Oxidation of diclofenac to reactive intermediates by neutrophils, myeloperoxidase, and hypochlorous acid

Article abstract of DOI:10.1021/tx960190k

Diclofenac is associated with a low, but significant, incidence of hepatotoxicity and bone marrow toxicity. It has been suggested that this could be due to a reactive acyl glucuronide. An alternative hypothesis is that an oxidative reactive metabolite could be responsible for such reactions and such metabolites formed by the enzymes present in neutrophils could be responsible for bone marrow toxicity. Others had reported the formation of 2,2'-dihydroxyazobenzene during the oxidation of diclofenac by myeloperoxidase/hydrogen peroxide. In contrast, in similar experiments we did not find evidence for the formation of 2,2'-dihydroxyazobenzene, but we did find several products, including a reactive iminoquinone. The same iminoquinone was formed by the oxidation of 5-hydroxydiclofenac. This iminoquinone was also formed by oxidation of diclofenac by HOCl or by activated neutrophils. It reacted with glutathione to form a conjugate. 5- Hydroxydiclofenac is also a major hepatic metabolite of diclofenac, and we found that rat hepatic microsomes oxidized 5-hydroxydiclofenac to the iminoquinone which was trapped with glutathione. This reactive metabolite represents another possible cause of the idiosyncratic reactions associated with the use of diclofenac.

Full text of DOI:10.1021/tx960190k

414
Chem. Res. Toxicol. 1997, 10, 414-419
Oxid a tion of Diclofen a c to Rea ctive In ter m ed ia tes by
Neu tr op h ils, Myelop er oxid a se, a n d Hyp och lor ou s Acid
Gohachiro Miyamoto,§,‡ Nasir Zahid,§ and J ack P. Uetrecht*,|,†,§
Faculties of Pharmacy and Medicine and University of Toronto Drug Safety Research Group,
University of Toronto, Toronto, Ontario M5S 2S2, Canada
Received November 19, 1996X
Diclofenac is associated with a low, but significant, incidence of hepatotoxicity and bone
marrow toxicity. It has been suggested that this could be due to a reactive acyl glucuronide.
An alternative hypothesis is that an oxidative reactive metabolite could be responsible for
such reactions and such metabolites formed by the enzymes present in neutrophils could be
responsible for bone marrow toxicity. Others had reported the formation of 2,2′-dihydroxy-
azobenzene during the oxidation of diclofenac by myeloperoxidase/hydrogen peroxide. In
contrast, in similar experiments we did not find evidence for the formation of 2,2′-dihydroxy-
azobenzene, but we did find several products, including a reactive iminoquinone. The same
iminoquinone was formed by the oxidation of 5-hydroxydiclofenac. This iminoquinone was
also formed by oxidation of diclofenac by HOCl or by activated neutrophils. It reacted with
glutathione to form a conjugate. 5-Hydroxydiclofenac is also a major hepatic metabolite of
diclofenac, and we found that rat hepatic microsomes oxidized 5-hydroxydiclofenac to the
iminoquinone which was trapped with glutathione. This reactive metabolite represents another
possible cause of the idiosyncratic reactions associated with the use of diclofenac.
Acyl glucuronides have relatively long biological half-
lives and can reach sites distant from where they are
formed, but this is not true for most reactive metabolites.
Therefore, it is logical to examine the formation of
reactive metabolites by the target of an idiosyncratic
reaction. In the case of bone marrow toxicity, the target
is the bone marrow cells, which include neutrophils and
neutrophil precursors. We have demonstrated that es-
sentially all of the drugs associated with the highest
incidence of agranulocytosis are oxidized to reactive
metabolites by activated neutrophils or by HOCl that is
generated by the myeloperoxidase (MPO) system of
neutrophils (19-21). Many of the drugs associated with
agranulocytosis are arylamines, and diclofenac is a
secondary arylamine.
Although the hepatotoxicity of diclofenac may be due
to an acyl glucuronide formed in the liver, agranulo-
cytosis may be due to a reactive metabolite formed by
activated neutrophils. Zuurbier et al. studied the oxida-
tion of diclofenac by MPO/H2O2 and found strong evi-
dence for the production of 2,2′-dihydroxyazobenzene (22);
however, it is difficult to understand how such a product
could be formed. In this study, we set out to determine
if diclofenac is oxidized by activated neutrophils as well
as simply by HOCl or MPO and, if it is, to determine if
a reactive intermediate is involved.
In tr od u ction
Diclofenac is a widely used nonsteroidal antiinflam-
matory drug (NSAID)1. All NSAIDs are associated with
a relatively high incidence of adverse reactions, especially
gastrointestinal bleeding. Although uncommon, diclofen-
ac appears to be associated with hepatitis (1, 2) and
hematological toxicity, such as aplastic anemia, neutro-
penia (3-6), hemolytic anemia, and thrombocytopenia (5,
7-9).
There is circumstantial evidence that idiosyncratic
reactions are often due to chemically reactive metabolites
(10, 11). One type of reactive metabolite associated with
many NSAIDs is an acyl glucuronide, and it has been
proposed that these acyl glucuronides are responsible for
many of the idiosyncratic reactions associated with
NSAIDs (12). It has been demonstrated that diclofenac
forms an acyl glucuronide, and covalent binding of
diclofenac, apparently due to the acyl glucuronide, has
been observed in the liver (13-16). Although NSAIDs,
as a class, are associated with a relatively high incidence
of idiosyncratic reactions, only a few, such as phenyl-
butazone, indomethacin, and diclofenac, appear to be
associated with a low, but statistically significant inci-
dence of bone marrow toxicity (4). Although diclofenac-
associated bone marrow toxicity may be due to the
reactive acyl glucuronide, based on its structure, the
glucuronide of phenylbutazone presumably is not reac-
tive. However, phenylbutazone is oxidized to reactive
intermediates by neutrophil-derived peroxidases (17, 18).
Ma ter ia ls a n d Meth od s
Ma ter ia ls. Diclofenac was purchased from Sigma Chemical
Co. (St. Louis, MO). NaOCl and 2,2′-dihydroxyazobenzene were
obtained from Aldrich Chemical Co. Inc. (Milwaukee, WI), and
the concentration of NaOCl was determined spectrophotometri-
cally (23). MPO was obtained from Cortex Biochem Inc. (San
Leandrow, CA), and 1 unit of MPO was defined as the amount
of enzyme that would decompose 1 µM H2O2/min at 25 °C and
pH 6. 5-Hydroxydiclofenac was a generous gift from Ciba-Geigy
Canada Ltd. (Calgary, Alberta).
* Address correspondence to this author at the Faculty of Pharmacy,
University of Toronto, 19 Russell St., Toronto, Ontario, Canada, M5S
2S2. Telephone: 416 978 8939. E-mail: jack.uetrecht@utoronto.ca.
† University of Toronto Drug Safety Research Group.
‡ Present address: Otsuka Pharmaceutical Co., Tokushima, J apan.
§ Faculty of Pharmacy.
| Faculty of Medicine.
X Abstract published in Advance ACS Abstracts, March 15, 1997.
1 Abbreviations: MPO, myeloperoxidase; PMA, phorbol 12-myristate
13-acetate; HBSS, Hanks’ balanced salt solution; NSAID, nonsteroidal
antiinflammatory drug.
An a lytica l P r oced u r es. HPLC was performed with UV
detection at a wavelength of 254 nm. The conditions utilized
S0893-228x(96)00190-7 CCC: $14.00 © 1997 American Chemical Society

Oxidation of Diclofenac
Chem. Res. Toxicol., Vol. 10, No. 4, 1997 415
for most of the HPLC consisted of a column packed with
Ultracarb 5 ODS 30 (100 mm × 2.0 mm i.d.; Phenomenex;
Torrance, CA) and a mobile phase of CH3CN/H2O/CH3COOH
(50:50:1, v/v) at a flow rate of 0.2 mL/min. A mobile phase
containing 35% acetonitrile was occasionally used to increase
the resolution of compounds with short retention times. For
preparative HPLC, the column had the same packing material
but the dimensions were 150 × 10 mm. Ultraviolet/visible
spectra were obtained with a Hewlett Packard 8452A diode
array spectrophotometer (Hewlett Packard Co.; Palo Alto, CA)
or, if done in conjunction with HPLC, with a Hewlett Packard
1040A diode array HPLC detector.
F igu r e 1. HPLC of the products of diclofenac oxidation by
HOCl. The concentration of diclofenac was 10 mM, and that of
HOCl was 100 mM.
Mass spectrometry was performed with a PE SCIEX API-III
Biomolecular Mass Analyzer in the ion spray mode (Sciex,
Toronto, Ontario). Most of the studies were done in the LC/
MS mode using the same HPLC column and solvent as described
above for simple HPLC.
LC/MS. A standard of 5-hydroxydiclofenac was also oxidized
with HOCl and glutathione added in the same manner.
Oxid a tion of 5-Hyd r oxyd iclofen a c by Ra t Hep a tic Mi-
cr osom es. Rat hepatic microsomes (Sprague Dawley, 1 mg/
mL) were suspended in 1 mL of phosphate buffer (pH 7.4), and
5-hydroxydiclofenac (final concentration 1 mM), glutathione
(final concentration 5 mM), and finally NADPH (1 mM) were
added. After incubation at 37 °C for 30 min, the mixture was
passed through a Supelclean tube (3 mL, LC-18 SPE; Supelco,
Mississauga, Ontario) and washed with water. The compounds
were then eluted with 50% methanol in water, and the solvent
was removed under a stream of nitrogen. The residue was
redisolved in 0.1 mL of 50% ethanol in water, and 5 µL of this
was analyzed by LC/MS as described above.
NMR spectra of M5 and M6 were obtained with a Varian
Unity Plus 500 MHz spectrometer (Varian Associates Inc., Palo
Alto, CA). NMR spectra of M1 and M2 were obtained with a
Bruker 500 MHz NMR spectrometer (Brucker Canada; Milton,
Ontario). Metabolites were dissolved in CD3OD.
Oxid a tion of Diclofen a c by HOCl. Diclofenac, dissolved
in methanol, was reacted with equimolar concentrations of
HOCl dissolved in water at final concentrations from 1 mM to
10 mM for 30 min, and the products were analyzed by HPLC.
Diclofenac was not sufficiently soluble in water to achieve this
concentration; however, when the oxidation was carried out in
the absence of methanol at lower concentrations, the pattern of
products was essentially the same (data not shown).
Resu lts
For isolation of larger quantities of products, diclofenac (100
mM) was oxidized with an equal volume of HOCl (0.7 M), and
the products were purified by preparative HPLC and analyzed
by MS/MS and NMR.
Oxid a tion of Diclofen a c by HOCl. The reaction of
diclofenac with HOCl was slow relative to the analogous
reaction of many other arylamines that we have studied,
and an excess of HOCl was utilized in order to obtain
products in a short period of time. A search was made
for 2,2′-dihydroxyazobenzene, which has a retention time
with our HPLC system of 29 min, but no peak with this
retention time was observed. We also used LC/MS with
selected ion monitoring at m/z 215 in case a small peak
was hidden in the base line, but this product was not
observed.
Oxidation of diclofenac by HOCl led to several products
as shown by HPLC in Figure 1. Some of these (M1, M2,
M3, M4, M5, M6) were tentatively identified by LC/MS
and/or NMR as described below, and this led to the
proposed pathway shown in Figure 2.
M1 had a retention time of 19 min on HPLC, and its
MS revealed a M+1 ion at m/z 330 which represents
substitution of a chlorine for a hydrogen atom on di-
clofenac. The proton NMR spectra of this product (Table
1) indicate that the chlorine is in the 3 position as shown
in Figure 2.
M2 was the major product and had a retention time of
25 min on HPLC. Its mass spectrum was similar to that
of M1 with a M+1 ion at m/z 330. The NMR spectra
(Table 1) indicate that the chlorine was in the 5 position.
M3, with a retention time of 39 min on HPLC, had a
M+1 ion on mass spectrometry at m/z 366, indicating the
substitution of two chlorines. From the structures of M1
and M2, it is likely that the chlorines were in the 3 and
5 positions; however, we did not purify sufficient material
to prove this by NMR.
The initial products of the reaction of diclofenac with HOCl
were analyzed with a flow system in which the reactants were
continuously pumped (Harvard infusion pumps, Model 55-2222,
Harvard Apparatus, South-Natick, MA) into a mixing chamber
(dead volume 3.1 µL; Upchurch Mixing Tee; Upchurch Scientific;
Oak Harbor, WA) and from there the products were fed into
the Sciex mass spectrometer. Infusion rates of each pump were
2-20 µL/min, and the flow rates of each pump were the same.
The concentration of diclofenac was 1 mM, and that of HOCl
was 100 mM. Diclofenac was dissolved in methanol, and HOCl
was diluted with water.
Oxid a tion of Diclofen a c by MP O/H 2O2/Cl-. Diclofenac
(from 0.1 to 1 mM) was incubated with MPO (2, 4, and 6 units)
and hydrogen peroxide (10 µL, 0.8 mM) in 1 mL of 0.1 M
phosphate buffer (pH 5 or pH 6) containing 150 mM KCl for
1.5 h. The products were analyzed by HPLC.
In order to duplicate the experiment of Zuurbier et al. (22),
the oxidation of diclofenac [200 µM in phosphate buffer, pH 7.2,
(without chloride)] by H2O2 (150 µM) in the presence of MPO (3
µM) was followed with a photodiode array spectrophotometer.
Oxid a tion of Diclofen a c by Activa ted Neu tr op h ils.
Blood was drawn from healthy subjects into a heparinized
syringe. Neutrophils were isolated by differential centrifugation
with Ficoll-Hypaque as described previously (24). Cell viability
was greater than 95% as determined by trypan blue exclusion.
Diclofenac (1, 5, 10, 50 and 100 µM) was incubated with
neutrophils [2 × 106, 4 × 106, and 6 × 106 cells/mL and 20 nmol
of phorbol myristate acetate (PMA) in 1 mL of Hanks’ balanced
salt solution (HBSS)]. Metabolites were identified by LC/MS
and comparison with the products produced by oxidation with
HOCl. We did not have enough material for a standard curve,
and so the concentrations of the metabolites were approximated
by using the same extinction coefficient as that for diclofenac.
This is only a gross approximation to the true concentration.
M4 had a retention time of 5 min, and its M+1 ion on
mass spectrometry was at m/z 310 which corresponds to
the addition of oxygen and loss of two hydrogens. MS/
MS gave fragments at m/z 292 (100%), 264 (14%), 235
(18%), 229 (23%), 201 (34%), 194 (30%), and 166 (20%).
The λmax of this metabolite was 455.5 nm. When formed
In t er m ed ia t es Tr a p p ed b y Glu t a t h ion e. Diclofenac (1
mM in methanol) was oxidized with HOCl (100 mM in water),
and after 10 min, glutathione (1 mL of a 100 mM solution in
PBS, pH 7.4) was added, and the products were analyzed by

416 Chem. Res. Toxicol., Vol. 10, No. 4, 1997
Miyamoto et al.
Other products were observed on the chromatogram
(Figure 1), but some of these products appear to be
unstable because the chromatographic pattern changed
with time although part of this was likely due to the
excess of HOCl causing further oxidation.
A flow system was utilized in which diclofenac and
HOCl were mixed and the products went directly into
the mass spectrometer in order to try to identify the
earliest intermediates in the reaction. By varying the
flow rate, one can get a crude estimate of the rate of the
reaction and the half-life of the intermediates, and in this
case, the range used was 1-20 µL/min. At a rate of 1
µL/min (in which case the reaction time was about 70 s),
a chlorinated product was observed; however, at faster
flow rates, the amount of product was not sufficient for
detection by MS. Thus, the reaction rate was too slow
for this method to provide further information about the
nature of initial products of the reaction.
Oxid a tion of Diclofen a c by MP O/H2O2/Cl-. Di-
clofenac was also oxidized by the combination of MPO/
H2O2/Cl-, and the major products were M4 and M6 as
shown by HPLC. The amount of M4 relative to M6 was
greater at pH 5 than at pH 6 (data not shown). The
chlorinated products formed by oxidation with HOCl were
not observed. We again searched for the presence of 2,2′-
dihydroxyazobenzene by HPLC using a synthetic stand-
ard of the compound to determine the retention time, but
none was observed.
When we tried to duplicate the conditions used by
Zuurbier, which did not include chloride, we also obtained
an absorption band at about 452 nm (Figure 3) that
increased with time, and it looks very similar to what
they observed; however, an overlay of the absorption
spectrum of 2,2′-dihydroxyazobenzene (dashed line) dem-
onstrates that there are significant differences between
the observed spectra generated during the oxidation of
diclofenac with MPO/H2O2 and that of 2,2′-dihydroxy-
azobenzene. Although M4 has a λmax close to 450 nm, it
also absorbs significantly between 300 and 400 nm, and
so it is difficult to ascribe this peak solely to M4. On
addition of ascorbic acid to the reaction mixture, the 452
nm peak very rapidly disappeared. In contrast, addition
of ascorbic acid to 2,2′-dihydroxyazobenzene did not
eliminate the absorption band at ∼450 nm. We were also
unable to detect 2,2′-dihydroxyazobenzene in this incuba-
tion using HPLC.
F igu r e 2. Proposed pathways for the oxidation of diclofenac
by HOCl. The proposed intermediates in brackets were not
observed.
by oxidation of diclofenac with HOCl, its concentration
peaked after about 10 min and then decreased (data not
shown). When the peak for M4 was collected from
preparative HPLC and attempts were made to isolate it,
it decomposed; however, if left in the acidic HPLC solvent
it spontaneously converted to M6 as well as a trace of
M5 over a period of days (data not shown). Its proposed
chemical structure is shown in Figure 2. Addition of
ascorbate to the reaction mixture led to the disappear-
ance of M4 and the appearance of 5-hydroxydiclofenac,
confirmed by comparison with the synthetic standard,
and it had a M+1 ion at m/z 312. When 5-hydroxy-
diclofenac was oxidized with HOCl, the major product
had the same retention time on HPLC, the same M+1
ion at m/z 310, and the same fragmentation pattern on
MS/MS as M4 produced from diclofenac.
Oxid a tion of Diclofen a c by Activa ted Neu tr o-
p h ils. Using the neutrophil system, metabolites of
diclofenac were detected by HPLC (Figure 4). The major
product had a retention time of 5 min. Using a mobile
phase containing 35% rather than 50% acetonitrile, the
retention time was 15.8 min, and when it was mixed with
the product of oxidation of 5-hydroxydiclofenac by HOCl,
there was only one peak on the chromatogram, indicating
that the major metabolite must be M4. LC/MS confirmed
that this product had a M+1 ion at m/z 310. The other
two products both had M+1 ions at m/z 282, and they
were presumably M5 and M6. M4 spontaneously con-
verted to the two products with M+1 ions at m/z 282.
Other metabolites were detected on the chromatogram
at an earlier retention time than M4, and these appear
to be unstable intermediates. Chlorinated metabolites,
M1, M2, and M3, which were detected by oxidation of
diclofenac by HOCl, were present in trace amounts or
nil. The concentration of major metabolite, M4, was
dependent on the concentration of diclofenac and the
number of neutrophils (Figures 5, 6).
M5 had a retention time of 7 min, and on mass
spectrometry, its M+1 ion was at m/z 282 with a chlorine
isotope peak at m/z 284 (69% of the M+1 ion). The
fragment ions were at m/z 264 (80%, 282-H2O), 229
(32%, 282-H2O-Cl), 201 (22%, 282-H2O,-CO,-Cl), and
194 (37%, 282-H2O-2Cl). The NMR data are shown in
Table 1. The COSY spectrum was consistent with the
proton assignments. The proposed structure of M5 is
shown in Figure 2.
M6 has a retention time of 9 min and also has a M+1
ion of m/z 282 (57%) and a chlorine isotope peak at m/z
284 (66% of the M+1 ion). Its fragment ions were at m/z
264 (90%, 282-H2O), 247 (72%, 282-Cl), and 230 (100%,
282-HOCl). Its NMR spectral data are shown in Table
1, and its proposed chemical structure is shown in Figure
2. The COSY spectrum was consistent with the proton
assignments. A C13 NMR spectrum had a major peak
at 195.75 ppm, which is consistent with the presence of
an aldehyde carbon. (The carboxylate carbon in di-
clofenac is at 180.62 ppm.)

Oxidation of Diclofenac
Chem. Res. Toxicol., Vol. 10, No. 4, 1997 417
Ta ble 1.1H-NMR Sp ectr a of Diclofen a c, M1, M2, M5, a n d M6
1H position
3
4
5
6
3′ & 5′
4′
CH2
diclofenac
δ (ppm)
multiplicity
J (Hz)
integration
M1
6.355
d,d
8.06, 1.1
1H
6.959
t,d
8.06, 1.65
1H
6.815
t,d
7.32, 1.28
1H
7.186
d,d
7.32, 1.29
1H
7.355
d
8.05
2H
6.966
t
7.87
1H
3.623
s
2H
δ (ppm)
multiplicity
J (Hz)
integration
M2
-
-
-
-
7.240
d,d
8.06, 1.47
1H
6.831
t
8.05
1H
7.214
d,d
7.69, 1.47
1H
7.242
d
8.06
2H
7.068
t
7.78
1H
3.766
s
2H
δ (ppm)
multiplicity
J (Hz)
integration
M5
6.369
d
8.06
1H
7.052
d,d
8.61, 2.56
1H
-
-
-
-
7.233
d
2.38
1H
7.404
d
8.06
2H
7.084
t
8.06
1H
3.738
s
2H
δ (ppm)
multiplicity
J (Hz)
integration
M6
δ (ppm)
multiplicity
J (Hz)
6.679
d
10.07
1H
6.468
d,d
10.07, 2.19
1H
-
-
-
-
6.795
q
2.20
1H
7.472
d
8.05
2H
7.167
t
8.05
1H
4.762
d
2.02
2H
CHO
9.886
s
6.215
d
8.97
1H
6.904
d,d
8.97, 2.93
1H
-
-
-
-
7.112
d
2.93
1H
7.490
d
8.24
2H
7.246
t
8.24
1H
integration
1H
F igu r e 3. Absorption spectra from the oxidation of diclofenac
with MPO/H2O2. The concentration of diclofenac was 200 µM,
that of MPO was 3 µM, and that of H2O2 was 150 µM. The
time between spectra was 1 min. The first spectrum is the
bottom spectrum. The spectrum represented by the dotted line
is that of authentic 2,2′-dihydroxyazobenzene.
F igu r e 5. Formation of M4 by oxidation of diclofenac with
activated neutrophils as a function of time and at different
concentrations of diclofenac. The absolute concentrations of M4
are only approximate and are based on the approximation that
the extinction coefficient of M4 is equal to that of diclofenac.
new peak was a glutathione adduct with a M+1 ion at
m/z 617. This corresponds to the adduct that would be
expected from the reaction between M4 and glutathione.
In addition, LC/MS/MS gave fragments at m/z 542 (19%,
617-Gly), 488 (11%, 617-Glu), 385 (33%, 617-Glu-
Cys), 342 (100%, 617-GSH+S), 324 (34%, 342-H2O),
and 310 (6%, 617-GSH). The major product formed by
oxidation of 5-hydroxydiclofenac with HOCl followed by
addition of glutathione also had a M+1 ion at m/z 617,
and by MS/MS, it also had the same fragmentation
pattern as that of the glutathione adduct formed from
oxidation of diclofenac.
Oxid a tion of 5-Hyd r oxyd iclofen a c by Ra t Hep a tic
Micr osom es. We suspected that M4 might not be
detected in the incubation of 5-hydroxydiclofenac with
hepatic microsomes because of reaction with the mi-
crosomes; therefore, we added glutathione to trap it. We
found the glutathione adduct with a M+1 ion at m/z 617,
and it had the same fragementation pattern on MS/MS
as the M4-glutathione conjugate formed by oxidation of
diclofenac with HOCl.
F igu r e 4. HPLC of the products of diclofenac oxidation by
activated neutrophils. The concentration of diclofenac was 100
µM, and the number of neutrophils was 6 × 106 cells/mL.
GSH Ad d u ct w ith Rea ctive Meta bolite. When
glutathione was added to the products of the reaction of
diclofenac with HOCl, M4 (retention time 5 min) disap-
peared, and a new peak with a retention time of 2 min
was observed. LC/MS of this product indicated that this

418 Chem. Res. Toxicol., Vol. 10, No. 4, 1997
Miyamoto et al.
metabolites generated may be responsible for some of the
idiosyncratic reactions associated with diclofenac, espe-
cially those involving the bone marrow. Although it
appears that several reactive species are formed during
the oxidation of diclofenac, including a presumed nitren-
ium ion intermediate, it is likely that M4 would make
the largest contribution to covalent binding, and it is
likely to react with protein sulfhydryl groups analogous
to its reaction with glutathione. The phenolic precursor
to M4, 5-hydroxydiclofenac, is also a significant hepatic
metabolite of diclofenac (28), and, therefore, M4 may be
formed by oxidation of this phenol in the bone marrow.
In addition, we found that M4 was formed by the
oxidation of 5-hydroxydiclofenac with rat hepatic mi-
crosomes, so it is likely that M4 is also formed by
oxidation of 5-hydroxydiclofenac in the liver of humans.
Hargus et al. found evidence for P-450-mediated covalent
binding of diclofenac in the liver (29); therefore, it is
possible that M4 could also be responsible for the
hepatotoxicity associated with the diclofenac.
F igu r e 6. Formation of M4 by oxidation of diclofenac with
activated neutrophils as a function of time and at different
concentrations of neutrophils. The absolute concentrations of
M4 are only approximate and are based on the approximation
that theextinction coefficient of M4 is equal to that of diclofenac.
Ack n ow led gm en t. This work was supported by a
grant from the Medical Research Council of Canada (MT-
13478). We thank Dr. Henrianna Pang for her expert
analysis of the mass spectra and Dr. R. A. McClelland
for help with the structural analysis.
Discu ssion
We failed to confirm the production of 2,2′-dihydroxy-
azobenzene in the oxidation of diclofenac by the MPO/
H2O2 system. Although the evidence provided in a
previous report appeared substantial, it is very difficult
to imagine a mechanism by which this product could be
formed. However, we did find a series of products.
Oxidation of diclofenac by HOCl produced primarily
chlorination of the benzene ring but also significant
amounts of other products. In contrast, M4 was the
major product observed with oxidation by MPO or
activated neutrophils. Both chlorination of the ring and
formation of the quinone-type products could involve
N-chlorination followed by loss of a chloride anion to form
a nitrenium ion that could react with either water or
chloride ion. The putative nitrenium ion may be stabi-
lized by the formation of an intramolecular ion pair as
shown in Figure 2. The reason for the difference in the
products observed in the MPO and neutrophil incubation
compared with oxidation by HOCl is unknown. Although
it seems as if neutophils generate free HOCl, the MPO
system also appears to be able to oxidize compounds
without the involvement of HOCl (25). It could be that
oxidation by the MPO system does not involve an
N-chloro intermediate. Alternatively, it could be that the
different environment of the putative N-chloro interme-
diate in the MPO and neutrophil systems leads to
preferential formation of the phenol rather than reaction
with chloride or rearrangement to form M1 or M2.
Despite the differences in the pattern of products formed,
M4 is a significant reactive intermediate formed by all
three systems.
Refer en ces
(1) Dunk, A. A., Walt, R. P., J enkins, W. J ., and Sherlock, S. S. (1982)
Diclofenac hepatitis. Br. Med. J . Clin. Res. Ed. 284, 1605-1606.
(2) Helfgott, S. M., Sandberg-Cook, J ., Zakim, D., and Nestler, J .
(1990) Diclofenac-associated hepatotoxicity [see comments]. J AMA,
J . Am. Med. Assoc. 264, 2660-2662.
(3) Eustace, S., O’Neill, T., McHale, S., and Molony, J . (1989) Fatal
aplastic anaemia following prolonged diclofenac use in an elderly
patient. Ir. J . Med. Sci. 158, 217.
(4) Kaufman, D. W., Kelly, J . P., Levy, M., and Shapiro, S. (1991)
The drug etiology of agranulocytosis and aplastic anemia, Oxford
University Press, New York.
(5) Kim, H. L., and Kovacs, M. J . (1995) Diclofenac-associated
thrombocytopenia and neutropenia. Ann. Pharmacother. 29, 713-
715.
(6) Salama, A., Schu¨tz, B., Kiefel, V., Breithaupt, H., and Mueller-
Eckhardt, C. (1989) Immune-mediated agranulocytosis related to
drugs and their metabolites: mode of sensitization and hetero-
geneity of antibodies. Br. J . Haematol. 72, 127-132.
(7) Kramer, M. R., Levene, C., and Hershko, C. (1986) Severe
reversible autoimmune haemolytic anaemia and thrombocyto-
penia associated with diclofenac therapy. Scand. J . Haematol.
36, 118-120.
(8) Epstein, M., Vickars, L., and Stein, H. (1990) Diclofenac induced
immune thrombocytopenia. J . Rheumatol. 17, 1403-1404.
(9) Salama, A., Gottsche, B., and Mueller-Eckhardt, C. (1991) Auto-
antibodies and drug- or metabolite-dependent antibodies in
patients with diclofenac-induced immune haemolysis. Br. J .
Haematol. 77, 546-549.
(10) Nelson, S. D., and Pearson, P. G. (1990) Covalent and noncovalent
interactions in acute lethal cell injury caused by chemicals. Annu.
Rev. Pharmacol. Toxicol. 30, 169-195.
(11) Hinson, J . A., and Roberts, D. W. (1992) Role of covalent and
noncovalent interactions in cell toxicity: effects on proteins. Annu.
Rev. Pharmacol. Toxicol. 32, 471-510.
(12) Spahn-Langguth, H., and Benet, L. Z. (1992) Acyl glucuronides
revisited: is the glucuronidation process a toxification as well as
a detoxification mechanism. Drug Metab. Rev. 24, 5-47.
(13) Pumford, N. R., Myers, T. G., Davila, J . C., Highet, R. J ., and
Pohl, L. R. (1993) Immunochemical detection of liver protein
adducts of the nonsteroidal anti-inflammatory drug diclofenac.
Chem. Res. Toxicol. 6, 147-150.
(14) Kretz-Rommel, A., and Boelsterli, U. A. (1994) Mechanism of
covalent adduct formation of diclofenac to rat hepatic microsomal
proteins. Retention of the glucuronic acid moiety in the adduct.
Drug Metab. Dispos. 22, 956-961.
It is likely that the reactions observed in this study
are involved in the ability of diclofenac to inhibit HOCl
formation by the MPO that was described by others (22).
Monocytes from patients receiving diclofenac have also
been shown to produce less superoxide when stimulated
(26). On the other hand, the reaction of diclofenac with
HOCl is relatively slow, and another study suggests that
the in vivo scavenging effects of diclofenac would be small
(27). Despite the observation that the reaction of di-
clofenac with HOCl is relatively slow, it is likely that a
small amount of diclofenac could be oxidized by activated
neutrophils or monocytes in vivo, and the reactive
(15) Kretz-Rommel, A., and Boelsterli, U. A. (1994) Selective protein
adducts to membrane proteins in cultured rat hepatocytes exposed
to diclofenac: radiochemical and immunochemical analysis. Mol.
Pharmacol. 45, 237-244.

Oxidation of Diclofenac
Chem. Res. Toxicol., Vol. 10, No. 4, 1997 419
(16) Hargus, S. J ., Martin, B. M., George, J . W., and Pohl, L. R. (1995)
Covalent modification of rat liver dipeptidyl peptidase IV (CD26)
by the nonsteroidal anti-inflammatory drug diclofenac. Chem.
Res. Toxicol. 8, 993-996.
(17) Kalyanaraman, B., and Sohnle, P. G. (1985) Generation of free
radical intermediates from foreign compounds by neutrophil-
derived oxidants. J . Clin. Invest. 75, 1618-1622.
(18) Ichihara, S., Tomisawa, H., Fukazawa, H., Tateishi, M., J oly, R.,
and Heintz, R. (1986) Involvement of leukocytes in the oxygen-
ation and chlorination reaction of phenylbutazone. Biochem.
Pharmacol. 35, 3935-3939.
(19) Uetrecht, J . P. (1992) The role of leukocyte-generated metabolites
in the pathogenesis of idiosyncratic drug reactions. Drug Metab.
Rev. 24, 299-366.
(20) Uetrecht, J . P., Ma, H. M., MacKnight, E., and McClelland, R.
(1995) Oxidation of aminopyrine by hypochlorite to a reactive
dication: possible implications for aminopyrine-induced agranulo-
cytosis. Chem. Res. Toxicol. 8, 226-233.
(21) Liu, Z. C., and Uetrecht, J . P. (1995) Clozapine is oxidized by
activated human neutrophils to a reactive nitrenium ion that
irreversibly binds to the cells. J . Pharmacol. Exp. Ther. 275,
1476-1483.
(22) Zuurbier, K. W. M., Bakkenist, A. R. J ., Fokkens, R. H., Nibbering,
N. M. M., Wever, R., and Muijsers, A. O. (1990) Interaction of
myeloperoxidase with diclofenac: inhibition of the chlorinating
activity of myeloperoxidase by diclofenac and oxidation of di-
clofenac to dihydroxyazobenzene by myeloperoxidase. Biochem.
Pharmacol. 40, 1801-1808.
(23) Hussain, A., Trudell, P., and Repta, A. J . (1970) Quantitative
spectrophotometric methods for determination of sodium hypo-
chloride in aqueous solutions. J . Pharm. Sci. 59, 1168-1170.
(24) Klebanoff, S. J ., Olszowski, S., Voorhis, W. C. V., Ledbetter, J .
A., Waltersdorph, A. M., and Schlechte, K. G. (1992) Effects of
g-interferon on human neutrophils: Protection from deterioration
on storage. Blood 80, 225-234.
(25) Marquez, L. A., and Dunford, H. B. (1994) Chlorination of taurine
by myeloperoxidase: kinetic evidence for an enzyme-bound
intermediate. J . Biol. Chem. 269, 7950-7956.
(26) Bell, A. L., Adamson, H., Kirk, F., McCaigue, M. D., and Rotman,
H. (1991) Diclofenac inhibits monocyte superoxide production ex
vivo in rheumatoid arthritis. Rheumatol. Int. 11, 27-30.
(27) Wasil, M., Halliwell, B., Morehouse, C. P., Hutchison, D. C. S.,
and Baum, H. (1987) Biologically-significant scavenging of the
myeloperoxidase-derived oxidant hypochlorous acid by some anti-
inflammatory drugs. Biochem. Pharmacol. 36, 3847-3850.
(28) Sawchuk, R. J ., Maloney, J . A., Cartier, L. L., Rackley, R. J ., Chan,
K. K., and Lau, H. S. (1995) Analysis of diclofenac and four of its
metabolites in human urine by HPLC. Pharm. Res. 12, 756-
762.
(29) Hargus, S. J ., Amouzedeh, H. R., Pumford, N. R., Myers, T. G.,
McCoy, S. C., and Pohl, L. R. (1994) Metabolic activation and
immunochemical localization of liver protein adducts of the
nonsteroidal anti-inflammatory drug diclofenac. Chem. Res. Toxi-
col. 7, 575-582.
TX960190K