One-electron oxidation of diclofenac by human cytochrome P450s as a potential bioactivation mechanism for formation of 2′-(glutathion-S-yl)- deschloro-diclofenac

Article abstract of DOI:10.1016/j.cbi.2013.11.001

Reactive metabolites have been suggested to play a role in the idiosyncratic hepatotoxicity observed with diclofenac (DF). By structural identification of the GSH conjugates formed after P450-catalyzed bioactivation of DF, it was shown that three types of

Full text of DOI:10.1016/j.cbi.2013.11.001

Chemico-Biological Interactions 207 (2014) 32–40
Contents lists available at ScienceDirect
Chemico-Biological Interactions
journal homepage: www.elsevier.com/locate/chembioint
One-electron oxidation of diclofenac by human cytochrome P450s
as a potential bioactivation mechanism for formation
of 20-(glutathion-S-yl)-deschloro-diclofenac
⇑
Jan Simon Boerma, Nico P.E. Vermeulen, Jan N.M. Commandeur
Division of Molecular Toxicology, Amsterdam Institute for Molecules, Medicines and Systems (AIMMS), Faculty of Sciences, VU University Amsterdam, De Boelelaan 1083,
1081 HV Amsterdam, The Netherlands
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 June 2013
Received in revised form 9 October 2013
Accepted 5 November 2013
Available online 15 November 2013
Reactive metabolites have been suggested to play a role in the idiosyncratic hepatotoxicity observed with
diclofenac (DF). By structural identification of the GSH conjugates formed after P450-catalyzed bioactiva-
tion of DF, it was shown that three types of reactive intermediates were formed: p-benzoquinone imines,
o-imine methide and arene-oxide. Recently, detection of 20-(glutathion-S-yl)-deschloro-diclofenac (DDF-
SG), resulting from chlorine substitution, suggested the existence of a fourth type of P450-dependent
reactive intermediate whose inactivation by GSH is completely dependent on presence of glutathione
S-transferase. In this study, fourteen recombinant cytochrome P450s and three flavin-containing mono-
oxygenases were tested for their ability to produce oxidative DF metabolites and their corresponding GSH
conjugates. Concerning the hydroxymetabolites and their GSH conjugates, results were consistent with
previous studies. Unexpectedly, all tested recombinant P450s were able to form DDF-SG to almost similar
extent. DDF-SG formation was found to be partially independent of NADPH and even occurred by heat-
inactivated P450. However, product formation was fully dependent on both GSH and glutathione-S-trans-
ferase P1-1. DDF-SG formation was also observed in reactions with horseradish peroxidase in absence of
hydrogen peroxide. Because DDF-SG was not formed by free iron, it appears that DF can be bioactivated
by iron in hemeproteins. This was confirmed by DDF-SG formation by other hemeproteins such as hemo-
globin. As a mechanism, we propose that DF is subject to heme-dependent one-electron oxidation. The
resulting nitrogen radical cation, which might activate the chlorines of DF, then undergoes a GST-cata-
lyzed nucleophilic aromatic substitution reaction in which the chlorine atom of the DF moiety is replaced
by GSH.
Keywords:
Diclofenac
Cytochrome P450
Bioactivation
Reactive metabolite
Glutathione S-transferase
GSH conjugate
Ó 2013 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
DF metabolism in humans occurs predominantly by glucuroni-
dation and oxidative biotransformation by P450s, both of which
The nonsteroidal anti-inflammatory drug diclofenac (DF) is
commonly prescribed for the treatment of a number of inflamma-
tory disorders, including rheumatoid arthritis. Therapy with DF is
associated with severe hepatotoxicity in a very small percentage
of patients [1]. The exact mechanism of this idiosyncratic toxicity
is still poorly understood. Nevertheless, metabolic activation of
DF to reactive metabolites (RMs) is generally considered to play
an important role in the development of hepatotoxicity [2].
contribute to the formation of RMs [3]. Glucuronidation of DF leads
to reactive acyl-glucuronides, which appeared to be responsible for
the majority of the protein modification observed in vitro and in test
animals [4–6]. Metabolism of DF by cytochrome P450s (CYPs) re-
sults in several hydroxylated metabolites. The main hydroxylation
product is 40-hydroxydiclofenac (40-OH-DF), whereas 5-hydrox-
ydiclofenac (5-OH-DF) is a minor metabolite [3,7], Fig. 1. CYP2C9
was found to be mainly responsible for 40-OH-DF formation
[8–11]. In contrast, 5-OH-DF may be formed by several P450s,
including CYP2C8 and CYP3A4 [9–11]. Furthermore, 40-hydroxyl-
ation of DF-acylglucuronides by CYP2C8 is also considered to play
an important role [12]. Further enzymatic or non-enzymatic oxida-
tion of 40-OH-DF and 5-OH-DF results in their corresponding
p-benzoquinone imines, which have been detected as GSH conju-
gates M1 and M3 from diclofenac-2,5-quinone imine (5-OH-DF-
QI) and GSH-conjugate M2 from diclofenac-10,40-quinone imine
Abbreviations: 40-OH-DF, 40-hydroxydiclofenac; 5-OH-DF, 5-hydroxydiclofenac;
CYP, cytochrome P450; DDF-SG, 20-(glutathion-S-yl)-deschloro-diclofenac; DF,
diclofenac; FMO, flavin-containing monooxygenase; hGST P1-1, human glutathi-
one-S-transferase P1-1; HRP, horseradish peroxidase; RM, reactive metabolite.
⇑
Corresponding author. Tel.: +31 20 5987595; fax: +31 20 5987610.
E-mail address: j.n.m.commandeur@vu.nl (J.N.M. Commandeur).
0009-2797/$ - see front matter Ó 2013 Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.cbi.2013.11.001

J.S. Boerma et al. / Chemico-Biological Interactions 207 (2014) 32–40
33
Fig. 1. Overview of the metabolic pathways of diclofenac bioactivation by P450 enzymes. DF, diclofenac; QI, quinone-imine; DC-DF-o-IM, decarboxylated diclofenac o-imine
methide; DPAB-SG, 2-(2,6-dichlorophenylamino)-benzyl-S-thioether glutathione; DF-o-IM, diclofenac o-imine methide; DDF-SG, 20-(glutathion-S-yl)-deschloro-diclofenac.
(4’-OH-DF-QI) [13–15]. Another conjugate, assigned M5, was
shown to result from GSH substitution of a chlorine atom of DF-
10,40-QI followed by reduction to restore aromaticity [16]. Another
GSH-conjugate, previously assigned M4, has exactly the same mass
as M5 and was tentatively proposed to result from conjugation of
GSH to diclofenac-20,30-oxide [17]. Another bioactivation pathway
involves the oxidative decarboxylation of DF by CYP3A4 which re-
sults in a reactive o-imine methide that was identified after GSH
conjugation (DPAB-SG) [18,19].
By application of a mass spectrometry-based method for the
high-throughput screening of RMs, Wen and co-workers identified
a new GSH conjugate of DF (20-(glutathion-S-yl)-deschloro-diclofe-
nac; DDF-SG) in incubations with human liver microsomes [20]. The
mass of the novel conjugate (m/z 567.1) can be explained by substi-
tution of one of the chlorines on the dichlorophenyl ring of DF by
GSH. Formation of this conjugate cannot be explained by GSH con-
jugation of the known P450-dependent reactive intermediates of DF
(p-benzoquinone imine, arene-oxide and o-imine methide resulting
from decarboxylation). Nevertheless, the authors did not investi-
gate nor discuss the possibility that an as yet uncharacterized
intermediate of DF might be responsible for DDF-SG formation.
Recently, by using mass spectrometry to characterize the struc-
ture of DF-protein adducts, two peptide adducts with mass incre-
ment of 259 Da were found which result from adduction of DF
by chlorine substitution. The corresponding GSH conjugate,
DDF-SG, was only observed when DF incubations of CYP2C9 and
CYP3A4 with GSH were supplemented with human glutathione-
S-transferase P1-1 (hGST P1-1) [21]. More recently, we showed
that also hGSTM1-1 and hGSTA1-1 were able to catalyze formation
of DDF-SG, although at lower activity than hGSTP1-1 [22].
Three possible RMs of DF were previously proposed to explain
the chlorine substitution reaction mechanistically, Fig. 1. An
electron-withdrawing nitrogen atom resulting from N-oxidation,
o-imine methide formation or one-electron oxidation might allow
for GST-catalyzed GSH substitution of the activated chlorine atoms
[21].
The aim of the present paper is to characterize the bioactivation
pathway resulting in the chlorine-substituted DF conjugate. It was
investigated which of the commercially available human CYPs and
flavin-containing monooxygenases (FMOs) is able to form DDF-SG
and the other known metabolites of DF. The results from this study
provide evidence for non-enzymatic heme-dependent bioactiva-
tion of DF into a radical cation, which is subject to GST P1-1-depen-
dent chlorine substitution by GSH.
2. Materials and methods
2.1. Materials
DF sodium salt, bovine hemin, horseradish peroxidase (HRP)
(type I), human hemoglobin, equine heart myoglobin and human

34
J.S. Boerma et al. / Chemico-Biological Interactions 207 (2014) 32–40
serum albumin were obtained from Sigma (Steinheim, Germany).
40-OH-DF and 5-OH-DF were from Toronto Research Chemicals
(North York, Canada). Commercially available recombinant P450
enzymes and FMOs (Supersomes) were purchased from Gentest
Corporation (Woburn, USA). CYP2A6, 2B6, 2C8, 2C9*1, 2C19, 2E1,
2J2, 3A4 and 3A5 were co-expressed with human P450 oxidore-
ductase and cytochrome b5. CYP1A1, 1A2, 1B1, 2C18 and 2D6*1
were co-expressed with human P450 oxidoreductase. Pooled hu-
man liver microsomes (20 mg/mL) were from Xenotech (lot No.
0710619). All other materials were from standard suppliers and
of analytical grade.
2.4. Incubations of diclofenac with horseradish peroxidase, heme-
containing proteins and hemin in presence of human glutathione-
S-transferase P1-1
Heat-inactivated P450s unexpectedly showed significant DDF-
SG formation. To investigate whether DDF-SG formation might
be the result from heme-iron, we also studied whether this GSH
conjugate could also be formed in reactions with other hemepro-
teins, such as horseradish peroxidase (HRP), hemoglobin, myoglo-
bin and hemin. As negative control corresponding incubations
were carried out with human serum albumin. Incubations of DF
with HRP were performed in a total volume of 250
at pH 7.4 supplemented with 0.5 mM Detapac. Samples contained
100 M DF, 20 U/mL HRP, 10 mM GSH and 8 M GST P1-1. Reac-
tions were started by addition of 50 M H2O2 and allowed to pro-
ceed for 4 h at 25 °C. Incubations were terminated by addition of
250 L ice-cold methanol containing 5 mM ascorbic acid and cen-
lL 0.05 M KPi
2.2. Enzyme expression
l
l
For mechanistic studies, CYP2C9 and CYP3A4 were also co-ex-
pressed with human NADPH cytochrome P450 reductase in Esche-
l
richia coli DH5
a
. The expression and isolation of the P450s was
l
conducted as previously described [23].
trifuged for 20 min at 20,000g to remove precipitated protein. To
investigate whether observed metabolites were HRP dependent,
the reversible inhibitor potassium cyanide was used at a final con-
centration of 100 lM. Non-enzymatic product formation by HRP
was evaluated by use of heat-inactivated enzyme (5 min at
An expression system composed of E. coli XL-1 Blue cells carry-
ing a plasmid encoding GST P1-1 was a kind gift from Professor B.
Mannervik (Department of Biochemistry and Organic chemistry,
Uppsala University, Sweden). Expression and purification of GST
P1-1 by GSH-affinity chromatography was performed as described
previously [24].
95 °C). Reactions of DF with hemoglobin, myoglobin and hemin
were conducted in 250
0.5 mM Detapac. Incubations were performed with 500
5 mM GSH and 8 M GST P1-1. The reactions were initiated by
addition of a saturated solution of hemin (100 M) or by hemepro-
lL 0.05 M KPi at pH 7.4 supplemented with
l
M DF,
2.3. Incubations of diclofenac with human P450s and FMOs in presence
of human glutathione-S-transferase P1-1
l
l
tein (typically 0.2 mg/mL final concentration) and then incubated
for 1 h at 37 °C. Incubations were stopped by the addition of an
equal volume of ice-cold methanol containing 2% 250 mM ascorbic
acid. Samples were centrifuged (20 min, 20,000g) and supernatants
were stored at ꢀ20 °C until analysis by HPLC–UV or LC–MS/MS.
Incubations were conducted in a total volume of 125 lL 0.1 M
potassium phosphate (KPi) buffer at pH 7.4. DF was added as a
stock solution in DMSO (1% of incubation volume). A final concen-
tration of 100
CYPs and FMOs and for incubations with isoform-specific inhibi-
tors, whereas other incubations were conducted with 500 M DF.
All reactions were performed in presence of 100 M GSH and
M hGST P1-1. Commercial P450s and FMOs were added to a fi-
lM DF was used for incubations with commercial
l
2.5. Analytical methods
l
8
l
The metabolites of DF were separated by reversed phase chro-
nal concentration of 100 nM and 0.5 mg/mL, respectively. Incuba-
tions with in-house expressed P450s were conducted with
250 nM enzyme. Reactions were initiated by the addition of
NADPH regenerating system; final concentrations were 50 lM
NADPH, 2.5 mM glucose-6-phosphate and 0.5 U/mL glucose-6-
phosphate dehydrogenase. First, the time course of product forma-
tion was investigated by analyzing aliquots (250 lL) taken from
large-scale incubations (2.5 mL) at regular time points over a per-
iod of 90 min. Based on the results, subsequent reactions were per-
formed for 1 h at 37 °C. All incubations were quenched by the
addition of an equal volume of ice-cold methanol supplemented
with 2% (v/v) 250 mM ascorbic acid in water to prevent further
non-enzymatic oxidation after reactions were terminated. Protein
fractions were pelleted by centrifugation (20 min, 20,000g) and
supernatants were subsequently analyzed by HPLC–UV or LC–
MS/MS. Incubations in absence of one of the reaction components
were also conducted to investigate enzyme, GSH, and NADPH
dependence of the formed DF metabolites.
matography using a C18 column (Symmetry Shield C18, 3.5 lm,
4.6 ꢁ 100 mm i.d.; Waters) at a flow rate of 0.5 mL/min. The gradi-
ent was composed of solvent A (98.8% water/1% acetonitrile/0.2%
formic acid) and solvent B (98.8% acetonitrile/1% water/0.2% formic
acid). The first 5 min were isocratic at 0% B. The gradient was linear
from 0% to 100% between 5 and 30 min, and the column was al-
lowed to re-equilibrate from 30 to 40 min at 0% B. Samples were
analyzed by HPLC–UV or by LC–MS/MS.
HPLC analysis was performed on a Shimadzu HPLC equipped
with two LC-20AD pumps, a SIL20AC autosampler and SPD20A
UV detector. Following injection of 50
graphed and detected by UV/Vis at 254 nm. A standard curve of DF
(0.05–20 M) was used to determine the concentrations of the
lL, samples were chromato-
l
formed GSH-conjugates of DF, assuming the extinction coefficients
of the conjugates are identical to that of DF. Peak areas of DF
metabolites were analyzed using the Shimadzu LC solution soft-
ware package (version 1.25).
For MS analysis, an injection volume of 25 lL was used. Sam-
To evaluate whether formation of DF metabolites was P450
dependent, incubations of CYP3A4 were performed in presence of
ples were analyzed on an Agilent 1200 Series Rapid resolution LC
equipped with a hybrid Quadrupole-Time-Of-Flight (Q-TOF) Agi-
lent 6520 mass spectrometer (Agilent technologies, Waldbronn,
Germany). Analytes were first detected by UV at 254 nm and sub-
sequently ionized by electrospray ionization. The mass spectrome-
ter was operated at a capillary voltage of 3500 V with nitrogen as
drying gas (12 L/min) and nebulizer gas (pressure 60 psig). The
gas temperature was 350 °C during operation. The Q-TOF was used
in the positive mode and data was acquired using the Mass Hunter
workstation software (version B.02.00).
2
lM ketoconazole, whereas CYP2C9 reactions were conducted
with 20 M sulfaphenazole. To study non-enzymatic formation of
l
DF metabolites by CYPs, also heat-inactivated (5 min, 95 °C) en-
zymes were used.
Previously, formation of an o-quinone methide was proposed as
one of the alternative bioactivation mechanisms for DDF-SG. To
investigate whether this mechanism is involved, incubations were
also performed using buffers and cofactors prepared in deuterium
oxide. Incorporation of a deuterium ion at the benzylic position
would be expected if an o-quinone methide is involved.
For identification of metabolites, samples were analyzed by
automated MS/MS analysis. In each round, one MS1 spectra (m/z

J.S. Boerma et al. / Chemico-Biological Interactions 207 (2014) 32–40
35
200–1000) was acquired, followed by fragmentation of the three
most abundant ions. MS/MS analysis was conducted at a collision
energy voltage of 25 V using nitrogen as the collision gas. For quan-
titative analyses, only MS1 spectra (m/z 300–700) were acquired in
Table A.1 of Appendix A. Consistent with previous data [8–10],
CYP2C9 appeared the major isoform involved in 40-OH-DF forma-
tion, whereas CYP2D6 was found to have 2.7-fold lower activity
(Fig. 3A). Compared to these enzymes, the contribution of the other
CYPs to 40-OH-DF formation was negligible. Conjugate M2 and M5
result from GSH conjugation to the p-benzoquinone imine of 40-
OH-DF. Similar to DF 40-hydroxylation, M2 formation was mainly
mediated by CYP2C9 (Fig. 3B).
order to obtain sufficient data points for each ion. Clozapine (4 lM)
was added to each of the samples as internal standard. All LC–MS/
MS data was analyzed using the Mass Hunter software package
(version B.01.03).
As shown in Fig. 4A, the minor hydroxylation product 5-OH-DF
was mainly catalyzed by CYP3A4 with 5- to 10-fold lower activity
of CYP2B6, CYP2C18, CYP2C19, CYP2D6 and CYP3A5. The GSH con-
jugation products of the quinone imine of 5-OH-DF (M1 and M3)
were also predominantly formed by CYP3A4, consistent with re-
sults of Tang and coworkers [14]. Evaluation of the CYPs which
have not been tested previously (1A1, 1B1, 2B6, 2C18, 2J2 and
3A5) showed minor formation of M1 and M3 in incubations with
CYP2B6 and CYP2C18, although activities were less than 15% of
CYP3A4 (Fig. 4B and C).
DPAB-SG, which results from DF decarboxylation, was only
formed in very low levels by recombinant CYP3A4 (data not
shown), which is in line with previous observations [19]. As ex-
pected, human FMOs were unable to form monohydroxylated DF
metabolites or GSH conjugates thereof.
3. Results
3.1. Time-dependent formation of GSH conjugates of diclofenac by
cytochrome P450s
The time course of product formation was evaluated with
CYP2C9 and CYP3A4, which were previously found to be responsi-
ble for formation of GSH conjugates resulting from 40-OH-DF and
5-OH-DF, respectively [14]. In addition, these P450s formed the
GSH conjugate resulting from chlorine substitution (DDF-SG) in
reactions performed with 100 lM GSH and 8 lM GST P1-1 [21].
When time-dependent product formation was investigated under
these conditions, DDF-SG formation increased linearly for approx-
imately 20 min in DF reactions of E. coli expressed CYP2C9 and
CYP3A4, respectively, Fig. 2A and B. DDF-SG formation reached a
plateau after 60 min of incubation, which might be explained by
product-inhibition of GST P1-1. Sigmoidal curves were obtained
when the time-course of formation of GSH conjugates M2 and
M3 was plotted, Fig. 2C and D. These sigmoidal curves can be ex-
plained by the fact that the primary metabolites 40-OH-DF and 5-
OH-DF must undergo a second oxidation to produce the GSH-reac-
tive quinone imines. Because at 60 min all GSH-conjugates of DF
were formed at significant extent, all further incubations were also
allowed to proceed for 60 min.
3.3. Formation of DDF-SG by human liver microsomes, recombinant
human P450s and FMOs
Formation of DDF-SG was only observed when DF incubations
with human liver microsomes were supplemented with GST P1-1
(Fig. A.1A in Appendix A). The mass spectrum of DDF-SG (m/z
567.1) exhibited a mono-chlorine isotope cluster (Fig. A.1B). Pres-
ence of a GSH-moiety in DDF-SG was confirmed by MS/MS frag-
ments of m/z 492 and 438, which correspond to loss to loss of
glycine and pyroglutamate, respectively [20,21].
3.2. Bioactivation of DF by recombinant human P450s and FMOs
Fig. 5 shows the formation of DDF-SG in the incubations with
cDNA expressed P450s and FMOs. Unexpectedly, all tested P450s
appeared to form DDF-SG in significant amounts, with CYP1A1 as
most active, and CYP2A6 as least active isoform. In line with our
previous findings [21], low levels of DDF-SG were also found in
FMO incubations. However, the abundance of DDF-SG in reactions
with FMOs appeared not to be different from the control micro-
To evaluate which P450s contribute to the formation of the dif-
ferent stable and reactive metabolites of DF, a panel of 14 commer-
cially available cDNA expressed P450s was incubated with 100
lM
DF in presence of 100 M GSH and 8 M GST P1-1, Figs. 3 and 4.
l
l
The specific activities of the product formation can be found in
B
A
CYP 3A4
3
CYP 2C9
1.0
0.8
0.6
0.4
0.2
0.0
2
1
0
0
20 40 60 80 100
Time (min)
0
20 40 60 80 100
Time (min)
C40
D
CYP 2C9
CYP 3A4
5
4
3
2
1
0
30
20
10
0
0
20 40 60 80 100
Time (min)
0
20 40 60 80 100
Time (min)
Fig. 2. Time-dependent formation of GSH conjugates of diclofenac by CYP2C9 and CYP3A4. (A) and (B): Formation of DDF-SG by CYP2C9 (A) and CYP3A4 (B). (C) and (D):
Formation of GSH conjugate M2 by CYP2C9 (C) and conjugate M3 by CYP3A4 (D).

36
J.S. Boerma et al. / Chemico-Biological Interactions 207 (2014) 32–40
Relative formation of 4'-OH-DF
A
A
Relative formation of 5-OH-DF
120
100
80
60
40
20
0
120
100
80
60
40
20
0
Relative formation of M2
B
B
Relative formation of M1
120
100
80
60
40
20
0
120
100
80
60
40
20
0
Fig. 3. Evaluation of recombinant human P450s and FMOs involved in the
formation of 40-OH-DF (A) and GSH conjugate M2 (B). Control microsomes (insect)
and microsomes expressing reductase and cytochrome b5 (reductase) are indicated.
C
Relative formation of M3
120
100
80
60
40
20
0
somes without expressed P450 or FMO. Not any DDF-SG was de-
tected when microsomes were omitted from incubations.
Since it appeared highly unlikely that all P450 isoforms contrib-
ute to DDF-SG formation to significant extent, additional mecha-
nistic studies were conducted with CYP2C9 and CYP3A4 to
further investigate how different components of the reaction af-
fected DDF-SG formation.
First, it was investigated whether DDF-SG and other DF metab-
olites were formed under conditions in which one of the incuba-
tion constituents was omitted from the reaction (Table 1).
CYP2C9 and CYP3A4 formed DDF-SG in DF reactions containing
P450, GSH, GST P1-1 and NADPH. DDF-SG and other DF metabo-
lites were not detected in absence of P450, whereas 40-OH-DF
and 5-OH-DF were the only products formed in absence of GSH.
The GST-dependent GSH conjugate DDF-SG was not found in
P450 incubations without GST P1-1. Other GSH conjugates could
be detected under these conditions but conjugate formation was
at least 4-fold lower in absence of GST P1-1. In case of CYP2C9,
addition of NADPH had no effect on DDF-SG formation. In contrast,
DDF-SG was about 1.5-fold less abundant when NADPH was omit-
ted in reactions of CYP3A4. In absence of NADPH, 40-OH-DF, 5-OH-
DF and their corresponding GSH conjugates were not observed.
To evaluate whether DDF-SG formation was dependent on
binding to the P450-active site, reactions were also conducted in
presence of isoform-specific inhibitors (Table 1). When CYP3A4
Fig. 4. Formation of 5-OH-DF (A) and its GSH conjugates M1 (B) and M3 (C) by
cDNA expressed P450s and FMOs. Controls consisting of insect cell microsomes
(insect) and microsomes expressing P450 reductase and cytochrome b5 (reductase)
are shown.
Relative formation of DDF-SG
120
100
80
60
40
20
0
Fig. 5. Formation of DDF-SG in incubations of recombinant P450s and FMOs.
Control microsomes without reductase and CYPs (insect) or expressing reductase
and cytochrome b5 (reductase) were also evaluated.
was incubated with ketoconazole (2 lM), DDF-SG formation was
inhibited by 54% when compared to the non-inhibited incubations.
Other DF metabolites of CYP3A4 were fully inhibited by ketocona-
zole. In contrast, DDF-SG formation in CYP2C9 incubations was not
inhibited by sulfaphenazole (20
l
M), whereas the formation of 40-
controls. Heat denaturation of CYP2C9 even increased DDF-SG
formation by more than 2-fold, clearly indicating that regular oxi-
dative biotransformation is not applicable in formation of this GSH
conjugate. No DDF-SG was detected when P450 reactions were
performed with heat-denatured GST P1-1, suggesting that the
native conformation of this enzyme is required for DDF-SG forma-
tion. Similar to incubations without GST P1-1, heat treatment of
GST P1-1 resulted in low levels of GSH conjugates derived from
the monohydroxylated metabolites of DF.
OH-DF and its GSH-conjugates was decreased by at least 50%.
To investigate the possibility of non-enzymatic DDF-SG forma-
tion by P450s and GST P1-1, incubations of DF were also performed
with heat-inactivated enzymes (Table 1). The formation of mono-
hydroxylated DF metabolites and their corresponding GSH conju-
gates was completely abolished upon heat treatment of the
P450s at 95 °C. Unexpectedly, in incubations with heat-inactivated
CYP3A4, DDF-SG formation was still 54% of the non-heated

J.S. Boerma et al. / Chemico-Biological Interactions 207 (2014) 32–40
37
Table 1
Relative amounts of formation of diclofenac metabolites under different experimental conditions.
Controls
CYP2C9
CYP3A4
40-OH-DF
M2 + M5
DDF-SG
5-OH-DF
M1 + M3
DDF-SG
Complete
ꢀP450
100( 4)a
100( 10)
ND
ND
25( 2)
ND
100( 3)
ND
ND
ND
108( 12)
100( 3)
ND
376( 109)
434( 127)
ND
100( 2)
ND
ND
6( 1)
ND
100( 3)
ND
ND
ND
63( 15)
NDb
ꢀGSH
94( 9)
92( 5)
ND
ꢀGST P1-1
ꢀNADPH
P450 inhibition
ꢀInhibitor
100( 3)
51( 13)
100( 7)
30( 5)
100( 4)
105( 2)
ND
ND
100( 3)
1( 2)
100( 1)
46( 1)
+Inhibitorc
Heat treatment
ꢀBoiling
100( 4)
ND
99( 6)
100( 2)
ND
23( 2)
100( 3)
242( 35)
ND
100( 21)
ND
814( 146)
100( 1)
6( 1)
7( 1)
100( 2)
54( 6)
ND
Boiled P450
Boiled GST
a
Formation of DF metabolites as percentage of the amount formed in control incubations (AUCmetabolite/AUCmetabolite, control*100%). Control reactions consisted of all
incubation components (complete), were conducted without inhibitor (ꢀinhibitor) or were not subjected to heat treatment (ꢀboiling).
b
ND, not detected.
c
Ketoconazole (2 lM) was used as inhibitor for CYP3A4, whereas sulfaphenazole (20 lM) was used to inhibit CYP2C9.
3.4. Formation of DDF-SG in incubations of horseradish peroxidase
3.5. Effect of iron and hemeproteins on the formation of DDF-SG
Previously, hemeproteins have been found to perform oxidative
reactions by mechanisms of peroxidases [25,26]. Therefore, incuba-
tions of DF were also conducted with HRP, which is known to cata-
lyze one-electron oxidations [27]. HRP did not form 40-OH-DF and
5-OH-DF or GSH conjugates resulting from these metabolites. Inter-
estingly, HRP was able to produce DDF-SG in DF reactions performed
in presence of GSH, GST P1-1 and H2O2 (Fig. 6). Although incubations
were performed with 10 mM GSH, no DDF-SG was detected in ab-
sence of HRP or GST P1-1, ruling out direct substitution of one of
the chlorines of DF by GSH. However, product formation in reactions
without H2O2 was still 77% of complete incubations, suggesting
DDF-SG formation by HRP is non-enzymatic. In line with this, nei-
ther boiling of HRP prior to the reaction nor addition of the reversible
HRP inhibitor potassium cyanide completely abolished product
formation. Under these conditions, DDF-SG formation was still
41% and 44% of controls, respectively. As observed in the P450 reac-
tions, no DDF-SG was detected upon heat inactivation of GST P1-1.
The observation that both P450s and HRP can form DDF-SG in
absence of cofactor raises the possibility that this conjugate results
from iron-mediated activation of DF. DDF-SG was not observed
when DF was incubated with 100 l
M Fe3+. Alternatively, coordi-
nated iron from the prosthetic heme groups in P450 and HRP
may be required for DDF-SG formation. To investigate this, several
hemeproteins were incubated with DF in presence of GSH and GST
P1-1 (Fig. 7). LC–MS analysis showed formation of DDF-SG when
reactions were conducted with hemoglobin, myoglobin or HRP.
No conjugate was found with human serum albumin or when
hemeproteins were incubated in absence of GST P1-1. To further
corroborate these findings, reactions were also performed with
iron-containing protoporphyrin IX (hemin), which is highly similar
to the heme-moiety of the tested hemeproteins. DDF-SG was de-
tected in DF reactions of hemin containing GSH and GST P1-1,
but not in similar reactions in which hemin or GST P1-1 was omit-
ted (Fig. 7).
Fig. 6. HPLC–UV analysis of diclofenac conjugate DDF-SG in incubations of HRP. Shown UV traces (254 nm) are from incubations in which HRP, GST P1-1 or peroxide was
omitted. A reaction containing all components required for catalytically active HRP (complete) is also shown.

38
J.S. Boerma et al. / Chemico-Biological Interactions 207 (2014) 32–40
present paper was to elucidate the mechanism by which DF bioac-
tivation might result in DDF-SG. Preliminary findings indicated
that recombinant human FMOs formed low amounts of DDF-SG,
which was interpreted as support for N-oxygenation as a possible
bioactivation pathway [21]. However, in the present study, no dif-
ference in product formation was observed between recombinant
FMOs and their corresponding control microsomes (Fig. 5). P450-
expressing microsomes showed significantly higher levels of
DDF-SG than FMOs.
To evaluate whether an imine-methide intermediate of DF is
responsible for DDF-SG formation (Fig. 1), P450 incubations were
also performed in phosphate buffer prepared with D2O. However,
no deuterium incorporation was observed at the benzylic position
in DDF-SG (data not shown), suggesting the postulated DF imine-
methide is not formed. Remarkably, formation of DDF-SG appeared
non-enzymatic in P450 incubations, because the product was ob-
served in absence of NADPH and in reactions of heat-inactivated
P450s (Table 1). The finding that other hemeproteins and protopor-
phyrin IX could also catalyze DDF-SG formation (Fig. 7) strongly
suggests that this conjugate results from a DF radical cation formed
by one-electron oxidation followed by GST P1-1 catalyzed GSH
conjugation. Unambiguous evidence for the involvement of a rad-
ical cation of DF may be obtained from electron spin resonance
studies.
Fig. 7. Ion traces of DDF-SG (m/z 567.1) from diclofenac incubations performed
with different (heme)proteins and hemin. Shown traces are from horseradish
peroxidase (A), hemoglobin (B), myoglobin (C), human serum albumin (D) and
hemin (E) incubated in presence of GSH and GST P1-1. (F) DF incubations of hemin
in which GST P1-1 was omitted.
Peroxidase-like activity of hemeproteins may be responsible for
one-electron oxidation of substrates. The higher oxidation states
(compound I and II) of HRP were found to accept electrons from
nitrogen compounds, resulting in formation of nitrogen-centered
radical cations [28,29]. Peroxidase-type mechanisms have also
been observed in reactions of the antipsychotic chlorpromazine
with methemoglobin, and were implicated in the bioactivation of
the antidepressant nomifensine by hemoglobin and myeloperoxi-
dase [26,30]. In the aforementioned studies, peroxide was used
to form higher oxidation states of the hemeproteins, which is nec-
essary for one-electron oxidation to occur [25,29]. In the present
study, however, DDF-SG was formed when reactions of HRP and
other hemeproteins were incubated in absence of hydrogen perox-
ide (Figs. 6 and 7). To evaluate a potential role for H2O2 in DDF-SG
formation, CYP2C9 incubations were also supplemented with cat-
alase (200 U/mL) or with 500 lM H2O2. Nevertheless, these condi-
Fig. 8. Proposed mechanism for the bioactivation route of diclofenac that results in
tions did not affect DDF-SG formation by CYP2C9 (data not shown).
A possible mechanism to explain DDF-SG formation is coupled
one-electron oxidation of DF and hemeprotein, resulting in elec-
tron transfer to molecular oxygen and formation of H2O2 (Fig. 8).
The DF radical cations formed may either be reduced back to DF
by GSH or may undergo GST-catalyzed chlorine substitution. Previ-
ously, similar mechanisms were proposed to rationalize one-elec-
tron oxidation of various substrates in incubations with
hemoglobin [31–34]. Alternatively, the radical intermediate of DF
might result from autoxidation. However, DDF-SG could not be de-
tected in absence of hemeproteins (Figs. 6 and 7), even after incu-
bation at room temperature for 4 h under air.
The formation of a nitrogen centered radical cation has been
proposed as the initial step in the electrochemical oxidation of
DF [35]. It was previously suggested that this radical might con-
tribute to DF toxicity [2]. Galati and co-workers showed that incu-
bation of DF and other nonsteroidal anti-inflammatory drugs with
HRP and H2O2 resulted in co-oxidation of NADH, which they ex-
plained by formation of prooxidant radicals from these drugs
[36]. The activation of molecular oxygen by NAD radicals might re-
sult in formation of reactive oxygen species which can induce
mitochondrial injury. In support of a role for prooxidant DF radicals
in cell toxicity is the finding that addition of noncytotoxic concen-
trations of HRP and H2O2 to DF incubations with rat hepatocytes
significantly increased lipid peroxidation and GSSG formation
[37]. Reduction of DF radical cations by concomitant oxidation of
formation of DDF-SG.
3.6. Effect of ascorbic acid on the formation of DDF-SG by P450 and
HRP
Because DF bioactivation by hemeproteins presumably involves
non-enzymatic one-electron oxidation resulting in formation of a
DF radical cation, it was investigated whether ascorbic acid could
inhibit product formation. DDF-SG was not observed when DF
incubations of P450 or HRP were supplemented with ascorbic acid.
GSH conjugates M1 and M3 were also strongly decreased when
ascorbic acid was added to P450 incubations (data not shown).
This is in line with previous observations that ascorbic acid reduces
p-benzoquinone imine back to 5-OH-DF [15].
4. Discussion
Recently, protein adducts of DF were identified in which the
chlorine atom of DF was substituted by protein thiol. The corre-
sponding GSH conjugate (DDF-SG) was subsequently only identi-
fied in incubations in presence of GST P1-1, suggesting that the
reactive entity involved is only reactive to activated thiol groups
[21]. As possible explanation for DDF-SG formation, three reactive
intermediates have been proposed in which an electron withdraw-
ing nitrogen atom activates the chlorines (Fig. 1). The aim of the

J.S. Boerma et al. / Chemico-Biological Interactions 207 (2014) 32–40
39
GSH regenerates DF, Fig. 8. The resulting redox cycling may deplete
References
cellular GSH pools, impairing cellular defenses and disrupting the
normal redox balance [38]. Besides reactions involving GSH, we
have previously shown that the putative radical cation of DF also
adducts to cysteine residues of a model protein [21]. Because of
its apparent preference for cysteine thiolates, the radical cation
of DF may react with highly nucleophilic cysteines in catalytic tri-
ads of proteins such as the molecular chaperone protein disulfide
isomerase. The essential dithiol pairs in this protein each contain
a reactive cysteine, modification of which was shown to result in
complete enzyme inactivation [39]. In the present study and in
our previous work, we have used an in vitro experimental approach
to investigate GSH conjugation and protein modification of the rad-
ical cation of DF [21]. Therefore, it remains to be established
whether thiol modification by the radical cation of DF is relevant
in vivo and whether this mechanism contributes to DF toxicity.
Contrary to other GSH conjugates of DF, DDF-SG was only de-
tected when DF incubations of recombinant P450s or human liver
microsomes were supplemented with GST P1-1 (Table 1, Fig. A1).
This may provide a likely explanation for the fact that DDF-SG
was previously not observed in previous DF studies using recombi-
nant P450s or human liver microsomes [14,16,17,19,35]. Neverthe-
less, by analysis of concentrated DF incubations using a sensitive
LC–MS method, DDF-SG was previously also detected in human li-
ver microsomes [20]. This finding suggests that microsomal GSTs
might also catalyze formation of this conjugate.
In conclusion, as the mechanism of DDF-SG formation we pro-
pose a nucleophilic aromatic substitution reaction of the chlorine
atom of the DF radical cation formed by heme-iron containing
proteins. GSTs can catalyze the substitution of chlorine atoms on
electrophilic compounds as is most well-known by the example
of 1-chloro-2,4-dinitrobenzene [40]. However, unlike GSH conju-
gation to 1-chloro-2,4-dinitrobenzene, which also occurs non-
enzymatically to significant extent at pH 7.4, the chlorine substitu-
tion of the proposed DF radical cation requires activation of GSH
(to GSꢀ) by GSTs. Interestingly, a fully GST P1-1-dependent chlo-
rine substitution reaction was previously observed for the cloza-
pine nitrenium ion [24]. Hence, this type of GST-catalyzed
substitution reactions could also apply to other chlorinated reac-
tive drug metabolites. The high selectivity of the radical cation of
DF for activated cysteine residues in proteins might be of toxico-
logical relevance, since activated cysteines generally have high
functional importance [41,42]. The cytotoxicity of DF might be
the result of the combination of the different P450-dependent
reactive intermediates. However, the occurrence of the novel
bioactivation mechanism in vivo remains to be established by ana-
lyzing DDF-SG and/or corresponding thioethers in bile or urine of
DF-exposed individuals, or by analyzing corresponding protein
adducts.
[1] S.M. Helfgott, J. Sandberg-Cook, D. Zakim, J. Nestler, Diclofenac-associated
hepatotoxicity, JAMA 264 (1990) 2660–2662.
[2] U.A. Boelsterli, Diclofenac-induced liver injury: a paradigm of idiosyncratic
drug toxicity, Toxicol. Appl. Pharmacol. 192 (2003) 307–322.
[3] H. Stierlin, J.W. Faigle, Biotransformation of diclofenac sodium (Voltaren) in
animals and in man. II. Quantitative determination of the unchanged drug and
principal phenolic metabolites, in urine and bile, Xenobiotica 9 (1979) 611–
621.
[4] A. Kretz-Rommel, U.A. Boelsterli, 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 (1993) 155–161.
[5] S.J. Hargus, H.R. Amouzedeh, N.R. Pumford, T.G. Myers, S.C. McCoy, L.R. Pohl,
Metabolic activation and immunochemical localization of liver protein adducts
of the nonsteroidal anti-inflammatory drug diclofenac, Chem. Res. Toxicol. 7
(1994) 575–582.
[6] L.T. Wade, J.G. Kenna, J. Caldwell, Immunochemical identification of mouse
hepatic protein adducts derived from the nonsteroidal anti-inflammatory
drugs diclofenac, sulindac, and ibuprofen, Chem. Res. Toxicol. 10 (1997) 546–
555.
[7] P.H. Degen, W. Dieterle, W. Schneider, W. Theobald, U. Sinterhauf,
Pharmacokinetics of diclofenac and five metabolites after single doses in
healthy volunteers and after repeated doses in patients, Xenobiotica 18 (1988)
1449–1455.
[8] T. Leemann, C. Transon, P. Dayer, Cytochrome P450TB (CYP2C):
a major
monooxygenase catalyzing diclofenac 40-hydroxylation in human liver, Life
Sci. 52 (1993) 29–34.
[9] R. Bort, K. Mace, A. Boobis, M.J. Gomez-Lechon, A. Pfeifer, J. Castell, Hepatic
metabolism of diclofenac: role of human CYP in the minor oxidative pathways,
Biochem. Pharmacol. 58 (1999) 787–796.
[10] A. Mancy, M. Antignac, C. Minoletti, S. Dijols, V. Mouries, N.T. Duong, P.
Battioni, P.M. Dansette, D. Mansuy, Diclofenac and its derivatives as tools for
studying human cytochromes P450 active sites: particular efficiency and
regioselectivity of P450 2Cs, Biochemistry 38 (1999) 14264–14270.
[11] S. Shen, M.R. Marchick, M.R. Davis, G.A. Doss, L.R. Pohl, Metabolic activation of
diclofenac by human cytochrome P450 3A4: role of 5-hydroxydiclofenac,
Chem. Res. Toxicol. 12 (1999) 214–222.
[12] S. Kumar, K. Samuel, R. Subramanian, M.P. Braun, R.A. Stearns, S.H. Chiu, D.C.
Evans, T.A. Baillie, Extrapolation of diclofenac clearance from in vitro
microsomal metabolism data: role of acyl glucuronidation and sequential
oxidative metabolism of the acyl glucuronide, J. Pharmacol. Exp. Ther. 303
(2002) 969–978.
[13] W. Tang, R.A. Stearns, S.M. Bandiera, Y. Zhang, C. Raab, M.P. Braun, D.C. Dean, J.
Pang, K.H. Leung, G.A. Doss, J.R. Strauss, G.Y. Kwei, T.H. Rushmore, S.H. Chiu,
T.A. Baillie, Studies on cytochrome P-450-mediated bioactivation of diclofenac
in rats and in human hepatocytes: identification of glutathione conjugated
metabolites, Drug Metab. Dispos. 27 (1999) 365–372.
[14] W. Tang, R.A. Stearns, R.W. Wang, S.H. Chiu, T.A. Baillie, Roles of human
hepatic cytochrome P450s 2C9 and 3A4 in the metabolic activation of
diclofenac, Chem. Res. Toxicol. 12 (1999) 192–199.
[15] G. Miyamoto, N. Zahid, J.P. Uetrecht, Oxidation of diclofenac to reactive
intermediates by neutrophils, myeloperoxidase, and hypochlorous acid, Chem.
Res. Toxicol. 10 (1997) 414–419.
[16] L.J. Yu, Y. Chen, M.P. Deninno, T.N. O’Connell, C.E. Hop, Identification of a novel
glutathione adduct of diclofenac, 40-hydroxy-20-glutathion-deschloro-
diclofenac, upon incubation with human liver microsomes, Drug Metab.
Dispos. 33 (2005) 484–488.
[17] Z. Yan, J. Li, N. Huebert, G.W. Caldwell, Y. Du, H. Zhong, Detection of a novel
reactive metabolite of diclofenac: evidence for CYP2C9-mediated bioactivation
via arene oxides, Drug Metab. Dispos. 33 (2005) 706–713.
[18] Y. Teffera, D.J. Waldon, A.E. Colletti, B.K. Albrecht, Z. Zhao, Identification of a
novel glutathione conjugate of diclofenac by LTQ-Orbitrap, Drug Metab. Lett. 2
(2008) 35–40.
[19] M.P. Grillo, J. Ma, Y. Teffera, D.J. Waldon, A novel bioactivation pathway for 2-
[2-(2,6-dichlorophenyl)aminophenyl]ethanoic acid (diclofenac) initiated by
cytochrome P450-mediated oxidative decarboxylation, Drug Metab. Dispos.
36 (2008) 1740–1744.
[20] B. Wen, L. Ma, S.D. Nelson, M. Zhu, High-throughput screening and
characterization of reactive metabolites using polarity switching of hybrid
triple quadrupole linear ion trap mass spectrometry, Anal. Chem. 80 (2008)
1788–1799.
Conflict of interest
The authors declare that there are no conflicts of interest.
Acknowledgements
[21] J.S. Boerma, S. Dragovic, N.P.E. Vermeulen, J.N.M. Commandeur, Mass
spectrometric characterization of protein adducts of multiple P450-
dependent reactive intermediates of diclofenac to human glutathione-S-
transferase P1-1, Chem. Res. Toxicol. 25 (2012) 2532–2541.
This research was performed within the framework of project
D3-201 of the Dutch Top Institute Pharma.
[22] S. Dragovic, J.S. Boerma, N.P.E. Vermeulen, J.N.M. Commandeur, Effect of
human glutathione S-transferases on glutathione-dependent inactivation of
cytochrome P450-dependent reactive intermediates of diclofenac, Chem. Res.
Toxicol. 26 (2013) 1632–1641.
Appendix A. Supplementary data
[23] R. Appiah-Opong, J.N. Commandeur, B. van Vugt-Lussenburg, N.P. Vermeulen,
Inhibition of human recombinant cytochrome P450s by curcumin and
curcumin decomposition products, Toxicology 235 (2007) 83–91.
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.cbi.2013.11.001.

40
J.S. Boerma et al. / Chemico-Biological Interactions 207 (2014) 32–40
[24] S. Dragovic, J.S. Boerma, L. van Bergen, N.P.E. Vermeulen, J.N.M.
Commandeur, Role of human glutathione S-transferases in the
inactivation of reactive metabolites of clozapine, Chem. Res. Toxicol. 23
(2010) 1467–1476.
[25] B.W. Griffin, P.L. Ting, Mechanism of N-demethylation of aminopyrine by
hydrogen peroxide catalyzed by horseradish peroxidase, metmyoglobin, and
protohemin, Biochemistry 17 (1978) 2206–2211.
[33] P. Eyer, M. Kiese, G. Lipowsky, N. Weger, Reactions of 4-dimethylaminophenol
with hemoglobin, and autoxidation of 4-dimethylaminophenol, Chem. Biol.
Interact. 8 (1974) 41–59.
[34] W.J. Wallace, W.S. Caughey, Mechanism for the autoxidation of hemoglobin by
phenols, nitrite and ‘‘oxidant’’ drugs. Peroxide formation by one electron
donation to bound dioxygen, Biochem. Biophys. Res. Commun. 62 (1975) 561–
567.
[26] P.P. Kelder, M.J. Fischer, N.J. de Mol, L.H. Janssen, Oxidation of chlorpromazine
by methemoglobin in the presence of hydrogen peroxide. Formation of
chlorpromazine radical cation and its covalent binding to methemoglobin,
Arch. Biochem. Biophys. 284 (1991) 313–319.
[27] P.R. Ortiz de Montellano, Catalytic sites of hemoprotein peroxidases, Annu.
Rev. Pharmacol. Toxicol. 32 (1992) 89–107.
[28] F.S. Sariaslani, M.W. Duffel, J.P. Rosazza, One-electron oxidation of vindoline
and 16-O-acetylvindoline catalyzed by peroxidase, J. Med. Chem. 28 (1985)
629–633.
[29] J. Van der Zee, D.R. Duling, R.P. Mason, T.E. Eling, The oxidation of N-
substituted aromatic amines by horseradish peroxidase, J. Biol. Chem. 264
(1989) 19828–19836.
[35] K.G. Madsen, C. Skonberg, U. Jurva, C. Cornett, S.H. Hansen, T.N. Johansen, J.
Olsen, Bioactivation of diclofenac in vitro and in vivo: correlation to
electrochemical studies, Chem. Res. Toxicol. 21 (2008) 1107–1119.
[36] G. Galati, S. Tafazoli, O. Sabzevari, T.S. Chan, P.J. O’Brien, Idiosyncratic NSAID
drug induced oxidative stress, Chem. Biol. Interact. 142 (2002) 25–41.
[37] S. Tafazoli, D.D. Spehar, P.J. O’Brien, Oxidative stress mediated idiosyncratic
drug toxicity, Drug Metab. Rev. 37 (2005) 311–325.
[38] D. Han, N. Hanawa, B. Saberi, N. Kaplowitz, Mechanisms of liver injury. III. Role
of glutathione redox status in liver injury, Am. J. Physiol. Gastrointest. Liver
Physiol. 291 (2006) G1–G7.
[39] H.C. Hawkins, R.B. Freedman, The reactivities and ionization properties of the
active-site dithiol groups of mammalian protein disulphide-isomerase,
Biochem. J. 275 (1991) 335–339.
[40] W.H. Habig, M.J. Pabst, W.B. Jakoby, Glutathione S-transferases. The first
enzymatic step in mercapturic acid formation, J. Biol. Chem. 249 (1974) 7130–
7139.
[30] R.S. Obach, D.K. Dalvie, Metabolism of nomifensine to a dihydroisoquinolinium
ion metabolite by human myeloperoxidase, hemoglobin, monoamine
oxidase A, and cytochrome P450 enzymes, Drug Metab. Dispos. 34 (2006)
1310–1316.
[31] K.A. Fletcher, P.F. Barton, J.A. Kelly, Studies on the mechanisms of oxidation in
the erythrocyte by metabolites of primaquine, Biochem. Pharmacol. 37 (1988)
2683–2690.
[32] C. Storle, P. Eyer, Formation and reactions of the Wurster’s blue radical cation
during the reaction of N,N,N0,N0-tetramethyl-p-phenylenediamine with
oxyhemoglobin, Chem. Biol. Interact. 78 (1991) 321–331.
[41] E. Weerapana, C. Wang, G.M. Simon, F. Richter, S. Khare, M.B. Dillon, D.A.
Bachovchin, K. Mowen, D. Baker, B.F. Cravat, Quantitative reactivity profiling
predicts functional cysteines in proteomes, Nature 468 (2010) 790–795.
[42] S.B. Wall, M.R. Smith, K. Ricart, F. Zhou, P.K. Vayalil, J.Y. Oh, A. Landar,
Detection of electrophile-sensitive proteins, Biochim. Biophys. Acta (2013)
[Epub ahead of print].