Partial purification and characterization of a soybean β-glucosidase with high specific activity towards isoflavone conjugates

Article abstract of DOI:10.1016/S0031-9422(01)00380-6

A β-glucosidase with high specific activity towards isoflavone conjugates was purified from soybean [Glycine max] roots by high salt extraction from a low speed centrifugal pellet and subsequent anion and cation exchange chromatography. Purification required stabilization throughout fractionation in 10% glycerol. The enzyme is most likely a dimer (approximate M_r 165 kDa) with potential subunits of M_r 80 and/or 75 kDa. The pH and temperature optima are pH 6 and 30 °C, respectively. The enzyme was highly heat-stable. Of the various potential effectors examined, silver and mercury ions were the most inhibitory. The IC₅₀ of silver ions was increased from 140 μM to 14 mM in the presence of 250 μM β-mercaptoethanol. Glucono-δ-lactone was not strongly inhibitory (IC₅₀ 24 mM). The activity was highly active against isoflavone conjugates, with a specificity constant 160-1000 fold higher for isoflavone conjugates over the generic chromogenic substrate, p-nitrophenyl β-glucoside. The enzyme was inactive against the flavonol glycosides tested. The partially purified enzyme had similar K_m and k_cat towards 7-O-glucosyl- and 7-O-glucosyl-6″-malonyl-isoflavones, suggesting that it may be able to cleave the esterified glucosyl conjugate. We hypothesize that the enzyme is involved in the release of daidzein and genistein, both of which play central roles in soybean defense.

Full text of DOI:10.1016/S0031-9422(01)00380-6

Phytochemistry 58 (2001) 995–1005
www.elsevier.com/locate/phytochem
Partial purification and characterization of a soybean
b-glucosidase with high specific activity towards isoflavone
conjugates
1
Ming-Ching Hsieh , Terrence L. Graham*
Department of Plant Pathology, The Ohio State University, Columbus, OH 43210, USA
Received 4 May 2001; received in revised form 6 August 2001
Abstract
A b-glucosidase with high specific activity towards isoflavone conjugates was purified from soybean [Glycine max] roots by high
salt extraction from a low speed centrifugal pellet and subsequent anion and cation exchange chromatography. Purification
required stabilization throughout fractionation in 10% glycerol. The enzyme is most likely a dimer (approximate M 165 kDa) with
r
ꢀ
potential subunits of M
r
80 and/or 75 kDa. The pH and temperature optima are pH 6 and 30 C, respectively. The enzyme was
highly heat-stable. Of the various potential effectors examined, silver and mercury ions were the most inhibitory. The IC50 of silver
ions was increased from 140 mM to 14 mM in the presence of 250 mM b-mercaptoethanol. Glucono-d-lactone was not strongly
inhibitory (IC50 24 mM). The activity was highly active against isoflavone conjugates, with a specificity constant 160–1000 fold
higher for isoflavone conjugates over the generic chromogenic substrate, p-nitrophenyl b-glucoside. The enzyme was inactive
m
against the flavonol glycosides tested. The partially purified enzyme had similar K and kcat towards 7-O-glucosyl- and 7-O-gluco-
0
0
syl-6 -malonyl-isoflavones, suggesting that it may be able to cleave the esterified glucosyl conjugate. We hypothesize that the
enzyme is involved in the release of daidzein and genistein, both of which play central roles in soybean defense. # 2001 Elsevier
Science Ltd. All rights reserved.
Keywords: Glycine max; Leguminosae; Soybean; Phenylpropanoid; Isoflavone; Genistein; Daidzein; b-Glucosidase; Partial purification
1
. Introduction
microbes, insects, or parasitic plants (Bell, 1981; Poul-
ton, 1990; Phillips and Streit, 1996), floral development
and pigmentation (Harborne and Mabry, 1982; Koes et
al., 1994), and they are also thought to have roles in
b-Glucosidases (b-d-glucoside glucohydrolase, EC
.2.1.21) comprise a heterogeneous group of enzymes
3
that are able to cleave the b-glucosidic linkages of di-
and/or oligo-saccharides, or other glucose conjugates.
b-Glucosidases are widely distributed in the living world
and play pivotal roles in many biological processes,
such as degradation of cellulosic biomass, hydrolysis of
glycolipids, cyanogenesis, and modification of second-
ary metabolites (Esen, 1993). In plants, b-glucosidase
activity is involved in processes such as the compart-
mentalization and activity of phytohormones (Klecz-
kowski and Schell, 1995), defense mechanisms against
¨
lignification and cell wall decomposition (Hosel et al.,
1978; Leah et al., 1995). Several b-glucosidases of plant
origin are highly substrate-specific, such as those having
selective specificity toward sapinin (Inoue and Ebizuka,
1996), hydrojuglone (Duroux et al., 1998), cinnamyl
alcohols (Dharmawardhana et al., 1995), cyanogenic
glycosides (Conn, 1993), flavones (Maier et al., 1993)
¨
and isoflavones (Hosel and Barz, 1975). Despite the
availability of crystal structures for several plant b-glu-
cosidases, the aglycone-binding site is still poorly
understood. In a recent study, the active site of the
maize b-glucosidase for the substrate DIMBOA-gluco-
side and its interaction with the cyanogenic glucoside
dhurrin was examined (Czjzek et al., 2000).
*
Corresponding author. Tel.: +614-292-1375; fax: +1-614-292-
455.
E-mail address: graham.1@osu.edu (T.L. Graham).
4
1
Isoflavone conjugates are sometimes constitutively
present in large quantities in legumes (see e.g. Graham,
Current address: Department of Chemistry, University of Penn-
sylvania, Philadelphia, PA 19104-6323, USA.
0031-9422/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0031-9422(01)00380-6

996
M.-C. Hsieh, T.L. Graham / Phytochemistry 58 (2001) 995–1005
1
991b). These isoflavone conjugates, including iso-
In plants, the accumulation of isoflavone conjugates
may be mediated by several enzymes (Fig. 1). The
hydrolytic enzymes, isoflavone 7-O-glucoside-6 -O-mal-
onate malonylesterase and isoflavone 7-O-glucoside b-
glucosidase, from chickpea have been purified and
0
0
flavone 7-O-glucosides and isoflavone 7-O-glucoside-6 -
O-malonates, may serve as pools for the release of free
aglycones, which in turn may play important roles in
plant-microbe interactions and defensive mechanisms
against pathogen infection (Graham et al., 1990; Kess-
mann et al., 1990; Mackenbrock and Barz, 1991;
Dakora and Phillips, 1996). Released isoflavones, for
instance, can be either incorporated into the biosyn-
thetic pathway of phytoalexins (Ebel, 1986; Barz and
Welle, 1992; Kleczkowski and Schell, 1995) or act as
preformed toxins, termed phytoanticipins (Johnson et
al., 1976; Kramer et al., 1984; Weidenborner et al.,
0
0
¨
characterized (Hosel and Barz, 1975; Hinderer et. al.,
1986). Although non-specific b-glucosidase activities
hydrolyzing isoflavone glucosides in the process of miso
preparation and a lactase phlorizin hydrolase have been
reported in soybean (Matsuura and Obata, 1993; Mat-
suura et al., 1995; Day et al., 2000), we know of no
reports of characterization of b-glucosidases in soybean
with preferential activity towards isoflavone conjugates.
Because of the critical roles daidzein and genistein may
play in soybean defense, we undertook the character-
ization of some of the enzymes involved in the turnover
of their conjugates. In this report, a soybean isoflavone
conjugate-hydrolyzing b-glucosidase (ICHG) was par-
tially purified and characterized. The enzyme has high
specific activity towards genistein/daidzein 7-O-gluco-
side (genistin and daidzin) and genistein/daidzein 7-O-
1
990; Rivera-Vargas et al., 1993). Free isoflavones may
also function as chemoattractants for both pathogens
Morris and Ward, 1992) and symbionts (Phillips et al.,
992) or as regulators of gene expression in Rhizobium
(
1
spp. (Hungria and Stacey, 1997). More recently, the
isoflavone genistein has been hypothesized to play a
regulatory role in the activation of a soybean Type II
NADH oxidase thought to participate in the establish-
ment of competency for response to pathogen elicitors
0
0
glucosyl 6 -O-malonate (MGD and MGG), with very
low Km and high kcat values. The specificity constant
(kcat/K ) is 160–1000 fold higher for the isoflavone
(see Graham and Graham, 1996a,b and reviews Graham
and Graham, 1999, 2000).
m
Fig. 1. Enzymatic regulation of the isoflavone conjugate pools. The scheme is based on the four purified enzymes from chickpea where mal-
onylesterase and b-glucosidase hydrolyze isoflavone 7-O-glucoside-6 -O-malonates and isoflavone 7-O-glucosides, respectively, and glucosyl-
transferase and malonyltransferase transfer glucose and malonate sequentially to the isoflavone moiety.
00

M.-C. Hsieh, T.L. Graham / Phytochemistry 58 (2001) 995–1005
997
conjugates over the chromogenic substrate p-nitrophe-
nyl b-glucoside. Isoflavone conjugate hydrolysis in soy-
bean may differ from that of chickpeas in that we found
no evidence for an esterase active on MGG or MGD.
Although we cannot preclude the existence of such an
esterase, the partially purified glucosidase characterized
in this report was able to hydrolyze both the glucosyl
and malonyl glucosyl isoflavones directly with similar
kinetics.
ware (Molecular Dynamics) determined that the pixel
volume of the 80 kD band was three to four fold enri-
ched in comparison to the other minor bands. This is
consistent with the separation from a majority of other
proteins for this fraction as shown in Fig. 2B. At this
point, the ICHG was purified approximately 20-fold
with a 20% recovery rate (Table 1). The most active
fractions were pooled and lyophilized until almost only
glycerol was left. Aliquots of this glycerol solution of
ꢀ
ICHG were stored at ꢁ20 C for further study. Under
these conditions, the enzyme aliquots did not freeze
solid and activity was stable for at least six months.
2
. Results and discussion
2
ꢀ
.1. Purification of the isoflavone conjugate-hydrolyzing
-glucosidase
2.2. Properties of the isoflavone conjugate-hydrolyzing
ꢀ-glucosidase
Many studies were performed with various soybean
The molecular mass of ICHG upon SDS PAGE
under non-reducing conditions was judged to be
approximately 165 kDa (data not shown), while under
reducing conditions a major band of mass 80 kDa and a
less intensely staining band of 75 kDa were seen (Fig. 3).
Though both of these bands were reproducibly asso-
ciated with activity throughout purification, it is not
clear whether the two bands represent different poly-
peptides or different post-translational modifications of
a single polypeptide. Thus, while it is likely that the
enzyme is a dimer, with monomers linked through a
disulfide bond, its precise subunit composition will
require further clarification. The ICHG tolerates very
high concentrations of b-mercaptoethanol, with 50%
inhibition of enzyme activity (IC ) at 1.1 M, suggesting
organs to develop the optimal purification protocol. For
example, vacuum infiltration of soybean cotyledons
with 0.2 M phosphate buffer, pH. 6, containing 200–500
uM NaCl and subsequent isolation of the intercellular
washing fluids by centrifugation led to a preparation
highly enriched for the glucosidase, suggesting a possi-
ble apoplastic localization. These intercellular washing
fluids contain no cytoplasmic marker enzymes (T.L.
Graham, unpublished). However, due to the relatively
low level of the ICHG in cotyledons and the difficulty in
preparing large amounts of the intercellular fluids, we
chose to work with roots. Roots contained the highest
specific activity of ICHG in the soybean seedling. For
example, a crude preparation from roots was about 800-
fold more enriched for the enzyme than a crude pre-
paration from cotyledons (data not shown). In roots,
activity was found to be associated with the low speed
centrifugal pellet.
5
0
that the monomer form may also be catalytically active.
Many b-glucosidases contain either identical or non-
identical subunits. The number of subunits varies from
2 to 10 (or more) depending on the enzyme. For
instance, a 60 kDa coniferin-hydrolyzing b-glucosidase
from lodgepole pine has two non-identical subunits with
Enzyme activity was solubilized from the pellet with
.2 M phosphate buffer, pH 6, containing 0.5 M NaCl
0
and fractionated by 40–60% ammonium sulfate satura-
tion. Due to empirical results regarding the instability of
the enzyme, a desalting column (Sephadex G-25) was
chosen over dialysis for salt removal, and 10% glycerol
was always present in the buffer from this step onward.
When the desalted protein solution was passed through
a DEAE-Sephadex A-50 column at pH 6.0, nearly all of
the enzyme activity was in the unbound fractions
M of 28 and 24 kDa (Dharmawardhana et al., 1995).
r
The furostanol glycoside 26-O-b-glucosidase from Cos-
tus speciosus is a 110 kDa protein with two subunits
with M of 54 and 58 kDa (Inoue and Ebizuka, 1996).
r
Furthermore, the isoflavone hydrolyzing b-glucosidase
from chickpea has apparent M of 125–135 kDa with
r
two identical subunits of 68 kDa (Ho
1975).
The ICHG showed a pH optimum at approximately
¨
sel and Barz,
(
Fig. 2A). The active fractions were pooled and 5 mM
ꢀ
ꢀ
final concentration of sodium ascorbate was introduced
to maximize stability. The protein sample was then loa-
ded onto a CM-Sephadex C-50 column and eluted with
a linear gradient of buffer F in buffer E. The activity
peak was eluted at 0.56 M NaCl (Fig. 2B). SDS poly-
acrylamide gel electrophoresis (PAGE) of various frac-
tions under reducing conditions is shown in Fig. 3.
Although the banding patterns for the fractions from
the DEAE and CM-Sephadex columns are qualitatively
similar, the analysis of the gel with ImageQuant soft-
pH 6 at 30 C. The optimum temperature was 30 C.
Both are typical for b-glucosidases (Esen, 1993). The
enzyme is highly heat stable and remains substantially
ꢀ
active (87%) after 45 min at 50 C.
Several chemicals reported to be effectors of b-gluco-
sidases were examined. Of all the chemicals tested for
the inhibition of enzyme activity, only silver and mer-
cury ions were inhibitory at the 1 mM level (Table 2),
suggesting that sulfhydryl groups may play an essential
role in enzyme activity. However, other sulfhydryl

998
M.-C. Hsieh, T.L. Graham / Phytochemistry 58 (2001) 995–1005
Fig. 2. The purification of the isoflavone-specific b-glucosidase. The chromatograms for anion exchange on DEAE-Sephadex (a), and cation
exchange on CM-Sephadex (b). Symbols: open circle and dashed line, A280; closed circle and solid line, total activity. The linear salt gradients are
shown.
Table 1
The purification scheme for the isoflavone conjugate-hydrolyzing b-glucosidase from soybean roots
Purification step
Total activitya
Total protein
(mg)
Specific activity
(mU/mg protein)
Yield
(%)
Purity
(fold)
(
mU)
Root cell wall wash
Ammonium sulfate
8780
10,540
124
50
70
220
100
120
1
3
(40–60%)
DEAE-Sephadex
CM-Sephadex
2960
1740
12
240
1340
34
20
3.4
20
1.3
a
The enzyme activity was measured by an HPLC method using a natural isoflavone conjugate mixture as substrate. The activity was calculated
from the concentration of the released daidzein.

M.-C. Hsieh, T.L. Graham / Phytochemistry 58 (2001) 995–1005
999
enzyme activity of b-glucosidases (Rotrekl et al., 1999;
Hakulinen et al., 2000), although most b-glucosidases in
Family 1 (Henrissat, 1991) require two essential car-
boxylates for their catalytic activity, contributing to the
general acid/base and nucleophilic catalysis (Zechel and
Withers, 2000). The transition state analogue inhibitor,
glucono-d-lactone, was not very effective. The IC50 with
this inhibitor was 24 mM (data not shown), while the
counterpart in chickpea was more sensitive to this inhi-
¨
bitor, with an IC50 of 1 mM (Hosel and Barz, 1975).
The activity of the ICHG on a variety of glycoside
substrates is summarized in Table 3. This enzyme is able
to hydrolyze a range of synthetic glucosides, but gen-
Fig. 3. SDS-PAGE analysis. Protein samples from each step of pur-
ification were subjected to SDS-PAGE (7.5% acrylamide) under
reducing conditions. lane 1, protein markers; lane 2, crude root pellet
extract; lane 3, 40–60% ammonium sulfate fraction; lane 4, activity
peak from DEAE-Sephadex; lane 5, activity peak from CM-Sephadex.
erally only with relatively low specificity constant (kcat/
K ). Although the enzyme has moderate affinity with
m
several other natural glucosides, the specificity constant
for those tested is far lower than that of the isoflavone
conjugates (Table 3). Interestingly, the enzyme showed
no measurable activity with several flavonol glycosides,
although we did not test the 7-O-substituted flavones or
flavonols. In contrast, commercial almond emulsin b-
glucosidase has much higher specificity toward the syn-
thetic substrate, p-nitrophenyl b-glucoside than toward
the isoflavone conjugates (Table 3). While almond
emulsin b-glucosidase does not hydrolyze the mal-
onylglucosyl isoflavone conjugates, consistent with the
fact that most b-glucosidases only cleave non-modified
terminal glucose residues, the partially purified ICHG
apparently hydrolyzes malonyl isoflavone glucosides as
well as isoflavone glucosides with similar kinetics
(Table 3). The breakdown of isoflavone conjugates is
mediated by two distinct enzymes in chickpea, a mal-
onylesterase and a b-glucosidase, which successively
group modifiers such as iodoacetate and p-chlor-
omercuribenzoate (PCMB) had little or no effect
(
Table 2), possibly due either to the different accessi-
bility of the protein SH groups to these compounds
Esen, 1992) or to their different redox potentials (Joce-
(
lyn, 1972). Consistent with the presence of critical sulf-
hydryl groups, inhibition by silver ions was prevented in
the presence of b-mercaptoethanol. In the presence of
+
2
50 mM b-mercaptoethanol, the IC50 for Ag inhibi-
tion increased 100 fold from 140 mM (without b-mer-
captoethanol) to 14 mM (data not shown). The
importance of SH groups in ICHG is consistent with the
fact that cysteine residues are involved in stability and
Table 2
The effect of inorganic and organic chemicals on the activity of the isoflavone-conjugate hydrolyzing b-glucosidase
Inorganic chemicals
Relative activitya
Organic chemicalsb
Relative
activitya
(
%)
AgNO
HgCl
CaCl
MgCl
3
1 mM
0 mM
1 mM
0 mM
1 mM
0 mM
1 mM
0 mM
1 mM
0 mM
1 mM
0 mM
1 mM
0 mM
1 mM
0 mM
35
0
Iodoacetate
1 mM
10 mM
0.1 mM
1 mM
1 mM
10 mM
1 mM
control
10 mM
Control
1 mM
5%
97
74
1
2
62
23
99
94
92
82
97
PCMB
102
103
99
1
2
Glucono-d-lactone
PMSF (in 95% EtOH)
1
93
89
2
1
97
41
CoCl
2
1
102
99
37
87
MnCl
2
Eserine
DMSO
1
74
99
117
109
120
CuSO
4
EGMME
EGDME
5%
5%
1
115
104
77
ZnSO
4
1
a
Values are the means of two determinations in two independent experiments.
Chemical abbreviations: PCMB, p-chrolomercuribenzoate; DMSO, dimethyl sulfoxide; PMSF, phenylmethylsulfonyl fluoride; EGGME,
ethylene glycol monomethyl ether; EGDME, ethylene glycol dimethyl ether.
b

1000
M.-C. Hsieh, T.L. Graham / Phytochemistry 58 (2001) 995–1005
Table 3
Reaction parameters and substrate specificity of the isoflavone conjugate-hydrolyzing glucosidase (ICHG) compared with that of the glucosidase
from almond emulsin (AEG)
ICHG
AEG
a
a
a
a
Substrates
Structures
K
m
k
cat
ꢁ
k
cat/K
m
K
(M)
m
k
cat
ꢁ1
k
cat/K
m
(s 1)
(M
ꢁ1 ꢁ1
s
)
(s
)
(M
ꢁ1 ꢁ1
s
)
(
M)
3
1
2.2ꢂ10ꢁ3
2
4
p-Nitrophenyl b-glucoside
1.3ꢂ10ꢁ
1.3ꢂ10ꢁ
N.R.b
0.11
8.1ꢂ10
7.2ꢂ10
–
1.2ꢂ10
5.2ꢂ10
2
1
2.6ꢂ10ꢁ3
N.R.
3
p-Nitrophenyl b-galactoside
p-Nitrophenyl b-mannoside
p-Nitrophenyl a-glucoside
p-Nitrophenyl a-galactoside
p-Nitrophenyl a-mannoside
o-Nitrophenyl b-glucoside
0.93
–
6.8
2.6ꢂ10
–
–
N.R.
–
–
N.R.
–
–
8.4ꢂ10ꢁ
9.9ꢂ10ꢁ
1.4ꢂ10ꢁ
2.3ꢂ10ꢁ
3.5ꢂ10ꢁ
8.8ꢂ10ꢁ
6.1ꢂ10ꢁ
3.3ꢂ10ꢁ
5.4ꢂ10ꢁ
9.1ꢂ10ꢁ
4
3
3
4
3
3
3
4
5
5
0.49
0.10
1.4
5.8ꢂ10
1.0ꢂ10
1.0ꢂ10
2.3ꢂ10
6.0ꢂ10
2.8
2
1
3
2
1
N.R.
–
–
c
–
–
3.9ꢂ10ꢁ3
2.3ꢂ10ꢁ3
1.1ꢂ10ꢁ2
5.0
1.3ꢂ10
2.0ꢂ10
7.6ꢂ10
3
4
1
1
4-Methylumbelliferyl b-glucoside
0.05
0.21
0.02
0.15
4.45
4.38
3.07
4.7ꢂ10
Salicin
0.83
Esculin
1
4
4
4
Arbutin
2.4ꢂ10
1.3ꢂ10
8.1ꢂ10
3.4ꢂ10
1
Genistein 7-O-glucoside (genistin)
Genistein 7-O-glucosyl 6-O-malonate (MGG)
Daidzein 7-O-glucoside (daidzin)
1.1ꢂ10ꢁ2
0.65
–
5.9ꢂ10
c
–
(
continued on next page)

M.-C. Hsieh, T.L. Graham / Phytochemistry 58 (2001) 995–1005
1001
Table 3 (continued)
ICHG
AEG
a
a
a
a
Substrates
Structures
K
m
k
cat
ꢁ
k
cat/K
m
K
(M)
m
k
cat
ꢁ1
k
cat/K
m
(s 1)
(M
ꢁ1 ꢁ1
s
)
(s
)
(M
ꢁ1 ꢁ1
s )
(
M)
1.2ꢂ10ꢁ
4
4.34
3.6ꢂ10
4
Daidzein 7-O-glucosyl 6’’-O-malonate (MDG)
Rutin
N.R.
–
–
–
Isoquercitrin
Phloridzin
N.R.
–
N.R.
–
–
–
–
c
Amygdalin
a
Values are means of determinations in three independent experiments.
ꢀ
b
c
N.R.: no reaction at 40 C for 30 min.
m
The products were too small to measure or difficult to calculate K and Vmax.
remove the malonyl and glucose moieties, respectively
Fig. 1). By using the isoflavone conjugate mixture cou-
3. Conclusions
(
pled with the HPLC assay, we are able to examine these
two enzyme activities simultaneously. However, a sepa-
rate malonylesterase activity was never observed in any
soybean tissue extract or at any stage of purification.
Moreover, the lack of esterase activity in the partially
purified IGHG preparation was also demonstrated
when malonylated isoflavone glucosides were the sole
substrates in the assay system with an excess of the glu-
cosidase inhibitor, glucono-d-lactone, over its IC50 value
The hydrolysis of glucosyl and malonylglucosyl iso-
flavone conjugates is biologically relevant. A relatively
high percentage of the total malonylated isoflavone
glucosides is found in the apoplast in soybean (T. L.
Graham, unpublished observation) and in chickpea.
The genistein aglycone is toxic to fungal pathogens,
such as Phytophthora sojae (Rivera-Vargas et al., 1993),
and genistein also may function to activate an extra-
cellular NADH oxidase that functions in defense
potentiation (Graham and Graham, 1999). The quick
release of genistein during pathogen attack by the apo-
plastic isoflavone conjugate-hydrolyzing b-glucosidase
may thus be critically important for several aspects of
disease resistance.
Although previous non-specific b-glucosidases were
reported to hydrolyze isoflavone glucosides in soybean
(Matsuura and Obata, 1993; Matsuura et al., 1995), the
currently purified soybean ICHG has higher specific
activity towards isoflavone conjugates with substrate
specificity constants 160-fold to 1000-fold over p-nitro-
phenyl b-glucoside (Table 3). The b-glucosidase B and C
isolated from soybean cotyledons had a relatively low
(
data not shown). While the current data certainly do
not preclude the existence of a separate isoflavone mal-
onyl esterase in soybean and do not unequivocally sup-
port the existence of both activities in ICHG, it is
possible that the soybean system may differ in this
regard from the chickpea system. In the literature, only
one other b-glucosidase, the flavone 7-O-glucoside-spe-
cific b-glucosidase from ligulate florets of Chamomilla
recutita, can also use apigenin 7-O-acetylglucoside as a
substrate, but the Vmax was reduced dramatically when
the glucose is esterified (Maier et al., 1993). Ultimately,
more detailed characterization of the various enzymes
will be needed to resolve these issues.

1002
M.-C. Hsieh, T.L. Graham / Phytochemistry 58 (2001) 995–1005
reaction rates (79 and 38%, respectively) with isoflavone
glucosides compared to p-nitrophenyl b-glucoside
1 ml of 10% acetonitrile (HPLC grade, Sigma Chemical
Co., MO). The 10% eluent was discarded. Washing
with 1 ml of 20% acetonitrile released bound daidzin,
genistin, MGD and MGG. This provided a preparation
of the isoflavone conjugates free of aglycones (which
remained bound to the column and are eluted only at
higher acetonitrile concentrations). For measurement of
enzyme activity in initial protein fractionations, this
isoflavone conjugate fraction was brought to dryness
and dissolved in an appropriate amount of 100 mM
phosphate–50 mM citrate buffer, pH 5.0 to make the
concentrations of each isoflavone conjugate greater than
2.5 mM as judged by HPLC (Graham, 1991a). For final
purification and characterization of the glucosidase,
individual isoflavone conjugates were isolated by semi-
preparative HPLC using a Merck Lichrosorb RP-18 10
mm C18 column (Alltech Associates, Deerfield, IL) with
non-acidic water and acetonitrile as described by Gra-
ham (1991a).
(
1
100%) (Matsuura and Obata, 1993; Matsuura et al.,
995). The substrate binding affinities of ICGH towards
isoflavone conjugates are comparable to the chickpea
isoflavone-specific b-glucosidase and flavone-glucoside-
cleaving b-glucosidase from Charmomilla recutita (in the
submillimolar range), suggesting favored binding to
¨
their natural substrates (Hosel and Barz, 1975; Maier et
al., 1993). However, chickpea isoflavone-specific b-glu-
cosidase has apparently higher specificity constants
6
toward formononetin and biochanin A glucosides (ꢃ10
ꢁ
1
ꢁ1
+
+
M
s ) than that of ICHG. Furthermore, Hg , Ag
and glucono-d-lactone are all inhibitory to ICGH, con-
sistent with other plant b-glucosidases, although ICGH
+
is less sensitive to Hg and glucono-d-lactone compared
to the chickpea enzyme. Although further characteriza-
tion will be required to firmly determine this, the ICHG
may hydrolyze isoflavone malonylglucosides directly.
This may distinguish it from the chickpea b-glucosidase.
Taken together, we conclude that the b-glucosidase
purified from soybean roots is a novel b-glucosidase
with high specificity toward isoflavone conjugates.
4.3. Buffers
Buffer A, 200 mM sodium phosphate buffer, pH 6.0;
buffer B, 200 mM sodium phosphate buffer, pH 6.0,
containing 0.5 M NaCl; buffer C, 20 mM sodium phos-
phate buffer, pH 6.0, containing 3 mM DTT and 10%
glycerol; buffer D, buffer C containing 1 M NaCl; buffer
E, buffer C containing 5 mM sodium ascorbate; and
buffer F, buffer D containing 5 mM sodium ascorbate.
4
. Experimental
4
.1. Plant material
Soybean (Glycine max L.) cv. Williams was obtained
from Dr. A. F. Schmitthnner (Department of Plant
Pathology, The Ohio State University and the Ohio
Agricultural Research and Development Center, Woos-
ter, OH). Seedlings were grown as described previously
4.4. Enzyme extraction and purification
Frozen roots were ground almost to a powder in a
ꢀ
mortar and pestle pre-chilled to 0 C and extracted three
(
Graham, 1991a). Soybean roots from 7–9 day-old
times with cold buffer A (1 ml buffer/ 1 g tissue). The
combined extracts were centrifuged at 10,000ꢂg for 10
min. The pellet was extracted with buffer B (1 ml buffer/
g tissue) overnight at 4 C. The root residue was further
extracted one more time with an equal volume of buffer
B for another 3 h. The crude extracts were pooled and
seedlings were washed with tap water to remove plant-
ing media and blotted with paper towels. The roots were
frozen at ꢁ20 C until used.
ꢀ
ꢀ
4
.2. Isolation and purification of isoflavone conjugates
1% (w/v) insoluble polyvinylpolypyrolidone (PVPP)
Isoflavone conjugate substrates were prepared as fol-
was added to remove phenolics. This mixture was stir-
red for 3 h on ice. The extract was then filtered through
four-layers of cheesecloth, and clarified by centrifuga-
tion at 10,000ꢂg for 20 min. The resulting solution was
slowly brought to 40–60% saturation with solid ammo-
lows. Soybean cv. Williams seeds (50 g) were ground to
a fine powder with a coffee grinder and the powder was
extracted with 100 ml chloroform for 30 min with stir-
ring. The slurry was filtered through Whatman No. 1
filter paper with suction and washed two more times
with chloroform. The defatted soybean flour was
extracted twice with 50 ml distilled water. The aqueous
extract (100 ml) was clarified by low speed centrifuga-
tion (International Equipment Co., Needham, MA).
The isoflavone conjugates of genistein and daidzein
were partially purified using a C18 PrepSep column
ꢀ
nium sulfate at 4 C followed by gentle stirring over-
night. The pellet was collected by centrifugation at
10,000ꢂg for 20 min, and then dissolved in a minimum
volume of buffer C. The protein solution was clarified
by centrifugation and desalted using a Sephadex G-25
column (1.7ꢂ20 cm) equilibrated with buffer C. The
desalted protein solution was applied to a DEAE-
Sephadex A-50 column (2.8ꢂ35 cm) equilibrated with
the same buffer system and washed with buffer C. The
enzyme was found in the unbound fractions. Sodium
(
Fisher Scientific, Springfield, NJ). The crude aqueous
substrate preparation was passed through the column,
and the column was washed successively with water and

M.-C. Hsieh, T.L. Graham / Phytochemistry 58 (2001) 995–1005
1003
ꢀ
ascorbate was added to the protein solution to a final
concentration of 5 mM, and the solution was then loa-
ded onto a column of CM-Sephadex C-50 (2.8ꢂ19 cm),
equilibrated with buffer E. The elution was carried out
with a 400 ml linear gradient of buffer F in buffer E.
Chromatographic fractions having ICHG activity were
pooled and lyophilized. The concentrated enzyme was
citrate buffer, pH 5.0, was incubated at 40 C for var-
ious time courses. The reaction was stopped by alkali-
zation of the assay mixture with 100 ml of sodium
carbonate, and the mix was then diluted to 1 ml with
distilled water for measurement of absorbance. For
quantitative calculations, the following wavelengths and
ꢁ
1
ꢁ1
" values (mM cm ) were used: p-nitrophenol, 400
nm, "=19.3; o-nitrophenol, 420 nm, "=4.55; 4-methy-
lumbelliferone, 360 nm, "=18.25; and salicyl alcohol,
295 nm, "=3.3. For the substrates, esculin and arbutin,
without known " values, the following method was
developed. The UV spectrum of esculin showed a peak
at 374 nm and the absorbance intensity decreased over
time with enzymatic activity. The activity assay was
conducted as described above, and the absorbance
change was converted to concentration by a linear
standard curve made from various concentrations of
esculin in 100 ml Na CO and 900 ml water. The enzyme
ꢀ
stored at ꢁ20 C for further study.
4
.5. Enzyme and protein assays
Protein concentration was determined by the Brad-
ford (1976) method with BSA as standard. During pur-
ification and unless otherwise noted, ICHG activity was
determined in 0.2 M phosphate–0.1 M citrate buffer, pH
ꢀ
5
.0 at 40 C in a total reaction volume of 10 ml con-
taining 5 ml of isoflavone conjugate mixture. The reac-
tion was stopped at 30 min by adding 90 ml of methanol
and activity was measured by quantifying the release of
aglycones by HPLC (Graham, 1991a). This enzyme
assay allowed us to simultaneously examine the relative
release of aglycones from both daidzein and genistein
conjugates, and also allowed us to measure the possible
esterase activities in fractions leading to the formation
of daidzin and genistin from MGD and MGG, respec-
tively. During the course of the fractionation, the glu-
cosidic activities towards daidzein and genistein
conjugates were always comparable and thus results are
expressed in terms of total aglycone release. One unit of
enzyme activity was defined as the amount of enzyme
that produced 1 mmol of product per minute. Specific
activity was expressed as enzyme units per mg protein.
No apparent conversion of MGD or MGG to daidzin
or genistin was observed in any extract or during any
fractionation step; that is, we never observed an iso-
flavone-specific malonylesterase activity as reported in
2
3
activity against arbutin was determined similarly at 267
nm.
Activity against several natural substrates was
assayed by the HPLC method for isoflavone conjugates
as described above for enzyme purification, except that
the total reaction volume was 15 ml and the reaction was
stopped at various times by adding 85 ml of methanol.
The substrates assayed by HPLC included the four iso-
flavone conjugates, two flavonol glycosides (rutin and
isoquercitrin), phloridzin (a chalcone glucoside), and
amygdalin (a cyanogenic di-glucoside). Commercial
almond emulsin b-glucosidase, a classic non-specific b-
glucosidase, was also examined to compare substrate
specificity. All Km and Vmax values were calculated
based on Lineweaver-Burk plots, and the kcat value is
equal to Vmax divided by the total enzyme concentration
(in molarity) which was calculated using M of 165 kDa
r
for ICHG and 135 kDa for almond emulsin b-glucosi-
dase. The kinetic parameters reported are the mean
values of two or three independent determinations.
The temperature optimum of the ICHG was resolved
by incubating the enzyme solution with isoflavone con-
¨
chickpea (Hosel and Barz, 1975).
In the study of substrate specificity, the concentration
of substrate was 2 mM unless otherwise noted. Daidzein
00
7
-O-glucoside, daidzein 7-O-glucoside-6 -O-malate,
0
0
ꢀ
genistein 7-O-glucoside, and genistein 7-O-glucoside-6 -
O-malate were assayed at 0.2, 0.18, 0.79 and 1.15 mM,
respectively. The appropriate concentration of the
ICHG used for each substrate was determined graphi-
cally from the linear portions of initial reaction velocity.
All substrates were dissolved in 0.2 M phosphate–0.1 M
citrate buffer, pH 5.0, except that p-nitrophenyl a-glu-
coside, 4-methylumbelliferyl b-glucoside, and genistein
jugates over a temperature range from 0 to 90 C for 30
min. Enzyme activity was monitored by the standard
HPLC method. To determine the pH optimum, 4-
methylumbelliferyl b-glucoside was used in various buf-
fers varying in pH values from 2.6 to 10.6. Enzyme
activity was determined spectrophotometrically as
described above.
7
-O-glucoside were dissolved in ethylene glycol mono-
4.6. SDS-polyacrylamide gel electrophoresis (PAGE)
ethyl ether at 50 mM and then diluted to final con-
centration in the same buffer.
Depending on the substrate, activity was measured
either spectrophotometrically or by HPLC. For those
substrates releasing an aglycone with known molar
coefficient (E), the assay solution containing the enzyme
SDS-PAGE was performed as described by Laemmli
(1970) using a Mini-Protein II electrophoresis system
(Bio-Rad). Analyses were carried out under non-redu-
cing or reducing (mercaptoethanol) conditions as speci-
fied. The molecular markers used were bovine serum
albumin (66 kD), ovalbumin (45 kD), and carbonic
(
10 ml) and substrate (90 ml) in 0.2 M phosphate–0.1 M

1004
M.-C. Hsieh, T.L. Graham / Phytochemistry 58 (2001) 995–1005
anhydrase (29 kD) all from Sigma Chemical Co. Protein
gels were stained with silver stain (Bio-Rad).
glucosidase-DIMBOA, -DIMBOAGlc, and -dhurrin complexes.
Proc. Natl. Acad. Sci. 97, 13555–13560.
Dakora, F.D., Phillips, D.A., 1996. Diverse functions of isoflavonoids
in legumes transcend antimicrobial definitions of phytoalexins.
Physiol. Mol. Plant Pathol. 49, 1–20.
4
.7. Inhibition assays
˜
Day, A.J., Canada, F.J., Dıaz, J.C., Kroon, P.A., Mclauchlan, R.,
´
Several inorganic and organic compounds and sol-
Faulds, C.B., Plumb, G.W., Morgan, M.R.A., Williamson, G., 2000.
Dietary flavonoids and isoflavone glycosides are hydrolysed by the
lactase site of lactase phlorizin dydrolase. FEBS Lett. 468, 166–170.
Dharmawardhana, D.P., Ellis, B.E., Carlson, J.E., 1995. A b-glucosi-
dase from lodgepole pine xylem specific for the lignin precursor
coniferin. Plant Physiol. 107, 331–339.
vents were examined for their inhibitory effects. The
reaction mixture was in 0.2 M phosphate–0.1 M citrate
buffer, pH 5.0 using a total reaction volume of 15 ml
ꢀ
incubated at 40 C for 30 min; the reaction was stopped
by adding 85 ml of methanol. The enzyme activity was
measured by HPLC and designated as 100% in the
absence of added chemicals. The activity reported is the
average from two independent measurements. A series
of concentrations of Ag , b-mercaptoethanol and glu-
cono-d-lactone, respectively, were further inspected to
Duroux, L., Delmotte, F.M., Lancelin, J.-M., Keravis, G., Jay-Alle-
´
mand, C., 1998. Insight into naphthoquinone metabolism: b-gluco-
sidase-catalysed hydrolysis of hydrojuglone b-d-glucopyranoside.
Biochem. J. 333, 275–283.
Ebel, J., 1986. Phytoalexin synthesis: the biochemical analysis of the
induction process. Annu. Rev. Phytopathol. 24, 235–264.
Esen, A., 1992. Purification and partial characterization of maize (Zea
mays L.) b-glucosidase. Plant Physiol. 98, 174–182.
Esen, A., 1993. b-Glucosidases, overview. In: Esen, A. (Ed.), b-Glu-
cosidase: Biochemistry and Molecular Biology. American Chemical
Society, Washington, DC, pp. 1–13.
+
determine the concentration for 50% inhibition (IC ).
For prevention of inhibition of activity by silver ions,
50 mM of b-mercaptoethanol was incubated with 5 ml
5
0
2
of isoflavone-conjugate substrate and 5 ml of various
concentrations of Ag for 20 min before the enzyme
Graham, T.L., 1991a. A rapid, high resolution HPLC profiling pro-
cedure for plant and microbial aromatic secondary metabolites.
Plant Physiol. 95, 584–593.
+
was introduced. IC50 values are obtained from the least
square fit to the Langmuir isotherm equation, v /v =1/
Graham, T.L., 1991b. Flavonoid and isoflavonoid distribution in
developing soybean seedling tissues and in seed and root exudates.
Plant Physiol. 95, 594–603.
i
0
(
Inc., OR), where v is the velocity in the presence of
1+[I]/IC ), using Igor Pro software (WaveMetrics,
50
i
Graham, T.L., Graham, M.Y., 1996a. Signaling in soybean phenyl-
propanoid responses. Dissection of primary, secondary, and con-
ditioning effects of light, wounding, and elicitor treatments. Plant
Physiol. 110, 1123–1133.
inhibitor at concentration [I] and v is the velocity in the
0
absence of inhibitor.
Graham, T. L., Graham, M. Y., 1996b. Cell surface reduction of tet-
razolium dyes induces elicitation competency in soybean. In: 8th
International Congress Molecular Plant–Microbe Interactions,
Knoxville, TN, 14-19 July 1996.
Acknowledgements
Graham, T.L., Graham, M.Y., 1999. Role of hypersensitive cell death
in conditioning elicitation competency and defense potentiation.
Physiol. Mol. Plant Pathol. 55, 13–20.
We would like to acknowledge frequent stimulating
conversations with and useful suggestions for the work
from Dr. Madge Graham. Funding for this project was
provided by grants from the USDA-NRI and the North
Central Biotechnical Consortium. Salaries and research
support were also provided by State and Federal Funds
appropriated to the Ohio Agricultural Research and
Development Center, the Ohio State University.
Graham, T.L., Graham, M.Y., 2000. Defense potentiation and elici-
tation competency: redox conditioning effects of salicylic acid and
genistein. In: Stacey, G., Keen, N.T. (Eds.), Plant Microbe Interac-
tions, Vol. 5. APS Press, St. Paul, MN, pp. 181–220.
Graham, T.L., Kim, J.E., Graham, M.Y., 1990. Role of constitutive
isoflavone conjugates in the accumulation of glyceollin in soybean
infected with Phytophthora megasperma. Mol. Plant–Microbe
Interactions 3, 157–166.
Hakulinen, N., Paavilainen, S., Korpela, T., Rouvinen, J., 2000. The
crystal structure of b-glucosidase from Bacillus circulans sp. alkalo-
philus: Ability to form long polymeric assemblies. J. Struct. Biol.
129, 69–79.
References
Barz, W., Welle, R., 1992. Biosynthesis and metabolism of isoflavones
and pterocarpan phytoalexins in chickpea, soybean, and phyto-
pathogenic fungi. Recent Adv. Phytochem. 26, 139–164.
Bell, E.A., 1981. The physiological role(s) of secondary (natural) pro-
ducts. In: Stumpf, P.K., Conn, E.E. (Eds.), The Biochemistry of
Plants, Secondary Plant Products, Vol. 7. Academic Press, New
York, pp. 1–19.
Harborne, J.B., Mabry, T.J., 1982. The Flavonoids: Advances in
Research. Chapman and Hall, London.
Henrissat, B., 1991. A classification of glycosyl hydrolases based on
amino acid sequence similarities. Biochem. J. 280, 309–316.
¨
Hinderer, W., Koster, J., Barz, W., 1986. Purification and properties
00
of a specific isoflavone 7-O-glucoside-6 -malonate malonylesterase
from roots of chickpea (Cicer arietinum L.). Arch. Biochem. Bio-
phys. 248, 570–578.
Bradford, M.M., 1976. A rapid and sensitive method for quantitation
of microgram quantities of protein utilizing the principle of protein-
dye binding. Anal. Biochem. 72, 248–254.
¨
Hosel, W., Barz, W., 1975. b-Glucosidases from Cicer arietinum L.:
Conn, E.E., 1993. b-Glucosidases in plants. In: Esen, A. (Ed.), b-Glu-
cosidase: Biochemistry and Molecular Biology. American Chemical
Society, Washington, DC, pp. 14–26.
purification and properties of isoflavone-7-O-glucoside-specific-b-
glucosidases. Eur. J. Biochem. 57, 607–616.
¨
Hosel, W., Surholt, E., Borgmann, E., 1978. Characterization of b-
Czjzek, M., Cicek, M., Zamboni, V., Bevan, D.R., Henrissat, B., Esen,
A., 2000. The mechanism of substrate (aglycone) specificity in b-
glucosidases is revealed by crystal structures of mutant maize b-
glucosidase isoenzymes possibly involved in lignification from
chickpea (Cicer arietinum L.) cell suspension cultures. Eur. J. Bio-
chem. 84, 487–492.

M.-C. Hsieh, T.L. Graham / Phytochemistry 58 (2001) 995–1005
1005
Hungria, M., Stacey, G., 1997. Molecular signals exchanged between
host plants and rhizobia: basic aspects and potential application in
agriculture. Soil Biol. Biochem. 29, 819–830.
Maier, R., Carle, R., Kreis, W., Reinhard, E., 1993. Purification and
characterization of a flavone 7-O-glucoside-specific glucosidase
from ligulate florets of Chamomilla recutita. Planta Med. 59, 436–
441.
Inoue, K., Ebizuka, Y., 1996. Purification and characterization of
furostanol glycoside 26-O-b-glucosidase from Costus speciosus rhi-
zomes. FEBS Lett. 378, 157–160.
Matsuura, M., Obata, A., 1993. b-Glucosidases from soybeans
hydrolyze daidzin and genistin. J. Food Sci. 58, 144–147.
Matsuura, M., Sasaki, J., Murao, S., 1995. Studies on b-glucosidases
from soybeans that hydrolyze daidzin and genistin: isolation and
characterization of an isozyme. Biosci. Biotech. Biochem. 59, 1623–
1627.
Johnson, G., Maag, D.D., Johnson, D.K., Thomas, R.D., 1976. The
role of phytoalexins in the resistance of sugarbeet (Beta vulgaris) to
Cercospora beticola. Physiol. Plant Pathol. 8, 225–230.
Jocelyn, P.C., 1972. Biochemistry of the SH group. The Occurrence,
Chemical Properties, Metabolism and Biological Function of Thiols
and Disulphides. Academic Press, New York, pp. 404.
Morris, P.F., Ward, E.W.B., 1992. Chemoattraction of zoospores of
the soybean pathogen, Phytophthora sojae, by isoflavones. Physiol.
Plant Pathol. 40, 17–22.
Kessmann, H., Edwards, R., Geno, P.W., Dixon, R.A., 1990. Stress
responses in alfalfa (Medicago sativa L.). V. Constitutive and eli-
citor-induced accumulation of isoflavonoid conjugates in cell sus-
pension cultures. Plant Physiol. 94, 227–232.
Phillips, D. A., Streit, W., 1996. Legume signals to rhizobial sym-
bionts: a new approach for defining rhizosphere colonization. In:
Stacey, G., Keen, N. (Eds.), Plant–Microbe Interactions, Vol. 1. pp.
236–271.
Kleczkowski, K., Schell, J., 1995. Phytohormone conjugates: nature
and function. Crit. Rev. Plant Sci. 14, 283–298.
Phillips, D.A., Stafford, H.A., Ibrahim, R.K., 1992. Flavonoids: plant
signals to soil microbes. Recent Adv. Phytochem. 26, 201–231.
Poulton, J.E., 1990. Cyanogenesis in plants. Plant Physiol 94, 401–405.
Rivera-Vargas, L.I., Schmitthenner, A.F., Graham, T.L., 1993. Soy-
bean flavonoid effects on and metabolism by Phytophthora sojae.
Phytochemistry 32, 851–857.
Koes, R.E., Quattrocchino, F., Mol, J.N.M., 1994. The flavonoid
biosynthetic pathway in plants: function and evolution. BioEssays
16, 123–132.
Kramer, R.P., Hindorf, H., Jha, H.C., Kallage, J., Zilliken, F., 1984.
Antifungal activity of soybean and chickpea isoflavones and their
reduced derivatives. Phytochemistry 23, 2203–2205.
Rotrekl, V., Nejedla, E., Kucera, I., Abdallah, F., Palme, K., Brzo-
bohaty, B., 1999. The role of cysteine residues in structure and
enzyme activity of a maize b-glucosidase. Eur. J. Biochem. 266,
1056–1065.
Laemmli, U.K., 1970. Cleavage of structural proteins during the
assembly of the head of the bacteriophage T4. Nature 227, 680–685.
Leah, R., Kigel, J., Svendsen, I., Mundy, J., 1995. Biochemical and
molecular characterization of a barley seed b-glucosidase. J. Biol.
Chem. 270, 15789–15797.
Weidenborner, M., Hindorf, H., Jha, H.C., Tsotsonos, P., Egge, H.,
1990. Antifungal activity of isoflavonoids in different reduced stages
on Rhizoctonia solani and Sclerotium rolfsii. Phytochemistry 29,
801–803.
Mackenbrock, U., Barz, W., 1991. Elicitor-induced formation of
pterocarpan phytoalexins in chickpea (Cicer arietinum L.) cell sus-
pension cultures from constitutive isoflavone conjugates upon inhi-
bition of phenylalanine ammonia lyase. Z. Naturforsch. 46c, 43–50.
Zechel, D.L., Withers, S.G., 2000. Glycosidase mechanisms: anatomy
of a finely tuned catalyst. Acc. Chem. Res. 33, 11–18.