Regioselectivity of phase II metabolism of luteolin and quercetin by UDP-glucuronosyl transferases

Article abstract of DOI:10.1021/tx0101705

The regioselectivity of phase II conjugation of flavonoids is expected to be of importance for their biological activity. In the present study, the regioselectivity of phase II biotransformation of the model flavonoids luteolin and quercetin by UDP-glucuronosyltransferases was investigated. Identification of the metabolites formed in microsomal incubations with luteolin or quercetin was done using HPLC, LC-MS, and ¹H NMR. The results obtained demonstrate the major sites for glucuronidation to be the 7-, 3-, 3′-, or 4′-hydroxyl moiety. Using these unequivocal identifications, the regioselectivity of the glucuronidation of luteolin and quercetin by microsomal samples from different origin, i.e., rat and human intestine and liver, as well as by various individual human UDP-glucuronosyltransferase isoenzymes was characterized. The results obtained reveal that regioselectivity is dependent on the model flavonoid of interest, glucuronidation of luteolin and quercetin not following the same pattern, depending on the isoenzyme of UDP-glucuronosyltransferases (UGT) involved. Human UGTIA1, UGTIA8, and UGTIA9 were shown to be especially active in conjugation of both flavonoids, whereas UGTIA4 and UGTIA10 and the isoenzymes from the UGTB family, UGT2B7 and UGT2B15, were less efficient. Due to the different regioselectivity and activity displayed by the various UDP- glucuronosyltransferases, regioselectivity and rate of flavonoid conjugation varies with species and organ. Qualitative comparison of the regioselectivities of glucuronidation obtained with human intestine and liver microsomes to those obtained with human UGT isoenzymes indicates that, in human liver, especially UGT1A9 and, in intestine, UGT1A1 and UGT1A8 are involved in glucuronidation of quercetin and luteolin. Taking into account the fact that the anti-oxidant action as well as the pro-oxidant toxicity of these catechol-type flavonoids is especially related to their 3′,4′-dihydroxyl moiety, it is of interest to note that the human intestine UGT's appear to be especially effective in conjugating this 3′,4′ catechol unit. This would imply that upon glucuronidation along the transport across the intestinal border, the flavonoids loose a significant part of these biological activities.

Full text of DOI:10.1021/tx0101705

662
Chem. Res. Toxicol. 2002, 15, 662-670
Regioselectivity of P h a se II Meta bolism of Lu teolin a n d
Qu er cetin by UDP -Glu cu r on osyl Tr a n sfer a ses
Marelle G. Boersma,†,‡ Hester van der Woude,† J an Bogaards,§ Sjef Boeren,‡
J acques Vervoort,‡ Nicole H. P. Cnubben,§,| Marlou L. P. S. van Iersel,
Peter J . van Bladeren,†,§,| and Ivonne M. C. M. Rietjens*,†,|
Division of Toxicology, Wageningen University, Tuinlaan 5, 6703 HE, Wageningen,
The Netherlands, Laboratory of Biochemistry, Wageningen University, Dreijenlaan 3,
6703 HA, Wageningen, The Netherlands, TNO Nutrition and Food Research, P.O. Box 360,
3700 AJ , Zeist, The Netherlands, WU/ TNO Centre for Food Toxicology, P.O. Box 8000,
6700 EA, Wageningen, The Netherlands, and N.V. Organon, P.O. Box 20,
5340 BH, Oss, The Netherlands
Received November 2, 2001
The regioselectivity of phase II conjugation of flavonoids is expected to be of importance for
their biological activity. In the present study, the regioselectivity of phase II biotransformation
of the model flavonoids luteolin and quercetin by UDP-glucuronosyltransferases was investi-
gated. Identification of the metabolites formed in microsomal incubations with luteolin or
1
quercetin was done using HPLC, LC-MS, and H NMR. The results obtained demonstrate the
major sites for glucuronidation to be the 7-, 3-, 3′-, or 4′-hydroxyl moiety. Using these
unequivocal identifications, the regioselectivity of the glucuronidation of luteolin and quercetin
by microsomal samples from different origin, i.e., rat and human intestine and liver, as well
as by various individual human UDP-glucuronosyltransferase isoenzymes was characterized.
The results obtained reveal that regioselectivity is dependent on the model flavonoid of interest,
glucuronidation of luteolin and quercetin not following the same pattern, depending on the
isoenzyme of UDP-glucuronosyltransferases (UGT) involved. Human UGT1A1, UGT1A8, and
UGT1A9 were shown to be especially active in conjugation of both flavonoids, whereas UGT1A4
and UGT1A10 and the isoenzymes from the UGTB family, UGT2B7 and UGT2B15, were less
efficient. Due to the different regioselectivity and activity displayed by the various UDP-
glucuronosyltransferases, regioselectivity and rate of flavonoid conjugation varies with species
and organ. Qualitative comparison of the regioselectivities of glucuronidation obtained with
human intestine and liver microsomes to those obtained with human UGT isoenzymes indicates
that, in human liver, especially UGT1A9 and, in intestine, UGT1A1 and UGT1A8 are involved
in glucuronidation of quercetin and luteolin. Taking into account the fact that the anti-oxidant
action as well as the pro-oxidant toxicity of these catechol-type flavonoids is especially related
to their 3′,4′-dihydroxyl moiety, it is of interest to note that the human intestine UGT’s appear
to be especially effective in conjugating this 3′,4′ catechol unit. This would imply that upon
glucuronidation along the transport across the intestinal border, the flavonoids loose a
significant part of these biological activities.
In tr od u ction
Luteolin and quercetin (Figure 1) are important mem-
bers of the class of flavonoid food components. These
flavonoids are polyphenols, which are important con-
stituents of fruit, vegetables, red wine, and tea. The
antioxidant properties of flavonoids are often claimed to
be responsible for the protective effects of these com-
pounds against cardiovascular disease, certain forms of
cancer, and/or photosensitivity diseases (1-4). Because
of these beneficial health claims, the interest for the use
of flavonoids such as quercetin and luteolin as functional
food ingredients is increasing rapidly. However, other
F igu r e 1. Chemical structures of luteolin and quercetin.
research data point at possible toxic pro-oxidant action
of flavonoids. This includes the formation of reactive
mutagenic quinoid-type metabolites from especially 3′,4′-
dihydroxy () catechol)-type polyphenol antioxidants such
as the flavonoid quercetin (1, 5-8). The anti- as well as
the pro-oxidant effects of flavonoids have been reported
to be highly dependent on the number and position of
the hydroxyl moieties in the flavonoid molecule. To
achieve an efficient antioxidant action, several studies
* To whom correspondence should be addressed. Fax: (31) (317)
484801. E-mail: ivonne.rietjens@algemeen.tox.wau.nl.
† Division of Toxicology, Wageningen University.
‡ Laboratory of Biochemistry, Wageningen University.
§ TNO Nutrition and Food Research.
| TNO Centre for Food Toxicology.
N.V. Organon.
10.1021/tx0101705 CCC: $22.00 © 2002 American Chemical Society
Published on Web 04/19/2002

Regioselectivity of Phase II Metabolism
Chem. Res. Toxicol., Vol. 15, No. 5, 2002 663
have reported the importance of a 3′,4′-dihydroxy moiety
in the B-ring, the 2,3-double bond in combination with a
4-oxo function in the C ring and the additional presence
of 3- and 5-hydroxyl groups in the flavonoid molecular
structure (2, 9). Data on pro-oxidant activities indicate
similar structural features to be essential for mutagenic
and other pro-oxidant activities (5, 8, 10, 11).
Exp er im en ta l P r oced u r es
Ma ter ia ls. Quercetin was obtained from Acros Organics
(New J ersey). Isoquercetrin, luteolin, luteolin-7-glucoside, and
luteolin-4′-glucoside were purchased from Indofine (Somerville,
PA). All flavonoid substrates were prepared fresh as a stock
solution in DMSO or DMSO-d6 in the case of 1H NMR measure-
ments. 4-Trifluoromethyl-7-hydroxycoumarin was purchased
from Enzyme Systems Products (Dublin, CA). Alamethicin,
uridine 5′-diphosphoglucuronic acid (UDPGA), and trifluopera-
zine were obtained from Sigma (St. Louis, MO). â-Glucuronidase
was from Boehringer (Mannheim, Germany). DMSO-d6 and D2O
were obtained from ARC Laboratories (Amsterdam, The Neth-
erlands). Trifluoroacetic acid was obtained from Merck (Darm-
stadt, Germany). Acetonitril was HPLC grade and purchased
from Lab-Scan Ltd. (Dublin, Ireland).
Rat liver microsomes were prepared from perfused livers from
Wistar rats as described before (30) and human liver microsomes
were prepared from human liver samples as described previ-
ously (31). Human intestine microsomes (pooled from 10 healthy
individuals) were obtained from In Vitro Technologies (Balti-
more, MD) and microsomes from rat intestine from Biopredic
International (Rennes, France). Microsomes (UGT1A6 and
UGT1A9) and Supersomes (UGT1A1, UGT1A3, UGT1A4,
UGT1A8, UGT1A10, UGT2B7, and UGT2B15) from cell lines
transfected with cDNAs expressing human UGTs were obtained
from GENTEST Corp. (Woburn, MA).
Since the hydroxyl substituent pattern significantly
affects the anti- as well as the pro-oxidant biological
activity of the flavonoids, the regioselectivity of their
conjugation can be expected to influence these biological
activities as well. Naturally occurring flavonoids are often
glycosylated or glucuronidated, but may become decon-
jugated by bacterial activity in the gut and/or during
transport over the gut wall (12). Upon and/or during
absorption, biotransformation of the flavonoids (meth-
ylation, glucuronidation, and sulfation) may result in
conjugated metabolites with different properties, depend-
ing on the position at which conjugation occurs. Cao et
al. (13) concluded that the more hydroxyl substituents,
the stronger the antioxidant but also the pro-oxidant
activity of flavonoids and that O-methylation of the
hydroxyl groups inactivates both these activities of
flavonoids. Glucuronidation of the 3′- or 4′-hydroxyl group
in the B-ring of flavonoids would increase the reduction
potential and therefore decrease the antioxidant activity,
whereas glucuronidation in the A-ring would affect
antioxidant activity to a lesser extent (14). With respect
to the influence of conjugation on toxic effects, it has been
reported that O-methylation of the 3′,4′-catechol moiety
results in detoxification of quercetin and fisetin (15).
Although the metabolism of flavonoids is a growing field
of interest (15-21), much about the uptake and metabo-
lism of dietary flavonoids remains unknown (22). This
holds especially for knowledge on the regioselectivity of
flavonoid conjugation.
Liver a n d In testin e Micr osom a l In cu ba tion s. Incubation
mixtures contained 0.1 M potassium phosphate, pH 7.4, 0.5 or
2 mg/mL microsomal protein (as indicated), 100 µM luteolin or
quercetin, added as a 100 times concentrated stock solution in
DMSO. The reaction was started by adding 4 mM (final
concentration) UDPGA and incubated at 37 °C for 15 min.
Samples were either frozen in liquid nitrogen before analysis
or analyzed directly by HPLC.
Quantification of the glucuronosyl metabolites formed by
microsomal incubations was done on the basis of the peak area
in the HPLC chromatogram at 345 nm for luteolin and 370 nm
for quercetin, assuming similar ꢀ value of flavonoids and their
glucuronosyl conjugates (22). The UV-vis absorption spectra
reveal this to be a reasonable assumption because the effects
of conjugation on the UV-vis spectra are not significant.
In cu ba tion s w ith In d ivid u a l UGT-Isoen zym es. Incuba-
tion mixtures contained 0.1 M Tris-HCl pH 7.5, 1 mM DTT, 10
mM MgCl2, 25 µg/mL alamethicin, 0.5 mg/mL microsomal
protein, 100 µM luteolin or quercetin, added as a 100 times
concentrated stock solution in DMSO and 1 mM UDPGA and
were incubated for 30 min at 37 °C. The reaction was stopped
by adding 100 µL of acetonitril:acetic acid (94:6) (v/v) to 200 µL
incubation mixture. After centrifugation (5 min at 2750g), the
samples were stored at - 20 °C until analyzed by HPLC.
Activities of the commercially obtained UGT1A1, UGT1A3,
UGT1A6, UGT1A8, UGT1A9, UGT1A10, UGT2B7, and
UGT2B15 isoenzyme preparations were checked on the basis
of activity toward the standard substrate 4-trifluoromethyl-7-
hydroxycoumarin and were as follows: 795, 730, 7430, 350,
2300, 90, 1510, and 1860 pmol/min/mg of protein, respectively.
The enzyme activity of UGT1A4 toward trifluoperazine was
1070 pmol/min/mg of protein. These assays were described by
the supplier using standard methods (GENTEST Corp.) and the
activities were in line with literature data and validate the
quality of the UGT enzyme preparations used.
En zym a tic Decon ju ga tion . To confirm the nature of the
glucuronosyl conjugates, medium samples were treated with
â-glucuronidase. To this end, 10 µL of sample was added to 90
µL of 200 mM potassium phosphate, pH 6.2 and after addition
of 3 µL of â-glucuronidase solution (0.8 units) the mixture was
incubated for 1 h at 37 °C. After the incubation, the mixtures
were either frozen in liquid nitrogen before analysis or analyzed
directly by HPLC.
HP LC An a lysis. HPLC was performed with a Waters M600
liquid chromatography system. Reaction product analysis was
achieved using an Alltima C18 5U column (4.6 × 150 mm;
Therefore, the objective of the present study was to
identify the major sites for glucuronidation of two model
flavonoids, luteolin and quercetin (Figure 1). Luteolin is
a member of the flavone subclass of flavonoids and
usually occurs as a glycosylated form in celery, perilla
leaves, green pepper, and camomile tea and as aglycone
in perilla seeds and in parsley (23). Recently, several
groups have reported intestinal absorption and metabo-
lism of luteolin in rats and humans (12, 20, 24), but these
studies did not identify the exact nature and regioselec-
tivity of the metabolites formed. Quercetin glycosides
represent the predominant flavonoid fraction present in
fruits and vegetables, especially onions, broccoli, and
apples (22, 25) as well as in tea and red wine. Several
studies report glucuronidated, sulfated, and/or meth-
ylated quercetin metabolites in plasma of man and rats
upon quercetin exposure (18, 19, 21, 26-28), but again
the exact regioselectivity of the conjugate formation was
not elucidated. Recently Day et al. (29) reported HPLC
UV-vis based identification of quercetin glucuronosyls
formed by human liver cell free extracts showing conju-
gation at 4′-, 3′-, 7-, and 3-OH, the highest rate of
conversion being observed for the 3′-position. In the
present study, the regioselectivity of quercetin but also
luteolin conjugation was investigated using LC-MS and
1
especially H NMR as the method of choice for identifica-
tion of the conjugates. In addition, the present study
reports on the regioselectivity of glucuronidation of the
two model flavonoids by individual UDP-glucuronosyl-
transferase isoenzymes.

664 Chem. Res. Toxicol., Vol. 15, No. 5, 2002
Boersma et al.
Alltech, Breda, The Netherlands), eluted at a flow of 1 mL/min
with a gradient starting at 17% acetonitril in Nanopure water
containing 0.1% trifluoroacetic acid for 2 min, followed by a
gradient of 17% to 25% acetonitril in 5 min, increasing to 35%
in 8 min, to 50% in 5 min and increasing to 100% acetonitril in
5 min, finally keeping this percentage for 2 min, followed by a
reequilibration at 17% acetonitril in Nanopure water containing
0.1% trifluoroacetic acid for 10 min.
In a typical run, aliquots of 10 µL of the incubation mixture
were injected. Detection was performed between 230 and 420
nm using a Waters 996 photodiode array detector. Chromato-
grams presented are based on detection at 345 nm for luteolin
and 370 nm for quercetin incubations.
To collect the different metabolites for product identification,
similar HPLC runs were performed except for the fact that
luteolin or quercetin were added at least three times to the
incubation mixtures followed by 10 min incubation at 37 °C,
before injecting 50 µL of the incubation mixture onto the HPLC
column. Product peaks of successive runs were collected, pooled,
freeze-dried and dissolved in DMSO-d6 for 1H NMR analysis.
LC-MS An a lysis. Incubation samples were analyzed by LC-
MS. Mass spectrometric analysis (Finnigan MAT 95, San J ose,
CA) was performed in the negative electrospray mode using a
spray voltage of 4.5 kV and a capillary temperature of 180 °C
with nitrogen as sheath and auxiliary gas. For LC, the same
conditions as described above were used, except that in eluent
A 0.1% trifluoroacetic acid was replaced by 0.1% acetic acid,
which had no effect on retention times.
1H NMR An a lysis. The glucuronosyl conjugates were char-
acterized by 1H NMR using a Bruker DPX 400 MHz or a Bruker
AMX 500 MHz spectrometer. A 1.5 s presaturation delay was
used, a 60° pulse angle and a 2.2 s acquisition time (7575 Hz
sweep width, 32 K data points). The data were processed using
an exponential multiplication of 0.5 or 1.0 Hz and zero filling
to 64 K data points. Resonances are reported relative to DMSO
at 2.50 ppm.
Resu lts
Id en tifica tion of Lu teolin Glu cu r on osyls. Figure
2a shows a typical HPLC chromatogram of an incubation
of 100 µM luteolin with rat liver microsomes. The
chromatogram shows that with rat liver microsomes,
upon 15 min of incubation with UDPGA, three metabo-
lites, with retention times at 11.4 ( 0.2, 13.9 ( 0.2, and
14.5 ( 0.2 min, are formed. HPLC analysis of the
incubations of luteolin with human liver microsomes
(Figure 2c) reveals formation of the same three metabo-
lites only at a different ratio (Table 1). Figure 2, panels
b and d, presents the HPLC chromatograms of similar
incubations using rat and human intestine microsomes.
The same three metabolites are formed at again a
different ratio (Table 1). Comparison of these ratios
reveals that regioselectivity of the glucuronidation varies
between organs as well as between species (Table 1).
Increasing the concentration of microsomal protein in the
incubations from 0.5 to 2 mg/mL increased the overall
conversion upon 15 min incubation 2.5 times but did not
affect the regioselectivity, i.e., the ratio between the
different conjugates formed.
F igu r e 2. HPLC chromatogram of incubations of luteolin with
(a) rat liver microsomes, (b) rat intestine microsomes, (c) human
liver microsomes, and (d) human intestine microsomes, (2 mg/
mL final concentration microsomal protein) after 15 min
incubation at 37 °C and based on detection at 345 nm.
the three metabolites a m/z value of 461 representing the
[M - H]-mass of a luteolin monoglucuronosyl.
To identify the exact position of the glucuronosyl
1
moiety in the luteolin conjugates, H NMR analysis was
1
performed. Table 2 presents the H NMR chemical shift
values and coupling constants of the reference compounds
luteolin, luteolin-7-O-glucoside and luteolin-4′-O-gluco-
side. Identification of the various 1H NMR resonances
has been achieved on the basis of comparison to literature
data (32) and of their splitting patterns. Table 2 also
summarizes the 1H NMR chemical shift values and
coupling constants of the three metabolites with retention
times at 11.4, 13.9, and 14.5 min. Comparison of the
chemical shift values of the aromatic protons of the
glucuronosyl adduct with retention time 11.4 min to the
1H NMR resonances of luteolin reveals a change of +0.24
ppm and +0.35 ppm, respectively, in the chemical shift
value of the NMR resonances of especially the C6-H and
Incubation of the samples with â-glucuronidase results
in complete elimination of the three metabolite peaks and
a concomitant increase of the luteolin peak (HPLC
chromatograms not shown), indicating that all metabolite
peaks represent glucuronidated luteolin derivatives.
4
the C8-H with their J HH coupling of 2.1 Hz (Table 2).
1
In contrast, the H NMR signals of C2′-H, C5′-H, C6′-H,
1
To identify the exact nature of the glucuronosyl con-
jugates, samples from the microsomal incubations were
also analyzed by LC-MS. The results obtained reveal for
and C3-H are unaffected in the H NMR spectrum of
this glucuronosyl metabolite as compared to luteolin. In
1
addition the H NMR characteristics of this metabolite

Regioselectivity of Phase II Metabolism
Chem. Res. Toxicol., Vol. 15, No. 5, 2002 665
Ta ble 1. P er cen ta ges of Meta bolites F or m ed fr om Lu teolin a n d Qu er cetin w ith Micr osom es fr om Differ en t Sou r ces
Ba sed on P ea k Ar ea Mea su r ed by HP LC a t 345 n m for Lu teolin a n d 370 n m for Qu er cetin a
Luteolin
tR 11.4 min
tR 13.9 min
tR 14.5 min
substrate
microsomes
rat liver
rat intestine
human liver
7-O-Gluc (%)
-
4′-O-Gluc (%)
3′-O-Gluc (%)
conversion (%)
28
32
77
5
-
11
3
6
61
65
17
51
21
27
27
74
-
-
-
human intestine
44
Quercetin
tR 10.5 min
tR 10.7 min
tR 13.7 min
tR 14.2 min
substrate
microsomes
7-O-Gluc (%)
3-O-Gluc (%)
4′-O-Gluc (%)
3′-O-Gluc (%)
conversion (%)
rat liver
40
41
65
4
0
9
13
20
18
4
3
42
46
19
49
18
25
22
61
rat intestine
human liver
human intestine
27
a
The final microsomal protein concentration was 0.5 mg/mL.
Ta ble 2. 1H NMR Ch em ica l Sh ift Va lu es a n d Cou p lin g Con sta n ts of Lu teolin a n d Som e Con ju ga tes Mea su r ed
a
in DMSO-d 6
C6-H
C8-H
C3-H
6.63
C5′-H
C2′-H
C6′-H
luteolin
6.14
6.40
6.84
7.35
7.37
J H6-H8 ) 2.1 Hz J H8-H6 ) 2.1 Hz
J H5′-H6′ ) 8.3 Hz J H2′-H6′ ) 2.1 Hz J H6′-H2′ ) 2.1 Hz
J H6′-H5′ ) 8.3 Hz
luteolin-7-O-glucoside
6.44 (+0.30)
6.79 (+0.39)
6.76 (+0.13) 6.90 (+0.06)
7.42 (+0.07)
7.45 (+0.08)
J H6-H8 ) 2.1 Hz J H8-H6 ) 2.1 Hz
J H5′-H6′ ) 8.4 Hz J H2′-H6′ ) 2.1 Hz J H6′-H2′ ) 2.1 Hz
J H6′-H5′ ) 8.4 Hz
luteolin-4′-O-glucoside 6.19 (+0.05) 6.49 (+0.09)
6.83 (+0.20) 7.23 (+0.39)
7.50 (+0.15)
7.52 (+0.15)
J H6-H8 ) 2.1 Hz J H8-H6 ) 2.1 Hz
6.38 (+0.24) 6.75 (+0.35)
J H5′-H6′ ) 8.5 Hz J H2′-H6′ ) 2.1 Hz J H6′-H2′ ) 2.1 Hz
J H6′-H5′ ) 8.5 Hz
metabolite
tR 11.4 min
6.72 (+0.09) 6.85 (+0.01)
7.39 (+0.04)
7.41 (+0.04)
J H6-H8 ) 2.1 Hz J H8-H6 ) 2.1 Hz
J H5′-H6′ ) 8.4 Hz J H2′-H6′ ) 2.1 Hz J H6′-H2′ ) 2.1 Hz
J H6′-H5′ ) 8.4 Hz
metabolite
tR 13.9 min
6.15 (+0.01) 6.45 (+0.05)
6.78 (+0.15) 7.21 (+0.37)
7.45 (+0.10)
7.49 (+0.12)
J H6-H8 ) 1.8 Hz J H8-H6 ) 1.8 Hz
6.17 (+0.03) 6.81 (+0.41)
J H5′-H6′ ) 8.9 Hz J H2′-H6′ ) 1.8 Hz J H6′-H2′ ) 1.8 Hz
J H6′-H5′ ) 8.9 Hz
metabolite
tR 14.5 min
6.55 (-0.08) 6.96 (+0.12)
7.64 (+0.29)
7.80 (+0.43)
a
The ppm shift of the protons in the conjugates compared to the same protons in luteolin are put into brackets. The ppm values of the
two reference compounds, luteolin-7-glucoside and luteolin-4′-glucoside, were assigned based on comparison to literature data (32). 1H
NMR identification of the metabolite of luteolin with retention time of 14.5 min was somewhat hampered by its instability, preventing
identification of all splitting patterns although major resonance chemical shift values could still be observed.
resemble those of the reference compound luteolin-7-
glucoside. All together these H NMR data identify the
glucuronosyl of luteolin, this metabolite can be either the
5-O- or the 3′-O-glucuronosyl of luteolin. On the basis of
the presence of the resonance of the 5-OH group in the
1H NMR spectrum of this metabolite at 12.95 ppm (data
not shown) being a 1H NMR visible OH resonance
because of lack of rapid exchange with solvent due to
strong hydrogen bonding of the OH-hydrogen to the
4-keto moiety, the metabolite cannot be the 5-O-glucu-
ronosyl of luteolin. This indicates that the metabolite can
be identified as the 3′-O-glucuronosyl of luteolin. This
conclusion is corroborated by the fact that the chemical
shift changes observed for the aromatic proton signals
of this metabolite as compared to luteolin itself reveal
1
microsomal luteolin monoglucuronosyl metabolite with
retention time of 11.4 min as the 7-O-glucuronosyl of
luteolin.
Comparison of the 1H NMR data obtained for the
metabolite with retention time of 13.9 min on HPLC with
1
the H NMR spectrum of luteolin reveals that the C5′-
3
H, represented by a doublet at 6.84 ppm with a J HH
coupling of 8.3 Hz, shifts by +0.37 ppm toward 7.21 ppm
for this monoglucuronidated luteolin metabolite (Table
1
2). The H NMR signal of C2′-H, shifts by +0.10 ppm,
and that of C6′-H by +0.12 ppm. The 1H NMR resonance
of C6-H and C8-H are almost similar to those of C6-H
and C8-H of luteolin. Together with the LC-MS results,
this reveals that the metabolite is a monoglucuronosyl
intermediate which can be identified as luteolin-4′-O-
glucuronosyl.
1
significant changes in especially the H NMR signal of
C2′-H (+0.29 ppm) and C6′-H (+0.43 ppm), whereas
the chemical shift value of C6-H and C3-H remain
unaffected. This is in line with the identification of the
metabolite as the 3′-O-glucuronosyl of luteolin. The
unexpected relatively large shift in the ppm value of
C8-H (+0.41 ppm) must then be ascribed to an intramo-
lecular interaction between this C8-H and the glucu-
ronosyl moiety attached at C3′-OH.
Id en tifica tion of Qu er cetin Glu cu r on osyls. Simi-
lar experiments were conducted with quercetin as the
model compound. Figure 3, panels a and c, shows the
HPLC chromatograms of incubations of rat and human
1H NMR identification of the third mono-glucuronosyl
metabolite of luteolin with retention time of 14.5 min was
somewhat hampered by its instability preventing iden-
tification of all splitting patterns although major reso-
nance chemical shift values could still be observed. On
the basis of LC-MS analysis, the metabolite is a mono-
glucuronosyl of luteolin. Together with the identification
of the other two metabolites as the 4′-O- and the 7-O-

666 Chem. Res. Toxicol., Vol. 15, No. 5, 2002
Boersma et al.
tions were also analyzed by LC-MS. The results obtained
reveal for the four metabolites a m/z value of 477
representing the [M - H]-mass of a quercetin monoglu-
curonosyl.
Table 3 summarizes the 1H NMR chemical shift values
and coupling constants of three quercetin monoglucu-
ronosyls eluting from HPLC with retention times 10.5,
13.7, and 14.2 min as well as of quercetin and the
reference compounds quercetin-3-O-glucoside (isoquer-
cetrin) and quercetin-4′-O-glucoside (spiraeoside).
1
Identification of the various H NMR resonances has
been achieved on the basis of chemical shift values and
their splitting patterns and for the quercetin monoglu-
curonosyl metabolites also on the basis of comparison
with the data of the reference compounds. Comparison
of the resonances of the aromatic protons of the metabo-
lite eluting from the HPLC at retention time 10.5 min to
the resonances of quercetin shows a shift of +0.22 ppm
for the C6-H signal and an even larger shift of +0.35
ppm for the C8-H chemical shift value. In contrast, the
chemical shift values of the protons on the B-ring are
almost unaffected (Table 3). This result identifies the
major metabolite of the quercetin incubations with liver
microsomes, with a retention time of 10.5 min, as the
7-O-glucuronosyl of quercetin.
1
The H NMR resonances of the metabolite with reten-
tion time 13.7 min on HPLC are similar to those
previously published for spiraeoside (4′-O-glucoside of
quercetin) (32) and show a shift of +0.37 ppm for the
1
C5′-H H NMR signal as the only difference compared
1
to the H NMR data of quercetin. Together with the LC-
MS results, these NMR data identify this metabolite as
quercetin 4′-O-glucuronosyl.
Comparison of the 1H NMR data obtained for the
metabolite with a retention time of 14.2 min on HPLC
with the 1H NMR resonances of quercetin reveals a
change of +0.37 and +0.31 ppm, respectively, in the
chemical shift value of the NMR resonances of the C6′-H
and C2′-H (Table 3), while the chemical shift value of
the C5′-H shows only a small shift of +0.09 ppm. This
metabolite can be ascribed to be the 3′-O-glucuronosyl
of quercetin. This is further supported by the similarity
of its UV-absorption spectrum to that of the metabolite
identified as the 4′-O-glucuronosyl of quercetin, the fact
that its retention time on HPLC is slightly larger than
that of the 4′-O-glucuronosyl conjugate, an elution order
as observed for the 3′-O- and the 4′-O-glucuronosyl
conjugate of luteolin (Figure 2), and by the LC-MS results
which reveal this metabolite to be a monoglucuronidated
intermediate. These results identify this metabolite as
quercetin-3′-O-glucuronosyl.
The metabolite with a retention time of 10.7 min on
HPLC is present in the quercetin incubation with human
intestine microsomes (Figure 3d), but is only a minor
product in the rat intestine and human liver incubations
(Figure 3, panels b and c) or is even absent in the rat
liver incubations (Figure 3a). The chemical instability of
this metabolite during freeze-drying and subsequent
NMR measurement overnight hampered its further
identification by 1H NMR (33). Since the LC-MS data
indicated it to be a monoglucuronosyl of quercetin, and
the 7-O-, 3′-O-, and 4′-O-monoglucuronosyls were already
identified, the one residual monoglucuronosyl can be the
3-O- or the 5-O-glucuronosyl of quercetin. It has been
published (34) that the 5-position only reacts if this is
the only hydroxyl available. Due to the strong hydrogen
F igu r e 3. HPLC chromatogram of incubations of quercetin
with (a) rat liver microsomes, (b) rat intestine microsomes, (c)
human liver microsomes, and (d) human intestine microsomes
(2 mg/mL final concentration microsomal protein) after 15 min
incubation at 37 °C and based on detection at 370 nm.
liver microsomes with 100 µM quercetin in the presence
of UDPGA. Figure 3, panels b and d, shows the HPLC
chromatograms of similar incubations with rat and
human intestine microsomes. The chromatograms reveal
that, in contrast with luteolin, not three but four major
metabolites, with retention times at 10.5 ( 0.2, 10.7 (
0.2, 13.7 ( 0.2, and 14.2 ( 0.2 min, are formed in
different ratios. Table 1 summarizes the ratios obtained
in the different incubations. Comparison of these ratios
reveals that also for quercetin, the regioselectivity of
glucuronidation varies between organs and between
species (Table 1). Increasing the concentration of mi-
crosomal protein in the incubations from 0.5 to 2 mg/mL
increased the overall conversion upon 15 min incubation
2.5 times but did not affect the regioselectivity i.e., the
ratio between the different conjugates formed.
Incubation of the samples with â-glucuronidase results
in complete elimination of the metabolite peaks and a
concomitant increase of the quercetin peak (HPLC chro-
matograms not shown), indicating that, as expected, all
four metabolite peaks represent glucuronidated quercetin
derivatives. To identify the exact nature of the glucu-
ronosyl conjugates, samples from the microsomal incuba-

Regioselectivity of Phase II Metabolism
Chem. Res. Toxicol., Vol. 15, No. 5, 2002 667
Ta ble 3. 1H NMR Ch em ica l Sh ift Va lu es a n d Cou p lin g Con sta n ts of Qu er cetin a n d Som e Con ju ga tes Mea su r ed
a
in DMSO-d 6
C6-H
C8-H
C5′-H
C2′-H
C6′-H
quercetin
6.18
6.40
6.87
7.67
7.53
J H6-H8 ) 2.1 Hz
J H8-H6 ) 2.1 Hz
J H5′-H6′ ) 8.5 Hz
J H2′-H6′ ) 2.1 Hz
J H6′-H2′ ) 2.1 Hz
J H6′-H5′ ) 8.5 Hz
7.58 (+0.05)
isoquercetrin (3-O-GlcQ)
spiraeoside (4′-O-GlcQ)
6.20 (+0.02)
6.40 (+0)
6.84 (-0.03)
7.57 (-0.10)
J H6-H8 ) 2.1 Hz
6.22 (+0.04)
J H8-H6 ) 2.1 Hz
6.46 (+0.06)
J H8-H6 ) 2.0 Hz
J H5′-H6′ ) 8.9 Hz
7.26 (+0.39)
J H6′-H5′ ) 8.9 Hz
7.63 (+0.10)
7.71 (+0.04)
J H2′-H6′ ) 2.2 Hz
J H6-H8 ) 2.0 Hz
J H5′-H6′ ) 8.7 Hz
J H6′-H2′ ) 2.2 Hz
J H6′-H5′ ) 8.7 Hz
7.57 (+0.04)
J H6′-H2′ ) 1.9 Hz
J H6′-H5′ ) 8.4 Hz
7.60 (+0.07)
metabolite
tR 10.5 min
6.40 (+0.22)
J H6-H8 ) 1.9 Hz
6.75 (+0.35)
J H8-H6 ) 1.9 Hz
6.89 (+0.02)
J H5′-H6′ ) 8.4 Hz
7.70 (+0.03)
J H2′-H6′ ) 1.9 Hz
metabolite
tR 13.7 min
6.18 (+0)
J H6-H8 ) 1.9 Hz
6.45 (+0.05)
J H8-H6 ) 1.9 Hz
7.24 (+0.37)
J H5′-H6′ ) 8.7 Hz
7.70 (+0.03)
J H2′-H6′ ) 1.9 Hz
J H6′-H2′ ) 1.9 Hz
J H6′-H5′ ) 8.7 Hz
7.90 (+0.37)
metabolite
tR 14.2 min
6.16 (-0.02)
6.59 (+0.19)
6.96 (+0.09)
J H5′-H6′ ) 8.3 Hz
7.98 (+0.31)
J H6′-H5′ ) 8.3 Hz
a
The ppm shift of the protons in the conjugates compared to the same protons in quercetin are put into brackets. The ppm values of
spiraeoside were taken from literature and the ones of isoquercetrin assigned based on comparison to literature data (32). 1H NMR
identification of the metabolite of quercetin with retention time of 14.2 min was somewhat hampered by its instability, preventing
identification of all splitting patterns although major resonance chemical shift values could still be observed.
bonding of the 5-OH-hydrogen with the 4-keto, conjuga-
tion of this 5-OH does not readily occur. This leaves the
3-position as the only remaining possibility for monoglu-
curonosyl conjugated quercetin metabolites. The UV-
absorption spectrum of the metabolite at 10.7 min shows
a relatively large shift of band I from 370 nm in quercetin
to 354 nm. It is reported (29) that this relatively large
shift of the hypsochromic Band I, with a value of 12-17
nm points at conjugation at the 3-hydroxyl of quercetin.
Taking these considerations into account, it can be
concluded that this metabolite with a retention time of
10.7 min on HPLC must be the 3-O-monoglucuronosyl
of quercetin.
of all three glucuronosyls, i.e., 7-O-, 3′-O-, and 4′-O- from
luteolin, but the highly preferential 3′-O-glucuronosyl
formation for quercetin by UGT1A1 and UGT1A8 (Figure
4, panels a and d, Table 4). In addition, efficient conver-
sion of one flavonoid apparently does not imply efficient
conversion of the related compound, as illustrated by the
fact that UGT1A6 hardly converts quercetin (Figure 4e),
but forms significant levels of the 7-O-glucuronosyl from
luteolin (Figure 4b).
Thus, the results obtained reveal that regioselectivity
is dependent on the model flavonoid of interest and
the UDP-glucuronosyltransferase involved, UGT1A1,
UGT1A8, and UGT1A9 being especially active in fla-
vonoid conjugation whereas the isoenzymes UGT1A4,
UGT1A10, and the isoenzymes from the UGTB family,
UGT2B7 and UGT2B15, are less efficient.
Con ver sion of Lu teolin a n d Qu er cetin by In d i-
vid u a l UDP -Glu cu r on osyltr a n sfer a ses. In a subse-
quent series of incubations, the model flavonoids of the
present study were incubated with Supersomes contain-
ing specific individual UDP-glucuronosyltransferases. As
an example, Figure 4 presents the HPLC chromatograms
of incubations of UGT1A1, UGT1A6, and UGT1A9 with
luteolin (Figure 4, panels a-c) and quercetin (Figure 4,
panels d-f). These chromatograms represent the ones
with the highest levels of metabolite formation or with
different behavior between luteolin and quercetin conver-
sion. Table 4 presents the relative percentages of sub-
strate conversion in 30 min incubations as well as the
regioselectivities observed for the incubations of luteolin
and quercetin with UGT1A1, 1A3, 1A4, 1A6, 1A8, 1A9,
1A10, 2B7, and 2B15.
Discu ssion
In this study, the identification of phase II metabolites,
resulting from glucuronidation of two flavonoid model
compounds, i.e., luteolin and quercetin, was carried out
1
using HPLC, LC/MS, and H NMR. Incubation of luteolin
and quercetin with rat as well as human liver microsomes
resulted in the formation of 7-O-glucuronosyl and 3′-O-
glucuronosyl conjugates of luteolin and -quercetin as
major metabolites and the 4′-O-glucuronosyl derivative
of both flavonoids as a third, but minor metabolite. In
addition, quercetin, having a hydroxyl group at C3, was
metabolized to its 3-O-glucuronosyl conjugate as well.
Human and rat liver differ in the ratio of the metabolites
produced. Using intestine microsomes from human and
rat, it could be demonstrated that the type of metabolites
formed are the same as those formed by liver microsomes,
although regioselectivities of the glucuronidation vary
significantly.
Major metabolites observed were similar to the ones
observed in the microsomal incubations and identified
as 3-O-glucuronosyl (only with quercetin), 7-O-, 3′-O-, and
4′-O-glucuronosyls of the respective flavonoid. The data
reveal that these isoenzymes generally do not make a
single glucuronosyl metabolite but rather make a mixture
of 3-O-, 7-O-, 3′-O-, and 4′-O-glucuronosyls in varying
ratios (Table 4). Also important to derive from these data
is that preference of an isoenzyme for a specific site of
glucuronidation with one flavonoid (for example, the
7-OH for luteolin by UGT1A3) does not necessarily imply
preferential glucuronidation at the same position for a
structurally related flavonoid, since UGT1A3 preferen-
tially generates the 3′-O-glucuronosyl of quercetin. An-
other example of a striking discrepancy is the generation
Several studies report glucuronidated quercetin me-
tabolites in in vitro microsomal studies. Crespy et al. (19)
reported four unidentified glucuronosyl type metabolites,
when investigating the quercetin glucuronidating capac-
ity of microsomes from jejunum and ileum of rats.
Morand et al. (18) reported four glucuronosyls of quer-
cetin upon incubation of rat liver and cecal microsomal
fractions. Although LC-MS data are required to substan-

668 Chem. Res. Toxicol., Vol. 15, No. 5, 2002
Boersma et al.
F igu r e 4. HPLC chromatogram of incubations of luteolin with (a) UGT1A1 (b) UGT1A6 and (c) UGT1A9 and of quercetin with (d)
UGT1A1 (e) UGT1A6 and (f) UGT1A9 after 30 min incubation at 37 °C and based on detection at 345 nm for luteolin and 370 nm
for quercetin.
Ta ble 4. P er cen ta ges of Meta bolites F or m ed fr om Lu teolin a n d Qu er cetin w ith In d ivid u a l
UDP -Glu cu r on osyltr a n sfer a ses, Ba sed on P ea k Ar ea Mea su r ed by HP LC a t 345 n m for Lu teolin a n d 370 n m for
Qu er cetin
Luteolin
tR 11.4 min
tR 13.9 min
tR 14.5 min
substrate
UGT
7-O-Gluc (%)
-
4′-O-Gluc (%)
3′-O-Gluc (%)
conversion (%)
UGT1A1
UGT1A3
UGT1A4
UGT1A6
UGT1A8
UGT1A9
UGT1A10
UGT2B7
UGT2B15
21
54
40
98
39
49
64
19
88
-
-
-
-
-
-
-
-
-
50
28
38
1
22
12
24
4
29
18
22
1
39
39
12
77
6
26
7
0.5
40
25
31
11
10
4
6
Quercetin
tR 10.5 min
tR 10.7 min
tR 13.7 min
tR 14.2 min
substrate
UGT
7-O-Gluc (%)
3-O-Gluc (%)
4′-O-Gluc (%)
3′-O-Gluc (%)
conversion (%)
UGT1A1
UGT1A3
UGT1A4
UGT1A6
UGT1A8
UGT1A9
UGT1A10
UGT2B7
UGT2B15
6
21
0
30
13
50
38
12
56
0
16
0
16
8
26
35
41
0
9
11
54
32
9
5
14
3
85
52
46
22
70
19
13
44
20
30
21
1
3
20
36
9
19
4
24
tiate the monoglucuronidated nature of these metabolites
reported in the literature, it is likely that, based on the
results of the present study, formation of 7-O-, 3′-O-, 4′-
O-, and 3-O-monoglucuronosyl quercetin was also ob-
served in these previous studies.
Comparison of the conversion percentages of both rat
and human microsomes (Table 1), expressed per mil-
ligram of microsomal protein, reveals that the human
intestine microsomes cause a more extensive metabolism
of both luteolin and quercetin than the liver microsomes.

Regioselectivity of Phase II Metabolism
Chem. Res. Toxicol., Vol. 15, No. 5, 2002 669
These results illustrate that nonhepatic organs such as
the small intestine should not be underestimated with
respect to their flavonoid metabolizing capacities (22).
This result is in accordance with the findings of Morand
et al. (18) who compared the quercetin metabolising
capacity of rat liver and cecum microsomal fractions and
calculated that the capacity of the cecal wall to produce
glucuronosyl conjugates of quercetin is significantly
greater than that of the liver.
ronosyl conjugation in the two tissues. The regioselec-
tivity of luteolin and quercetin conjugation by human
liver microsomes, correlates best with the pattern ob-
tained with UGT1A9 and the human intestine incuba-
tions result in patterns of glucuronosyl conjugates, which
match the UGT1A1 and UGT1A8 metabolite patterns
best.
Cheng et al. (35) found that human UGT1A6 and
UGT1A9 were not expressed in small intestine and colon,
so not contributing to the first-line defense against
phenolic xenobiotic compounds that are presented via the
gut. They also reported that human UGT1A8 may be
involved in flavonoid metabolism and displayed a higher
glucuronidation rate than UGT1A10 toward these sub-
strates. This is in line with the results obtained in the
present study for UGT1A8 and UGT1A10 catalyzed
conversion of quercetin and luteolin. Green et al. (36)
found that UGT1A3 catalyzed the glucuronidation of
flavonoids, e.g., quercetin. Results of the present study
reveal that for luteolin and quercetin, the role of UGT1A9
(liver) and UGT1A1 and UGT1A8 (intestine) may be even
more pronounced. King et al. (37) found that rat and
human UGT1A1 exhibit comparable substrate specificity
and efficiency of glucuronidation for flavonoids. In this
study, it was found that the efficiency of glucuronidation
of human UGT1A1 for luteolin and quercetin is indeed
comparable, but that this UGT isoenzyme produces
significant differences in the regioselectivity of glucu-
ronidation for both flavonoids.
Finally, it is important to note that the position at
which glucuronidation is occurring is of importance for
the anti-oxidant as well as the pro-oxidant capacity of
the flavonoids (20). Glucuronidation at the 7-O-position,
is not thought to influence the anti-oxidant or pro-oxidant
activity of the two flavonoids to a significant effect
because this leaves the catechol moiety intact (18).
Taking into account the fact that the anti-oxidant action
(18, 20, 38) as well as the pro-oxidant toxicity (10) of these
catechol-type flavonoids is especially related to their
unconjugated 3′,4′-dihydroxyl moiety, it is of interest to
note that the human intestine UGT(s) appear to be
especially effective in conjugating this 3′,4′ catechol unit.
This would imply that upon glucuronidation along the
transport across the intestinal border the flavonoids loose
a significant part of these biological activities.
Several studies on the metabolism of flavonoids by
intestine models have been performed. Shimoi et al. (12)
investigated the absorption of luteolin by rat everted
small intestine. Luteolin was recovered in rat plasma as
two metabolites, which were not identified in detail, but
proved to be glucuronidated or sulfated forms of O-
methylated conjugates. Only a small part of the com-
pound remained unconjugated. Spencer et al. (20) per-
fused the jejunum and ileum from rats with quercetin
and luteolin. Luteolin yielded six glucuronosyls which
were not identified. In the present study three glucu-
ronosyl conjugates of luteolin, the 7-O-, the 3′-O-, and
the 4′-O-glucuronosyl luteolin, could be identified using
intestine microsomes from rat and man. Possibly, the
additional conjugates found by Spencer et al. (20) were
mixed conjugates arising from the combined action of
UDP-glucuronosyltransferases and catechol-O-methyl-
transferases, also known to be present in the intestine
and active in flavonoid phase II type metabolism. Zhu et
al. (15) even suggested that the rapid 3′-O-methylation
of quercetin by catechol-O-methyl transferase found in
vivo in hamsters could be an explanation for its lack of
carcinogenicity in vivo.
Day et al. (29) reported formation of four glucuronosyl
conjugates formed by incubation of quercetin with human
liver microsomes. They determined the position of each
quercetin glucuronosyl by using UV absorbance spectra
and known shift reagents. In the present study, the
glucuronosyl conjugates are identified with 1H NMR,
which is the method of choice for unambiguous identifi-
cation. Similar sites for glucuronidation of quercetin were
observed but the ratio of the formed products is different
from that reported by Day et al. (29). In the present
study, the 7-O-glucuronosyl is the major metabolite of
the human liver incubation with quercetin, whereas Day
et al. (29) have found that the 4′-OH is the preferential
position for glucuronosyl conjugation. This difference in
regioselectivity can be ascribed to a difference in the
quercetin concentration used and the kinetic parameters
(vmax app. and Km) for the different glucuronidations.
Glucuronidation at the 7-OH having a high vmax app. and
a Km of 6.5 µM, the 100 µM concentration of the present
study as compared to 1 µM in the study of Day et al.,
explains the relatively high contribution of 7-O-glucu-
ronidation in the present study.
Ack n ow led gm en t. This work was supported by the
“NWO Stimuleringsprogramma Verantwoorde voeding”
Project 014-12-012.
Refer en ces
(1) Middleton, E., J r., and Kandaswami, C. (1994) The impact of plant
flavonoids on mammalian biology: implications for immunity,
inflammation and cancer. In The flavonoids: advances in research
since 1986 (Harborne, J . B., Ed.) pp 619-652, Chapman & Hall,
London.
(2) Rice-Evans, C. A., Miller, N. J ., and Paganga, G. (1996) Structure-
antioxidant activity relationships of flavonoids and phenolic acids.
Free Radical Biol. Med. 20, 933-956.
(3) Diplock, A. T., Charleux, J .-L., Crozier-Willi, G., Kok, F. J ., Rice-
Evans, C. A., Roberfroid, M., Stahl, W., and Vina-Ribes, J . (1998)
Functional food science and defence against reactive oxygen
species. Br. J . Nutr. 80 (Suppl. 1), S77-S112.
(4) Penning, T. M., Burczynski, M. E., Hung, K. D., McCoull, K. D.,
Palackal, N. T., and Tsuruda, L. S. (1999) Dihydrodiol dehydro-
genases and polycyclic aromatic hydrocarbon activation: genera-
tion of reactive and redox active o-quinones. Chem. Res. Toxicol.
12, 1-18.
The regioselectivity of glucuronosyl conjugation is not
only dependent on the flavonoid of interest but also
on the UDP-glucuronosyltransferase involved. Human
UGT1A1, UGT1A8, and UGT1A9 were shown to be
especially active in conjugation of both flavonoids whereas
UGT1A4, UGT1A10, and the isoenzymes from the UGTB
family, UGT2B7 and UGT2B15, were less efficient.
Qualitative comparison of the results obtained with
human liver and intestine microsomal incubations and
the regioselectivities found for the various human UGT
isoenzymes may reveal to some extent which isoenzymes
play an important role in the preferential site of glucu-

670 Chem. Res. Toxicol., Vol. 15, No. 5, 2002
Boersma et al.
(5) MacGregor, J . T., and J urd, L. (1978) Mutagenicity of plant
flavonoids: structural requirements for mutagenic activity in
Salmonella typhymurium. Mutat. Res. 54, 297-309.
(6) Mitchel, R. E. J ., and McCann, R. (1993) Vitamin E is a complete
tumor promotor in mouse skin. Carcinogenesis 14, 659-662.
(7) Podmore, I. D., Griffiths, H. R., Herbert, K. E., Mistry, N., Mistry,
P., and Lunec, J . (1998) Vitamin C exhibits pro-oxidant properties.
Nature 392, 559-559.
(8) Galati, G., Chan, T., Wu, B., and O’Brien, P. (1999) Glutathione-
dependent generation of reactive oxygen species by the peroxi-
dase-catalysed redox cycling of flavonoids. Chem. Res. Toxicol. 12,
521-525.
(9) Bors, W., Heller, W., Michel, C., and Saran, M. (1990) Flavonoids
as antioxidants: determination of radical-scavenging efficiencies.
Methods Enzymol. 186, 343-355.
(10) Boersma, M. G., Vervoort, J ., Szymusiak, H., Lemanska, K.,
Tyrakowska, B., Cenas, N., Segura-Aguilar, J ., and Rietjens, I.
M. C. M. (2000) Regioselectivity and reversibility of the glu-
tathione conjugation of quercetin quinone methide. Chem. Res.
Toxicol. 13, 185-191.
(11) Awad, H. M., Boersma, M. G., Vervoort, J ., and Rietjens, I. M. C.
M. (2000) Peroxidase-catalyzed formation of quercetin quinone
methide-glutathione adducts. Arch. Biochem. Biophys. 378, 224-
233.
(12) Shimoi, K., Okada, H., Furugori, M., Goda, T., Takase, S., Suzuki,
M., Hara, Y., Yamamoto, H., and Kinae, N. (1998) Intestinal
absorption of luteolin and luteolin 7-O-â-glucoside in rats and
humans. FEBS Lett. 438, 220-224.
(13) Cao, G., Sofic, E., and Prior, R. L. (1997) Antioxidant and
prooxidant behavior of flavonoids structure-activity relation-
ships. Free Radical Biol. Med. 22, 749-760.
(14) Brown, J . E., Khidr, H., Hider, R. C., and Rice-Evans, C. A. (1998)
Structural dependence of flavonoid interactions with Cu2+ ions:
implications for their antioxidant properties. Biochem. J . 330,
1173-1178.
(15) Zhu, B. T., Ezell, E. L., and Liehr, J . G. (1994) Catechol-O-
methyltransferase-catalyzed rapid O-methylation of mutagenic
flavonoids. J . Biol. Chem. 269, 292-299.
(16) Piskula, M. K., and Terao, J . (1998) Accumulation of (-)-
epicatechin metabolites in rat plasma after oral administration
and distribution of conjugation enzymes in rat tissues. J . Nutr.
128, 1172-1178.
(17) Piskula, M. K., and Terao, J . (1998) Quercetin’s solubility affects
its accumulation in rat plasma after oral administration. J . Agric.
Food Chem. 46, 4313-4317.
(18) Morand, C., Crespy, V., Manach, C., Besson, C., Demigne, C., and
Remesy, C. (1998) Plasma metabolites of quercetin and their
antioxidant properties. Am J . Physiol. 275, R212-R219.
(19) Crespy, V., Morand, C., Manach, C., Besson, C., Demigne, C., and
Remesy, C. (1999) Part of quercetin absorbed in the small
intestine is conjugated and further secreted in the intestinal
lumen. Am. J . Physiol. 40, G120-G126.
(20) Spencer, J . P., Chowrimootoo, G., Choudhury, R., Debnam, E. S.,
Srai, S. K., and Rice-Evans, C. (1999) The small intestine can
both absorb and glucuronidate luminal flavonoids. FEBS Lett 458,
224-230.
(21) Manach, C., Texier, O., Morand, C., Crespy, V., Tegerat, F.,
Demigne, C., and Remesy, C. (1999) Comparison of the bioavail-
ability of quercetin and catechin in rats. Free Radical Biol. Med.
27, 1259-1266.
(22) Williamson, G., Day, J ., Plumb, G. W., and Couteau, D. (2000)
Human metabolic pathways of dietary flavonoids and cinnamates.
Biochem. Soc. Transact. 28, 16-22.
(23) Hempel, J ., Pforte, H., Raab, B., Engst, W., Bohm, H., and
J acobasch, G. (1999) Flavonols and flavones of parsley cell
suspension culture change the antioxidative capacity of plasma
in rats. Nahrung 43, 201-204.
(24) Liu, C. S., Song, Y. S., Zhang, K. J ., Ryu, J . C., Kim, M., and
Zhou, T. H. (1995) Gas chromatographic/mass spectrometric
profiling of luteolin and its metabolites in rat urine and bile. J .
Pharm. Biomed. Anal. 13, 1409-1414.
(25) Hertog, M. G. L., Hollman, P. C. H., Katan, M. B., and Kromhout,
D. (1993) Intake of potentially anticarcinogenic flavonoids and
their determinants in adults in The Netherlands. Nutr. Cancer
20, 21-29.
(26) Manach, C., Texier, O., Regerat, F., Agullo, G., Demigne, C., and
Remesy, C. (1996) Dietary quercetin is recovered in rat plasma
as conjugated derivatives of isorhamnetin and quercetin. Nutr.
Biochem. 7, 375-380.
(27) Manach, C., Morand, C., Demigne, C., Texier, O., Regerat, F., and
Remesy, C. (1997) Bioavailability of rutin and quercetin in rats.
FEBS Lett. 409, 12-16.
(28) Manach, C., Morand, C., Crespy, V., Demigne, C., Texier, O.,
Regerat, F., and Remesy, C. (1998) Quercetin is recovered in
human plasma as conjugated derivatives which retain antioxidant
properties. FEBS Lett. 426, 331-336.
(29) Day, A. J ., Bao, Y., Morgan, M. R. A., and Williamson, G. (2000)
Conjugation position of quercetin glucuronides and effect on
biological activity. Free Radical Biol. Med. 29, 1234-1243.
(30) Rietjens, I. M. C. M., and Vervoort, J . (1989) Microsomal
metabolism of fluoroanilines. Xenobiotica 19, 1297-1305.
(31) Soffers, A. E. M. F., Ploemen, J . H. T. M., Moonen, M. J . H.,
Wobbes, T., van Ommen, B., Vervoort, J ., van Bladeren, P. J .,
and Rietjens, I. M. C. M. (1996) Regioselectivity and quantitative
structure activity relationships for the conjugation of a series of
fluoronitrobenzenes by purified glutathione S-transferase en-
zymes from rat and man. Chem. Res. Toxicol. 9, 638-646.
(32) Markham, K., Geiger, H. (1993) 1H nuclear magnetic resonance
spectroscopy of flavonoids and their glycosides in hexadeu-
terodimethylsulfoxide. In The flavonoids: advances in research
since 1986 (Harborne, J . B., Ed.) pp 441-497, Chapman and Hall,
London.
(33) Wolfender, J .-L, Rodriguez, S., Hostettmann, K., and Hiller, W.
(1997) Liquid chromatography/ultra violet/mass spectrometric and
liquid chromatography/nuclear magnetic resonance spectroscopic
analysis of crude extracts of gentianaceae species. Phytochem.
Anal. 8, 97-104.
(34) Boutin, J . A., Meunier, F., Lambert, P., Hennig, P., Bertin, D.,
Serkiz, B., and Volland, J . (1993) In vivo and in vitro glucuronida-
tion of the flavonoid diosmetin in rats. Drug Metab. Dispos. 21,
1157-1161.
(35) Cheng, Z. Q., Radominska Pandya, A., and Tephly, T. R. (1998)
Cloning and expression of human UDP-glucuronosyltransferase
(UGT) 1A8. Arch. Biochem. Biophys. 356, 301-305.
(36) Green, M. D., King, C. D., Mojarrabi, B., Mackenzie, P. I., and
Tephly, T. R. (1998) Glucuronidation of amines and other xeno-
biotics catalyzed by expressed human UDP-glucuronosyltrans-
ferase 1A3. Drug Metab. Dispos. 26, 507-512.
(37) King, C. D., Green, M. D., Rios, G. R., Coffman, B. L., Owens, I.
S., Bishop, W. P., and Tephly, T. R. (1996) The glucuronidation
of exogenous and endogenous compounds by stably expressed rat
and human UDP-glucuronosyltransferase 1.1. Arch. Biochem.
Biophys. 332, 92-100.
(38) van Acker, S. A. B. E., Plemper-van Balen, G., van den Berg, D.-
J ., Bast, A., and van der Vijgh, W. J . F. (1998) Influence of iron
chelation on the antioxidant activity of flavonoids. Biochem.
Pharmacol. 56, 935-943.
TX0101705