Inhibition of extrahepatic human cytochromes P450 1A1 and 1B1 by metabolism of isoflavones found in Trifolium pratense (red clover)

Article abstract of DOI:10.1021/jf049418x

Biochanin A and formononetin are the predominant isoflavones in red clover. In a previous study (J. Agric. Food Chem. 2002, 50, 4783-4790), it was demonstrated that human liver microsomes converted biochanin A and formononetin to genistein and daidzein. This paper now shows CYP1B1-catalyzed O-demethylation of biochanin A and formononetin to produce genistein and daidzein, respectively, which inhibit CYP1B1. Recombinant human CYP1A1 or CYP1B1 was incubated with biochanin A or formononetin. CYP1A1 catalyzed isoflavone 4′-O-demethylation and hydroxylations with similar efficiency, whereas CYP1B1 favored 4′-O-demethylation over hydroxylations. Three of the biochanin A metabolites (5,7,3′-trihydroxy-4′-methoxyisoflavone, 5,7,8-trihydroxy-4′-methoxyisoflavone, and 5,6,7-trihydroxy-4′- methoxyisoflavone) were characterized by ¹H NMR spectroscopy and mass spectrometry. Daidzein (K_i = 3.7 μM) exhibited competitive inhibition of CYP1B1 7-ethoxyresorufin O-deethylase activity, and genistein (K_i = 1.9 μM) exhibited mixed inhibition. Biochanin A and/or formononetin may exert anticarcinogenic effects directly by acting as competitive substrates for CYP1B1 or indirectly through their metabolites daidzein and genistein, which inhibit CYP1B1.

Full text of DOI:10.1021/jf049418x

J. Agric. Food Chem. 2004, 52, 6623−6632 6623
Inhibition of Extrahepatic Human Cytochromes P450 1A1 and
1B1 by Metabolism of Isoflavones Found in Trifolium pratense
(Red Clover)
DEAN W. ROBERTS,† DANIEL R. DOERGE,† MONA I. CHURCHWELL,†
GONC¸ ALO GAMBOA DA COSTA,§ M. MATILDE MARQUES,§ AND
WILLIAM H. TOLLESON*,†
Division of Biochemical Toxicology, National Center for Toxicological Research, 3900 NCTR Road,
Jefferson, Arkansas 72079, and Centro de Qu´ımica Estrutural, Complexo I, Instituto Superior Te´cnico,
Av. Rovisco Pais, 1049-001 Lisboa, Portugal
Biochanin A and formononetin are the predominant isoflavones in red clover. In a previous study
(J. Agric. Food Chem. 2002, 50, 4783-4790), it was demonstrated that human liver microsomes
converted biochanin A and formononetin to genistein and daidzein. This paper now shows CYP1B1-
catalyzed O-demethylation of biochanin A and formononetin to produce genistein and daidzein,
respectively, which inhibit CYP1B1. Recombinant human CYP1A1 or CYP1B1 was incubated with
biochanin A or formononetin. CYP1A1 catalyzed isoflavone 4′-O-demethylation and hydroxylations
with similar efficiency, whereas CYP1B1 favored 4′-O-demethylation over hydroxylations. Three of
the biochanin A metabolites (5,7,3′-trihydroxy-4′-methoxyisoflavone, 5,7,8-trihydroxy-4′-methoxy-
isoflavone, and 5,6,7-trihydroxy-4′-methoxyisoflavone) were characterized by 1H NMR spectroscopy
and mass spectrometry. Daidzein (Ki ) 3.7 µM) exhibited competitive inhibition of CYP1B1
7-ethoxyresorufin O-deethylase activity, and genistein (Ki ) 1.9 µM) exhibited mixed inhibition.
Biochanin A and/or formononetin may exert anticarcinogenic effects directly by acting as competitive
substrates for CYP1B1 or indirectly through their metabolites daidzein and genistein, which inhibit
CYP1B1.
KEYWORDS: Biochanin A; CYP1A1; CYP1B1; cytochrome P450; demethylation; enzyme inhibition;
extrahepatic metabolism; formononetin; hydroxylation; isoflavones; Trifolium pratense
INTRODUCTION
and 6′′-malonylglucoside conjugates (10). Both the glucoside
conjugates and their isoflavone aglycons are bioavailable (11-
13). Isoflavone glycosides in the intestinal tract are deconjugated
to the corresponding aglycons by bacterial â-glucosidases and
by membrane-bound â-glucosidase expressed by enterocytes of
the small intestine. Isoflavone aglycons are taken up by
enterocytes and converted to glucuronide or sulfate conjugates
by phase II enzymes before their delivery into the circulation.
Enterocytes also express cytochrome P450 (CYP) 3A4 abun-
dantly, as well as CYP2C9 and CYP2C19 at moderate levels,
and lower or variable amounts of the inducible isoforms
CYP1A1, CYP1B1, and CYP2E1 (trace) (14-16). In addition,
CYP2E1 mRNA expression has been reported in the human
small intestine (17).
Previous studies in vitro showed that human liver microsomal
enzymes catalyze isoflavone 4′-O-demethylation (6, 9). How-
ever, the major hepatic CYP450 isoform, CYP3A4, does not
catalyze biochanin A or formononetin demethylation, and
recombinant human CYP2C9 and CYP2C19 are only weakly
competent to catalyze this transformation (9). These findings
suggest that other CYP450 isoforms, namely, the CYP1 family,
catalyze the conversion of formononetin and biochanin A to
Dietary consumption of foods and food additives containing
isoflavone phytoestrogens has been associated with a variety
of health benefits, including relief from symptoms of menopause
(1) and reduced risk of hormonal cancer (2). Extracts of
Trifolium pratense (red clover) have been proposed as an
alternative to traditional hormone replacement therapy in
postmenopausal women (3) even though clinical trials showed
no effect (4). The isoflavones formononetin and biochanin A,
and their corresponding 4′-O-demethylated derivatives daidzein
and genistein (Figure 1), comprise approximately 52, 40, 2,
and 5%, respectively, of the total isoflavone content of T.
pratense extracts (5). Further oxidative metabolism of daidzein
and genistein (6, 7) and of formononetin and biochanin A
(8, 9) by rat and human liver microsomes to yield polyphenolic
derivatives has recently been described.
Isoflavone metabolism in humans occurs stepwise (reviewed
in ref 8). Dietary isoflavonoids consist mainly of 7-â-D-glucoside
* Corresponding author [telephone (870) 543-7645; fax (870) 543-7136;
e-mail wtolleson@nctr.fda.gov].
† National Center for Toxicological Research.
§ Instituto Superior Te´cnico.
10.1021/jf049418x CCC: $27.50
© 2004 American Chemical Society
Published on Web 09/25/2004

6624 J. Agric. Food Chem., Vol. 52, No. 21, 2004
Roberts et al.
Figure 1. Hydroxylation and demethylation reactions catalyzed by cytochromes P450 1A1 and 1B1 for (A) biochanin A and (B) formononetin.
daidzein and genistein, respectively. In fact, human CYP1A2
is the primary hepatic CYP450 isoform responsible for this
activity (9, 18), provoking the hypothesis that extrahepatic CYP1
isoforms are also important for isoflavone metabolism.
Extensive demethylation of formononetin and biochanin A
was indicated by transient low levels (<10 ng/mL) of for-
mononetin and biochanin A and much higher and sustained
levels of daidzein and genistein (>100 ng/mL) in the plasma
of a human volunteer following consumption of a commercially
available red clover extract dietary supplement containing 23.7
mg of biochanin A and 14.3 mg of formononetin (10). This
evidence is consistent with efficient CYP450-dependent 4′-O-
demethylation. However, it also suggests that consumption of
a sufficiently large dose of formononetin and biochanin A may
saturate metabolic capacity, enabling detectable amounts of the
methylated isoflavones to accumulate in the peripheral circula-
tion.
co-workers (19) demonstrated that six common flavonoids
present in citrus juices at ∼70-200 mg/L are potent inhibitors
of CYP1A1, CYP1A2, and CYP1B1 activities in vitro. Dios-
metin (5,7,3′-trihydroxy-4′-methoxyflavone) and hesperetin
(5,7,3′-trihydroxy-4′-methoxyflavanone) are potent inhibitors of
the CYP1 family and share structural features with 3′-hydroxy-
biochanin A, a metabolite described in this study. Collectively,
these observations suggest that consumption of fruits and
vegetables containing flavonoids (e.g., acacetin, diosmetin,
naringenin, hesperetin, eriodictyol, or homoeriodictyol) at the
same time as food additives containing biochanin A or for-
mononetin could inhibit CYP1 family enzymes in the gut and
liver and increase the concentration of biochanin A or for-
mononetin in blood.
In view of the putative health benefits associated with
phytoestrogen consumption, the potential metabolic interaction
between phytoestrogens and endogenous estrogens in target
tissues for estrogen carcinogenesis is of particular interest. The
major source of estrogen is a three-step reaction sequence
catalyzed by CYP19 (aromatase), which converts androstene-
Consumption of foods containing flavonoids, the 2-substituted
analogues of isoflavonoids, may inhibit gut enzymes that
normally catalyze isoflavone O-demethylation. Doostdar and

Inhibition of CYP1A1 and CYP1B1 by Isoflavone Metabolism
J. Agric. Food Chem., Vol. 52, No. 21, 2004 6625
was evaporated under vacuum, and the samples were dissolved in 0.1
mL of methanol for further purification by HPLC until >90% purity
was achieved. Unstable isoflavone metabolites were sealed under argon
dione and testosterone to estrone and estradiol, respectively (20).
Normal human breast and human breast tumor tissues are known
to express various CYP450 isoforms, including CYP1B1, a
catalytically efficient estrogen 4-hydroxylase (21). The co-
localization of aromatase and CYP1B1 in breast tissue suggests
that the local concentration of putatively carcinogenic 4-hy-
droxy-estrogen products may be higher than that indicated by
the circulating levels of these metabolites (21). It is conceivable
that alternate substrates or inhibitors of CYP1B1 could be
associated with a reduced risk of estrogen-induced carcinogen-
esis.
1
and stored at -80 °C until H NMR analyses could be performed.
Isoflavone Analyses Using HPLC-EC and HPLC-MS. Analyses
using high-pressure liquid chromatography with electrochemical detec-
tion (HPLC-EC) and HPLC with mass spectrometry (MS) or tandem
mass spectrometry (MS/MS) detection were conducted as described
previously (9).
1H NMR Analyses. 1H NMR analyses of the isoflavone metabolites
isolated from rat liver microsomal incubations were performed in
dimethyl sulfoxide-d6 solutions on a Varian Gemini 300 spectrometer
operated at 300 MHz. Chemical shifts are reported in parts per million
downfield from tetramethylsilane, and coupling constants are reported
in hertz. Proton assignments were based on comparison with the spectra
of the parent biochanin A or formononetin, as well as their previously
characterized metabolites (8, 9). Whenever necessary, 2D 1H-1H-COSY
sequences were run for additional confirmation of the assigned
resonances.
In the work presented here we examine the metabolism of
formononetin and biochanin A by CYP1A1 and CYP1B1 to
yield daidzein, genistein, and hydroxylation products and their
effects on the catalytic activities of both isoforms.
MATERIALS AND METHODS
Enzyme Kinetic Constants for Human Cytochromes P450 1A1
and 1B1. Michaelis-Menten enzyme kinetic constants were calculated
using the isoflavone metabolite concentrations determined in the ethyl
acetate extracts from 0.5 mL reactions. Nonlinear regression analysis
was performed using the Prism software program (GraphPad Software
Inc., San Diego, CA).
Chemicals and Biochemicals. Human cytochrome P450 isoforms
1A1 and 1B1 and human cytochrome P450 oxidoreductase, expressed
by a baculovirus microsomal expression system, and an NADPH
regeneration system were obtained from Gentest Corp. (Woburn, MA).
Genistein [CAS Registry No. 446-72-0], daidzein [486-66-8], for-
mononetin [485-72-3], and biochanin A [491-80-5] were obtained from
Indofine Chemical Co. (Somerville, NJ). Isoflavone stock solutions were
prepared in DMSO and diluted to 0.1 mM with absolute ethanol (0.2%
DMSO, final). The purity of isoflavone stock solutions was verified
by HPLC with UV and electrochemical detection, and concentrations
were determined by UV spectrophotometry using the molar absorptivity
coefficients reported by Franke et al. (22).
Microsomal Incubations with Human CYP450 Isoforms. Each
0.5 mL reaction mixture included 0-10 nmol of isoflavones, mi-
crosomes containing 1 pmol of recombinant human CYP450 isoform
1A1 or 1B1, and microsomes expressing recombinant human cyto-
chrome P450 oxidoreductase (0.09 mg). The protein concentration was
adjusted to 0.1 mg/mL by adding control microsomes devoid of
CYP450 activity (Gentest). Reactions also contained 50 mM potassium
phosphate (pH 7.4), 1 mM EDTA, 1 mg/mL ascorbic acid, 1.33 mM
NADP+, 3.3 mM glucose-6-phosphate, and 3.3 mM magnesium
chloride. Reactions were equilibrated for 3 min at 37 °C and then
initiated by adding 0.4 unit/mL glucose-6-phosphate dehydrogenase.
Incubations were conducted at 37 °C with gentle mixing for the times
given in the appropriate figure legends. The reactions were quenched
by extracting twice with 1.0 mL of ethyl acetate. The organic extracts
were combined, dried in vacuo, and stored at -80 °C. The samples
were reconstituted in 100 µL of methanol assisted by ultrasonic
agitation, and 20 µL aliquots were injected into the HPLC detection
system.
Inhibition of Cytochrome P450 1B1-Dependent 7-Ethoxyresoru-
fin Deethylase (EROD) Activity by Genistein and Daidzein. EROD
activity was determined as described by the enzyme supplier using a
Fluoroskan 96-well plate reader (Thermo Labsystems Inc., Franklin,
MD) equipped with 550 nm excitation and 590 nm emission filters.
Each 250 µL reaction mixture contained 0.1 M potassium phosphate
(pH 7.4), 2 mM EDTA, 1 mg/mL ascorbic acid, 4 nM CYP1B1, 0.1
mg/mL microsomal protein, 0.4 nM cytochrome P450 oxidoreductase,
5-1280 nM 7-ethoxyresorufin, 1.33 mM NADP+, 3.3 mM glucose-
6-phosphate, and 3.3 mM magnesium chloride. Genistein or daidzein
(0-20 µM) was added to each mixture, which was then incubated for
3 min at 37 °C. The reactions were initiated by adding 0.4 unit/mL
glucose-6-phosphate dehydrogenase, and the increase in fluorescence
at 590 nm was recorded. Product formation was determined by
comparing fluorescence measurements to those obtained from resorufin
standard curves. Enzyme inhibition constants were calculated using
Sigma Plot (SPSS, Inc., Chicago, IL).
Statistical Analyses. Statistical analyses were performed using Prism
software. Student’s t test and one-way ANOVA were used to compare
enzyme kinetic and inhibition parameters. The level of statistical
significance was set at P < 0.05 for all data comparisons.
RESULTS
Rat Liver Microsome Reactions. Bulk reactions (100 mL) were
used to produce sufficient quantities of isoflavone metabolites for H
Identification of Metabolites Formed by O-Demethylation
and Hydroxylation of Formononetin and Biochanin A. In a
previous paper (9) we demonstrated that formononetin and
biochanin A are O-demethylated by human hepatic microsomal
enzymes to yield daidzein and genistein, respectively. In
1
NMR analysis. The incubation mixtures contained either 0.5 mM
formononetin or biochanin A and 1.0 mg/mL â-naphthoflavone-induced
rat liver microsomes (In Vitro Technologies, Baltimore, MD) with the
NADPH generating system and buffer described above. The mixtures
were incubated overnight at 37 °C and extracted with ethyl acetate
(3 × 100 mL). The extracts were passed through a mixed bed ion-
exchange resin (Bio-Rad AG 501 X8 resin; 5 g) preequilibrated with
ethyl acetate to remove charged lipids and concentrated to 4 mL under
vacuum. The volume of the concentrated extract was adjusted to 8 mL
using methanol and then further diluted to 200 mL with high-purity
water. The diluted extract was divided into 100 mL portions and applied
to separate Waters Sep-Pak cartridges (C-18 resin, 6 mL) preequilibrated
with aqueous 5% methanol. The cartridges were washed with 15 mL
of 5% methanol and then eluted using 5 mL of 75% methanol. The
eluted samples were dried under vacuum and reconstituted to 0.5 mL
using methanol. Isoflavone metabolites were isolated by preparative
reverse phase HPLC as described previously (9). Metabolite-containing
HPLC fractions were diluted with high-purity water to a final
concentration of 4% acetonitrile, adsorbed to C18 Sep-Pak cartridges
as described above, and eluted with 4 mL of methanol. The solvent
1
addition, we used HPLC-MS and H NMR spectroscopy to
identify 7,3′-dihydroxy-4′-methoxyisoflavone, 6,7-dihydroxy-
4′-methoxyisoflavone, and 7,8-dihydroxy-4′-methoxyisoflavone
as metabolites of formononetin formed by human liver mi-
crosomal enzymes, particularly cytochrome P450 isoforms 1A2
and 2C9*1. Besides genistein, three additional metabolites of
biochanin A formed by human liver microsomal enzymes were
detected. One of these metabolites was identified as 3′-
hydroxygenistein. The other two metabolites had protonated
molecular ions (m/z 301) consistent with hydroxylated biochanin
A regioisomers, but insufficient quantities for detailed structural
analyses were generated.
In this paper we show that extrahepatic human cytochrome
P450 isoforms 1A1 and 1B1 also efficiently catalyze 4′-O-
demethylation and hydroxylation reactions of formononetin and

6626 J. Agric. Food Chem., Vol. 52, No. 21, 2004
Roberts et al.
Table 1. 1H NMR Data for Biochanin A (BA) and BA Metabolitesa
biochanin A
R2 R3
R4 CH3)
genisteinb
R2 R3
R4 H)
3
′
-OH-biochanin A
(R1 R2 H;
CH3)
8-OH-biochanin A
(R1 R3 H;
CH3)
6-OH-biochanin A
OH; R2 R3
R4 CH3)
(R1
)
)
)
)
H;
(R1
)
)
)
)
)
OH; R4
)
)
)
OH; R4
)
)
(R1
)
)
) H;
δ
R3
)
R2
)
)
CH3
3.79
3.79
3.79
3.79
(3H, s)
(3H, s)
(3H, s)
(3H, s)
H2
H6
H8
8.38
(1H, s)
8.26
(1H, s)
8.32
(1H, s)
8.41
(1H, s)
8.35
(1H, s)
6.24
(1H, s)
6.21
6.22
6.30
(1H, s)
(1H, d, J
)
)
1.95)
1.95)
8.2)
(1H, d, J
6.38
(1H, d,J 2.0)
)
2.0)
6.40
(1H, s)
6.39
(1H, d, J
6.50
(1H, s)
H2
H3
H2
H5
H6
′
′
′
′
′
,6
′
′
7.50
7.38
(2H, d, J
7.51
7.51
(2H, d, J
)
8.6)
8.6)
)
(2H, d, J
)
8.8)
8.8)
(2H, d, J
7.00
(2H, d, J
)
8.7)
8.7)
,5
7.01
6.82
(2H, d, J
7.00
(2H, d, J
)
)
8.2)
(2H, d, J
)
)
7.03
(1H, d, J
)
1.7)
6.96
(1H, bs)
6.95
(1H, m)
5-OH
6-OH
7-OH
8-OH
12.93
(1H, s)
12.98
(1H, s)
12.97
(1H, bs)
12.35
(1H, s)
12.76
(1H, s)
8.79
(1H, bs)
11.04
(1H, bs)
10.92
(1H, s)
NDc
10.55
(1H, bs)
10.54
(1H, bs)
8.76
(1H, bs)
3
′
′
-OH
-OH
9.06
(1H, bs)
4
9.60
(1H, s)
a Spectra were recorded in DMSO-d6. Abbreviations: bs, broad singlet; d, doublet; m, multiplet; s, singlet; J, coupling constant (Hz). b Data from Wang, H.; Nair, M. G.;
Strasburg, G. M.; Booren, A. M.; Gray, J. I. Antioxidant polyphenols from tart cherries (Prunus cerasus). J. Agric. Food Chem. 1999, 47, 840
844. c Not detected.
−
biochanin A. We further show that daidzein and genistein are
the major reaction products of CYP1A1- and 1B1-catalyzed
reactions with formononetin and biochanin A, respectively, and
that hydroxylation products of formononetin and biochanin A
are formed at lower levels. In addition, we characterize three
hydroxybiochanin A metabolites.
Large-scale incubations of biochanin A with â-naphthofla-
vone-induced rat liver microsomes were used to produce
sufficient quantities of the hydroxybiochanin A metabolites for
1H NMR analysis. In these bulk reactions, biochanin A was
predominantly O-demethylated to form genistein (2-3% con-
version), whereas each of the hydroxylated biochanin A products
was produced in yields of 0.5-2%. Increased reaction times
and increased ratios of NADPH regeneration system to enzyme
did not markedly increase the yields of these products, sug-
gesting that end-product inhibition limited reaction progress.
Moreover, two of the three hydroxylated biochanin A isomers
were unstable and were stored under argon to prevent further
oxidation.
HPLC-ES-MS/MS analysis confirmed that the three hydroxy-
biochanin A metabolites exhibited protonated molecules (m/z
301), consistent with monohydroxylation. 1H NMR spectroscopy
(Table 1) was performed to assign the structures. All three
metabolites exhibited a three-proton singlet at 3.79 ppm
consistent with the presence of a 4′-methoxy substituent, and
the H2 proton was detected as a sharp downfield singlet (8.32-
8.41 ppm), indicating conservation of the chromone substructure.
The NMR spectra of 8-hydroxybiochanin A and 6-hydroxy-
biochanin A displayed very similar features. In both instances,
the presence of two mutually ortho-coupled two-proton doublets
in the aromatic region (7.00 and 7.51 ppm), characteristic of
an AB coupling pattern and virtually identical to those observed
in biochanin A (Figure 2 and Table 1), indicated that no further
substitution had occurred in the 3-(4′-methoxy)phenyl substitu-
ent. Similarly to biochanin A, both 8-hydroxybiochanin A and
6-hydroxybiochanin A exhibited two downfield one-proton
resonances, one as a sharp singlet at δ >12.3 and the other as

Inhibition of CYP1A1 and CYP1B1 by Isoflavone Metabolism
J. Agric. Food Chem., Vol. 52, No. 21, 2004 6627
Table 2. Kinetic Constants for CYP1A1 and CYP1B1
enzyme, substrate, and
products
ka0pp
(min
ka0pp/Km
-
1
-
-1
Km
(
µ
M)
)
(min 1 µ
M
)
Incubations of Formononetin with CYP1A1
daidzein
-hydroxyformononetin
8-hydroxyformononetin
6-hydroxyformononetin
5.9
7.5
5.8
4.5
±
±
±
±
1.4
1.9
1.1
0.8
13.7
1.6
6.7
7.3
±
±
±
±
1.7
0.2
0.5
0.5
2.3
0.21
1.2
1.6
±
±
±
±
0.7
0.08
0.3
3
′
0.4
Incubations of Biochanin A with CYP1A1
genistein
-hydroxybiochanin A
8-hydroxybiochanin A
6-hydroxybiochanin A
5.2
7.3
6.0
2.7
±
±
±
±
0.9
0.8
0.7
0.4
8.4
1.2
2.9
2.6
±
±
±
±
0.6
0.06
0.2
0.1
1.6
1.6
0.48
0.96
±
±
±
0.4
0.03
0.08
0.18
3
′
±
Incubations of Formononetin with CYP1B1
daidzein
-hydroxyformononetin
8-hydroxyformononetin
6-hydroxyformononetin
0.27
nda
nd
±
0.06
3.6
nd
nd
nd
±
0.2
13
nd
nd
nd
±
4
3
′
nd
Incubations of Biochanin A with CYP1B1
genistein
-hydroxybiochanin A
8-hydroxybiochanin A
6-hydroxybiochanin A
1.3
1.5
±
±
13
0.1
0.1
4.9
0.1
1.2
nd
±
±
±
0.1
0.01
0.3
3.8
0.6
0.03
nd
±
±
±
0.4
0.01
0.02
3
′
40
±
nd
a
Not determined. Iterative nonlinear regression methods failed to converge on
enzyme kinetic constants for these metabolites.
lation occurred in the 3-aryl substituent. This was confirmed
by the absence of an AB coupling profile in that ring, which
contained only three almost superimposed aromatic protons at
approximately δ 7.0, in a pattern closely resembling that of 7,3′-
dihydroxy-4′-methoxyisoflavone (9). Furthermore, two broad
exchangeable resonances, assigned to phenolic protons, were
detected at δ 12.97 (5-OH) and δ 9.06, with the latter being
assigned to the additional hydroxyl proton in the 3-aryl
substituent. The 7-OH resonance, expected at ∼10.5-11.0 by
comparison with the spectrum of biochanin A, was presumably
too broad to be detected as a result of the high dilution of the
sample. Taken together, the 1H NMR data are fully compatible
with characterization of this metabolite as 3′-hydroxybiochanin
A.
Enzyme Kinetics for the Metabolism of Formononetin and
Biochanin A Catalyzed by Human Cytochrome P450 Iso-
forms 1A1 and 1B1. Human cytochromes 1A1 and 1B1
catalyzed the 4′-O-demethylation of formononetin and biochanin
A to yield daidzein and genistein, respectively. Figures 3 and
4 show that the O-demethylated metabolites daidzein and
genistein accumulated more rapidly than the hydroxylation
products, particularly for CYP1B1-catalyzed reactions.
Figure 5 and Table 2 show that Km values are similar for
the formation of daidzein (5.9 ( 1.4 µM) and genistein (5.2 (
0.9 µM) catalyzed by CYP1A1. Comparison of the apparent
rate constants, ka0pp, for these CYP1A1-catalyzed reactions
(Table 2) reveals that the rate of formation of daidzein (13.7
( 1.7 min-1) is greater than that for genistein formation (8.4
( 0.6 min-1). The enzyme specificity constants k0app/Km for
CYP1B1-catalyzed demethylation reactions show that daidzein
formation (13 ( 4 min-1 µM-1) is favored over genistein
formation (3.8 ( 0.4 min-1 µM-1). Notably, the Km value for
4′-O-demethylation of formononetin by CYP1B1 (0.27 ( 0.06
µM) is 22-fold lower than that for the same reaction catalyzed
by CYP1A1.
Figure 2. Aromatic region (
A and the monohydroxylated biochanin A metabolites characterized in
this study.
δ 6−
8.5) of the 1H NMR spectra of biochanin
a broad singlet at δ ∼10.5. These signals were assigned to the
5-OH and 7-OH protons, respectively, on the basis of the
expected deshielding caused by intramolecular hydrogen bond-
ing of the 5-hydroxyl proton to the peri carbonyl oxygen. In
contrast with biochanin A, for which two signals (H6 and H8)
were observed in the upfield zone of the aromatic region (6.24
and 6.40 ppm, respectively), only one sharp singlet, accounting
for one proton, was observed for each of the two metabolites
(Figure 2 and Table 1). This observation, combined with the
presence, in both instances, of a broad singlet at approximately
8.8 ppm (Table 1), consistent with a phenolic proton, further
corroborated that hydroxylation had occurred at either C6 or
C8. On the basis of the relative chemical shifts of the upfield
aromatic protons (Table 1), 8-hydroxybiochanin A [δ (H6) )
6.30] and 6-hydroxybiochanin A [δ (H8) ) 6.50] were
identified.
The 1H NMR spectrum of the 3′-hydroxybiochanin A
metabolite (Figure 2 and Table 1) indicated the presence of
the H6 and H8 protons as two mutually meta-coupled one-proton
doublets (δ 6.22 and 6.38, respectively), implying that hydroxy-
Comparison of enzyme specificity constants in Table 2 shows
that CYP1A1 catalyzes hydroxylation at the C3′ position of
formononetin (0.21 ( 0.08 min-1 µM-1) less efficiently than
4′-O-demethylation (2.3 ( 0.7 min-1 µM-1) or hydroxylation

6628 J. Agric. Food Chem., Vol. 52, No. 21, 2004
Roberts et al.
Figure 3. Time course for metabolism of biochanin A by human cytochromes P450 1A1 and 1B1: (A) time course for formation of primary metabolites
of biochanin A by CYP1A1; (B) time course for formation of primary metabolites of biochanin A by CYP1B1; genistein ( ); 3 -hydroxybiochanin A ( );
8-hydroxybiochanin A ( ); 6-hydroxybiochanin A ( ). Reactions (0.5 mL) consisted of human cytochrome P450 (10 pmol) with 0.5 nmol of biochanin
O
′
3
0
]
A, initiated by adding the NADPH regeneration system, and quenched as described under Materials and Methods.
Figure 4. Time course for metabolism of formononetin by human cytochromes P450 1A1 and 1B1: (A) time course for formation of primary metabolites
of formononetin by CYP1A1; (B) time course for formation of primary metabolites of formononetin by CYP1B1; daidzein (
8-hydroxyformononetin ( ); 6-hydroxyformononetin ( ). Reactions (0.5 mL) consisted of human cytochrome P450 (10 pmol) with 0.5 nmol of formononetin,
initiated by adding the NADPH regeneration system, and quenched as described under Materials and Methods.
O); 3′-hydroxyformononetin (3);
0
]
Figure 5. Lineweaver
demethylation of biochanin A (
initiated by adding the NADPH regeneration system and incubated for 1 h at 37
−
Burk plots for isoflavone 4
′
-O-demethylation reactions catalyzed by human cytochromes P450 1A1 (A) and 1B1 (B):
). Reactions (0.5 mL) consisted of 0 20 M isoflavones with CYP1A1 or CYP1B1 (1 pmol)
C. Extracts were dissolved in methanol and analyzed by HPLC-EC.
4′-O-
3
) or of formononetin (
O
−
µ
°
at either C6 (1.6 ( 0.4 min-1 µM-1) or C8 (1.2 ( 0.3 min-1
µM-1), but much more efficiently than CYP1B1. Although all
known monohydroxylated formononetin and biochanin A
metabolites were detectable in reactions containing CYP1B1,
Km and ka0pp values could not be determined reliably for
hydroxylation at C6, C8, or C3′ of formononetin or for
hydroxylation of biochanin A at C6. Hydroxylation at C3′ of
biochanin A was catalyzed by CYP1B1 with a Km value similar
to that for 4′-O-demethylation, albeit at a 49-fold lower rate.
Inhibition of Cytochrome P450 1B1 7-Ethoxyresorufin
Deethylase Activity by Genistein and Daidzein. Microsomal
incubations containing CYP1A1 or CYP1B1 and 0.2-200 µM
genistein or daidzein were also conducted as described for the
biochanin A and formononetin incubations. As noted above,

Inhibition of CYP1A1 and CYP1B1 by Isoflavone Metabolism
J. Agric. Food Chem., Vol. 52, No. 21, 2004 6629
Figure 6. Inhibition of human cytochrome P450 1B1-dependent 7-ethoxyresorufin deethylase activity by genistein and daidzein: (A
CYP1B1 activity by genistein; (D F) inhibition of CYP1B1 activity by daidzein; (A, D) Dixon plots [( ) 80 nM, ( ) 160 nM, ( ) 320 nM, (
7-ethoxyresorufin vs isoflavone inhibitor concentrations]; (B, E) slopes ( ) from Lineweaver Burk analyses vs concentration of isoflavone inhibitor; (C,
F) Eadie Hofstee plots [( ) 0.0, ( ) 2.5, ( ) 5.0, ( ) 10, and ( ) 20 M genistein; or ( ) 0.0, ( ) 2.0, ( ) 4.0, ( ) 10, or ( ) 20 M daidzein].
Reaction progress was monitored in real time using a 96-well fluorescent plate reader (550 nm excitation, 590 nm emmission).
−C) inhibition of
−
O
3
0
]
) 640 nM
b
−
−
b
1
9
[
2
µ
b
1
9
[
2
µ
CYP1B1 exhibited greater selectivity for isoflavone demethy-
lation than CYP1A1. This observation led us to investigate the
ability of the demethylated isoflavones genistein and daidzein
to inhibit CYP1B1 EROD activity. The data are summarized
in Figure 6. We found evidence for inhibition of CYP1B1
EROD activity by both genistein (Ki ) 1.9 ( 0.2 µM; Figure
6A-C) and daidzein (Ki ) 3.7 ( 0.4 µM; Figure 6D-F).
Enzyme inhibition constants, Ki, were determined by replotting
slopes from Lineweaver-Burk analyses versus inhibitor con-
centrations (Figure 6B,E). Dixon and Eadie-Hofstee plots
(Figure 6D,F, respectively) indicated competitive inhibition of
CYP1B1 EROD activity by daidzein. The evidence for simple
competitive inhibition of CYP1B1 EROD activity by genistein
(Figure 6A,C) was not as strong as for daidzein and was more
consistent with mixed-type inhibition. These results reveal a
potential negative feedback loop in which the end products of
4′-O-demethylation of biochanin A or formononetin act as potent
CYP1B1 inhibitors.
DISCUSSION
Although it is generally acknowledged that isoflavones have
1
only a fraction of the estrogenic activity (∼ /1000) of endoge-
neous estrogens, individuals consuming diets or nutritional
supplements rich in isoflavones may have circulating levels of
isoflavones that are orders of magnitude higher than the levels

6630 J. Agric. Food Chem., Vol. 52, No. 21, 2004
Roberts et al.
of endogeneous estrogens (21, 23). This is particularly evident
higher that those determined for the inhibition of CYP1B1
in postmenopausal women.
EROD activity in the present study. Similarly, equol (IC50 )
1.7 ( 0.8 mM) and genistein (IC50 ) 5.6 ( 1.8 mM)
noncompetitively inhibit mouse liver microsomal CYP1A and
human CYP1A2 EROD activities (27). Noncompetitive inhibi-
tion of CYP2E1-dependent p-nitrophenol hydroxylation by equol
(IC50 ) 560 µM) and genistein (IC50 ) 10 mM) has also been
established (27).
Approximately 90% of the isoflavones in red clover extract
are formononetin and biochanin A, and the remaining isofla-
vones are the corresponding 4′-O-demethylated isoflavones
daidzein and genistein. If 4′-O-demethylation by gut enterocytes
or the liver occurs during absorption and enterohepatic circula-
tion, then the isoflavones reaching peripheral blood and organs
are predominantly genistein and daidzein. However, if the
mechanisms that would normally 4′-O-demethylate formo-
nonetin and biochanin A are overwhelmed or inhibited, the local
metabolism of the parent 4′-methoxyisoflavones may produce
an altered constellation of effects associated with the local
production of hydroxylated metabolites of biochanin A and
formononetin. Alternatively, local 4′-O-demethylation products
of formononetin and biochanin A, daidzein and genistein, could
compete with endogeneous estrogens for local CYP1B1,
potentially reducing the formation of estrogen 3,4-catechols.
In addition to their estrogenic activity, ingested isoflavones
or their metabolites may modulate hormonal carcinogenesis by
influencing the metabolism and local availability of endogeneous
hormones. Humans express at least seven isoforms of 17â-
hydroxysteroid dehydrogenase (17â-HSD), which are key
enzymes in the formation of both androgens and estrogens (28).
Inactivation of 17â-HSD type 3 is considered to be critical for
male pseudohermaphroditism. Biochanin A (10.3 µM) and
baicalein (9.3 µM) are potent inhibitors of recombinant human
17â-HSD type 3 (29). Human 17â-HSD type 5 is inhibited
effectively by quercetin (3,5,7,3′,4′-pentahydroxyflavone) and
biochanin A (30). The isoflavones genistein, biochanin A, and
equol inhibit human steroid 5R reductase isoforms 1 and 2,
which convert testosterone to dihydrotestosterone, a more potent
androgen (31). 3â-Hydroxysteroid dehydrogenase (3â-HSD,
progesterone reductase) functions as a dehydrogenase and a
steroid 4-ene/5-ene isomerase. Both 3â-HSD activities are
strongly inhibited by genistein, daidzein, biochanin A, and
formononetin (IC50 values of 0.4-11 µM) (32).
In humans, isoflavones have been shown to affect the
metabolism of endogeneous estrogens and, specifically, the
formation of mutagenic estrogen quinone metabolites (33). The
current study provides mechanistic insight into this important
observation. Biochanin A and formononetin and their metabo-
lites have the potential not only to inhibit CYP1B1 or to act as
alternative substrates for CYP1B1, affecting the production of
mutagenic estrogen 3,4-catechols, but also to compete with
catechol-O-methyltransferase (34, 35) which normally detoxifies
estrogen catechols. The catechol and pyrogallol metabolites of
biochanin A and formononetin need to be further considered,
not only in the context of potential health benefits but also in
the context of their potential toxicity. These metabolites are
candidates for further metabolism to quinones and associated
redox cycling (36, 37), and their potential to form carcinogenic
depurinating DNA adducts (38, 39) should also be considered.
Genistein and daidzein are the principal metabolites from red
clover isoflavone extracts produced by human CYP450 iso-
forms. Among the human hepatic CYP450 isoforms, CYP1A2
is clearly the most efficient catalyst for the formation of genistein
from biochanin A (9, 18). Hu and co-workers found that human
CYP1A2 exhibited Vmax ) 903 pmol/min/mg of protein for
biochanin A demethylation (18), 9-fold higher than that for
pooled human liver microsomes (9). In the present study, we
found that human CYP1A1 catalyzes isoflavone 4′-O-demethy-
lation and hydroxylations with similar efficiency, whereas
CYP1B1 strongly favors isoflavone 4′-O-demethylation over
hydroxylation (Table 2). Furthermore, CYP1B1 exhibits Km
values for the 4′-O-demethylation of biochanin A and formo-
nonetin that are ∼11-fold lower than those determined for
similar reactions catalyzed by pooled human liver microsomes
(9). Notably, Vmax values (derived from ka0pp determinations)
for the synthesis of genistein from biochanin A catalyzed by
CYP1A1 and CYP1B1 are 150- and 90-fold greater, respec-
tively, than those measured for CYP1A2. Thus, these extrahe-
patic CYP450 isoforms represent higher efficiency catalysts,
exhibiting lower Km and higher kapp for the 4′-O-demethylation
0
reactions than related hepatic enzymes.
The data (Figure 6) also suggest that isoflavone 4′-O-
demethylation may create negative feedback loops modulating
CYP1A1 and CYP1B1 activity. This conclusion is substantiated
by our observation that CYP1B1 EROD activity was inhibited
by genistein (Ki ) 1.9 ( 0.2 µM) and daidzein (Ki ) 3.7 ( 0.4
µM). A similar effect was deduced for human CYP1A2, the
phenacetin O-demethylation activity of which was inhibited by
genistein (IC50 ) 16 µM) (18). Recent studies have shown that
genistein, but not daidzein, added to cultured MCF-7 human
breast carcinoma cells treated with 7,12-dimethylbenz[a]an-
thracene (DMBA) prevented the formation of DMBA-DNA
adducts (24), and this protective effect was reproduced in similar
cultures treated with biochanin A (25). The same authors found
that genistein and biochanin A are competitive inhibitors of
recombinant human CYP1A1 (Ki ) 15 and 4 µM, respectively)
and CYP1B1 (Ki ) 0.68 and 0.59 µM, respectively) EROD
activity using assay conditions comparable to those used in this
study (24, 25).
ABBREVIATIONS USED
ka0pp, apparent enzyme-catalyzed rate constant; COSY, cor-
relation spectroscopy; CYP450, cytochrome P450; Ki, enzyme
inhibition constant; k0app/Km, enzyme selectivity constant;
EROD, 7-ethoxyresorufin deethylase; HPLC-EC, high-pressure
liquid chromatography with electrochemical detection; HPLC-
MS, high-pressure liquid chromatography with mass spectro-
metric detection; IC50, concentration of an agent in vitro
producing 50% inhibition of the biological effect; Km, Michae-
lis-Menten constant; Vmax, maximal enzyme-catalyzed reaction
velocity.
LITERATURE CITED
Interestingly, human CYP1B1 EROD activity appears to be
more sensitive to inhibition by daidzein or genistein than other
CYP450 isoforms. For instance, both daidzein and genistein
have been found to inhibit mouse CYP1A1-dependent hydroxy-
lation of benzo[a]pyrene noncompetitively, but the Ki values
(325 and 140 µM, respectively) (26) were 2 orders of magnitude
(1) Brzezinski, A.; Debi, A. Phytoestrogens: the “natural” selective
estrogen receptor modulators? Eur. J. Ob. Gyn. Rep. Biol. 1999,
85, 47-51.
(2) Adlercreutz, C. H.; Goldin, B. R.; Gorbach, S. L.; Hockerstedt,
K. A.; Watanabe, S.; Hamalainen, E. K.; Markkanen, M. H.;

Inhibition of CYP1A1 and CYP1B1 by Isoflavone Metabolism
J. Agric. Food Chem., Vol. 52, No. 21, 2004 6631
Makela, T. H.; Wahala, K. T.; Adlercreutz, T. Soybean phyto-
estrogen intake and cancer risk. J. Nutr. 1995, 125, 757S-770S.
(3) Dornstauder, E.; Jisa, E.; Unterrieder, I.; Krenn, L.; Kubelka,
W.; Jungbauer, A. Estrogenic activity of two standardized red
clover extracts (Menoflavon) intended for large scale use in
hormone replacement therapy. J. Steroid Biochem. Mol. Biol.
2001, 78, 67-75.
(4) Tice, J. A.; Ettinger, B.; Ensrud, K.; Wallace, R.; Blackwell,
T.; Cummings, S. R. Phytoestrogen supplements for the treatment
of hot flashes: the Isoflavone Clover Extract (ICE) Study: a
randomized controlled trial. JAMA 2003, 290, 207-214.
(5) Burdette, J. E.; Liu, J.; Lantvit, D.; Lim, E.; Booth, N.; Bhat, K.
P.; Hedayat, S.; Van Breemen, R. B.; Constantinou, A. I.;
Pezzuto, J. M.; Farnsworth, N. R.; Bolton, J. L. Trifolium
pratense (red clover) exhibits estrogenic effects in vivo in
ovariectomized Sprague-Dawley rats. J. Nutr. 2002, 132, 27-
30.
(6) Kulling, S. E.; Honig, D. M.; Metzler, M. Oxidative metabolism
of the soy isoflavones daidzein and genistein in humans in vitro
and in vivo. J. Agric. Food Chem. 2001, 49, 3024-3033.
(7) Kulling, S. E.; Honig, D. M.; Simat, T. J.; Metzler, M. Oxidative
in vitro metabolism of the soy phytoestrogens daidzein and
genistein. J. Agric. Food Chem. 2000, 48, 4963-4972.
(8) Kulling, S. E.; Lehmann, L.; Metzler, M. Oxidative metabolism
and genotoxic potential of major isoflavone phytoestrogens. J
Chromatogr. B Anal. Technol. Biomed. Life Sci. 2002, 777, 211-
218.
(9) Tolleson, W. H.; Doerge, D. R.; Churchwell, M. I.; Marques,
M. M.; Roberts, D. W. Metabolism of biochanin A and
formononetin by human liver microsomes in vitro. J. Agric. Food
Chem. 2002, 50, 4783-4790.
(20) Zhu, B. T.; Conney, A. H. Functional role of estrogen metabolism
in target cells: review and perspectives. Carcinogenesis 1998,
19, 1-27.
(21) Jefcoate, C. R.; Liehr, J. G.; Santen, R. J.; Sutter, T. R.; Yager,
J. D.; Yue, W.; Santner, S. J.; Tekmal, R.; Demers, L.; Pauley,
R.; Naftolin, F.; Mor, G.; Berstein, L. Tissue-specific synthesis
and oxidative metabolism of estrogens. J. Natl. Cancer Inst.
Monogr. 2000, 27, 95-112.
(22) Franke, A. A.; Custer, L. J.; Cerna, C. M.; Narala, K. Rapid
HPLC analysis of dietary phytoestrogens from legumes and from
human urine. Proc. Soc. Exp. Biol. Med. 1995, 208, 18-26.
(23) Duncan, A. M.; Underhill, K. E.; Xu, X.; Lavalleur, J.; Phipps,
W. R.; Kurzer, M. S. Modest hormonal effects of soy isoflavones
in postmenopausal women. J. Clin. Endocrinol. Metab. 1999,
84, 3479-3484.
(24) Chan, H. Y.; Leung, L. K. A potential protective mechanism of
soya isoflavones against 7,12-dimethylbenz[a]anthracene tumour
initiation. Br. J. Nutr. 2003, 90, 457-465.
(25) Chan, H. Y.; Wang, H.; Leung, L. K. The red clover (Trifolium
pratense) isoflavone biochanin A modulates the biotransforma-
tion pathways of 7,12-dimethylbenz[a]anthracene. Br. J. Nutr.
2003, 90, 87-92.
(26) Shertzer, H. G.; Puga, A.; Chang, C.; Smith, P.; Nebert, D. W.;
Setchell, K. D.; Dalton, T. P. Inhibition of CYP1A1 enzyme
activity in mouse hepatoma cell culture by soybean isoflavones.
Chem. Biol. Interact. 1999, 123, 31-49.
(27) Helsby, N. A.; Chipman, J. K.; Gescher, A.; Kerr, D. Inhibition
of mouse and human CYP 1A- and 2E1-dependent substrate
metabolism by the isoflavonoids genistein and equol. Food
Chem. Toxicol. 1998, 36, 375-382.
(28) Luu-The, V. Analysis and characteristics of multiple types of
human 17â-hydroxysteroid dehydrogenase. J. Steroid Biochem.
Mol. Biol. 2001, 76, 143-151.
(10) Setchell, K. D. R.; Brown, N. M.; Desai, P.; Zimmer-Nechemias,
L.; Wolfe, B. E.; Brashear, W. T.; Kirschner, A. S.; Cassidy,
A.; Heubi, J. E. Bioavailability of pure isoflavones in healthy
humans and analysis of commercial soy isoflavone supplements.
J. Nutr. 2001, 131, 1362S-1375S.
(29) Le Lain, R.; Nicholls, P. J.; Smith, H. J.; Maharlouie, F. H.
Inhibitors of human and rat testes microsomal 17â-hydroxys-
teroid dehydrogenase (17â-HSD) as potential agents for prostatic
cancer. J. Enzyme Inhib. 2001, 16, 35-45.
(30) Krazeisen, A.; Breitling, R.; Moller, G.; Adamski, J. Phytoestro-
gens inhibit human 17â-hydroxysteroid dehydrogenase type 5.
Mol. Cell. Endocrionol. 2001, 171, 151-162.
(11) King, R. A.; Broadbent, J. L.; Head, R. J. Absorption and
excretion of the soy isoflavone genistein in rats. J. Nutr. 1996,
126, 176-182.
(31) Evans, B. A.; Griffiths, K.; Morton, M. S. Inhibition of
5R-reductase in genital skin fibroblasts and prostate tissue by
dietary lignans and isoflavonoids. J. Endocrinol. 1995, 147, 295-
302.
(32) Ohno, S.; Shinoda, S.; Toyoshima, S.; Nakazawa, H.; Makino,
T.; Nakajin, S. Effects of flavonoid phytochemicals on cortisol
production and on activities of steroidogenic enzymes in human
adrenocortical H295R cells. J. Steroid Biochem. Mol. Biol. 2002,
80, 355-363.
(33) Xu, X.; Duncan, A. M.; Merz, B. E.; Kurzer, M. S. Effects of
soy isoflavones on estrogen and phytoestrogen metabolism in
premenopausal women. Cancer Epidemiol. Biomarkers PreV.
1998, 7, 1101-1108.
(34) Lavigne, J. A.; Goodman, J. E.; Fonong, T.; Odwin, S.; He, P.;
Roberts, D. W.; Yager, J. D. The effects of catechol-O-
methyltransferase inhibition on estrogen metabolite and oxidative
DNA damage levels in estradiol-treated MCF-7 cells. Cancer
Res. 2001, 61, 7488-7494.
(35) Thompson, P. A.; Ambrosone, C. Molecular epidemiology of
genetic polymorphisms in estrogen metabolizing enzymes in
human breast cancer. J. Natl. Cancer Inst. Monogr. 2000, 27,
125-134.
(36) Bolton, J. L.; Shen, L. p-Quinone methides are the major
decomposition products of catechol estrogen o-quinones. Car-
cinogenesis 1996, 17, 925-929.
(37) Chang, M. S.; Zhang, F. G.; Shen, L.; Pauss, N.; Alam, I.; van
Breemen, R. B.; Blond, S. Y.; Bolton, J. L. Inhibition of
glutathione S-transferase activity by the quinoid metabolites of
equine estrogens. Chem. Res. Toxicol. 1998, 11, 758-765.
(12) Setchell, K. D. Soy isoflavonessbenefits and risks from nature’s
selective estrogen receptor modulators (SERMs). J. Am. Coll.
Nutr. 2001, 20, 354S-362S.
(13) Zubik, L.; Meydani, M. Bioavailability of soybean isoflavones
from aglycone and glucoside forms in American women. Am.
J. Clin. Nutr. 2003, 77, 1459-1465.
(14) Sutter, T. R.; Tang, Y. M.; Hayes, C. L.; Wo, Y. Y.; Jabs, E.
W.; Li, X.; Yin, H.; Cody, C. W.; Greenlee, W. F. Complete
cDNA sequence of a human dioxin-inducible mRNA identifies
a new gene subfamily of cytochrome P450 that maps to
chromosome 2. J. Biol. Chem. 1994, 269, 13092-13099.
(15) Wacher, V. J.; Salphati, L.; Benet, L. Z. Active secretion and
enterocytic drug metabolism barriers to drug absorption. AdV.
Drug DeliVery ReV. 2001, 46, 89-102.
(16) Obach, R. S.; Zhang, Q. Y.; Dunbar, D.; Kaminsky, L. S.
Metabolic characterization of the major human small intestinal
cytochrome P450s. Drug Metab. Dispos. 2001, 29, 347-352.
(17) Zhang, Q. Y.; Dunbar, D.; Ostrowska, A.; Zeisloft, S.; Yang,
J.; Kaminsky, L. S. Characterization of human small intestinal
cytochromes P-450. Drug Metab. Dispos. 1999, 27, 804-809.
(18) Hu, M.; Krausz, K.; Chen, J.; Ge, X.; Li, J.; Gelboin, H. L.;
Gonzalez, F. J. Identification of CYP1A2 as the main isoform
for the phase I hydroxylated metabolism of genistein and a
prodrug converting enzyme of methylated isoflavones. Drug
Metab. Dispos. 2003, 31, 924-931.
(19) Doostdar, H.; Burke, M. D.; Mayer, R. T. Bioflavonoids:
selective substrates and inhibitors for cytochrome P450 CYP1A
and CYP1B1. Toxicology 2000, 144, 31-38.

6632 J. Agric. Food Chem., Vol. 52, No. 21, 2004
Roberts et al.
(38) Cavalieri, E.; Frenkel, K.; Liehr, J. G.; Rogan, E.; Roy, D.
Estrogens as endogenous genotoxic agentssDNA adducts
and mutations. J. Natl. Cancer Inst. Monogr. 2000, 27, 75-
93.
(39) Cavalieri, E. L.; Rogan, E. G.; Chakravarti, D. Initiation of cancer
and other diseases by catechol ortho-quinones: a unifying
mechanism. Cell. Mol. Life Sci. 2002, 59, 665-681.
Received for review April 9, 2004. Accepted August 6, 2004. This work
was supported in part by separate awards to W.H.T. and D.W.R.
provided by the U.S. Food and Drug Administration’s Office of
Women’s Health. G.G.C. thanks Fundac¸a˜o para a Cieˆncia e a Tecno-
logia (FCT), Portugal, for a postdoctoral fellowship.
JF049418X