Oxidation of Flavone, 5-Hydroxyflavone, and 5,7-Dihydroxyflavone to Mono-, Di-, and Tri-Hydroxyflavones by Human Cytochrome P450 Enzymes

Article abstract of DOI:10.1021/acs.chemrestox.9b00078

Biologically active plant flavonoids, including 5,7-dihydroxyflavone (57diOHF, chrysin), 4′,5,7-trihydroxyflavone (4′57triOHF, apigenin), and 5,6,7-trihydroxyflavone (567triOHF, baicalein), have important pharmacological and toxicological significance, e.g., antiallergic, anti-inflammatory, antioxidative, antimicrobial, and antitumorgenic properties. In order to better understand the metabolism of these flavonoids in humans, we examined the oxidation of flavone, 5-hydroxyflavone (5OHF), and 57diOHF to various products by human cytochrome P450 (P450 or CYP) and liver microsomal enzymes. Individual human P450s and liver microsomes oxidized flavone to 6-hydroxyflavone, small amounts of 5OHF, and 11 other monohydroxylated products at different rates and also produced several dihydroxylated products (including 57diOHF and 7,8-dihydroxyflavone) from flavone. We also found that 5OHF was oxidized by several P450 enzymes and human liver microsomes to 57diOHF and further to 567triOHF, but the turnover rates in these reactions were low. Interestingly, both CYP1B1.1 and 1B1.3 converted 57diOHF to 567triOHF at turnover rates (on the basis of P450 contents) of >3.0 min^-1, and CYP1A1 and 1A2 produced 567triOHF at rates of 0.51 and 0.72 min^-1, respectively. CYP2A13 and 2A6 catalyzed the oxidation of 57diOHF to 4′57triOHF at rates of 0.7 and 0.1 min^-1, respectively. Our present results show that different P450s have individual roles in oxidizing these phytochemical flavonoids and that these reactions may cause changes in their biological and toxicological properties in mammals.

Full text of DOI:10.1021/acs.chemrestox.9b00078

Subscriber access provided by Bibliothèque de l'Université Paris-Sud
Article
Oxidation of Flavone, 5-Hydroxyflavone, and 5,7-Dihydroxyflavone to
Mono-, Di-, and Tri-Hydroxyflavones by Human Cytochrome P450 Enzymes
Haruna Nagayoshi, Norie Murayama, Kensaku Kakimoto, Masaki Tsujino,
Shigeo Takenaka, Jun Katahira, Young-Ran Lim, Donghak Kim, Hiroshi
Yamazaki, Masayuki Komori, F. Peter Guengerich, and Tsutomu Shimada
Chem. Res. Toxicol., Just Accepted Manuscript • DOI: 10.1021/acs.chemrestox.9b00078 • Publication Date (Web): 09 Apr 2019
Downloaded from http://pubs.acs.org on April 10, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 54
Chemical Research in Toxicology
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Oxidation of Flavone, 5-Hydroxyflavone, and 5,7-Dihydroxyflavone to Mono-, Di-,
and Tri-Hydroxyflavones by Human Cytochrome P450 Enzymes
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
†
¶
†
†
‡
Haruna Nagayoshi, Norie Murayama, Kensaku Kakimoto, Masaki Tsujino, Shigeo Takenaka, Jun
§
#
#
¶
§
Katahira, Young-Ran Lim, Donghak Kim, Hiroshi Yamazaki, Masayuki Komori, F. Peter
£
§,*
Guengerich, and Tsutomu Shimada
§Laboratory of Cellular and Molecular Biology, Veterinary Sciences, Osaka Prefecture University, 1-
8 Rinku-Orai-Kita, Izumisano, Osaka 598-8531, Japan
5
†Osaka Institute of Public Health, 1-3-69 Nakamichi, Higashinari-ku, Osaka 537-0025, Japan
¶Laboratory of Drug Metabolism and Pharmacokinetics, Showa Pharmaceutical University, Machida,
Tokyo 194-8543, Japan
‡Graduate School of Comprehensive Rehabilitation, Osaka Prefecture University, 3-7-30, Habikino,
Osaka 583-8555, Japan
#Department of Biological Sciences, Konkuk University, Seoul 05029, Korea
£Department of Biochemistry, Vanderbilt University School of Medicine, Nashville, Tennessee
37232-0146, USA
Address correspondence to: Tsutomu Shimada, Ph.D., Laboratory of Cellular and Molecular Biology,
Veterinary Sciences, Osaka Prefecture University, 1-58 Rinku-Orai-Kita, Izumisano, Osaka 598-8531,
Japan. E-mail: t.shimada@vet.osakafu-u.ac.jp
1
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 2 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table of Contents Graphics
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
ACS Paragon Plus Environment

Page 3 of 54
Chemical Research in Toxicology
1
2
3
4
ABSTRACT: Biologically active plant flavonoids, including 5,7-dihydroxyflavone (57diOHF,
chrysin), 4´,5,7-trihydroxyflavone (4´57triOHF, apigenin), and 5,6,7-trihydroxyflavone (567triOHF,
baicalein), have important pharmacological and toxicological significance, e.g. anti-allergic, anti-
inflammatory, anti-oxidative, anti-microbial, and anti-tumorgenic properties. In order to better
understand the metabolism of these flavonoids in humans, we examined the oxidation of flavone, 5-
hydroxyflavone (5OHF), and 57diOHF to various products by human cytochrome P450 (P450 or CYP)
and liver microsomal enzymes. Individual human P450s and liver microsomes oxidized flavone to 6-
hydroxyflavone, small amounts of 5OHF, and 11 other mono-hydroxylated products at different rates
and also produced several di-hydroxylated products (including 57diOHF and 7,8-dihydroxyflavone)
from flavone. We also found that 5OHF was oxidized by several P450 enzymes and human liver
microsomes to 57diOHF and further to 567triOHF, but the turnover rates in these reactions were low.
Interestingly, both CYP1B1.1 and 1B1.3 converted 57diOHF to 567triOHF at turnover rates (on basis
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-1
of P450 contents) of >3.0 min , and CYP1A1 and 1A2 produced 567triOHF at rates of 0.51 and 0.72
-1
min , respectively. CYP2A13 and 2A6 catalyzed the oxidation of 57diOHF to 4´57triOHF at rates of
-1
0
.7 and 0.1 min , respectively. Our present results show that different P450s have individual roles in
oxidizing these phytochemical flavonoids and that these reactions may cause changes in their
biological and toxicological properties in mammals.
3
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 4 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
INTRODUCTION
Flavonoids constitute a set of over 5,000 structurally different natural chemicals in the
environment, and many of these flavonoids have been shown to cause various biological benefits in
preventing a number of diseases including cancer, heart disorders, and bone loss in humans.1-5 The
biological properties of flavonoids also include toxicities per se. A search of the literature (PubMed,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
April 2019) showed 10,590 items in a search “toxicity of flavonoids.” Further, the search “toxicity
of quercetin” yielded 1,789 entries, and much remains to be known about the toxicity of many
flavonoids and the role of metabolism in changing the biological properties. The flavonoids 5,7-
dihydroxyflavone (57diOHF, chrysin), 4´5,7-trihydroxyflavone (4´57triOHF, apigenin), and 5,6,7-
trihydroxyflavone (567triOHF, baicalein) have been reported to show significant anti-allergic, anti-
inflammatory, anti-oxidative, anti-microbial, anti-tumorgenic, and anti-mutagenic properties in
mammals, including humans.6-22 Collectively, these three flavonoids yielded 593 hits in the PubMed
searches (“toxicity of [each flavonoid],” 1 April 2019) and are the subject of the present investigation.
Recent studies have shown that cytochrome P450 (P450 or CYP) Family 1 enzymes in
human breast cancer MDA-MB-468 cell lines are involved in the activation of flavonoids,
eupatorine, diosmetin, and nobiletin, to more active metabolites, e.g. cirisiliol and luteolin,23-29
which have been reported to show various antitumor activities in ovarian cancer30,31 and gastric
4
ACS Paragon Plus Environment

Page 5 of 54
Chemical Research in Toxicology
1
2
3
4
3
2
33
34
cancer cells, and to show suppression of metastasis in breast cancer and malignant melanoma.
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Such activation of these flavonoids was not seen in normal mammary MCF-10A cells, which do not
express CYP1A1 and 1B1, indicating that these P450s have critical roles in the activation of these
flavonoids.27,29,23,35
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We recently reported that flavone is oxidized by different forms of human P450 enzymes to
3
6
several mono-hydroxy (mono-OH) and further to dihydroxy (di-OH) products and that CYP2A6
has specific roles in oxidizing flavone to yet unidentified mono-OH product 6 (OHFM6) and
3
7
OHFM9 and in oxidizing flavanone to 2´-hydroxyflavanone. During course of these studies, we
noted that flavone is oxidized to 5-hydroxyflavone (5OHF) and 5,7-dihydroxyflavone (57diOHF)
by different human P450s and liver microsomes, although the rates of formation of two products by
these enzymes were not very high.36,37
In this study, we further examined roles of human P450s and liver microsomes to oxidize
flavone, 5OHF, and 57diOHF to form respective mono-OH, di-OH, and tri-OH products by
analyzing product formation using LC-MS/MS. Purified human P450 enzymes used included
CYP1A1, 1A2, 1B1.1, 1B1.3, 2A6, 2C9, and 3A4, and we also used recombinant CYP2B6
expressed in microsomes of Trichoplusia ni cells in which human NADPH-P450 reductase was co-
expressed. Among liver microsomes prepared from human samples HH2, HH47, and HH54, the
5
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 6 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
HH2 sample was defective in coumarin 7-hydroxylation activity, thus categorized to be a poor
3
7
metabolizer of CYP2A6. Molecular docking analysis of the interaction of 57diOHF with different
forms of human P450s, particularly CYP1B1.1 and 1B1.3, were studied to examine the basis of
catalytic differences found in these P450s in oxidizing these flavonoids.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
EXPERIMENTAL PROCEDURES
Chemicals. Flavone, 5-, 6-, and 7-hydroxyflavone (5OHF, 6OHF, and 7OHF), and 5,7- and
7
,8-dihydroxyflavone (57diOHF and 78diOHF, respectively) were obtained from Sigma-Aldrich (St.
Louis, MO, USA) and Wako Pure Chemicals (Osaka, Japan). 4´,5,7-Trihydroxyflavone (apigenin;
4
´57triOHF) and 5,6,7-trihydroxyflavone (baicalein; 567triOHF) were also purchased from Sigma-
Aldrich. Other chemicals and reagents used in this study were obtained from sources described
previously or were of the highest quality commercially available. 36-38
Enzymes. Purified preparations of wild type of human CYP1A1, CYP1A2, CYP1B1,
CYP2A6, CYP2A13, CYP2C9, and CYP3A4 expressed in Escherichia coli were obtained by
methods described previously.38,39,40,41 In this study, we were also used variant form CYP1B1.3 and
compared it with (wild-type) CYP1B1.1 for their catalytic differences. NADPH-P450 reductase and
cytochrome b (b ) were also purified from membranes of recombinant E. coli as described
5
5
6
ACS Paragon Plus Environment

Page 7 of 54
Chemical Research in Toxicology
1
2
3
4
elsewhere.39,40 Recombinant CYP2B6 expressed in microsomes of T. ni cells, infected with
baculovirus containing CYP2B6 and NADPH-P450 reductase cDNA inserts, was obtained from
Corning-GENTEST (Woburn, MA).
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Liver microsomes prepared from human samples HH2 (Cat No. 452002), HH47 (Cat No.
4
52047), HH54 (Cat No. 452054), and HG95 (Cat No. 452095) were obtained from Corning-
GENTEST. The data sheets provided by the manufacturer indicated that these microsomes contained
0
.19, 0.26, 0.35, and 0.25 nmol P450/mg protein, respectively. The oxidation activities towards
typical P450 substrates such as (S)-mephenytoin, paclitaxel, diclofenac, bufuralol, chlorzoxazone,
and testosterone were reported in these liver microsomes, except that HH2 had no detectable
coumarin 7-hydroxylation activity and the data sheets indicated that this individual was classified a
poor metabolizer (PM) for CYP2A6 activity.37
Oxidation of Flavone, 5OHF, and 57diOHF by P450 Enzymes by Human Liver
Microsomes. Oxidation of flavone, 5-hydroxyflavone (5OHF), and 5,7-dihydroxyflavone
(
57diOHF) by P450 enzymes was determined by the methods described previously.36,37 Briefly, liver
microsomes (50 pmol of P450), baculosome-derived CYP2B6 (20 pmol of P450), or reconstituted
monooxygenase systems consisting of each purified P450 (50 pmol), NADPH-P450 reductase (100
pmol), b (100 pmol, in the cases of CYP2A6, 2A13, 2C9, and 3A4 enzymes), and L--1,2 dilaouryl-
5
7
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 8 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
sn-glycero-3-phosphocholine (DLPC) (50 µg) were incubated (0.25 mL of total volume) with 60
µM flavone, 5OHF, or 57diOHF at 37 °C for 20 min, following a pre-incubation of 1 min before
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
adding an NADPH-generating system (0.5 mM NADP , 5 mM glucose 6-phosphate, and 0.5 unit
of yeast glucose 6-phosphate dehydrogenase/mL). Each reaction was terminated by the addition of
0
.5 mL of ice-cold CH CN. The mixture was mixed vigorously (with a vortex device) and
3
centrifuged at 10,000 × g for 5 min, and an aliquot of the upper CH CN layer was injected and
3
analyzed with LC-MS/MS as described.36,37
LC-MS/MS analyses were performed using an HPLC system (ACQUITY UPLC I-Class
system; Waters, Milford, MA) coupled to a tandem quadruple mass spectrometer (XevoTQ-S;
Waters) by the methods as described previously.36,37 MS/MS analysis was performed in the positive
electrospray ionization mode with a capillary voltage of 3000 V and cone voltage of 30 V as
described previously.36,37
Kinetic parameters were determined by non-linear regression analysis (mean ± standard
error) employing the Michaelis-Menten equation, v = Vmax × [S]/(K + [S]) or the allosteric Hill
m
n
n
n
equation model, v = Vmax × [S] /( S + [S] ) using the program Kaleida-Graph (Synergy Software,
50
Reading, PA, USA) or GraphPad Prism (GraphPad, La Jolla, CA, USA).
8
ACS Paragon Plus Environment

Page 9 of 54
Chemical Research in Toxicology
1
2
3
4
5
6
7
Other Assays. P450 and protein contents were determined by the methods described
previously.42,43
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Docking Simulations of Flavone, 5OHF, and 57diOHF into P450 Enzymes. Crystal
structures of CYP1A1 (Protein Data Bank 418V), 1A2 (2H14), 1B1.1 (3PMO), 2A6 (2FDV), 2A13
(
3T3S), 2B6 (3QOA), 2C9 (1R9O), and 3A4 (5TE8) have been reported and were used in this
study.44-52 CYP1B1.3 variant (L432V) model was established by incorporating the mutation into
CYP1B1.1 (3PMO). Chemical structures of 5OHF and 57diOHF were taken from PubChem (an
open chemistry database at the Untied States National Institutes of Health) and were optimized in
MOE software (ver. 2018.0101, Computing Group, Montreal, Canada). Simulations were carried
out in the MOE by the methods as described previously.36,37 Ligand-interaction energies (U values)
were obtained by use of the program ASEdock in MOE.
9
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 10 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
RESULTS
Different Roles of Human P450 Enzymes in the Oxidation of Flavone. Oxidations of
flavone to mono-OH flavones (Figure 1A-1G) and further to di-OH flavones (Figure 1H-1N) by
purified P450 enzymes (Figure 1A-1E and 1H-1L) and by human liver microsomes (Figure 1F, 1G,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
M, and 1N) were examined. As we have reported earlier,36,37 CYP2A6 had important roles in the
oxidation of flavone to form unidentified mono-OH flavones OHM6 and OHFM9 in liver
microsomes (Figure 1E) because these products were produced slowly by liver microsomes from
3
8
sample HH2, a poor metabolizer of CYP2A6, as compared with those catalyzed by HH47, an
extensive metabolizer of CYP2A6 (Figures 1F and 1G). Our present studies also showed that
OHFM10 was produced by CYP2A6 in liver microsomes, because formation of this product was
slowly catalyzed by microsomes from liver sample HH2 (Figures 1M and 1N).
CYP1B1.3 was shown to be unique in oxidizing flavone to form mono-OH flavones
OHFM1, M2. M3, and M4; these products were not produced by other human P450s and liver
microsomes (Figure 1D). It was also found that OHFM11 was not formed by CYP1B1.3 but that
formation was catalyzed by CYP1B1.1 and other human P450s and liver microsomes.
CYP1A2 was more active than CYP2A6 in catalyzing the oxidation of flavone to 6OHF
1
0
ACS Paragon Plus Environment

Page 11 of 54
Chemical Research in Toxicology
1
2
3
4
and both liver microsomal HH2 and HH47 samples were found to catalyze formation of 6OHF at
higher rates (on a P450 basis) than CYP2A6 did, indicating that CYP2A6 does not have a major role
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
in this reaction (Figure 1). We also found that 5OHF (t 11.2 min) was produced from flavone by
R
-1
CYP2A6 and human liver microsomes, but turnover rates were low (< 0.01 min ).
Human P450 enzymes and liver microsomes were able to form di-OH flavones, as well as
mono-OH flavones, when flavone was used as a substrate (Figure 1H-1N). Major di-OH products
were diOHFM3, diOHFM4, and diOHFM4 with CYP1A2 and CYP2A6 and liver microsomes
(Figure 1I, 1L, 1M, and 1N). CYP1A1 and 1B1.1 produced diOHFM5 and formed small amounts
of diOHFM3 and diOHFM4. CYP1B1.3 did not produce di-OH flavones at significant levels (Figure
1
7
K). An interesting finding was that CYP2A6 and human liver microsomes produced 57diOHF (tR
.9 min) and that the latter enzymes formed 78diOHF (t 4.1 min); these di-OH products formed
R
were characterized by co-chromatography with standards (vide infra).
LC-MS/MS chromatograms of mono-OH products produced following incubation of
flavone with HH2 and HH54 microsomes showed that 5OHF (t 11.2 min) (as well as 6OHF) was
R
detected when examined using the transition m/z 239>137 (but not m/z 239>121) (Figure 2C and
2
D). The product ion spectra of 5OHF and 6OHF produced from liver microsomes of sample HH54
(Figure 2F) and of 6OHF from the HH2 microsomes (Figure 2H) were similar to those of standard
1
1
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 12 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
OHF (Figure 2E) and 6OHF (Figure 2G). We did not detect formation of 7OHF following
incubation of flavone with any P450s or human liver microsomal samples.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Chemical Structures of Unidentified Mono- and Di-OH Flavones. The product ion
spectra of unidentified products of OHFM1-OHFM12, as well as standards 5OHF, 6OHF, and 7OHF,
were examined with LC-MS/MS (Figure 3). By analyzing the product ion spectra of 11 unidentified
mono-OH flavones (OHFM1-OHFM12, except for OHFM8 which was a very minor product), we
were able to tentatively classify these mono-OH flavones into two groups (Figure 3). The first group
included OHFM4, OHFM5, OHFM7, and OHFM12, which showed a characteristic m/z 128.9 ion;
of the products detected, OHFM4 and M7 were also found to contain m/z 102.9 and 136.9 ions as
well as an m/z 128.9 ion, indicative of similarity to the spectra of the standard 5OHF, 6OHF, and
7
OHF products (Figure 3). OHFM5 contained m/z 102.9 and 128.9 fragments, but not 136.9, while
OHFM12 had characteristic m/z 128.9 and 136.9 fragments but not one at m/z 102.9. Among the
products determined, OHM1, OHFM6, OHFM9, and OHFM10 contained a characteristic m/z 144.8
ion and OHFM11 was found to contain an m/z 139.0 ion as well as an m/z 120.9 ion (Figure 3).
We also determined the product ion spectra of unidentified di-OH flavones following
incubation of flavone with these P450s and liver enzymes (Figure 4). Standard spectra of flavone,
5
7diOHF, and 78diOHF were included for comparison (Figure 4A, 4B, and 4D, respectively). A
1
2
ACS Paragon Plus Environment

Page 13 of 54
Chemical Research in Toxicology
1
2
3
4
product of 57diOHF produced from HH54 showed a similar spectrum to standard 57diOHF, having
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
m/z 128.9 and 152.9 ions, but dissimilar in the presence of an m/z 135.0 ion and the absence of m/z
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
46.9 (Figure 4C). The product 78diOHF, formed from HH54 microsomes, was also similar to the
standard, except that the former had a different ion, m/z 141.0, in the spectrum (Figure 4E). The
unidentified diOHFM1 and OHFM4 products, having an m/z 120.9 ion, were found to constitute one
group, and the other included diOHFM3 and diOHFM5, which contained an m/z 128.9 ion (Figure
4
1
). The diOHFM6 spectrum could be indicative of another group, having m/z 125.0 as well as m/z
02.9 ions (Figure 4J).
Oxidation of 5OHF by Human P450s and Liver Microsomes. Human CYP1A2, 2A6,
and 2A13 and liver microsomes from sample HH54 converted 5OHF into 57diOHF and further to
5
67triOHF (Figure 5). Sample HH54 was highly active in oxidizing 5OHF to 57diOHF, followed
by CYP1A2, 2A6, and 2A13, although the oxidation activities for this reaction were not very high
among these human enzymes examined (vide infra). CYP2A13 produced diOHFM6 from flavone
at low rates (Figure 5C). These P450s and microsomal sample HH54 catalyzed formation of
5
67diOHF (and triOHFM1 in CYP1A2 and 2A13) from flavone, at low rates (vide infra).
Oxidation of 57diOHF by Human P450s and Liver Microsomes. CYP1A1, 1A2, and
1
B1.1 (and also 1B1.3, results not shown) were found to be active in converting 57dOHF into
1
3
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 14 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
67triOHF; the turnover rates (on P450 basis) were higher with CYP1B1.1 and 1B1.3 than CYP1A1
and 1A2 (vide infra) when the concentrations of P450 and 57diOHF used were 0.2 µM and 60 µM,
respectively (Figure 6). CYP1A1, but not CYP1A2, produced small amounts of 4´57triOHF and
triOHFM3 (Figure 6A).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In contrast to the CYP1 enzymes, CYP2A6 and 2A13 were more active in forming 4´57triOHF
than 567triOHF from 57diOHF and the opposite was the case for liver microsomes from human
sample HH54; it should be mentioned that the peak areas obtained with the samples from CYP2A6
and 2A13 and human liver microsomes (Figure 6D-6F) were 10-fold lower than those obtained with
CYP1 enzymes at the same concentrations of P450 and 57diOHF (0.2 and 60 µM, respectively) used
for the assays (Figure 6A-6C). CYP2A6 and 2A13 and HH54 microsomes produced triOHFM3 in
small amounts. Product ion spectra were determined with these metabolites obtained on incubation
of 57diOHF with human P450 enzymes (Figure 7). The reference peak m/z 122.8 in the spectrum of
5
67triOHF observed from the incubation with CYP1B1.3 was similar to that of the
authentic standard (Figure 7A and 7B).
The 4´57triOHF product isolated from CYP2A13 (Figure 7E) had a similar spectrum to
the standard (Figure 7D), except that the latter was missing m/z 168.9. The unidentified triOHFM1
(from CYP1A2) and triOHFM3 products (from CYP1A1) were also characterized from their
1
4
ACS Paragon Plus Environment

Page 15 of 54
Chemical Research in Toxicology
1
2
3
4
spectra; the former was similar to 567triOHF and the latter was to 4´57triOHF, based on their spectra.
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Rates of Oxidation of 5OHF and 57diOHF by Human P450s and Liver Microsomes.
Oxidations of 5OHF (Figure 8A and 8B) and 57diOHF (Figure 8C and 8D) by human P450s and
liver microsomes were determined under standard reaction condition (0.2 µM P450 and 60 µM
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
OHF or 57diOHF), and rates of product formation was calculated (nmol product formed/min/nmol
P450).
The rates of oxidation of 5OHF to form 57diOHF and 567triOHF by human P450s were
-1
low, < 0.1 min , although the rates measured with human liver microsomes were higher than with
purified P450 enzymes (Figure 8A and 8B). In contrast, the oxidation of 57diOHF to 4´57triOHF
(Figure 8C) and 567triOHF (Figure 8D) was more evident with the P450 enzymes than with human
liver microsomes. CYP2A13 was most active in converting 57diOHF to 4´57triOHF, at rates of 0.7
-1
-1
min , followed by CYP2A6 (0.13 min ) (Figure 8C). Both CYP1B1.1 and 1B1.3 were most active
-1
in oxidizing 57diOHF to 567triOHF at rates of >3.0 min , followed by CYP1A1 and 1A2 at rates
-1
of 0.5 and 0.7 min , respectively (Figure 8D).
Kinetic Analysis of Oxidation of 57diOHF to 4´57triOHF and 567diOHF by Human
P450s. The K values for the formation of 4´57triOHF from 57diOHF by CYP1A1 and 1A2 were
m
1
5
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 16 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
8
5 and 97 µM and K values for the formation of 567triOHF by these P450s were 70 and 26 µM,
m
respectively, with higher k values in the latter reactions (Figure 9A, 9B, 9E, and 9F). CYP2A13
cat
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
showed higher activity in oxidizing 57diOHF to 4´57triOHF than CYP2A6, having a K value of
m
-1
1
5
4 µM and a k value of 1.0 min . CYP2A13 also had a low K value for the formation of
cat m
67triOHF, compared with CYP2A6 (K 200 µM) (Figure 9).
m
One interesting finding was that both CYP1B1 enzymes caused increases in the formation
of 567triOHF from 57diOHF with sigmoidal character at lower substrate concentrations (Figure 9x
and 9y). Homotropic cooperativity of wild-type CYP1B1.1 and variant CYP1B1.3 was seen in the
formation of 567triOHF from substrate 57diOHF (1-80 µM), with Hill coefficients of 1.3 ± 0.4 and
1
.2 ± 0.2 (means ± SEM) respectively (Figure 9X and 9Y). Moreover, the 95% confidential intervals
of Hill coefficients for CYP1B1.1 (1.9 ± 0.2) and CYP1B1.3 (5.3 ± 1.1) were 1.5-2.6 and 1.8-8.8,
respectively, in the substrate ranges of 1-16 µM.
Molecular Docking Analysis of Interaction of Flavonoids with Human P450s. In order
to gain information of molecular interaction of these P450 enzymes with flavonoids examined in
this study, we performed molecular docking analysis as described in Experimental Procedures.
Because homotropic cooperativity of CYP1B1.1 and CYP1B1.3 was evident for the oxidation of
5
7diOHF (vide supra, Figure 9), docking simulation was first performed with the substrate 57diOHF
1
6
ACS Paragon Plus Environment

Page 17 of 54
Chemical Research in Toxicology
1
2
3
4
using the CYP1B1.1 structure and a CYP1B1.3 model. Both wild type CYP1B1.1 and variant
CYP1B1.3 accommodated two 57diOHF molecules in the substrate-binding pocket (Figure 10). The
first substrate molecule was stabilized at a site far from the heme of CYP1B1.1 (U energy of -52.2
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
-1
kcal mol , Fig. 10A) by trapping the phenyl B ring and A ring of 57diOHF in interactions with Leu-
1
91 and His-216, respectively. Similarly, the first substrate molecule was also stabilized at a site far
-1
from the heme of CYP1B1.3 (U energy of -46.5 kcal mol , Figure 10B) by trapping the phenyl B
ring and A ring of 57diOHF in interactions with Arg-194 and Glu-229, respectively. The second
substrate molecule of 57diOHF was able to dock with both of the heme of CYP1B1.1 and CYP1B1.3
-1
with U energy values of -76.8 and -18.7 kcal mol , respectively.
Flavone oxidation differed in CYP1B1.1 and 1B1.3, with the latter enzyme showing
characteristic patterns in forming OHFM1-OHFM4. CYP1B1.1 did not form these products but was
highly active in producing OHFM6-M12 (Figure 1). Docking analysis of interaction of flavone with
these P450 enzymes suggested that there was a different orientation in these cases, one in which the
active site of CY1B1.1 faced the C4´ position at the B-ring, while CY1B1.3 showed interaction with
C5 of A-ring of flavone in its active site (Supporting information Figure S1). Such differences in the
two P450 enzymes were not seen in the interactions with 5OHF and 57diOHF (Supporting
Information Figure S1).
1
7
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 18 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Molecular docking analysis was done for the interaction of CYP1A1, 1A2, 2A6, and 2A13
with 5OHF (Supporting Information Figure S2A-S2D) and 57diOHF (Supporting Information
Figure S2E-S2H). The results showed that each P450 enzyme had similarities in interaction with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
OHF and 57diOHF. The A-ring of 5OHF and 57diOHF was close to the active sites of CYP1A1
and CYP2A13, but CYP1A2 and 2A6 interacted with the B-ring of 5OHF and 57diOHF in their
molecular docking analysis (Figure S2).
The interaction of the active sites of CYP2B6, 2C9, and 3A4 with 5OHF and 57diOHF
was also studied in molecular docking analysis (Supporting Information Figure S3).
1
8
ACS Paragon Plus Environment

Page 19 of 54
Chemical Research in Toxicology
1
2
3
4
DISCUSSION
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Our present results can be summarized as follows. (i) Flavone was oxidized by human
P450 and liver microsomal enzymes to several mono-OH and di-OH products, and CYP1B1.3 was
found to be unique in forming mono-OHFM1, OHFM2, OHFM3, and OHFM4 but not di-OH
products, compared with other P450s and liver microsomes. (ii) CYP2A6 was shown to catalyze the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
7
site-specific oxidation of flavone to mono-OHFM6 and OHFM9, as described previously, and we
also found that OHFM10 and diOHFM4 were produced by CYP2A6. (iii) Human liver microsomes
were able to oxidize flavone to 5OHF, 57diOHF, and 78diOHF, as well as other mono-OH and di-
OH flavones described above. (iv) 5OHF was converted by human P450 and liver microsomal
enzymes to 57diOHF and further to 567diOHF, but the rates of formation were low. (v) CYP1B1.1
and 1B1.3 were highly active in oxidizing 57diOHF to 567triOHF at rates (on a P450 basis) of >3.0
-1
min , followed by CYP1A1 and 1A2. CYP1B1.1 and 1B1.3 showed sigmoidal patterns in the
formation of 567triOHF from 57diOHF at lower substrate concentrations, indicative of homotropic
cooperativity in the reaction. (vi) Finally, CYP2A6 and 2A13 were not very active in catalyzing the
oxidation of 57diOHF to 567triOHF but were active in producing 4´57triOHF at rates of 0.7 and 0.1
-1
min , respectively.
As described above, CYP1B1.1 and 1B1.3 were found to be active in catalyzing oxidation
1
9
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 20 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
5
3
of 57diOHF to 567triOHF, followed by CYP1A1 and 1A2. Williams et al. have also reported that
CYP1A1 is able to catalyze the oxidation of 57diOHF to 567triOHF in yeast cells expressing human
CYP1A1. Although both 57diOHF (chrysin) and 567triOHF (baicalein) have been shown to have
several biological benefits (e.g. anti-proliferative, anti-microbial, anti-allergic, anti-inflammatory,
and anti-oxidative properties, 6-10,14-16,22 it is not known which flavonoid is more biologically active.
It is interesting in this connection to note recent studies showing that P450 Family 1 enzymes are
involved in the activation of flavonoids eupatorine, diosmetin, and nobiletin to more active
metabolites, e.g. cirisiliol and luteolin.23,26,28,29 Such activation could not be observed in normal
mammary MCF-10A cells which do not express CYP1A1 and 1B1,23,27,29 indicating that CYP1
family enzymes have significant roles of in the activation of flavonoids in such human cancer cells.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CYP2A13 and 2A6 catalyzed oxidation of 57diOHF to form 4´57triOF at turnover rates
-
1
of 0.7 and 0.1 min , respectively; these P450s showed formation of 567triOHF at the relatively slow
-1
rates of 0.013 and 0.019 min , respectively. Apigenin (4´57triOHF) has also been reported to be
highly active in several biological effects described above. 6,9,14,19
We obtained evidence for homotropic cooperativity in the oxidation of 57diOHF to
5
67triOHF catalyzed by purified CYP1B1.1 and 1B1.3. A number of studies have shown the
presence of both homotropic and heterotropic oxidation of substrates by several P450 enzymes,
2
0
ACS Paragon Plus Environment

Page 21 of 54
Chemical Research in Toxicology
1
2
3
4
including human and rabbit CYP3A enzymes and several other P450s.54-58 The mechanisms by
which CYP1B1 enzymes display homotropic (and also heterotropic) cooperativity in the oxidation
of 57diOHF are not known; however, our molecular docking studies showed that there were two
ligand binding sites in the interaction of 57diOHF with CYP1B1.1 and 1B1.3 active sites and it was
suggested that binding of the first ligand might affect the activation of the second substrate binding
to cause increases in the oxidation of 57diOHF by these P450 enzymes. Molecular docking results
also showed differences in the orientation of flavone in the active sites of CYP1B1.1 and 1B1.3
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(Supporting Information Figure S1). These P450s showed different patterns of oxidation of flavone
to form mono-OH and di-OH products; the former P450 catalyzed oxidation of flavone essentially
similar patterns to form these oxidative metabolites while CYP1B1.3 produced unique products such
as OHFM1-M4. Such products were not detected in other P450s and liver microsomes (Figure 1).
More studies will be needed to determine if these molecular differences in the CYP1B1 variants
affect their catalytic differences in flavone oxidation activities.
Some assignments of mono-, di-, and tri-OH flavone structures to the products obtained
in the present study could be made based on the literature on mass spectrometry of flavones (the
amounts of products formed were too low for NMR analysis). All assignments were made on the
basis of positive ion mass spectra.59-61 The preferred fragmentation is through the C ring, both with
2
1
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 22 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
flavones and 3-hydroxyflavones (Figure 10), and accordingly hydroxylations on the A and B rings
can be distinguished from each other.60
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We detected 11 unidentified mono-OF products on incubation of flavone with human
P450 and liver microsomal enzymes (Figure 1). On the basis of the mass spectral literature,60
OHFM11 was assigned as 3-OH flavone. The m/z 121 and 93 fragments are indicative of a non-A-
ring hydroxylation (i.e., leaving M1, M2, M3, M6, 9, M10, and M11), and M11 has no m/z 145 ion
(ruling at M1, M6, M9, and M10), but does have a strong m/z 115 ion, as expected for 3-OH
6
0
flavone. Mono-OH flavones M4, M5, M7, and M12 resemble standard 5-OH-, 6-OH-, and 7-OH
flavones in showing characteristic m/z 137, 129, and 103 peaks (Figure 3A-3G). These patterns are
6
0
indicative of hydroxylations on the A ring. The products M1, M2, M3, M6, M9, and M10 showed
6
0
strong m/z 121, 93, and 91 peaks, indicative of B ring hydroxylation. However, we were unable to
define specific sites of hydroxylation other than with 5-, 6-, and 7-OH flavone (for which standards
were available). The enzymes formed 5- and 6-OH flavone, but not 7-OH flavone.36,37 Thus, we
have at least six A-ring hydroxy products (one of which could be 8-OH flavone) and six B-ring
hydroxy products. The numbers of products raise a dilemma in that each ring only has three
possibilities (because 7-hydroxylation on the A ring was not observed). Flavone and its mono-
hydroxylated products are not chiral, so the explanation cannot be stereochemistry.
2
2
ACS Paragon Plus Environment

Page 23 of 54
Chemical Research in Toxicology
1
2
3
4
5
6
7
We detected 57diOHF and 78diOHF and five unidentified di-OH flavones on incubation
of flavone with human P450 and liver microsomal enzymes (Figure 1). The 57diOHF and diOHFM6
were also detected on incubation of 5OHF with these P450 enzymes and the minor product
diOHFM6 could be derived from 5OHF (Figure 4). The mass spectrum of diOHFM6 was very
similar to that of 57diOHF in terms of the fragments (Figure 4) and only the A ring should be
hydroxylated, suggesting that this might be either 5,6- or 5,8-dihydroxyflavone (known as rimetin).
The other four unidentified di-OH flavones were assigned based on comparison of their mass
spectral fragmentation patterns with those in the literature. Hydroxyl groups in the A ring show
strong m/z 129 and 137 fragments (Figures 2, 3, vide infra) and hydroxyl groups in the B ring show
strong m/z 103 and 121 ions. On the basis of comparisons, we assigned di-OHFM1 as 3´,4´-diOH
flavone (NIST 1207822), di-OHFM3 as 4´,5-diOH flavone (NIST 1207763), di-OHFM4 as 3,4´-di-
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
OH
flavone59
and
di-OHFM5
as
6,4´-di-OH
flavone
(PubChem
NP_C9_15_p5_E04_iTree_POS_05).
Two of the tri-OH products (Figure 6) were readily identified by their co-chromatography
and the similarity of their fragmentation with commercial standards, i.e. 567triOHF (baicalein) and
4
5
´57triOHF (apigenin). TriOHFM1 (formed with CYP1A2) had a mass spectrum similar to
67triOHF (Figure 7), dominated by m/z 123, apparently the remnant of the tri-hydroxy A ring
2
3
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 24 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
Figure 6). This was formed from 57diOHF, so it can only be 5, 7, 8-tri-OH flavone. The other
product (triOHFM3) fragmented in a very similar manner to 4´57triOHF (Figure 7), dominated by
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
m/z 153, presumably MH -118. The pattern indicates that the hydroxylation is on the B ring. It
cannot be at the 4´ position, because that compound is known and separates (Figure 6). The 2´ and
3
´ sites are possible. The spectrum resembles that of 5,7,2´-triOH flavone (NIST 1210448),
dominated by m/z 153. We prefer this assignment but cannot rule out the possibility of 5,7,3´-tri-OH
flavone, in that a mass spectrum of this has not been published nor is otherwise available.
An explanation is provided in that there is the potential for P450s to induce regiochemical
changes during the oxidation process. A precedent already exists in the isomerization of flavonone
to isoflavone, catalyzed by both plant62-65 and rat and human P450s. A mechanism for the P450-
catalyzed stereo isomerization of the drug (4-hydroxy) tamoxifen has already been proposed 67 and
a similar mechanism, based on that outlined by Kagawa et al. 66 is shown in Fig. 11.
66
We are currently investigating to verify this hypothesis that human P450s could convert
flavone to isoflavone as shown in Figure 12. We propose that the mono-OH flavone products (Figure
1
D-1M) are mixtures of flavones and isoflavones, which have very similar UV and mass spectra.
We cannot discern, with the data available, whether isomerization occurred before or after each
hydroxylation.
2
4
ACS Paragon Plus Environment

Page 25 of 54
Chemical Research in Toxicology
1
2
3
4
5
6
7
In conclusion, the present studies showed that flavone, 5OHF, and 57diOHF were
converted to mono-OH, di-OH, and tri-OH flavones, respectively, by human P450 and liver
microsomal enzymes and that individual human P450s had different roles in oxidizing these
flavonoids. Human P450 and liver microsomal enzymes produced various oxidized flavones,
including 6OHF and 5OHF and at least 11 unidentified products whose chemical structures are
proposed in this study by the analysis of product ion spectra (mass fragmentation) and comparison
with previously reported data. Formation of 5OHF from flavone by P450s and liver microsomes was
not very rapid, but the turnover rates of formation of 567triOHF and 4’57triOHF from 57diOHF
were relatively high with CYP Family 1 enzymes and CYP2A13 and 2A6, respectively, and these
phenomena may be relevant in the metabolic activation of numerous flavonoids into biologically
active products with beneficial properties in humans, e.g. prevention of cancer due to anti-
proliferative, anti-oxidant, and anti-inflammatory effects.
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
5
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 26 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ASSOCIATED CONTENT
SUPPORTING INFORMATION
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Molecular docking analysis of interaction of flavone, 5OHF, and 57diOHF with CYP1B1.1,
CYP1B1.3, CYP1A1, CYP1A2, CYP2A6, CYP2A13, CYP2B6, CYP2C9, and CYP3A4.
AUTHOR INFORMATION
Corresponding authors
*
(T.S.) Telephone: 72-463-5326; Fax: 72-463-5326
E-mail: t.shimada@vet.osakafu-u.ac.jp.
*
(H.Y.) Telephone: 42-721-1406; Fax: 42-721-1406
E-mail: hyamazak@ac.shoyaku.ac.jp
*
(F.P.G.) Telephone: 615-322-2261; Fax: 615-322-4349
E.-mail: f.guengerich@vanderbilt.edu
2
6
ACS Paragon Plus Environment

Page 27 of 54
Chemical Research in Toxicology
1
2
3
4
Funding Sources
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
This study was supported in part by JSPS KAKENHI [16K21710] (to H. N), [16K09119] (to K. K.),
[15K07770] (to S. T.), [17K08630] (to J. K.), [JP18K0600 ] (to M. K.), [17K08426] (to N. M.), and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
[17K08425] (to H. Y.), National Research Foundation of Korea [NRF-2016R1D1A1B03932002] (to
D. K.), and United States Public Health Service grant [R01 GM118122] (to F. P. G.). The content is
solely the responsibility of the authors and does not necessarily represent the official views of the
National Institutes of Health.
Notes
The authors declare no competing financial interests.
ABBREVIATIONS: P450 or CYP, cytochrome P450; b , cytochrome b ; DLPC, L-α-1,2 dilaouryl-
5
5
sn-glycero-3-phosphocholine; mono-OH, di-OH, and tri-OH flavone, general term for mono-, di-, and
tri-hydroxy flavone, respectively; 5OHF, 6OHF, 7OHF, 57diOHF, 78diOHF, 4’57triOHF, and
5
5
5
67triOHF, individual form for 5-hydroxyflavone (primuletin), 6-hydroxyflavone, 7-hydroxyflavone,
,7-dihydroxyflavone (chrysin), 7,8-dihydroxyflavone, 4´,5,7-trihydroxyflavone (apigenin), and
,6,7-trihydroxyflavone (baicalein), respectively. OHFM1-OHFM12, diOHFM1-diOHFM6,
triOHFM1 and triOHF M3, individual terms for unidentified products of mono-, di-, and triOH
flavones.
2
7
ACS Paragon Plus Environment

Chemical Research in Toxicology
Page 28 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
REFERENCES
(1) Zhang, S., Yang, X., Coburn, R. A., and Morris, M. E. (2005) Structure activity relationships
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
and quantitative structure activity relationships for the flavonoid-mediated inhibition of breast cancer
resistance protein. Biochem. Pharmacol. 70, 627-639.
(2) Walle, T., Ta, N., Kawamori, T., Wen, X., Tsuji, P. A., and Walle, U. K. (2007) Cancer
chemopreventive properties of orally bioavailable flavonoids—methylated versus unmethylated
flavones. Biochem. Pharmacol. 73, 1288-1296.
(3) Arct, J. and Pytkowska, K. (2008) Flavonoids as components of biologically active
cosmeceuticals. Clin. Dermatol. 26, 347-357.
(4) Kale, A., Gawande, S., and Kotwal, S. (2008) Cancer phytotherapeutics: Role for flavonoids
at the cellular level. Phytother Res. 22, 567-577.
(5) Hostetler, G. L., Ralston, R. A., and Schwartz, S. J. (2017) Flavones: Food sources,
bioavailability, metabolism, and bioactivity. Adv. Nutr. 8, 423-435.
(6) Patel, D., Shukla, S., and Gupta, S. (2007) Apigenin and cancer chemoprevention: Progress,
potential and promise (review). Int. J. Oncol. 30, 233-245.
(7) Li, Y. W., Xu, J., Zhu, G. Y., Huang, Z. J., Lu, Y., Li, X. Q., Wang, N., and Zhang, F. X.
(2018) Apigenin suppresses the stem cell-like properties of triple-negative breast cancer cells by
2
8
ACS Paragon Plus Environment

Page 29 of 54
Chemical Research in Toxicology
1
2
3
4
inhibiting YAP/TAZ activity. Cell Death Discov. 4, 105-14.
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
8) Shukla, S. and Gupta, S. (2010) Apigenin: a promising molecule for cancer prevention.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Pharm. Res. 27, 962-978.
(9) Clere, N., Faure, S., Martinez, M. C., and Andriantsitohaina, R. (2011) Anticancer
properties of flavonoids: Roles in various stages of carcinogenesis. Cardiovasc. Hematol. Agents Med.
Chem. 9, 62-77.
(10) Samarghandian, S., Afshari, J. T., and Davoodi, S. (2011) Chrysin reduces proliferation and
induces apoptosis in the human prostate cancer cell line pc-3. Clinics (Sao Paulo) 66, 1073-1079.
(11) Samarghandian, S., Azimi-Nezhad, M., Borji, A., Hasanzadeh, M., Jabbari, F., Farkhondeh,
T., and Samini, M. (2016) Inhibitory and cytotoxic activities of chrysin on human breast
adenocarcinoma cells by induction of apoptosis. Pharmacogn. Mag. Suppl. 4, S436-S440.
(12) Samarghandian, S., Farkhondeh, T., and Azimi-Nezhad, M. (2017) Protective effects of
chrysin against drugs and toxic agents. Dose Response 15, 1-10.
(13) Li, X. X., He, G. R., Mu, X., Xu, B., Tian, S., Yu, X., Meng, F. R., Xuan, Z. H., and Du, G.
H. (2012) Protective effects of baicalein against rotenone-induced neurotoxicity in PC12 cells and
isolated rat brain mitochondria. Eur. J. Pharmacol. 674, 227-233.
(14) Nabavi, S. M., Habtemariam, S., Daglia, M., and Nabavi, S. F. (2015) Apigenin and breast
2
9
ACS Paragon Plus Environment