Chulabhorn Research Institute, Bangkok 10210, Thailand
g
b, c, e, *
a
b
c
d
e
f
National Center for Genetic Engineering and Biotechnology, Pathum Thani 12120, Thailand
Research Faculty of Agriculture, Hokkaido University, Kita-9, Nishi-9, Kita-ku, Sapporo 060-8589, Japan
g
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 11 June 2015
Received in revised form
Gibberellin 1-O-
enzyme responsible for this activity has ever been purified and identified. Therefore, gibberellin A4
glucosyl ester (GA -GE) -glucosidase activity was purified from ten-day rice seedling stems and
leaves. The family 1 glycoside hydrolase Os4BGlu13 was identified in the final purification fraction. The
Os4BGlu13 cDNA was amplified from rice seedlings and expressed as an N-terminal thioredoxin-tagged
fusion protein in Escherichia coli. The purified recombinant Os4BGlu13 protein (rOs4BGlu13) had an
b-D-glucose ester hydrolysis activity has been detected in rice seedling extracts, but no
4
b-D
29 July 2015
Accepted 30 July 2015
Available online 1 August 2015
optimum pH of 4.5, for hydrolysis of p-nitrophenyl
b
-
D
-glucopyranoside (pNPGlc), which was the best
. rOs4BGlu13 hydrolyzed helicin best among natural
ꢀ1
s ). Os4BGlu13 was previously designated tuberonic acid
Keywords:
ꢀ
1 ꢀ1
substrate identified, with a kcat/K
glycosides tested (kcat/K of 74.4 mM
glucoside (TAG) -glucosidase (TAGG), and here the kcat/K
while that for GA
rOs4BGlu13 also hydrolyzed oligosaccharides, with preference for short
linked glucooligosaccharides. The enzymatic data suggests that Os4BGlu13 may contribute to TAG, SAG,
m
of 637 mM
s
b-glucosidase
ꢀ
1
Rice
m
ꢀ
1
ꢀ1
Gibberellin glucosyl ester
Recombinant expression
Phytohormone conjugates
b
m
of rOsBGlu13 for TAG was 6.68 mM
s ,
s .
ꢀ
1
ꢀ1
ꢀ1 ꢀ1
4
-GE was 3.63 mM
s
and for salicylic acid glucoside (SAG) is 0.88 mM
-(1 / 3)-linked over -(1 / 4)-
b
b
oligosaccharide and GA
its primary target.
4
-GE hydrolysis in the rice plant, although helicin or a similar compound may be
©
2015 Elsevier Inc. All rights reserved.
1. Introduction
conjugates are inactive reserve forms of active phytohormones in
plants, which can be hydrolyzed to release active phytohormones
Phytohormones serve as chemical messengers to regulate
by b-glucosidases. The b-glucosidases that hydrolyze these conju-
cellular activities, vegetative and reproductive development, and
stress responses. These plant hormones include auxins, cytokinins,
gibberellins, ethylene and abscisic acid (ABA), brassinosteroids,
strigolactones, jasmonates and salicylates. Many of these phyto-
hormones can be conjugated to glucose. Phytohormone glucosyl
gates have been found in several plants [1e4].
The gibberellins (GAs) are a class of plant hormones that play
important roles in various plant growth phenomena, including
seed germination, stem elongation, and flower development. They
are highly functionalized diterpenoids and are distributed widely in
plants, fungi and bacteria [5e8]. Among the 136 known GAs, only a
few of them are biologically active in plants. These include GA
1 3
, GA ,
GA , GA , GA and GA . Gibberellic acid (GA ), which was the first
gibberellin to be structurally characterized, has been widely used to
regulate plant growth and development.
4
5
6
7
3
*
Corresponding author. School of Biochemistry, Institute of Science, Suranaree
University of Technology, Nakhon Ratchasima 30000, Thailand.
E-mail address: cairns@sut.ac.th (J.R. Ketudat Cairns).
http://dx.doi.org/10.1016/j.abb.2015.07.021
0
003-9861/© 2015 Elsevier Inc. All rights reserved.

Y. Hua et al. / Archives of Biochemistry and Biophysics 583 (2015) 36e46
37
The bioactive GAs exist in the plants together with many inac-
tive GAs and their glucosyl conjugates, which may be inactive
precursors or inactivation products of the active forms, and are
closely regulated by the GA biosynthetic and catabolic pathways
[3,17]. Recently, two closely related rice glycoside hydrolase
family 1 -glucosidase isoenzymes, Os4BGlu12 and Os4BGlu13,
were found to hydrolyze tuberonic acid beta-glucoside (TAG,
b
Fig. 1a), and were designated TAG
TAGG2 [4,22].
b-glucosidase 1 (TAGG1) and
[9]. Both 1-O-acyl glucosyl esters and glucosides of GAs are found in
plants. It has been suggested that GA glucosyl esters are deactivated
GAs that can be enzymatically reconverted to active GAs, thus
serving as a reserve form of biologically active GAs [10]. The me-
Although b-glucosidases have been proposed to hydrolyze the
GA conjugates to active GAs, the molecular identification of the
responsible enzymes and investigation of their modes of action
13
13
13
13
tabolites,
GA29-2-O-glucoside, [ C]GA
identified in the extracts of the seedlings made 24 h after the in-
jection of [13C]GA20
-glucosyl ester into light-grown maize
[
C]GA20
,
[
C]GA29
,
[
C]GA20-13-O-glucoside,
[
C]
have yet to be reported. Schliemann [1] reported that
dases extracted from mature rice seeds and seedlings have different
hydrolytic activities toward GA -2-O-glucoside, GA -3-O-glucoside
and 1-O-GA -glucosyl ester, but purification and characterization of
the -glucosidase activities was not performed. In this study, we
extracted a family 1 glycoside hydrolase, Os4BGlu13, which has
relatively high activity to hydrolyze GA -GE (Fig. 1b). We also
b-glucosi-
1
3
13
8
and [ C]GA -2-O-glucoside were
8
8
3
-
b-
D
3
seedlings [11]. This showed that the endogenous hydrolysis of the
introduced conjugate and its reconjugation led to the three new
b
3
3
3
glucosides. In rice (Oryza sativa), [ H]GA1, [ H]GA2, [ H]GA34, the
4
3
3
3
3
glucosides of [ H]GA
2
, [ H]GA
4
, [ H]GA
8
and [ H]GA34, and the
expressed Os4BGlu13 in Escherichia coli, purified the expressed
rOs4BGlu13 and characterized its hydrolytic activities to natural
and synthetic glycosides.
3
glucosyl ester of [ H]GA
4
(GA -GE) have been found after applica-
4
3
tion of [ H]GA
Beta-glucosidases (
.2.1.21) are essential enzymes in all living organisms and play-
ing important roles in biomass conversion in microorganisms
13], glycolipid metabolism in animals [14], and activation of
4
to cell suspension cultures of cv. nipponbare [12].
b-D
-glucopyranoside glucohydrolases, E.C.
3
2
. Materials and methods
[
2
.1. Materials
defense compounds [15,16], phytohormones [17,18], lignin pre-
cursors [19], aromatic volatiles [20], and metabolic intermediates
Rice (O. sativa cv. Suphan buri 1) was grown in a field in Sikiu
[
21] in plants. Beta-glucosidases hydrolyze the
bond at a non-reducing terminal -glucosyl moiety through a
double-displacement mechanism to release -glucose from
b-O-glycosidic
district, Nakhon Ratchasima province, Thailand, during December,
011. After 10 days, seedling stems and leaves were collected and
used for protein purification.
GA -GE was synthesized from GA
neering Co.Ltd, P. R. China) following the previously described
methods [23,24]. The acetylated and deacetylated GA -GE were
D
2
b-D
glycosides and oligosaccharides. In doing so, they can release
glucose blocking groups from their inactive glucosides in plants.
4
4
(Jiangsu Fengyuan Bioengi-
To achieve specificity for these various functions, different
b-
4
glucosidases must bind to distinct aglycones with a wide variety
of structures, in addition to the glucosyl moiety of the substrate.
These substrates may include phytohormone glycoconjugates,
such as cytokinin glucoside and abscissic acid 1-O-glucose ester
confirmed by NMR spectra on a 300 MHz NMR spectrometer with a
Varian 300 ID/PFG probe at a frequency of 299.986 MHz (Unity
INOVA, Varian, USA). Tuberonic acid
b-D-glucopyranoside (TAG)
and salicylic acid -glucoside (SAG) were synthesized as
b-D
Fig. 1. Reactions of phytohormone
b
-glucosidases of interest. A) Tuberonic acid -glucoside is hydrolyzed by TAG
b
b-glucosidase 1 (TAGG1) to release tuberonic acid and glucose. B)
4
GA -GE is hydrolyzed by gibberellin glucose ester
b
-glucosidase to release GA and glucose.
4

38
Y. Hua et al. / Archives of Biochemistry and Biophysics 583 (2015) 36e46
previously described [25,26]. Other glycoside substrates were
purchased from SigmaeAldrich (St. Louis, MO, USA).
with buffer A by centrifugal filtration, as described above, before
they were tested for pNPGlc and GA -GE hydrolysis.
4
Finally, the active fractions were pooled and loaded to a
Superdex-200 gel-filtration column (10/300, 24 ml, GE Healthcare)
which was equilibrated with buffer C, and eluted with the same
2
.2. Extraction and purification of
b-glucosidase from 10-day-old
rice seedlings
4
buffer. The protein concentrations and GA -GE hydrolysis activities
Ten kilograms of 10-day-old rice seedling stems and leaves were
cut to small pieces with a blender, and the protein was extracted
with McIlvaine buffer (0.1 M citric acide0.2 M disodium hydrogen
of the fractions were measured, and their purity checked with SDS-
PAGE.
phosphate (Na
2
HPO
4
), pH 5.0) supplemented with 1 mM phenyl-
2.3. Identification of b-glucosidases with LC-MS
ꢁ
methylsulfonyl fluoride (PMSF) at 4 C overnight. The ratio of
seedlings and buffer was 100 g per 600 ml. A crude extract was
The protein from the Superdex-200 column chromatography
was separated by 8% SDS-PAGE, and the gel stained with Coomassie
brilliant blue R-250. The two main Coomassie-brilliant-blue-
stained bands were excised separately and destained with 25 mM
obtained by filtering and centrifugation at 12,000 ꢂ g for 20 min at
ꢁ
4
C. The protein was precipitated with 80% saturated ammonium
4 2 4
sulfate (NH ) SO and the pellet was collected by centrifugation.
The protein pellet was redissolved in 4-fold diluted McIlvaine
buffer, pH 5, (buffer A) and dialyzed overnight against buffer A
twice, then centrifuged to remove precipitate. The supernatant
4 3
ammonium bicarbonate (NH HCO )/50% methanol (v/v) and
washed with water followed by 100% acetonitrile (ACN). Then the
gel plugs were dried at room temperature and disulfide bonds were
was tested for hydrolysis activities toward p-nitrophenylꢀ
glucopyranoside (pNPGlc) and GA -GE. The dialyzed protein was
loaded onto a CM-Sepharose fast flow column (HiPrep CM FF 16/
0, 20 ml, GE Healthcare, Little Chalfont, UK) equilibrated with
buffer A. The unbound proteins were washed out with buffer A
and the bound proteins were eluted with a linear gradient of
b
-
D
-
reduced by reaction with 10 mM dithiothreitol in 10 mM NH
After removal of the dithiothreitol solution, 20 l of 100 mM
iodoacetamide in 10 mM NH HCO was added to the gels, which
were then kept in the dark at room temperature for 1 h, and
washed twice with 200 l of 100% ACN.
The gels were infiltrated with twenty microliters of 10 ng/
4
HCO3.
4
m
4
3
1
m
ml
0
e1.0 M sodium chloride (NaCl) in buffer A at a flow rate of 2.0 ml/
min (200 ml total). The fractions were collected and tested for
hydrolysis activities toward pNPGlc and GA -GE. The fractions
with activities to GA -GE were pooled and precipitated with
trypsin (Promega, sequencing grade) and incubated at room tem-
perature for 20 min, then at 37 C for 3 h. Thirty microliters of 50%
ꢁ
4
ACN/0.1% formic acid (v/v) was added to the sample, and the
resulting peptides were extracted by shaking at room temperature
4
ꢁ
(
NH
4
)
2
SO
4
(80% saturation), followed by centrifugation at
for 10 min. The extracted solution was dried at 40 C overnight. The
ꢁ
ꢁ
1
6,000 ꢂ g for 20 min at 4 C, then dialyzed against 20 mM
samples were kept at ꢀ80 C until analysis.
Tris(hydroxymethyl) aminomethane hydrochloride (TriseHCl),
pH 7, containing 0.5 M NaCl (buffer B).
The separation of tryptic peptides was performed with a
NanoAcquity system (Waters Corp., Milford, MA, USA) equipped
The dialyzed protein was loaded onto a packed Con A-Sepharose
with a Symmetry C18
5
m
m, 180
m
m ꢂ 20 mm trap column and a
4
B column (C10/20, 10 ml, GE Healthcare) equilibrated with buffer
B. The column was washed with 4 column volumes (CV) of buffer B,
and then eluted with 4 CV of 0.5 M -mannose in buffer B. The
BEH130C18 1.7 m, 100 m
m
m ꢂ 100 mm analytical reverse phase
column (Waters Corp., Milford, MA, USA). Analysis of tryptic pep-
tides was performed on a SYNAPT™ HDMS mass spectrometer
(Waters Corp., Manchester, UK). All analyses were performed in
positive ion nanoelectrospray mode. The peptides were separated
with a gradient of 15e50% acetonitrile in 0.1% formic acid over
15 min at a flow rate of 600 nl/min, then the column was washed 3-
min with 80% acetonitrile, 0.1% formic acid. The column tempera-
D
collected fractions were tested for hydrolysis of pNPGlc and the
active fractions were combined and concentrated with centrifugal
filters (Amicon Ultra, regenerated cellulose, 30,000 MWCO) at
ꢁ
7
00 ꢂ g, 4 C. The buffer of the concentrate was exchanged twice
with 4-fold diluted McIlvaine buffer, pH 7, containing 0.2 M NaCl
buffer C), before testing activity with GA -GE and further purifi-
cation by gel-filtration chromatography.
ꢁ
(
4
ture was maintained at 35 C. The time-of-flight analyzer of the
mass spectrometer was externally calibrated with [Glu ]fibrino-
1
The concentrated protein obtained by the Con A-Sepharose
column chromatography was loaded to a Superdex-75 gel-filtration
column (XK 26/40, 150 ml, GE Healthcare) equilibrated with buffer
C on an AKTA Protein Purifier system (GE Healthcare), and eluted at
a flow rate of 1.0 ml/min. The fractions were tested for pNPGlc and
peptide B from m/z 50 to 1600 with acquisition lock mass corrected
using the monoisotopic mass of the doubly charged precursor of
1
[Glu ]fibrinopeptide B. The MS/MS survey was over the range of
€
50e1990 Da and the scan time was 0.5 s. The peptide mass results
were used in a MASCOT search of the Genbank non-redundant (nr)
protein database.
4
GA -GE hydrolysis activities and their purities checked with so-
dium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-
PAGE). The active fractions were pooled and concentrated, and then
the buffer was exchanged with buffer A.
2.4. Determination of b-glucosidase activity from protein fractions
The concentrated protein from the Superdex-75 gel-filtration
column was loaded onto a 1 ml HiTrap SP Sepharose XL column (GE
Healthcare) equilibrated with buffer A on an AKTA Protein Purifier
The activities of protein fractions to hydrolyze pNPGlc were
tested in a manner similar to previously published methods
[27e30]. Aliquots of enzyme solutions were incubated with 4 mM
€
system. The column was eluted with a linear gradient of 0e1.0 M
NaCl in buffer A at a flow rate of 1.0 ml/min. The fractions with
hydrolysis activities were concentrated and the buffer exchanged
with 50 mM sodium phosphate buffer, pH 7, containing 1.7 M
pNPGlc in 50 mM sodium acetate (NaOAc) buffer, pH 5.0, (total
ꢁ
reaction volume 50
ml) at 30 C for 20 min. The reactions were
stopped by adding 150
2 3
ml of 2 M sodium carbonate (Na CO ). The
released p-nitrophenol (pNP) was quantified by measuring the
absorbance at 405 nm (A405) with a microplate reader (Thermo
Labsystems, Helsinki, Finland), and comparing it to that of a pNP
(
NH
4 2 4
) SO (buffer D), by centrifugal filtration as described above.
The protein from SP chromatography was loaded onto a 1 ml
HiTrap octyl FF column (Sepharose 4, GE Healthcare), which was
pre-equilibrated with buffer D. The protein was eluted with a linear
4
standard curve. The hydrolysis of GA -GE was determined by a
similar reaction in 50 mM NaOAc, followed with a peroxidase/
glucose oxidase-based glucose assay (PGO assay, SigmaeAldrich),
as previously described [24]. In all cases, control reactions
4 2 4
gradient of 1.7e0 M (NH ) SO in 50 mM sodium phosphate buffer,
pH 7.0. The buffer of the collected fractions was exchanged twice

Y. Hua et al. / Archives of Biochemistry and Biophysics 583 (2015) 36e46
39
incubated without
concentrations were determined with a Bio-Rad Bradford assay
with bovine serum albumin (BSA) as a standard.
b
-glucosidase enzyme served as blanks. Protein
concentrated and depleted of imidazole by centrifugal ultrafiltra-
tion (Amico Ultra, 30k MWCO) at 700 ꢂ g, 4 C.
ꢁ
2.7. Determination of pH optimum for Os4BGlu13
2.5. Cloning of Os4BGlu13
The optimum pH of rOs4BGlu13 hydrolysis of pNPGlc was
determined by incubating 0.4
mg of enzyme with 1 mM pNPGlc in
Total mRNA was extracted from 7 week old rice cv. KDML105
80 ml of 100 mM McIlvaine buffers [31], pH 2.5 to 8.0 in 0.5-pH-unit
seedlings with a Spectrum Plant RNA extraction kit (Sigma-
eAldrich). The mRNA was reverse transcribed from a poly dT18
primer with Superscript III reverse transcriptase (Invitrogen,
Carlsbad, CA, USA) and the derived cDNA used as a template
for PCR. The Os4BGlu13 cDNA was amplified by nested PCR with
ꢁ
increments (total reaction volume 100
reactions were stopped by adding 100
The released pNP was quantified as described for activity
m
l), at 30 C for 10 min. The
ml of 2 M sodium carbonate.
determination.
0
0
the Os4BGlu13StartF (5 -ATCCTTGCACAGCCATATGTC-3 ) and
2.8. Measurement of hydrolysis activities of Os4BGlu13 toward
0
0
Os4BGlu13_3pR (5 -AGCCGTGCTCAA GGAGAATCG-3 ) primers in
the first round and Os4BGlu13MatStNcoI (5 -CACCATG-
GAGCCGCCGCCGATCAG-3 ) and Os4BGlu13StopR (CCAAGCTT-
natural and synthetic glycosides
0
0
The kinetic parameters of rOs4BGlu13 with natural and syn-
thetic glycosides were determined by the method for determina-
tion of b-glucosidase activity for extracted protein fractions, but
variable reaction times, enzyme amounts and substrate concen-
trations were tested to obtain the initial velocities.
CATTTCTGGAGGAACTCCTTG) primers in the second round. The
final product was gel-purified, digested with NcoI and HindIII,
cloned into the corresponding sites in the pET32a(þ) expression
vector (Novagen, Merck-Millipore, Darmstadt, Germany), and
sequenced thoroughly in both directions by automated DNA
sequencing (Macrogen, Seoul, Korea).
For pNP glycosides substrates pNPGlc, pNP-
pNP- -fucoside, pNP- -mannoside and oNP-
.005e1.0 g of rOs4BGlu13 were incubated with concentrations of
substrates that bracketed the apparent K in 50 mM sodium ace-
l of BSA), pH 5.0, in a total volume
b-
D-galactoside,
b
-D
b-
D
b-D-glucoside,
0
m
2
.6. Recombinant expression and purification of Os4BGlu13
m
tate buffer (which included 1
m
g/m
ꢁ
The recombinant pET32a/Os4BGlu13 plasmids were introduced
into E. coli strain Rosetta-gami(DE3). The transformants were
of 140
70
measuring A405 and comparing it to that of a pNP standard curve.
The Vmax, K and kcat were calculated from nonlinear regression of
ml, at 30 C for 15 min. The reactions were stopped by adding
ml of 2 M sodium carbonate. The released pNP was quantified by
selected on an LB plate containing 12.5
m
g/ml chloramphenicol and
ꢁ
5
0
m
g/ml ampicillin at 37 C overnight. For expression, these cells
m
ꢁ
were grown continuously at 37 C with shaking at 220 rpm until
MichaeliseMenten plots with Grafit 5.0 software (Erithacus Soft-
the OD600 reached 0.4e0.6. Protein expression was induced with
ware, Horley, UK).
ꢁ
0
.2 mM IPTG (final concentration) at 20 C for 18 h. Cells were
For natural
0.01e1.0 g of rOs4BGlu13 was incubated with concentrations of
substrates that bracketed the apparent K in 50 mM sodium ace-
l of BSA), pH 5.0, in a total volume
b-D-glucoside and glucooligosaccharide substrates,
ꢁ
collected by centrifugation at 2240 ꢂ g for 30 min at 4 C and the
m
ꢁ
pellet was stored at ꢀ80 C until protein extraction. The pellets
m
were thawed at room temperature. Cells were resuspended with
tate buffer (which included 1
mg/m
ꢁ
extraction buffer (20 mM TriseHCl, pH 8.0, 600
mg/ml lysozyme, 1%
of 50
m
l, at 30 C for 15 min. The reactions were stopped by heating
ꢁ
Triton-X100, 0.1 mM PMSF, 5 g/ml DNase I and 0.1 mg/ml soybean
m
at 100 C for 5 min (except for GA -GE, which was only heated
4
trypsin inhibitor) and incubated at room temperature for 30 min.
Soluble proteins were separated from cell debris by centrifugation
1 min), and then the released glucose was determined with the
PGO assay as described above. The kinetic parameters were calcu-
lated as described for pNP glycosides.
ꢁ
at 12,000 ꢂ g for 10 min at 4 C.
Recombinantly expressed Os4BGlu13 (rOs4BGlu13) was purified
with 2 steps of immobilized metal affinity chromatography (IMAC)
on GE IMAC resin (GE Healthcare) bound with cobalt. The crude
protein was mixed with IMAC resin pre-equilibrated with equili-
2.9. Sequence analysis and homology modeling of Os4BGlu13
Secretory signal sequence prepeptide production was done with
SignalP [32], while N-linked glycosylation sites were predicted with
NetNGlc 1.0 [33]. Sequence alignments with closely related proteins
were conducted with CLUSTAL 2.1 [34].
The Os3BGlu13 structure was modeled based on the structure of
the closely related Os4BGlu12 [35]. Protein Data Bank Accession
[PDB: 3PTK] was used as template for homology modeling and the
three-dimensional model of Os4BGlu13 was generated by SWISS-
MODEL via the ExPASy web server [36].
ꢁ
bration buffer (20 mM TriseHCl, pH 8.0, and 150 mM NaCl) at 4 C
for 30 min. The resin with crude protein was loaded into a column
and loosely bound proteins were washed out with 5 CV of equili-
bration buffer, 5 CV of 5 mM imidazole in equilibration buffer and 5
CV of 10 mM imidazole in equilibration buffer. The bound proteins
were eluted with 5 CV of 250 mM imidazole in equilibration buffer
and 5 CV of 500 mM imidazole in equilibration buffer. The fractions
with activity were combined, concentrated and imidazole removed
by centrifugal ultrafiltration (Amico Ultra, 30k MWCO) at 700 ꢂ g,
ꢁ
4
4
C; and the buffer exchanged with 50 mM TriseHCl, pH 8.0, at
3. Results and discussions
ꢁ
C. The concentrated proteins from the 1st IMAC step were
incubated with enterokinase (New England Biolabs), with a ratio of
4
3.1. Extraction, purification and characterization of GA -GE b-
glucosidase from rice
ꢁ
0
.01
the N-terminal thioredoxin, His
The N-terminal thioredoxin, His
m
g enterokinase per 10 mg of protein at 20 C for 20 h to cut off
6
and S tags.
and S tags were removed with
6
Ten kilograms of 10-day rice seedling stems and leaves were
extracted and the GA -GE -glucosidase was purified with seven
purification steps, while monitoring the activities for hydrolysis of
pNPGlc and GA -GE (Table 1). Finally, 0.15 mg of protein was ob-
tained after the seven steps of purification in which protein frac-
tions with high hydrolysis activities toward GA -GE were pooled.
a second IMAC purification, in which the unbound proteins were
washed out with 10 CV of equilibration buffer, 10 CV of 5 mM
imidazole in equilibration buffer, and 10 CV of 10 mM imidazole in
equilibration buffer, respectively. The flow-through and wash
4
b
4
fractions containing
b
-glucosidase activity were combined,
4

40
Y. Hua et al. / Archives of Biochemistry and Biophysics 583 (2015) 36e46
Table 1
Summary of purification of b-glucosidase in rice.
Purification step
Total activity (
m
mol/min.)
GA -GE
11.0
Protein (mg)
Specific activity (
m
mol/min.mg)
Purification fold
Yielda (%)
pNP-Glc
4
pNP-Glc
GA -GE
4
pNP-Glc
4
GA -GE
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
2
4
4
3
2
2
1
2
3
4
5
6
7
8
. Crude extract
5710
2220
1470
281
27,200
12,000
1215
140
57
6
0.21
0.18
1.21
2.01
2.73
5.07
4.88
10.9
4.03 * 10
5.12 * 10
3.47 * 10
1.12 * 10
2.62 * 10
6.0 * 10
8.1 * 10
0.136
1
0.9
5.8
9.6
13
24
23
52
1
1.3
8.6
27.8
65
149
100
56
38
14
14
3.3
0.45
0.18
. After (NH4)2SO4 precipitation & dialysis
. CM-Sepharose 20 ml column
. Con A-Sepharose column
. Sephadex 75 column
. SP xl column
. Octyl column
6.17
4.22
1.57
1.50
0.36
0.05
0.02
157
ꢀ
30.4
3.07
1.63
ꢀ
2
0.63
0.15
201
337
. Superdex 200 column
4
mM pNPGlc and 1.72 mM of GA
Yield was calculated based on GA
4
-GE were used for hydrolysis activity measurements.
-GE hydrolysis activity.
a
4
The purification of GA
4
-GE-hydrolyzing
b
-
D
-glucosidase was 337
protein by SignalP 4.0 [32]; and 3 N-glycosylation sites at residues
33, 168 and 326 in the amino acid sequence of the precursor protein
were predicted [33] (Supplementary Fig. S1). The molecular mass
and theoretical pI of the mature protein were predicted to be
54,150 Da and 8.74, respectively. Since this protein bound to the
Con A column, it is likely to be glycosylated, which may easily ac-
count for its slightly larger apparent size of 60 kDa on the SDS-
PAGE. Although this protein is a member of the GlcD (Genbank:
COG0277) family of FMN/FAD-containing dehydrogenases, the
function of this protein is as yet unknown.
fold, while the recovery of activity was only 0.18%. During the pu-
rification, some of the enzyme molecules may have been degraded
and lost their activity, which may be one main reason for the low
4
purification yield. In addition, some isoenzymes that have GA -GE
hydrolyzing ability (many with lower specific activity than the
purified protein) were likely to have been removed during the
purification, resulting in a decrease in yield.
3
.2. Identification of b-glucosidase by LC-MS
The peptide masses from the 57 kDa protein matched the GH1
b
-glucosidase Os4BGlu13. This precursor protein was calculated to
Two major protein bands at approximately 60 kDa and 57 kDa
were observed in SDS-PAGE of the final GA -GE -glucosidase pu-
have a molecular mass of 57,386 Da and pI of 6.44 when its signal
peptide was included; and the molecular mass and theoretical pI
after the signal peptide was removed were predicted to be
4
b
rified from rice (Fig. 2) and were identified by LC-MS of tryptic
peptides generated from their SDS-PAGE gel bands. The peptide
masses from the 60 kDa band matched Genbank accession
EAZ01286.1, a hypothetical protein, OsI_23311 [Oryza sativa indica
Group], with the matching score of 851. The peptide masses from
the 57 kDa band matched Genbank accession NO_00105307.1, the
protein product from the Os04g0474900 gene locus [Oryza sativa
japonica Group], which corresponds to the glycoside hydrolase
5
4,761 Da and 6.66, respectively. Considering the fact that there are
five putative N-linked glycosylation sites in the amino acid
sequence of the precursor protein, this molecular mass matched
the size seen in the SDS-PAGE (Fig. 2).
The
viously reported by Wakuta et al. [4] as a
drolyzing tuberonic acid glucoside (TAG). The protein was
previously purified from rice panicles by (NH SO fractionation
and five different types of column chromatography. The purified
OsTAGG1 migrated as a single band on native PAGE, but was
present as two polypeptides with molecular masses of 42 kDa
and 26 kDa under denaturing conditions. The two apparent
subunits were caused by proteolytic cleavage of the initial
mature protein to produce two bands in SDS-PAGE, based on the
match of the two N-terminal sequences to the same gene. The
sum of the molecular masses of the two polypeptides was
b
-glucosidase Os4BGlu13, also called OsTAGG1, was pre-
b-D
-glucosidase hy-
family GH1
006 and sequence coverage of 49%.
The 60 kDa protein OsI_23311 [Oryza sativa Indica Group] was
b-glucosidase Os4BGlu13, with a matching score of
)
2
4
1
4
predicted to have a molecular mass of 57,128 Da, and calculated pI
of 8.67 for the precursor, including its signal peptide. The first 28
residues were predicted as a signal sequence in the precursor
6
5
8 kDa, which is different from the predicted molecular mass of
5 kDa. This discrepancy was explained by authors as likely to be
due to the fact that there are five N-linked glycosylation sites in
the amino acid sequence of the precursor protein, so the mo-
lecular size of the mature protein could be increased by addition
of sugar chains.
Our enzyme was extracted from 10-day old rice seeding stems
and leaves, while the OsTAGG1 of Wakuta et al. [4] was extracted
from rice panicles during grain filling. So, the different tissue
sources may be the main cause for obtaining two different molec-
ular sizes for the same mature protein, if the protein undergoes
proteolytic processing in the panicle that does not occur in the
seedling stems and leaves, and different levels of glycosylation
occur in the different tissues. Since different purification methods
were used, the previously purified OsTAGG1 may also have been
proteolytically cleaved during the purification, while proteolysis at
the same position did not appear to occur to the same extent in the
current work, since the major Coomassie-brilliant-blue-stained
band migrated at 57 kDa.
Fig. 2. 8% SDS-PAGE of fraction from Superdex S200 after staining with Coomassie
brilliant blue. Lane M, protein marker; lane 1, purified fraction. The two major
Coomassie-brilliant blue-stained bands were excised separately and submitted for
tryptic digestion followed by LC-MS analysis.

Y. Hua et al. / Archives of Biochemistry and Biophysics 583 (2015) 36e46
41
3
.3. Hydrolysis activities of rOs4BGlu13 toward natural and
rOs4BGlu13 had the highest hydrolysis activity toward pNPGlc,
while oNPGlc gives only 18.8% of activity relative to pNPGlc, pNP-
-fucopyranoside yields 14.9% relative activity, and pNP- -man-
nopyranoside, pNP- -galactopyranoside, pNP- -xylopyrano-
synthetic glycosides
b-
D
b-D
To test whether the Os4BGlu13 band was responsible for the
b
-D
b-D
GA
4
-GE hydrolysis, we cloned the cDNA from rice seedlings and
- and S-tagged
side gave relative activities of 4.4%, 1.5% and 0.7%, respectively.
Among glucooligo-saccharides, rOs4BGlu13 hydrolyzed lami-
used it to express N-terminally thioredoxin-, His
6
fusion protein in E. coli. The Rosetta-gami(DE3) strain was found to
give the best expression of four E. coli strains tested (also including
BL21(DE3), Origami(DE3) and Origami B(DE3), data not shown).
Although the expression level was low, such that the protein band
was not obvious in the crude extract and the protein was still quite
impure after the first IMAC step (Fig. 3), the two step IMAC puri-
fication provided 340-fold purification and 37% yield to give a
naribiose [b-(1 / 3)-linked disaccharide] best (19.1% relative to
pNPGlc), followed by laminaritriose (12.4%), cellotetraose (3.6%),
cellotriose (2.0%) and cellopentaose (1.9%). rOs4BGlu13 could not
hydrolyze longer
laminaritetraose and laminaripentaose. By comparison, gentiobiose
(6-O- -glucopyranosyl- -glucose was hydrolyzed with only 0.4%
relative activity, although this was higher than the 0.1% of cello-
b-(1 / 3) linked glucooligo-saccharides, such as
b-
D
D
protein with specific activity of 19.2
m
mol/min/mg protein (Table 2).
biose. Thus, rOs4BGlu13 prefers short
glucooligosaccharides.
Natural glycosides with different size of aglycones (Fig. 4) were
compared to the putative natural substrates TAG and GA -GE
b
-(1
/ 3)-linked
The purified Os4BGlu13 protein appeared as a single Coomassie
brilliant blue stained band on SDS-PAGE, to allow further charac-
terization (Fig. 3).
4
The rOs4BGlu13 was found to have high pNPGlc hydrolysis ac-
tivity between pH 4.0 and 5.2, with highest value at pH 4.5, and
with 50% of maximal activity at approximately pH 3.4 and 6.2
(Table 3). The rOs4BGlu13 enzyme preferred smaller aromatic
aglycones and with good leaving groups to large aglycones, espe-
cially those that are poorly ionized leaving group. Helicin (salicy-
(
Supplementary Fig. S2). This is similar to the pH optimum of
laldehyde
activity, followed by n-octyl-
delonitrile 6-O- -glucosyl-
n-heptyl- -glucoside, tuberonic acid glucoside (TAG), GA
daidzein-7-O- -glucopyranoside, salicylic acid glucoside (SAG)
and methyl- -glucoside. Hydrolysis of helicin results in release of
b
-
D
-glucoside) gave the highest relative hydrolysis
-glucoside, -amygdalin ( -man-
-glucoside), indoxyl -glucoside,
-GE,
approximately 4 reported for TAGG1 purified from rice panicle [4].
To evaluate the substrate specificity of the purified rOs4BGlu13,
the hydrolysis activities of Os4BGlu13 toward natural and synthetic
glycosides were determined (Table 3). Among pNP glycosides,
b
-
D
D
D
b
-
D
b-
D
b-D
b
-
D
4
b-D
b-D
salicylaldehyde, which is toxic to nematodes and bacteria and
adapted by specialist insects as a defense compound [37]. The hy-
drolysis of daidzein, an isoflavonoid found in legumes, suggests
that it may hydrolyze the structurally related flavonoid glycosides,
although hydrolysis of the flavonoid apigenin 7-O-glucoside was
very low. Arabidopsis thaliana BGLU15, which has been reported to
fall in the same protein sequence-based phylogenetic cluster
[28,38,39], has been shown to hydrolyze flavonoid glucosides that
accumulate in vacuoles in response to abiotic stress [38].
The substrates oNPGlc, helicin, SAG have similar sized agly-
cones, but with different functional groups on the ortho-position
a
of the phenyl ring. The pK values of the phenolic oxygen on these
three aglycones are 7.2 for oNP, 8.2 for salicyladehyde and 13.8 for
the salicylic acid. The hydrolysis activities of rOs4BGlu13 towards
these three substrates are helicin > oNPGlc > SAG, which suggests
the polarity of the ortho group may be more critical than the pK
itself. Both daidzein-7-O- -glucopyranoside and GA -GE have
a
b
-
D
4
very large aglycones, which may limit the substrate binding to
the active site, contributing to the low activities for these two
substrates.
The kinetic parameters of Os4BGlu13 with substrates of interest
ꢀ1
ꢀ1
are shown in Table 4. Its kcat/K
highest, while that of oNPGlc is 8.5-fold lower and for pNP-
fucopyranoside is over 30-fold lower. The kcat/K for helicin is
, which is 20-fold higher than the kcat/K of GA -GE.
Among the phytohormone glucoconjugates tested, TAG had the
highest kcat/K , while that of GA -GE was slightly lower and SAG
was lowest. Among oligosaccharides, the most efficiently hydro-
lyzed oligosaccharide tested, laminaritriose, had a kcat/K is
m
for pNPGlc (637 mM
s
) is
b-D-
m
ꢀ
1 ꢀ1
74.4 mM
s
m
4
Fig. 3. SDS-PAGE profiles of Os4BGlu13 recombinant protein expressed in Rosetta-
gami(DE3) E. coli. The 8% SDS-PAGE gel was stained with Coomassie brilliant blue.
Lanes: M, standard marker (Bio-RAD); 1, Crude protein extract; 2, Protein purified by
m
4
1st immobilized metal affinity chromatography (IMAC); 3, The Os4BGlu13 protein after
m
cutting off the fusion tags with enterokinase and removing them by a 2nd IMAC step.
Table 2
Summary of purification of
b-glucosidase from recombinant expression.
Purification step
Total activity (
m
mol/min.)
Protein (mg)
Specific activity (
m
mol/min.mg)
Purification fold
Yielda (%)
1
2
3
. Crude extract
. After first IMAC and buffer exchange
. After second IMAC, buffer exchange and
151
98.0
56.4
2674
23.4
2.94
0.0564
4.18
19.2
1
74
340
100
65
37
concentration
a
Purification parameters were calculated based on pNPGlc hydrolysis activity.

42
Y. Hua et al. / Archives of Biochemistry and Biophysics 583 (2015) 36e46
Table 3
Relative activities of Os4BGlu13 toward natural and synthetic glycosides.
Substrates
Relative activity with standard deviation (%)
pNP glycosides
pNP-
pNP-
pNP-
pNP-
pNP-
oNP-
b
b
b
b
b
-
-
-
-
-
D
D
D
D
D
D
-glucopyranoside
-fucopyranoside
-mannopyranoside
-galactopyranoside
-xylopyranoside
-glucopyranoside
100 ± 1
14.9 ± 0.7
4.45 ± 0.12
1.48 ± 0.12
0.67 ± 0.12
18.8 ± 1.2
b-
Glucooligosaccharides
Laminaribiose [
Laminaritriose
Laminaritetraose
Laminaripentaose
b
-(1 / 3)-linked]
19.1 ± 3.5
12.4 ± 1.4
0.73 ± 0.07
0.19 ± 0.02
0.13 ± 0.02
1.99 ± 0.15
3.56 ± 0.18
1.90 ± 0.07
1.13 ± 0.14
0.43 ± 0.03
Cellobiose [
Cellotriose
b-(1 / 4)-linked]
Cellotetraose
Cellopentaose
Cellohexaose
Gentiobiose (6-O-
b-
D
-Glucopyranosyl-
D
-glucose)
Alkyl
n-Octyl-
n-Heptyl-
Methyl-
b
-
b
D
-glucosides
-D-glucoside
-D-glucoside
-D-glucoside
-glucosides
10.0 ± 0.5
3.82 ± 0.29
0.42 ± 0.05
b
b
Natural
b
-D
Helicin (salicylaldehyde
-Amygdalin ( -mandelonitrile 6-O-
Indoxyl -glucoside
GA -GE (gibberellin A4 b-D-glucosyl ester)
Tuberonic acid glucoside (TAG)
Salicylic acid glucoside (SAG)
b
-
D
-glucoside)
35.6 ± 2.2
D
D
b
-D
-glucosido-
b
-D
-glucoside)
9.60 ± 0.85
8.93 ± 2.19
1.54 ± 0.04
1.67 ± 0.02
0.50 ± 0.003
0.54 ± 0.01
0.044 ± 0.004
b-D
4
Daidzin (daidzein-7-O-
b-D-glucopyranoside)
Apigenin 7-O- -glucopyranoside
b-D
1
mM substrates were used for relative activity measurement. The standard deviations were calculated based on three replicates.
.57 mM 1
-GE. These data support the identification of Os4BGlu13
-glucosidase, but suggest it may play other functions in
the plant as well, including helicin and perhaps GA -GE hydrolysis.
ꢀ
s , which is lower than that of TAG, but higher than
ꢀ1
catalytic nucleophile, respectively, based on previous studies,
including those on Os3BGlu6 [29], Os3BGlu7 [27] and Os4BGlu12
[28]. Although Os4BGlu12 and Os4BGlu13 have 85% amino acid
sequence identity, these two enzymes have shown different spe-
5
that of GA
as a TAG
4
b
4
The hydrolysis of TAG, SAG and helicin by Os4BGlu13 support a
role in wound response and regrowth, although Os4BGlu13 gene
expression is not induced in wounding [22]. Inspection of the
microarray data in the Rice ExPro database [40] (http://ricexpro.
4 4
cific activity, especially to GA -GE ester. The hydrolysis of GA -GE
by partially purified Os4BGlu13 (approximately 50% purity from
SDS-PAGE) was 1% of that of pNPGlc, while for recombinant
Os4BGlu13 gave 1.5%. By comparison, the closely related Os4BGlu12
dna.affrc.go.jp/field-development.php?featurenum¼21285)
in-
4
activity toward GA -GE was 0.03% of that toward pNPGlc, while
Os3BGlu6, which had the highest relative activity compared to
pNPGlc, had 7% [24]. Thus, relative to activity toward pNPGlc, the
dicates that the Os4BGlu13 locus Os4g0474900 is most highly
expressed in root, leaf sheath during the vegetative stage and early
stages of ovary growth. It is most highly expressed during the
rapidly growing stage of roots (21e56 days after transplantation)
and the expression drops off when root growth is essentially
complete (63 days after transplantation). Since Os4BGlu13 is most
highly expressed in the development of roots, leaf sheath and ovary
hydrolysis of GA
However, the rOs4BGlu13 rate of hydrolysis of pNPGlc is quite high,
similar to Os4BGlu12. In fact, the kcat/K values of rOs4BGlu13, for
both pNPGlc and GA -GE are much higher, 637 and 3.63 mM
respectively, compared to 6.2 and 0.13 mM
respectively [24]. Therefore, Os4BGlu13 is the most efficient
4
-GE by Os3BGlu6 appears better than Os4BGlu13.
m
ꢀ
1 ꢀ1
4
s
ꢀ
1
ꢀ1
s for Os3BGlu6,
4
during their rapid growth, Os4BGlu13 hydrolysis of GA -GE may
play a role in regulating growth in these tissues, although other
Os4BGlu13 substrates may also be important in development of
these tissues or providing for their defense.
enzyme for GA
tested.
4
-GE hydrolysis among the enzymes that have been
The four enzymes also showed different substrate preferences
toward oligosaccharides. Os3BGlu7, Os4BGlu12 and rOs4BGlu13
3
.4. Comparison of the protein structures and substrate binding for
hydrolyze
but differences were observed in their specificities for substrate
chain lengths. The rates of hydrolysis of -(1,4)-linked oligosac-
b-(1,4)-linked oligosaccharides of 3e6 glucose residues,
Os3BGlu6, Os3BGlu7, Os4BGlu12 and Os4BGlu13
b
To study the relationship between the protein structure and
substrate binding, the amino acid sequences of Os3BGlu6, Os3B-
Glu7 and Os4BGlu12 were aligned with Os4BGlu13. The alignment
in Fig. 5 shows that Os4BGlu12 and Os4BGlu13 have 85% identity,
while Os3BGlu6 and Os3BGlu7 have 49% identity with Os4BGlu13
and each other.
The glutamate residues at positions 174 and 388 (Os4BGlu13
numbering) in the consensus sequences T(F/L)NEP and ITENG,
respectively (Fig. 5), were predicted to be the acid-base catalyst and
charides by Os3BGlu7 increase with increasing chain length of the
substrate, while those of Os4BGlu12 and rOs4BGlu13 remain
approximately constant [24,25]. Os3BGlu6 had very low hydrolysis
activities towards
b-(1,4)-linked oligosaccharides [29]. Among the
b-1,3-linked laminarioligosaccharides, Os4BGlu12 only hydrolyzes
laminaribiose, while Os3BGlu6 Os3BGlu7, rOs4BGlu13 also hydro-
lyze laminaritriose, although at a lower rate than laminaribiose.
By comparing the active sites of Os3BGlu6, Os4BGlu12, and
Os4BGlu13 (Fig. 6), we can see that the residues at the ꢀ1 site are

Y. Hua et al. / Archives of Biochemistry and Biophysics 583 (2015) 36e46
43
Fig. 4. Structures of natural b-D-glucoside substrates hydrolyzed by Os4BGlu13.

4
4
Y. Hua et al. / Archives of Biochemistry and Biophysics 583 (2015) 36e46
Table 4
Apparent kinetic parameters of Os4BGlu13 for hydrolysis of natural and synthetic glycosides.
a
cat (sꢀ1)
k
cat/K
(mM
ꢀ1 ꢀ1)
s
Substrate (mM)
V
max
K
m
k
m
pNP-
pNP-
pNP-
pNP-
oNP-
Helicin
Cellotriose
Cellotetraose
Cellopentaose
Laminaribiose
Laminaritriose
b
b
b
b
-
-
-
-
D
D
D
D
D
-glucopyranoside
-fucoside
-mannoside
-galactoside
-glucoside
157 ± 7
0.234 ± 0.017
0.183 ± 0.016
1.22 ± 0.15
1.50 ± 0.10
7.19 ± 0.05
0.835 ± 0.065
1.90 ± 0.17
3.44 ± 0.11
4.71 ± 0.06
31.8 ± 1.1
149 ± 8
637 ± 26
20.4 ± 4.0
0.927 ± 0.065
2.34 ± 0.17
71.8 ± 0.4
3.87 ± 0.10
1.19 ± 0.03
3.69 ± 0.03
543 ± 2
65.4 ± 1.7
3.27 ± 0.02
5.06 ± 0.07
8.42 ± 0.03
147 ± 2
54.9 ± 1.2
3.01 ± 0.05
2.86 ± 0.04
1.58 ± 0.04
3.68 ± 0.11
1.13 ± 0.03
3.51 ± 0.03
516 ± 2
62.1 ± 1.6
3.10 ± 0.02
4.80 ± 0.07
8.00 ± 0.06
140 ± 2
52.1 ± 1.2
2.86 ± 0.04
2.71 ± 0.03
1.50 ± 0.04
b-
74.4 ± 2.6
1.64 ± 0.08
1.39 ± 0.02
1.70 ± 0.03
4.40 ± 0.09
5.57 ± 0.19
3.63 ± 0.13
6.68 ± 0.34
0.878 ± 0.022
9.36 ± 0.38
0.788 ± 0.038
0.407 ± 0.026
1.71 ± 0.09
4
GA -GE
Tuberonic acid glucoside
Salicylic acid glucoside
a
The values were derived from nonlinear regression of the MichaeliseMenton graphs and the standard errors are calculated from the curves, which included 3 replicates at
m
no less than 6 substrate concentrations covering the range of at least 0.3 to 3 times the K .
Fig. 5. Partial sequence alignment for four rice b-glycosidases, Os3BGlu6, Os3BGlu7, Os4BGlu12 and Os4BGlu13. The alignment was generated with Clustal 2.1 software [37]. The
region shown includes the two catalytic amino acids, which fall in the consensus sequences that are boxed in rectangles, as well as most sugar-binding residues. The residues at
the þ2 site are indicated with black arrows.
well conserved for the three enzymes, but the residues at the þ1
and þ2 sites are different. For Os4BGlu12, two residues at the þ2
site, W181 and H252, are bulkier and occupy more space than the
residues at the same positions in Os3BGlu6 (H180 and M251) and
Os4BGlu13 (L176 and N247), so they may lead to a smaller binding
relative to Os4BGlu13 and Os3BGlu6, so that Os4BGlu12 hydrolyzed
it poorly in comparison to Os4BGlu13 and Os3BGlu6.
This work suggested that the amino acid residues around
binding cleft are very important for the substrate binding; the
charge, the polarity and the sizes of the amino acid residues all
contribute to the enzyme's substrate specificity. The enzymatic
4
cleft. This may result in poor binding of GA -GE for Os4BGlu12,

Y. Hua et al. / Archives of Biochemistry and Biophysics 583 (2015) 36e46
45
Fig. 6. The superimposition of the active sites of the Os3BGlu6 complex with n-octyl-
Os4BGlu13. The carbons of octyl- -D-thiogluco-pyranoside are gray, while the residues with pink carbons are from Os3BGlu13, those with yellow carbons are from Os3BGlu6 and
those with green carbons are from Os4BGlu12. The subsites, relative to the positions of binding of
b-D-thiogluco-pyranoside (PDB: 3GNP), Os4BGlu12 (PDB: 3PTK) and the homology model of
b
b
-1,4-linked glucosyl units, are indicated by the numbers (ꢀ1 for the nonreducing
glucosyl residue, þ1 for the next, and þ2 for the next, going from the nonreducing to reducing end). The figures were generated by Pymol (Schr o€ dinger LLC). An interactive version is
available in the supplementary data as Fig. S3. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
data suggests that Os4BGlu13 may contribute to helicin, TAG, SAG,
oligosaccharide and GA -GE hydrolysis in the rice plant. This may
[6] J.D. Mauseth, Botany: an Introduction to Plant Biology, Saunders, Philadelphia,
991, pp. 348e415.
1
4
[7] P.H. Raven, R.F. Evert, S.E. Eichhorn, Biology of Plants, Worth, New York, 1992,
indicate a combined response including defense, wounding and
systematic acquired resistance, recycling of damaged cell wall and
regrowth, although such coordination of wounding response re-
quires further study and other roles for Os4BGlu13 are possible.
pp. 545e572.
[8] F.B. Salisbury, C.W. Ross, Plant Physiology, Wadsworth, Belmont, CA, 1992, pp.
57e407, 531e548.
3
[
9] V.M. Sponsel, P. Hedden, in: P.J. Davies (Ed.), Plant Hormones: Biosynthesis,
Signal Transduction, Action!, Kluwer, Dordrecht, 2004, pp. 62e98.
[
[
[
10] G. Schneider, in: A. Crazier (Ed.), The Biochemistry and Physiology of Gib-
berellins, vol. 1, Praeger Publishers, New York, 1983, pp. 389e456.
11] G. Schneider, E. Jensen, C.R. Spray, B.O. Phinney, Proc. Natl. Acad. Sci. U. S. A.
Acknowledgments
89 (1992) 8045e8048.
12] M. Koshioka, E. Minami, H. Saka, R.P. Pharis, L.N. Mander, In: 13th Interna-
tional Conference on Plant Growth Substances, Edited by: R.P. Pharis and S. B.
Rood, 217.
We thank Ms. Chiraporn Saetang for assistance with protein
purification and Ms. Daranee Chokchaichamnankit from the Chu-
labhorn Research Institute (CRI) and Ms. Atchara Paemanee from
the National Center for Genetic Engineering and Biotechnology
[13] T. Fowler, in: A. Esen (Ed.), Biochemistry and Molecular Biology, American
Chemical Society, Washington DC, 1993, pp. 56e65. ACS Symposium Series
533.
(
BIOTEC) for their help on peptide analysis by LC-MS. This work was
supported by Suranaree University of Technology (SUT1-102-55-
6-01), the National Research University Project of the Commis-
[14] Y. Yildiz, H. Matern, B. Thompson, J.C. Allegood, R.L. Warren, D.M. Ramirez,
R.E. Hammer, F.K. Hamra, S. Matern, D.W. Russell, J. Clin. Invest. 116 (11)
(2006) 2985e2994.
3
[
[
15] J.E. Poulton, Plant Physiol. 94 (1990) 401e405.
16] L. Duroux, F.M. Delmotte, J.M. Lancelin, G. Keravis, C. Jay-Alleand, Biochem. J.
333 (1998) 275e283.
sion on Higher Education of Thailand and the Thailand Research
Fund (Grant BRG5680012).
[17] B. Brzobohatý, I. Moore, P. Kristoffersen, L. Bako, N. Campos, J. Schell, K. Palme,
Science 262 (1993) 1051e1054.
18] A. Falk, L. Rask, Plant Physiol. 108 (1995) 1369e1377.
19] D.P. Dharmawardhana, B.E. Ellis, J.E. Carlson, Plant Physiol. 107 (1995)
Appendix A. Supplementary data
[
[
3
31e339.
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.abb.2015.07.021.
[
[
20] M. Reuveni, Z. Sagi, D. Evnor, A. Hetzroni, Plant Sci. 147 (1999) 19e24.
21] L. Barleben, X. Ma, J. Koepke, G. Peng, H. Michel, J. Stockigt, Biochim. Biophys.
Acta 1747 (2005) 89e92.
€
[
[
22] S. Wakuta, S. Hamada, H. Ito, R. Imai, H. mori, H. Matsuura, K. Nabeta,
H. Matsui, J. Appl. Glycosci. 58 (2011) 67e70.
23] K. Hiraga, T. Yokota, N. Takahashi, Phytochemistry 13 (1974) 2371e2376.
References
[
[
[
1] W. Schliemann, J. Plant Physiol. 116 (1984) 123e132.
[24] Y. Hua, S. Sansenya, C. Saetang, S. Wakuta, J.R. Ketudat Cairns, Arch. Biochem.
Biophys. 537 (1) (2013) 39e48.
[25] H. Matsuura, F. Ohmori, M. Kobayashi, A. Sakurai, T. Yoshihara, Biosci. Bio-
technol. Biochem. 64 (2000) 2380e2387.
2] K. Kai, K. Wakasa, H. Miyagawa, Phytochemistry 68 (2007) 2512e2522.
3] K.H. Lee, H.L. Piao, H.Y. Kim, S.M. Choi, F. Jiang, W. Hartung, L. Hwang,
J.M. Kwak, I.J. Lee, I. Hwang, Cell 126 (2006) 1109e1120.
[
[
4] S. Wakuta, S. Hamada, H. Ito, H. Matsuura, K. Nabeta, H. Matsui, Phytochem-
istry 71 (2010) 1280e1288.
5] P.J. Davies, Plant Hormones: Physiology, Biochemistry and Molecular Biology,
Publisher, Kluwer Academic Publishes, 1995, ISBN 0792329848.
[26] G. Grynkiewicz, O. Achmatowicz, J. Hennig, J. Indulski, D.F. Klessig, Pol. J.
Chem. 67 (1993) 1251e1254.
[27] R. Opassiri, J.R. Ketudat Cairns, T. Akiyama, O. Wara-Aswapati, J. Svasti,
A. Esen, Plant Sci. 165 (2003) 627e638.

46
Y. Hua et al. / Archives of Biochemistry and Biophysics 583 (2015) 36e46
[
[
[
28] R. Opassiri, B. Pomthong, T. Onkoksoong, T. Akiyama, A. Esen, J.R. Ketudat
Cairns, BMC Plant Biol. 6 (2006) 33.
29] S. Seshadri, T. Akiyama, R. Opassiri, B. Kuaprasert, J.R. Ketudat Cairns, Plant
Physiol. 151 (1) (2009) 47e58.
30] T. Kuntothom, S. Luang, A.J. Harvey, G.B. Fincher, R. Opassiri, M. Hrmova,
J.R. Ketudat Cairns, Arch. Biochem. Biophys. 491 (1e2) (2009) 85e95.
31] T.C. McIlvaine, J. Biol. Chem. 49 (1921) 183e186.
[35] S. Sansenya, R. Opassiri, B. Kuaprasert, C.J. Chen, J.R. Ketudat Cairns, Arch.
Biochem. Biophys. 510 (2011) 62e72.
[36] M. Biasini, S. Bienert, A. Waterhous, K. Arnold, G. Studer, T. Schmidt, F. Kiefer,
T. Gallo Cassarino, M. Bertoni, L. Bordoli, T. Schwede, Nucl. Acids Res. 42
(2014) W252eW258.
[37] P. Caboni, N. Aissani, T. Cabras, A. Falqui, R. Marotta, B. Liori, N. Ntalli, G. Sarais,
N. Sasanelli, G. Tocco, J. Agric. Food Chem. 61 (2013) 1794e1803.
[38] J. Roepke, G.G. Bozzo, Phytochemistry 109 (2015) 14e24.
[39] C. Rouyi, S. Baiya, S.K. Lee, B. Mahong, J.S. Jeon, J.R. Ketudat-Cairns, M. Ketudat-
Cairns, PLoS One 9 (2014) e96712.
[40] Y. Sato, H. Takehisa, K. Kamatsuki, H. Minami, N. Namiki, H. Ikawa,
H. Ohyanagi, K. Sugimoto, B. Antonio, Y. Nagamura, Nucl. Acids Res. 41 (2013)
D1206eD1213.
[
[
32] T.N. Petersen, S. Brunak, G. von Heijne, H. Nielson, Nat. Methods 8 (2011)
785e786.
[
33] R. Gupta, E. Jung, S. Brunak, NetNGlyc: Prediction of N-glycosylation Sites in
Human Proteins, Software, 2005. http://www.cbs.dtu.dk/services/NetNGlyc/.
34] M.A. Larkin, G. Blackshields, N.P. Brown, R. Chenna, P.A. McGettigan,
H. McWilliam, F. Valentin, I.M. Wallace, A. Wilm, R. Lopez, J.D. Thompson,
T.J. Gibson, D.G. Higgins, Bioinformatics 23 (2007) 2947e2948.
[