Characterization of moclobemide N-oxidation in human liver microsomes

Article abstract of DOI:10.1080/00498250110055488

1. Moclobemide undergoes morpholine ring N-oxidation to form a major metabolite in plasma, Ro12-5637. 2. The kinetics of moclobemide N-oxidation in human liver microsomes (HLM) (n = 6) have been investigated and the mixed-function oxidase enzymes catalysi

Full text of DOI:10.1080/00498250110055488

xenobiotica, 2001, vol. 31, no. 7, 387±397
Characterization of moclobemide N-oxidation in
human liver microsomes
J. HOSKINSy*, G. SHENFIELDy, M. MURRAYz and A. GROSSy§
y Department of Clinical Pharmacology, Royal North Shore Hospital, St Leonards,
NSW 2065, Australia
z School of Physiology and Pharmacology, University of New South Wales, Randwick
NSW 2052, Australia
§ Present address: James Lance GlaxoSmithKline Medicines Research Unit, Parkes 10 E,
Prince of Wales Hospital, Randwick NSW 2031, Australia
Received 11 January 2001
1. Moclobemide undergoes morpholine ring N-oxidation to form a major metabolite
in plasma, Ro12-5637.
2. The kinetics of moclobemide N-oxidation in human liver microsomes (HLM)
…n ˆ 6† have been investigated and the mixed-function oxidase enzymes catalysing this
reaction have been identi®ed using inhibition, enzyme correlation, altered pH and heat
pretreatment experiments.
3. N-oxidation followed single enzyme Michealis-Menten kinetics (0.02±4.0 mM).
Km app and Vmax ranged from 0.48 to 1.35 mM (mean § SD 0:77 § 0:34 mM) and 0.22 to
1
1
1
2.15 nmol mg min (1:39 § 0:80 nmol mg min 1), respectively.
4. The N-oxidation of moclobemide strongly correlated with benzydamine N-
oxidation, a probe reaction for ¯avin-containing monooxygenase (FMO) activity, (0.1 mM
moclobemide, rS ˆ 0:81; p < 0:005; 4 mM moclobemide, rS ˆ 0:94; p ˆ 0:0001). Correla-
tions were observed between moclobemide N-oxidation and speci®c cytochrome P450
(CYP) activities at both moclobemide concentrations (0.1 mM moclobemide, CYP2C19
rS ˆ 0:66; p < 0:05; 4 mM moclobemide, CYP2E1 rS ˆ 0:56; p < 0:05).
5. The general P450 inhibitor, N-benzylimidazole, did not a€ect the rate of Ro12-5637
formation (0% inhibition versus control) at 1.3 mM moclobemide. Furthermore, the rate of
Ro12-5637 formation in HLM was una€ected by inhibitors or substrates of speci®c P450s
(<10% inhibition versus control).
6. Heat pretreatment of HLM in the absence of NADPH (inactivating FMOs)
resulted in 97% inhibition of Ro12-5637 formation. N-oxidation activity was greatest
when incubated at pH 8.5. These results are consistent with the reaction being FMO
mediated.
7. In conclusion, moclobemide N-oxidation activity has been observed in HLM in
vitro and the reaction is predominantly catalysed by FMOs with a potentially small
contribution from cytochrome P450 isoforms.
Introduction
Moclobemide is a reversible selective inhibitor of monoamine oxidase-A
(MAO-A) that is widely prescribed for the treatment of depression. It undergoes
extensive ®rst-pass metabolism and following oral administration <0.5% of the
parent drug is excreted unchanged in the urine (Jauch et al. 1990). Moclobemide
undergoes N- and C-oxidation to form its two major metabolites in plasma, Ro12-
5637 and Ro12-8095, respectively (Jauch et al. 1990) (®gure 1). Ro12-5637 is a
* Author for correspondence: Department of Pharmacology, Bosch Building D05, University of
Sydney, NSW 2006, Australia; e-mail: janelle@med.usyd.edu.au
#
Xenobiotica ISSN 0049±8254 print/ISSN 1366±5928 online
2001 Taylor & Francis Ltd
http://www.tandf.co.uk/journals
DOI: 10.1080/00498250110055488

388
J. Hoskins et al.
Figure 1. Primary metabolic pathways of moclobemide.
weak inhibitor of MAO-A, whereas Ro12-8095 has no MAO-A inhibitory activity
(Da Prada et al. 1989). About 35% of an orally administered dose of moclobemide
is recovered in urine as Ro12-5637, whereas Ro12-8095 accounts for <1% of the
dose (Mayersohn and Guentert 1995).
The enzymes catalysing moclobemide oxidation have not been characterized.
In vivo evidence suggests that the C-oxidation pathway is partially mediated by
cytochrome P450 2C19 (Gram et al. 1995). N-oxidation of the tertiary amines of
many psychoactive drugs, including imipramine, clozapine and olanzapine, is
predominantly catalysed by FMOs rather than by cytochrome P450s (Lemoine
et al. 1990, Ring et al. 1996, Tugnait et al. 1997). N-oxidation of moclobemide
occurs at the tertiary amine of the morpholine ring suggesting that this reaction
may be FMO rather than P450 mediated.
Cytochrome P450s and FMOs are present in the microsomal fraction prepared
by di€erential centrifugation of homogenized hepatic cells. The hepatic oxidation
of a drug can be characterized in vitro using microsomes prepared from human
liver samples. Straightforward experiments can be employed to determine whether
a reaction is FMO- or P450-mediated (Grothusen et al. 1996) and enzyme
correlation and chemical inhibition experiments can be performed to identify
the speci®c P450 isoforms involved (Chang et al. 1994, Newton et al. 1995, Bourrie

Moclobemide N-oxidaton and FMOs
389
et al. 1996, Ono et al. 1996). The objective of the current study was to characterize
moclobemide N-oxidation in human liver microsomes (HLM) in vitro and to
identify the mixed-function oxidase(s) that catalyse(s) this reaction.
Materials and methods
Chemicals
Dextromethorphan hydrobromide, coumarin, ¬-naphtho¯avone, phenacetin, quinidine, salicylic
acid, tolbutamide, troleandomycin (TAO) and - -NADP‡ (- -nicotinamide adenine dinucleotide phos-
phate, potassium salt) were purchased from Sigma Chemical Co. (St Louis, MO, USA). S-mepheny-
toin and (§)-40-hydroxymephenytoin were purchased from Salford Ultra®ne Chemicals and Research
Ltd (Manchester, UK) and didiethylcarbamate (DDC) and N-benzylimidazole from Aldrich Chemical
Co. (Milwaukee, WI, USA). Other drugs and drug metabolites were generously provided as follows:
moclobemide, Ro12-8095, Ro12-5637 and dextrorphan tartrate from Roche Pty Ltd (Sydney, NSW,
Australia), ketoconazole from Janssen-Cilag Pty Ltd (North Ryde, NSW, Australia), sulphaphenazole
from Ciba-Geigy Australia Ltd (Pendle Hill, NSW, Australia), and benzydamine and benzydamine N-
oxide from 3M Pharmaceuticals Pty Ltd (Sydney, NSW, Australia). Glucose 6-phosphate and glucose
6-phosphate dehydrogenase were purchased from Boehringer Mannheim (Sydney, NSW, Australia).
Methanol and triethylamine were HPLC grade and purchased from BDH (Kilsyth, VIC, Australia).
Potassium phosphate salts, 70% perchloric acid, 85% orthophosphoric acid and magnesium chloride
were analytical grade and purchased from Ajax Chemicals (Auburn, NSW, Australia). Methanolic
solutions were prepared of moclobemide and inhibitors and stored at 208C. Ro12-5637 was prepared
in double-distilled water.
Ro12-5637 high-performance liquid chromatography (HPLC) assay
The reversed-phase HPLC method of Geschke et al. (1987) was modi®ed to measure moclobemide
and Ro12-5637 in microsomal incubates. A mobile phase consisting of methanol:triethylamine:33.5 mM
potassium dihydrogen orthophosphate (25 : 0.1 : 74.9 v/v/v), adjusted to pH 3 using orthophosphoric
1
acid, was pumped at a ¯ow rate of 1.3 ml min using an LC-10AT pump (Shimadzu Oceania Pty Ltd,
Rydalmere, NSW, Australia) with a back pressure of 150±180 kg cm 2. An SIL-10AXL Autoinjector
(Shimadzu Oceania) was used to inject 0.13 ml of the incubates and calibration standards onto a
Spherisorb 5 mm C6 column 150 £ 4:6 mm i.d. (Phenomenex, Torrance, CA, USA). An SPD-10A UV-
Vis detector (Shimadzu Oceania) set at 240 nm detected moclobemide and Ro12-5637. The system was
controlled and chromatograms integrated using CLASS LC10 (v.1.60) software (Shimadzu Oceania).
Calibration standards of 0.04-12.3 mM Ro12-5637 in 0.2 ml 0.1 M potassium phosphate bu€er (pH 7.4)
and 10 ml 17.5% perchloric acid were prepared in 1.7 ml centrifuge tubes (Edwards Instrument Co.,
Narellan, NSW, Australia). Standards were vortexed and centrifuged for 8 min at 6800g. Standard
curves were estimated by least-squares linear regression of Ro12-5637 concentration versus peak height.
Concentrations of Ro12-5637 in the microsomal incubates were calculated from the regression lines.
Microsomal preparation
Microsomes from HL2-32 (n ˆ 11) were prepared as described by Sutton et al. (Sutton et al. 1997).
HLM1-6 were prepared from healthy hepatic tissue obtained at biopsy that would otherwise have been
discarded as surgical waste. After tissue removal it was immediately placed into liquid nitrogen and
stored at 708C. The cytochrome P450 content and protein concentration were determined as
previously described (Rieutord et al. 1995).
Moclobemide N-oxidation microsomal assay
Methanolic solutions of moclobemide and inhibitor solutions (prepared in methanol and stored at
208C) were added to 10 ml polypropylene tubes (Sarstedt Australia Pty Ltd, Technology Park, SA,
Australia) and the methanol was evaporated using a Speedivac Rotary Evaporator (Savant Instruments,
Inc., Farmingdale, NY, USA) prior to incubation. A typical incubation (®nal volume of 0.2 ml)
contained moclobemide, potassium phosphate bu€er (0.1 M, pH 7.4) and an NADPH-generating
system (®nal concentrations: 10 mM glucose 6-phosphate, 10 mM MgCl2, 1 mM NADP‡, 0.4 U glucose
6-phosphate dehydrogenase). Samples were pre-incubated for 5 min at 378C in a constant-temperature
SW-208C agitated water bath (Julabo Labortechnik, Seelbach, Germany) and reactions were initiated
by adding HLM. The reaction proceeded without a lid so that the incubate was in equilibrium with

390
J. Hoskins et al.
oxygen in the air. The reaction was stopped by adding 10 ml 17.5% perchloric acid and samples were
then vortexed and centrifuged for 8 min at 6800g.
The relationship between rate of N-oxide formation and HLM protein concentration was assessed
1
from 0.025 to 0.3 mg ml
at 378C for 60 min. The relationship between rate of formation and
incubation time was assessed from 5 to 120 min at 378C at a protein concentration of 0.1 mg ml 1. To
determine the reproducibility of the microsomal assay, moclobemide was incubated with HLM2 in
duplicate on four separate occasions to assess between-day reproducibility and in quadruplicate on a
single occasion to assess within-day reproducibility. The conditions selected for investigation were a
1
protein concentration of 0.1 mg ml and an incubation time of 60 min. For kinetic studies, samples
from HLM1, 2±6 and HL27 were incubated with increasing concentrations of moclobemide (0.02±4.0
mM). For activity correlations the rate of Ro12-5637 formation was measured at 0.1 and 4 mM
moclobemide in 11 HLM samples. All other experiments were performed at 1.3 mM moclobemide
(average apparent Km). To investigate the contribution of FMO to the N-oxidation of moclobemide,
microsomes (HLM2) were pre-incubated with and without the NADPH-regenerating system at 508C
for 1 min and stored on ice for <10 min before being incubated with moclobemide. To assess the e€ect
of pH on moclobemide N-oxidation, the rate of Ro12-5637 formation was measured at pHs 7.0, 7.4, 8.0,
8.5, 9.0 and 9.5 in HLM2. For these incubations, 0.2 M Tris-HCl was prepared and adjusted to the
appropriate pH by the addition of 0.1 M HCl. All incubations were performed in duplicate and the
average value is reported.
Activity of speci®c P450 isoforms and FMOs
Eleven HLM samples were assessed for the following activities: testosterone 6- -hydroxylation
(CYP3A4; 50 mM), 7-ethoxyresoru®n-O-deethylation (CYP1A2; 2.5 mM), N-dimethylnitrosoamine N-
demethylation (CYP2E1; 4 mM), aniline 4-hydroxylation (CYP2E1; 5 mM) and tolbutamide 4-hydro-
xylation (CYP2C9; 0.3 mM) as described (Sutton et al. 1997). Dextromethorphan O-demethylation
(CYP2D6; 40 mM) (Kerry et al. 1993), S-mephenytoin 40-hydroxylation (CYP2C19; 0.2 mM) (Chiba et
al. 1993) and benzydamine N-oxidation (FMO; 0.5 mM) (Baldock et al. 1990, Lang et al. 1998) were
also assessed.
Chemical inhibition experiments
Potential chemical inhibitors were pre-incubated with moclobemide as described above and the
reactions were initiated by adding HLM2. The mechanistic inhibitors, TAO and DDC, were pre-
incubated with HLM and an NADPH-regenerating system for 30 min at 378C and the moclobemide N-
oxidation reaction initiated by adding moclobemide in 0.1 M phosphate bu€er (pH 7.4). The ®nal
concentrations of putative P450 inhibitors and substrates were 1 mM N-benzylimidazole (general P450
inhibitor), 10 and 25 mM coumarin (CYP2A6), 10 and 50 mM DDC (CYP2E1), 2 and 5 mM ketoconazole
(CYP3A4/5), 50 and 250 mM S-mephenytoin (CYP2C19), 0.1 and 1.0 mM ¬-naphtho¯avone (CYP1A1/
2), 50 and 250 mM phenacetin (CYP1A1/2), 0.5 and 5 mM quinidine (CYP2D6), 10 and 20 mM
sulphaphenazole (CYP2C9), 0.2 and 0.5 mM tolbutamide (CYP2C9), and 10 and 50 mM TAO
(CYP3A4). Salicylic acid (0.25 mM) was used as a negative control. The ®nal concentrations were
selected on the basis of literature reports of each compound’s maximum inhibition selectivity
(Grothusen et al. 1996) and the Km of the substrates (Chiba et al. 1993, Chang et al. 1994, Newton
et al. 1995, Bourrie et al. 1996, Ono et al. 1996).
Data analysis
The rate of metabolite formation (v) was calculated as nanomoles of Ro12-5637 formed per
1
milligram of protein per min (nmol mg min 1). Apparent Km and Vmax derived from Eadie-Hofstee
plots were used as initial estimates for ®tting the kinetic data to the single-enzyme Michaelis-Menten
equation using extended least-squares non-linear regression. The relationships between moclobemide
N-oxidation and the activities of FMOs and speci®c P450s were assessed by calculating the magnitude
and signi®cance of the Spearman rank correlation coecients.
Results
Using the HPLC technique described, the retention times of moclobemide and
Ro12-5637 were 5.0 and 6.9 min, respectively, with a ®nal run time of 20 min. A
representative chromatogram of moclobemide incubated with HLM and an
NADPH-regenerating system is shown in ®gure 2A. The calibration curve was
linear for the measurement of Ro12-5637 over the concentration range studied
with regression coecients >0.99. A typical equation for the Ro12-5637 calibra-

Moclobemide N-oxidaton and FMOs
391
Figure 2. Representative chromatograms of moclobemide incubated with human liver microsomes
(0.1 mg protein ml 1) at 378C for 60 min. Chromatogram (A) is an incubation of moclobemide
(1.3 mm) with HLM2 and (B) is an incubation of moclobemide (1.3 mM) with HLM2 pre-
incubated at 508C for 1 min in the absence of an NADPH-regenerating system. FMOs are
unstable at 508C in the absence of NADPH.
tion curve was y ˆ 5555x 540 The limit of quanti®cation for the Ro12-5637
:
assay was 0.04 mM.
Formation of Ro12-5637 was observed in the presence of an NADPH-
regenerating system but not in the absence of NADPH under the same conditions
(table 1). Over the range of moclobemide concentrations studied, <10% of the
parent drug was metabolized. At a moclobemide concentration of 1.3 mm the
within-day coecient of variation of Ro12-5637 formation (n ˆ 4) was <4.0%.
The between-day coecient of variation for the incubation assay determined at 0.1
mM moclobemide on four separate occasions was <4.0%.
Ro12-5637 formation was linear with time from 5 to 120 min and was also
1
linear with protein concentration from 0.025 to 0.3 mg ml (data not shown). The

392
J. Hoskins et al.
Table 1. Velocity of moclobemide N-oxidation to Rol2-5637 under
di€erent experimental conditions. Moclobemide (1.3 mM) was
incubated with human liver microsomes (HLM2) for 60 min
at 378C.
HLM pre-incubation
conditions
Incubatio‡n
co-factor
Velocity
(nmol mg min
1
1
z
)
No preincubation
No preincubation
Heat§
+NADPH
NADPH
+NADPH
+NADPH
1.26
0.00
0.03
1.12
Heat, NADPH§
‡Cofactor was an NADPH regenerating system consisting of
glucose-6-phosphate, MgCl2, NADP‡ and glucose-6-phosphate
dehydrogenase.
zEach value represents the average of duplicate measurements.
§Microsomal protein was preincubated with moclobemide for 1
min at 508C in the presence or absence of NADPH regenerating
system and stored on ice <10 min before being incubated with
moclobemide as per usual.
rate of Ro12-5637 formation was subsequently determined over the concentration
range of 0.02-4 mM moclobemide. Eadie-Hofstee plots for metabolite formation
were monophasic for all six livers tested so kinetic data was ®tted to the Michaelis-
Menten equation for a single enzyme system (®gure 3). The apparent Km varied
2.8-fold from 0.48 to 1.35 mM (mean § SD ˆ 0 77 § 0 34 mM) and V
varied
:
:
max
1
1
9.8-fold from 0.22 to 2.15 nmol mg min …mean § SD ˆ 1 39 § 0 80 nmol
:
:
1
mg min 1).
In a bank of 11 HLM samples, large interindividual variability in moclobemide
N-oxidation was observed at both 0.1 mM (26-fold) and 4 mM (65-fold). One
Figure 3. Moclobemide (0.02-4.0 mM) incubated with human liver microsomes (HLM6; 0.1 mg
protein ml 1) at 378C for 60 min. Data are presented as Michaelis-Menten (A) and Eadie-
Hofstee (B) plots. Each point represents the average of duplicate measurements.

Moclobemide N-oxidaton and FMOs
393
sample had very low N-oxidation activity (moclobemide 0.1 mM, v ˆ 0 007 nmol
:
1
1
mg min 1; 4 mM, 0.023 nmol mg min 1) relative to other HLM samples and
so was considered an outlier and excluded from further analysis. The activities of
the 10 human liver preparations varied 4-fold at 0.1 mM moclobemide and 6-fold
at 4 mM. The Spearman rank correlations between moclobemide N-oxidation and
isoform speci®c P450 and FMO activities are presented in table 2. At both
moclobemide concentrations high correlations were observed between the rate of
Ro12-5637 formation and the N-oxidation of benzydamine, a probe reaction for
FMO activity. At 0.1 mM moclobemide, but not 4.0 mM, a weaker correlation was
observed between moclobemide N-oxidation and S-mephenytoin 40-hydroxyla-
tion. At the high moclobemide concentration a weak correlation was observed
between moclobemide N-oxidation and one of the CYP2E1 probe reactions,
aniline 4-hydroxylation, but not for another CYP2E1 mediated reaction, N-
dimethylnitrosoamine N-demethylation. No signi®cant relationships between
moclobemide N-oxidation and 7-ethoxyresoru®n O-deethylation, dextromethor-
phan O-demethylation, tolbutamide 4-hydroxylation and testosterone 6 -hydro-
-
xylation were observed suggesting that CYP1A2, CYP2D6, CYP2C9 and
CYP3A4 respectively do not mediate this reaction at the moclobemide concentra-
tions tested.
At 1.3 mM moclobemide, the general P450 inhibitor, N-benzylimidazole, had
no e€ect on Ro12-5637 formation by HLM. In addition, all of the isoform speci®c
inhibitors and substrates tested produced <10% inhibition of N-oxidation (data
not shown). Heat treatment of HLM in the absence of an NADPH-regenerating
system (Grothusen et al. 1996) decreased Ro12-5637 formation to 2% of control in
HLM (table 1). The NADPH-generating system protected the reaction from
thermal inactivation (88% of control activity) (table 1). At pH 7.4 the rate of
moclobemide N-oxidation was greater for moclobemide incubated with HLM in
1
1
Tris-HCl bu€er rather than phosphate bu€er (1.14 versus 1.00 nmol mg min
respectively). Ro12-5637 formation was greater at pH>7.4 and maximal at pH 8.5
,
1
(2.10 nmol mg min 1) (®gure 4).
Table 2. Correlations (rS) between moclobemide N-oxidation and ¯avin-containing monooxygenase
(FMO) activity and cytochrome P450 (CYP) isoform-speci®c activities in a panel of human liver
microsomal preparations (n ˆ 10).
Correlation coecient (rS† with
moclobemide N-oxidation
Isoform-selective activity
0.1 mM
4 mM
Benzydamine N-oxidation (FMO)
0.806**
0.527
0.358
0.661*
0.376
0.134
0.430
0.394
0.939***
0.394
0.503
0.249
0.394
0.146
0.564*
0.297
7-Ethoxyresoru®n O-deethylation (CYP1A2)
Tolbutamide 4-hydroxylation (CYP2C9)
S-mephenytoin 40-hydroxylation (CYP2C19)
Dextromethorphan O-demethylation (CYP2D6)
N-dimethylnitrosomaine N-demethylation (CYP2E1)
Analine 4-hydroxylation (CYP2E1)
Testosterone 6- -hydroxylation (CYP3A4)
* p < 0:05; **p < 0:005; ***p < 0:0005.

394
J. Hoskins et al.
Figure 4. E€ect of pH on N-oxidation of moclobemide. Moclobemide (1.3 mM) was incubated for 60
min at 378C with HLM2, an NADPH-generating system and 0.2 M Tris-HCl bu€er at increasing
pH. Each point represents the average of duplicate measurements.
Discussion
The formation of the N-oxide metabolite of moclobemide, Ro12-5637, was
NADPH dependent indicating that the reaction is FMO and/or cytochrome P450-
mediated. FMOs are inactivated selectively by pre-incubating HLM at 508C for 1
min (Grothusen et al. 1996) in the absence of an NADPH-regenerating system.
These heat conditions greatly diminished the formation of Ro12-5637 whereas
HLM heat-treatment in the presence of an NADPH-regenerating system protects
the enzyme from inactivation. Furthermore, the reaction velocity is maximal at pH
8.5, a characteristic of FMO-mediated reactions (Cashman et al. 1993, Brunelle et
al. 1997), and the rate of Ro12-5637 formation is strongly correlated with the N-
oxidation of benzydamine, an FMO substrate (Lang et al. 1998). Together these
observations are consistent with FMOs catalysing the N-oxidation of moclobe-
mide in human liver, in vitro. Five isoforms of FMO (1±5) have been identi®ed in
humans. FMO3 and FMO5 are expressed in the adult human liver and FMO3 is
the major isoform (Overby et al. 1997). A recent study using recombinant human
FMOs showed that FMO3 catalyses benzydamine N-oxidation with greater
eciency than FMO5 (FMO3: Km ˆ 60 mM; Vmax ˆ 46 min 1, FMO5: Km
2
>
mM; Vmax 1 min 1) (Lang and Rettie 2000), suggesting benzydamine N-oxida-
<
tion is a speci®c probe reaction for FMO3 activity in HLM. Given the strong
relationship between benzydamine and moclobemide N-oxidation it is likely that
the formation of Ro12-5637 is also mediated by FMO3 in HLM.
Large variability in moclobemide N-oxidation activity was observed (26- and
65-fold), however, when the HLM sample with the lowest activity was excluded,
the variability (6-fold) was similar to that reported for other FMO mediated

Moclobemide N-oxidaton and FMOs
395
reactions (Overby et al. 1997, Tugnait et al. 1997, Lang et al. 1998). Individuals
homozygous for a missense Pro153 to Leu153 mutation in exon 4 of FMO3 have
impaired N-oxidation of trimethylamine (Dolphin et al. 1997, Zschocke et al.
1999). These individuals su€er from `®sh-odour’ syndrome because they excrete
the ®shy smelling trimethylamine in their sweat, urine and breath. A number of
other mutations of FMO3 have been identi®ed, some of which result in amino acid
changes that may contribute to interindividual variation in FMO catalytic activity
(Cashman et al. 2000, Kang et al. 2000). The variability observed in the group of
human liver microsomal samples tested in this study may be due to interindividual
variability in FMO3 genotypes, a hypothesis which remains to be con®rmed.
To the authors’ knowledge there is one report of benzydamine N-oxidation
activities in HLM (Lang et al. 1998). Using the same incubation conditions, the
benzydamine N-oxidation activities observed in the present study are lower than
1
those previously reported (0.15±2.3 versus 2.8±9.5 nmol mg min 1) (Lang et al.
1998). These lower activities may re¯ect di€erent microsomal preparation tech-
niques (Sadeque et al. 1993). FMOs are very sensitive to thermal inactivation in
the absence of NADPH and labile to mechanical disruption (Cashman 1999).
Although our HLM have lower FMO catalytic activities possibly re¯ecting lower
levels of viable FMOs, a high correlation between benzydamine and moclobemide
N-oxidation was still observed supporting a major role for FMOs in moclobemide
N-oxidation.
The general P450 inhibitor, N-benzylimidazole, did not a€ect the formation of
Ro12-5637 suggesting a lack of P450 involvement in moclobemide N-oxidation.
Enzyme correlation experiments, however, indicate minor roles for CYP2C19 and
CYP2E1 in catalysis of N-oxidation at low and high moclobemide concentrations,
respectively. However, these observations are not supported by the chemical
inhibition results. Experiments using heterologously expressed P450 enzymes
would assist in elucidating the speci®c P450 isoforms involved in catalysing the
reaction and to what extent they contribute to the formation of Ro12-5637.
In conclusion, the N-oxidation of moclobemide of human liver microsomes
highly correlates with the ¯avin-containing monooxygenase mediated N-oxidation
of N-benzydamine, indicating that moclobemide N-oxidation is predominantly
catalysed by FMOs. The role of this group of enzymes was con®rmed by the
observation that Ro12-5637 formation is maximal at pH 8.5 and that heat
pretreatment of the human liver microsomes at 508C for 1 min in the absence of
NADPH abolishes the formation of Ro12-5637. Enzyme correlation and chemical
inhibition experiments suggest only minor roles for cytochrome P450s in catalys-
ing the reaction.
Acknowledgements
J. H. was funded by a Dora Lush (Biomedical) Scholarship awarded by the
National Health and Medical Research Council of Australia.
References
BALDOCK, G. A., BRODIE, R. R., CHASSEAUD, L. F. and TAYLOR, T., 1990, Determination of
benzydamine and its N-oxide in biological ¯uids by high-performance liquid chromatography.
Journal of Chromatography, 529, 113±123.

396
J. Hoskins et al.
BOURRIE, M., MEUNIER, V., BERGER, Y. and FABRE, G., 1996, Cytochrome P450 isoform inhibitors as a
tool for the investigation of metabolic reactions catalyzed by human liver microsomes. Journal of
Pharmacology and Experimental Therapeutics, 277, 321±332.
BRUNELLE, A., BI, Y. A., LIN, J., RUSSELL, B., LUY, L., BERKMAN, C. and CASHMAN, J., 1997,
Characterization of two human ¯avin-containing monooxygenase (form 3) enzymes expressed in
Escherichia coli as maltose binding protein fusions. Drug Metabolism and Disposition, 25, 1001±
1007.
CASHMAN, J., 1999, In T. F. Woolf (ed.), Handbook of Drug Metabolism (New York: Dekker), pp. 477±
505.
CASHMAN, J. R., AKERMAN, B. R., FORREST, S. M. and TREACY, E. P., 2000, Population-speci®c
polymorphisms of the human FMO3 gene: signi®cance for detoxication. Drug Metabolism and
Disposition, 28, 169±173.
CASHMAN, J. R., YANG, Z., YANG, L. and WRIGHTON, S. A., 1993, Stereo- and regioselective N- and S-
oxidation of tertiary amines and sul®des in the presence of adult human liver microsomes. Drug
Metabolism and Disposition, 21, 492±501.
CHANG, T. K., GONZALEZ, F. J. and WAXMAN, D. J., 1994, Evaluation of triacetyloleandomycin, alpha-
naphtho¯avone and diethyldithiocarbamate as selective chemical probes for inhibition of human
cytochromes P450. Archives of Biochemistry and Biophysics, 311, 437±442.
CHIBA, K., MANABE, K., KOBAYASHI, K., TAKAYAMA, Y., TANI, M. and ISHIZAKI, T., 1993,
Development and preliminary application of a simple assay of S-mephenytoin 4-hydroxylase
activity in human liver microsomes. European Journal of Clinical Pharmacology, 44, 559±562.
DA PRADA, M., KETTLER, R., KELLER, H. H., BURKARD, W. P., MUGGLI-MANIGLIO, D. and HAEFELY,
W. E., 1989, Neurochemical pro®le of moclobemide, a short-acting and reversible inhibitor of
monoamine oxidase type A. Journal of Pharmacology and Experimental Therapeutics, 248, 400±
414.
DOLPHIN, C. T., JANMOHAMED, A., SMITH, R. L., SHEPHARD, E. A. and PHILLIPS, I. R., 1997, Missense
mutation in ¯avin-containing mono-oxygenase 3 gene, FMO3, underlies ®sh-odour syndrome.
Nature Genetics, 17, 491±494.
GESCHKE, R., KORNER, J. and EGGERS, H., 1987, Determination of the new monoamine oxidase
inhibitor moclobemide and three of its metabolites in biological ¯uids by high-performance
liquid chromatography. Journal of Chromatography, 420, 111±120.
GRAM, L. F., GUENTERT, T. W., GRANGE, S., VISTISEN, K. and BROSEN, K., 1995, Moclobemide, a
substrate of CYP2C19 and an inhibitor of CYP2C19, CYP2D6, and CYP1A2: a panel study.
Clinical Pharmacology and Therapeutics, 57, 670±677.
GROTHUSEN, A., HARDT, J., BRAUTIGAM, L., LANG, D. and BOCKER, R., 1996, A convenient method to
discriminate between cytochrome P450 enzymes and ¯avin-containing monooxygenases in
human liver microsomes. Archives of Toxicology, 71, 64±71.
JAUCH, R., GRIESSER, E., OESTERHELT, G., ARNOLD, W., MEISTER, W., ZIEGLER, W. H. and GUENTERT,
T. W., 1990, Biotransformation of moclobemide in humans. Acta Psychiatrica Scandinavica,
Supplementum, 360, 87±90.
KANG, J. H., CHUNG, W. G., LEE, K. H., PARK, C. S., KANG, J. S., SHIN, I. C., ROH, H. K., DONG, M.
S., BAEK, H. M. and CHA, Y. N., 2000, Phenotypes of ¯avin-containing monooxygenase activity
determined by ranitidine N-oxidation are positively correlated with genotypes of linked FMO3
gene mutations in a Korean population. Pharmacogenetics, 10, 67±78.
KERRY, N. L., SOMOGYI, A. A., MIKUS, G. and BOCHNER, F., 1993, Primary and secondary oxidative
metabolism of dextromethorphan. In vitro studies with female Sprague-Dawley and Dark Agouti
rat liver microsomes. Biochemical Pharmacology, 45, 833±839.
LANG, D. H. and RETTIE, A. E., 2000, In vitro evaluation of potential in vivo probes for human ¯avin-
containing monooxygenase (FMO): metabolism of benzydamine and ca€eine by FMO and P450
isoforms. British Journal of Clinical Pharmacology, 50, 311±314.
LANG, D. H., YEUNG, C. K., PETER, R. M., IBARRA, C., GASSER, R., ITAGAKI, K., PHILPOT, R. M. and
RETTIE, A. E., 1998, Isoform speci®city of trimethylamine N-oxygenation by human ¯avin-
containing monooxygenase (FMO) and P450 enzymes: selective catalysis by FMO3. Biochemical
Pharmacology, 56, 1005±1012.
LEMOINE, A., JOHANN, M. and CRESTEIL, T., 1990, Evidence for the presence of distinct ¯avin-
containing monooxygenases in human tissues. Archives of Biochemistry and Biophysics, 276, 336±
342.
MAYERSOHN, M. and GUENTERT, T. W., 1995, Clinical pharmacokinetics of the monoamine oxidase-A
inhibitor moclobemide. Clinical Pharmacokinetics, 29, 292±332.
NEWTON, D. J., WANG, R. W. and LU, A. Y., 1995, Cytochrome P450 inhibitors. Evaluation of
speci®cities in the in vitro metabolism of therapeutic agents by human liver microsomes. Drug
Metabolism and Disposition, 23, 154±158.
ONO, S., HATANAKA, T., HOTTA, H., SATOH, T., GONZALEZ, F. J. and TSUTSUI, M., 1996, Speci®city of
substrate and inhibitor probes for cytochrome P450s: evaluation of in vitro metabolism using
cDNA-expressed human P450s and human liver microsomes. Xenobiotica, 26, 681±693.

Moclobemide N-oxidaton and FMOs
397
OVERBY, L. H., CARVER, G. C. and PHILPOT, R. M., 1997, Quantitation and kinetic properties of hepatic
microsomal and recombinant ¯avin-containing monooxygenases 3 and 5 from humans. Chemico-
Biological Interactions, 106, 29±45.
RIEUTORD, A., STUPANS, I., SHENFIELD, G. M. and GROSS, A. S., 1995, Gliclazide hydroxylation by rat
liver microsomes. Xenobiotica, 25, 1345±1354.
RING, B. J., CATLOW, J., LINDSAY, T. J., GILLESPIE, T., ROSKOS, L. K., CERIMELE, B. J., SWANSON, S.
P., HAMMAN, M. A. and WRIGHTON, S. A., 1996, Identi®cation of the human cytochromes P450
responsible for the in vitro formation of the major oxidative metabolites of the antipsychotic
agent olanzapine. Journal of Pharmacology and Experimental Therapeutics, 276, 658±666.
SADEQUE, A. J., THUMMEL, K. E. and RETTIE, A. E., 1993, Puri®cation of macaque liver ¯avin-
containing monooxygenase: a form of the enzyme related immunochemically to an isozyme
expressed selectively in adult human liver. Biochimica et Biophysica Acta, 1162, 127±134.
SUTTON, D., BUTLER, A. M., NADIN, L. and MURRAY, M., 1997, Role of CYP3A4 in human hepatic
diltiazem N-demethylation: inhibition of CYP3A4 activity by oxidized diltiazem metabolites.
Journal of Pharmacology and Experimental Therapeutics, 282, 294±300.
TUGNAIT, M., HAWES, E. M., MCKAY, G., RETTIE, A. E., HAINING, R. L. and MIDHA, K. K., 1997, N-
oxygenation of clozapine by ¯avin-containing monooxygenase. Drug Metabolism and Disposition,
25, 524±527.
ZSCHOCKE, J., KOHLMUELLER, D., QUAK, E., MEISSNER, T., HOFFMANN, G. F. and MAYATEPEK, E.,
1999, Mild trimethylaminuria caused by common variants in FMO3 gene. Lancet, 354, 834±835.