Multiple cytochrome P450 isoforms are involved in the generation of a pharmacologically active thiol metabolite, whereas paraoxonase 1 and carboxylesterase 1 catalyze the formation of a thiol metabolite isomer from ticlopidine

Article abstract of DOI:10.1124/dmd.113.053017

Ticlopidine is a first-generation thienopyridine antiplatelet drug that prevents adenosine 5'-diphosphate (ADP)-induced platelet aggregation. We identified the enzymes responsible for the two-step metabolic bioactivation of ticlopidine in human liver microsomes and plasma. Formation of 2-oxo-ticlopidine, an intermediate metabolite, was NADPH dependent and cytochrome P450 (CYP) 1A2, 2B6, 2C19, and 2D6 were involved in this reaction. Conversion of 2-oxo-ticlopidine to thiol metabolites was observed in both microsomes (M1 and M2) and plasma (M1). These two metabolites were considered as isomers, and mass spectral analysis suggested that M2 was a thiol metabolite bearing an exocyclic double bond, whereas M1 was an isomer in which the double bond was migrated to an endocyclic position in the piperidine ring. The conversion of 2-oxo-ticlopidine to M1 in plasma was significantly increased by the addition of 1 mM CaCl₂. In contrast, the activity in microsomes was not changed in the presence of CaCl₂. M1 formation in plasma was inhibited by EDTA but not by other esterase inhibitors, whereas this activity in microsomes was substantially inhibited by carboxylesterase (CES) inhibitors such as bis-(p-nitrophenyl)phosphate (BNPP), diisopropylphosphorofluoride (DFP), and clopidogrel. The conversion of 2-oxo-ticlopidine to M1 was further confirmed with recombinant paraoxonase 1 (PON1) and CES1. However, M2 was detected only in NADPHdependent microsomal incubation, and multiple CYP isoforms were involved in M2 formation with highest contribution of CYP2B6. In vitro platelet aggregation assay demonstrated that M2 was pharmacologically active. These results collectively indicated that the formation of M2 was mediated by CYP isoforms whereas M1, an isomer of M2, was generated either by human PON1 in plasma or by CES1 in the human liver. Copyright

Full text of DOI:10.1124/dmd.113.053017

Supplemental Material can be found at:
http://dmd.aspetjournals.org/content/suppl/2013/10/29/dmd.113.053017.DC1.html
1521-009X/42/1/141–152$25.00
http://dx.doi.org/10.1124/dmd.113.053017
DRUG METABOLISM AND DISPOSITION
Drug Metab Dispos 42:141–152, January 2014
Copyright ª 2013 by The American Society for Pharmacology and Experimental Therapeutics
Multiple Cytochrome P450 Isoforms Are Involved in the Generation
of a Pharmacologically Active Thiol Metabolite, whereas
Paraoxonase 1 and Carboxylesterase 1 Catalyze the Formation
of a Thiol Metabolite Isomer from Ticlopidines
Min-Jung Kim, Eun Sook Jeong, Jung-Soon Park, Su-Jun Lee, Jong Lyul Ghim, Chang-Soo Choi,
Jae-Gook Shin, and Dong-Hyun Kim
Department of Pharmacology and PharmacoGenomics Research Center, College of Medicine (M.-J.K., E.S.J, J.-S.P., S.-J.L., J.L.G,
J.-G.S., D.-H.K.), and Department of General Surgery, Busan Paik Hospital (C.-S.C.), Inje University, Busan, South Korea
Received May 22, 2013; accepted October 29, 2013
ABSTRACT
Ticlopidine is a first-generation thienopyridine antiplatelet drug that microsomes was not changed in the presence of CaCl2. M1
prevents adenosine 59-diphosphate (ADP)-induced platelet aggre-
gation. We identified the enzymes responsible for the two-step
metabolic bioactivation of ticlopidine in human liver microsomes
and plasma. Formation of 2-oxo-ticlopidine, an intermediate meta-
bolite, was NADPH dependent and cytochrome P450 (CYP) 1A2,
2B6, 2C19, and 2D6 were involved in this reaction. Conversion
of 2-oxo-ticlopidine to thiol metabolites was observed in both
microsomes (M1 and M2) and plasma (M1). These two metabolites
were considered as isomers, and mass spectral analysis sug-
formation in plasma was inhibited by EDTA but not by other
esterase inhibitors, whereas this activity in microsomes was
substantially inhibited by carboxylesterase (CES) inhibitors such
as bis-(p-nitrophenyl)phosphate (BNPP), diisopropylphosphorofluor-
ide (DFP), and clopidogrel. The conversion of 2-oxo-ticlopidine
to M1 was further confirmed with recombinant paraoxonase 1
(PON1) and CES1. However, M2 was detected only in NADPH-
dependent microsomal incubation, and multiple CYP isoforms
were involved in M2 formation with highest contribution of
gested that M2 was a thiol metabolite bearing an exocyclic double CYP2B6. In vitro platelet aggregation assay demonstrated that M2
bond, whereas M1 was an isomer in which the double bond was was pharmacologically active. These results collectively indicated
migrated to an endocyclic position in the piperidine ring. The that the formation of M2 was mediated by CYP isoforms whereas
conversion of 2-oxo-ticlopidine to M1 in plasma was significantly
increased by the addition of 1 mM CaCl2. In contrast, the activity in
M1, an isomer of M2, was generated either by human PON1 in
plasma or by CES1 in the human liver.
Introduction
thienopyridine antiplatelet agent is a safer, better-tolerated alternative
to ticlopidine.
Ticlopidine [5-(2-chlorophenyl)methyl-4,5,6,7-tetrahydrothieno
[3,2-c] pyridine] (Fig. 1) was the first thienopyridine antiplatelet
agent with potent and long-acting inhibition of platelet aggregation
(Noble and Goa, 1996). It inhibits adenosine 59-diphosphate
(ADP)-induced platelet aggregation by irreversible binding of an
active thiol metabolite to the 2-methylthio-ADP-binding receptor.
Although it is effective in preventing atherothrombotic events in
cardiovascular, cerebrovascular, and peripheral vascular disease,
administration of ticlopidine results in a relatively high incidence
of hematologic toxicities (Lesesve et al., 1994; Love et al., 1998)
such as agranulocytosis (Ono et al., 1991), thrombotic thrombocy-
topenic purpura (Steinhubl et al., 1999), and aplastic anemia
(Mataix et al., 1992). Therefore, clopidogrel, a second-generation
A large inter-individual variability of platelet responsiveness was
observed in patients receiving a standard regimen of clopidogrel
(Gurbel et al., 2003), and this variation was related to CYP2C19
genetic polymorphisms (Shuldiner et al., 2009; Hulot et al., 2011; Yin
and Miyata, 2011; Cuisset et al., 2012). Clopidogrel is converted to
a biologically active thiol metabolite via a two-step enzymatic activation
that involves an initial thiolactone metabolite formation by CYP3A, 2B6,
and 2C19, followed by a thiol metabolite formation produced primarily
by CYP2C19 (Kazui et al., 2010). Recently, paraoxonase 1 (PON1) was
shown to be involved in the generation of a thiol metabolite (Bouman
et al., 2011; Dansette et al., 2012a).
Ticlopidine is reported to be a useful alternative therapy for the
treatment of patients showing resistance to clopidogrel (Aleil et al.,
2007). In addition, the antiplatelet effects of ticlopidine are not af-
fected by CYP2C19 genetic polymorphisms (Farid et al., 2010; Maeda
et al., 2011), suggesting that other drug metabolizing enzymes may be
involved in the formation of a thiol metabolite from 2-oxo-ticlopidine.
This work was supported by a National Research Foundation of Korea (NRF)
grant funded by the Korean Government (MEST) [No. R13-2007-023-00000-0].
dx.doi.org/10.1124/dmd.113.053017.
s This article has supplemental material available at dmd.aspetjournals.org.
ABBREVIATIONS: BNPP, bis-(p-nitrophenyl)phosphate; CES, carboxylesterase; CLint, intrinsic clearance; CMBL, carboxymethylenebutenolidase;
CYP, cytochrome P450; DFP, diisopropylphosphorofluoride; HPLC, high-performance liquid chromatography; LC-MS/MS, liquid chromatography
coupled with tandem mass spectrometry; 2MeSADP, 2-methylthioadenosine diphosphate; MPB, 39-methoxyphenacyl bromide; PCR, polymerase
chain reaction; PMSF, phenylmethylsulfonyl fluoride; PON, paraoxonase.
141

142
Kim et al.
Fig. 1. Postulated pathways for the metabolism
of ticlopidine to thiol metabolites.
parenchymal tissue confirmed as histopathologically normal. The liver tissues
and their clinical information were obtained from Inje Biobank at the Inje
University Busan Paik Hospital. The use of human tissues was approved by the
institutional review board of Busan Paik Hospital.
Even though the initial formation of 2-oxo-ticlopidine from ticlopidine
is known to be mediated by multiple CYP isoforms (Farid et al.,
2010), the pathways that activate ticlopidine have not been identified.
Therefore, the characterization of enzymes involved in the bioactivation
of ticlopidine could be useful in understanding in vivo pharmacologic
activity and differences in the effects of CYP2C19 polymorphisms
among thienopyridine antiplatelet agents.
Our study characterized enzymes responsible for the formation of
a thiolactone and an active thiol metabolite in human using in vitro
systems. Our results showed that multiple CYP isoforms were in-
volved in the formation of 2-oxo-ticlopidine from ticlopidine and the
formation of an active thiol metabolite (exo-form) from 2-oxo-ticlopidine
in human liver microsomes. In addition, a ring-opened thiol metabolite
isomer (endo-form) was generated mainly by carboxylesterase 1 (CES1)
in human liver microsomes and primarily by PON1 in human plasma.
Isolation and Characterization of MPB-Derivatized Ticlopidine Thiol
Metabolite. Because a MPB-derivatized ticlopidine metabolite standard was
not available, the MPB-derivatized metabolite was separated and fractionized
by HPLC. First, the MPB-derivatized ticlopidine active metabolite peak was
identified using liquid chromatography coupled with tandem mass spectrometry
(LC-MS/MS). Samples of microsomes incubated with 200 mM 2-oxo-
ticlopidine were subjected to LC-MS/MS, and the [M+H]+ ion at m/z 446
was consistent with an MPB-derivatized ticlopidine metabolite structure (Fig. 3).
Fractionation was performed using an Agilent 1100 series HPLC system (Agilent,
Palo Alto, CA) and CAPCELL PAK C18 (UG120) column (4.6 ꢀ 250 mm,
5 mm; Shiseido, Tokyo, Japan) with a mobile phase consisting of 0.1% formate
solution/acetonitrile (72/27, v/v) at a flow rate of 1.2 ml/min. The peaks were
repeatedly collected, lyophilized, and analyzed by nuclear magnetic resonance
(NMR). The concentrations of MPB-thiol metabolite were estimated by comparing
the peak signal intensity of the methoxy group to those of tetramethylsilane spiked
in nuclear magnetic resonance solvent (d3-methanol).
Materials and Methods
Chemicals and Reagents. Ticlopidine hydrochloride, clopidogrel hydrogen
sulfate, coumarin, diethyldithiocarbamate, furafylline, ketoconazole, montelu-
kast, quinidine, sulfaphenazole, thio-TEPA, b-nicotinamide adenine dinucle-
otide phosphate, loperamide, glutathione (GSH), bis-(p-nitrophenyl)phosphate
(BNPP), diisopropylphosphorofluoride (DFP), physostigmine sulfate (eserine),
phenylmethylsulfonyl fluoride (PMSF), 39-methoxyphenacyl bromide (MPB),
2-methylthioadenosine diphosphate (2MeSADP), fibrinogen, glucose 6-phosphate,
and glucose-6-phosphate dehydrogenase were purchased from Sigma-Aldrich
(St. Louis, MO). We obtained 2-oxo-ticlopidine, 2-oxo-clopidogrel hydrochloride,
clopidogrel carboxylic acid, clopidogrel carboxylic acid-d4, and S-benzylnirvanol
from Toronto Research Chemicals (Toronto, Canada). The solvents were high-
performance liquid chromatography (HPLC)-grade (Fisher Scientific Co.,
Pittsburgh, PA), and the other chemicals were of the highest quality available.
Baculovirus-insect cell expressed CYP1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6,
3A4 (Supersomes), CES1 (2300 nmol/min/mg for 4-nitrophenyl acetate assay),
CES2 (1300 nmol/min/mg for 4-nitrophenyl acetate assay), and control
microsomes were purchased from BD Gentest Co. (Woburn, MA). Recombi-
nant human carboxymethylenebutenolidase (CMBL) was kindly donated by
Daiichi Sankyo Co. (Tokyo, Japan).
Construction of Recombinant Vectors for PON1 Variants. The wild-type
PON1 cDNA cloned in the pCR4-TOPO plasmid (Thermo Fisher Scientific,
Portsmouth, NH) was used as a polymerase chain reaction (PCR) template
for amplification of the PON1 open-reading frame with an additional 6ꢀ-His
tag to the C-terminal region. PCR primers included a forward primer 59-
CACCATGGCGAAGCTGATTGCGCTCACCCTCTTG-39, and a reverse primer
59-TCAATGGTGATGGTGATGGTGGAGCTCACAGTAAAGAGCT-39. The
amplified PON1 cDNA was cloned into a TOPO entry vector (Invitrogen,
Carlsbad, CA) and subsequently subcloned into a pcDNA-DEST40 (In-
vitrogen) using the Gateway cloning system according to the manufacturer’s
protocol. After construction of wild-type PON1 cDNA in the pcDNA-
DEST40 vector, site-directed mutagenesis was performed using the
QuikChange XL site-directed mutagenesis kit (Stratagene, La Jolla, CA). The
PCR primers for the site-directed mutagenesis were as follows: PON1L55MF
(59-CTGGCTCTGAAGACATGGAGATACTGCCT-39), PON1L55MR (59-
AGGCAGTATCTCCATGTCTTCAGAGCCAG-39), PON1Q192RF (59-
CTTGACCCCTACTTACGATCCTGGGAGATG-39), and PON1Q192RR
(59-CATCTCCCAGGATCGTAAGTAGGGGTCAAG-39). The nucleotides marked
as bold type are mismatches with the PON1 reference sequence. The entire open-
reading frame region was sequenced in both directions, and changes were confirmed
before expression.
Human Plasma and Liver Microsomes. Human plasma was prepared from
fresh blood collected from healthy male subjects (20–30 years old) who had
fasted overnight. Blood (20 ml) was taken and transferred to a heparin
Vacutainer. After centrifugation at 1000g for 10 minutes, plasma was taken
Expression and Purification of Human PON1 in a Mammalian Cell
and stored at 280°C. Human liver microsomes were individually prepared from Line. FreeStyle 293-F cells were transfected with 30 mg of pcDNA-DEST40
empty vector, pcDNA-DEST40-PON1-His-WT, or pcDNA-DEST40-PON1-
removal of metastatic tumors at the Department of General Surgery, Busan Paik His-variants using DNA-293fectin, according to the manufacturer’s instructions
Hospital (Busan, South Korea). The samples were of non-tumor-bearing (Invitrogen). The transfected cells were cultured for 1 day in FreeStyle 293
20 male human liver tissue donors, patients undergoing partial hepatectomy for

Metabolic Conversion of Ticlopidine to Thiol Metabolites.
143
expression medium (Invitrogen). The cells were harvested and resuspended in inhibitors, and the thiol metabolite formation rates from 50 mM 2-oxo-ticlopidine
lysis buffer containing 50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole, measured. The esterase inhibitors used were BNPP and DFP for general CES
and 0.05% Tween 20 (pH 8.0). After centrifugation at 10,000g for 10 minutes, (Heymann and Krisch, 1967), eserine for choline esterase (ChE), CES1, and
the supernatant was mixed with NI-NTA slurry (Qiagen Inc., Valencia, CA).
The overexpressed histidine-tagged proteins were purified by Ni-affinity chro-
matography with an Econo-Column (cat no. 737-1052; Bio-Rad Laboratories,
Hercules, CA.). The elutes were collected and dialyzed against 1 liter of 50 mM
Tris-HCl buffer containing 5% glycerol (pH 7.5); the resultant proteins were
stored at 280°C until use. Protein concentration was determined by using the
method of Bradford with bovine serum albumin as the standard.
CES2 (Preuss and Svensson, 1996; Takahashi et al., 2009), EDTA for PON
(Gonzalvo et al., 1997), clopidogrel for CES1 (Shi et al., 2006), loperamide
for CES2 (Quinney et al., 2005), and PMSF for serine hydrolase (Johnson
and Moore, 2000). Each inhibitor was used at 0.01, 0.1, and 1.0 mM. The
experimental procedure was similar to those for metabolism in human plasma
and liver microsomes. Activities of all inhibitors were compared with those
of inhibitor-free controls. CYP-isoform selective inhibitors were used as
Immunoblotting. Human liver microsomes (30 mg), cellular proteins trans- described earlier.
fected with empty vector (30 mg/lane), and purified PON1 proteins (1 mg/lane)
were separated by electrophoresis in a NuPAGE 4%–12% Bis-Tris gel (In-
vitrogen) and transferred to a polyvinylidene difluoride (PVDF) membrane (GE
Healthcare Bio-Sciences, Piscataway, NJ). The membranes were blocked with
5% nonfat milk in 10 mM Tris-HCl, pH 7.4, including 0.9% NaCl (TBS).
Dialysis. To evaluate the effect of dialysis, human plasma and microsomes
were placed in 10,000 molecular-weight-cutoff dialysis tubes (Slide-A-Lyzer
MINI Dialysis Unit; Thermo Scientific, Waltham, MA) and dialyzed for 20
hours at 4°C against 2-l 0.1 M Tris-HCl buffer (pH 8.0). The dialyzed plasma
and microsomes were incubated with 2-oxo-ticlopidine in the presence and
Specific anti-PON1 IgG (Abcam, Cambridge, MA) was used at a 1:2500 dilution absence of 1.0 mM CaCl2 for 5 minutes. After adding MPB and the internal
as the primary antibody. The membranes were washed thoroughly with TBS plus standard, the samples were evaluated using the LC-MS/MS system.
0.1% Tween 20 (TBST). Immunoreactive proteins were detected using the
enhanced electrochemiluminescence system (GE Healthcare Bio-Sciences).
Formation of 2-Oxo-Ticlopidine in Human Liver Microsomes and
cDNA-Expressed CYP. The incubation mixture consisted of 0.2 mg/ml pooled
Correlation Experiments. For correlation analysis, active thiol metabolite
formation rates from 2-oxo-ticlopidine in 20 human liver microsomes were
determined by incubating 2-oxo-ticlopidine (50 mM) with 0.2 mg/ml
microsomal protein for 5 minutes. Clopidogrel hydrolysis activity as a marker
microsomes or 20 pmol/ml recombinant CYP, ticlopidine (2 and 20 mM), and of CES1 was measured by LC/MS/MS, as described elsewhere (Sato et al.,
an NADPH-generating system (1.3 mM b-nicotinamide adenine dinucleotide 2012). Paraoxon hydrolysis activity was measured as a marker activity of
phosphate, 3.3 mM glucose 6-phosphate, 3.3 mM MgCl2, and 1.0 U/ml
PON1, as described elsewhere (Richter et al., 2009). The correlation
glucose-6-phosphate dehydrogenase) in a total volume of 100 mL phosphate coefficients between active thiol metabolite formation rates and those of
buffer (0.1 M, pH 7.4). The reaction was initiated by the addition of the
NADPH-generating system and continued in a water bath at 37°C for 10
minutes. The reaction was terminated by the addition of the same volume of
acetonitrile-containing internal standard (1 mM of 2-oxo-clopidogrel). After
centrifugation at 16,000g for 5 minutes, a 2-mL aliquot of the supernatant was
injected directly into the LC-MS/MS system.
clopidogrel and paraoxon hydrolysis in the microsomes were evaluated using
Pearson’s correlation coefficient analysis (SAS version 4.3; SAS Institute Inc.,
Cary, NC). P , 0.05 was considered statistically significant.
LC-MS/MS Analysis. HPLC was performed using an ACQUITY UPLC
system (Waters, Milford, MA). The analytical column was a reverse-phase
Kinetex C18 (2.1 ꢀ 100 mm i.d., 2.6 mm; Phenomenex, Torrance, CA). The
Chemical Inhibition of 2-Oxo-Ticlopidine Formation. We used the mobile phases consisted of 0.1% formic acid solution (A) and 0.1% formic acid
following CYP isoform-selective inhibitors: 10 mM furafylline for CYP1A2; in acetonitrile (B). A gradient program was used for HPLC separation with
100 mM coumarin for CYP2A6; 5 mM thio-TEPA for CYP2B6; 1 mM montelukast
a flow rate of 0.4 ml/min. The initial composition of mobile phase B was 15%,
for CYP2C8; 10 mM sulfaphenazole for CYP2C9; 1 mM S-benzylnirvanol for linearly ramped to 45% in 2 minutes, then 95% in 2.5 minutes and maintained
CYP2C19; 10 mM quinidine for CYP2D6; 10 mM diethyldithiocarbamate for for 1.5 minutes, followed by re-equilibration to the initial condition for 1
CYP2E1; and 1 mM ketoconazole for CYP3A. The CYP isoform selective in- minute. The total run time was 5 minutes. HPLC was coupled to a Quattro
hibitors and their concentrations were described elsewhere (Seo et al., 2012). Premier XE LC-MS/MS System (Micromass Ltd., Manchester, UK) equipped
Incubation mixtures consisted of ticlopidine (2 and 20 mM), an inhibitor, pooled with a Turbo ion spray source. Electrospray ionization was performed in the
human liver microsomes (0.2 mg/ml), and an NADPH-generating system in positive ion mode with nitrogen as the drying (700 l/h) and nebulizing (50 l/h)
0.1 ml of 0.1 M potassium phosphate buffer (pH 7.4). The mechanism-based gas. The capillary voltage used was 3500 V and extraction cone voltage 3 V for
inhibitors, furafylline and diethyldithiocarbamate, were preincubated for 15 all compounds. The desolvation temperature was 350°C and source temperature
minutes with microsomes in the presence of an NADPH-generating system 120°C. Quadrupoles Q1 and Q3 were set on unit resolution.
before addition of ticlopidine to initiate the reaction.
Multiple reaction monitoring detection was performed using argon as the
Formation of Thiol Metabolite from 2-Oxo-Ticlopidine. The incubation collision gas. Quantitation of metabolites were performed by monitoring
mixture (100 mL final volume) contained 0.1 M Tris-HCl buffer (pH 8.0),
5 mM glutathione, and various enzyme sources (human liver microsomes,
the transitions of m/z 280→125 for 2-oxo-ticlopidine, m/z 338→155 for
2-oxo-clopidogrel (internal standard for 2-oxo-ticlopidine), m/z 446→154
0.2 mg/ml; human plasma, 1 mL/ml; recombinant CYP isoforms, 20 pmol/ml; for MBP-endo thiol metabolite (M1), m/z 446→266 for MBP-exo-thiol
recombinant PON1 proteins, 0.05 mg/ml; CES1 and CES2, 0.1 mg/ml; CMBL,
0.2 mg/ml). An NADPH-generating system was added to the mixture in
microsomes and recombinant CYP reactions. When assessing the effects of
CaCl2 on the formation of a thiol metabolite, it was used at the concentration of
metabolite (M2), and m/z 308→198 for clopidogrel carboxylic acid (internal
standard for MBP-thiol metabolite. The analytical data were processed using
the MassLynx software (version 4.1; Waters).
In Vitro Platelet Aggregation. Washed platelets were prepared from fresh
1 mM. The mixtures were prewarmed for 5 minutes at 37°C, and the reaction blood taken from healthy male subjects as described elsewhere (Abell and Liu,
was initiated by adding various concentrations of 2-oxo-ticlopidine (0– 2011). Final isolated platelets were responded in modified Tyrode’s buffer
1000 mM). After a 5-minute incubation, the reaction was terminated by adding
100 ml acetonitrile containing 4 mM MPB. The mixture was incubated at 37°C
for 30 minutes to complete the derivatization reaction. After adding the internal
standard (1 mM of clopidogrel carboxylic acid), the mixture was centrifuged at
16,000g for 5 minutes, and an aliquot of the supernatant was injected into the
containing 2 mM calcium chloride and 1 mg/ml human fibrinogen with cell
density between 200,000 and 250,000 platelets/mL. Washed platelets (870 mL)
were transferred into Eppendorf tubes. Then, 10 mL of 2-oxo-ticlopidine stock
solution was added to the final concentration of 5 and 20 mM, and the contents
were mixed by inversion. The reaction was initiated by adding 120 mL of
LC-MS/MS for determination of the MPB-derivatized thiol metabolites. The prewarmed microsomal incubation mixture (2.0 mg/ml microsomes, 25 mM
reactivity of M1 and M2 toward MPB was different in derivatization reaction, glucose-6-phosphate, 1 mM NADP+, 3 U/ml glucose-6-phosphate dehydroge-
and optimal concentration of MPB for the reaction was determined to be 4 mM
(Supplemental Fig. 1).
nase in 0.1 M phosphate buffer, pH 7.4). After 10 minutes of incubation at 37°C,
140 mL of the reaction mixture was pipetted into the wells of a clear-bottomed
Chemical Inhibition of Active Thiol Metabolite Formation. Various microtiter plate followed by the immediate addition of 10 mL of 15 mM
esterase inhibitors were used to determine the enzymes involved in thiol
metabolite formation from 2-oxo-ticlopidine in human plasma and liver
2MeSADP to start the aggregation reaction.
Platelet aggregation was measured by reading the optical density at 595 nm
microsomes. Reaction mixtures were incubated in the presence and absence of in a preheated (37°C) plate reader (SpectraMax, Molecular Devices, Sunnyvale,

144
Kim et al.
CA) every 10 seconds for 3 minutes with mixing between each reading. The
slope for the linear portion of the aggregation time course was used to calculate
the rate of plate aggregation. The percentage inhibition of platelet aggregation
(%IPA) was calculated using the following equation:
Characterization of Thiol Metabolite Formation in Human Liver
Microsomes and Plasma. The identification and quantitation of
thiol metabolites of ticlopidine were performed after derivatization
with MPB. 2-Oxo-ticlopidine was incubated for 5 minutes at 37°C
with human liver microsomes, an NADPH-generating system, and
5 mM glutathione as a reducing agent. Then HPLC analysis was
performed after derivatization of the thiol metabolite by MPB
treatment. Two thiol metabolites (M1 and M2) were identified in
microsomes, but only one thiol metabolite (M1) was detected when
incubated without an NADPH-generating system (Fig. 3, A and B).
M1 was also detected when 2-oxo-ticlopidine was incubated with
human plasma.
Both metabolites showed a protonated ion at m/z 446 and 448 with
the ratio expected for the Cl35 and Cl37 isotopes (Fig. 3, C and D).
However, M1 and M2 produced different product ion mass spectra, as
shown in Fig. 3, E and F. The MS/MS spectrum of protonated M1
showed the fragment ions at m/z 154 and 125. The fragment ions at
m/z 154 (chlorobenzyl methyliminium moiety in ticlopidine) and 125
(chlorobenzyl group) were the ions typically observed in ticlopidine
(Talakad et al., 2011). On the other hand, the MS/MS spectrum
of protonated M2 contained the ions at m/z 296, 266, 154, 140, and
125. The product ion at m/z 296 was generated by the loss of
39-methoxyphenone moiety (MH+-150), and the ion at m/z 266 was
postulated due to the loss of 39-methoxyphenone containing sulfur
moiety (MH+-180).
Previous reports demonstrated that the thiol metabolite bearing
exocyclic double bond (exo-form) showed the same fragment pattern,
yielding the corresponding product ion at m/z 354 (MH+–150) in
clopidogrel (Dansette et al., 2012a) and at m/z 348 (MH+–150) in
prasugrel (Dansette et al., 2012b). In addition, the thiol metabolite
in which the double bond was migrated to an endocyclic position
in the piperidine ring (endo-form) did not produce these fragment
ions. Instead, the ions at m/z 212 (clopidogrel) and 206 (prasugrel),
equivalent to the corresponding ion at m/z 154 in ticlopidine, were
observed as the most abundant ion in clopidogrel and prasugrel,
respectively. Based on MS/MS spectra, M1 and M2 were pro-
visionally assigned to be an endo-form and exo-form of the thiol
metabolite, respectively.
%IPA ¼ ð1 2 ðRreaction=RcontrolÞ ꢀ 100
where Rreaction equals the slope of the aggregation assay containing 2-oxo-
ticlopidine and microsomal mixture and Rcontrol equals the slope of the
aggregation assay containing only microsomal mixture.
Data Analysis. The apparent kinetic parameters of 2-oxo ticlopidine
metabolism were determined by fitting the unweighted kinetic data from plasma,
human liver microsomes, and human esterases to an one-enzyme Michaelis-
Menten equation or a Hill equation [V = Vmax ꢀ [S]n/(Km + [S]n)]. Calculated
parameters included the maximum rate of metabolite formation (Vmax),
Michaelis-Menten constant (apparent Km), the intrinsic clearance (CLint
=
Vmax/Km), and Hill coefficient (n). The percentages of inhibition were calculated
by the ratio of the rate of metabolite formation with and without the specific
inhibitor. Calculations were performed using WinNonlin, version 2.1 (Pharsight,
Mountain View, CA).
Results
Identification of Enzymes Responsible for the Formation of
2-Oxo-Ticlopidine. Ticlopidine was metabolized to 2-oxo-ticlopidine
in human liver microsomes only when NADPH was added to the
incubation. To characterize CYP isoforms involved in the formation of
2-oxo-ticlopidine, chemical inhibition with CYP isoform selective
inhibitors and metabolism with cDNA-expressed CYP were performed.
The formation of 2-oxo-ticlopidine in human liver microsomes was
not substantially inhibited by any of CYP-selective inhibitors used at
2 and 20 mM ticlopidine (data not shown). No significant inhibition
of 2-oxo-ticlopidine formation by CYP isoform selective inhibitors
might be due to the involvement of multiple CYP isoforms. The
formation rates of 2-oxo-ticlopidine from ticlopidine (2 and 20 mM)
using cDNA-expressed CYP1A2, 2A6, 2B6, 2C8, 2C8, 2C19, 2D6,
and 3A4 are shown in Fig. 2. Multiple CYP isoforms mediated
2-oxo-ticlopidine formation. Among them, CYP2D6 showed the
highest activity at 2 mM whereas CYP2B6 and 2C19 showed
relatively high activity at 20 mM ticlopidine.
Fig. 2. Formation of 2-oxo-ticlopidine from ticlopidine by human
cDNA-expressed CYP isoforms. Ticlopidine at two different
concentrations of 2 (j) and 20 mM (u) were incubated with each
recombinant CYP isoform (CYP1A2, 2A6, 2B6, 2C8, 2C9, 2C19,
2D6, 3A4, 3A5, and 2J2) and an NADPH-generating system at 37°C
for 5 minutes. Each bar represents the mean 6 S.D. of triplicate
determinations.

Metabolic Conversion of Ticlopidine to Thiol Metabolites.
145
Fig. 3. HPLC chromatograms of incubations of 2-oxo-ticlopidine with human liver microsomes with (A) and without (B) an NADPH-generating system with MS detection
at m/z 446 after derivatization with 39-methoxy-phenacyl bromide, mass spectra of M1 (C) and M2 (D), and product ion mass spectra of M1 (E) and M2 (F).
Formation of thiol Metabolites in Human Liver Microsomes to M1 might be mediated by PON1 (Fig. 4B). The involvement of
and Plasma. Conversion of 2-oxo-ticlopidine to M1 was observed PON1 in the formation of the thiol metabolite from 2-oxo-clopidogrel
regardless of the presence of NADPH in microsomal incubation has been reported previously (Bouman et al., 2011; Dansette et al.,
whereas M2 was generated only in the presence of NADPH (Fig. 4A). 2012a). The addition of CaCl2 to dialyzed plasma restored the acti-
In addition, the formation of M1 was not inhibited by 1-amino- vity. In contrast, M1 formation from 2-oxo-ticlopidine in microsomes
benzotriazole, a general CYP inhibitor (data not shown). To assess was not significantly changed in the presence of CaCl2. Dialysis of
whether another tissue source was responsible for the catalytic microsomes for 24 hours resulted in a 37% loss of activity, which was
conversion of 2-oxo-ticlopidine to M1, 2-oxo-ticlopidine was in- not restored by addition of CaCl2. These results collectively indicated
cubated with human plasma. M1 formation was also observed in an that M1 formation was mediated by enzymes other than CYP or
incubation of 2-oxo-ticlopidine with human plasma. In addition, the PON1.
formation rate was increased by 2;2.6-fold in the presence of 1 mM
Kinetic analyses for the formation rates of M1 and M2 from
CaCl2 whereas it was abolished almost completely after dialysis of 2-oxo-ticlopidine were conducted using human plasma and liver
plasma, suggesting that the catalytic conversion of 2-oxo-ticlopidine microsomes (Fig. 5). The parameters of enzyme kinetics are summarized
Fig. 4. Formation rates of thiol metabolites after incubation of 2-oxo-ticlopidine with human liver microsomes in the presence and absence of an NADPH-generating system
(A), and effects of Ca2+ and dialysis on M1 formation rates in human plasma and microsomes (B). Each data point represents the mean 6 S.D. of triplicate determinations.

146
Kim et al.
Fig. 5. Kinetics of M1 formation from 2-oxo-ticlopidine in pooled human plasma (A), pooled human liver microsomes (B) in the absence (d) or presence (s) of 1 mM
CaCl2, and M2 formation from 2-oxo-ticlopidine in pooled human liver microsomes (C). An increasing concentration of 2-oxo-ticlopidine was incubated with microsomes
(0.2 mg/ml) or plasma (1 mL/ml) at 37°C for 5 minutes. Each data point represents the mean 6 S.D. of triplicate determinations.
in Table 1. Conversion of 2-oxo-ticlopidine to M1 and M2 followed possible involvement of CES, especially CES1, in M1 formation in
a single enzyme Michaelis-Menten kinetics. The Km and Vmax for M1 human liver microsomes.
formation in plasma were calculated to be 306 mM and 3.72 nmol/
Identification of the Human Esterases Responsible for M1
min/mg, respectively, and the kinetic values in microsomes were Formation. The enzymes involved in the formation of M1 from
382 mM and 5.81 nmol/min/mg, respectively. Addition of 1 mM CaCl2 2-oxo-ticlopidine were further confirmed using various cDNA-
caused a 2.3-fold increase in Vmax value without affecting Km value in expressed enzymes including PON1 (wild-type, Q192R, and L55M),
plasma whereas it did not change the Km and Vmax values in case of CES1, CES2, and CMBL. M1 formation rates from 2-oxo-ticlopidine
microsomes. The Km and Vmax for M2 formation in microsomes were (50 mM) using PON1, CES1, CES2 and CMBL are shown in Fig. 7.
calculated to be 22.8 mM and 0.13 nmol/min/mg, respectively. In vitro PON1 catalyzed the conversion of 2-oxo-ticlopidine to M1 at the
intrinsic clearance (CLint) for M1 formation was approximately 3-fold highest rate. CES1 also showed considerable catalytic activity with
higher than that for M2 formation in microsomes.
;15% of the formation rate of M1 by PON1. In contrast, CES2 and
Chemical Inhibition of Formation of M1 in Human Liver CMBL showed negligible catalytic conversion of 2-oxo-ticlopidine
Microsomes and Plasma. To further characterize the enzymes to M1.
involved in the formation of M1 from 2-oxo-ticlopidine, chemical
To determine whether a PON1 polymorphism affects M1 formation
inhibition studies were performed using representative esterase from 2-oxo-ticlopidine, M1 formation rates were measured after the
inhibitors; BNPP and DFP as CES inhibitors; eserine as an inhibitor of incubation with cDNA-expressed PON1 variants (wild-type, Q192R,
ChE, CES1, and CES2; EDTA as a PON inhibitor; clopidogrel as and L55M). The PON1 gene has a common polymorphism within
a CES1 inhibitor; loperamide as a CES2 inhibitor; and PMSF as coding region causing amino acid substitution that results in a change
a serine hydrolase inhibitor. Concentrations of each esterase inhibitor in catalytic activity (Mackness et al., 1998; Bhattacharyya et al.,
were 0.01, 0.1, and 1.0 mM. Catalytic activity was completely inhibited 2008). Expression of PON1 proteins was confirmed by immunoblot-
by .0.1 mM EDTA but not by other esterase inhibitors in plasma ting with a monoclonal anti-PON1 antibody (Fig. 8A). PON1 192Q
(Fig. 6A). However, M1 formation in microsomes was substantially showed the highest catalytic activity (5.06 6 0.24 nmol/min/mg) with
inhibited by CES inhibitors such as BNPP, DFP, clopidogrel, 50 mM 2-oxo-ticlopidine in the presence of 1 mM CaCl2. The catalytic
and PMSF in a concentration-dependent manner. In addition, the activities of PON1 192R and 55M were 85% and 50%, respectively,
conversion of 2-oxo-ticlopidine to M1 in microsomes was mildly that of PON1 192Q (Fig. 8B). These results suggested that the genetic
inhibited by eserine whereas no inhibition was observed by EDTA and polymorphisms of PON1 Q192R and L55M might affect the for-
loperamide (Fig. 6B). The chemical inhibition results indicated the mation of M1 from 2-oxo-ticlopidine.
TABLE 1
Kinetic parameters of M1 and M2 formation by human tissue preparations
M1
M2
Vmax
Enzyme Source
Km
Vmax
CLint
Km
CLint
mM
nmol/min/mg
protein
ml/min/mg protein
mM
nmol/min/mg
protein
ml/min/mg
protein
Plasma
Plasma + CaCl2
Microsome
Microsome + CaCl2
Microsome + NADPH
274
297
210
274
280
3.36
6.03
6.90
8.03
5.32
12.3
20.3
32.9
29.3
19.0
—
—
—
—
—
—
a
—
—
—
—
—
—
a
22.8
0.13
5.83
a
Concentration of CaCl2 was 1 mM.

Metabolic Conversion of Ticlopidine to Thiol Metabolites.
147
Fig. 6. Effects of carboxyesterase, choline esterase, and paraoxonase 1 inhibitors on M1 formation in human plasma (A) and liver microsomes (B). Pooled microsomes
(0.2 mg/ml) or plasma (1 ml/ml) were incubated with 50 mM 2-oxo-ticlopidine in the absence or presence of various chemical inhibitors at 37°C for 5 minutes. Each data
point represents the mean 6 S.D. of triplicate determinations.
We performed kinetic analyses for M1 formation by cDNA- decreased by 38% and 71%, respectively. Kinetic analyses showed
expressed PON1 variants and CES1 to determine the substrate affinity that genetic polymorphisms in PON1 gene did not affect substrate
and intrinsic clearance of PON1 and CES1 (Fig. 7B and Fig. 8C). The binding affinity but decreased the catalytic conversion rates of 2-oxo-
parameters of enzyme kinetics are summarized in Table 2. M1 ticlopidine to M1.
formation in recombinant PON1 variants (wild-type, Q192R, and
To further characterize the role of CES1 in M1 formation in human
L55M) and CES1 also exhibited single-enzyme Michaelis-Menten liver microsomes, M1 formation rates from 2-oxo-ticlopidine were
kinetics. The Km and Vmax values of CES1 were 277 mM and compared with clopidogrel hydrolysis activities, a known marker
2.16 nmol/min/mg, respectively, resulting in CLint value of 7.8 mL/ activity of CES1 (Sato et al., 2012), in 20 human liver microsomes
min/mg. In contrast, the Km and Vmax values of PON1 192Q were 338 mM (Fig. 9). M1 formation rates were statistically significantly correlated
and 145 nmol/min/mg, respectively. The Km values of PON1 192R with the rates of clopidogrel hydrolysis (r = 0.9110, P , 0.0001). M1
and 55M were similar to that of PON1 192Q, but Vmax values were formation rates were also considerably correlated with paraoxon
Fig. 7. M1 formation rates (A), and kinetics of M1 formation from 2-oxo-ticlopidine (B) by human cDNA-expressed PON1, CES1, CES2, and CMBL. An increasing
concentration of 2-oxo-ticlopidine was incubated with recombinant human PON1 (0.05 mg/ml), CES1, CES2 (0.1 mg/ml), and CMBL (0.2 mg/ml) at 37°C for 5 minutes.
Each data point represents the mean 6 S.D. of triplicate determinations.

148
Kim et al.
Fig. 8. The effects of PON1 polymorphisms (Q192R and L55M) on M1 formation from 2-oxo-ticlopidine. Immunoblotting of recombinant PON1 192Q, 192R, and 55M
expressed in Free Style 293-F cells (A). Formation rates of M1 from 2-oxo-ticlopidine by mock, recombinant human PON1 192Q, 192R, and 55M. The proteins were
incubated with 50 mM 2-oxo-ticlopidine (B). Kinetics of M1 formation from 2-oxo-ticlopidine by PON1 192Q (d), 192R (s), and 55M (.) (C). Each data point represents
the mean 6 S.D. of triplicate determinations.
hydrolysis rates, but a lesser degree of correlation was observed (r = considerable activity, but the M2 formation rates were 8% and 28% of
0.5999, P = 0.0066).
CYP2C19 catalytic activity, respectively. However, relative formation
Identification of CYP Isoforms Responsible for M2 Formation. rates of M2 by CYP1A2, 2D6, and 3A4 compared with that by
M2 formation was observed in human liver microsomes only when an CYP2C19 were increased while CYP2B6-mediated rate was not much
NADPH-generating system was added. To characterize CYP isoforms changed at a 50 mM 2-oxo-ticlopidine.
involved in the formation of M2, chemical inhibition with CYP
Kinetic analyses for the formation of M2 from 2-oxo-ticlopidine
isoform selective inhibitors and metabolism with cDNA-expressed were performed using cDNA-expressed CYP1A2, 2B6, 2C19, 2D6,
CYP were performed. The formation of M2 in microsomes was not and 3A4 (Fig. 11 and Table 3). Conversion of 2-oxo-ticlopidine to M2
substantially inhibited by any of CYP-selective inhibitors used at 5 by CYP2C19 and 2D6 followed a single enzyme Michaelis-Menten
and 50 mM 2-oxo-ticlopidine (data not shown). No significant kinetics. On the other hand, the reaction mediated by CYP1A2, 2B6,
inhibition of M2 formation by CYP isoform selective inhibitors might and 3A4 was best fitted by a Hill equation. Eadie-Hofstee plots of
be due to the involvement of multiple CYP isoforms. The formation M2 formation by CYP1A2, 2B6, and 3A4 showed the convex re-
rates of M2 from 2-oxo-ticlopidine (5 and 50 mM) by CYP1A2, 2A6, lationship, indicating positive cooperativity (n . 1). The saturation of
2B6, 2C8, 2C9, 2C19, 2D6, and 3A4 are shown in Fig. 10. CYP2B6 CYP2B6- and 2C19-mediated M2 formation at relatively low con-
and 2C19 catalyzed the conversion of 2-oxo-ticlopidine to M2 with centrations of 2-oxo-ticlopidine might be due to metabolism-dependent
similar rates at both concentrations. CYP1A2 and 2D6 also showed inactivation of these enzymes by 2-oxo-ticlopidine. The metabolism-
dependent inactivation of CYP2B6 and CYP2C19 by ticlopidine is
well characterized (Nishiya et al., 2009a,b). Although the Vmax values
for M2 formation by CYP2C19, CYP2D6, and CYP3A4 were greater
TABLE 2
Kinetic parameters of M1 formation from 2-oxo-ticlopidine by human cDNA-
(2.0-, 3.4- and 2.3-fold, respectively) compared with that of CYP2B6,
the Km values obtained from CYP2C19, CYP2D6, and CYP3A4 were
much larger (7.4-, 38.3- and 337-fold, respectively) than that of
CYP2B6. Consequently, in vitro CLint for CYP2B6-catalyzed M2
formation was 3.6-, 11.3-, and 155-fold faster compared with
CYP2C19, CYP2D6, and CYP3A4, respectively.
expressed esterases
Enzyme Source
Km
Vmax
CLint
mM
nmol/min/mg protein
145
90.1
42.2
2.16
ml/min/mg protein
PON1 wild type
PON1 Q192R
PON1 L55M
CES1
388
346
355
277
374
260
119
7.80
In Vitro Inhibition of Platelet Aggregation. To clarify the
pharmacologic activity of M1 and M2, in vitro platelet aggregation

Metabolic Conversion of Ticlopidine to Thiol Metabolites.
149
Fig. 9. Correlation analysis of the rates of thiol metabolite formation from 2-oxo-ticlopidine and the hydrolysis rates of clopidogrel (A) and paraoxon (B) in 20 human liver
microsomes. The concentrations of 2-oxo-ticlopidine, clopidogrel, and paraoxon were 50, 5, and 3000 mM, respectively.
assay was performed. Because M2 was generated by CYP isoforms
It is generally accepted that clopidogrel, a structural analog of
and M1 was formed by NADPH-independent esterases in microsomes, ticlopidine, is converted to a thiol metabolite by two-step enzymatic
washed platelets were incubated with 2-oxo-ticlopidine and micro- conversion and that CYP isoforms are responsible for both steps of the
somes in the presence and absence of an NADPH-generating system. enzymatic reaction (Kazui et al., 2010). CYP2C19 is the principal
Inhibition of 2MeSADP-induced platelet aggregation was observed enzyme responsible for the second step enzymatic conversion of
when 2-oxo-ticlopidine was incubated with microsomes in the pre- 2-oxo-clopidogrel to the thiol metabolite. Resistance to clopidogrel
sence of an NADPH-generating system (24 and 38% IPA in 5 and 20 mM, therapy is due primarily to CYP2C19 polymorphisms (Hulot et al.,
respectively) whereas no substantial inhibition was noted in the incubation 2011; Yin and Miyata, 2011; Cuisset et al., 2012). Recently, PON1
without an NADPH-generating system (Fig. 12).
was reported to be a major enzyme responsible for the formation
of a thiol metabolite of clopidogrel (Bouman et al., 2011). However,
the contribution of PON1 to thiol metabolite formation was limited
to the hydrolysis of the endo-form of 2-oxo-clopidogrel while CYP
are primarily responsible for the hydrolysis of the exo-form of
2-oxo-clopdogrel, leading to the generation of two cis diastereomers
(Dansette et al., 2012a). 2-Oxo-ticlopidine can also exist as two
rotamers (endo- and exo-forms), and at least two ticlopidine thiol
metabolite isomers could be expected in in vitro incubation.
Two ring-opened thiol metabolites were identified in our system,
and their structures were characterized on the basis of their product ion
mass spectra and the fragmentation pattern of thiol metabolites of
clopidogrel and prasugrel, structural analogs of ticlopidine, as
presented elsewhere (Dansette et al., 2012a,b). M1 observed in
plasma and microsomal incubation was characterized as a thiol
metabolite in which the double bond was migrated to an endocyclic
position in the piperidine ring (endo-form). On the other hand, M2
generated only in microsomal incubation in the presence of an NADPH-
generating system was identified as a thiol metabolite bearing an exocyclic
double bond (exo-form).
Discussion
Our results suggest that ticlopidine is metabolized to thiol
metabolites via two-step enzymatic reactions. Multiple CYP isoforms
were involved in the conversion of ticlopidine to 2-oxo-ticlopidine,
which was further metabolized to ring-opened thiol metabolites either
by PON-1 and CES1 (M1 formation), or by CYP2B6, 2D6, and 2C19
(M2 formation) (Fig. 1).
PON1 is the major enzyme responsible for the metabolic conversion
of 2-oxo-ticlopidine to M1 in plasma, and CES1 is the primary
mediator of the reaction in human liver microsomes. Involvement of
plasma PON1 was verified using several approaches. First, the M1
formation rate in plasma was increased in the presence of 1 mM
CaCl2. In addition, dialysis of plasma completely abolished the
catalytic activity and addition of 1 mM CaCl2 to dialyzed plasma
restored the activity to the original level, indicating that M1 formation
in plasma is dependent on Ca2+. Second, plasma activity was
completely inhibited by EDTA but not by other esterase inhibitors
such as BNPP, eserine, and clopidogrel. Finally, cDNA-expressed
Fig. 10. Formation of M2 from 2-oxo-ticlopidine by human cDNA-expressed CYP
isoforms. 2-Oxo-ticlopidine at two different concentrations of 5 (j) and 50 mM (u)
were incubated with each recombinant CYP isoform (CYP1A2, 2A6, 2B6, 2C8,
2C9, 2C19, 2D6, 3A4, 3A5, and 2J2). Each bar represents the mean 6 SD of
triplicate determinations.

150
Kim et al.
Fig. 11. Kinetics for the formation of M2 from 2-oxo-ticlopidine in human cDNA-expressed CYP 1A2 (A), 2B6 (B), 2C19 (C), 2D6 (D), and 3A4 (E). An increasing
concentration of 2-oxo-ticlopidine was incubated with each recombinant CYP (20 pmol/ml) and an NADPH-generating system at 37°C for 5 minutes. The corresponding
Eadie-Hofstee plots are shown in the insert, respectively. The kinetic data were fitted by a Hill equation. Each data point represents the mean 6 S.D. of triplicate
determinations.
PON1 showed strong activity toward the formation of M1. These The activity in microsomes was also not inhibited by 1 mM EDTA,
results collectively indicated that PON1 was primarily involved in M1 a concentration that resulted in complete loss of the activity in hu-
formation from 2-oxo-ticlopidine. PON1 is known to be involved in man plasma. The involvement of CES1 in M1 formation was further
the hydrolysis of lactone or thiolactone derivatives such as pilocarpine demonstrated using esterase inhibitors such as BNPP, eserine, and
(Billecke et al., 2000; Hioki et al., 2011) and homocysteine thiolactone clopidogrel. Clopidogrel is hydrolyzed primarily by CES1 (Hagihara
(Borowczyk et al., 2012).
et al., 2009; Farid et al., 2010) and eserine is a more selective inhibitor
The primary contribution of CES1 to thiol metabolite formation in
liver microsomes was somewhat unexpected when considering the
distribution of PON1 in the liver and the structure of 2-oxo-ticlopidine.
PON1 is known to be distributed in liver microsomes and in the
cytosol (Gonzalvo et al., 1998), and considerable paraoxon hydrolysis
activity was detected in our system. However, microsomal conversion
of 2-oxo-ticlopidine to M1 was not affected by the addition of CaCl2.
TABLE 3
Kinetic parameters of M2 formation from 2-oxo-ticlopidine by human cDNA-
expressed CYP isoforms
CYP Isoform
Km
Vmax
CLint
na
mM
pmol/min/pmol CYP
mL/min/pmol CYP
CYP1A2
CYP2B6
CYP2C19
CYP2D6
CYP3A4
128
0.59
4.39
22.6
199
0.40
0.73
1.49
2.50
1.67
0.003
1.24
0.34
0.11
0.008
1.61
1.69
—
Fig. 12. Inhibition of in vitro 2MeSADP-induced aggregation of washed human
platelets by 2-oxo-ticlopidine. Washed platelets were incubated with 5 mM (j) and
20 mM (u) 2-oxo-ticlopidine and microsomes in the presence and absence of an
NADPH-generating system. Each data point represents the mean 6 S.D. of triplicate
determinations.
—
1.54
a
Hill coefficient.

Metabolic Conversion of Ticlopidine to Thiol Metabolites.
151
of CES1 than CES2 (Takahashi et al., 2009; Sato et al., 2012). 2-oxo-ticlopidine was observed in microsomal incubation only in the
Furthermore, recombinant CES1 showed marked activity in terms of presence of an NADPH-generating system, suggesting that M2 was
M1 formation, whereas recombinant CES2 did not. A significant mainly responsible for the inhibition of platelet aggregation like other
correlation (r = 0.9110, P , 0.0001) between clopidogrel hydrolysis thienopyridine antiplatelet agents. Therefore, contribution of CYP2B6
rates and M1 formation rates in a panel of human liver microsomes to the formation of M2 along with CYP2C19 and 2D6 may explain how
additionally supported the role of CES1 in microsomes. M1 formation ticlopidine is effective in patients resistant to clopidogrel therapy, as
rates were also correlated with paraoxon hydrolysis rates (r = 0.5999, evidenced by Aleil et al. (2007), and the antiplatelet effects of ticlopidine
P = 0.0066) but the degree of correlation was less than that with are not affected by CYP2C19 genetic polymorphisms (Farid et al., 2010;
clopidogrel hydrolysis rates. Our results collectively demonstrated that Maeda et al., 2011).
CES1 was primarily responsible for the conversion of 2-oxo-ticlopidine
In conclusion, our results have demonstrated that multiple CYP
to M1 in human liver microsomes. Considering chemical inhibition isoforms are involved in the formation of 2-oxo-ticlopidine from
data, other esterases such as PON-1 might also be involved in the ticlopidine and the formation of a pharmacologically active thiol meta-
formation of M1 from 2-oxo-ticlopidine but their contribution seemed bolite (exo-form) from 2-oxo-ticlopidine in human liver microsomes.
to be much smaller compared with CES-1. It is generally known that In addition, a ring-opened thiol metabolite isomer (endo-form) is gen-
drugs containing ester and thioester linkage can be served as good erated mainly by CES1 in human liver microsomes and primarily by
substrates for CES (Williams, 1985). Hydrolysis of cyclic lactone or PON1 in human plasma. These results support that ticlopidine could
thiolactone by CES1 has not been reported to our best knowledge, be an alternative therapy in case of pharmacologic resistance to
although lactone ring-containing statins such as simvastatin, lovastatin, clopidogrel, showing that its antiplatelet effects are affected by
and mevastatin can inhibit the activities of recombinant CES1. The CYP2C19 genetic polymorphisms.
presence of a lactone ring in the statin backbone is essential for
inhibition of CES1 (Fleming et al., 2005; Fukami et al., 2010).
PON1 has two common polymorphisms, Q192R and L55M. The
Q192R polymorphism affects the hydrolysis of various substrates
Authorship Contributions
Participated in research design: Lee, Shin, D.-H. Kim.
Conducted experiments: M.-J. Kim, Jeong, Park.
as the kinetics of some substrates are accelerated (e.g., paraoxon)
whereas others are slowed down (e.g., soman and diazoxon) or
unaffected (e.g., phenylacetate) (Billecke et al., 2000). Olmesartan
medoxomil and prulifloxacin, which are converted to an active
metabolite by PON1, are known to be more efficiently bioactivated by
PON1 192R (Tougou et al., 1998; Ishizuka et al., 2010). In addition,
a lactone derivative pilocarpine is also hydrolyzed more rapidly by
PON1 192R (Hioki et al., 2011). However, our results showed that the
CLint value of the 2-oxo-ticlopidine hydrolyase activity by recombinant
PON1 192Q was 1.65- and 3.62-fold greater than that of recombinant
PON1 192R and PON1 55M, respectively. 2-Oxo-ticlopidine, which
contains a thiolactone moiety, may bind differently to the polymorphic
PON1 than pilocarpine and olmesartan medoxomil.
Unlike to the formation of M1 from 2-oxo-ticlopidine, M2 was
exclusively generated by CYP isoforms. Experiments with cDNA-
expressed CYP isoforms showed that multiple CYP isoforms were
involved in the formation of M2 although their catalytic activities
varied. Kinetic analyses showed that the substrate binding affinity of
CYP2B6 (Km value of 0.59 mM) was 7.4- to 337-fold greater than
those of other CYP isoforms. In addition, in vitro CLint for CYP2B6-
catalyzed M2 formation from 2-oxo-ticlopidine was 413-, 3.6-, 11.3-,
and 155-fold faster than the corresponding values for CYP1A2, 2C19,
2D6, and 3A4, respectively (Table 3). These results indicate that
CYP2B6 plays a major role in the conversion of 2-oxo-ticlopidine to
M2. CYP2C19 and CYP2D6 also contribute to the formation of M2 to
a considerable extent, but the role of CYP1A2 and 3A4 seems to be
negligible in human liver microsomes. The ratios of formation rate of
an active thiol metabolite (exo-form) to its isomer (endo-form) in
human liver microsomes were 8.6 and 3.9 times for clopidogrel and
prasugrel, respectively (Dansette et al., 2012a,b). However, a reverse
phenomenon was noted in ticlopidine, and our results showed that
the formation rate of M2 was 4-fold higher than that of M1 in
microsomes. The physiologic role of endo-metabolites in thienopyridine
antiplatelet agents may need to be further clarified.
Performed data analysis: M.-J. Kim, Jeong, Choi, Ghim, Shin, D.-H. Kim.
Wrote or contributed to the writing of the manuscript: M.-J. Kim, D.-H. Kim.
References
Abell LM and Liu EC (2011) Dissecting the activation of thienopyridines by cytochromes P450
using a pharmacodynamic assay in vitro. J Pharmacol Exp Ther 339:589–596.
Aleil B, Rochoux G, Monassier JP, Cazenave JP, and Gachet C (2007) Ticlopidine could be an
alternative therapy in the case of pharmacological resistance to clopidogrel: a report of three
cases. J Thromb Haemost 5:879–881.
Bhattacharyya T, Nicholls SJ, Topol EJ, Zhang R, Yang X, Schmitt D, Fu X, Shao M, Brennan
DM, and Ellis SG, et al. (2008) Relationship of paraoxonase 1 (PON1) gene polymorphisms and
functional activity with systemic oxidative stress and cardiovascular risk. JAMA 299:1265–1276.
Billecke S, Draganov D, Counsell R, Stetson P, Watson C, Hsu C, and La Du BN (2000) Human
serum paraoxonase (PON1) isozymes Q and R hydrolyze lactones and cyclic carbonate esters.
Drug Metab Dispos 28:1335–1342.
Borowczyk K, Shih DM, and Jakubowski H (2012) Metabolism and neurotoxicity of homo-
cysteine thiolactone in mice: evidence for a protective role of paraoxonase 1. J Alzheimers Dis
30:225–231.
Bouman HJ, Schömig E, van Werkum JW, Velder J, Hackeng CM, Hirschhäuser C, Waldmann
C, Schmalz HG, ten Berg JM, and Taubert D (2011) Paraoxonase-1 is a major determinant of
clopidogrel efficacy. Nat Med 17:110–116.
Cuisset T, Morange PE, and Alessi MC (2012) Recent advances in the pharmacogenetics of
clopidogrel. Hum Genet 131:653–664.
Dansette PM, Rosi J, Bertho G, and Mansuy D (2012a) Cytochromes P450 catalyze both steps of
the major pathway of clopidogrel bioactivation, whereas paraoxonase catalyzes the formation
of a minor thiol metabolite isomer. Chem Res Toxicol 25:348–356.
Dansette PM, Rosi J, Debernardi J, Bertho G, and Mansuy D (2012b) Metabolic activation of
prasugrel: nature of the two competitive pathways resulting in the opening of its thiophene ring.
Chem Res Toxicol 25:1058–1065.
Farid NA, Kurihara A, and Wrighton SA (2010) Metabolism and disposition of the thienopyridine
antiplatelet drugs ticlopidine, clopidogrel, and prasugrel in humans. J Clin Pharmacol 50:
126–142.
Fleming CD, Bencharit S, Edwards CC, Hyatt JL, Tsurkan L, Bai F, Fraga C, Morton CL, Howard-
Williams EL, and Potter PM, et al. (2005) Structural insights into drug processing by human
carboxylesterase 1: tamoxifen, mevastatin, and inhibition by benzil. J Mol Biol 352:165–177.
Fukami T, Takahashi S, Nakagawa N, Maruichi T, Nakajima M, and Yokoi T (2010) In vitro
evaluation of inhibitory effects of antidiabetic and antihyperlipidemic drugs on human car-
boxylesterase activities. Drug Metab Dispos 38:2173–2178.
Gonzalvo MC, Gil F, Hernández AF, Villanueva E, and Pla A (1997) Inhibition of paraoxonase
activity in human liver microsomes by exposure to EDTA, metals and mercurials. Chem Biol
Interact 105:169–179.
Gonzalvo MC, Gil F, Hernandez AF, Rodrigo L, Villanueva E, and Pla A (1998) Human liver
paraoxonase (PON1): subcellular distribution and characterization. J Biochem Mol Toxicol 12:
61–69.
Gurbel PA, Bliden KP, Hiatt BL, and O’Connor CM (2003) Clopidogrel for coronary stenting:
response variability, drug resistance, and the effect of pretreatment platelet reactivity. Circu-
lation 107:2908–2913.
Hagihara K, Kazui M, Kurihara A, Yoshiike M, Honda K, Okazaki O, Farid NA, and Ikeda T
(2009) A possible mechanism for the differences in efficiency and variability of active me-
tabolite formation from thienopyridine antiplatelet agents, prasugrel and clopidogrel. Drug
Metab Dispos 37:2145–2152.
Heymann E and Krisch K (1967) Phosphoric acid-bis-(p-nitro-phenylester), a new inhibitor
of microsomal carboxylesterases [article in German]. Hoppe Seylers Z Physiol Chem 348:
609–619.
Because only the clopidogrel thiol metabolite that possesses the cis
configuration with the double bond outside the pyridine ring is known
to be pharmacologically active (Tuffal et al., 2011), we assessed the
pharmacologic activity of M1 and M2 using in vitro platelet aggre-
gation assay. Inhibition of 2MeSADP-induced platelet aggregation by

152
Kim et al.
Hioki T, Fukami T, Nakajima M, and Yokoi T (2011) Human paraoxonase 1 is the enzyme
responsible for pilocarpine hydrolysis. Drug Metab Dispos 39:1345–1352.
Hulot JS, Collet JP, Cayla G, Silvain J, Allanic F, Bellemain-Appaix A, Scott SA,
and Montalescot G (2011) CYP2C19 but not PON1 genetic variants influence clopidogrel
pharmacokinetics, pharmacodynamics, and clinical efficacy in post-myocardial infarction
patients. Circ Cardiovasc Interv 4:422–428.
Ishizuka T, Fujimori I, Kato M, Noji-Sakikawa C, Saito M, Yoshigae Y, Kubota K, Kurihara A,
Izumi T, and Ikeda T, et al. (2010) Human carboxymethylenebutenolidase as a bioactivat-
ing hydrolase of olmesartan medoxomil in liver and intestine. J Biol Chem 285:11892–
11902.
Quinney SK, Sanghani SP, Davis WI, Hurley TD, Sun Z, Murry DJ, and Bosron WF (2005)
Hydrolysis of capecitabine to 59-deoxy-5-fluorocytidine by human carboxylesterases and in-
hibition by loperamide. J Pharmacol Exp Ther 313:1011–1016.
Richter RJ, Jarvik GP, and Furlong CE (2009) Paraoxonase 1 (PON1) status and substrate
hydrolysis. Toxicol Appl Pharmacol 235:1–9.
Sato Y, Miyashita A, Iwatsubo T, and Usui T (2012) Conclusive identification of the oxybutynin-
hydrolyzing enzyme in human liver. Drug Metab Dispos 40:902–906.
Seo KA, Lee SJ, Kim KB, Bae SK, Liu KH, Kim DH, and Shin JG (2012) Ilaprazole, a new
proton pump inhibitor, is primarily metabolized to ilaprazole sulfone by CYP3A4 and 3A5.
Xenobiotica 42:278–284.
Johnson G and Moore SW (2000) Cholinesterase-like catalytic antibodies: reaction with sub-
strates and inhibitors. Mol Immunol 37:707–719.
Shi D, Yang J, Yang D, LeCluyse EL, Black C, You L, Akhlaghi F, and Yan B (2006) Anti-
influenza prodrug oseltamivir is activated by carboxylesterase human carboxylesterase 1, and
the activation is inhibited by antiplatelet agent clopidogrel. J Pharmacol Exp Ther 319:
1477–1484.
Shuldiner AR, O’Connell JR, Bliden KP, Gandhi A, Ryan K, Horenstein RB, Damcott CM,
Pakyz R, Tantry US, and Gibson Q, et al. (2009) Association of cytochrome P450 2C19
genotype with the antiplatelet effect and clinical efficacy of clopidogrel therapy. JAMA 302:
849–857.
Steinhubl SR, Tan WA, Foody JM, and Topol EJ (1999) Incidence and clinical course of
thrombotic thrombocytopenic purpura due to ticlopidine following coronary stenting. EPIS-
TENT Investigators. Evaluation of platelet IIb/IIIa inhibitor for stenting. JAMA 281:806–810.
Takahashi S, Katoh M, Saitoh T, Nakajima M, and Yokoi T (2009) Different inhibitory effects in
rat and human carboxylesterases. Drug Metab Dispos 37:956–961.
Talakad JC, Shah MB, Walker GS, Xiang C, Halpert JR, and Dalvie D (2011) Comparison of in
vitro metabolism of ticlopidine by human cytochrome P450 2B6 and rabbit cytochrome P450
2B4. Drug Metab Dispos 39:539–550.
Kazui M, Nishiya Y, Ishizuka T, Hagihara K, Farid NA, Okazaki O, Ikeda T, and Kurihara A
(2010) Identification of the human cytochrome P450 enzymes involved in the two oxidative
steps in the bioactivation of clopidogrel to its pharmacologically active metabolite. Drug Metab
Dispos 38:92–99.
Lesesve JF, Callat MP, Lenormand B, Monconduit M, Noblet C, Moore N, Caron F, Humbert G,
Stamatoullas A, and Tilly H (1994) Hematological toxicity of ticlopidine. Am J Hematol 47:
149–150.
Love BB, Biller J, and Gent M (1998) Adverse haematological effects of ticlopidine. Prevention,
recognition and management. Drug Saf 19:89–98.
Mackness B, Mackness MI, Arrol S, Turkie W, Julier K, Abuasha B, Miller JE, Boulton AJ,
and Durrington PN (1998) Serum paraoxonase (PON1) 55 and 192 polymorphism and para-
oxonase activity and concentration in non-insulin dependent diabetes mellitus. Atherosclerosis
139:341–349.
Maeda A, Ando H, Asai T, Ishiguro H, Umemoto N, Ohta M, Morishima M, Sumida A,
Kobayashi T, and Hosohata K, et al. (2011) Differential impacts of CYP2C19 gene poly-
morphisms on the antiplatelet effects of clopidogrel and ticlopidine. Clin Pharmacol Ther 89:
229–233.
Tougou K, Nakamura A, Watanabe S, Okuyama Y, and Morino A (1998) Paraoxonase has
a major role in the hydrolysis of prulifloxacin (NM441), a prodrug of a new antibacterial agent.
Drug Metab Dispos 26:355–359.
Mataix R, Ojeda E, Perez MC, and Jimenez S (1992) Ticlopidine and severe aplastic anaemia. Br
J Haematol 80:125–126.
Tuffal G, Roy S, Lavisse M, Brasseur D, Schofield J, Delesque Touchard N, Savi P, Bremond N,
Rouchon MC, and Hurbin F, et al. (2011) An improved method for specific and quantitative
determination of the clopidogrel active metabolite isomers in human plasma. Thromb Haemost
105:696–705.
Williams FM (1985) Clinical significance of esterases in man. Clin Pharmacokinet 10:392–403.
Yin T and Miyata T (2011) Pharmacogenomics of clopidogrel: evidence and perspectives.
Thromb Res 128:307–316.
Nishiya Y, Hagihara K, Kurihara A, Okudaira N, Farid NA, Okazaki O, and Ikeda T (2009a)
Comparison of mechanism-based inhibition of human cytochrome P450 2C19 by ticlopidine,
clopidogrel, and prasugrel. Xenobiotica 39:836–843.
Nishiya Y, Hagihara K, Ito T, Tajima M, Miura S, Kurihara A, Farid NA, and Ikeda T (2009b)
Mechanism-based inhibition of human cytochrome P450 2B6 by ticlopidine, clopidogrel, and
the thiolactone metabolite of prasugrel. Drug Metab Dispos 37:589–593.
Noble S and Goa KL (1996) Ticlopidine. A review of its pharmacology, clinical efficacy and
tolerability in the prevention of cerebral ischaemia and stroke. Drugs Aging 8:214–232.
Ono K, Kurohara K, Yoshihara M, Shimamoto Y, and Yamaguchi M (1991) Agranulocytosis
caused by ticlopidine and its mechanism. Am J Hematol 37:239–242.
Address correspondence to: Dr. Dong-Hyun Kim, Department of Pharmacology
and PharmacoGenomics Research Center, Inje University College of Medicine,
#633-165 Gaegum-Dong, Jin-Gu, Busan 614-735, South Korea. E-mail: dhkim@
inje.ac.kr
Preuss CV and Svensson CK (1996) Arylacetamide deacetylase activity towards mono-
acetyldapsone. Species comparison, factors that influence activity, and comparison with 2-
acetylaminofluorene and p-nitrophenyl acetate hydrolysis. Biochem Pharmacol 51:1661–1668.