Purification and characterization of a flavonol 3-O-β heterodisaccharidase from the dried herb of Fagopyrum esculentum Moench

Article abstract of DOI:10.1016/S0031-9422(03)00418-7

A flavonol-3-O-β-heterodisaccharide glycosidase (FHG I) was isolated from dried aerial tissues of Fagopyrum esculentum Moench (Fagopyri herba). It has a specific enzyme activity of ca. 3.5 nkat mg^-1 protein in buffered extracts when rutin (quercetin-3-O-rutinoside) was used as substrate and an optimal enzyme activity was seen at around pH 4.8 and 30 °C. FHG I was purified about 156-fold to apparent homogeneity by hydrophobic interaction, anion exchange and size exclusion chromatographic steps. The apparent molecular mass of FHG I was 74.5 ± 2 kDa as determined by SDS-PAGE and it is a monomeric glycoprotein with a carbohydrate content of 23%. The isoelectric point as determined by isoelectric focusing was 5.7 and the energy of activation was 32 kJ mol^-1. FHG I exhibits a high substrate specificity, preferring flavonol 3-O-glycosides comprising the disaccharide rutinose. The K_m and V_max values for the natural substrate rutin were calculated to be 0.561 μM and 745 nkat mg ^-1 protein, respectively. Two oligopeptide fragments obtained after enzymatic digestion of FHG I were sequenced and showed similarities to sequences of β-glucohydrolases from other plant species. Polyclonal antibodies were raised and their specificities determined. Another flavonol 3-O-β -heterodisaccharide glycosidase (FHG II) could also be detected in buckwheat herb, having a molecular mass of 85.3 ± 2 kDa and an isoelectric point between pH 6.0 and 6.5.

Full text of DOI:10.1016/S0031-9422(03)00418-7

Phytochemistry 64 (2003) 411–418
www.elsevier.com/locate/phytochem
Purification and characterization of a flavonol 3-O-b-
heterodisaccharidase from the dried herb of Fagopyrum esculentum
Moench
a
b
a
a,
Andreas Baumgertel , Rudi Grimm , Wilhelm Eisenbeiß , Wolfgang Kreis *
a
Lehrstuhl fu¨r Pharmazeutische Biologie, Friedrich-Alexander-Universita¨t Erlangen-Nu¨rnberg, Staudtstrasse 5, 91058 Erlangen, Germany
bTOPLAB Gesellschaft fu¨r angewandte Biotechnologie, Am Klopferspitz 19, 82152 Martinsried, Germany
Received 14 January 2003; received in revised form 19 May 2003
Dedicated to the memory of Professor Jeffrey B. Harborne
Abstract
A flavonol-3-O-b-heterodisaccharide glycosidase (FHG I) was isolated from dried aerial tissues of Fagopyrum esculentum
Moench (Fagopyri herba). It has a specific enzyme activity of ca. 3.5 nkat mg protein in buffered extracts when rutin (quercetin-3-
ꢀ1
ꢁ
O-rutinoside) was used as substrate and an optimal enzyme activity was seen at around pH 4.8 and 30 C. FHG I was purified
about 156-fold to apparent homogeneity by hydrophobic interaction, anion exchange and size exclusion chromatographic steps.
The apparent molecular mass of FHG I was 74.5ꢂ2 kDa as determined by SDS-PAGEand it is a monomeric glycoprotein with a
carbohydrate content of 23%. The isoelectric point as determined by isoelectric focusing was 5.7 and the energy of activation was
3
2 kJ mol 1. FHG I exhibits a high substrate specificity, preferring flavonol 3-O-glycosides comprising the disaccharide rutinose.
ꢀ
ꢀ1
The Km and Vmax values for the natural substrate rutin were calculated to be 0.561 mM and 745 nkat mg
protein, respectively.
Two oligopeptide fragments obtained after enzymatic digestion of FHG I were sequenced and showed similarities to sequences of b-
glucohydrolases from other plant species. Polyclonal antibodies were raised and their specificities determined. Another flavonol 3-
O-b-heterodisaccharide glycosidase (FHG II) could also be detected in buckwheat herb, having a molecular mass of 85.3ꢂ2 kDa
and an isoelectric point between pH 6.0 and 6.5.
#
2003 Elsevier Ltd. All rights reserved.
Keywords: Fagopyrum esculentum; Polygonaceae; Rutinoside; Rutin; b-Glycosidase; Purification
1
. Introduction
from common and tartary buckwheat seeds (Hay et al.,
961; Kurusawa et al., 1973; Bourbouze et al., 1974,
1
Rutin (quercetin 3-O-rutinoside) 1 is the dominant
flavonol glycoside in the leaves and flowers of the com-
mon buckwheat, Fagopyrum esculentum Moench
1976; MacDonald et al., 1984; Yasuda and Nakagawa,
1994). However, rutin-cleaving enzymes have not been
investigated so far in leaves of Fagopyrum esculentum
Moench or in Fagopyri herba, which is the commer-
cially available dried buckwheat herb, although it is well
known that the rutin 1 content of the drug may decrease
after harvest (Nguyen-Hiep et al., 1964; Noguchi and
Mori, 1969). From previous studies on the stability of
rutin 1 in Fagopyri herba it may be deduced that the
disappearance of the most important ingredient of
Fagopyri herba, namely rutin 1, and the transient
appearance of quercetin 2 is a result of enzyme-cata-
lyzed reactions, involving glycosidases and oxidases.
In the present paper we report for the first time the
purification to apparent homogeneity and the molecular
(
Couch et al., 1946). Rutin 1 exhibits antioxidative,
antihemorrhagic and blood vessel protecting properties
e.g., Gabor, 1972; Wojcicki et al., 1995a, 1995b; Rice-
(
Evans et al., 1997). Therefore, the dried aerial parts of
common buckwheat are used as tea preparations or in
tablets for the treatment of vascular diseases and circu-
latory disorders (Ihme et al., 1996). Rutin 1 can be
hydrolysed by microorganisms and enzymes isolated
*
Corresponding author. Tel.: +49-9131-8528241; fax: +49-9131-
528243.
E-mail address: wkreis@biologie.uni-erlangen.de (W. Kreis).
8
0
031-9422/03/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0031-9422(03)00418-7

412
A. Baumgertel et al. / Phytochemistry 64 (2003) 411–418
characterization of a flavonol-3-O-b-heterodisaccharide
glycosidase isolated from dried buckwheat herb (Fago-
pyri herba)
2
. Results
2
.1. Enzyme purification
After incubating Fagopyri herba in water or different
buffer solutions, rutin 1 was degraded fairly rapidly
Fig. 1), and quercetin 2 and rutinose 3 were identified
(
as degradation products (data not shown). Rutin 1 was
not degraded under the same conditions in drug samples
heated prior to fermentation (Fig. 1) indicating that the
degradation seen in the untreated samples can be
attributed to an enzymatic process. The activity
observed in crude protein extracts reached about 3.5
nkat/mg protein. Acetone powders prepared from
Fagopyri herba were still quite active and served as the
starting material for enzyme purification. Rutin 1, a
flavonol 3-O-heterodisaccharide, was the preferred sub-
strate of the two enzymes characterized here, which
were hence termed flavonol-3-O-heterodisaccharide glu-
cohydrolases I and II (FHG I and II).
The major flavonol-3-O-heterodisaccharide glucohy-
drolase (FHG I) was successfully purified 156-fold from
acetone powders prepared from dried buckwheat herb,
with a total recovery of 16.4% (Table 1). Hydrophobic
interaction chromatography on Phenylsepharose was
the first chromatographic step used to separate FHG I
from other proteins. The activity peaked at 0.34 M
(NH ) SO (Fig. 2A). Fractions showing FHG activity
were pooled, dialysed against water and subjected to
anion exchange chromatography on Resource Q for
4
2
4

A. Baumgertel et al. / Phytochemistry 64 (2003) 411–418
413
further purification. A NaCl gradient between 0 and 0.7
M was then applied, and FHG I was released from the
matrix at a NaCl concentration of 0.32 M (Fig. 2B).
The active fractions were again combined, concentrated
by membrane filtration and subjected to gel filtration,
where the rutin-hydrolysing activity eluted as a single
peak between 10 and 18 ml (Fig. 2C). The size of the
holoenzyme was estimated to be approximately 80 kDa.
masses of the native protein and its deglycosylated
derivative seen in SDS-PAGE. The molecular mass of
the deglycosylated protein was 58 kDa indicating a
sugar content of 23% (Fig. 4).
The pI of the FHG I as determined by isoelectric
focusing was 5.7 (Fig. 5).
Using rutin 1 as a substrate, pH and temperature
dependencies of the purified enzyme were investigated.
FHG I activity was optimal at pH 4.8 with half maximal
values at pH 3.9 and 7.0 and proved to be stable over a
2.2. Molecular mass, isoelectric point, pH-optimum,
temperature optimum and energy of activation
ꢁ
wide temperature range, namely from 4 to 50 C. The
ꢁ
ꢀ1
enzyme worked best at around 30 C (42 nkat mg pro-
tein), but surprisingly high enzyme activity was seen at
The purification steps were accompanied by SDS-
PAGEin order to control the efficacy of the respective
purification steps. The active fraction 14 obtained after
gel filtration showed a single polypeptide band at
ꢁ
ꢀ1
ꢁ
4 C (37 nkat mg protein) and was still active at 70 C
(11 nkat mg protein). Enzyme activities determined
ꢀ1
7
4.5ꢂ2 kDa (Fig. 3). SDS-PAGEunder reducing con-
ditions yielded similar results, suggesting that FHG I is
a monomer (data not shown).
Purified FHG I was treated with trifluoromethane
sulphonic acid to remove sugar residues covalently
attached to the protein. The carbohydrate content of
FHG I was then estimated by comparing the molecular
Fig. 1. Rutin 1 and quercetin 2 concentration after incubation of
ꢁ
powdered crude drug in McIlvaine buffer pH 6.0 at 25 C.
Table 1
Purification of flavonol 3-O-b-heterodisaccharidase I (FHG I) from
the dried herb of Fagopyrum esculentum (buckwheat)
Stage
Protein Total
[
Specific Recovery Purification
(fold)
mg] activity activity [%]
[
nkat] [nkat/mg
protein]
Fig. 2. Elution profiles of flavonol 3-O-b-heterodisaccharidase I
(
FHG I) purified from the dried herb of Fagopyrum esculentum
buckwheat). Acetone-precipitated protein was subjected to a three-
(
Crude extract 21.1
Phenylsepharose 1.4
66
33.1
3.1
23.6
29.5
100
50
1
7.6
9.5
156
step chromatographic procedure. (A) Hydrophobic interaction chro-
matography on Phenylsepharose, (B) anion exchange chromatography
on Resource Q and (C) size exclusion chromatography on Superdex
Resource Q
Superdex 200
0.98 28.9
0.023 10.8
44.7
16.4
468
2
00. Enzyme activity was checked using 1 mM rutin 1 as the substrate.

414
A. Baumgertel et al. / Phytochemistry 64 (2003) 411–418
Fig. 4. SDS-PAGEanalysis of native and deglycosylated FHG I.
Tracks 1 and 2 show molecular weight markers; track 3 is FHG I
before, track 4 FHG I after trifluoromethane sulphonic acid treat-
ment.
Fig. 3. SDS-PAGEanalysis of the purification steps of FHG I. Track
and 2 show molecular weight markers. Track 3 shows the crude
1
extract, track 4 active pooled fractions of hydrophobic interaction
chromatography, track 5 the active pooled fractions of anion exchange
chromatography, and track 6 the most active fraction (fraction 14) of
size exclusion chromatography (see Fig. 2).
ꢁ
between 4 and 30 C were used to calculate the energy
ꢀ
of activation, which was 32 kJ mol 1
.
2
.3. Enzyme inhibition and substrate specifity
The substrate preferences of FHG I were studied with
various other flavonol 3-O-glycosides 4–8. Table 2
summarizes the kinetic parameters determined. Taking
low Km values as an indication of affinity, FHG I
showed its highest affinity for kaempferol 3-O-mono-
glycosides, with Km values being up to 19 times lower
than those with kaempferol 3-O-disaccharides (5).
Surprisingly, quercetin 3-O-monosaccharides and
quercetin 3-O-sophoroside were not hydrolyzed at all
by FHG I (data not shown), and sugars attached to
the phenolic hydroxyl group at C7 of the flavonoid
structure were not removed by FHG I. Cardiac gly-
cosides, cyanogenic glycosides, oligosaccharides and
p-nitrophenyl b-d-glucoside were not accepted as sub-
strates either. Flavonol 3-O-rutinosides (1,4,5) were all
cleaved by FHG I, however, the affinity of the enzyme
seems to depend on the substitution pattern of the
flavonol B-ring.
Fig. 5. Determination of the isoelectric point of FHG I using
isoelectric focusing.
Vmax/K ratios, which may be taken as a measure of the
m
efficacy of the enzyme in the turn-over of a given sub-
strate, were higher for disaccharides than for mono-
saccharides. Rutin 1, probably the natural substrate of
FHG I, was degraded with a Vmax/Km ratio of 1291.
Using the relative velocity as a measure of specificity,
kaempferol rutinoside 5 was hydrolysed best, among the
other possible substrate tested.
With regard to reaction velocities and relative activ-
ities, flavonol 3-O-disaccharides were better substrates
for FHG I than the respective monosaccharides. The
Rutinose 1 was the only sugar released upon incu-
bation of rutin 1 in the presence of FHG I; glucose or
rhamnose could not be detected.
Table 2
Kinetic constants and specific activity of the FHG I from buckwheat herb. (Amounts calculated as kaempferol equivalents)
Substrate
Product
K
m
V
max
V
mM] [nkat/mg] [nkat/mg] [nkat/mg protein] [%]
max/K
m
Specific activity Specific activity
[
Quercetin 3-O-rutinoside (Rutin) (1)
Isorhamnetin 3-O-rutinoside (4)
Kaempferol 3-O-rutinoside (5)
Quercetin (2)
Isorhamnetin (10)
Kaempferol (11)
0.561 724
0.375 21
0.950 1211
0.533 572
0.050 42
1291
56
468
15
100
3
1275
1073
840
650
374
38
134
80
8
Kaempferol 3-O-sophoroside-7-glucoside (6)
Kaempferol 3-O-glucoside (7)
Kaempferol 7-O-glucoside (12)
Kaempferol (11)
Kaempferol 3-O-glucoside-7-O-rhamnoside (8) Kaempferol 7-O-rhamnoside (13) 0.060 50
833
47
10

A. Baumgertel et al. / Phytochemistry 64 (2003) 411–418
415
The degradation of rutin 1 was only weakly inhibited
by quercitrin (quercetin 3-O-rhamnoside 9) or the semi-
synthetic flavonol ether troxerutin (data not shown).
The influence on FHG I of alcohols used as solvents for
the preparation of herbal extracts was also investigated
degrade the compounds in question (Nguyen-Hiep et
al., 1964; Noguchi and Mori, 1969). Endogenous plant
enzymes may in this way influence the quality and ther-
apeutic value of drugs and drug preparations.
We started our investigations by fermenting heat-
treated and untreated samples of Fagopyri herba in
buffer solutions and found a dramatic decrease in rutin
1 in the untreated samples. This degradation was still
seen in crude protein extracts containing 33% MeOH or
and methanol–H O at 1:2 (v/v) led to a loss of activity
2
of about only 20%. However, the enzyme was com-
pletely inactivated by methanol–H O at > 1:1 (v/v) and
2
with ethanol–H O > 1:2 (v/v).
2
ꢁ
in buffered protein extracts stored at 4 C. This obser-
2
.4. FHG I Antibodies and protein sequence
vation is of special pharmaceutical interest since it
shows that stability of rutin 1 cannot be achieved under
standard extraction and storage conditions. Assuming
that the degradation of rutin 1 starts with the initial
removal of the sugar moieties, we aimed at the isolation
of a flavonol 3-O-glycosidase. Using the 3-step purifica-
tion method described here, we succeeded in isolating
two enzymes, termed FHG I and FHG II, which are
responsible for rutin 1 hydrolysis in Fagopyri herba.
Comparing the properties of FHG I with those of
enzymes isolated from the seeds of common and tartary
buckwheat (Bourbouze et al., 1974, 1976; Yasuda and
Nakagawa, 1994), we found that the buckwheat glyco-
sides described so far share some characteristics. The
molecular masses of a rhamnodiastase and an a-l-
rhamnosidase from Fagopyrum esculentum seeds, and
the rutin-degrading enzymes RDEI and RDEII from
Fagopyrum tartaricum are similar to the masses calcu-
lated for FHG I and FHG II in the present investiga-
tion. Other plant b-glucosidases of similar molecular
masses are also known (see Schliemann, 1987, for a
review). FHG I seems to be a monomer which corre-
sponds to the previous findings of monomeric organi-
sation of RDEI and II in tartary buckwheat seeds.
RDEI and II contained 11 and 14% carbohydrates,
respectively, as compared to 23% determined for FHG
I in the present study. Isoelectric points have not been
published for RDEI and II from tartary buckwheat
seeds. The isoelectric points for glycosidases from com-
mon buckwheat seeds were around 3.7 (Bourbouze et
al., 1974), which indicates that these enzymes are not
identical with the FHGs described here. But, Bourbouze
et al. (1974) did not use rutin 1 as the substrate to trace
glycosidase activity (see below). On the other hand, iso-
electric points determined here for the two FHGs are
well in the range of those reported for other plant b-
glucosidases (Schliemann, 1987; Esen, 1993). Somewhat
striking was the observation that FHG I does not
hydrolyse cyanogenic glycosides, although it shares
considerable sequence homology with cyanogenic b-
glucosidases (Oxtoby et al. 1991; Li et al., 1992; Zheng
and Poulton, 1995; Zhou et al., 2002).
FHG I purified from Fagopyri herba was used to
raise polyclonal antibodies in mice. The antisera
obtained contained antibodies recognizing FHG I in
unreduced and reduced samples of purified FHG I and
crude protein extracts (data not shown). Under the
conditions used, immune globulins present in the pre-
immune sera did not bind to buckwheat proteins. In
addition to FHG I, a 87-kDa protein was recognized by
the immune globulins present in the sera after immuni-
zation of the experimental animals with purified FHG I.
This particular protein (appearing in fraction 18 toge-
ther with FGH I) could be separated from FHG I in
SDS-PAGE, and it was found that it also was capable
of hydrolyzing rutin 1 (data not shown). Therefore, this
enzyme was named FHG II. The isoelectric point of
FHG II, as determined by isoelectric focusing, was at
about pH 6.3. FHG II was not characterized further.
Finally, the purified FHG I was partially sequenced.
The protein, which was blocked at the N-terminus, was
digested using the endoproteinase LysC from Lyso-
bacter enzymogenes. This digest yielded several oligo-
peptides, two of which were sequenced: a 15 amino acid
fragment (fragment 1: LIVGDRLPTFTDEQT) and a
1
0 amino acid fragment (fragment 2: VSGGINQAGI).
Comparing amino acid sequences of known glucosi-
dases with the information obtained from our sequen-
cing experiments revealed high homology of FHG I
with prunasin hydrolase isoforms from Prunus serotina
(
E-values <0.1 for fragment 1) and various other
putative) b-glucohydrolases (E-values <1.0 for frag-
(
ment 1). The properties described so far let us assume
that the FHG I isolated here is a member of the Family
1
of glycoside hydrolases [EC 3.2.1.21.].
3
. Discussion
The importance of endogenous enzymes on the stabi-
lity of herbal drugs and their biological active ingre-
dients has rarely been investigated in the past. After
harvest and during drying and processing of a given
drug the destruction of cell compartments may result in
the degradation of important drug constituents, because
enzymes still active in the drug may modify or even
The substrate preferences of rutin-hydrolysing
enzymes have been investigated in some detail. The gly-
cosidases purified from common buckwheat by Bour-
bouze et al. (1974, 1976) were capable of hydrolysing

416
A. Baumgertel et al. / Phytochemistry 64 (2003) 411–418
a-rhamnosides, but also the disaccharide rutinose,
amygdalin, various p-nitrophenylglycosides and querce-
tin-3-sophoroside 9. In contrast, only a small number of
substrates tested were accepted by the FHG I described
here. Cyanogenic glycosides, 7-O-glycosides, p-nitro-
phenyl b-d-glucoside, cardiac glycosides, and oligo-
saccharides were not hydrolysed by FHG I. The
glycosidases described more recently (Yasuda and
Nakagawa, 1994) were tested with rutin as the sub-
strate, and both rutin-degrading enzymes showed a high
specificity for this particular substrate.
(Mannheim, Germany). Rutinose hepatacetate was
obtained from Sigma (Steinheim, Germany). All other
reagents used were obtained from Merck (Darmstadt,
Germany) or Sigma (Steinheim, Germany).
The dried aerial tissues of Fagopyrum esculentum
Moench (buckwheat herb, Fagopyri herba; Poly-
gonaceae) were obtained from Fink GmbH (Glaxo
SmithKline), Herrenberg, Germany.
4.2. Flavonoid and sugar analysis
As expected, the sugar residues attached to the 3-b-O-
position have an influence on reaction velocity and affi-
nity. Moreover, we found here that the structure of the
Flavonoids were extracted with MeOH-H O (4:1) as
2
described by Hagels et al. (1995) and the extracts then
analyzed by HPLC. The flavonoids were separated on
an RP 18 column (Lichrosphere 100, 5 mm, 125x 4,
Merck) equilibrated with solution A (aq. phosphoric
acid 0.85% (v/v)+4% (v/v) acetonitrile). The flavo-
noids were eluted with solution B (acetonitrile contain-
ing 12.5% (v/v) phosphoric acid 0.85% (v/v)) using the
following gradient: 0 min 6% B, 5 min 15% B, 11 min
15% B, 16.5 min 24% B, 22 min 41% B, 25 min 41% B,
33 min 6% B and detected by their UV-absorption at
340 nm. Flavonoids were identified on the basis of their
retention times in HPLC by direct chromatographic
comparison with authentic samples.
0
aglycone, especially the group attached in the 3 position
of the B-ring, influences the catalytic activity of FHG I.
Different enzymes may be involved in the degradation
or modification of genuine flavonol glycosides in Fago-
pyri herba. Glycosidases responsible for the hydrolysis
of secondary plant products have been shown to possess
a high degree of specificity. The cleavage of rutin 1 into
quercetin 2 and rutinose 3 is the first step in enzymatic
rutin 1 degradation (Surholt and Hosel, 1978). The
¨
aglycone released by this reaction may be further
degraded by so-called ‘‘flavonol oxidases‘‘ (Barz and
Ko
¨
ster, 1981) yielding products such as 3,4-dihydroxy-
Sugars were analysed by TLC on silica gel 60 using
toluene:acetone (4:1; v/v) as the mobile phase and
thymol-H SO for detection. Sugars were acetylated
(Blakeney et al., 1983) prior to analysis since rutinose
was commercially available only as its heptaacetate.
benzoic acid and 2,4,5-trihydroxybenzoic acid. These
compounds were not detected until after prolonged
incubation of quercetin 2 by protein extracts prepared
from buckwheat herb (data not shown). This may
explain the observation that no increase in the quercetin
2
4
2
content of buckwheat herb is seen although rutin 1
4.3. Enzyme assays
eventually disappears. Enzymatic degradation of biolo-
gically active drug constituents may occur at various
stages of a pharmaceutical process. As shown here,
specific enzymes may even be active in ethanol extracts
or at low temperatures. Therefore, we feel that a
knowledge of enzymes active or becoming active in
crude drugs is an important issue on our way to well-
defined processes for the production of phytopharma-
ceuticals.
The standard assay for testing FHG I contained 1
mM substrate and about 15 nkat enzyme in a total
volume of 105 ml buffer. 1.5-ml Eppendorf tubes con-
ꢁ
taining this mixture were incubated for 15 min at 30 C,
and the enzyme reaction was terminated by the addition
of 315 ml of methanol. After centrifugation the mixture
was subjected to HPLC. Specific activities of FHG I
were calculated from the amount of flavonol released
from the corresponding flavonol-glycoside used as sub-
strate.
4
. Experimental
Protein was determined according to Bradford (1976)
with bovine serum albumin as the standard protein for
establishing calibration curves.
4
.1. Materials
Some rare flavonol glycosides, such as isorhamnetin
-O-rutinoside 4, kaempferol 3-O-rutinoside 5, kaemp-
4.4. Purification of FHG I
3
TM
A High Load system (Pharmacia) and a FPLC
TM
ferol 3-O-glucoside-7-O-rhamnoside 8, kaempferol 3-
sophoroside-7-O-glucoside and quercetin 3-O-
sophoroside were generous gifts from Dr. M. Veit,
Universitat Wurzburg, Germany. Other glycosides and
6
system (Pharmacia) were utilized for the chromato-
graphic purification steps. All operations were carried
ꢁ
¨
¨
out at 4 C. Acetone powders prepared from buckwheat
reference compounds used in the present study were
purchased from Carl Roth GmbH & Co KG (Karls-
ruhe, Germany) or Boehringer Mannheim GmbH
herb were extracted with 10 volumes (w/v) of McIlvaine
buffer, pH 6.0 and homogenized using an Ultra-Turrax.
After centrifugation at 15,000 g for 20 min, the super-

A. Baumgertel et al. / Phytochemistry 64 (2003) 411–418
417
natant was removed and filtered through MiraclothR
temperature optimum and the energy of activation were
determined at pH 5.0; otherwise the above conditions
(
Calbiochem). Proteins were precipitated by the addi-
ꢁ
tion of three volumes (w/v) cold acetone (ꢀ18 C) and
pelleted at 3000 g for 10 min. Precipitated protein was
dissolved in buffer A (0.1 M potassium phosphate buffer
pH 6.4+1.7 M (NH ) SO ). This solution was loaded
were applied. The K and V
m
values of the FHG I for
the different substrates were determined using the
hyperbolic regression method in the presence of about 1
max
ꢀ1
mg protein mL and an incubation time of 15 min.
Relative reaction velocities were determined with 1 mM
of the respective substrate.
4
2
4
on a Phenylsepharose 6 FF column (Amersham Bios-
ciences) equilibrated with buffer A. The column was
washed with the same buffer until the eluate was pro-
tein-free. Proteins were eluted with a linear gradient
from 1.7 to 0 M (NH ) SO at a flow rate of 1.33 ml
The carbohydrate content of FHG I was determined
after deglycosylation of the purified protein using the
Glycofree Deglycosylation Kit K-500 (Oxford Glyco-
Systems) according to the manufacturer’s instruction.
The sequencing of oligopeptides obtained after LysC
digestion was carried out by Toplab Martinsried, Ger-
many.
4
2
4
min 1. The eluate was collected in 3-ml fractions that
were then assayed for FHG I using rutin 1 as the sub-
strate. The most active fractions were combined and
dialysed against double-distilled water overnight.
ꢀ
The dialysed enzyme solution was adjusted to pH 6.5
and then loaded onto a Resource Q ion exchange col-
umn (Amersham Biosciences) previously equilibrated
with buffer B (0.02 M sodium acetate buffer, pH 6.5).
Proteins were eluted with a linear gradient from 0 to 0.7
4.7. Polyclonal antibodies
Polyclonal antibodies against purified FHG I were
raised in white mice. Two weeks after the collection of
pre-immune sera, three mice were immunized at differ-
ent subcutaneous sites with FHG I homogenates con-
taining 80 mg protein and 0.9% (w/v) NaCl. Three
subsequent boost immunizations were administered
after 16 (62 mg), 49 (10 mg) and 107 (50 mg). Antisera
were collected 10 days after the final boost. IgG-enri-
ched antisera were obtained from whole blood by cen-
trifugation at 1000 g for 10 min and stored in small
ꢀ
1
M NaCl at a flow rate of 1.33 ml min and collected in
-ml fractions. The most active fractions were combined
3
and concentrated by ultra-filtration using 30-kDa Mac-
rosep membrane filters (Filtron).
The FHG I was further purified by size exclusion
chromatography on Superdex 200 HR 10/30 (Amer-
sham Biosciences). The column was equilibrated with
McIlvaine buffer, pH 4.8. The proteins were eluted at a
flow rate of 0.5 ml min 1, and fractions of 1 ml each
were collected. After having checked their activity and
purity, fractions containing FHG I were pooled, con-
centrated, and used for the studies reported here.
ꢀ
ꢁ
aliquots at ꢀ20 C.
Antisera titers were determined by ELISA. Antisera
specifities were analyzed by immunoblotting techniques
(Western Blot). Crude protein extracts from buckwheat
herb as well as homogeneous FHG I were subjected to
SDS-PAGE. Proteins were electroblotted onto nitro-
cellulose membranes and probed using sera containing
polyclonal antibodies raised against FHG I (see above).
4
.5. Gel electrophoresis
Discontinuous SDS-PAGEwas performed according
to Laemmli (1970). Gels (10% polyacrylamide) were
developed for 15 min at 80 V, then for 75 min at 120 V.
Molecular weight standards were obtained from Biorad
Acknowledgements
(
dards and protein samples were heated in sample buffer
at 100 C for 3 min prior to SDS-PAGE. Gels were sil-
ver-stained (Silver Staining Kit, Amersham Biosciences)
according to the instructions of the manufacturer.
Low and High Molecular Weight Marker Kit). Stan-
This work was financially supported by Fink (Glaxo
SmithKline), Herrenberg. The authors wish to express
their gratitude to PD Dr M. Veit, Universitat Wurz-
¨
¨
burg, for the gift of special flavonol glycosides and to
Dr Ch. Theurer at Fink (Glaxo SmithKline) for initiat-
ing this study and following it with growing interest.
ꢁ
4
.6. Enzyme properties
The isoelectric point of the purified FHG I was
determined on PhastGel IEF 3-9 using the PhastSys-
temTM (Amersham Biosciences). Proteins were silver-
stained according to the instructions of the manu-
facturer. Proteins of the Amersham Biosciences pI
Marker Kit served as the pI standards. The pH-opti-
mum was determined in McIlvaine buffer in the pre-
sence of about 100 mg protein per ml and 1 mM rutin as
the substrate. The incubation time was 15 min. The
References
¨
Barz, W., Koster, J., 1981. Turnover and degradation of secondary
natural) products. In: Stumpf, P.K., Conn, E.E. (Eds.), The Bio-
(
chemistry of Plants. Secondary Plant Products, Academic Press,
New York, London, Toronto, Sydney, San Francisco, pp. 35–80.
Blakeney, A.B., Harris, P.J., Henry, R.J., Stone, B.A., 1992. A simple
and rapid preparation of alditol acetates for monosaccharide ana-
lysis. Carbohydrate Research 113, 291–299.

418
A. Baumgertel et al. / Phytochemistry 64 (2003) 411–418
Bourbouze, R., Percheron, F., Courtois, J.E., 1976. a-l-Rhamnosi-
dase de Fagopyrum esculentum. Eur. J. Biochem. 63, 331–337.
Bourbouze, R., Pratviel-Sosa, F., Percheron, F., 1974. Purification,
L.V., 1984. Rutin-induced b-glucosidase activity in Streptococcus
faecium VGH-1 and Streptococcus sp. Strain FRP-17 isolated from
human feces: formation of the mutagen, quercetin, from rutin.
Appl. Environ. Microbiol. 47, 350–355.
proprietes et specifite d’une heteroglycosidase des graines de sarra-
´
sin. Biochimie 56, 1305–1313.
Nguyen-Hiep, P., Delveau, G., Paris, R.R., 1964. Sur la dessiccation et
decinales. I.—Plantes a heteroside
´ ´ ´
Bradford, M.M., 1976. A rapid and sensitive method for the quanti-
tation of microgram quantities of protein utilizing the principle of
protein-dye binding. Analytical Biochemistry 72, 248–254.
Couch, J.F., Naghski, J., Krewson, C.F., 1946. Buckwheat as a source
of rutin. Science 103, 197–198.
la stabilisation des plantes me
´
flavoniques; Annales pharmaceutiques franc
¸
aise 22, 573–580.
Noguchi, I., Mori, S., 1969. Enzymatic degradation in Fagopyrum
vulgare leaves. Arch Biochem Biophy 132, 353–354.
Oxtoby, E., Dunn, M.A., Pancoro, A., Hughes, M.A., 1991. Nucleo-
tide and derived amino acid sequence of the cyanogenic b-glucosi-
dase (linamarase) from white clover (Trifolium repens L.). Plant
Mol. Biol. 17, 209–219.
Esen, A., 1993. b-Glycosidases: Biochemistry and Molecular Biology.
ACS Symposium Series. (ISSN 0097-6156; 533, 1–13).
Gabor, M., 1972. Pharmagologic effects of flavonoids on blood ves-
sels. Angiologica 9, 355–374.
Rice-Evans, C.A., Miller, N.J., Paganga, G., 1997. Antioxidant prop-
erties of phenolic compounds. Trends Plant Science 2, 152–159.
Schliemann, W., 1987. b-D-Glucosidasen von Pflanzen und Pflan-
zlichen Zellkulturen. Die Pharmazie 42 (7), 225–239.
Hagels, H., Wagenbreth, D., Schilcher, H., 1995. Phenolic compounds
of buckwheat herb and influence of plant and agricultural factors
(
Fagopyrum esculentum Moench and Fagopyrum tataricum (L.)
Gaertn.). In: Matano, T., Ujihara, A. (Eds.), Current Advances in
Buckwheat Research, Vol II. Shinshu University Press, Matsumoto,
Japan, pp. 801–809.
¨
Surholt, E., Hosel, W., 1978. Screening for flavonol 3-glycoside spe-
cific b-glycosidases in plants using a spectrophotometric enzymatic
assay. Phytochemistry 17, 873–877.
Hay, G.W., Westlake, D.W.S., Simpson, F.J., 1961. Degradation of
rutin by Aspergillus flavus. Purification and characterization of
rutinase. Can. J. Microbiol. 7, 921–932.
Wojcicki, J., Barcew-Wisniewska, B., Samochowiec, L., Rozewicka,
L., 1995a. Extractum Fagopyri reduces atherosclerosis in high-fat
diet fed rabbits. Pharmazie 50, 560–562.
Ihme, N., Kiesewetter, H., Jung, F., Hoffmann, K.H., Birk, A., Mu
¨
l-
ler, A., Grutzner, K.I., 1996. Leg oedema protection from a buck-
Wojcicki, J., Samochowiec, L., Gonet, B., Juzwiak, S., Dabrowska-
Zamojcin, E., Katdonska, M., Tustanowski, S., 1995b. Effect of
buckwheat extract on free radical generation in rabbits administered
high-fat diet. Phytother. Res. 9, 323–326.
¨
wheat herb tea in patients with chronic venous insufficiency: a single
centre, randomised, double-blind, placebo-controlled clinical trial.
Eur. J. Clin. Pharmacol. 50, 443–447.
Yasuda, T., Nakagawa, H., 1994. Purification and characterization of
the rutin-degrading enzymes in tartary buckwheat seeds. Phy-
tochemistry 37, 133–136.
Kurusawa, Y., Ikeda, K., Egami, F., 1973. a-L-Rhamnosidases of the
liver of Turbo cornutus and Aspergillus niger. J. Biochem. 73, 31–
3
7.
Zheng, L., Poulton, J.E., 1995. Temporal and spatial expression of
amygdalin hydrolase and (R)-(+)-mandelonitrile lyase in black
cherry seeds. Plant Physiol. 109, 31–39.
Laemmli, U.K., 1970. Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature (London) 227,
6
80–685.
Zhou, J., Hartmann, S., Shepherd, B.K., Poulton, J.E., 2002. Investi-
gation of the microheterogeneity and aglycone specificity-conferring
residues of black cherry (Prunus serotina) prunasin hydrolases. Plant
Physiol. 129, 1252–1264.
Li, C.P., Swain, E., Poulton, J.E., 1992. Prunus serotina amygdalin
hydrolase and prunasin hydrolase. Plant Physiol. 100, 282–290.
MacDonald, I.A., Bussard, R.G., Hutchinson, D.M., Holdeman,