In vitro metabolic fate of alizapride: Evidence for the formation of reactive metabolites based on liquid chromatography-tandem mass spectrometry

Article abstract of DOI:10.1002/jms.3011

The study of the formation of reactive metabolites during drug metabolism is one of the major areas of research in drug development since the link between reactive metabolites and drug adverse effects was well recognized. In particular, it has been shown that acrolein, a reactive carbonyl species sharing carbonylating and alkylating properties, binds covalently to nucleophilic sites in proteins, causing cellular damage. Alizapride, (±)-6-methoxy-N-{[1- (prop-2-en-1-yl)-pyrrolidin-2-yl] methyl}-1H-benzotriazole-5-carboxamide, is a N-allyl containing dopamine antagonist with antiemetic properties for which no data concerning its metabolic fate are so far reported. The study of the in vitro metabolism of alizapride showed the formation of acrolein during the oxidative N-deallylation. Moreover, the formation of an epoxide metabolite has been also described suggesting its role as a putative structural alert. The reactivity of the acrolein and the epoxide generated in alizapride metabolism was demonstrated by the formation of the corresponding adducts with nucleophilic thiols. Overall, ten metabolites have been identified and characterized by electrospray ionization tandem mass spectrometry analysis allowing to propose an in vitro metabolic scheme for alizapride. At the best of our knowledge, this is the second case of a drug involved in the generation of acrolein during its metabolism being the first represented by cyclophosphamide. Copyright

Full text of DOI:10.1002/jms.3011

Research Article
Received: 6 December 2011
Revised: 6 April 2012
Accepted: 6 April 2012
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI 10.1002/jms.3011
In vitro metabolic fate of alizapride: evidence
for the formation of reactive metabolites based
on liquid chromatography-tandem mass
spectrometry
Silvio Aprile,* Erika Del Grosso and Giorgio Grosa
The study of the formation of reactive metabolites during drug metabolism is one of the major areas of research in drug
development since the link between reactive metabolites and drug adverse effects was well recognized. In particular, it has
been shown that acrolein, a reactive carbonyl species sharing carbonylating and alkylating properties, binds covalently to
nucleophilic sites in proteins, causing cellular damage. Alizapride, (Æ)-6-methoxy-N-{[1-(prop-2-en-1-yl)-pyrrolidin-2-yl]
methyl}-1H-benzotriazole-5-carboxamide, is a N-allyl containing dopamine antagonist with antiemetic properties for which
no data concerning its metabolic fate are so far reported. The study of the in vitro metabolism of alizapride showed the
formation of acrolein during the oxidative N-deallylation. Moreover, the formation of an epoxide metabolite has been also
described suggesting its role as a putative structural alert. The reactivity of the acrolein and the epoxide generated in
alizapride metabolism was demonstrated by the formation of the corresponding adducts with nucleophilic thiols. Overall,
ten metabolites have been identified and characterized by electrospray ionization tandem mass spectrometry analysis
allowing to propose an in vitro metabolic scheme for alizapride. At the best of our knowledge, this is the second case of a drug
involved in the generation of acrolein during its metabolism being the first represented by cyclophosphamide. Copyright ©
2012 John Wiley & Sons, Ltd.
Supporting information may be found in the online version of this article.
Keywords: alizapride; acrolein; reactive metabolites; microsomes; LC-MS
that acrolein, which shares carbonylating and alkylating pro-
perties, was generated through the CYP-mediated oxidation of
INTRODUCTION
[
7,8]
cyclophosphamide,
being also responsible for the serious
Investigation into the role of bioactivation in drug-induced toxicity
is one of the major areas of research since the link between reactive
metabolites and drug adverse effects was well recognized. There
have been a lot of reports clearly demonstrating that a drug can
be ‘metabolic activated’ to toxic metabolites. For this reason, in
the drug development, several procedures aimed at identifying
structural moieties that undergo metabolic activation to electro-
philic and reactive species are employed. Typically, these efforts
involve a combination of in vitro and in vivo studies in which
reactive metabolites are identified indirectly through structural
characterization of the stable ‘end products’ that are formed upon
bladder toxicity associated with this drug treatment. Differently,
it has not so far been reported the formation of acrolein during
the metabolic biotransformation of drugs containing the N-allyl
[
9,10]
[
1]
moiety such as naloxone and levallorphan.
Alizapride (AL) (Æ)-6-methoxy-N-{[1-(prop-2-en-1-yl)-pyrroli-
din-2-yl]methyl}-1H-benzotriazole-5-carboxamide (Table 1) is a
dopamine antagonist with prokinetic and antiemetic effects
administered as a racemic mixture in the treatment of cancer
chemotherapy-induced nausea and vomiting, including postop-
[
11]
erative nausea and vomiting.
Although pharmacodynamic
[
2]
studies of AL in mice and rats demonstrated little toxicity
capture by nucleophilic scavengers.
[
12]
(
particularly after parenteral administration),
only few data
Liquid chromatography coupled to mass spectrometry (LC-MS)
technique is increasingly becoming the analytical tool of
choice in drug metabolism studies. Since the introduction of
electrospray ionization (ESI) and atmospheric pressure chemical
ionization, LC-MS has been the preeminent analytical tool used in
small molecule drug metabolism investigation. This is primarily
due to the high sensitivity and selectivity of this combined tech-
are available about AL toxicity in man. It is well known that AL
is mainly excreted as unmodified drug after oral or parenteral
administration; however, a share of about 25% undergoes
metabolic transformation before deletion. While the pharmacokinetics
[
3,4]
nique and its ability to separate, detect and identify metabolites.
* Correspondence to: Silvio Aprile, Dipartimento di Scienze del Farmaco, Università
degli Studi del Piemonte Orientale ‘A. Avogadro’, Largo Donegani 2, 28100
Novara, Italy. E-mail: silvio.aprile@pharm.unipmn.it
Reactive carbonyl species consisting in a,b-unsaturated
aldehydes (e.g. acrolein), bind covalently to nucleophilic sites in
proteins, particularly to the free amino group of lysine residues,
Dipartimento di Scienze del Farmaco, Università degli Studi del Piemonte
Orientale ‘A. Avogadro’, Largo Donegani 2, 28100 Novara, Italy
[
5,6]
causing cellular damage.
It has been reported in the literature
J. Mass. Spectrom. 2012, 47, 737–750
Copyright © 2012 John Wiley & Sons, Ltd.

S. Aprile, E. Del Grosso and G. Grosa
n
Table 1. Proposed structures of the MS fragment ions of standard alizapride (AL) (precursor ion m/z 316) and reference compounds 3, 4–5 and 6
precursor ions m/z 276, 332 and 302, respectively)
(
2
3
Compound
Proposed structures for MS and MS fragment ions
wileyonlinelibrary.com/journal/jms
Copyright © 2012 John Wiley & Sons, Ltd.
J. Mass. Spectrom. 2012, 47, 737–750

Alizapride metabolism
[
13]
of AL has been studied, no data are available in literature about
its metabolic fate; indeed, the structures of the metabolites and the
related metabolic pathways are so far unknown. Since the structure
of AL shows a N-allyl moiety which potentially could undergo
N-deallylation and/or epoxidation reactions, the formation of chem-
ically reactive metabolites can be expected. In particular acrolein
Units. Data were acquired in full-scan and multi stage product
ion scan (MS ) modes using mass scan range m/z 105–600.
The collision energy was optimized at 32%. When the UV
a
detector was used, the analytical wavelength (l ) was 225 nm.
n
1
13
(
Ac) could arise from N-deallylation while a reactive epoxide could
H / C NMR
result from the N-allyl moiety epoxidation.
1
13
H and C attached proton test experiments were performed on
a JEOL ECP 300 FT MHz spectrophotometer (JEOL, Tokyo, Japan).
Chemical shifts are reported in parts per million (ppm).
These considerations prompted us to study the in vitro metab-
olism of AL using rat and human liver subcellular fractions. First,
the formation of Ac from N-deallylation was demonstrated.
Second, the formation of an epoxide metabolite has been also
described suggesting its role as a putative structural alert. Finally,
ten AL metabolites have been identified in the incubation
mixtures, and their structures have been elucidated by compari-
son of their mass spectral data with those of the available
synthetic reference compounds.
Human intestinal and liver microsomes
The cryopreserved Pooled Human Intestinal Microsomes (HIM),
(
0
(
2
pooled mixed sex, protein concentration: 10 mg/ml, total P450:
.26 nmol/mg) and Pooled Human Liver Microsomes (HLM),
pooled mixed sex, ten individual donors, protein concentration:
0 mg/ml, total P450: 0.36 nmol/mg) were purchased from BD
MATERIAL AND METHODS
™
Gentest (Woburn, MA) and used throughout this study.
Reagents and chemicals
Acrolein, methanol (HPLC grade), (S)-(+)-2-(aminomethyl)pyrrolidine,
semicarbazide hydrochloride, anhydrous sodium acetate and thionyl
chloride were purchased from Sigma–Aldrich (Milano, Italy). Water
Rat liver microsomes
Male Wistar rat liver microsomes (RLM), (protein concentration:
25 mg/ml, total P450: 0.64 nmol/mg protein) were used throughout
(
HPLC grade) was obtained from ELGA PURELAB ultra (M-medical,
Milano, Italy). Column chromatography was performed on silica gel
Merck Kieselgel 70–230 mesh ASTM) using the indicated eluants.
[
16]
this study and prepared using the previously described protocol.
The rat liver microsomal P450 CYP concentration was determined
(
[17]
using the Omura and Sato method.
carried out using a horizontal DUBNOFF shaking thermostatic bath
Dese Lab Research, Padova, Italy) and were protected from light.
The incubations were all
Thin layer chromatography (TLC) was carried out on plates with a
layer thickness of 0.25 mm (Merck silica gel 60 F254). Alizapride
hydrochloride (AL) was obtained as gift sample from Pharmafar
S.r.l. (Torino, Italy); 6-methoxy-1H-benzotriazole-5-carboxylic acid 1
and 6-methoxy-N-{[1-oxy-1-(prop-2-en-1-yl)pyrrolidin-2-yl]methyl}-1H-
benzotriazole-5-carboxamide(diastereomersA4andB5)weresyn-
(
In vitro alizapride incubations
[
14]
thesized according to the procedure cited in the literature. 6-
The standard incubation mixture (500-ml final volume), in 1.5-ml
eppendorf tubes, was carried out in a 50-mM TRIS buffer (pH 7.4)
containing 3.3-mM MgCl , 1.3-mM NADPNa , 3.3-mM glucose
Hydroxy-N-{[1-(prop-2-en-1-yl)pyrrolidin-2-yl]methyl}-1H-benzotriazole-
[
15]
5
-carboxamide 6 was synthesized as previously reported.
2
2
6-phosphate, 0.4-Units/ml glucose 6-phosphate dehydrogenase
LC-MS
and 1mM AL. After pre-equilibration of the mixture, an appropriate
volume of microsomal suspension was added to give a final protein
A Thermo Finningan LCQ Deca XP plus ion trap mass spectrometer
equipped with a quaternary pump, a Surveyor AS autosampler, a
Surveyor photodiode array detector and a vacuum degasser was
used for LC-MS analysis (Thermo Electron Corporation, Waltham,
MA). All the chromatographic separations were performed on a
Polaris C18-A (150 Â 4.6 mm i.d., 5-mm particle size; Varian, USA)
column with a C18 SecurityGuard column (Phenomenex srl, Castel
Bolognese, Italy). Aliquots (20 ml) of supernatants obtained from
incubations were injected onto the HPLC system and eluted with
a mobile phase (flow rate 0.7 ml/min) consisting of a 20-mM ammo-
nium formate buffer, pH 4.0 (solvent A) and methanol (solvent B).
The following gradient elution was used: 0 to 15 min B = 14–33%,
concentration of 1 mg/ml. The mixture was shaken for 60 min at
ꢀ
3
7 C, and the reaction was quenched by diluting the samples with
150 ml of ice-cold formic acid solution (25% v/v). The samples were
then centrifuged at 11,000 rpm for 5 min, and the supernatant was
directly injected onto the column for analysis. Control incubations
without the presence of substrate, NADP(H)-regenerating system,
or microsomes were also carried out.
Incubation of alizapride in the presence of derivatizing
agents
1
5 to 15.5 min B = 33–14%, 15.5 to 20 min B = 14%. The eluants
Reactivity of AL toward nucleophilic thiols was studied by adding
2-mercaptoethanol (3 mM) to the incubation mixture. In
cubations were also carried out in the presence of the aldehyde
trapping agent semicarbazide hydrochloride (3 mM). In the case
of the amino acid metabolite, 500 ml of crude mixture was ester-
ified by treating, after incubation, with 200 ml of 10% sulphuric
acid in ethanol. The mixture was allowed to stand at room
temperature for 1 h. After neutralization with 10 N NaOH aqueous
solution, the mixture was centrifuged to remove insoluble matter,
and a 20-ml aliquot was injected onto the HPLC column.
were filtered through a 0.45-mm pore size polyvinylidene
difluoride membrane filter before use. The eluate was injected
into the electrospray ion source (ESI) with a splitting of 40%.
MS spectra were acquired and processed using Xcalibur
software. The operating conditions on the ion trap mass
spectrometer were as follows: spray voltage, 5.30 kV; source
current, 80 mA; capillary temperature, 350 C; capillary voltage,
1
multipole 2 offset, À12.00 V; sheath gas flow (N ), 40 Arbitrary
W
ꢀ
5.00 V; tube lens offset, 10.00 V; multipole 1 offset, À4.00 V;
2
J. Mass. Spectrom. 2012, 47, 737–750
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jms

S. Aprile, E. Del Grosso and G. Grosa
Urine sample collection
ESI-MS Calculated for C H N O , M = 207.19; Found m/z
M + H] = 208
9
9 3 3
+
[
Urine was collected after (6 h) a single-dose administration of a
W
5
4
0-mg Limican ampoules (Pharmafar S.r.l. Torino, Italy) to a
7-year-old female patient who was prescribed alizapride hydro-
6
-Methoxy-N-(pyrrolidin-2-ylmethyl)-1H-benzotriazole-5-carboxa-
mide 3
chloride by the patient’s physician. The patient received a full
explanation of the procedure and gave written consent to partic-
ipate in the excretion study. Urine blank was collected after a
(
S)-(+)-2-(Aminomethyl)pyrrolidine (290 mg, 2.90 mol) and
compound 2 (120 mg, 0.58 mmol) were dissolved in methanol
(
24 h. After solvent evaporation under reduced pressure, the residue
was purified by column chromatography using dichloromethane/
methanol/glacial acetic acid 80:20:2 and then 75:25:2 as eluants to
give pure compound 3 (104 mg, 65% yield).
ꢀ
3 ml), and the resulting reaction mixture was stirred at 70 C for
1
-week wash-out period. The samples were stored in a freezer
ꢀ
at À80 C before analysis.
Synthesis of authentic reference compounds
Acrolein semicarbazone (Ac-S) and 6-methoxy-N-(pyrrolidin-2-
ylmethyl)-1H-benzotriazole-5-carboxamide 3 were synthesized
as shown in Scheme 1. A brief description of the synthesis proce-
dure follows below.
ESI-MS Calculated for C H N O , M = 275.31; Found m/z
13 17 5 2
+
[
M + H] = 276
1
H-NMR: (300 MHz, D O) d (ppm): 8.1 (s, H4); 7.0 (s, H7); 3.85
2
(
s, CH ); 3.7- 3.72 (m, H2’); 3.32-3.41 (m, 2H5’ + 2H6’); 1.9-2.08
3
(
m, 2H3’ + 2H4’) (supplemental 1).
1
3
Acrolein semicarbazone (Ac-S)
C-NMR: (300 MHz, D O) d (ppm): 169.7 (C8), 157.6 (C6), 137.9
2
(
(
C7a), 137.2 (C3a), 122.8 (C5), 120.7 (C4), 93.8 (C7), 61.3 (C2’), 57.3
C10), 49.9 (C5’), 41.8 (C6’), 28.3 (C3’), 23.6 (C4’) (supplemental 1).
UV–vis: (CH OH, conc. 10 mg/ml) l : 220, 295 nm.
Anhydrous sodium acetate (732 mg, 8.90 mmol) and semicarbazide
hydrochloride (1.1 g, 9.82 mmol) were dissolved in methanol (23 ml).
Acrolein (592 ml, 8.90 mmol) was added, and the resulting reaction
mixture was stirred at r.t. for 2 h. After solvent evaporation
under reduced pressure, the residue was reconstituted in 50 ml of
water and extracted with a mixture of chloroform/ethyl acetate/
3
max
-
1
FT-IR: (CH OH): 1650, 1630, 1450, 660 cm .
3
RESULTS AND DISCUSSION
2
-propanol 45:45:10 (50 ml  3). The pooled organic layers were
dried over anhydrous sodium sulphate, filtered and then evaporated
under reduced pressure. The residue was crystallized from ethyl
acetate to give acrolein semicarbazone Ac-S (150 mg, 15% yield).
Fragmentation behavior of reference compounds
Since in the literature, the mass spectral data of AL were not avail-
able, ESI-MS analysis of AL and the available synthetic compounds
n
ESI-MS Calculated for
M + H] = 114
4 7 3
C H N O, M = 113.12; Found m/z
+
(3, 4–5, 6) were performed. Indeed, the study of their collision-
induced dissociation behavior was very useful to establish if the
metabolic transformation occurred on the 1H-benzotriazole ring
or at the level of N-allyl-2-methylene pyrrolidine region. Direct
infusions of AL, compound 3, diastereomers 4–5 and compound
6 were carried out generating the protonated molecules [M+ H]
at m/z 316, 276, 332 and 302 respectively. Hence, the product ion
spectra of these ions (Fig. 1) and the proposed MS fragmentation
pathways (Table 1) are shown and discussed. AL product ion
spectrum showed the presence of several diagnostic ions at:
m/z 288 (loss of N2 from the 1H-benzotriazole ring), m/z 274
(N-deallylation), m/z 176 and 141 (amide cleavage), m/z 167 (neutral
loss of the 1H-benzotriazole ring), m/z 148 (loss of the neutral N-{[1-
[
6
-Methoxy-1H-benzotriazole-5-carboxylic acid methyl ester 2
Thionyl chloride (900 ml, 12.4 mmol) was dissolved in ice-cold
methanol (20 ml), and then 6-methoxy-1H-benzotriazole-5-
carboxylic acid 1 (430 mg, 2.23 mmol) was added to the mixture.
The mixture was heated to reflux and stirred for 4 h. The solution
was then cooled to r.t., and the solvent evaporated under
reduced pressure to give a yellow pale residue. The crude
product was purified by silica column chromatography, using
dichloromethane/methanol 9:1 mixture as eluant to give pure
compound 2 (439 mg, 95% yield).
+
n
(
(
prop-2-en-1-yl)pyrrolidin-2-yl]methyl}formamide) and m/z 124
base peak, loss of the neutral 6-methoxy-1H-benzotriazole-5-
carboxamide). A similar fragmentation pattern was observed for
compound 6 where the 6-methoxy group on the 1H-benzotriazole
ring is replaced by a hydroxyl group. The product ion spectrum of
ion at m/z 276 (compound 3) gave only two fragment ions: the first
at m/z 259 (base peak) generated by the neutral loss of NH
3
(which
3
at the MS stage further lost 28 Da (CH = CH ), m/z 231), and the
2 2
second at m/z 176 (amide cleavage). Finally, the precursor ion
at m/z 332 of the alizapride N-oxide diastereomers (compounds
4
and 5) gave the fragment ions at m/z 176 (base peak, amide
cleavage), m/z 291 (consistent with the loss of the allyl radical),
m/z 274 (neutral loss of C O) and m/z 259 (neutral loss of
NHOH-CH -CH = CH ).
3 6
H
2
2
Alizapride metabolite identification by LC-MS analysis
Alizapride incubations carried out in the presence of microsomes
and NADP(H)-regenerating system were analyzed in both positive
Scheme 1. Synthesis of the compounds Ac-S and M1 (3).
wileyonlinelibrary.com/journal/jms
Copyright © 2012 John Wiley & Sons, Ltd.
J. Mass. Spectrom. 2012, 47, 737–750

Alizapride metabolism
2
Figure 1. Mass spectra from MS analysis of standard alizapride (AL) (precursor ion m/z 316) (a) and reference compounds 3, 4–5 and 6 (precursor ions
m/z 276, 332 and 302, respectively) (b, c, d).
n
full-scan and product ion scan MS modes, and the obtained data
sets compared with that of the control incubations. Indeed, for
the discrimination of metabolic versus non-metabolic transforma-
tions, different blank samples were prepared: a substrate blank
reported in Supplemental 6). Moreover, the formation of M1,
through the oxidative N-deallylation of AL, was also confirmed by
trapping acrolein in RLM incubation mixture with semicarbazide
hydrochloride. As expected, the peak of acrolein semicarbazone
adduct (Ac-S) was observed in product ion scan (m/z 114) chro-
matogram (Fig. 4b) of RLM incubation while it did not in control in-
cubation performed in the absence of semicarbazide hydrochloride
(
everything except the substrate), a cofactor blank (everything
but not NADP(H)-regenerating system) and an enzyme blank.
Once a putative metabolite ion was found in full-scan analysis of
n
2
the incubation mixtures, further MS acquisition of the detected
(Fig. 4a). On the contrary, the absence of the Ac-S peak in LC-MS
ions followed. The metabolites were characterized based on their
predicted mass shift from the alizapride protonated molecule
chromatograms of HLM and HIM incubations could be attributed
to their lower enzymatic activities compared to that of RLM. With
the aim to assess the in vivo formation of M1, a human urine
sample collected after a single dose administration of alizapride
hydrochloride, was analyzed by LC-MS monitoring the ions at
m/z 316 (AL) and 276 (M1). The chromatograms depicted in
Supplemental 4 showed the presence of both the peaks referable
to AL and its metabolite M1. Taken together, these features
suggested that oxidative N-deallylation could be the main in vitro
metabolic pathway of AL leading to the formation of acrolein, a
known reactive and toxic metabolite. Even if the presence of M1
was also observed in human urine, the very limited experimental
features employed did not allow to delineate the in vivo metabolic
fate of AL. To the best of our knowledge, this is the first demonstra-
tion of the acrolein formation from the metabolism of a drug
containing a N-allyl moiety. Indeed, the studies on the metabolic
fate of alclofenac, an anti-inflammatory agent bearing in its struc-
ture an O-allyl moiety, did not allow to establish the formation of
acrolein during the metabolism. On the contrary, the formation of
a reactive alclofenac epoxide arising from the oxidation of allylic
+
n
([M + H] ; m/z 316) and the interpretation of their MS spectra.
As a matter of fact, the LC-MS analysis of AL incubations,
performed in the presence of RLM and HLM, revealed the pres-
ence of several metabolite-related ions which were summarized
2
in Table 2. The LC-MS chromatograms obtained from the analysis
of HLM and RLM incubations were reported in Figs. 2 and 3,
2
respectively. Finally, LC-MS analysis of HIM incubations gave
the metabolite-related ions at m/z 276 (M1), 302 (M2), 314 (M3)
and 332 (M4) (Supplemental 2).
Metabolite M1 and acrolein
Metabolite M1 ion (m/z 276) was detected in LC-MS chromato-
2
grams of HLM, HIM and RLM incubations being its MS spectrum
(Supplemental 3) identical to that of the compound 3 (Fig. 1). This
metabolite was one of the most abundant in all incubation carried
out, especially with RLM where M1 peak area contributed for
1
4% after metabolic biotransformation (see LC-UV chromatograms
J. Mass. Spectrom. 2012, 47, 737–750
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jms

S. Aprile, E. Del Grosso and G. Grosa
n
Table 2. Alizapride metabolites identified from different microsomes source and structurally characterized by mass spectrometry LC-MS analysis
+
Metabolite
Protein source
[M + H]
276
Mass shift (Da)
Metabolic reaction
M1
RLM
HLM
HIM
RLM
HLM
HIM
RLM
HLM
HIM
RLM
RLM
RLM
HLM
HIM
RLM
HLM
HIM
RLM
RLM
HLM
RLM
HLM
À40
N-Deallylation with formation of acrolein
M2
M3
302
314
À14
À2
O-Demethylation
Desaturation of the N-allyl pyrrolidine
M7
M8
M4
314
330
332
À2
+14
+16
Desaturation of the N-allyl pyrrolidine
N-allyl oxidation to a,b-unsaturated amide
Hydroxylation of the 1H-benzotriazole ring
M9
332
+16
N-allyl double bond oxidation to epoxide
M10a, M10b
332
348
+16
+32
N-allyl pyrrolidine N-oxide formation
M5
N-allyl pyrrolidine oxidation to g-aminobutyric acid derivative
M6a, M6b
350
+34
vic-Diol formation on the N-allyl chain
2
Figure 2. LC/MS chromatograms of AL incubation (HLM) performed in positive product ion scan mode (MS ): m/z 316 (a), m/z 276 (b), m/z 302 (c),
m/z 314 (d), m/z 332 (e), m/z 348 (f) and m/z 350 (g).
wileyonlinelibrary.com/journal/jms
Copyright © 2012 John Wiley & Sons, Ltd.
J. Mass. Spectrom. 2012, 47, 737–750

Alizapride metabolism
2
Figure 3. LC/MS chromatograms of AL incubation (RLM) performed in positive product ion scan mode (MS ): m/z 316 (a), m/z 276 (b), m/z 302 (c), m/z
14 (d), m/z 330 (e), m/z 332 (f), m/z 348 (g) and m/z 350 (h).
3
[
18]
double bond was evidenced.
Noticeably, alclofenac was with-
double bond on pyrrolidine ring; indeed, the same fragmentation
pathway of AL gave the ion at m/z 124. Unlike the metabolite M3,
metabolite M7 product ion spectrum showed the presence of the
drawn from the market due to the emergence of an array of
adverse effects (skin rash, cutaneous vasculitis) probably related
to the formation of its metabolites.
[
19,20]
+
most intense fragment ion at m/z 296 [M + H-18] which is in
agreement with the loss of a 18 Da (H
presence of this diagnostic ion, we suggested the structure of
-methoxy-N-{[1-(prop-2-en-1-yl)-2,3-dihydro-1H-pyrrol-2-yl]
methyl}-1H-benzotriazole-5-carboxamide for this metabolite; on
the contrary, for metabolite M3, we were unable to define the
exact position of the unsaturation by the MS-based approach.
2
O), Fig. 5. Based on the
Metabolite M2
Metabolite M2 was identified by mass spectrometry to have a
6
+
[M+ H] ion at m/z 302 representing a loss of 14 Da from AL and
showing mass spectral (Supplemental 3) and chromatographic
properties identical to those of the synthetic compound 6 (Fig. 1).
Hence the structure of 6-hydroxy-N-{[1-(prop-2-en-1-yl)pyrrolidin-
Metabolite M8
2-yl]methyl}-1H-benzotriazole-5-carboxamide was assigned.
2
The LC-MS analysis at m/z 330 of RLM incubations, revealed the
2
presence of a metabolite (M8) eluting at 6.25 min (Fig. 3); the MS
Metabolite M3 and M7
+
spectrum of protonated molecule [M + H] (Fig. 6) produced four
2
When LC-MS analysis was carried out monitoring the ion at
main fragment ions at: m/z 312, 257, 176 and 138. In particular,
both ion at m/z 176, (arising from amide cleavage) and m/z 138
(corresponding to the ion at m/z 124 of AL + 14 Da) indicated
that the oxidative biotransformation occurred on the N-allyl-2-
methylen pyrrolidine substructure. Thus, the structure of the
a,b-unsaturated amide, N-[(1-acryloylpyrrolidin-2-yl)methyl]-6-
methoxy-1H-benzotriazole-5-carboxamide, was hypothesized
for metabolite M8. In order to support our hypothesis, we
exploited the reactivity of the a,b-unsaturated carbonyl moiety
towards nucleophiles, monitoring (at m/z 408) the adduct
formation in RLM incubation performed in the presence of
m/z 314, the presence of two peaks M3 (t
tR = 9.30 min) was observed only in RLM incubations (Fig. 3),
whereas only M3 was detected in HLM and HIM incubations
R
= 7.01 min), and M7
(
2
(
Fig. 2 and Supplemental 2) being their MS spectra reported in
+
Fig. 5. For both metabolites, the [M + H] ion at m/z 314 (2 Da less
than AL) indicated that the metabolic oxidative dehydrogenation
occurred, whereas the presence of the fragment ions at m/z 176
and 148 clearly demonstrated that the enzymatic transformation
occurred on the N-allyl-2-methylen pyrrolidine region. Further-
more, the fragment ion at m/z 122 suggested the presence of a
J. Mass. Spectrom. 2012, 47, 737–750
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jms

S. Aprile, E. Del Grosso and G. Grosa
2
Figure 4. Positive LC/MS chromatograms of AL incubation (RLM) monitoring the ion at m/z 114: incubation without the presence of semicarbazide
hydrochloride (a), incubation with the presence of semicarbazide hydrochloride (b). LC/MS analysis (m/z 114) of the synthetic acrolein semicarbazone
2
(
Ac-S) (c) and its mass spectra in full-scan (d) and MS (e) modes.
2
-mercaptoethanol. As expected, the presence of a new peak
oxidation occurred at C-4 or C-7 of the benzotriazole ring; obvi-
ously, the mass spectral data did not allow to indicate the exact
position of the hydroxyl function. The assignment of the structure
of the metabolite M4 was also supported by the relevant
literature; indeed, aromatic hydroxylation on benzene ring in
M8-d (Fig. 7) eluting at 10.62 min and the complete disappear-
ance of M8 were observed in rat liver microsomal incubation
analysis confirming the formation of the Michael acceptor
a,b unsaturated amide M8.
[
21]
1
H-benzotriazole carboxylic acids was reported by Hoffmann
[
22]
and co-workers.
Metabolite M4
The structure of metabolite M4 was also postulated based on the
MS and MS data. In particular, its product ion spectrum (Fig. 8)
Metabolites M9, M10a and M10b
2
3
2
showed the presence of two ions at m/z 304 (N neutral loss) and
LC-MS analysis of the RLM incubation revealed the presence of
2
2
90 (CH -CH = CH neutral loss). Further fragmentation of the
three additional metabolites (at m/z 332) (M9, M10a and
M10b) which were not found in HLM and HIM incubations (Fig. 3):
the protonated molecules suggested the insertion of an oxygen
3
2
3
fragment ion at m/z 304 at the MS stage gave rise to the diag-
nostic ion at m/z 262 (CH -CH = CH neutral loss) and the ions
at m/z 167, 141 and 124 already present in the AL fragmentation
3
2
2
atom on AL structure. The MS spectrum of metabolite M9 at
3
+
pathways. In addition, the product ion at m/z 290 gave at the MS
m/z 332 (Fig. 9) produced the ion at m/z 314 [M + H-18] (base
+
stage the fragment ion at m/z 192 arising from the cleavage of
amide function; since the same fragmentation pathway of AL
gave a fragment ion at m/z 176, we concluded that the enzymatic
peak), and the diagnostic at: m/z 288 [M + H-C H O] , m/z 259
2
4
+
[M + H-C H NO] and m/z 176, (amide cleavage) all indicating
3
7
the presence of an oxygen atom on the N-allyl chain. Based on
wileyonlinelibrary.com/journal/jms
Copyright © 2012 John Wiley & Sons, Ltd. J. Mass. Spectrom. 2012, 47, 737–750

Alizapride metabolism
2
n
Figure 5. Positive MS (m/z 314) spectra and the proposed structures of the MS fragment ions of the metabolites M3 (a) and M7 (b).
2
n
Figure 6. Positive MS (m/z 330) spectrum and the proposed structures of the MS fragment ions of the metabolite M8.
the observed fragmentation pathway for metabolite M9, the
epoxide structure was proposed for this metabolite. In order to
confirm our hypothesis, incubations using RLM, HLM and HIM
were also carried out in the presence of the nucleophilic reagent
Furthermore, the partial disappearance of M9 peak was observed
in LC-MS chromatogram at m/z 332 in RLM incubation in the
presence of 2-mercaptoethanol.
2
Other minor metabolites, M10a and M10b, (Fig. 3) were
observed and characterized only in RLM incubation analysis:
2
-mercaptoethanol, and the formation of the putative conjugate
was monitored at m/z 410. As a result, the presence of a peak
M9-d) (Fig. 10) was demonstrated in all microsomal incubation
2
indeed, they show the same tR and MS spectra (Fig. 1 and
(
Supplemental 5) of the synthetic alizapride N-oxide (mixture of
n
analysis confirming the formation of the epoxide metabolite.
diastereomers; compounds 4 and 5) whose MS fragmentation
J. Mass. Spectrom. 2012, 47, 737–750
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jms

S. Aprile, E. Del Grosso and G. Grosa
2
Figure 7. Positive LC/MS chromatograms of AL incubation (RLM) monitoring the ion at m/z 408 (M8-d): incubation without the presence of 2-
2
mercaptoethanol (a), incubation with the presence of 2-mercaptoethanol (b). Positive MS (m/z 408) spectrum (c) and the proposed structures of the
n
MS fragment ions of the adduct M8-d.
2
n
Figure 8. Positive MS (m/z 332) spectrum and the proposed structures of the MS fragment ions of the metabolite M4.
+
pathways were depicted in Fig. 1 and Table 1. On the contrary,
the formation of alizapride N-oxide has not been observed in
HLM and HIM incubation analysis.
a protonated molecule [M + H] at m/z 348 which is 32 Da higher
than that of AL; these features suggested the insertion of two
oxygen atoms into the parent compound. The MS spectrum of
2
M5 protonated molecule (Fig. 11) furnished several diagnostic
+
ions at: m/z 330 [M + H-18] (base peak), m/z 291 (neutral loss
of NH CH -CH = CH ) and m/z 176 (amide cleavage) the latter
2 2 2
Metabolites M5
When both HLM and RLM incubations were analyzed, a peak
indicating that benzotriazole ring was not affected by metabolism.
Moreover, the treatment of HLM incubation mixture with 10%
(Figs. 2, 3) eluting at 8.67 min (M5) was observed giving rise to
wileyonlinelibrary.com/journal/jms
Copyright © 2012 John Wiley & Sons, Ltd.
J. Mass. Spectrom. 2012, 47, 737–750

Alizapride metabolism
2
n
Figure 9. Positive MS (m/z 332) spectrum and the proposed structures of the MS fragment ions of the metabolite M9.
2
Figure 10. Positive LC/MS chromatograms of AL incubation (RLM) monitoring the ion at m/z 410 (M9-d): incubation without the presence of 2-
2
mercaptoethanol (a), incubation with the presence of 2-mercaptoethanol (b). Positive MS (m/z 410) spectrum (c) and the proposed structures of the
n
MS fragment ions of the adduct M9-d.
2
n
Figure 11. Positive MS (m/z 348) spectrum and the proposed structures of the MS fragment ions of the metabolite M5.
J. Mass. Spectrom. 2012, 47, 737–750
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jms

S. Aprile, E. Del Grosso and G. Grosa
2
Figure 12. Positive LC/MS chromatograms of AL incubation (HLM) monitoring the ion at m/z 376 (M5-d): incubation before (a) and after (b) the treat-
2
n
ment with 10% sulfuric acid in ethanol. Positive MS (m/z 410) spectrum (c) and the proposed structures of the MS fragment ions of the adduct M5-d.
sulfuric acid in ethanol led to the partial disappearance of M5 peak
and the concomitant detection of a new peak at 17.10 min (M5-d)
giving the protonated molecule [M+ H] at m/z 376 (Fig. 12). This
structure of N-{[1-(2,3-dihydroxypropyl)pyrrolidin-2-yl]methyl}-6-
methoxy-1H-benzotriazole-5-carboxamide was proposed for both
the metabolites which are diastereomers. The presence of the diolic
metabolites possibly reflects the activity of rat and human liver
epoxide hydrolases (EHs) which catalyzed the addition of water to
epoxides yielding the corresponding 1,2-diols. Moreover, it is
important to point out that epoxide M9 was only revealed as 2-
mercaptoethanol conjugate in HLM and HIM incubations, suggest-
ing that it was completely hydrolyzed by human microsomal EHs.
+
ion (28 Da higher than M5), and its fragment ion pattern at
m/z 319 (28 Da higher than m/z 291), 330 and 176 was in agree-
ment with the formation of the carboxyester function. Taken
together, these data suggested for M5 the structure of 5-{[(6-
methoxy-1H-benzotriazol-5-yl)carbonyl]amino}-4-(prop-2-en-1-yla-
mino)pentanoic acid. Finally, no traces of this metabolite were
detected in the HIM incubation analysis monitoring the same ion.
The assignment of the structure of the metabolite M5 was also
supported by the relevant literature. Indeed, Hukkanen and co-
workers reported the formation of the (S)-nicotine amino acid
CONCLUSIONS
[
23]
metabolite, the open-chain form of cotinine.
In this work, a detailed experimental approach to detection and
subsequent structural characterization of unknown metabolites
of AL using LC-MS techniques was presented; moreover, the
metabolic scheme for in vitro metabolism of AL was proposed
and reported for the first time (Scheme 2). Indeed, no data on
the metabolic fate of alizapride were reported so far in literature.
In general, the oxidative metabolic fate of AL appeared quite
similar in rat and human microsomes, and at the same time, it
is characterized by a marked regioselectivity being the oxidative
transformations mainly directed toward the N-allyl-pyrrolidinyl-2-
methylene moiety.
Metabolites M6a and M6b
Metabolites M6a and M6b eluting at 6.15 and 6.49 min, respec-
+
tively, gave protonated molecules [M + H] at m/z 350 when both
2
HLM and RLM incubations were analysed by LC-MS (Fig. 2, 3). No
traces of these metabolites were detected when incubation was
carried out using HIM. The product ion spectra (Fig. 13) of these
metabolites were identical and produced the most intense frag-
ment ion at m/z 158 corresponding to the [1-(2,3-dihydroxypropyl)
pyrrolidin-2-yl]methylium ion. Other diagnostic fragment ions were
m/z 259 (neutral loss NH CH -CHOH = CH OH), m/z 288 (neutral
In particular, alongside a number of metabolites, the formation
of the reactive and toxic metabolite acrolein (Ac) was assessed
2
2
2
2
loss CH OH= CH OH) and m/z 176 (amide cleavage) which
through LC-MS analysis, demonstrating (1) the formation of
2
2
confirmed that the 1H-benzotriazole nucleus was not affected by
enzymatic oxidation. Since metabolites M6a and M6b were charac-
the corresponding semicarbazone (Ac-S) in rat liver microsomal
incubation, (2) the formation of the metabolite M1, arising from
N-deallylation of AL in human liver and intestinal microsomal
terized by a very similar t and identical mass spectral data, the
R
wileyonlinelibrary.com/journal/jms
Copyright © 2012 John Wiley & Sons, Ltd.
J. Mass. Spectrom. 2012, 47, 737–750

Alizapride metabolism
2
n
Figure 13. Positive MS (m/z 350) spectra and the proposed structures of the MS fragment ions of the metabolites M6a and M6b.
Scheme 2. Proposed in vitro metabolic pathways for AL.
incubations. This finding was further validated by the detection of
M1 metabolite in human urine after a single-dose oral administra-
tion of AL. At the best of our knowledge, this is second case of a
drug involved in the generation of acrolein during its metabolic
transformation being the first represented by cyclophosphamide
where the toxic metabolite arises from the oxaphosphoranic ring
J. Mass. Spectrom. 2012, 47, 737–750
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/jms

S. Aprile, E. Del Grosso and G. Grosa
n
[
[
[
4] M. Shuguang, R. Subramanian. Detecting and characterizing reactive
metabolites by liquid chromatography/tandem mass spectrometry.
J. Mass Spectrom. 2006, 41, 1121.
5] G. Aldini, I. Dalle-Donne, R. Maffei Facino, A. Milzani, M. Carini. Interven-
tion strategies to inhibit protein carbonylation by lipoperoxidation-
derived reactive carbonyls. Med. Res. Rev. 2007, 27, 817.
6] L. M. Sayre, D. Lin, Q. Yuan, X. Zhu, Y. Tang. Protein adducts gener-
ated from products of lipid peroxidation: focus on the HNE and
ONE. Drug Metab. Rev. 2006, 38, 651.
cleavage. Furthermore, LC-MS analysis revealed the formation of
the electrophilic and reactive epoxide M9. Indeed, its presence
has been demonstrated: (1) by revealing the corresponding ions
in the RLM LC-MS traces (m/z 332), (2) by demonstrating the
formation of the corresponding adducts (m/z 410) when 2-mercap-
toethanol was added in the incubation mixture.
Interestingly, differently from alclofenac, the formation of
these reactive metabolites did not result in the appearance of
significant adverse effects during the therapeutic use of aliza-
pride; this could be explained taking into account its pharmaco-
kinetic properties. Indeed, alizapride was mainly eliminated as
unchanged drug by renal pathway being metabolized for only
2
[
7] S. M. Ludeman. The chemistry of the metabolites of cyclophospha-
mide. Curr. Pharm. Des. 1999, 5, 627.
[
8] J. Zhang, Q. Tian, S. Yung Chan, S. Chuen Li, S. Zhou, W. Duan, Y. Z.
Zhu. Metabolism and transport of oxazaphosphorines and the
clinical implications. Drug Metab. Rev. 2005, 37, 611.
9] B. Bonn, C. M. Masimirembwa, N. Castagnoli Jr. Exploration of cata-
lytic properties of CYP2D6 and CYP3A4 through metabolic studies
of levorphanol and levallorphan. Drug Metab. Dispos. 2010, 38, 187.
[
[
13]
2
5% of the administered dose. Moreover, alizapride was used
at a typical dose of 50 mg as antiemetic in the treatment of acute
nausea and vomiting associated with chemotherapy and with
postoperative status. On the contrary, alclofenac has been given
[
10] W. B. Fang, Y. Chang, E. F. McCance-Katz, D. E. Moody. Determina-
tion of naloxone and nornaloxone (noroxymorphone) by high-
performance liquid chromatography-electrospray ionization-tandem
mass spectrometry. J. Anal. Toxicol. 2009, 33, 409.
[
18]
in doses of 1.5 to 3 g daily
to treat rheumatoid arthritis a
chronic disease requiring a long-term drug therapy. Taking
together, these features result for alclofenac in a greater exposition
to the reactive metabolites than that experienced with alizapride.
The data arising from AL and alclofenac metabolic studies
showed that the N-allyl moiety was enzymatically oxidized to
electrophilic and reactive species, acrolein and epoxide, whose
interaction with biological targets could cause adverse effects.
Even if there is not simple correlation between in vitro bioactiva-
tion and adverse effects in clinic, these data suggest that the N-
allyl moiety should be used with caution in the synthesis of
chemical libraries for drug lead discovery. Finally, since AL is used
as a racemic mixture, a future development of the work will
involve the evaluation of difference in the metabolic activation
of the AL enantiomers.
[11] H. Bleiberg, B. Gerard, O. Dalesio, N. Crespeigne, M. Rozencweig.
Activity of a new antiemetic agent: alizapride. A randomized double-
blind crossover controlled trial. Cancer Chemother. Pharmacol.
1
988, 22, 316.
[
12] C. Laville, J. Margarit. Pharmacodynamics of alizapride. Sem. Hop.
1982, 58, 323.
[13] E. Rey, E. Stettler, P. D’athis, G. Pons. Pharmacokinetics of alizapride
in children receiving chemotherapy for solid tumour. Fundam. Clin.
Pharmacol. 2001, 15, 217.
[
14] I. Tamaro, S. Aprile, G. B. Giovenzana, G. Grosa. Development and
validation of a stability-indicating HPLC-UV method for the determi-
nation of alizapride and its degradation products. J. Pharm. Biomed.
Anal. 2010, 51, 1024.
[
15] I. Monkovic, D. Willner, M. A. Adam, M. Brown, R. R. Crenshaw, C. E.
Fuller, P. F. Juby, G. M. Luke, J. A. Matiskella, T. A. Montzka.
Substituted benzamides.1. Potential nondopaminergic antagonists
of chemotherapy-induced nausea and emesis. J. Med. Chem.
1
988, 31, 1548.
[
16] G. Grosa, U. Galli, B. Rolando, R. Fruttero, G. Gervasio, A. Gasco. Iden-
tification of 2,3-diaminophenazine and of o-benzoquinone dioxime
as the major in vitro metabolites of benzofuroxan. Xenobiotica
Acknowledgements
This work was supported by M.I.U.R.-PRIN 2009, Italy.
2
004, 34, 345.
[
17] T. Omura, R. Sato. The carbon monoxide binding pigment of liver
microsomes. J. Biol. Chem. 1964, 39, 3137.
Supporting Information
[
18] R. N. Brogden, R. C. Heel, T. M. Speight, G. S. Avery. Alclofenac: a
review of its pharmacological properties and therapeutic efficacy
in rheumatoid arthritis and allied rheumatic disorders. Drugs
Supporting information may be found in the online version of
this article.
1
977, 14, 241.
19] R. D. Mann. Withdrawal of alclofenac (Prinalgin). Br. Med. J. 1979,
, 133.
[
[
2
REFERENCES
20] J. F. Hort. Adverse reactions to alclofenac. Curr. Med. Res. Opin.
1975, 3, 333.
[21] D. Mulliner, D. Wondrousch, G. Schuurmann. Predicting Michael-
acceptor reactivity and toxicity through quantum chemical transition-
state calculations. Org. Biomol. Chem. 2011, 9, 8400.
[22] H. Hoffmann, R. Pooth. Biotransformation of 1H-benzotriazole and
N-1-alkylbenzotriazoles in vitro. Arch. Pharm. 1982, 315, 422.
[23] J. Hukkanen, P. Jacob III, N. L. Benowitz. Metabolism and Disposition
Kinetics of Nicotine. Pharmacol. Rev. 2005, 57, 115.
[1] A. P. Li. Overview: Evaluation of metabolism-based drug toxicity in
drug development. Chem. Biol. Interact. 2009, 179, 1.
[
2] T. A. Baillie. Future of toxicology-metabolic activation and drug
design: challenges and opportunities in chemical toxicology. Chem.
Res. Toxicol. 2006, 19, 889.
[
3] R. Kostiainen, T. Kotiaho, T. Kuuranne, S. Auriola. Liquid chromatog-
raphy/atmospheric pressure ionization-mass spectrometry in drug
metabolism studies. J. Mass Spectrom. 2003, 38, 357.
wileyonlinelibrary.com/journal/jms
Copyright © 2012 John Wiley & Sons, Ltd.
J. Mass. Spectrom. 2012, 47, 737–750