N-demethylation of N-methyl-4-aminoantipyrine, the main metabolite of metamizole

Article abstract of DOI:10.1016/j.ejps.2018.05.003

Metamizole is an old analgesic used frequently in some countries. Active metabolites of metamizole are the non-enzymatically generated N-methyl-4-aminoantipyrine (4-MAA) and its demethylation product 4-aminoantipyrine (4-AA). Previous studies suggested that 4-MAA demethylation can be performed by hepatic cytochrome P450 (CYP) 3A4, but the possible contribution of other CYPs remains unclear. Using human liver microsomes (HLM), liver homogenate and HepaRG cells, we could confirm 4-MAA demethylation by CYPs. Based on CYP induction (HepaRG cells) and CYP inhibition (HLM) we could identify CYP2B6, 2C8, 2C9 and 3A4 as major contributors to 4-MAA demethylation. The 4-MAA demethylation rate by HLM was 280 pmol/mg protein/h, too low to account for in vivo 4-MAA demethylation in humans. Since peroxidases can perform N-demethylation, we investigated horseradish peroxidase and human myeloperoxidase (MPO). Horse radish peroxidase efficiently demethylated 4-MAA, depending on the hydrogen peroxide concentration. This was also true for MPO; this reaction was saturable with a K_m of 22.5 μM and a maximal velocity of 14 nmol/min/mg protein. Calculation of the entire body MPO capacity revealed that the demethylation capacity by granulocyte/granulocyte precursors was approximately 600 times higher than the liver capacity and could account for 4-MAA demethylation in humans. 4-MAA demethylation could also be demonstrated in MPO-expressing granulocyte precursor cells (HL-60). In conclusion, 4-MAA can be demethylated in the liver by several CYPs, but hepatic metabolism cannot fully explain 4-MAA demethylation in humans. The current study suggests that the major part of 4-MAA is demethylated by circulating granulocytes and granulocyte precursors in bone marrow.

Full text of DOI:10.1016/j.ejps.2018.05.003

European Journal of Pharmaceutical Sciences 120 (2018) 172–180
Contents lists available at ScienceDirect
European Journal of Pharmaceutical Sciences
journal homepage: www.elsevier.com/locate/ejps
N-demethylation of N-methyl-4-aminoantipyrine, the main metabolite of
metamizole
Fabio Bachmanna,b, Urs Duthalera,b, Deborah Rudina,b, Stephan Krähenbühla,b,⁎
Manuel Haschke
,
c
a
Division of Clinical Pharmacology & Toxicology, University Hospital Basel, Switzerland
Department of Biomedicine, University of Basel, Switzerland
b
c
Clinical Pharmacology and Toxicology, Department of General Internal Medicine, Inselspital, Bern University Hospital, University of Bern and Institute of Pharmacology,
University of Bern, Switzerland
A R T I C L E I N F O
A B S T R A C T
Keywords:
Metamizole
N-methyl-4-aminoantipyrine
N-demethylation
Cytochrome P450 (CYP)
Myeloperoxidase
Metamizole is an old analgesic used frequently in some countries. Active metabolites of metamizole are the non-
enzymatically generated N-methyl-4-aminoantipyrine (4-MAA) and its demethylation product 4-aminoanti-
pyrine (4-AA). Previous studies suggested that 4-MAA demethylation can be performed by hepatic cytochrome
P450 (CYP) 3A4, but the possible contribution of other CYPs remains unclear. Using human liver microsomes
(HLM), liver homogenate and HepaRG cells, we could confirm 4-MAA demethylation by CYPs. Based on CYP
induction (HepaRG cells) and CYP inhibition (HLM) we could identify CYP2B6, 2C8, 2C9 and 3A4 as major
contributors to 4-MAA demethylation. The 4-MAA demethylation rate by HLM was 280 pmol/mg protein/h, too
low to account for in vivo 4-MAA demethylation in humans. Since peroxidases can perform N-demethylation, we
investigated horseradish peroxidase and human myeloperoxidase (MPO). Horse radish peroxidase efficiently
demethylated 4-MAA, depending on the hydrogen peroxide concentration. This was also true for MPO; this
reaction was saturable with a K of 22.5 μM and a maximal velocity of 14 nmol/min/mg protein. Calculation of
m
the entire body MPO capacity revealed that the demethylation capacity by granulocyte/granulocyte precursors
was approximately 600 times higher than the liver capacity and could account for 4-MAA demethylation in
humans. 4-MAA demethylation could also be demonstrated in MPO-expressing granulocyte precursor cells (HL-
60). In conclusion, 4-MAA can be demethylated in the liver by several CYPs, but hepatic metabolism cannot fully
explain 4-MAA demethylation in humans. The current study suggests that the major part of 4-MAA is de-
methylated by circulating granulocytes and granulocyte precursors in bone marrow.
1. Introduction
2003), stimulation of the endogenous opioid system (Tortorici et al.,
1996; Vanegas et al., 1997; Vasquez and Vanegas, 2000), and interac-
Metamizole (dipyrone) is an analgesic, antipyretic and spasmolytic
tion with glutamate transmission (Beirith et al., 1998; Siebel et al.,
2004) or with the endocannabinoid system (Escobar et al., 2012;
Maione et al., 2015). The spasmolytic effect of metamizole is experi-
mentally established, but the mechanism is also not fully clarified.
Opening of ATP sensitive potassium channels (Valenzuela et al., 2005)
and inhibition of G protein-coupled receptor (GPCR) mediated con-
striction (Gulmez et al., 2006) of vascular smooth muscle cells are
mechanisms currently discussed.
Metamizole is a prodrug that is spontaneously converted to 4-MAA,
which has a high oral bioavailability (> 80%) (Ergün et al., 2004). As
shown in Figs. 1, 4-MAA can be demethylated to 4-AA or formylated to
N-formyl-4-aminoantipyrine (4-FAA). 4-AA can be acetylated to N-
acetyl-4-aminoantipyrine (4-AAA), which is excreted mainly in the
drug that was introduced in Germany almost 100 years ago. Its mode of
action has been investigated in numerous studies but is still not com-
pletely elucidated. One study has shown that the two major metabolites
of metamizole, N-methyl-4-aminoantipyrine (4-MAA) and aminoanti-
pyrine (4-AA) inhibit COX-1 and COX-2 by interfering with Fe3+ con-
tained in the heme of the cyclooxygenases (Pierre et al., 2007). How-
ever, while the analgesic effect of metamizole is similar to
indomethacin, its anti-inflammatory effect is weaker (Tatsuo et al.,
1
994), suggesting also COX-independent pathways. Multiple additional
modes of action have been proposed, among them interaction with the
adrenergic nervous system (Silva et al., 2015), modulation of potassium
channels (Alves and Duarte, 2002; Duarte et al., 1992; Ortiz et al.,
⁎
Corresponding author at: Clinical Pharmacology & Toxicology, University Hospital, 4031 Basel, Switzerland.
E-mail address: stephan.kraehenbuehl@usb.ch (S. Krähenbühl).
https://doi.org/10.1016/j.ejps.2018.05.003
Received 21 January 2018; Received in revised form 17 April 2018; Accepted 6 May 2018
Available online 08 May 2018
0928-0987/ © 2018 Elsevier B.V. All rights reserved.

F. Bachmann et al.
European Journal of Pharmaceutical Sciences 120 (2018) 172–180
Fig. 1. Metabolism of metamizole. Metamizole is a prodrug efficiently converted non-enzymatically in the intestinal tract to N-methyl-4-aminoantipyrine (4-MAA).
-MAA has a good oral bioavailability (> 80%) and can be converted by C-oxidation to N-formyl-4-aminoantiyprine (4-FAA) or by N-demethylation to 4-ami-
noantipyrine (4-AA). 4-AA can be N-acetylated by NAT2 to N-acetyl-4-aminoantipyrine (4-AAA).
4
urine. Only 3% of the administered metamizole is excreted in the urine
as 4-MAA, the rest is excreted mainly as 4-AA, 4-AAA and 4-FAA as well
as additional minor metabolites (Flusser et al., 1988; Levy Micha and
Bernd, 1995; Noda et al., 1976a; Noda et al., 1976b; Volz and Kellner,
with impaired liver function, the half-life of 4-MAA is prolonged, sup-
porting the concept that 4-MAA is metabolized by the liver (Agundez
et al., 1995; Levy Micha and Bernd, 1995).
Although in most studies concerning N-demethylation of 4-DMAA
and 4-MAA the involvement of hepatic CYPs was investigated, it has to
be taken into account that also other enzymes can perform N-de-
methylation of xenobiotics (Benedetti, 2001; Benedetti et al., 2009). N-
demethylation of 4-DMAA has for instance been shown for horseradish
peroxidase (HRP) (Stiborova et al., 1997) and for soybean lipoxygenase
(Perez-Gilabert et al., 1997; Yang and Kulkarni, 1998).
Considering the uncertainties regarding N-demethylation of 4-MAA,
the aim of the current study was to investigate this step in the meta-
bolism of metamizole in more detail. Therefore, we studied the meta-
bolism of 4-MAA using different hepatic systems such as human liver
microsomes (HLM), differentiated HepaRG cells and human liver
homogenate as well as in the presence of isolated enzymes (horse radish
peroxidase, soybean lipoxygenase and human myeloperoxidase) and
human promyelotic leukemia cells (HL 60), which contain human
myeloperoxidase (Maseneni et al., 2012).
1980; Zylber-Katz et al., 1992).
Although the four main metabolites of metamizole have been well-
described, only the enzyme responsible for the formation of 4-AAA, a
polymorphous N-acetyltransferase (Agundez et al., 1995; Levy et al.,
1984), later identified as N-acetyltransferase type 2 (NAT2) (Agundez
et al., 1995), has been identified. In contrast, the enzymes involved in
the demethylation and formylation of 4-MAA are so far not known.
Metabolic studies with the structurally closely related N,N-dimethyl-4-
aminoantipyrine (4-DMAA) have shown that it can be converted at a
slow rate to 4-AA in rat liver microsomes (Bast and Noordhoek, 1981;
Imaoka et al., 1988; Inoue et al., 1983). Later experiments revealed that
approximately 1–2% of 4-MAA could be converted to 4-AA by human
liver microsomes (Geisslinger et al., 1996). This process could be par-
tially inhibited by 10 μM ketoconazole (Geisslinger et al., 1996), im-
plicating involvement of CYP 3A4 (Eagling et al., 1998). In humans
173

F. Bachmann et al.
European Journal of Pharmaceutical Sciences 120 (2018) 172–180
2. Material and Methods
substrate solution and stopped as described above. All samples were
stored at −20 °C until analysis.
2.1. Chemicals and Reagents
2.3. Assays With HepaRG Cells
Metamizole, N-methyl-4-aminoantipyrine (4-MAA), 4-aminoanti-
pyrine (4-AA), N-formyl-4-aminoantipyrine (4-FAA), N-acetyl-4-ami-
noantipyrine (4-AAA), ticlopidine, sulfaphenazole, (+)-N-3-benzylnir-
vanol, methylpyrazole, montelukast, PF-06281355, rifampicin and 3-
methylcholanthrene were purchased from Sigma-Aldrich (Buchs,
Switzerland). 4-MAA-d3.
HepaRG cells were obtained from Biopredic International (Rennes,
France) as undifferentiated cryopreserved cells. The cells were cultured
in a 5% CO and 95% humidified air environment at 37 °C. Freshly split
2
cells were seeded in 96 well plates (10,000 cells/well) and treated for
4 weeks as previously described (Aninat et al., 2006). After differ-
entiation, cells were induced with rifampicin (20 μM) or 3-methylcho-
lanthrene (5 μM) for 3 days, while control cells were treated with ve-
hicle (0.1% DMSO). Final vehicle concentration was 0.1%. Assays were
started by replacing the induction medium with 50 μL fresh William's E
medium containing 4-MAA or a control substrate (midazolam for ri-
fampicin, tizanidine for 3-methylcholanthrene). During the assay, the
4
-AA-d3, 4-AAA-d3, bisoprolol-d5, 8′-hydroxyefavirenz, efavirenz-
d4, flurbiprofen, 4′-hydroxyflurbiprofen, flurbiprofen-d4, omeprazole,
′-hydroxyomeprazole, omeprazole-d3, metoprolol,
5
α-hydro-
xymetoprolol, metoprolol-d7, tizanidine, hydroxytizanidine, tizanidine-
d4, furaphylline, paclitaxel, 6α-hydroxypaclitaxel, chlorzoxazone, and
6
1
-hydroxychlorzoxazone were acquired from TRC (Toronto, Canada).
′-hydroxymidazolam and midazolam-d6 were obtained from Lipomed
2
plates were kept at 37 °C and 5% CO in the cell culture incubator. The
(
Arlesheim, Switzerland). Midazolam and efavirenz were kindly pro-
reaction was stopped by adding 150 μL of ice cold IS solution at selected
time points (0, 15, 30, 45, 60, 75, 90, 105, 120 min). HepaRG cells were
gently scratched from the bottom of the plate with a pipette tip and
150 μL were transferred into tubes. Samples were shaken for 1 min and
stored at −20 °C until analysis.
vided by their manufacturers. Formic acid, HPLC grade methanol and
HPLC grade water were purchased from Merck (Darmstadt, Germany).
Cell culture media, fetal bovine serum, penicillin/streptomycin and
Glutamax® were bought from GIBCO (Lucerne, Switzerland).
Hydrocortison was obtained from Sigma-Aldrich (Buchs, Switzerland).
Stock solutions for the calibration curves were prepared in methanol
2.4. Assays With Human Liver Homogenate
(
4-MAA, 4-AA, 4-FAA, 4-AAA, midazolam and 1′-hydroxymidazolam)
or DMSO (tizanidine, hydroxytizanidine, efavirenz, 8′-hydro-
xyefavirenz, flurbiprofen, 4′-hydroxyflurbiprofen, omeprazole, 5′-hy-
droxyomeprazole, metoprolol, α-hydroxymetoprolol, paclitaxel, 6α-
hydroxypaclitaxel, chlorzoxazone, and 6-hydroxychlorzoxazone). Stock
solutions for the assays were prepared in water (4-MAA), methanol
Part of a human liver biopsy (generously provided in anonymized
form by Prof. Dr. med. L. Terracciano, Department of Pathology,
University Hospital Basel) was cut into small cubes with a razor blade,
weighed and transferred into a manual glass tissue grinder. Ice cold
potassium phosphate buffer (100 mM, pH 7.4) containing NADPH (final
concentration 1 mM) was added (3 times the weight of the liver cubes).
The liver was homogenized on ice, transferred into Eppendorf tubes and
put on ice. The substrates, 4-MAA (50 μM) and midazolam (12 μM),
were dissolved in potassium phosphate buffer (100 mM, pH 7.4). The
reaction was started by mixing 4 parts of liver homogenate (37 °C) and
1 part of the substrate solution (v/v), which was preheated to 37 °C
(Graham, 2002). The final volume of the assay was 1 mL. Samples were
withdrawn after 0, 15, 30, 45, 60, 75, 90, 105, 120 min and subse-
quently precipitated in 300 μL ice cold IS solution in order to terminate
enzymatic reactions. Samples were shaken for 10 min and kept at
(
midazolam) or DMSO (tizanidine, hydroxytizanidine, efavirenz, 8′-
hydroxyefavirenz, flurbiprofen, 4′-hydroxyflurbiprofen, omeprazole, 5′-
hydroxyomeprazole, metoprolol, α-hydroxymetoprolol, paclitaxel, 6α-
hydroxypaclitaxel, chlorzoxazone, and 6-hydroxychlorzoxazone, ri-
fampicin, 3-MC). The deuterated internal standards (IS) were prepared
in methanol (4-MAA-d3: 150 ng/mL; 4-AA-d3: 20 ng/mL; 4-AAA-d3:
5
0 ng/mL; midazolam-d6: 100 ng/mL; metoprolol-d6: 12.5 ng/mL; bi-
soprolol-d5: 2.5 ng/mL; efavirenz-d5: 50 ng/mL; flurbiprofen-d3:
0 ng/mL; omeprazole-d3: 10 ng/mL; tizanidine-d4: 50 ng/mL; pacli-
taxel-d5: 200 ng/mL).
5
−
20 °C until analysis.
2.2. Microsomal Assay
2.5. Assays With Horseradish Peroxidase and Soy Bean Lipoxygenase
Human liver microsomes (HLM; 0.5 mg microsomal protein),
+
NADPH regenerating solutions A and B (A, contains NADP , glucose-6-
4-MAA (50 μM) was preincubated in potassium phosphate buffer
phophate, and MgCl
2
; B, contains glucose-6-phosphate dehydrogenase)
(50 mM, pH 7.4) containing hydrogen peroxide (final concentration 0,
10, 25, and 100 μM for assays with horse radish peroxidase, 1 mM for
assays with soy bean lipoxygenase) for 5 min at 37 °C. Reactions were
started by addition of purified enzyme (2 units/mL horse radish per-
oxidase, 10.8 μg/mL soy bean lipoxygenase). Samples were taken at 0,
5, 10, 15, 30, 60, 90 and 120 min and precipitated with ice cold me-
thanol in a ratio of 1:3 (v/v). Samples were mixed for 1 min, diluted
with IS solution (1:1 v/v), and then stored at −20 °C until analysis.
were obtained from Corning Life science (Woburn, MA, USA). HLM
were preincubated in potassium phosphate buffer (pH 7.4, final con-
centration 100 mM), and NADPH generating system for 5 min at 37 °C.
The assay was started by the addition of 4-MAA (50 μM) or a control
substrate (tizanidine (5 μM) for 1A2, efavirenz (10 μM) for 2B6, pacli-
taxel for (10 μM) for 2C8, flurbiprofen (5 μM) for 2C9, omeprazole
(
17 μM) for 2C19, metoprolol (20 μM) for 2D6, chlorzoxazone (1 μM)
for 2E1, and midazolam (12 μM) for 3A4). The final volume of the assay
was 1 mL. At time points 0, 1 h, 2 h, 4 h and 6 h, samples were taken and
the reaction immediately stopped by adding 300 μL ice cold IS solution
and subsequent vortex mixing.
In the case of inhibition assays, the HLM, the NAPDH regenerating
system, and the respective CYP inhibitor were preincubated in po-
tassium phosphate buffer (pH 7.4, final concentration 100 mM) for
2.6. Assays With Human Myeloperoxidase
Human myeloperoxidase (MPO, Sigma-Aldrich, 0.5 units/mL) was
dissolved in potassium phosphate buffer (50 mM, pH 7.4), 4-MAA was
added (1, 2.5, 5, 10, 25, or 50 μM), and the mixture was preheated to
37 °C. Reactions were started by the addition of 100 μM hydrogen
peroxide (100 μM final concentration). Samples (50 μL) were taken at 0,
2, 4, 8, 12, 16, 20 min. Enzyme kinetics were assessed within the first
4 min of the assay, because the 4-AA formation was linear in this time
period. Reactions were terminated by the addition of ice cold methanol
(150 μL). Samples were diluted (1:1 v/v) with IS solution, vortex mixed,
and stored at −20 °C until analysis.
5
min at 37 °C. Inhibitors were furaphylline for CYP1A2 (10 μM), ti-
clopidine for CYP2B6 (1 μM), montelukast for CYP2C8 (0.5 μM), sul-
faphenazole for CYP2C9 (10 μM), (+)-N-3-benzylnirvanol for CYP2C19
(
(
10 μM), quinidine for CYP2D6 (1 μM), methylpyrazole for CYP2E21
1 μM), and ketoconazole for CYP3A4 (1 μM). For complete CYP in-
hibition, all inhibitors were combined. Assays were started by adding
174

F. Bachmann et al.
European Journal of Pharmaceutical Sciences 120 (2018) 172–180
2
.7. Assay With HL60 Cells
4-AA within 2 h of incubation (Fig. 2, 1a). The time to reach the
maximal 4-AA concentration (1.5 μM) was 1.5 h. Midazolam was ra-
pidly metabolized to 1-hydroxymidazolam; about 25% of MDZ was
converted to 1-hydroxymidazolam and the maximal concentration of
about 2.5 μM was reached after 30 min (Fig. 2, 1b). This proves the
metabolic integrity of the liver homogenate.
Human liver microsomes converted approximately 1% of the 4-MAA
to 4-AA and the metabolic process was not completed within the 6 h
incubation time. A maximal 4-AA concentration of 0.5 μM was reached
at the end of the experiment (Fig. 2, a). In comparison, 60% of the
initial amount of midazolam was converted to 1-hydroxymidazolam
and most of the metabolite was formed within the first 2 h of the in-
cubation (Fig. 2, b).
Human promyelocytic leukemia cells (HL60, CCL-240, lot number
703261, ATCC, Wesel, Germany) were cultured in a 12 well plate
7
(
2,000,000 cells/well) using RPMI medium. Three different sets of the
assay were applied. Set 1 consisted of viable HL60 cells, set 2 was
carried out with viable HL60 cells but in the presence of a selective
MPO inhibitor (PF-06281355, 30 μM), and set 3 was completed with
heat inactivated HL60 cells (94 °C for 20 min). For each assay set, hy-
drogen peroxide was either added at time point zero, at the time points
0
and 60 min, or not at all. 4-MAA was added to the cells, which were
kept at 37 °C, to start the assay. Samples were taken at 0, 5, 10, 15, 30,
0, 65, 70, 75, 90 and 120 min. Reactions were terminated by mixing
6
the samples with ice cold methanol in a ratio of 1:3 (v/v) The samples
were further 1:1 (v/v) diluted with IS solution and kept at −20 °C until
analysis.
HepaRG cells were differentiated to hepatocyte-like cells and their
cytochrome content was induced with either rifampicin (CYP2B6, 2C9,
2C19, 3A4) or 3-MC (CYP 1A2) (Berger et al., 2016). The extent of
induction was tested with midazolam and tizanidine, which are specific
substrates for CYP 3A4 and 1A2, respectively. Rifampicin-induced He-
paRG cells converted approximately 0.15% 4-MAA to 4-AA and the
maximal 4-AA concentration (0.075 μM) was reached after 6 h, whereas
non-induced cells were about 3 times less productive (Fig. 2, 3a). In
contrast, 3-MC-treatment did not increase the 4-AA levels compared to
non-treated cells (Fig. 2, 4a). Formation of 1-hydroxymidazolam and
hydroxytizanidine was clearly induced by rifampicin and 3-MC treat-
ment, respectively, demonstrating the suitability of the model. Forma-
tion of 4-FAA and 4-AAA was not detectable in any of the hepatic
systems used (data not shown).
In the experimental systems used, 4-AA concentrations were highest
in HLH followed by HLM and HepRG cells (Fig. 2). Since the micro-
somal protein content is about 40 mg protein per gram human liver
(Zhang et al., 2015), the incubations with liver homogenate contained
approximately 8 mg microsomal protein per assay (200 mg liver tissue
per assay). In comparison, 0.5 mg microsomal protein was used per
HLM assay. After 2 h of incubation, approximately 280 pmol and
1600 pmol 4-AA was formed in HLM and HLH, respectively (Fig. 2).
Taking into account the microsomal protein content per assay, the 4-AA
formation rate per mg microsomal protein per hour was about 280 pmol
in HLM and 100 pmol in HLH. Hence, the microsomes were roughly
three times more efficient in producing 4-AA than the HLH, suggesting
that in the liver, the conversion of 4-MAA to 4-AA is performed within
the microsomes.
2.8. LC-MS/MS Analysis
Assay samples were thawed, vortex mixed, and centrifuged at 15 °C
and 3220 g for 30 min. The analyte concentrations were determined by
high performance liquid chromatography (HPLC, Shimadzu, Kyoto,
Japan) tandem mass spectrometry (ABSciex, Ontario, Canada). An API
5500 Qtrap was used to analyze tizanidine, hydroxytizanidine, and ti-
zanidine-d4. All other analytes were measured on an API 4000 mass
spectrometer. All compounds were analyzed by electrospray ionization
with multiple reaction monitoring (MRM) in positive polarity mode
except chlorzoxazone, 6-hydroxychlorzoxazone, efavirenz, 8′-hydro-
xyefavirenz, efavirenz-d4, flurbiprofen, 4′-hydroxyflurbiprofen, and
flurbiprofen-d3, which were negatively charged. Nitrogen was used in
the ion source, and as curtain and collision gas. The LC-MS/MS systems
were operated using Analyst software 1.6.2 (AB Sciex, Framingham,
MA, USA).
Three LC methods (A, B, and C) were applied to analyze the com-
pounds (summarized in Table 1). Water (mobile A) and methanol
(
mobile B), both supplemented with 0.1% formic acid, were used as
mobile phases for the chromatography in method A and C whereas
.01% acetic acid in water (mobile A) and methanol (mobile B) were
0
used in method B.
Calibration lines were prepared in assay buffer and consisted of a
blank sample and at least seven calibrators, which were injected twice,
once at the beginning and once at the end of the run. Quality control
samples (medium, high and low) were analyzed between the calibration
lines along with the assay samples to ensure the integrity of the ana-
3.2. Inhibition of 4-AA Formation in Human Liver Microsomes
lysis. Sample concentrations were quantified by linear regression
CYP inhibition studies were carried out in HLM, because the con-
version of 4-MAA to 4-AA could be determined with a robust assay and
microsomes were more readily available than the other systems in-
vestigated. CYP inhibition studies were carried out in order to identify
the responsible enzymes for 4-MAA demethylation. Hence, the meta-
bolism of 4-MAA was studied with and without specific inhibitors of
CYP1A2, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, and 3A4. Moreover, the in-
hibitory effect of all eight inhibitors combined was investigated. In
parallel, the inhibitors were incubated also with the respective CYP-
specific substrate in order to verify their inhibitory activity. All in-
hibitors visually decreased the metabolism of the specific CYP sub-
strates (Fig. 3A). In comparison, formation of 4-AA appeared to be only
slightly inhibited in the presence of inhibitors of CYP2B6, 2C8, 2C9,
2E1 and 3A4, but not by CYP1A2, 2C19 and 2D6. Accordingly, a mix-
ture of all eight inhibitors resulted in a distinct inhibition of 4-MAA
metabolism. The AUC0-6h was calculated with and without inhibitor.
The inhibitors reduced the metabolism between 30% (1A2) and up to
90% (2D6) in the case of specific CYP substrates. The % inhibition of 4-
MAA demethylation to 4-AA was around 15% for 2B6 and 3A4, about
20% for 2C9, 25% for 2C8 and 7% for 2E1 (Fig. 3B). All inhibitors
together decreased the 4-MAA demethylation of by about 65%. These
results show that the demethylation of 4-MAA to 4-AA is not specific for
2
(
weighting 1/x ) of the nominal analyte concentration (x) against the
peak area ratio of the analyte and IS (y).
2.9. Statistics
Results are displayed as mean ± standard deviation of at least 3
independent experiments. Statistical analyses including Michaelis-
Menten kinetics of the 4-AA formation were performed using GraphPad
Prism 6 (GraphPad Software, La Jolla, CA, USA).
3. Results
3.1. 4-MAA Metabolism in Human Hepatic Systems
4-MAA was incubated in the presence of human liver homogenate
(
HLH), human liver microsomes (HLM), and HepaRG cells at a con-
centration of 50 μM and the formation of the metabolites 4-AA, 4-FAA
and 4-AAA was monitored (Fig. 2). Metabolism of midazolam to 1-
hydroxymidazolam was used to assure the functionality of the different
hepatic systems.
The liver homogenate converted approximately 3% of the 4-MAA to
175

F. Bachmann et al.
European Journal of Pharmaceutical Sciences 120 (2018) 172–180
Table 1
LC-MS methods used in this study.
LC method
A
Analyte
MRM (Q1 → Q3) [m/z]
221.2 → 56.1
Collision energy [V]
Polarity mode [+/−]
Calibration range [nM]
4-MAA
31
46
71
46
46
46
61
86
+
+
+
+
+
+
+
+
+
+
−
−
−
−
−
−
−
−
+
+
+
+
+
+
+
+
+
+
+
+
0.1–100 μM
−
0.01–10 μM
−
0.01–10 μM
0.01–10 μM
−
0.01–10 μM
−
0.01–10 μM
0.025–10 μM
0.001–2.5 μM
−
0.005–10 μM
0.005–10 μM
−
0.001–10 μM
0.001–2.5 μM
0.001–5 μM
0.001–10 μM
−
0.0025–10 μM
0.0025–2.5 μM
−
4
4
4
4
4
4
-MAA-d3
-AA
-AA-d3
-FAA
-AAA
204.2 → 56.2
207.2 → 56.2
232.1 → 204.1
246.2 → 83.1
249.2 → 84.1
326.2 → 291.3
331.6 → 296.0
342.2 → 324.0
313.8 → 246.6
329.9 → 286.1, 257.8
317.9 → 68.9
243.0 → 199.0
259.1 → 214.9
246.0 → 202.0
167.709 → 131.8
183.782 → 119.8
346.2 → 198.2
362.2 → 214.1
349.2 → 198.1
268.2 → 116.2
284.2 → 116.2, 98.2
275.307 → 123.3
854.287 → 105.0, 569.2
870.203 → 104.9, 286.1
859.297 → 291.2, 569.2
255.14 → 44.1
270.0 → 60.1
-AAA-d3
Midazolam
Midazolam-d6
Hydroxymidazolam
Efavirenz
41
61
B
−105
−105
−70
−45
−50
−50
−30
−28
46
41
41
11
66
8′-hydroxyefavirenz
Efavirenz-d4
Flurbiprofen
4′-hydroxyflurbiprofen
Flurbiprofen-d3
Chlorzoxazone
6
-hydroxychlorzoxazone
Omeprazole
′-hydroxyomeprazole
5
Omeprazole-d3
Metoprolol
α-hydroxymetoprolol
Metoprolol-d7
Paclitaxel
31
107, 15
93, 23
27, 15
81
81
136
0.05–10 μM
0.05–10 μM
−
0.01–10 μM
0.001–5 μM
6
α-hydroxypaclitaxel
Paclitaxel-d5
Tizanidine
C
Hydroxytizanidine
Tizanidine-d4
259.1 → 48.2
LC method A: Analytical column: Atlantis® T3, 2.1 × 50 mm, 3 μm; Flow rate: 0.85 mL/min; Temperature: 45° C.
Gradient program: 0.5 min, 2% B; 0.51 min, 30% B; 2 min, 50% B; 2.5 min, 98% B; 3 min, 98% B; 3.01 min, 2% B; 3.5 min, 2% B.
LC method B: Analytical column: Atlantis® T3, 2.1 × 50 mm, 3 μm; Flow rate: 0.8 mL/min; Temperature: 45 °C.
Gradient program: 0–0.5 min, 0% B; 1.5 min, 95% B; 2.5 min, 95% B; 2.51 min, 0% B; 3 min, 0% B.
A tee union was used to inline dilute the sample with mobile phase A during the first 0.5 min of each run.
LC method C: Analytical column: Luna PFP column, 50 × 2 mm, 3 μm; Flow rate: 0.8 mL/min; Temperature: 45 °C.
Gradient program: 0–0.2 min, 5% B; 0.2 min, 50% B; 2 min, 95% B; 2.7 min, 95% B; 2.71 min, 5% B; 3 min, 5% B.
Fig. 2. In vitro metabolism of 4-MAA to 4-AA (-●-) mediated by different hepatic systems (upper row), namely liver homogenate (1), microsomes (2), and HepRG
cells (3 and 4). HepRG were either not treated (-●-), or induced with rifampicin (-○-, 3) or 3-MC (-○-, 4). Metabolic activity of the hepatic systems was confirmed by
the formation of hydroxy-midazolam (-■-, 1b-3b) and hydroxy-tizanidine (-■-, 4b) from midazolam or tizanidine, respectively (lower row). Hydroxy-midazolam (3b,
-
□-) and hydroxy-tizanidine (4b, -□-) formation was additionally monitored in rifampicin (3b) and 3-MC (4b) induced HepRG cells. Error bars correspond to the
standard deviation (n = 3). Initial concentrations were 50 μM for 4-MAA, 12 μM for midazolam and 5 μM for tizanidine.
176

F. Bachmann et al.
European Journal of Pharmaceutical Sciences 120 (2018) 172–180
Fig. 3. A. Formation of 4-AA (-●-) and specific cytochrome substrates (-■-, 1A2: tizanidine, 2B6: efavirenz, 2C9:flurbiprofen, 2C19: omeprazole, 2D6: metoprolol,
A4: midazolam, mixture: midazolam) in human liver microsomes. Specific cytochrome isoform inhibitors (1A2: furafylline, 2B6: ticlopidine, 2C8: montelukast, 2C9:
3
sulfaphenazole, 2C19: (+)-N-3-benzylnirvanol, 2D6: quinidine, 2E1: methylpyrazole, 3A4: ketoconazole) and a mixture of all eight inhibitors were used to inhibit the
formation of 4-AA (-○-) and of the known cytochrome substrates (-□-). B. Inhibition of 4-AA formation (white bars) and metabolite formation of specific cytochrome
AUCwith inhibitor
without inhibitor
substrates (black bars). The % inhibition was calculated as 100 −
× 100 . The area under the concentration-time curve (AUC) from time point zero to
(
AUC
)
6
h was determined using the linear trapezoidal rule. The error bars correspond to the standard deviation (n = 3).
one CYP.
min/mg protein (Fig. 4B).
Overall, none of the investigated peroxidases was capable of pro-
ducing 4-FAA or 4-AAA (data not shown).
3.3. 4-MAA Demethylation Catalyzed by Peroxidases
N-demethylation of xenobiotics can also be performed by other
enzymes than cytochrome P450 oxidases, e.g. by peroxidases
Benedetti, 2001; Benedetti et al., 2009). We therefore incubated 4-
3.4. 4-MAA Demethylation by HL-60 Cells
(
In a next step, we investigated whether 4-MAA demethylation is
possible by HL-60 cells, which are known to express MPO (Maseneni
et al., 2012). 4-MAA was incubated with viable HL60 cells, with viable
HL60 cells plus a selective MPO inhibitor, and with heat inactivated
HL60 cells (Fig. 5). For each setup hydrogen peroxide was either not
added, or added once (T = 0 h) or twice (T = 0 h and 1 h). Without the
addition of hydrogen peroxide, almost no 4-AA was formed, whereby
viable cells produced more 4-AA than MPO-inhibited or heat in-
activated cells. HL60 cells converted approximately 5% of 4-MAA to 4-
AA when hydrogen peroxide was added at the beginning of the ex-
periment. A second addition of hydrogen peroxide after 1 h incubation
almost doubled the conversion of 4-MAA to 4-AA within 15 min. HL60
cells treated with a MPO inhibitor formed less than half of 4-AA com-
pared with uninhibited cells (2% vs. 5%). Heat inactivation almost
nullified the 4-AA formation so that 95% less 4-AA was produced
compared to viable and untreated HL60 cells. 4-FAA and 4-AAA were
not detected following incubation of 4-MAA in HL60 cells (data not
shown).
MAA in the presence of HRP and different concentrations of hydrogen
peroxide (0, 10, 25, and 100 μM), which acts as a cofactor of the en-
zyme. The decrease of 4-MAA depended visibly on the hydrogen per-
oxide concentration (Fig. 4A). 100 μM hydrogen peroxide led to com-
plete degradation of 50 μM 4-MAA within 10 min, while 25, 10 and
0
μM produced a drop from 50 μM to 10, 30, and 45 μM, respectively.
The 4-AA formation was also connected to the hydrogen peroxide
concentration. Maximal 4-AA concentrations of about 1 μM, 6 μM and
10 μM were observed when 0, 10 or 25 μM hydrogen peroxide was
added to the reaction mixture. Interestingly, 4-AA reached a maximal
concentration of approximately 12 μM within 5 min of incubation when
using 100 μM hydrogen peroxide and then decreased and had com-
pletely disappeared after 30 min. Hence, 4-AA is presumably an inter-
mediate metabolite considering also that the mass balance between 4-
MAA and 4-AA did not come out even. Importantly, hydrogen peroxide
alone at a concentration of 100 μM formed only negligible amounts of
4
-AA (~0.1 μM), showing that this reaction depends on the presence of
HRP.
4-MAA was additionally incubated in the presence of soybean li-
4. Discussion
poxygenase and 100 μM hydrogen peroxide. In contrast to HRP, under
these conditions, the 4-AA concentration increased steadily over
Metamizole is a commonly used analgesic drug; however, it has
been withdrawn from market in several countries because of the risk of
agranulocytosis (Hedenmalm and Spigset, 2002). Although the com-
pound is marketed since almost 100 years, neither the mode of action,
nor the mechanism of toxicity is entirely elucidated. Moreover, the
enzymes involved in the metabolism of metamizole are mostly un-
known. Considering this lack of knowledge, more data are needed to
ensure a safe clinical use of metamizole. Here, we investigated the N-
demethylation process of 4-MAA in different metabolic systems in order
to increase our knowledge about metamizole metabolism.
120 min and the 4-MAA concentration decreased in a similar manner as
described above for HRP (data not shown).
Finally, the demethylation capability of a human peroxidase,
namely myeloperoxidase, was tested for 4-MAA. Again, 100 μM hy-
drogen peroxide was used as a cofactor. Approximately 2% of 4-MAA
was converted to 4-AA within 20 min of incubation, which did not lead
to a measurable decrease of the 4-MAA concentration (50 μM). The
increase of 4-AA was linear for the first 4 min but leveled off afterwards.
Thus, Michaelis-Menten kinetics was determined based on the initial
reaction rate (0–4 min), resulting in a K
of 0.140 ± 0.009 nmol/min/0.5 U (equal to 10 μg protein) or 14 nmol/
m
of 22.5 ± 3.7 μM and a Vmax
First, 4-MAA demethylation was investigated in different hepatic
models such as HLH, HLM, and HepaRG cells, since it has been
177

F. Bachmann et al.
European Journal of Pharmaceutical Sciences 120 (2018) 172–180
Fig. 4. A. Demethylation of 4-MAA (left graph) to 4-AA (right graph) catalyzed by horse radish peroxidase. Hydrogen peroxide at a concentration of 0 μM (−○-),
1
0 μM (-◒-), 25 μM (-◓-) and 100 μM (-●-) was used as co-factor. In addition, MAA was incubated with 100 μM hydrogen peroxide but without horse radish
peroxidase (-■-). The error bars correspond to the standard deviation (n = 5). B. Demethylation of 4-MAA (-○-) to 4-AA (-●-) in the presence of human myelo-
peroxidase and 100 μM hydrogen peroxide (left graph, n = 5). Michaelis-Menten kinetics of the 4-AA formation catalyzed by human myeloperoxidase (right graph,
n = 4). The error bars correspond to the standard deviation.
Fig. 5. A. Formation of 4-AA in the presence of HL60 cells. Hydrogen peroxide was either added at time point zero (-●-), or at the time points 0 and 60 min (-●-), or
not at all (-○-). 4-AA formation in the presence of viable HL60 cells (left panel), of viable HL60 cells plus a selective MPO inhibitor (PF-06281355, 30 μM, middle
panel), and of heat inactivated HL60 cells was investigated. B. AUC0-2 h determined for the different conditions. Coloring of the bars agrees with the concentration
time profiles. The error bars correspond to the standard deviation.
178

F. Bachmann et al.
European Journal of Pharmaceutical Sciences 120 (2018) 172–180
demonstrated that 4-MAA can be demethylated by human liver mi-
crosomes (Geisslinger et al., 1996). All investigated hepatic models
investigated were functional and able to form 4-AA. In relation to the
estimated content of microsomes, HLM was more efficient in the de-
methylation of 4-MAA than HLH. In HLH, the metabolic activity may
have been compromised by the thawing and homogenization process
and/or the demethylation may have been impaired due to protein
binding of 4-MAA (Zylber-Katz et al., 1985). In support of these as-
sumptions, a reduced activity was not only present for 4-MAA de-
methylation but also for midazolam hydroxylation, which was about 30
times lower in HLH compared to HLM when calculated per mg micro-
somal protein. On the other hand, the data suggested that the hepatic
demethylation of 4-MAA can largely be explained by microsomal me-
tabolism.
granulocytes and granulocyte precursors. 4-MAA demethylation by
MPO could be described by Michaelis-Menten kinetics with a K
m
of
22.5 μM and a maximal reaction velocity of 14 nmol/min/mg protein.
The 4-MAA plasma concentrations following a single oral dose of 1 g
reached serum concentrations in the range of the K (Levy Micha and
m
Bernd, 1995), suggesting that MPO may play a role in the demethyla-
tion of 3-MAA in vivo. To further investigate this hypothesis, we roughly
calculated the amount of MPO in human bone marrow in order to es-
timate the demethylation capacity by MPO. Assuming that the bone
marrow corresponds to about 4% of the body weight and that about
50% of the bone marrow consists of hematopoietic cells, a person of
70 kg body weight has approximately 1.4 kg hematopoietic bone
marrow. Of that, MPO expressing granulocytes and granulocyte pre-
cursors (promyelocytes, myelocytes, metamyelocytes and granulocytes)
represent approximately 50% or 700 g (Donohue et al., 1958). Sup-
posing an MPO protein content of 17 mg/g granulocytes (Schultz and
Kaminker, 1962), the total amount in bone marrow would be about
12 g MPO protein. Taking into account the in vitro activity of 14 nmol/
min/mg MPO protein, this results in a total activity of approximately
170 μmol/min or 10 mmol/h. Compared to the liver, the 4-MAA de-
methylation capacity of granulocytes and granulocyte precursors is
approximately 600 times higher, underscoring the important role of
extrahepatic demethylation of 4-MAA.
Rifampicin, which induces CYP3A4, 2B6, 2C9 and 2C19 (Berger
et al., 2016; Derungs et al., 2016), led to an approximately 4-fold higher
4-AA formation in treated compared to untreated HepaRG cells, which
supports a CYP-dependent metabolism of 4-MAA (Berger et al., 2016).
In contrast, up-regulation of CYP1A2 by 3-MC did not affect the 4-AA
formation. CYP inhibition assays in HLM confirmed these results, since
selective inhibition of CYP 1A2, 2C19, and 2D6 did not reduce 4-AA
formation. In contrast, co-incubation with inhibitors for CYP3A4, 2B6,
2C8, 2C9 or 2E1 was associated with a small reduction of 4-AA for-
mation. Inhibition of 4-MAA demethylation with the CYP3A4 inhibitor
ketoconazole is in line with data described in literature (Geisslinger
et al., 1996). Overall, inhibition of 4-MAA metabolism was generally
low compared to the inhibition of the respective model compounds
used; only a mixture of all eight CYP inhibitors was capable of in-
hibiting the demethylation rate by > 50%. Hence, the data indicate
that hepatic 4-MAA demethylation is not performed by a specific CYP,
but that several CYPs are contributing.
Based on the data of the current study, we can estimate the hepatic
capacity for 4-MAA demethylation and compare it with the in vivo
metabolism of 4-MAA. Using HLM, we observed a 4-MAA demethyla-
tion rate of 280 pmol/mg protein/h. A human liver weighs approxi-
mately 1500 g and 1 g liver contains approximately 40 mg microsomal
protein (Zhang et al., 2015). Hence, assuming that 4-MAA is mainly
metabolized in hepatic microsomes by CYPs, the in vivo capacity of 1 g
liver is 40 × 280 pmol/h or 11.2 nmol/h and the hepatic capacity
We also investigated 3-MAA demethylation in a more complex
model than isolated MPO, namely in HL-60 cells, a granulocyte pre-
cursor cell line containing myeloperoxidase (Maseneni et al., 2012).
Our experiments demonstrate that HL-60 cells are able to metabolize 4-
MAA in the presence of hydrogen peroxide. The addition of a selective
MPO inhibitor reduced 4-MAA demethylation by HL-60 cells sig-
nificantly, proving the involvement of MPO in 4-MAA demethylation.
To the best of our knowledge, it has so far neither been shown that
human MPO is capable of N-demethylation of small molecules, nor that
it has a role in the metabolism of xenobiotics. Usually, plasma and
blood concentrations of hydrogen peroxide are in the low μM range
(Forman et al., 2016), but in granulocytes, hydrogen peroxide con-
centrations are higher and can rise up to 100 μM (Droge, 2002), sug-
gesting that sufficiently high concentrations for 4-MAA demethylation
can be reached.
Typically, peroxidase-derived oxidations are connected to radical
formation (Campomanes et al., 2015; Hiner et al., 2002; Perez-Gilabert
et al., 1997). It has for instance been shown that degradation of clo-
zapine and amodiaquine by myeloperoxidase can lead to covalent
binding of reactive metabolites to proteins in neutrophils, possibly ex-
plaining why these drugs are associated with agranulocytosis (Lobach
and Uetrecht, 2014). Assuming that N-demethylation of 4-MAA may
also take place in granulocyte precursors and granulocytes in the bone
marrow, radical formation by this reaction could explain myelotoxicity
of metamizole in susceptible persons. In support of this assumption,
granulopoiesis has been described to be impaired on the promyelocyte
stage in patients with metamizole-associated myoletoxicity (Kummer
et al., 2006), the stage where myeloperoxidase has been shown to be
present (Tsuruta et al., 1999).
(
1500 g) equals approximately 16.8 μmol/h. Following a single oral
dose of 1 g metamizole (molecular weight: 311 g/mol) and assuming a
bioavailability of 80%, approximately 2.6 mmol 4-MAA reaches the
systemic circulation. The half-life of 4-MAA is about 2.7 h (Berger et al.,
2016; Levy Micha, 1995), suggesting that at least 1.3 mmol 4-MAA can
be eliminated during 2.7 h, corresponding to approximately 500 μmol/
h. Since almost all 4-MAA is eliminated by demethylation, this number
can be compared with the hepatic microsomal capacity of 16.8 μmol/h.
This comparison suggests a large extrahepatic demethylation capacity
for 4-MAA.
It has been shown that HRP is able to N-demethylate xenobiotic
compounds and that soybean lipoxygenase is capable to demethylate
aminopyrine, which differs only by a single methyl group from 4-MAA
(
1
Perez-Gilabert et al., 1997; Stiborova et al., 1997; Yang and Kulkarni,
998). Hence, we incubated 4-MAA in the presence of HRP or soybean
In conclusion, the cytochrome P450 inhibition assays in HLM re-
vealed that 4-MAA demethylation is not catalyzed by a specific CYP,
but mainly by CYP3A4, 2B6, 2C8 and 2C9. Translation of our in vitro
data into in vivo conditions suggest a dominant extra-hepatic metabo-
lism of 4-MAA. In the presence of hydrogen peroxide, isolated human
lipoxygenase, which both were capable to N-demethylate 4-MAA and to
form 4-AA. Interestingly, in the presence of HRP, the 4-MAA con-
centration dropped to a higher extent than 4-AA was produced, sug-
gesting the formation of metabolites other than 4-AA that we did not
detect with our analytical method. It has been shown that, in the pre-
sence of HRP and hydrogen peroxide, the pyrazole ring of 4-MAA can
be opened to form oxalic acid hydrazides, which could explain the
mismatch of 4-MAA degradation and 4-AA formation (Wessel et al.,
m
MPO was able to demethylate 4-MAA and the K of the reaction was in
the same range as plasma concentrations of 4-MAA after therapeutic
doses of metamizole. The demethylation capacity of granulocytes and
granulocyte precursors was found to be considerably larger than the
hepatic capacity and sufficiently large to account for the entire de-
methylation of 4-MAA to 4-AA in humans.
2
006).
Based on these findings, we speculated that peroxidases could play a
role in the metabolism of 4-MAA. Therefore, we tested human myelo-
peroxidase, which has high activity in mature neutrophilic
a
179

F. Bachmann et al.
European Journal of Pharmaceutical Sciences 120 (2018) 172–180
Conflict of interest
demethylation in liver microsomes by high-performance liquid chromatography. J.
Chromatogr. 274, 201–208.
Kummer, O., Haschke, M., Tuchscherer, D., Lampert, M., Martius, F., Krahenbuhl, S.,
None of the authors has a conflict of interest regarding this study.
Financial support
2
006. Agranulocytosis in a patient treated with metamizole and clopidogrel. In:
Praxis (Bern 1994). 95. pp. 1743–1745 (quiz 1746-1747).
Levy Micha, Z.-K.E., Bernd, Rosenkranz, 1995. Clinical pharmacokinetics of dipyrone and
its metabolites. Clin. Pharmacokinet. 28, 216–234.
Levy, M., Flusser, D., Zylber-Katz, E., Granit, L., 1984. Plasma kinetics of dipyrone me-
tabolites in rapid and slow acetylators. Eur. J. Clin. Pharmacol. 27, 453–458.
Lobach, A.R., Uetrecht, J., 2014. Involvement of myeloperoxidase and NADPH oxidase in
the covalent binding of amodiaquine and clozapine to neutrophils: implications for
drug-induced agranulocytosis. Chem. Res. Toxicol. 27, 699–709.
Maione, S., Radanova, L., De Gregorio, D., Luongo, L., De Petrocellis, L., Di Marzo, V.,
Imming, P., 2015. Effects of metabolites of the analgesic agent dipyrone (metamizol)
on rostral ventromedial medulla cell activity in mice. Eur. J. Pharmacol. 748,
The study was supported by a grant from the Swiss National Science
foundation to SK (SNF 31003A_156270).
References
Agundez, J.A., Martinez, C., Benitez, J., 1995. Metabolism of aminopyrine and derivatives
in man: in vivo study of monomorphic and polymorphic metabolic pathways.
Xenobiotica 25, 417–427.
Alves, D., Duarte, I., 2002. Involvement of ATP-sensitive K(+) channels in the peripheral
antinociceptive effect induced by dipyrone. Eur. J. Pharmacol. 444, 47–52.
Aninat, C., Piton, A., Glaise, D., Le Charpentier, T., Langouet, S., Morel, F., Guguen-
Guillouzo, C., Guillouzo, A., 2006. Expression of cytochromes P450, conjugating
enzymes and nuclear receptors in human hepatoma HepaRG cells. Drug Metab.
Dispos. 34, 75–83.
Bast, A., Noordhoek, J., 1981. Product inhibition during the hepatic microsomal N-de-
methylation of aminopyrine in the rat. Biochem. Pharmacol. 30, 19–24.
Beirith, A., Santos, A.R., Rodrigues, A.L., Creczynski-Pasa, T.B., Calixto, J.B., 1998. Spinal
and supraspinal antinociceptive action of dipyrone in formalin, capsaicin and glu-
tamate tests. Study of the mechanism of action. Eur. J. Pharmacol. 345, 233–245.
Benedetti, M.S., 2001. Biotransformation of xenobiotics by amine oxidases. Fundam. Clin.
Pharmacol. 15, 75–84.
Benedetti, M.S., Whomsley, R., Poggesi, I., Cawello, W., Mathy, F.X., Delporte, M.L.,
Papeleu, P., Watelet, J.B., 2009. Drug metabolism and pharmacokinetics. Drug
Metab. Rev. 41, 344–390.
1
15–122.
Maseneni, S., Donzelli, M., Taegtmeyer, A.B., Brecht, K., Krahenbuhl, S., 2012. Toxicity of
clopidogrel and ticlopidine on human myeloid progenitor cells: importance of me-
tabolites. Toxicology 299, 139–145.
Noda, A., Goromaru, T., Tsubone, N., Matsuyama, K., Iguchi, S., 1976a. In vivo formation
of 4-formylaminoantipyrine as a new metabolite of aminopyrine. I. Chem. Pharm.
Bull. (Tokyo) 24, 1502–1505.
Noda, A., Tsubone, N., Mihara, M., Goromaru, T., Iguchi, S., 1976b. Formation of 4-
formylaminoantipyrine as a new metabolite of aminopyrine. II. Enzymatic de-
methylation and oxidation of aminopyrine and 4-monomethylaminoantipyrine.
Chem. Pharm. Bull. (Tokyo) 24, 3229–3231.
Ortiz, M.I., Castaneda-Hernandez, G., Granados-Soto, V., 2003. Possible involvement of
potassium channels in peripheral antinociception induced by metamizol: lack of
participation of ATP-sensitive K+ channels. Pharmacol. Biochem. Behav. 74,
4
65–470.
Perez-Gilabert, M., Sanchez-Ferrer, A., Garcia-Carmona, F., 1997. Oxidation of amino-
pyrine by the hydroperoxidase activity of lipoxygenase: a new proposed mechanism
of N-demethylation. Free Radic. Biol. Med. 23, 548–555.
Pierre, S.C., Schmidt, R., Brenneis, C., Michaelis, M., Geisslinger, G., Scholich, K., 2007.
Inhibition of cyclooxygenases by dipyrone. Br. J. Pharmacol. 151, 494–503.
Schultz, J., Kaminker, K., 1962. Myeloperoxidase of the leucocyte of normal human
blood. I. Content and localization. Arch. Biochem. Biophys. 96, 465–467.
Siebel, J.S., Beirith, A., Calixto, J.B., 2004. Evidence for the involvement of metabotropic
glutamatergic, neurokinin 1 receptor pathways and protein kinase C in the anti-
nociceptive effect of dipyrone in mice. Brain Res. 1003, 61–67.
Silva, L.C.R., Miranda e Castor, M.G., Souza, T.C., Duarte, I.D.G., Romero, T.R.L., 2015.
NSAIDs induce peripheral antinociception by interaction with the adrenergic system.
Life Sci. 130, 7–11.
Stiborova, M., Hansikova, H., Schmeiser, H.H., Frei, E., 1997. An investigation of the
metabolism of N-nitrosodimethylamine and N-nitrosomethylaniline by horseradish
peroxidase in vitro. Gen. Physiol. Biophys. 16, 285–297.
Berger, B., Donzelli, M., Maseneni, S., Boess, F., Roth, A., Krahenbuhl, S., Haschke, M.,
2
016. Comparison of liver cell models using the Basel phenotyping cocktail. Front.
Pharmacol. 7, 443.
Campomanes, P., Rothlisberger, U., Alfonso-Prieto, M., Rovira, C., 2015. The molecular
mechanism of the catalase-like activity in horseradish peroxidase. J. Am. Chem. Soc.
1
37, 11170–11178.
Derungs, A., Donzelli, M., Berger, B., Noppen, C., Krahenbuhl, S., Haschke, M., 2016.
Effects of cytochrome P450 inhibition and induction on the phenotyping metrics of
the Basel cocktail: a randomized crossover study. Clin. Pharmacokinet. 55, 79–91.
Donohue, D.M., Reiff, R.H., Hanson, M.L., Betson, Y., 1958. Quantitative measurement of
the erythrocytic and granulocytic cells of the marrow and blood. J. Clin. Invest. 37,
1
571–1576.
Droge, W., 2002. Free radicals in the physiological control of cell function. Physiol. Rev.
2, 47–95.
Tatsuo, M.A., Carvalho, W.M., Silva, C.V., Miranda, A.E., Ferreira, S.H., Francischi, J.N.,
8
1
994. Analgesic and anti-inflammatory effects of dipyrone in rat adjuvant arthritis
Duarte, I.D., dos Santos, I.R., Lorenzetti, B.B., Ferreira, S.H., 1992. Analgesia by direct
antagonism of nociceptor sensitization involves the arginine-nitric oxide-cGMP
pathway. Eur. J. Pharmacol. 217, 225–227.
Eagling, V.A., Tjia, J.F., Back, D.J., 1998. Differential selectivity of cytochrome P450
inhibitors against probe substrates in human and rat liver microsomes. Br. J. Clin.
Pharmacol. 45, 107–114.
model. Inflammation 18, 399–405.
Tortorici, V., Vasquez, E., Vanegas, H., 1996. Naloxone partial reversal of the anti-
nociception produced by dipyrone microinjected into the periaqueductal gray of rats.
Possible involvement of medullary off- and on-cells. Brain Res. 725, 106–110.
Tsuruta, T., Tani, K., Hoshika, A., Asano, S., 1999. Myeloperoxidase gene expression and
regulation by myeloid cell growth factors in normal and leukemic cells. Leuk.
Lymphoma 32, 257–267.
Valenzuela, F., García-Saisó, S., Lemini, C., Ramírez-Solares, R., Vidrio, H., Mendoza-
Fernández, V., 2005. Metamizol acts as an ATP sensitive potassium channel opener to
inhibit the contracting response induced by angiotensin II but not to norepinephrine
in rat thoracic aorta smooth muscle. Vasc. Pharmacol. 43, 120–127.
Vanegas, H., Tortorici, V., Eblen-Zajjur, A., Vasquez, E., 1997. PAG-microinjected di-
pyrone (metamizol) inhibits responses of spinal dorsal horn neurons to natural
noxious stimulation in rats. Brain Res. 759, 171–174.
Vasquez, E., Vanegas, H., 2000. The antinociceptive effect of PAG-microinjected dipyrone
in rats is mediated by endogenous opioids of the rostral ventromedical medulla. Brain
Res. 854, 249–252.
Volz, M., Kellner, H.M., 1980. Kinetics and metabolism of pyrazolones (propyphenazone,
aminopyrine and dipyrone). Br. J. Clin. Pharmacol. 10 (Suppl. 2), 299S–308S.
Wessel, J.C., Matyja, M., Neugebauer, M., Kiefer, H., Daldrup, T., Tarbah, F.A., Weber, H.,
Ergün, H., Frattarelli, D.A.C., Aranda, J.V., 2004. Characterization of the role of physi-
cochemical factors on the hydrolysis of dipyrone. J. Pharm. Biomed. Anal. 35,
4
79–487.
Escobar, W., Ramirez, K., Avila, C., Limongi, R., Vanegas, H., Vazquez, E., 2012.
Metamizol, a non-opioid analgesic, acts via endocannabinoids in the PAG-RVM axis
during inflammation in rats. Eur. J. Pain 16, 676–689.
Flusser, D., Zylber-Katz, E., Granit, L., Levy, M., 1988. Influence of food on the phar-
macokinetics of dipyrone. Eur. J. Clin. Pharmacol. 34, 105–107.
Forman, H.J., Bernardo, A., Davies, K.J., 2016. What is the concentration of hydrogen
peroxide in blood and plasma? Arch. Biochem. Biophys. 603, 48–53.
Geisslinger, G., Bocker, R., Levy, M., 1996. High-performance liquid chromatographic
analysis of dipyrone metabolites to study their formation in human liver microsomes.
Pharm. Res. 13, 1272–1275.
Graham, J.M., 2002. Homogenization of mammalian tissue. Sci. World J. 2, 1626–1629.
Gulmez, S.E., Gurdal, H., Tulunay, F.C., 2006. Airway smooth muscle relaxations induced
by dipyrone. Pharmacology 78, 202–208.
2
006. Characterization of oxalic acid derivatives as new metabolites of metamizol
(dipyrone) in incubated hen's egg and human. Eur. J. Pharm. Sci. 28, 15–25.
Hedenmalm, K., Spigset, O., 2002. Agranulocytosis and other blood dyscrasias associated
with dipyrone (metamizole). Eur. J. Clin. Pharmacol. 58, 265–274.
Hiner, A.N.P., Raven, E.L., Thorneley, R.N.F., Garcia-Canovas, F., Rodriguez-Lopez, J.N.,
Yang, X., Kulkarni, A.P., 1998. N-dealkylation of aminopyrine catalyzed by soybean li-
poxygenase in the presence of hydrogen peroxide. J. Biochem. Mol. Toxicol. 12,
1
75–183.
2
002. Mechanisms of compound I formation in heme peroxidases. J. Inorg. Biochem.
Zhang, H., Gao, N., Tian, X., Liu, T., Fang, Y., Zhou, J., Wen, Q., Xu, B., Qi, B., Gao, J., Li,
H., Jia, L., Qiao, H., 2015. Content and activity of human liver microsomal protein
and prediction of individual hepatic clearance in vivo. Sci. Rep. 5, 17671.
Zylber-Katz, E., Granit, L., Levy, M., 1985. Plasma protein binding of dipyrone metabo-
lites in man. Eur. J. Clin. Pharmacol. 29, 67–71.
9
1, 27–34.
Imaoka, S., Inoue, K., Funae, Y., 1988. Aminopyrine metabolism by multiple forms of
cytochrome P-450 from rat liver microsomes: simultaneous quantitation of four
aminopyrine metabolites by high-performance liquid chromatography. Arch.
Biochem. Biophys. 265, 159–170.
Inoue, K., Fujimori, K., Mizokami, K., Sunouchi, M., Takanaka, A., Omori, Y., 1983.
Simultaneous determination of aminopyrine hydroxylation and aminopyrine N-
Zylber-Katz, E., Granit, L., Levy, M., 1992. Formation and excretion of dipyrone meta-
bolites in man. Eur. J. Clin. Pharmacol. 42, 187–191.
180