Cytochrome P450 mediated metabolic activation of chrysophanol

Article abstract of DOI:10.1016/j.cbi.2018.04.015

Chrysophanol, a major anthraquinone component occurring in many traditional Chinese herbs, is accepted as important active component with various pharmacological actions such as antibacterial and anticancer activity. Previous studies demonstrated that exposure to chrysophanol induced cytotoxicity, but the mechanisms of the toxic effects remain unknown. In the present metabolism study, three oxidative metabolites (M1-M3, aloe-emodine, 7-hydroxychrysophanol, and 2-hydroxychrysophanol) and five GSH conjugates (M4-M8) were detected in rat and human liver microsomal incubations of chrysophanol supplemented with GSH, and the formation of the metabolites was NADPH dependent except M4 and M5. M4 and M5 were directly derived from parent compound chrysophanol, M6 arose from M2, and M7 and M8 resulted from the oxidation of M4 and M5. Metabolites M5 and M6 were also observed in bile of rats after exposure to chrysophanol, M1-M3 and one NAC conjugate (M9) were detected in urine of rats administrated chrysophanol, and urinary metabolite M9 originated from the degradation of biliary GSH conjugation M6. Recombinant P450 enzyme incubation and microsome inhibition studies demonstrated that P450 1A2 was the primary enzyme responsible for the metabolic activation of chrysophanol and that P450 2B6 and P450 3A4 also participated in the generation of the oxidative metabolites. These findings helped us to understand the mechanisms of chrysophanol-induced cytotoxicity.

Full text of DOI:10.1016/j.cbi.2018.04.015

Accepted Manuscript
Cytochrome P450 mediated metabolic activation of chrysophanol
Ying Sun, Xin Xin, Kehan Zhang, Tiantian Cui, Ying Peng, Jiang Zheng
PII:
S0009-2797(17)30699-3
10.1016/j.cbi.2018.04.015
CBI 8279
DOI:
Reference:
To appear in: Chemico-Biological Interactions
Received Date: 30 June 2017
Revised Date: 28 March 2018
Accepted Date: 12 April 2018
Please cite this article as: Y. Sun, X. Xin, K. Zhang, T. Cui, Y. Peng, J. Zheng, Cytochrome P450
mediated metabolic activation of chrysophanol, Chemico-Biological Interactions (2018), doi: 10.1016/
j.cbi.2018.04.015.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
1
Cytochrome P450 Mediated Metabolic Activation of Chrysophanol
2
3
Ying Suna, Xin Xina, Kehan Zhanga, Tiantian Cuia, Ying Penga,*, and Jiang Zhenga,b,*
aWuya College of Innovation, Shenyang Pharmaceutical University, Shenyang, Liaoning,
4
5
110016, PR China
6
bState Key Laboratory of Functions and Applications of Medicinal Plants, Key
Laboratory of Pharmaceutics of Guizhou Province, Guizhou Medical University,
Guiyang, Guizhou, 550025, PR China.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
1

ACCEPTED MANUSCRIPT
1
2
Corresponding Authors:
Jiang Zheng, PhD
3
§Wuya College of Innovation, Shenyang Pharmaceutical University, Shenyang, Liaoning,
110016, PR China；¶State Key Laboratory of Functions and Applications of Medicinal
Plants, Key Laboratory of Pharmaceutics of Guizhou Province, Guizhou Medical
University, Guiyang, Guizhou, 550025, PR China;
4
5
6
7
E-mail: zhengneu@yahoo.com
8
Tel: +86-24-23986361
9
Fax: +86-24-23986510
10
11
12
13
14
15
16
17
18
19
20
Ying Peng, PhD
Wuya College of Innovation, Shenyang Pharmaceutical University, Shenyang, PR China.
E-mail: yingpeng1999@163.com
Tel: +86-24-23986361
Fax: +86-24-23986510
§¶: The two corresponding units contributed equally to this work.
2

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
Abbreviations：CE, collision energy; DDQ, 2,3-dichloro-5,6-dicyano-1,4-benzoquinone;
DMSO, dimethyl sulfoxide; DP, declustering potential; EPI, enhanced product ion; GSH,
glutathione;
IDA,
information-dependent
acquisition;
LC-MS/MS,
liquid
chromatography-tandem mass spectrometry; MRM, multiple-reaction monitoring; NAC,
N-acetyl cysteine; NADPH, β-nicotinamide adenine dinucleotide 2′-phosphate reduced
tetrasodium salt; NMR, nuclear magnetic resonance.
3

ACCEPTED MANUSCRIPT
1
2
Abstract
Chrysophanol, a major anthraquinone component occurring in many traditional
Chinese herbs, is accepted as important active component with various pharmacological
actions such as antibacterial and anticancer activity. Previous studies demonstrated that
exposure to chrysophanol induced cytotoxicity, but the mechanisms of the toxic effects
remain unknown. In the present metabolism study, three oxidative metabolites (M1-M3,
aloe-emodine, 7-hydroxychrysophanol, and 2-hydroxychrysophanol) and five GSH
conjugates (M4-M8) were detected in rat and human liver microsomal incubations of
chrysophanol supplemented with GSH, and the formation of the metabolites was NADPH
dependent except M4 and M5. M4 and M5 were directly derived from parent
compound chrysophanol, M6 arose from M2, and M7 and M8 resulted from the oxidation
of M4 and M5. Metabolites M5 and M6 were also observed in bile of rats after
exposure to chrysophanol, M1-M3 and one NAC conjugate (M9) were detected in urine
of rats administrated chrysophanol, and urinary metabolite M9 originated from the
degradation of biliary GSH conjugation M6. Recombinant P450 enzyme incubation and
microsome inhibition studies demonstrated that P450 1A2 was the primary enzyme
responsible for the metabolic activation of chrysophanol and that P450 2B6 and P450
3A4 also participated in the generation of the oxidative metabolites. These findings
helped us to understand the mechanisms of chrysophanol-induced cytotoxicity.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
4

ACCEPTED MANUSCRIPT
1
2
1. Introduction
Chrysophanol
(1,8-dihydroxy-3-methylanthracenedione),
a
3
4
1,8-dihydroxyanthraquinone derivative, is a natural ingredient of many Chinese herbal
medicines, such as Rheum palmatum L., Polygonum multiflorum, and Polygonum
cuspidatum [1-4]. Chrysophanol has been identified as a component responsible for
antibacterial activity, anticancer action, anti-diabetic property and shortening effect on
blood coagulation time [5-11]. Given its high pharmaceutical values, chrysophanol has
been widely used in clinic in China. In addition, chronic use of anthraquinone drugs is
not recommended due to their adverse effects including an increased risk of colorectal
carcinoma [12-13]. Anthraquinones as major nephrotoxic components of rhubarb
induced cytotoxicity in human carcinoma cells [14]. Chrysophanol also exhibited
genotoxicity by micronucleus test and mutation assay in mouse lymphoma L5178Y tk+/-
cells [7]. Thus, any health risks associated with exposure to chrysophanol should be
carefully evaluated.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Studies have demonstrated that the biotransformation of chrysophanol is cytochrom
P450-dependent, and chrysophanol is transformed to aloe-emodine as the major product
formed. Besides, metabolite aloe-emodine reportedly induced significantly higher
micronucleus frequencies than did chrysophanol in mouse lymphoma L5178Y cells [7].
In addition, hydroxyl substitution on the phenyl ring is another predominant metabolic
reaction for chrysophanol [15]. It has been reported that the oxidation of the
chrysophanol seems to be catalyzed by several P450 enzymes, and 1A2 plays a major
5

ACCEPTED MANUSCRIPT
1
2
role in hydroxylating planar polycyclic aromatic compounds and several phenols among
these enzymes [16-17]. Notably, numbers of hepatic toxicity cases associated with
metabolic activation were mediated by cytochrome P450 enzymes [18] and quinones are
highly reactive compounds that exert cytotoxicity through reaction with nucleophiles of
macromolecules. This encouraged us to characterize potential reactive intermediates of
chrysophanol and to define the metabolic pathways of chrysophanol, which may facilitate
the understanding of the mechanisms of the reported chrysophanol toxicities. The
objectives of the present studies were 1) characterization of reactive quinone metabolites
of chrysophanol using in vitro and in vivo models; 2) elucidation of the pathways of
chrysophanol bioactivation; and 3) identification of cytochrome P450 enzymes involved
in the generation of the metabolites of chrysophanol.
3
4
5
6
7
8
9
10
11
12
6

ACCEPTED MANUSCRIPT
1
2
2. Materials and Methods
2.1 Chemicals and Materials
3
Chrysophanol (> 98%) and aloe-emodine (> 98%) were purchased from Chengdu
Pufeide Biologic Technology Co., Ltd. (Chengdu, China). Glutathione (GSH),
N-acetylcysteine (NAC), S-hexylglutathione, reduced nicotinamide adenine dinucleotide
phosphate (NADPH), 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ), ticlopidine,
α-naphthoflavone, sulfaphenazole, quinidine, disulfiram, methoxsalen, pilocarpine, and
ketoconazole were obtained from Sigma-Aldrich Co. (St. Louis, MO). Recombinant
human P450 enzymes and human liver microsomes were obtained from BD Gentest
(Woburn, MA). Rat liver microsomes were prepared from Sprague-Dawley rats (male),
based on the procedure we published earlier [19]. The C18 solid-phase extraction (SPE)
columns were purchased from Thermo Fisher Scientific Co., Ltd. ( Shanghai ,China).
Distilled water was acquired from Wahaha Co., Ltd. (Hangzhou, China). All organic
solvents were from Fisher Scientific (Springfield, NJ). All reagents and solvents were
of either analytical or LC grade.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
2.2 Microsomal Incubations
Chrysophanol was dissolved in dimethyl sulfoxide (DMSO) as stock solution, and the
total volume applied in the incubations did not exceed 1% of the total incubation volume
[20]. The incubation mixtures were prepared in a final volume of 250 µL of phosphate
buffer (pH 7.4) containing 100 µM chrysophanol, 20 mM GSH or 10 mM NAC, rat or
7

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
human liver microsomes (1.0 mg protein/mL), and 3.2 mM MgCl2. The reactions were
launched by addition of NADPH (1.0 mM). The control groups without NADPH were
included to ensure that the metabolism of chrysophanol was NADPH dependent. The
reactions were terminated by adding ice-cold acetonitrile (1:1 v/v) after 45 min
incubation at 37 °C. Incubations were conducted in duplicate. The reaction mixtures
were vortex mixed and centrifuged at 19,000 g for 10 min at 4 °C to remove precipitated
protein. The supernatants (10 µL) were subjected to LC-MS/MS for analysis.
8
9
2.3 Chemical Synthesis of Chrysophanol Metabolites
2-Hydroxychrysophanol (M2) and 7-hydroxychrysophanol (M3) were chemically
synthesized from chrysophanol, according to a published protocol [22]. The brief
operation is as follows: a solution of chrysophanol (20 mg) in sulfuric acid (1 mL) was
mixed with potassium persulfate (52 mg). When the resulting deep red solution was
stirred for 10 min at 22 ± 4 , a further quantity of potassium persulfate (52 mg) was
added with additional 50 min stirring. The resultant violet solution was diluted with
water (30 mL). The resulting orange suspension was mixed with sodium metabisulfite
(0.2 g), followed by extraction with butanol (3 mL). The deep orange organic phase
was washed with dilute hydrochloric acid (3 × 5 mL) and then by water (3 × 5 mL).
After evaporated to dryness under a stream of nitrogen, the residues were redissolved
with the initial mobile solvent and submitted to the LC-MS/MS system. Synthetic M2
and M3 were purified by semipreparative liquid chromatography. All nuclear magnetic
10
11
12
13
14
15
16
17
18
19
20
8

ACCEPTED MANUSCRIPT
1
2
resonance (NMR) measurements were obtained at 600 MHz on a BRUKER-ARX-600
spectrometer.
3
GSH conjugates M4 and M5 were synthesized by direct reaction of GSH (100 µmol,
20 mM) with chrysophanol (0.5 µmol, 100 µM) in potassium phosphate buffer (5 mL).
The mixture was stirred at 37 °C for 1 h. The resulting reaction solution was centrifuged
at 19,000 g for 10 min. The supernatants (10 µL) were then collected and analyzed by
LC-MS/MS. Synthetic M4 and M5 were purified by high performance liquid
chromatography.
4
5
6
7
8
9
Chemical synthesis of GSH conjugate M6 was performed as follows: synthetic M2
(1.0 mg, 3.7 µmol) was dissolved in dichloromethane (1.5 mL), followed by slow
addition of DDQ (3.6 mg, 15.8 µmol), and the reaction mixture was stirred for 1 h at
room temperature and mixed with 0.8 mL H2O. The resulting mixture was concentrated
by blowing with a stream of nitrogen, reconstituted in tetrahydrofuran (0.6 mL), and
mixed with GSH (20 µmol) aqueous solution (1.0 mL). The reaction mixtures were
stirred at 37 °C for 5 h and then dried with a gentle stream of nitrogen gas. The
resulting concentrates were reconstituted with 200 µL of 10% acetonitrile in water and
then centrifuged at 19,000 g for 10 min. A 10 µL aliquot of the supernatants was
analyzed by LC-MS/MS system.
10
11
12
13
14
15
16
17
18
19
20
21
GSH conjugates M7 and M8 were synthesized by automatic oxidation of purified M4
and M5 in the air as follows: purified M4 and M5 were redissolved in 500 µL potassium
phosphate buffer (pH 7.4), followed by 1 h stirring at 37 °C in the air and centrifugation
9

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
at 19,000 g for 10 min. A 10 µL aliquot of the resulting supernatants was analyzed by
LC-MS/MS.
NAC conjugate M9 was synthesized by direct reaction of synthetic M2 with NAC
(100 mM) in potassium phosphate buffer, followed by 1 h stirring at 37 °C and
centrifugation at 19,000 g for 10 min. The supernatants (10 µL) were then collected and
analyzed by LC-MS/MS.
7
8
2.4 Animal Experiments
All animal manipulations met the requirements approved by the Ethics Review
Committee for Animal Experimentation of Shenyang Pharmaceutical University
(Shenyang, China). Male Sprague−Dawley rats (180 ± 20 g) were obtained from the
Animal Center of Shenyang Pharmaceutical University. All rats were fed in a controlled
environment (temperature of 22±4 °C，moderate humidity) and acclimatized at 12 h
light/dark cycle. Animals were divided into two groups, and each group contained three
rats. Before administration, the experimental animals were deprived of food for 24 h.
The animals were anesthetized with 10% chloral hydrate (3.0 mL/kg, i.p.), and their bile
ducts were cannulated with PE-10 tubing. Rats were intraperitoneally administered with
chrysophanol (50 mg/kg) dissolved in corn oil, and bile was collected from 0–12 h after
administration. Control bile was also collected before the treatment. The other group
of rats was placed in metabolism cages and administered with the same dose of
chrysophanol. Urine samples were collected during the time periods of 0−24 h after the
9
10
11
12
13
14
15
16
17
18
19
20
10

ACCEPTED MANUSCRIPT
1
2
3
procedure. Control urine was collected prior to the i.p. administration. During the
experiment, the rats were allowed free access to food and water. Both the collected bile
and urine samples were stored at −80 °C until analysis.
4
5
2.5 Sample Preparation
The collected bile and urine samples were cleaned up by using the C18 SPE columns.
First, the SPE columns were activated with 10 mL of methanol and balanced by 10 ml of
water containing glacial acetic acid (pH 3-4). Then, 300 µL bile or urine sample at pH
3.5 was loaded onto the upper of SPE columns. The samples on the SPE columns were
washed with 10 mL of deionized water and eluted with a gradient methanol solution.
The specific concentrations were 20, 30, 40, 50, 60, and 80%, respectively. The eluates
were concentrated under a stream of nitrogen gas and reconstituted in 200 µL of 10%
acetonitrile in water and centrifuged at 19,000 g for 10 min at 4 °C. An aliquot (10 µL)
of the supernatant was analyzed by LC-MS/MS.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
2.6 Recombinant Human P450 Incubations
To determine the specific P450 enzymes involved in the formation of reactive
metabolites of chrysophanol, a total of nine human recombinant P450s, including
cytochrome P450s 1A2, 2A6, 2B6, 2C9, 2C19, 2D6, 2E1, 3A4, and 3A5 (100 nM in each
incubation), were tested. Incubation mixtures contained 100 nM recombinant human
P450 enzyme, 100 µM chrysophanol, 1.0 mM NADPH, and 20 mM GSH (final volume:
250 µL). The incubation time was 45 min. Reactions were quenched by addition of an
11

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
equal volume (250 µL) of ice-cold acetonitrile which contained S-hexylglutathione (20
ng/mL) as the internal standard with similar structure of the GSH conjugates analyzed.
The formation of chrysophanol metabolites was monitored by an LC-MS/MS method.
The rates of metabolite formation in individual incubations with recombinant P450
enzymes were multiplied by the mean specific content of the corresponding P450 enzyme
in human liver microsomes to obtain the normalized reaction rates of each enzyme [21].
7
8
2.7 Metabolizing Enzyme Inhibition Studies
The incubation systems contained human liver microsomes (1.0 mg protein/mL),
GSH (20 mM) or NAC (10 mM), chrysophanol (100 µM), and MgCl2 (3.2 mM) in
potassium phosphate buffer (final volume: 250 µL), along with selective P450 inhibitors
[α-naphthoflavone (5.0 µM for P450 1A2), sulfaphenazole (20 µM for P450 2C9),
ticlopidine (100 µM for P450s 2B6 and 2C19), quinidine (10 µM for P450 2D6),
disulfiram (100 µM for P450 2E1), methoxsalen (20 µM for P450s 2A13 and 2A6),
pilocarpine (100 µM for P450 2A1), and ketoconazole (10 µM for P450 3A)]. All
inhibitors were dissolved individually in DMSO as stock solution. The inhibitor
concentrations were selected according to the Food and Drug Administration Guidance
for Industry on Drug Interaction Studies. Incubation mixtures consisted of human liver
microsomes (1.0 mg protein/mL), chrysophanol (100 mM), NADPH (1.0 mM), GSH (20
mM), and the individual inhibitors at the designed concentrations. The final volume
was 250 µL. Incubations were performed at 37 °C for 45 min, and reactions were
9
10
11
12
13
14
15
16
17
18
19
20
12

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
terminated by an equal volumeof ice-cold acetonitrile containing internal standard
S-hexylglutathione (20 ng/mL). Control groups lacked the inhibitors. The reaction
mixtures were vortex mixed and centrifuged at 19,000 g for 10 min at 4 °C to remove
precipitated protein. The production of the GSH conjugate was monitored by an
LC-MS/MS method. A comparison was made relative to the controls in the absence of
the inhibitors, and P450 activity was expressed as the percentage of the control activity.
The experiments were performed in duplicate.
8
9
2.8 Determination of Air-Labile Reactions
To examine the reactive intermediates in the reaction system, microsomal
incubations were purged with nitrogen, and the reactions were started by mixing with
NADPH (1.0 mM). The control groups were performed under the same conditions
except for the absence of NADPH. After 45 min incubation at 37 °C under nitrogen, the
reactions were quenched by adding equal volumes of ice-cold acetonitrile. The
reactions were vortex-mixed for 3 min and centrifuged at 19,000 g for 10 min to remove
protein. The supernatants (10 µL) were injected onto LC-MS/MS for analysis. The
rest of samples were placed in the air for 24 h, vortexed in the air for 10 min, and
resubmitted to LC-MS/MS for further analysis. In a separate study, chrysophanol (100
µM) and GSH (20 mM) were mixed directly in potassium phosphate buffer (pH 7.4).
The final volume was 250 µL. The mixture was stirred in a gentle shaking water bath at
37 °C for 1 h. The mixtures were vortexed for 3 min and centrifuged at 19,000 g for 10
10
11
12
13
14
15
16
17
18
19
20
13

ACCEPTED MANUSCRIPT
1
2
3
min. A 10 µL aliquot of the resulting supernatants was analyzed by LC-MS/MS. The
rest of samples were resubmitted to LC-MS/MS after exposed to the air for 24 h and
vortexed in the air for 10 min.
4
5
2.9 LC-MS/MS Method for Metabolite Identification
LC-MS/MS analyses were performed on an AB SCIEX Instruments 4000 hybrid
triple quadrupole-linear ion trap (Q-Trap) mass spectrometer (Applied Biosystems, Foster
City, CA) equipped online with a 1260 infinity LC system (Agilent Technologies,
Biblingen, Germany). Data acquisition and processing were performed using Applied
Biosystems/SCIEX Analyst software (version 1.6.2). The separations of analytes were
achieved on a Accuore C18 column (4.6 × 150 mm, 5µm; Thermo Fisher, Inc., Pittsburgh,
PA). Gradient elution was performed with acetonitrile containing 0.1% formic acid
(solvent A) and water containing 0.1% formic acid (solvent B). The gradient elution
was optimized as follows: 0-2 min, 10% solvent A; 2-12 min, 10%-95% solvent A; 12-17
min, 95% solvent A; 17-18 min, 95%-10% solvent A; 18-21 min, 10% solvent A. The
flow rate of the mobile phase and the volume of injection were set at 0.8 mL/min and 10
µL, respectively. Mass spectrometric analyses were conducted using an electrospray
ionization source in positive-ion mode, and all analytes were performed by
multiple-reaction monitoring (MRM) and enhanced product ion (EPI) scanning mode
with the following conditions: ion spray voltage, 5500 V; source temperature, 650 °C;
curtain gas, 25 psi; ion source gas 1, 50 psi; ion source gas 2, 50 psi; declustering
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
14

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
potential, 115 V; entrance potential, 10 V; and cell exit potential, 15 V. The
characteristics of ion pairs (corresponding to collision energy) were m/z 255 → 153 (43)
for chrysophanol; 271 → 225 (38) for hydroxylation metabolites of chrysophanol; m/z
560 → 285 (55) and 576 → 327 (47) for GSH conjugates; m/z 432 → 301 (40) for NAC
conjugate; and m/z 392 → 246 (30) for S-hexylglutathione (internal standard). The EPI
scan was run in positive mode at a scan range for product ions from m/z 50 to 600. The
collision energy was set at 42 eV with a spread of 15 eV.
15

ACCEPTED MANUSCRIPT
1
2
3. Results
3.1 MS Behavior of Chrysophanol
3
Considering chemical structure, parent drugs and their metabolites often share great
4
similarity.
We started with investigating the mass spectrometric property of
5
chrysophanol before metabolite identification. Parent compound chrysophanol showed
fragment ions of m/z 237 ([M + H − H2O ]+), 227 ([M + H − CO]+), 209 ([M + H − H2O
− CO]+), 181 ([M + H − H2O − 2CO]+), 153 ([M + H − H2O − 3CO]+), and 127 ([M + H
− H2O − 3CO − C2H2]+), along with [M + H]+ of m/z 255 (Fig. 1). The characteristic
product ions from the parent compound can be used to assist the characterization of their
metabolites.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
3.2 Oxidative Metabolites of Chrysophanol
Chrysophanol was incubated in rat or human liver microsomes supplemented with
NADPH and GSH as a trapping agent. Three hydroxylation metabolites (M1−M3) were
detected in both rat and human liver microsomes (Fig. 2B and S1A,B), and only minor
such metabolites were observed in the absence of NADPH (Fig. 2A), which indicates that
the formation of these metabolites was NADPH-dependent and P450 enzymes played an
important role in the formation of these metabolites. M1, M2, and M3, sharing the same
molecular ions, were isomers of oxidative metabolites resulting from hydroxylation of
chrysophanol at three different positions (Scheme 1). The three metabolites exhibited
characteristic fragment ions at m/z 253 ([M + H − H2O]+), 225 ([M + H − H2O − CO]+),
16

ACCEPTED MANUSCRIPT
1
2
197 ([M + H − H2O − 2CO]+), 187 ([M + H − 3CO]+), 169 ([M + H −H2O − 3CO]+), 141
([M + H −H2O − 4CO]+), and 115 ([M + H − H2O − 4CO − C2H2]+) (Fig. 2G-I). The
observed product ions of m/z 225, 197, and 169 received mass gain of 16 Da relative to
the corresponding product ions of chrysophanol at m/z 209, 181 and 153 (Fig. 1),
indicating that the three oxidative metabolites were derived from oxidation of
chrysophanol.
3
4
5
6
7
M1 (retention time = 11.55 min) showed the same retention time, protonated
molecular ion, and fragmentation pattern as authentic standard of aloe-emodine (Fig. 2B,
C, G and J). This made us propose that M1 is aloe-emodine. Chemical synthesis was
conducted to further characterize the other two oxidative metabolites generated in the
microsomal incubations. Chrysophanol was oxidized by potassium persulfate in
sulfuric acid, followed by addition of sodium metabisulfite. The resulting products
showed the same chromatographic and mass spectrometric identities as M2 (retention
time = 12.32 min) and M3 (retention time = 12.70 min) generated in the microsomal
incubations (Fig. 2B, D, E, H, I, K and L).
8
9
10
11
12
13
14
15
16
17
18
19
20
21
13
We succeeded in obtaining 1H NMR and C NMR spectra of synthetic M2 and M3
1
as follows: H NMR (CDCl3, 600 MHz) δH: 12.33 (1H, s, 1-OH), 7.08 (1H, br s, H-2),
2.47 (3H, s, 3-CH3), 7.67 (1H, br s, H-4), 7.83 (1H, d, J=8.4 Hz, H-5), 7.24 (1H, d, J=8.4
Hz, H-6), 11.94 (1H, s, 8-OH); 13C NMR (CDCl3, 150 MHz) δC: 162.8 (C-1), 123.9 (C-2),
149.9 (C-3), 121.6 (C-4), 122.1 (C-5), 120.7 (C-6), 151.5 (C-7), 149.5 (C-8), 193.0 (C-9),
181.2 (C-10), 22.3 (C-11), 114.0 (C-9a), 134.0 (C-4a), 125.4 (C-10a), 116.2 (C-8a). The
17

ACCEPTED MANUSCRIPT
1
1
2
triplet of ring C [23] observed in H NMR spectrum disappeared in the spectrum of the
product, and a pair of double peaks at δH 7.83 (1H, d, J=8.4 Hz, H-5) and δH 7.24 (1H, d,
J=8.4 Hz, H-6) was detected. This was further confirmed by the observed chemical
3
13
4
shift of C-7 migrating to low field (δC: 124.8 to 151.5) in the C NMR spectra. The
5
observed NMR spectra led us to assign M2 as 7-hydroxychrysophanol.
6
1H NMR and 13C NMR spectra of synthetic M3 were recorded for structural
characterization as follows: 1H NMR (CDCl3, 600 MHz) δH: 12.20 (1H, s, 1-OH), 2.38
(3H, s, 3-CH3), 7.71 (1H, s, H-4), 7.82 (1H, d, J=8.4 Hz, H-5), 7.24 (1H, t, J=8.4 Hz,
H-6), 7.27 (1H, d, J=8.4 Hz, H-7), 12.03 (1H, s, 8-OH); 13C NMR (CDCl3, 150 MHz) δC:
148.7 (C-1), 149.2 (C-2), 132.4 (C-3), 123.9 (C-4), 120.1 (C-5), 137.5 (C-6), 124.1 (C-7),
162.5 (C-8), 193.2 (C-9), 181.2 (C-10), 16.4 (C-11), 114.4 (C-9a), 124.2 (C-4a), 134.3
7
8
9
10
11
12
13
14
15
16
1
(C-10a), 116.1 (C-8a). The H NMR spectrum of rings B and C of M3 were similar to
that of chrysophanol [23], except that the broad singlet of H-4 in ring A was replaced by
one sharp singlet. Additionally, the 13C NMR spectrum revealed that the chemical shift
of C-2 moved to low field (δC: 125.1 to 149.2). This indicates that the hydroxyl group
was attached to C2 of chrysophanol.
17
18
19
20
3.3 Identification of Chrysophanol GSH Conjugates
Metabolites M4 and M5 eluted out at 8.73 and 8.94 min, respectively (Fig. 3B).
The two metabolites shared the same [M+H]+ at m/z 560, and the spectra showed the
indicative characteristic fragment ions associated with the cleavage of GSH moiety (Fig.
18

ACCEPTED MANUSCRIPT
1
2
3E and F). The product ions at m/z 485 and 414 were derived from the loss of the
glycinyl moiety (−75 Da) and glutamine moiety (−146 Da) from m/z 560. The fragment
ion at m/z 311 arose from the loss of the N-formylglycine moiety (−103 Da) from m/z 414.
The primary product ion at m/z 285 was derived from the cleavage of the C−S bond of
m/z 311 (−26 Da). In addition, both M4 and M5 were detected in microsomal
incubations in the absence of NADPH (Fig. 3A). Direct reaction of chrysophanol with
GSH in PBS buffer solution produced two products which revealed the same
chromatographic and mass spectral behaviors as those of M4 and M5 produced in the
microsomal incubations (Fig. 3A-C and E-H). Unfortunately, we were unable to obtain
enough amounts of both the resulting products for nuclear magnetic resonance (NMR)
characterization, due to the poor yields of the reaction.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Three peaks assigned as M6 (retention time = 8.64 min), M7 (retention time = 9.10
min), and M8 (retention time = 9.31 min) responsible for GSH conjugates ([M+H]+ at m/z
576) derived from hydroxylation metabolites of chrysophanol were observed by MRM
scanning mode (m/z 576→327) in the incubation mixtures (Fig. 4B, S1C and D).
MS/MS analysis was performed with the ion trap instrument to further characterize the
structure of GSH conjugates. The fragment ion at m/z 501 and 430 derived from the
loss of the glycinyl moiety (−75 Da) and glutamine moiety (−146 Da) from m/z 576,
respectively. The product ion at m/z 327 arose from the loss of the N-formylglycine
moiety (−103 Da) from m/z 430. The primary product ion at m/z 301 was derived from
the cleavage of the C−S bond of m/z 327 (−26 Da) (Fig. 4E, 5D and E). The observed
19

ACCEPTED MANUSCRIPT
1
2
fragment ions of M6, M7 and M8 at m/z 501, 430, 327 and 301 were 16 Da more than the
product ions of M4 and M5 at m/z 485, 414, 311 and 285 (Fig. 3E and F). This suggests
that the formation of M6, M7 and M8 resulted from hydroxylation of chrysophanol and
then addition of a molecule of GSH.
3
4
5
Chemical synthesis was performed to further characterize metabolite M6 detected in
rat and human liver microsomal incubations. Synthetic M2 with purity of > 98% was
sequentially oxidized with DDQ, followed by reaction with GSH. The GSH conjugate
formed in the reaction was found to correspond to M6 observed in the microsomal
incubations, based on the chromatographic and mass spectrometric identities (Fig. 4B, C,
E and F).
6
7
8
9
10
11
12
13
14
15
16
To determine the origins of M7 and M8, synthetic M4 and M5 with purity of > 98%
were oxidized spontaneously by exposure to the air，and the oxidation products
demonstrated the similar chromatographic and mass spectrometric properties as those of
M7 and M8 found in the microsomal incubations (Fig. 5B-G). Unfortunately, we were
unable to obtain enough amounts of the metabolites/products for NMR characterization,
due to the complexity and poor yields of the reactions.
17
18
19
20
3.4 Biliary and Urinary Metabolites of Chrysophanol
To investigate the bioactivation of chrysophanol in vivo, bile and urine samples were
collected in chrysophanol-exposed rats and analyzed by LC-MS/MS system in MRM-EPI
mode. Two biliary GSH conjugates (M5 and M6) with molecular ions of m/z 560 and
20

ACCEPTED MANUSCRIPT
1
2
576 were found in the bile of rats treated with chrysophanol (Fig. 3D and 4D), and no
such metabolites were observed in the control group. The metabolites found in vivo
studies showed the same chromatographic and mass spectrometric behaviors (Fig. 3D, 3I,
4D and 4G) as those formed in the microsomal incubations. However, we failed to find
the other three chrysophanol GSH conjugates (M4, M7, and M8) in the bile samples.
Three oxidative metabolites (M1, M2, and M3) were found in the urine of rats after
administration of chrysophanol (Fig. 2F), and no such metabolites were observed in the
urine before the treatment with chrysophanol. The urinary metabolites showed the same
retention time and major fragment ions as that for those produced in microsomal
incubations (Fig. 2B, F, G-I and M-O). No chrysophanol-derived GSH conjugates were
detected in urine of rats given chrysophanol. Instead, we found one urinary metabolite
(M9) with molecular ion of m/z 432 which matches the molecular weight of the NAC
conjugate derived from hydroxylation metabolite of chrysophanol (Fig. 6B, D, E and G).
As expected, no such conjugate was detected in the blank urine.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Urinary metabolite M9 (retention time = 10.28 min) was detected by monitoring the
ion of m/z 432 in positive ion mode (Fig. 6D). The product ion at m/z 327 was derived
from the loss of acetyl amino group and carboxylic acid. The product ion at m/z 303
corresponded to the characteristic neutral loss of 129 Da for the NAC conjugates. The
fragment ions at m/z 162 and 130 corresponded to the NAC moiety with sulfur and
without sulfur, respectively. These results suggest that NAC conjugate M9 was derived
from a hydroxylation metabolite of chrysophanol. In a separate study, an NAC
21

ACCEPTED MANUSCRIPT
1
2
3
4
5
6
7
8
9
conjugate (retention time = 10.34 min) was found in incubations containing chrysophanol,
rat liver microsomes, NADPH, and NAC. The observed NAC conjugate showed the
same chromatographic and mass spectral properties as those found in urine sample (Fig.
6B, D, E and G). Chemical synthesis was performed to further characterize the
metabolite detected in both the urine of rats and microsomal incubations after exposure to
chrysophanol. Synthetic M2 reacted with NAC aqueous solution in potassium
phosphate buffer, followed by 1 h stirring at 37 °C. The NAC conjugate (retention time
= 10.29 min）formed in the reaction was found to correspond to M9, based on the
chromatographic and mass spectrometric identities (Fig. 6C and F).
10
11
12
13
14
15
16
17
18
19
20
3.5 Identification of P450 Enzymes Involved in Bioactivation of Chrysophanol
To identify which individual human recombinant P450 enzymes are mainly
responsible for the bioactivation of chrysophanol, chrysophanol was incubated with
individual human recombinant P450 enzymes, including 1A2, 2A6, 2C9, 2C19, 2B6,
2D6, 2E1, 3A4, and 3A5 supplemented with GSH as the trapping agent in the presence of
NADPH, followed by monitoring the formation of M1−M3 and M6−M8. We found that
multiple P450 enzymes mediated the metabolism of chrysophanol. As shown in Figures
7A and B, P450 2B6 was found to be the major enzyme was involved in the production of
M1, P450 1A2 was the foremost enzyme participating in the generation of M2, M6, M7
and M8, and P450 3A4 was the primary enzyme contributing to the formation of M3.
P450 3A4 catalyzed the formation of M1−M3, and M6−M8 of chrysophanol.
22

ACCEPTED MANUSCRIPT
1
2
The identification of P450 enzymes in the bioactivation of chrysophanol was also
performed by incubation of chrysophanol in human liver microsomes mixed with
individual P450 enzyme-selective inhibitors，including α-naphthoflavone (P450 1A2
inhibitor), pilocarpine (P450 2A1 inhibitor), methoxsalen (P450s 2A13 and 2A6
inhibitor), sulfaphenazole (P450 2C9 inhibitor), ticlopidine (P450s 2B6 and 2C19
inhibitors), quinidine (P450 2D6 inhibitor), disulfiram (P450 2E1 inhibitor), and
ketoconazole (P450 3A inhibitor). Among the inhibitors applied, α-naphthoflavone
showed a significant inhibitory effect on the formation of M6, M7 and M8 (Fig. 7C).
No or minor inhibition was observed in the microsomal incubations in the presence of the
other inhibitors. In addition, the production of M3 in human liver microsomes was
significantly lower than that in rat liver microsomes, indicating that there was species
difference in the bioactivation of chrysophanol (Fig. S1). The above results verified that
P450 1A2 was the foremost enzyme involved in the bioactivation of chrysophanol in
human liver microsomes.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
23

ACCEPTED MANUSCRIPT
1
2
4. Discussion
In recent years, as the global interest in natural medicines is increasing, more and
more cases of poisonings resulting from exposure to herbal medicines have been being
reported. Drug induced liver injury is not rare in clinical trials and clinical practice.
Chrysophanol is one of the major components in rhubarb which is widely used as a
traditional herbal medicine. Although chrysophanol was reported to cause cytotoxicity
[24,25], the mechanisms of toxicities induced by chrysophanol remain unknown. Our
study was designed and performed to elucidate the metabolism pathways of chrysophanol
to better understand the mechanisms of the toxic action. As an initial effort for
metabolite identification, mass spectrometric behavior and fragmentation pattern of
chrysophanol were determined (Fig. 1), which assisted the structural characterization of
chrysophanol metabolites.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Successively, we performed microsomal incubations with chrysophanol
supplemented with GSH and NADPH. Three hydroxylation metabolites (M1-M3) were
detected in both rat and human microsomal incubations, and the formation of the three
metabolites was NADPH-dependent. Apparently, hydroxylation of chrysophanol at the
alkyl carbon and aromatic carbons was the primary metabolism pathway. The identified
putative activation processes of chrysophanol are shown in Scheme 1. The retention
time and fragmentation pattern of M1 proved that chrysophanol was biotransformed into
aloe-emodine. Based on the previous study, the biotransformation of chrysophanol by
cytochromes P450 to yield aloe-emodine is regarded as a bioactivation reaction, because
24

ACCEPTED MANUSCRIPT
1
2
aloe-emodine is a much more potent inducer of micronuclei than chrysophanol[16]
.
Additionally, chemical synthesis was performed to characterize the structural identities of
7-hydroxychrysophanol (M2) and 2-hydroxychrysophanol (M3). We all know that
ortho-quinone is an electrophilic species which is generally considered as an important
structural alert. The ortho-quinone intermediates derived from M2 could induce
cytotoxicity.
3
4
5
6
7
M4 and M5 were found in the reaction system without the catalysis of cytochrome
P450 enzymes (Fig. 3C), which demonstrates chrysophanol can directly react with GSH.
To probe the mechanism, we mixed chrysophanol with GSH in a buffer system in a full
of nitrogen. In addition to M4 and M5, we detected two hydroquinone compounds
(hydroquinone 1 and hydroquinone 2 with m/z 562 → 287) (Fig. S2) in the mixture. As
expected, the amount of hydroquinone 1 and hydroquinone 2 gradually declined in the
mixture as the time of exposure to the air increased (Fig. S2A and B). This suggests that
the hydroquinone intermediates are air-labile and auto-oxidized to quinones M4 and M5
(Scheme 1). Unfortunately, due to the instibility of M4 and M5, we failed to gain
sufficient amounts to further characterize the structure of M4 and M5 by NMR. We
speculated that M4 and M5 may be positional isomers resulting from the reaction of GSH
with chrysophanol at different sites of the aromatic system.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Two pathways are considered to be responsible for the formation of M2-originated
M6 (7-hydroxychrysophanol-GSH conjugate). One is that direct reaction of M2 with
GSH (Michael addition) offers the corresponding GSH conjugate (a hydroquinone
25

ACCEPTED MANUSCRIPT
1
2
derivative), followed by auto-oxidation to quinone M6. The other is that the reaction
starts with oxidation of M2 to the corresponding benzoquinone intermediate which
3
sequentially reacts with GSH.
To probe the mechanism, we determined the
4
hydroquinone derivative (m/z 578 → 329) in the microsomal incubation system operated
under nitrogen. No hydroquinone intermediates were detected (date not shown). In a
separate study, we oxidized M2 using DDQ, followed by reaction with GSH. The
reactions produced M6. The results above suggest that formation of M6 started with
hydroxylation of chrysophanol to hydroxychrysophanol which was sequentially oxidized
to the corresponding benzoquinone intermediate, followed by reaction with GSH
(Scheme 1).
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
M7 and M8 were tentatively assigned to be hydroxylated products of M4 and M5.
To determine the origins of M7 and M8, a solution of synthetic M4 and M5 was stirred in
the air. The auto-oxidation reaction produced two oxidative products with the same
chromatographic and mass spectrometric properties as those of M7 and M8 found in the
microsomal incubations. This indicates that GSH conjugates M7 and M8 resulted from
auto-oxidation of M4 and M5. In addition, the production of M7 and M8 in rat liver
microsomal incubations containing NADPH was significantly higher than the control
group without NADPH (Fig. 5A and B), indicating that P450 enzymes were also involved
in the oxidation of M4 and M5 to M7 and M8.
GSH conjugates M5 and M6 were detected in bile of rats administered chrysophanol
as observed in microsomal incubations, but not M4 for some reason. In addition, we
26

ACCEPTED MANUSCRIPT
1
2
failed to detect M7 and M8 in the bile of chrysophanol-treated rats. Biliary GSH
conjugates are often used as biomarkers of exposure to electrophilic species [26-27]. In
our study, the observation of GSH conjugates in vitro and in vivo implies that the reactive
metabolites might bind to proteins or other biologic macromolecules, which could initiate
toxicities. Glutathione conjugates are often converted to mercapturic acids before
excretion in urine. This process involves hydrolytic removal of glutamic acid and
glycine leaving the cysteine conjugates, followed by N-acetylation to produce
mercapturic acids (NAC conjugates). In our study, we successfully detected NAC
conjugate M9 in urine of chrysophanol-treated rats. Our study demonstrated that M9
was derived from the degradation of GSH conjugate M6.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
Three hydroxylation metabolites were formed both in vitro and in vivo. Apparently,
three different P450 enzymes were responsible for the formation of the three metabolites
with some preferences. P450 2B6 was found to be the major enzyme involved in the
production of M1, P450 1A2 was the foremost enzyme participating in the generation of
M2, and P450 3A4 was the primary enzyme contributing to the metabolic activation of
M3. Recombinant enzyme study demonstrated that CYP1A2 was the major enzyme
involved in the generation of M6-M8, followed by CYPs3A4 and 3A5. Furthermore,
coincubation with α-naphthoflavone (CYP1A2 inhibitor) in the microsomal incubations
significantly attenuated the generation of M6-M8. Interestingly, ketoconazole (P450 3A
inhibitor) failed to slow down the formation of M6-M8. This might result from the
dominant role of CYP1A2 in the formation of M6-M8 in the microsomal reactions. The
27

ACCEPTED MANUSCRIPT
1
2
enzyme identification work will facilitate our ongoing mechanistic study of chrysophanol
cytotoxicity.
3
In summary, we investigated metabolic bioactivation pathways of chrysophanol.
Three hydroxylation metabolites were detected in chrysophanol-exposed rats and human
liver microsomes: aloe-emodine, 7-hydroxychrysophanol, and 2-hydroxychrysophanol.
Both the parent compound chrysophanol and the oxidative metabolites reacted with GSH
in vitro and in vivo. M2 (7-hydroxychrysophanol) was further oxidized to the
corresponding quinone which reacts with GSH. P450 1A2 was the primary enzyme
catalyzing the bioactivation of chrysophanol. P450 2B6 and P450 3A4 also participated
in the generation of the oxidative metabolites of chrysophanol. These findings allowed
us to better understand the mechanisms of cytotoxicity induced by chrysophanol.
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
28

ACCEPTED MANUSCRIPT
1
2
References
[1] P. Li, Q. Lu, W. Jiang, X. Pei, Y. Sun, H. Hao, K. Hao, Pharmacokinetics and
pharmacodynamics of rhubarb anthraquinones extract in normal and disease rats, Biomed
Pharmacother. 91 (2017) 425-435.
3
4
5
[2] D. Aichner, M. Ganzera, Analysis of anthraquinones in rhubarb (Rheum palmatum
and Rheum officinale) by supercritical fluid chromatography, Talanta. 144 (2015)
1239-1244.
6
7
8
[3] S. Yao, Y. Li, L. Kong, Preparative isolation and purification of chemical constituents
from the root of Polygonum multiflorum by high-speed counter-current chromatography,
J Chromatogr A. 1115 (2006) 64-71.
9
10
11
12
13
14
15
16
17
18
19
20
21
[4] S.F. Yang, Y.S. Sun, J.H. Wang, Optimization of ionic liquid based microwave
assisted extraction of five anthraquinones from Polygonum cuspidatum by response
surface method, Zhong Yao Cai. 37 (2014) 871-875.
[5] C. Lu, X. Luo, R. Luo, X. Chen, L. Xing, Z. Tang, H. Li, Assessment of antibacterial
properties and the active ingredient of plant extracts and its effect on the performance of
crucian carp (Carassius auratus gibelio var. E'erqisi, Bloch), J Sci Food Agric. 93 (2013)
902-909.
[6] T.C. Lo, H.C. Nian, K.H. Chiu, A.Y. Wang, B.Z. Wu, Rapid and efficient purification
of chrysophanol in Rheum Palmatum LINN by supercritical fluid extraction coupled with
preparative liquid chromatography in tandem, J Chromatogr B Analyt Technol Biomed
Life Sci. 893-894 (2012) 101-106.
29