0090-9556/99/2703-0365–372$02.

00/0
DRUG METABOLISM AND DISPOSITION Vol. 27, No. 3
Copyright © 1999 by The American Society for Pharmacology and Experimental Therapeutics Printed in U.S.A.

STUDIES ON CYTOCHROME P-450-MEDIATED BIOACTIVATION OF DICLOFENAC IN

RATS AND IN HUMAN HEPATOCYTES: IDENTIFICATION OF GLUTATHIONE
CONJUGATED METABOLITES

WEI TANG, RALPH A. STEARNS, STELVIO M. BANDIERA, YONG ZHANG, CONRAD RAAB, MATTHEW P. BRAUN,
DENNIS C. DEAN, JIANMEI PANG, KWAN H. LEUNG, GEORGE A. DOSS, JOHN R. STRAUSS, GLORIA Y. KWEI,
THOMAS H. RUSHMORE, SHUET-HING L. CHIU, AND THOMAS A. BAILLIE

Department of Drug Metabolism, Merck Research Laboratories, Rahway, New Jersey (W.T., R.S., Y.Z., C.R., M.B., D.D., J.P., K.H.L., G.A.D.,
J.R.S., G.Y.K., S-H.L.C., T.A.B.); Faculty of Pharmaceutical Sciences, University of British Columbia, Vancouver, British Columbia, Canada
(S.M.B.); and Department of Drug Metabolism, Merck Research Laboratories, West Point, Pennsylvania (T.H.R., T.A.B.)

(Received September 4, 1998; accepted December 2, 1998)

This paper is available online at http://www.dmd.org

Downloaded from dmd.aspetjournals.org at ASPET Journals on April 13, 2018

ABSTRACT:

The nonsteroidal anti-inflammatory drug diclofenac causes a rare
but potentially fatal hepatotoxicity that may be associated with the
formation of reactive metabolites. In this study, three glutathione
(GSH) adducts, namely 5-hydroxy-4-(glutathion-S-yl)diclofenac
(M1), 4*-hydroxy-3*-(glutathion-S-yl)diclofenac (M2), and 5-hy-
droxy-6-(glutathion-S-yl)diclofenac (M3), were identified by liquid
chromatography-tandem mass spectrometry analysis of bile from
Sprague-Dawley rats injected i.p. with a single dose of diclofenac
(200 mg/kg). These adducts presumably were formed via hepatic
cytochrome P-450 (CYP)-catalyzed oxidation of diclofenac to re-
active benzoquinone imines that were trapped by GSH conjuga-
tion. In support of this hypothesis, M1, M2, and M3 were generated
from diclofenac in incubations with rat liver microsomes in the
presence of NADPH and GSH. Increases in adduct formation were
observed when incubations were performed with liver microsomes
from phenobarbital- or dexamethasone-treated rats. Adduct for-
mation was inhibited by polyclonal antibodies against CYP2B,
CYP2C, and CYP3A (40–50% inhibition at 5 mg of IgG/nmol of CYP)
but not by an antibody against CYP1A. Maximal inhibition was
obtained when the three inhibitory antibodies were used in a cock-
tail fashion (70–80% inhibition at 2.5 mg of each IgG/nmol of CYP).
These data suggest that diclofenac undergoes biotransformation
to reactive metabolites in rats and that CYP isoforms of the 2B, 2C,
and 3A subfamilies are involved in this bioactivation process. With
respect to CYP2C isoforms, rat hepatic CYP2C7 and CYP2C11
were implicated as mediators of the bioactivation based on immu-
noinhibition studies using antibodies specific to CYP2C7 and
CYP2C11. Screening for GSH adducts also was carried out in
human hepatocyte cultures containing diclofenac, and M1, M2,
and M3 again were detected. It is possible, therefore, that reactive
benzoquinone imines may be formed in vivo in humans and con-
tribute to diclofenac-mediated hepatic injury.

Diclofenac is a nonsteroidal anti-inflammatory drug that in rare
cases causes severe liver injury (Breen et al., 1986; Helfgott et al.,
1990; Purcell et al., 1991). The hepatotoxicity has been described as
idiosyncratic in nature and possibly associated with metabolism of the
drug (Banks et al., 1995; Boelsterli et al., 1995). Conjugation with
glucuronic acid represents a major biotransformation pathway for
diclofenac and the resulting acyl glucuronide has been implicated as
a mediator of the hepatotoxicity (Boelsterli et al., 1995). With immu-
nochemical detection, diclofenac was found to form protein adducts in
the liver of treated mice and rats (Pumford et al., 1993; Hargus et al.,
1994) as well as in hepatocyte cultures (Kretz-Rommel and Boelsterli,
1993; Gil et al., 1995). The formation of diclofenac-protein adducts
was dependent upon uridine diphosphate glucuronosyltransferase ac-
tivity and the covalently modified proteins were shown to retain the
glucuronic acid moiety (Hargus et al., 1994; Kretz-Rommel and
A preliminary account of this study was presented at Experimental Biology,
San Francisco, CA, April 1998.
Send reprint requests to: Dr. Wei Tang, Department of Drug Metabolism,
Merck & Co., P.O. Box 2000, RY80L-109, Rahway, NJ 07065. E-mail:
wei tang@merck.com
365
Boelsterli, 1994). In rat liver, the protein targets for diclofenac binding
included 110-, 140-, and 200-kDa plasma membrane proteins and a
60-kDa microsomal protein. The 110-kDa protein was identified as
dipeptidyl peptidase IV (Hargus et al., 1995). Data also were obtained
that suggest that reactive metabolite(s) of diclofenac could be gener-
ated through a cytochrome P-450 (CYP)1-mediated pathway (Kretz-
Rommel and Boelsterli, 1993; Shen et al., 1997a, b). For example, a
51-kDa protein was found to be modified covalently in male rats
dosed with diclofenac and this modification, which also was observed
in incubations with rat liver microsomes, was a NADPH-dependent
process (Hargus et al., 1994; Shen et al., 1997a). The 51-kDa protein
subsequently was identified as CYP2C11 (Shen et al., 1997a). In rat
hepatocyte cultures, the diclofenac-induced leakage of lactate dehy-
drogenase was reduced markedly in the presence of CYP2C-selective
inhibitors, suggesting that a CYP2C enzyme(s) was (were) involved in
1
Abbreviations used are: CYP, cytochrome P-450; CID, collision-induced
dissociation; Dex, dexamethasone; GSH, glutathione; LC/MS/MS, liquid chroma-
tography-tandem mass spectrometry; M1, 5-OH-4-GS-diclofenac or 5-hydroxy-
4-(glutathion-S-yl)diclofenac; M2, 49-OH-39-GS-diclofenac or 49-hydroxy-39-(glu-
tathion-S-yl)diclofenac; M3, 5-OH-6-GS-diclofenac or 5-hydroxy-6-(glutathion-S-
yl)diclofenac; PB, phenobarbital; TFA, trifluoroacetic acid.
366 TANG ET AL.
FIG. 1. LC/MS/MS detection of the GSH adducts of diclofenac by constant
neutral loss scanning (loss of 129 Da).
An aliquot of bile (50 ml) from the rat treated with diclofenac (200 mg/kg) was
acidified and injected onto a Zorbax C8 column.
the drug-associated cytotoxicity (Kretz-Rommel and Boelsterli,
1993).
The CYP-catalyzed biotransformation of diclofenac gives rise to
two monohydroxylated metabolites, namely, 49-hydroxydiclofenac
and 5-hydroxydiclofenac (Leemann et al., 1993; Shen et al., 1997b).
It was proposed that the 5-hydroxy derivative underwent further
oxidation to a putative benzoquinone imine (Shen et al., 1997b). By
virtue of its electrophilic nature, this intermediate could react with
proteins or cellular glutathione (GSH). From an analytical point of
view, characterization of the GSH adduct(s) of such reactive metab-
olites would provide insight into the nature of the short-lived electro-
philic species (Baillie and Davis, 1993). In this paper, we describe the
identification of GSH-conjugated metabolites in the bile from rats
treated with diclofenac. The role of rat hepatic CYP enzymes in the
formation of these metabolites was investigated using rat liver micro-
somes and antibodies against CYP enzymes and screening for GSH
adducts also was carried out in human hepatocyte cultures treated with
diclofenac.
Materials and Methods
Chemicals. Diclofenac, dimethyl sulfoxide, GSH, and NADPH were pur-
chased from Sigma Chemical Co. (St. Louis, MO). Trifluoroacetic acid (TFA)
was obtained from Fisher Scientific (Fair Lawn, NJ) and bis(pinacolato)dibo-
ron was purchased from Strem Chemical (Newburyport, MA). All other
chemicals were obtained from Aldrich Chemical Co. (Milwaukee, WI). Bon-
dElut C18 solid phase extraction cartridge columns were obtained from Varian
Chromatography Systems (Walnut Creek, CA).
Instrumentation and Analytical Methods. Liquid chromatography-tan-
dem mass spectrometry (LC/MS/MS) was carried out on a SCIEX API III1
tandem mass spectrometer (Perkin-Elmer, Toronto, Canada) interfaced to a
HPLC system consisting of two LC-10A pumps and a static-bed mixer (Shi-
madzu Scientific Instruments, Kyoto, Japan). LC/MS/MS experiments were
performed with an ion spray interface with positive ion detection at a voltage
of 5 kV. The orifice potential was 65 V. Collision-induced dissociation (CID)
used argon as the collision gas at a thickness of 1.3 3 1014 atoms/cm2 and the
collision energy was 30 eV. Chromatography was performed on a Zorbax
Rx-C8 column (4.6 3 250 mm, 5 mm; DuPont, Wilmington, DE) and samples
were delivered at a flow rate of 1 ml/min with a 1:25 split. The mobile phase
consisted of aqueous acetonitrile containing 10% methanol and 0.05% TFA
and was programmed by a linear increase from 10 to 70% acetonitrile during
a 30-min period.
HPLC purification of synthetic products was performed either on a Dy-
namax SD-200 liquid chromatograph (Rainin Instruments, Woburn, MA) with
a Zorbax RX-C18 column (21.2 3 250 mm, 5 mm) at a flow rate of 20 ml/min
or a LC-10AD liquid chromatograph (Shimadzu Scientific Instruments) with a
Zorbax Rx-C8 column (4.6 3 250 mm, 5 mm) at a flow rate of 1 ml/min. UV
detection was at 210 nm.
NMR spectra of synthetic products were obtained on either a Varian
Inova400 or a Varian Inova500 spectrometer operating at 400 and 500 MHz,
Downloaded from dmd.aspetjournals.org at ASPET Journals on April 13, 2018
respectively, and chemical shifts were expressed relative to tetramethylsilane.
Synthesis of Diclofenac Metabolites. 49-Hydroxy-39-(glutathion-S-yl)di-
clofenac (49-OH-39-GS-diclofenac, M2). 49-Hydroxydiclofenac was generated
from diclofenac in incubations with a baculovirus-insect cell line expressing
CYP2C9. The product was purified by the Dynamax HPLC system (Rainin
Instruments; mobile phase, isocratic 35% acetonitrile containing 0.1% TFA;
retention time, 65 min). LC/MS, m/z 312 (MH1). 1H NMR (CD3OD), d 3.6 (s,
CH2COOH); 6.22 (d, J 5 8.5 Hz, 3-CH); 6.74 (t, J 5 8.5 Hz, 5-CH); 6.86 (s,
39 and 59-CH); 6.94 (t, J 5 8.5 Hz, 4-CH); 7.16 (d, J 5 8.5 Hz, 6-CH).
Fifty-eight milligrams of silver(I) oxide was added to 3 mg of 49-
hydroxydiclofenac in 1 ml of benzene. The reaction mixture was stirred at
room temperature for 22 h and the resulting solution was treated with 36 mg
of GSH in 1 ml of phosphate buffer (pH 7.4). The two-phase reaction mixture
was stirred vigorously at 37°C for 8 h. The aqueous phase was separated and
the crude product purified by the Shimadzu HPLC system (mobile phase,
aqueous acetonitrile containing 10% methanol and 0.05% TFA, linear increase
from 10% to 70% acetonitrile during a 30 min period; retention time, 14 min).
LC/MS: m/z 617 (MH1). [1H] NMR (D2O), d 2.0 to 2.2 (m, Glu, CH2); 2.4 to
2.5 (m, Glu, CH2); 3.2 to 3.3 and 3.4 to 3.6 (m, Cys, CH2); 3.7 to 3.9 (m, Gly,
CH2, Glu, CH, CH2COOH); 4.3 (m, Cys, CH); 6.36 (d, J 5 8.4 Hz, 3-CH);
6.94 (t, J 5 8.4 Hz, 5-CH); 7.16 (t, J 5 8.4 Hz, 4-CH); 7.19 (s, 59-CH); 7.28
(d, J 5 8.4 Hz, 6-CH).
5-Hydroxy-4-(glutathion-S-yl)diclofenac (5-OH-4-GS-diclofenac; M1) and
5-hydroxy-6-(glutathion-S-yl)diclofenac (5-OH-6-GS-diclofenac; M3). A
877-mg quantity of iodine monochloride in 4 ml of acetic acid was added to
1.6 g of diclofenac in 35 ml of acetic acid. The mixture was stirred for 1.5 h
and then mixed with 6 ml of 20% aqueous sodium persulfate. The solid
product, 5-iododiclofenac, was collected and washed with 8 ml of 50%
ethanol.
Seventy milligrams of bis(pinacolato)diboron, 5.6 mg of palladium(II)
chloride, and 74 mg of potassium acetate under nitrogen were added to 5.5 mg
of 5-iododiclofenac in dimethyl sulfoxide. The mixture was stirred at 60 –70°C
for 15 h, acidified to pH 3 to 4, and extracted with ethyl acetate. The product,
5-(tetramethyldioxaboron)diclofenac, was purified by the Dynamax HPLC
system (mobile phase, isocratic 47% acetonitrile containing 0.1% TFA; reten-
tion time, 45 min).
Seven milligrams of sodium bicarbonate and 0.2 ml of acetone were added
to 4 mg of 5-(tetramethyldioxaboron)diclofenac in 0.5 ml of 1.2% aqueous
sodium hydroxide. The mixture was cooled to 0°C and a total of 0.5 ml of 6
mg of Oxone in 0.4 mM EDTA solution was added dropwise. The mixture was
stirred for an additional 5 min at 0°C, mixed with 21 ml of 20% aqueous
sodium persulfate, acidified to pH 2 to 3, and extracted with ethyl acetate. The
product, 5-hydroxydiclofenac, was purified by the Dynamax HPLC system
(mobile phase, isocratic 37% acetonitrile containing 0.1% TFA; retention time,
14.3 min). LC/MS, m/z 312 (MH1); 1H NMR (CD3OD): d 3.6 (s, CH2COOH);
6.31 (d, J 5 8.5 Hz, 3-CH); 6.46 (dd, J 5 1.3 and 8.5 Hz, 4-CH); 6.70 (d, J 5
1.3 Hz, 6-CH); 6.92 (t, J 5 8.0 Hz, 49-CH); 7.31 (d, J 5 8.0 Hz, 39 and 59-CH).
Fifty-eight milligrams of silver(I) oxide was added to 3 mg of 5-hydroxy-
 BIOACTIVATION OF DICLOFENAC IN RATS 367

Downloaded from dmd.aspetjournals.org at ASPET Journals on April 13, 2018

FIG. 2. a, LC/MS/MS product ion spectra of isomeric GSH adducts detected in bile from rats treated with diclofenac at 200 mg/kg: A, 5-OH-4-GS-diclofenac (M1)
and B, 5-OH-6-GS-diclofenac (M3).
The spectra were obtained by CID of the MH1 ions at m/z 617 and the proposed origins of key fragment ions are as indicated. b, LC/MS/MS product ion spectrum of
a GSH adduct detected in bile from rats treated with diclofenac at 200 mg/kg: 49-OH-39-GS-diclofenac (M2). The spectrum was obtained by CID of the MH1 ion at m/z
617 and the proposed origins of key fragment ions are as indicated.

diclofenac in 1 ml of benzene. The reaction mixture was stirred at room by the Shimadzu HPLC system (mobile phase, aqueous acetonitrile containing
temperature for 22 h and the resulting solution was treated with 36 mg of GSH 10% methanol and 0.05% TFA, linear increase from 10% to 70% acetonitrile
in 1 ml of phosphate buffer (pH 7.4). The two-phase reaction mixture was during a 30-min period; retention time, 13.8 min for M1 and 15.0 min for M3).
stirred vigorously at 37°C for 8 h. Two products (M1 and M3) were purified M1, LC/MS, m/z 617 (MH1); 1H NMR (CD3OD), d 2.0 to 2.1 (m, Glu, CH2),
368 TANG ET AL.
FIG. 3. LC/MS/MS detection of GSH adducts in incubations containing diclofenac, rat liver microsomes, NADPH, and GSH.
Four mass transitions were used as criteria for metabolite identification: m/z 617 3 524, 617 3 488, 617 3 342, and 617 3 324. Substrate concentrations were A,
1 mM and B, 1000 mM.
2.3 to 2.5 (m, Glu, CH2), 3.1 to 3.2, and 3.4 to 3.5 (m, Cys, CH2), 3.7 to 3.9
(m, Gly, CH2, Glu, CH, CH2COOH), 4.5 to 4.6 (m, Cys, CH), 6.54 (s, 3-CH),
6.82 (s, 6-CH), 7.0 (t, J 5 8.5 Hz, 49-CH), 7.37 (d, J 5 8.5 Hz, 39 and 59-CH).
M3, LC/MS: m/z 617 (MH1); 1H NMR (CD3OD): d 2.1 to 2.2 (m, Glu, CH2),
2.4 to 2.6 (m, Glu, CH2), 3.2 to 3.3, 3.4 to 3.6 (m, Cys, CH2), 3.6 to 3.8 (m,
Gly, CH2, Glu, CH, CH2COOH), 4.3 to 4.4 (m, Cys, CH), 6.44 (d, J 5 8.5 Hz,
3-CH), 6.70 (d, J 5 8.5 Hz, 6-CH), 6.98 (t, J 5 8.0 Hz, 4-CH), 7.34 (d, J 5
8.0 Hz, 39 and 59-CH).
Animal Experiments. Experiments were performed according to proce-
dures approved by the Merck Institutional Animal Care and Use Committee.
Male and female Sprague-Dawley rats purchased from Harlan Laboratories
(Indianapolis, IN) and weighing 270 to 360 g were allowed free access to
commercial rat chow and water. They were anesthetized with sodium pento-
barbital (nembutal) and their bile ducts were cannulated with PE-10 tubing.
Bile was collected before treatment. An aqueous solution (pH 7) of diclofenac
was then administered at either 10 or 200 mg/kg by i.p. injection and bile was
collected for an additional 8 h.
A group of seven female rats was dosed with aqueous phenobarbital (PB) at
100 mg/kg per day for 3 days. Another group of 20 male rats was dosed with
dexamethasone (Dex) in 2% Tween 80 at 200 mg/kg/day for 3 days. These rats
were used for the isolation of liver microsomes.
Biological Preparations. Rat liver microsomes were isolated by differential
centrifugation (Raucy and Lasker, 1991) and pooled from 1) 40 naive male rats
(control); 2) 30 naive female rats; 3) 7 PB-treated rats; and 4) 20 Dex-treated
rats, respectively.
Polyclonal antibodies directed against rat CYP enzymes were prepared in
rabbits by immunization with the electrophoretically homogeneous proteins
(Bandiera and Dworschak, 1992; Levine et al., 1998; Wong and Bandiera,
1998). Monospecific antibodies were prepared by passing the polyspecific
antibodies through a series of columns containing partially purified CYP
enzymes excluding those CYPs the antibodies were raised against (Bandiera
and Dworschak, 1992; Levine et al., 1998).
Incubations with Rat Liver Microsomes. Diclofenac and GSH in phos-
phate buffer (pH 7.4) were added to rat liver microsomes (0.5–2.0 nmol of
CYP/ml) suspended in 0.1 M phosphate buffer (pH 7.4) containing EDTA (1
mM). Substrate concentrations were 1, 5, 10, 25, 50, 200, and 1000 mM and
the GSH concentration was 5 mM. The mixture was incubated at 37°C for 5
min before adding NADPH in phosphate buffer (1 mg/ml final concentration)
Downloaded from dmd.aspetjournals.org at ASPET Journals on April 13, 2018
to initiate the reaction. After an additional 30-min incubation, the reaction was
quenched with 10% aqueous TFA.
In immunoinhibition experiments, rat liver microsomes (0.5 nmol of CYP/
ml) were preincubated with each antibody (0.5, 1.0, 2.5, and 5.0 mg of
IgG/nmol of CYP) for 15 min at room temperature. Control incubations
contained IgG from untreated rabbits. Substrate concentration was 5 mM.
Diclofenac, GSH, and NADPH were added thereafter and the incubations were
performed in a manner similar to that described above.
Incubations with Human Hepatocyte Cultures. The human liver samples
were obtained from the Pennsylvania Regional Tissue Bank (Exton, PA). An
agreement was made between the tissue bank and Merck & Co. for research
use of the samples. The death of one donor, a 25-year-old male (ID no.
1130971), had been caused by an accident; the second donor, a 65-year-old
male (ID no. 0514981), died from anoxia. Hepatocytes were isolated based on
a two-step perfusion procedure (Pang et al., 1997). Upon isolation, the hepa-
tocytes were seeded on Matrigel-coated 6-well plates at 2 3 106 cells/ml and
cultured in Willams’ medium containing 0.1 mM Dex for 48 h before meta-
bolic studies.
Diclofenac dissolved in dimethyl sulfoxide was added to the hepatocyte
cultures to give a final drug concentration of 300 mM. Dimethyl sulfoxide
concentration was 0.1% (v/v). After incubation for 24 h, the culture medium
was acidified with 10% aqueous TFA.
Detection of GSH-Conjugated Metabolites. Rat bile (200 ml) was acidi-
fied with 10% aqueous TFA and precipitates were removed via centrifugation
at 13,600g for 5 min. Aliquots of bile (20 –50 ml) were injected onto the
Zorbax Rx-C8 column and analyzed by LC/MS/MS. Metabolites were iden-
tified based on their fragmentation upon CID and on their HPLC retention
times compared with those of reference compounds obtained by synthesis.
Samples from microsomal incubations or from hepatocyte cultures were
applied to a C18 extraction cartridge column that was prewashed with methanol
and water. The column was washed consecutively with water and methanol.
The methanol eluate was evaporated to dryness under a stream of nitrogen and
the residue was reconstituted in 300 ml of 60% aqueous acetonitrile containing
0.05% TFA. An 80-ml aliquot of the solid phase extracts was injected onto the
Zorbax Rx-C8 column and analyzed by LC/MS/MS. Identification of the
metabolites was based on multiple reaction monitoring detection of four
transitions, namely, m/z 617 3 542, 617 3 488, 617 3 342, and 617 3 324.
 BIOACTIVATION OF DICLOFENAC IN RATS 369
TABLE 1
Inhibition by anti-CYP antibodies of the metabolic activation of diclofenac in
incubations with rat liver microsomesa
Rat Liver % Controlb
Antibody IgG Diclofenac
Microsomes Total (M1, M3)

mg/nmol CYP mM
Anti-1A IgG 5 Male 5 100 (100, 100)
Anti-2B IgG 5 Male 5 67 (70, 63)
Anti-2C IgG 5 Male 5 45 (50, 40)
Anti-3A IgG 5 Male 5 55 (58, 52)
Anti-2B/2C/3A IgG 2.5 each Male 5 22 (20, 23)
Anti-2C11 IgG 5 Male 5 62 (64, 59)
Anti-2C7 IgG 5 Male 5 46 (47, 45)
Anti-1A IgG 5 Female 5 100 (100, 100)
Anti-2B IgG 5 Female 5 47 (47, 48)
Anti-2C IgG 5 Female 5 50 (55, 45)
Anti-3A IgG 5 Female 5 58 (59, 57)
Anti-2B/2C/3A IgG 2.5 each Female 5 30 (30, 31)
Anti-2C11 IgG 5 Female 5 97 (95, 100)
Anti-2C7 IgG 5 Female 5 42 (44, 40)
a
Diclofenac and GSH in phosphate buffer were added to rat liver microsomes suspended in

Downloaded from dmd.aspetjournals.org at ASPET Journals on April 13, 2018

phosphate buffer (0.1 M; pH 7.4) containing EDTA. Each anti-CYP IgG was preincubated with
microsomes for 15 min at room temperature. Control incubations contained preimmune IgG.
Reactions were initiated by adding NADPH and proceeded for an additional 30 min. The
products were analyzed by LC/MS/MS.
b
% Control (total) was calculated based on decreases in the formation of M1 and M3. %
Control also was calculated individually for M1 and M3 and the data are presented in paren-
theses.

lites were arbitrarily designated as M1, M2, and M3 according to their

FIG. 4. Inhibition of diclofenac metabolism by polyclonal antibodies against
CYP enzymes in incubations with liver microsomes isolated from untreated male
rats (top) and female rats (bottom).
Percentage of control was calculated based on decreases in the formation of M1
and M3.
Results
Characterization of the GSH Adducts Formed In Vivo in Rats.
LC/MS/MS screening for GSH adducts was performed by constant
neutral loss scan monitoring of ions that lose 129 Da upon CID
(Baillie and Davis, 1993). Three components that exhibited this re-
sponse were detected in a bile sample from the rat treated with
diclofenac (Fig. 1) and were assigned as diclofenac metabolites based
on their characteristic chlorine isotope cluster (m/z 617/619). The
associated parent ions were all at m/z 617, consistent with these
metabolites being GSH adducts of hydroxydiclofenac. The metabo-
relative HPLC retention times.
Subsequent CID of the MH1 ions at m/z 617 produced product ions
at m/z 542 and 488 resulting from neutral losses of glycine (75 Da)
and pyroglutamate (129 Da), respectively (Fig. 2). These neutral
losses are characteristic for xenobiotic GSH adducts (Baillie and
Davis, 1993). One major common fragment ion in the spectra of
adducts was found at m/z 342, which could be derived from cleavage
adjacent to the thioether moiety and charge retention on the aromatic
moiety (Fig. 2). An additional loss of water from the m/z 342 ion
would give rise to the observed product ion at m/z 324.
Further structural assignment for the metabolites was obtained via
comparison with synthetic reference compounds. Thus, identical
MS/MS fragmentation patterns and HPLC retention times were ob-
served for M1 and 5-OH-4-GS-diclofenac. The same held true for M2
and 49-OH-39-GS-diclofenac and for M3 and 5-OH-6-GS-diclofenac.
Metabolite M2 was detected only in bile from rats treated with a
high dose of diclofenac (200 mg/kg), whereas M1 and M3 were found
in rats treated with either the low or high dose (10 or 200 mg/kg).
Similar results were observed for both male and female rats.
Metabolite Formation in Incubations with Rat Liver Micro-
somes. The detection by LC/MS/MS of diclofenac metabolites formed
in microsomal incubations was based on multiple reaction monitoring
of four characteristic mass transitions. These transitions coincided
with the HPLC retention times for M1, M2, and M3 upon analysis of
samples derived from incubations of diclofenac with rat liver micro-
somes fortified with NADPH and GSH (Fig. 3). The metabolite
profiles were similar throughout a wide range of substrate concentra-
tions from 1 to 1000 mM (Fig. 3). Adduct M2 was a minor metabolite
in incubations at all substrate concentrations. The formation of M1
and M3 increased with increasing substrate concentration, reaching a
plateau at approximately 5 mM diclofenac (data not shown). Similar
to results observed in vivo, the GSH adducts were detected in incu-
bations containing liver microsomes isolated from either male or
female rats.
As compared with liver microsomes from untreated rats, an approx-
imately 2-fold increase in the yield of metabolites was observed when
diclofenac (50 mM) was incubated with liver microsomes isolated
370
FIG. 5. Inhibition of diclofenac metabolism by monospecific antibodies against
CYP2C7 and CYP2C11 in incubations with liver microsomes isolated from
untreated male rats (top) and female rats (bottom).
Percentage of control was calculated based on decreases in the formation of M1
and M3.
from PB- or Dex-treated rats. The comparison was made based on
microsomal protein concentrations.
Metabolite M2 also was detected in incubations of 49-hydroxydi-
clofenac with rat liver microsomes in the presence of NADPH and
GSH; M1 and M3 were formed in incubations containing 5-hydroxy-
diclofenac (data not shown).
Immunoinhibition of Rat Hepatic CYP-Mediated Bioactiva-
tion. The formation of metabolites in incubations containing 5 mM
diclofenac and 0.25 nmol of rat liver microsomal CYP (untreated) was
linear over a period of 30 min, with M1 and M3 being the predomi-
nant products. Inhibition experiments were performed using 30 min as
the end point and the percentage of inhibition was calculated based on
decreases in the formation of M1 and M3 relative to controls that
lacked the antibody.
TANG ET AL.
Downloaded from dmd.aspetjournals.org at ASPET Journals on April 13, 2018
FIG. 6. LC/MS/MS detection of GSH adducts in human hepatocyte cultures
treated with diclofenac.
Four mass transitions were used as criteria for metabolite identification: m/z 617
3 524, 617 3 488, 617 3 342, and 617 3 324. The donor was a 25-year-old male.
Metabolite formation in incubations with rat liver microsomes was
inhibited by polyclonal antibodies against CYP2B, CYP2C, and
CYP3A (Fig. 4), whereas no effect was observed with the antibody
against CYP1A (Table 1). For liver microsomes from male rats,
maximal inhibition occurred at the highest IgG concentration used
(Fig. 4, top). For liver microsomes from female rats, maximal inhi-
bition was achieved at approximately 2.5 mg of IgG/ nmol of CYP
(Fig. 4, bottom). When antibodies against CYP2B, CYP2C, and
CYP3A were used in a cocktail fashion, the inhibition was 70 to 80%,
which was greater than that achieved by any single antibody (Table 1).
The metabolism of diclofenac was inhibited by a monospecific
antibody against CYP2C7 (Fig. 5). In this case, maximal inhibition
occurred at the highest IgG concentration used (Fig. 5 and Table 1).
Diclofenac metabolism in incubations with liver microsomes from
male rats also was inhibited by a monospecific antibody against
CYP2C11 (Fig. 5 and Table 1). No effect was observed with the
antibody against CYP2C11 when liver microsomes from female rats
were used (Fig. 5). This finding is consistent with the specificity of the
antibody, because CYP2C11 is known to be expressed only in male
rats.
Formation of GSH Adducts in Human Hepatocyte Cultures.
 BIOACTIVATION OF DICLOFENAC IN RATS
Scheme 1. Proposed metabolic pathways leading to the formation of GSH-conjugated metabolites of diclofenac through oxidative biotransformation.
Similarly, the detection of diclofenac metabolites in hepatocyte cul-
tures was based on MS/MS multiple reaction monitoring coupled with
HPLC separation. Hepatocytes isolated from the two human liver
samples exhibited a viability of greater than 80% as determined by the
trypan blue exclusion test. Metabolites M1, M2, and M3 were de-
tected readily in incubations of diclofenac with human hepatocyte
cultures (Fig. 6).
Discussion
Protein adducts have been identified in rats treated with diclofenac,
but information on the structures of the reactive intermediates formed
via CYP-mediated oxidation of diclofenac has been elusive. In this
study, LC/MS/MS data were obtained indicating the presence of
GSH-conjugated metabolites in bile from either male or female rats
dosed with diclofenac. These conjugated metabolites subsequently
were identified as 5-OH-4-GS-diclofenac (M1), 49-OH-39-GS-di-
clofenac (M2), and 5-OH-6-GS-diclofenac (M3).
Two types of intermediate electrophiles that potentially could react
with GSH during the metabolism of diclofenac are benzoquinone
imines (Brune and Lindner, 1992) and arene oxides (Blum et al.,
1996; Scheme 1). Analysis of the metabolites formed in vivo in rats
suggested that diclofenac most likely was oxidized on the dichloroa-
niline ring to the 19,49-benzoquinone imine that was trapped by GSH
because only M2 was present in the bile. A similar process could
occur on the phenylacetic acid moiety to give M1 and M3 via the
common intermediate diclofenac-2,5-quinone imine (Scheme 1). This
hypothesis was further supported by the fact that M2 was synthesized
chemically from 49-hydroxydiclofenac and M1 and M3 were prepared
from 5-hydroxydiclofenac. In vitro incubation of 49-hydroxydiclofe-
nac with rat liver microsomes in the presence of NADPH and GSH
resulted in the formation of M2, whereas incubation of 5-hydroxydi-
clofenac produced M1 and M3. The 49- and 5-hydroxylated deriva-
tives are viewed as the biological precursors of benzoquinone imines.
Differences in the formation of M1 and M3 (Table 1) are most likely
the result of preferential attack by GSH on one of the reactive centers
of diclofenac-2,5-quinone imine.
371
Downloaded from dmd.aspetjournals.org at ASPET Journals on April 13, 2018
It is proposed that the metabolic activation of diclofenac through an
oxidative pathway is catalyzed by CYP enzymes (Scheme 1). Con-
sistent with this hypothesis was the observation that the formation of
GSH adducts in incubations with rat hepatic microsomes from either
sex was NADPH-dependent and was inhibited by antibodies against
CYP enzymes. Inhibitory antibodies included polyclonal antibodies
against CYP2B, CYP2C, and CYP3A but not an antibody against
CYP1A, suggesting that the bioactivation process was catalyzed by
CYP isoforms in rat hepatic 2B, 2C, and 3A subfamilies. Among these
enzymes, CYP2B1/2 are inducible upon treatment of rats with PB and
CYP3A1/2 can be induced by Dex as well as by PB (Correia, 1995).
These data are in accord with the observations that adduct formation
was higher in incubations with liver microsomes from PB- or Dex-
treated rats.
With respect to CYP2C isoforms, male-specific CYP2C11 was
shown previously to be modified covalently by unknown reactive
metabolites of diclofenac (Shen et al., 1997a). In this study, CYP2C11
was found to be involved in catalyzing the oxidation of diclofenac to
reactive benzoquinone imines because the formation of GSH adducts
was inhibited by the antibody against CYP2C11. It is tempting to
speculate that the benzoquinone imine intermediates may serve to
arylate CYP2C11. Similarly, CYP2C7 also was implicated in diclofe-
nac bioactivation based on immunoinhibition studies with the anti-
body against CYP2C7. Whether CYP2C7 is modified by the reactive
metabolite(s) remains to be determined.
It is of interest to note that the GSH-conjugated metabolites were
detected in both male and female rats as well as in incubations with
liver microsomes from rats of either sex. In other words, gender
difference was not apparent in terms of the CYP-catalyzed bioactiva-
tion of diclofenac in rats as measured by GSH adduct formation. In an
earlier study, the CYP-mediated formation of a 51-kDa protein adduct
was found to occur only in male rats treated with diclofenac and in
vitro protein adduct formation was not attenuated by adding GSH to
the incubations (Shen et al., 1997a). Further investigation is warranted
to determine whether the formation of GSH conjugated metabolites
372 TANG ET AL.
and the formation of protein adducts share the same bioactivation
pathway(s).
Primary cell cultures are widely regarded as good models for the
corresponding in vivo systems (Li and Kedderis, 1997). In this inves-
tigation, the GSH adducts detected in rats also were identified in
human hepatocyte cultures containing diclofenac, suggesting that the
CYP-mediated bioactivation of diclofenac could occur in vivo in
humans. The resulting electrophilic benzoquinone imines probably are
capable of arylating proteins, potentially leading to toxic conse-
quences either by altering protein functions or by provoking immune
responses. Alternatively, the reactive intermediates could be scav-
enged by their reaction with GSH. However, GSH conjugation not
only represents a detoxification process but may also lead to depletion
of cellular GSH pools. The resulting GSH deficiency, caused by the
exposure to reactive compounds, is known to result in cell injury and
death due largely to the impaired antioxidant function of the GSH
redox system (Reed, 1990). In view of the idiosyncratic nature of
diclofenac-mediated hepatotoxicity, potential damage caused by re-
active metabolites through GSH depletion should be considered, par-
ticularly in patients who, because of genetic and/or environmental
factors, may be already compromised with respect to GSH.
In summary, three GSH adducts were identified in bile from rats
treated with diclofenac. The metabolites most likely were formed
through rat hepatic CYP-catalyzed oxidation of diclofenac to putative
benzoquinone imine intermediates followed by conjugation with
GSH. These findings may be relevant to diclofenac hepatotoxicity
because the same adducts were detected in human hepatocyte cultures
incubated with the drug. Current studies are in progress to investigate
the involvement of human hepatic CYP isoforms in diclofenac bio-
activation.
Acknowledgments. We thank Dr. Anthony Y. H. Lu (Rutgers
University) and Regina W. Wang (Merck Research Laboratories) for
valuable discussions.
References
Baillie TA and Davis MR (1993) Mass spectrometry in the analysis of glutathione conjugates.
Biol Mass Spectrom 22:319 –325.
Bandiera SM and Dworschak C (1992) Effects of testosterone and estrogen on hepatic levels of
cytochromes P450 2C7 and P450 2C11 in the rat. Arch Biochem Biophys 196:186 –195.
Banks AT, Zimmerman HJ, Ishak KG and Harter JG (1995) Diclofenac-associated hepatotox-
icity: Analysis of 180 cases reported to the food and drug administration as adverse reactions.
Hepatology 22:820 – 827.
Blum W, Faigle JW, Pfaar U and Sallmann A (1996) Characterization of a novel diclofenac
metabolite in human urine by capillary gas chromatography-negative chemical ionization mass
spectrometry. J Chromatogr B 685:251–263.
Boelsterli UA, Zimmerman HJ and Kretz-Rommel A (1995) Idiosyncratic liver toxicity of
nonsteroidal antiinflammatory drugs: Molecular mechanisms and pathology. Crit Rev Toxicol
25:207–235.
Breen EG, McNicholl J, Cosgrove E, McCabe J and Stevens FM (1986) Fatal hepatitis associated
with diclofenac. Gut 27:1390 –1393.
Brune K and Lindner J (1992) Increased liver toxicity of diclofenac by paracetamol: Results and
possible mechanism, in Side-Effects of Anti-inflammatory Drugs (Rainsford KD and Velo GP
eds) vol 3, 198 –203, Kluwer Academic Publishers, Dordrecht.
Correia MA (1995) Rat and human liver cytochromes P450: Substrate and inhibitor specificities
and functional markers, in Cytochrome P450: Structure, Mechanism, and Biochemistry, 2nd
ed. (Ortiz de Montellano PR ed) pp 607– 630, Plenum Press, New York.
Gil ML, Ramirez MC, Terencio MC and Castell JV (1995) Immunochemical detection of protein
adducts in cultured human hepatocytes exposed to diclofenac. Biochim Biophys Acta 1272:
140 –146.
Hargus SJ, Amouzedeh HR, Pumford NR, Myers TG, McCoy SC and Pohl LR (1994) Metabolic
activation and immunochemical localization of liver protein adducts of the nonsteroidal
anti-inflammatory drug diclofenac. Chem Res Toxicol 7:575–582.
Hargus SJ, Martin BM, George JW and Pohl LR (1995) Covalent modification of rat liver
dipeptidyl peptidase IV (CD26) by the nonsteroidal anti-inflammatory drug diclofenac. Chem
Res Toxicol 8:993–996.
Helfgott SM, Sandberg-Cook J, Zakim D and Nestler J (1990) Diclofenac-associated hepato-
toxicity. J Am Med Assoc 264:2660 –2662.
Downloaded from dmd.aspetjournals.org at ASPET Journals on April 13, 2018
Kretz-Rommel A and Boelsterli UA (1993) Diclofenac covalent protein binding is dependent on
acyl glucuronide formation and is inversely related to P450-mediated acute cell injury in
cultured rat hepatocytes. Toxicol Appl Pharmacol 120:155–161.
Kretz-Rommel A and Boelsterli UA (1994) Mechanism of covalent adduct formation of diclofe-
nac to rat hepatic microsomal proteins: Retention of the glucuronic acid moiety in the adduct.
Drug Metab Dispos 22:956 –961.
Leemann T, Transon C and Dayer P (1993) Cytochrome P450TB (CYP2C): A major monoox-
ygenase catalyzing diclofenac 49-hydroxylation in human liver. Life Sci 52:29 –34.
Levine M, Law EYW, Bandiera SM, Chang TKH and Bellward GD (1998): In vivo cimetidine
inhibits hepatic CYP2C6 and CYP2C11 but not CYP 1A1 in adult male rats. J Pharmacol Exp
Ther 284:493– 499.
Li AP and Kedderis GL (1997) Primary hepatocyte cultures as an experimental model for the
evaluation of interactions between xenobiotics and drug-metabolizing enzymes. Chem-Biol
Interact 107:1–3.
Pang J-M, Zaleski J and Kauffman FC (1997) Toxicity of allyl alcohol in primary cultures of
freshly isolated and cryopreserved hepatocytes maintained on hydrated collagen gels. Toxicol
Appl Pharmacol 142:87–94.
Pumford NR, Myers TG, Davila JC, Highet RJ and Pohl LR (1993) Immunochemical detection
of liver protein adducts of the nonsteroidal antiinflammatory drug diclofenac. Chem Res
Toxicol 6:147–150.
Purcell P, Henry D and Melville G (1991) Diclofenac hepatitis. Gut 32:1381–1385.
Raucy JL and Lasker JM (1991) Isolation of P450 enzymes from human liver. Methods in
Enzymology 206:577–587.
Reed DJ (1990) Glutathione: Toxicological implications. Annu Rev Pharmacol Toxicol 30:603–
631.
Shen S, Hatgus SJ, Martin BM and Pohl LR (1997a) Cytochrome P4502C11 is a target of
diclofenac covalent binding in rats. Chem Res Toxicol 10:420 – 423.
Shen S, Marchick MR, Davis MR and Pohl LR (1997b) 5-Hydroxylation of diclofenac by human
cytochrome P4503A4: Possible role in idiosyncratic hepatotoxicity. Proceedings of Sixth
European ISSX Meeting, p 102, Gothenburg, Sweden.
Wong A and Bandiera SM (1998) Induction of hepatic cytochrome P450 2B and P450 3A
isozymes in rats by zolazepam, a constituent of Telazol. Biochem Pharmacol 55:201–207.