Functional activity of the mouse flavin-containing monooxygenase forms 1, 3, and 5

Article abstract of DOI:10.1002/jbt.20176

Three functional mouse flavin-containing monooxygenases (mFMOs) (i.e., mFMO1, mFMO3, and mFMO5) have been reported to be the major FMOs present in mouse liver. To examine the biochemical features of these enzymes, recombinant enzymes were expressed as mal

Full text of DOI:10.1002/jbt.20176

J BIOCHEM MOLECULAR TOXICOLOGY
Volume 21, Number 4, 2007
Functional Activity of the Mouse Flavin-Containing
Monooxygenase Forms 1, 3, and 5
Jun Zhang, Matt A. Cerny, Matt Lawson, Ruzbeh Mosadeghi, and John R. Cashman
Human BioMolecular Research Institute, 5310 Eastgate Mall, San Diego, CA 92121, USA; E-mail: jcashman@hbri.org
Received 19 March 2007; revised 1 April 2007; accepted 1 May 2007
ABSTRACT: Three functional mouse flavin-containing
monooxygenases (mFMOs) (i.e., mFMO1, mFMO3,
and mFMO5) have been reported to be the ma-
jor FMOs present in mouse liver. To examine the
biochemical features of these enzymes, recombi-
nant enzymes were expressed as maltose-binding
protein fusion proteins (i.e., MBP-mFMO1, MBP-
mFMO3, and MBP-mFMO5) in Escherichia coli and
isolated and purified with affinity chromatography.
The substrate specificity of these three mouse hep-
INTRODUCTION
The mouse is one of the most commonly used an-
imal models in preclinical testing during early drug
development. Understanding absorption, distribution,
metabolism, excretion, and toxicology data obtained in
the mouse may help accurately predict the outcome of
drug candidates in humans, but this requires a detailed
knowledge of the function of specific mouse genes
and proteins, especially those enzymes that are in-
volved in drug metabolism and disposition. The flavin-
containing monooxygenases (FMOs) are a family of en-
zymes involved in the metabolism of drugs, pesticides,
dietary-derived compounds, and endogenous chemi-
cals (see reviews [1,2]). In both human and mouse, five
functional forms of FMO have been identified, namely,
FMO1, 2, 3, 4, and 5, based on nucleotide sequence
and consensus amino acids. A number of other FMO
genes and pseudogenes have also been reported in
both the human and mouse genome [3,4], but appar-
ently no additional functional FMO proteins have been
characterized.
A comparison of FMO steady-state mRNA profiles
in human and mouse liver showed several features in-
cluding (1) FMO5 is transcribed at the highest level
in both human and mouse liver, however, due to a
lack of an optimal substrate for FMO5, the contribu-
tion of this isoform to liver metabolism is not clear [5,6];
(2) expression of FMO2 and FMO4 in both human and
mouse liver is very low, and likely does not significantly
contribute to hepatic metabolism of most drugs [5–7];
(3) expression of the other two major FMO isoforms
(i.e., FMO1 and FMO3) display substantial differences
in the liver of humans and mice; and (4) expression
of mouse FMOs display significant gender specificity,
whereas in human, no gender-specific expression of
FMO has been reported [5,8]. In adult human liver, ex-
pression of FMO1 is absent, whereas mouse FMO1 is
expressed at a substantial level in liver of both male and
female mice. Expression of FMO3 is prominent in adult
human liver, while in mouse, hepatic FMO3 expression
atic FMO enzymes were examined using
a va-
riety of substrates, including mercaptoimidazole,
trimethylamine, S-methyl esonarimod, and an analog
thereof, and a series of 10-(N,N-dimethylaminoalkyl)-
2-(trifluoromethyl)phenothiazine analogs. The kinetic
parameters of the three mouse FMOs for these sub-
strates were compared in an attempt to explore sub-
strate structure–function relationships specific for each
mFMO. Utilizing a common phenothiazine substrate
for all three enzymes, we compared the pH dependence
for the recombinant enzymes under similar conditions.
In addition, thermal stability for mFMO1, mFMO3,
and mFMO5 enzymes was examined in the presence
and absence of NADPH. The results revealed unique
features for mFMO5, suggesting possible impact on
the functional significance of this abundantly ex-
pressed FMO5 isoform in both human and mouse liver.
ꢀ
C
2007 Wiley Periodicals, Inc. J Biochem Mol Toxicol
21:206–215, 2007; Published online in Wiley InterScience
(www.interscience.wiley.com). DOI 10:1002/jbt.20176
KEYWORDS: Mouse Flavin-containing Monooxygenase;
mFMO1; mFMO3; mFMO5; Expression and Purifi-
cation; Substrate Specificity; pH Profile; Thermal
Stability
Correspondence to: John R. Cashman.
This work is dedicated to Randy Rose who inspired work in
this area by his careful and thoughtful craftsmanship.
Contract Grant Sponsor: NIH.
Contract Grant Number: DK 59618.
2007 Wiley Periodicals, Inc.
c
ꢀ
206

Volume 21, Number 4, 2007
FUNCTION OF MOUSE FMO1, FMO3, FMO5
207
is significant in female animals only [8,9]. The presence
of FMO1 and absence of FMO3 in male mouse liver
represents a profile similar to that observed in fetal hu-
man liver, in which expression of FMO1 is replaced by
FMO3 in the adult human liver. Despite the important
differences between mouse FMO and human FMO ex-
pression profiles and the general application of mouse
liver microsomes and mouse animal models to current
drug discovery research, biochemical functional stud-
ies of individual enzymes of mouse FMOs have been
reported only in a few studies [10–12].
In this study, we cloned, expressed, and purified
individual mouse FMO1, FMO3, and FMO5 enzymes
as recombinant proteins and examined their substrate
specificity, pH dependence, and thermal stability. The
availability of purified recombinant mouse FMO en-
zymes provide essential tools for further characteriza-
tion of FMO enzyme function and substrate specificity,
and further our understanding of the potential as well
as limitations of using mouse animal models or mouse
liver microsomes in both in vivo and in vitro model
systems, respectively, for drug metabolism as well as
pharmacokinetics and predictive toxicology studies for
new drug candidates.
Cloning and Expression of Mouse FMO1,
3, and 5
Mouse FMO1 and mouse FMO5 genes were
RT-PCR amplified from male C57BL6 liver RNA
using the following primers: 5ꢀ-mFMO5 gatctctaga-
atggccaaaaaaaggattgct (XbaI site italicized); 3ꢀ-mFMO5
gatcctgcagctaaaaataagccaggatgacagc (PstI site ital-
icized); 5ꢀ-mFMO1 gatctctagaatggtgaagcgagtggcaat
(XbaI site italicized); and 3ꢀ-mFMO1 gatcctgcagttacagg-
aaaatcaagaagacagc (PstI site italicized). Mouse FMO3
was subcloned from vector PJL-2 kindly pro-
vided by Dr. Randy Rose (North Carolina State
University, Raleigh, NC) using primers 5ꢀ-mFMO3
gatcgaattcatgaagaagaaagtggccatca (EcoRI site itali-
cized) and 3ꢀ-mFMO3 gatcaagcttcagatcaacacaaggaac-
aaag (HindIII site italicized). The mFMO1, mFMO3,
and mFMO5 genes were then subcloned into the ex-
pression vector pMAL-c2 (New England Biolab, Ip-
swich, MA) through the corresponding cloning sites
(XbaI/PstI for mFMO1 and mFMO5, EcoRI/HindIII for
mFMO3). FMO gene inserts in selected clones were se-
quenced in both directions to confirm that no mutation
was introduced during PCR subcloning procedures.
Plasmids were freshly transformed in DH1α compe-
tent cells for recombinant expression. Expression of
the recombinant fusion proteins MBP-mFMO1, MBP-
mFMO3, and MBP-mFMO5 was done similarly as pre-
viously described [16].
MATERIALS AND METHODS
Materials
All chemicals and reagents were purchased from
Aldrich Chemical Co. (Milwaukee, WI) in the highest
purity commercially available form. The components
of the NADPH-generating system were obtained from
Sigma Chemical Co. Buffers and other agents were pur-
chased from VWR Scientific, Inc., (San Diego, CA). The
synthesis of test compounds 3a, 3b, 3-DPT, 5-DPT, and
8-DPT has been previously reported [13–15], as well
as is reported in the following. For chemical synthesis,
moisture-sensitive reactions were carried out in flame-
dried glassware under an argon atmosphere. Tetrahy-
drofuran and toluene was freshly distilled from cal-
cium hydride under an argon atmosphere. Analyti-
cal thin-layer chromatography was done on K6F silica
Purification of MBP-mFMO Proteins
All purification procedures were carried out at
4◦C. Bacterial cells were harvested by centrifugation
at 6,000 × g for 10 min and resuspended in lysis buffer
(50 mM Na2HPO4, pH 8.4; 0.5% Triton X-100) con-
taining 0.2% L-α-phosphatidylcholine, 0.5 mM phenyl-
methylsulfonylfluoride, and 100 mM flavin adenine
dinucleotide. After an incubation for 30 min on ice, the
resuspended cells were disrupted by sonication (i.e.,
three 2-min bursts separated by periods of cooling).
The solution was centrifuged at 18,000 × g for 30 min
at 4◦C. The resulting supernatants containing the sol-
ubilized MBP-mFMOs were applied (0.5 mL/min) to
an amylose column (New England Biolab) (35 mL)
equilibrated with Buffer A (50 mM Na2HPO4, pH
8.4; 0.5% Triton X-100). The column was washed with
5 volumes of Buffer A containing 100 mM flavin ade-
nine dinucleotide. Bound protein was eluted with
10 mM maltose in Buffer A containing 100 mM flavin
adenine dinucleotide. Eluted fractions were analyzed
by SDS-PAGE. MBP-mFMO1, MBP-mFMO3, and MBP-
mFMO5 were quantified by Coommassie blue stain-
ing and compared with a bovine serum albumin (BSA)
˚
gel 60 A (Whatman) glass-backed plates. Compounds
were detected using UV absorbance at 254 nm and/or
stained with I2 (iodine). Flash chromatography was
˚
done on Merck (60 A) pore silica. NMR spectra were
recorded at 300 MHz on a Varian Mercury spectrom-
eter. Chemical shifts are reported in parts per million
(ppm, δ) using residual solvent signals as internal stan-
dards. Mass spectroscopy was done using electrospray
ionization (ESI) on a Hitachi M-8000 3DQMS (ion-trap)
mass spectrometer. UV spectral data were acquired on
a VarianCary 1E UV-visible spectrophotometer.
J Biochem Molecular Toxicology DOI 10:1002/jbt

208
ZHANG ET AL.
Volume 21, Number 4, 2007
standard. Briefly, MBP-mFMO proteins and different
quantities of standard BSA (2, 1.5, 1.0, 0.5, and 0.1 µg
per lane) were fractionated by electrophoresis on a
10% polyacrylamide gel under denaturing conditions
and stained with Coommasie Blue. After destaining,
FMO quantification was done by densitometry analy-
sis employing Scion Image software (public domain,
http://www.scioncorp.com). The relationship between
the relative intensity of the detected signal and amount
of mFMO was linear (r2 = 0.9).
2-Acetylsulfanylmethyl-4-(4-methoxyphenyl)-
4-oxobutyric Acid (2b)
Compound 1b (1.65 g, 7.5 mmol) was treated with
thioacetic acid (0.65 ml, 9.1 mmol) and Et3N (0.23 mL,
1.6 mmol) in toluene and heated at 60◦C with stirring
for 4 h. The mixture was then diluted with EtOAc
(10 mL), acidified with 2M H2SO4 (10 mL), washed
with water (2 × 10 mL) and brine (10 ml), and finally
the product was dried with MgSO4, filtered, and con-
centrated in vacuo. The product was recrystallized with
isopropanol and hexane to afford compound 2b (0.59 g,
1
1.99 mmol) (26% yield). H NMR (CDCl3): δ 1.16 (1H,
Chemical Synthesis
d), 2.34 (3H, s), 3.33 (5H, m), 3.87 (3H, s), 6.95 (2H, d),
7.94 (2H, m).
Chemical synthesis of ( )-2-methylthiomethyl-
4-(4-methylphenyl)-4-oxobutanoic acid (S-methyl-
esonarimod, compound 3a, Scheme 1) has been
previously described [13]. An analog of S-methyl-
esonarimod (i.e., ( )-4-(4-methoxyphenyl)-2-methyl-
sulfanylmethyl-4-oxobutyric acid (compound 3b)) was
also synthesized with the same procedure by a slight
modification of the starting material (Scheme 1) [13].
4-(4-Methoxyphenyl)-2-methylsulfanylmethyl-4-
oxobutyric Acid (3b)
Compound 2b (0.59 g, 1.99 mmol) was combined
with 2.1 M KOH (3 mL, 6.24 mmol) and tetrabutylam-
monium chloride (0.32 g, 1.0 mmol) at room temper-
ature. Methyliodide (0.14 mL, 2.22 mmol) was added
dropwise and stirred at room temperature for 30 min.
The mixture was washed with ether and a white solid
precipitate thus formed was removed. The remaining
material was acidified with 10 mL 2M HCl and re-
extracted with ether, washed with MgSO4, and con-
centrated in vacuo. The combined organic material was
chromatographed with ethyl acetate/hexane to afford
3b (4 mg, 4% yield). 1H NMR (CDCl3): δ 7.94 (m, 2H),
6.95–6.90 (m, 2H), 3.86 (s, 3H), 3.44–3.29 (m, 3H), 2.91
(dd, J = 5.4, 8.1 Hz, 1H), 2.75 (dd, J = 5.4, 7.5 Hz, 1H),
2.12 (s, 3H).
Compound 3a was fully characterized spectrally
and the data were consistent with that previously pub-
lished [13]. Compound 3b was characterized spectrally
and gave consistent 1H NMR (CDCl3): δ 7.94 (m, 2H),
6.95–6.90 (m, 2H), 3.86 (s, 3H), 3.44–3.29 (m, 3H), 2.91
(dd, 1H), 2.75 (dd, 1H), 2.12 (s, 3H). The purity of
2-[2-(4-Methoxyphenyl)-2-oxoethyl]-acrylic acid
(1b)
To a stirred solution of nitrobenzene (3.4 mL) in
CH2Cl2 (10 mL) was added AlCl3 (2.67 g, 20 mmol)
in portions, and then allowed to stir for 10 min after
which itaconic anhydride (1.12 g, 10 mmol) was added.
Anisole (1.23 mL, 11.3 mmol) was then added dropwise
over 10 min to afford a red coloration. After 1 h, the
reaction was stopped by the addition of cold 2.4 M
aqueous HCl (10 mL) at 4◦C. Hexane (10 mL) was then
added and the mixture was allowed to stir at 4◦C for 20
min. The resulting precipitate was washed with 2.4 M
aqueous HCl, then H2O, and finally with isopropanol
to give compound 1b (1.95 g, 8.83 mmol) (88% yield).
1H NMR (CDCl3): δ 3.874 (3H, s), 3.952 (2H, s), 5.785
(2H, s), 6.059 (1H, s), 6.942 (2H, d), 7.962 (2H, d).
SCHEME 1. Chemical synthesis of S-methyl-esonarimod and analog, as well as compounds 3a and 3b.
J Biochem Molecular Toxicology DOI 10:1002/jbt

Volume 21, Number 4, 2007
FUNCTION OF MOUSE FMO1, FMO3, FMO5
209
oxygenation of these substrates. The assay contained
50 mM sodium phosphate buffer (pH 8.4), 0.5 mM di-
ethylenetriaminepentaacetic acid (DETAPAC), 0.2 mM
NADPH, and 10–150 µg MBP-mFMO. After obtaining a
stable baseline, reactions were initiated by the addition
of substrate and monitored at 340 nm. Kinetic param-
eters for oxygenation were obtained from individual
experiments after the addition of decreasing amounts
of substrates to the sample cuvette. The final substrate
concentrations used were 100, 40, 20, 10, and 5 µM for
MMI, TMA, and S-methyl-esonarimod (compound 3a),
respectively, as well as 250, 125, 80, 62.5, and 50 µM for
compound 3b.
The N-oxygenation of 3-DPT, 5-DPT, and 8-DPT
was determined by HPLC analysis as previously de-
scribed [17]. Briefly, a standard incubation mixture con-
tained 50 mM potassium phosphate buffer (pH 8.4),
0.4 mM NADP+, 0.4 mM glucose-6-phosphate, 1 U
glucose-6-phosphate dehydrogenase, 0.2 mM DETA-
PAC, and 4–40 µg MBP-mFMO. Reactions were initi-
ated by the addition of 3-DPT, 5-DPT, or 8-DPT to a final
concentration of 200 µM, the highest substrate concen-
tration that provided minimum solubility in the assay
system. Enzyme reactions were incubated at 37◦C in
air with shaking for 10 to 20 min, and then stopped
by the addition of 4 volumes of cold dichloromethane.
After addition of 20 mg of Na2CO3, the incubations
were mixed and centrifuged to partition metabolites
and remaining substrate into the organic fraction. The
organic fraction was collected and evaporated under a
stream of argon. Metabolites and remaining substrate
were dissolved in methanol, mixed thoroughly, cen-
trifuged, and analyzed by the HPLC method described
above for 3-DPT, 5-DPT, 8-DPT, and their correspond-
ing N-oxides, respectively.
To determine the pH profile for the recombinant
mFMO1, mFMO3, and mFMO5, N-oxygenation of
8-DPT was assayed using the HPLC method described
above, except that the buffer used in each reaction mix
was replaced with 100 mM potassium phosphate buffer
at pH 7, 8, 9, 10, and 11. The mean standard devi-
ation of the velocity calculated for pH 8 was desig-
nated as 100% for each recombinant enzyme, and the
velocity under other pH conditions was normalized
accordingly.
To determine the thermal stability for recombi-
nant mFMO1, mFMO3, and mFMO5, N-oxygenation of
8-DPT was performed using the HPLC method de-
scribed above for the standard reaction mix. Prior to
the incubation with substrate, recombinant mFMO1,
mFMO3, and mFMO5 enzymes were incubated at 40◦C
for various periods in the presence or absence of cofac-
tors (0.4 mM NADP+, 0.4 mM glucose-6-phosphate,
1 IU glucose-6-phosphate dehydrogenase, and 0.2 mM
DETAPAC) and then transferred to ice. N-Oxygenation
SCHEME 2. Structure of phenothiazine analogs.
compounds 3a and 3b was determined by reverse-
phase HPLC (RP-HPLC) (i.e., Supelco HS F5 pentaflu-
orophenyl column, 4.6 mm × 250 mm, 5 µm), with UV
detection at 290 nm and a mobile-phase gradient of
solvent A (acetonitrile) and solvent B (acetonitrile/50
mM potassium phosphate buffer, pH 3, v/v 50:50).
Compounds 3a and 3b had retention times of 5.8 and
6.5 min, respectively. Compound 3b was also analyzed
with ESI-LCMS (Waters Micromass 2Q, using a 4.6 mm
×50 mm C18 RP-HPLC column and a 5-min solvent
gradient, from 5% to 95% CH3CN) showing a retention
time of 3.96 min and m/z for C13H17O4S [M + H]+ cal-
culated 269, found 269. A prominent fragment at m/z
251 for [MH − H2O]+ was also observed.
In addition, phenothiazine analogs 3-DPT, 5-DPT,
and 8-DPT (Scheme 2) were synthesized by a pre-
viously described method [14,15]. Synthesis of the
N-oxide of 3-DPT, 5-DPT, and 8-DPT was achieved
by biosynthesis using pig liver microsomes and the
products were characterized by ESI-MS [14]. The pu-
rity of 3-DPT, 5-DPT, 8-DPT, and their corresponding
N-oxides was determined by HPLC analysis. These
analyses were performed with a Hitachi HPLC sys-
tem (Hitachi L-7200 autosampler and L-6000 pump
interfaced to a Hitachi L-4200H UV detector). Chro-
matographic separation of analytes was performed
with an Axxi-Chrom normal phase analytical column
(250 mm × 4.6 mm, 5 µm silica) with a mobile-phase
MeOH/isopropanol/HClO4 (40:60:0.01%) for 3-DPT,
MeOH/isopropanol/HClO4 (55:45:0.01%) for 5-DPT,
and MeOH/isopropanol/HClO4 (55:45:0.02%) for 8-
DPT. The flow rate was set at 1.5 mL/min−1 and the
total run time was 7 min. The wavelength for UV de-
tection was set at 254 nm. The retention times for 3-DPT,
3-DPT N-oxide, 5-DPT, 5-DPT N-oxide, 8-DPT, and 8-
DPT N-oxide were 4.9, 3.9, 4.5, 3.5, 3.54, and 2.94 min,
respectively.
Enzyme Assays
The oxygenation of trimethylamine (TMA), mer-
captoimidazole (MMI), as well as compounds 3a,
3b, and 8-DPT was determined by monitoring the
oxidation of NADPH associated with the N- or S-
J Biochem Molecular Toxicology DOI 10:1002/jbt

210
ZHANG ET AL.
Volume 21, Number 4, 2007
of 8-DPT was determined with the HPLC assay as de-
scribed above. The mean velocity calculated for en-
zyme not treated at 40◦C was designated as 100% for
each recombinant enzyme, and the velocity for en-
zymes treated under other conditions was normalized
accordingly.
Data Analysis
FIGURE 1. SDS-PAGE of MBP-mFMO1, MBP-mFMO3, and MBP-
mFMO5. Fusion proteins MBP-mFMO1, MBP-mFMO3, and MBP-
mFMO5 were expressed in Escherichia coli and purified as described
in the Materials and Methods section. The purified proteins were ana-
lyzed on 10% SDS-PAGE and stained with Coommassie Blue.
The estimation of kinetic parameters was achieved
by examining the data from the incubation of at
least five different substrate concentrations with MBP-
mFMOs. Incubations were done in duplicate or trip-
licate. Data were analyzed with a nonlinear regres-
sion curve fit tool from Graphpad software (Graphpad
Prism, San Diego, CA) using a Michaelis-Menten
model. Data were presented as the mean standard
deviation on the basis of results from repeated separate
experiments (n = 2–4). Statistical analysis was done us-
ing Prism Graphpad software.
procedures using affinity chromatography on amylose
columns as described in the Methods and Materials sec-
tion. The final products were analyzed with SDS-PAGE
followed with Coommassie Blue staining (Figure 1). All
three recombinant fusion proteins showed molecular
weights of approximately 105 kD. This is in agreement
with the estimated molecular weight of the fusion pro-
tein with the addition of the MBP protein (45.6 kD).
Recombinant FMO5 migrated on PAGE slightly above
the other FMOs (see Figure 1). We believe this is due
to a slight difference in PAGE migration caused by pri-
mary sequence differences. The purification of all three
enzymes showed similar yield and purity, indicating
similar biophysical features of the three fusion proteins
under the preparation conditions used. The three en-
zymes also showed similar stability after freeze-thaw
cycles and storage at −80◦C (data not shown).
RESULTS
Cloning and Expression of Recombinant
MBP-mFMO1, MBP-mFMO3, and
MBP-mFMO5
We examined the contribution of each mouse FMO
isoform to selective functional substrate metabolism in
vitro. We characterized substrate oxygenation of spe-
cific mFMO enzymes by expressing the individual re-
combinant enzymes mFMO1, mFMO3, and mFMO5.
The cloned full-length mFMO1, mFMO3, and mFMO5
(as described in the Method and Materials section) were
expressed in the expression vector pMAL-2c. The MBP-
fusion system is an efficient expression system for
the production of stable and highly active FMO en-
zymes and has been used extensively to character-
ize human FMO3 and various polymorphic variants
[16,17]. The clones were re-sequenced and this con-
firmed that they were identical to the mouse FMO
gene reported in gene bank (i.e., mFMO1 (MMU87456),
mFMO3 (MMU87147), and mFMO5 (MMU90535), re-
spectively). The three enzymes showed 83.3%, 79.2%,
and 84.1% identity with their human counterparts, re-
spectively. To examine the biochemical properties for
the three FMO isoforms, these enzymes were expressed
as MBP-fusion proteins analogous to that previously
described [16,17]. This expression system is routinely
used in our laboratory for expression and purification
of human FMO enzymes and their genetic variants for
biochemical characterization. All three enzymes were
expressed in Escherichia coli with similar efficiencies.
The enzymes were purified with essentially identical
Comparison of Substrate Specificity for
Mouse FMOs Using Representative
Substrates
To examine the functional activities of the recom-
binant enzymes, standard substrates MMI and TMA
were used in the NADPH consumption assay (Table 1).
While mFMO1 and mFMO3 showed similar catalytic
activity toward MMI (i.e., Km = 43–53 µM and Vmax
=
187–204 nmol min−1 mg−1), mFMO5 showed no de-
tectable activity. TMA was shown to be a selective sub-
strate for mFMO3 with a Km in the micromolar range
similar to human FMO3. Both mFMO1 and mFMO5
showed no activity toward TMA under the testing con-
ditions used (Table 1).
To characterize the functional substrate activity
of mFMO5, S-methyl-esonarimod and an analog (i.e.,
compounds 3a and 3b) were chemically synthesized
(Scheme 1) as described in the Materials and Methods sec-
tion because S-methyl-esonarimod has been reported
to be a substrate for human FMO5 [18]. S-Methyl-
esonarimod (compound 3a) served as a substrate for
J Biochem Molecular Toxicology DOI 10:1002/jbt

Volume 21, Number 4, 2007
FUNCTION OF MOUSE FMO1, FMO3, FMO5
211
TABLE 1. Oxygenation of Mercaptoimidazole (MMI) and Trimethylamine (TMA) by Mouse FMO1, 3, and 5
mFMO1
mFMO3
mFMO5
MMI
TMA
Km (µM)
52.65 8.70
203.77 16.57
3.87
43.74 6.26
186.78 12.43
4.27
8.75 0.71
206.00 4.66
23.54
ND
Vmax (nmol min−1 mg−1
)
)
V
max/Km (10−3 min−1 mg−1
)
)
Km (µM)
ND
ND
Vmax (nmol min−1 mg−1
V
max/Km (10−3 min−1 mg−1
The oxygenation of MMI and TMA was determined by monitoring consumption of NADPH associated with the S-oxygenation of MMI and N-oxygenation of
TMA as described in the Materials and Methods section. Data are mean standard deviation (n = 4). ND, not detectable.
mFMO1 and mFMO5, but not for mFMO3 (Table 2).
This result was comparable to a previous report for
this compound with human FMOs [18]. While the
Km value for compound 3a was similar for mFMO1
and mFMO5 (3.82 and 3.23 µM, respectively), Vmax for
mFMO1 was more than 3-fold than that for mFMO5.
Compound 3b represents a close analog of S-methyl-
esonarimod, with replacement of a distal methyl group
with a methoxy group. However, such modification
dramatically changed the substrate interaction with the
FMO enzymes examined, as indicated by significant in-
creases of Km values (18-fold increase for mFMO1 and
approximately 30-fold or more increase for mFMO5).
While detectable NADPH consumption was observed
for mFMO3 and mFMO5 in the presence of compound
3b, due to limited solubility of compound 3b in the
reaction buffer, Km and Vmax values for mFMO3 and
mFMO5 toward compound 3b could not be precisely
determined. We therefore estimated Km to be more than
100 µM for mFMO3 and mFMO5 (Table 2).
zymes. Specific activity was calculated on the basis of
an HPLC assay quantifying formation of N-oxide of the
corresponding substrate. Specific activities toward the
three compounds were determined at a final substrate
concentration of 200 µM. The substrate concentration
was chosen on the basis of highest substrate solubility
in the assay system to provide maximum assay sensi-
tivity and enzyme substrate saturation. Incubation time
and enzyme amount input were controlled to ensure
substrate conversion within the linear range of each re-
action. For all three compounds examined (i.e., 3-DPT,
5-DPT, and 8-DPT), mFMO1 showed greater functional
activity than with mFMO3 and mFMO5 (Table 3). For
mFMO3, an increase in specific activity was observed
when the carbon chain linker length increased. A signif-
icant difference was observed for N-oxygenation of 5-
DPT and 3-DPT by mFMO3. N-Oxygenation of 8-DPT
and 5-DPT by mFMO3 was not significantly different.
For mFMO5, no activity was observed for 3-DPT. When
the substrate aliphatic chain length increased from 5 to
8, the specific activity also increased for mFMO5. How-
ever, mFMO5 gave the lowest specific activity for all
three substrates (Table 3).
In addition, we used a series of aliphatic tertiary
amines related to 5-DPT [16,17] to characterize the
mFMO enzymes (Scheme 1). Previously, we and others
[1,19] showed that changing the length of the aliphatic
carbon chain linker connecting the bulky phenoth-
iazine group and the tertiary amine markedly altered
the efficiency of N-oxygenation by human FMO en-
For 8-DPT, the common substrate investigated for
all three mFMO isoforms, Km and Vmax were deter-
mined using an NADPH consumption assay (Table 4).
The results indicated that Km for 8-DPT was in the mi-
cromolar range for mFMO1, 3, and 5. The differences
in Km are the largest contributors to the differences in
catalytic efficiency. Even though the catalytic efficiency
TABLE 2. S-Oxygenation of S-Methyl-esonarimod and Ana-
log (Compounds 3a and 3b) by Mouse FMO1, 3, and 5
mFMO1
mFMO3
mFMO5
TABLE 3. N-Oxygenation of Aliphatic Tertiary Amines by
3a Km (µM)
Vmax (nmol min−1 mg−1
max/Km
(10−3 min−1 mg−1
3b Km (µM)
Vmax (nmol min−1 mg−1
max/Km
(10−3 min−1 mg−1
3.82 2.11
43.16 4.64
11.30
ND
3.23 0.35
13.38 0.26
4.14
Mouse FMO1, 3, and 5
)
)
mFMO1
mFMO3
mFMO5
V
)
3-DPT
5-DPT
8-DPT
195.2 17.8
407.2 20.5
293.7 35.4
2.2 0.4
40.6 2.9
68.9 1.1
ND
1.5 0.5
11.8 0.3
71.05 12.91 >100
220.98 18.81
>100
NA
NA
V
3.11
The N-oxygenation of phenothiazine analogs 3-DPT, 5-DPT, and 8-DPT, was
determined by the HPLC procedure as described in the Materials and Methods
section. Data represents mean standard deviation values of specific activ-
ity (nmol tertiary amine N-oxygenation min−1 mg−1 of FMO) at a substrate
concentration of 200 µM. Each determination was done in triplicate. ND, not
detectable.
)
The oxygenation of S-methyl-esonarimod and analog (compounds 3a and
3b) was determined by monitoring NADPH consumption as described in the
Materials and Methods section. Data are mean standard deviation (n = 4). ND,
not detectable; NA, not available.
J Biochem Molecular Toxicology DOI 10:1002/jbt

212
ZHANG ET AL.
Volume 21, Number 4, 2007
TABLE 4. N-Oxygenation of 8-DPT by Mouse FMO1, 3,
and 5
I
mFMO1
mFMO3
mFMO5
Km (µM)
6.80 0.86 24.54 9.83 80.82 23.24
421.4 16.3 185.9 27.0 107.5 17.6
Vmax (nmol min−1 mg−1
)
V
max/Km (10−3 min−1 mg−1
)
62.0
7.6
1.3
The oxygenation of 8-DPT was determined by monitoring consumption
of NADPH associated with the N-oxygenation of 8-DPT as described in the
Materials and Methods section. Data are mean standard deviation (n = 4).
FIGURE 3. Thermal stability of recombinant mFMO1, mFMO3, and
mFMO5. Recombinant mFMO1, mFMO3, and mFMO5 were incu-
bated at 40◦C for the indicated length of time in the absence or
presence of NADPH (+NADPH 1 min) and then transferred to ice.
N-Oxygenation of 8-DPT was determined with an HPLC assay as de-
scribed in the Materials and Methods section. Mean velocity calculated
for enzyme not treated at 40◦C was designated as 100% for each re-
combinant enzyme and the velocity for enzymes treated under other
conditions was normalized accordingly. Mean normalized activity
and standard error (n = 2) are shown.
of mFMO5 for 8-DPT was significantly lower than that
of mFMO1 (48-fold) and mFMO3 (5.8-fold), 8-DPT was
efficiently oxygenated by all three mouse FMOs under
the same condition. This is the first report that a tertiary
amine serves as an FMO5 substrate.
To determine the effect of pH on the enzyme ac-
tivity of mFMOs, we examined the N-oxygenation
of 8-DPT by all three mouse FMO recombinant en-
zymes under similar conditions. As shown in Figure 2,
mFMO1 and mFMO3 displayed similar profiles with
an optimal pH in the range from 9 to 10 without signif-
icant differences between pH values 8, 9, and 10. For
mFMO1 and mFMO3, enzyme activity at pH 7 and 11
were significantly lower, at about 50% the level of activ-
ity at pH 8. For mFMO5, 8-DPT N-oxygenation activity
continued to increases from pH 7 to 11. In comparison
with the activity at pH 8, the activity of mFMO5 at pH
9, 10, and 11 were 110%, 119%, and 130%, respectively.
In addition, to compare the thermal stability of the
mouse FMOs, recombinant enzymes were incubated at
40◦C for 1, 2, and 5 min in the absence of NADPH,
chilled on ice, and then the remaining enzyme activity
was determined by quantifying N-oxygenation of 8-
DPT in the presence of NADPH. As shown in Figure 3,
mFMO1, mFMO3, and mFMO5 displayed similar ther-
mal sensitivity, with activity decreased to 76% to 83%
after 1 min preincubation at 40◦C, 61% to 71% after
2 min preincubation, and 41% to 55% after 5 min prein-
cubation. When the enzyme was preincubated in the
presence of NADPH, the enzyme was stable with more
than 85% activity remaining after 30-min treatment at
40◦C (data not shown). For mFMO1 and mFMO3, incu-
bation at 40◦C for 1 min in the presence of NADPH did
not decrease the enzyme activity significantly (94%–
97%). In addition, we reproducibly observed increased
activity (105%–110%) for mFMO5 after preincubation
at 40◦C for 1 min in comparison with mFMO5/NADPH
preincubated on ice.
DISCUSSION
In this study, we examined the enzyme activity pro-
file of mFMO1, mFMO3, and mFMO5, the three domi-
nant FMO enzymes shown to be present in male and/or
female mouse liver. Mouse FMO2 and FMO4 were not
included in this study because they were previously
FIGURE 2. Effect of pH on 8-DPT N-oxygenation activity by recombinant mFMO1, mFMO3, and mFMO5. N-Oxygenation of 8-DPT by
recombinant mFMO1, mFMO3, and mFMO5 was determined with an HPLC-based assay as described in the Materials and Methods section using
200 µM 8-DPT and 100 mM potassium phosphate buffer at the indicated pH. Mean velocity determined at pH 8 was designated as 100% for
each recombinant enzyme and the velocity under other pH conditions was normalized accordingly. Mean normalized activity and standard
error (n = 2 for mFMO1 and mFMO3, n = 4 for mFMO5) are shown.
J Biochem Molecular Toxicology DOI 10:1002/jbt

Volume 21, Number 4, 2007
FUNCTION OF MOUSE FMO1, FMO3, FMO5
213
reported to be present at extremely low levels in mouse
liver and reportedly did not contribute significantly to
mouse liver xenobiotic metabolism [5,7]. We utilized a
bacterial expression system previously used to study
human FMO enzymes for the expression and purifica-
tion of recombinant mFMO1, mFMO3, and mFMO5.
All three enzymes were expressed efficiently to gener-
ate 20–40 mg of purified recombinant enzymes from
each 6-L bacterial culture. The availability of purified
recombinant mFMO1, mFMO3, and mFMO5 allowed
us to characterize the enzyme profile for these enzymes
individually and compare among the isoforms to iden-
tify unique activities of the enzyme.
A series of different substrates were used to exam-
ine both the N- and S-oxygenation activities of these
enzymes. As presented previously, standard FMO1 and
FMO3 substrates including MMI and TMA are not oxy-
genated by mFMO5 [11,20]. In fact, one challenge for
FMO5 research is to identify a good substrate and de-
fine substrate structure–function relationships for this
enzyme [18,21]. Interestingly, compound 3a, S-methyl-
esonarimod, serves as an efficient substrate for mFMO5
and mFMO1, but not for mFMO3. The Km value for
mFMO1 and mFMO5 for compound 3a is similar (i.e.,
in the micromolar range), indicating good access to the
binding site domain of the enzymes by the substrate.
However, a small modification of compound 3a to re-
place the p-methyl group with a p-methoxyl group sig-
nificantly increased Km for both mFMO1 and mFMO5,
indicating that modification of the substrate structure
significantly altered the accessibility to the binding site,
possibly through interaction with a peripheral site be-
cause the modified group is on the opposite side of the
molecule 3b away from the oxygenated sulfur atom.
An alternative explanation is that the para-substituent
alters the extent of lactonization of 3a and 3b. It is pos-
sible that the lactonized version of 3a or 3b is the pre-
ferred substrate for FMO and not the carboxylic acid.
However, we did not investigate this possibility.
mFMO3 and mFMO5 was dependent on the aliphatic
carbon chain length for substrate access. For mFMO3,
an 18-fold difference in specific activity between 5-DPT
and 3-DPT was observed and indicated that the active
˚
site is likely more than 5–6 A below the surface of the
mouth of the enzyme, requiring a molecule with five
or more carbon atoms to deliver the tertiary amine to
the active site domain. For mFMO5, the active site may
˚
be even deeper, likely more than 6 A below the sur-
face because an 8-fold increase in specific activity was
observed between 8-DPT and 5-DPT.
Identification of a tertiary amine as an efficient sub-
strate for mFMO1, 3, and 5 allowed us to use this sub-
strate to compare in parallel the pH profile and ther-
mal stability of all three mouse FMO enzymes. The
pH profile for mFMO1 and mFMO3 is very similar
and give optimal activity over the pH range from 8
to 10. This suggests that mFMO1 and mFMO3 ion-
izable residues that are involved in the enzyme cat-
alytic mechanism (e.g., presumably the rate-limiting
steps of the catalytic cycle) for mFMO1 and mFMO3
are probably conserved. A different pH profile is ob-
served for mFMO5, with a continuous increase in ac-
tivity up to pH 11. It is possible that mFMO5 has a
similar catalytic mechanism and rate-limiting step as
that of mFMO1 and mFMO3, but mFMO5 uses differ-
ent ionizable residue(s) compared with mFMO1 and
mFMO3, possibly involving arginine(s), which has a
pKa above pH 11 (without considering neighboring
residues and micro environments). In addition, given
the much slower overall turnover rate for mFMO5, it
is possible that the rate-limiting step for mFMO5 dur-
ing the catalytic cycle is different from that of mFMO1
and mFMO3, and therefore involves a different set of
residues favoring an extremely alkaline condition.
Data from heat treatment of mFMO enzymes fol-
lowed with enzyme activity determinations give clues
that mFMO5 interacts with NADPH differently from
that of mFMO1 and mFMO3, because we consistently
observe a higher activity for mFMO5 if it is allowed
to be preincubated with NADPH at 40◦C (Figure 3).
While release of NADP+ is believed to be the rate-
limiting step for FMO3 and FMO1, it is possible that
access to NADPH is an even slower step for mFMO5. It
was previously believed that access to the stable FMO
4α-hydroperoxyflavin intermediate is largely responsi-
ble for differences in substrate specificity for FMO en-
zymes [23]. However, it has been suggested that FMO5
has low or nondetectable catalytic efficiency for certain
substrates because of lack of the 4α-hydroperoxyflavin
intermediate [24], stability of FMO5 may function in
a similar manner to that of para-hydroxybenzoate hy-
droxylase [24,25], where the reduction of the FAD in
the enzyme by NADPH is dependent on the binding
We used a series of phenothiazine analogs with
varying lengths of aliphatic carbon chains attached to
the tertiary amine of the bulky phenothiazine group
to investigate the requirement for substrate access and
enzyme activity. Similar to previous reports on FMO1
enzymes from other species [22], mFMO1 accepts all
three substrates giving functional activities in the 200–
400 nmol min−1 mg−1 range, indicating that substrate
access for mFMO1 is relatively easy and not directly
affected by the length of the carbon chain. It has been
estimated that the hydroperoxyflavin in hFMO1 lies
˚
no more that 5 A below the surface of amino acid
side chains that control access of nucleophiles into the
substrate site [22], and we believe this also applies
to mFMO1. On the other hand, the activity of both
J Biochem Molecular Toxicology DOI 10:1002/jbt

214
ZHANG ET AL.
Volume 21, Number 4, 2007
6. Zhang J, Cashman JR. Quantitative analysis of FMO
gene mRNA levels in human tissues. Drug Metab Dis-
pos 2006;34(1):19–26.
7. Falls JG, Blake BL, Cao Y, Levi PE, Hodgson E. Gen-
der differences in hepatic expression of flavin-containing
monooxygenase isoforms (FMO1, FMO3, and FMO5) in
mice. J Biochem Toxicol 1995;10(3):171–177.
8. Koukouritaki SB, Simpson P, Yeung CK, Rettie AE, Hines
RN. Human hepatic flavin-containing monooxygenases 1
(FMO1) and 3 (FMO3) developmental expression. Pediatr
Res 2002;51(2):236–243.
9. Cherrington NJ, Cao Y, Cherrington JW, Rose RL,
Hodgson E. Physiological factors affecting protein ex-
pression of flavin-containing monooxygenases 1, 3 and
5. Xenobiotica 1998;28(7):673–682.
10. Falls JG, Cherrington NJ, Clements KM, Philpot RM,
Levi PE, Rose RL, Hodgson E. Molecular cloning, se-
quencing, and expression in Escherichia coli of mouse
flavin-containing monooxygenase 3 (FMO3): compari-
son with the human isoform. Arch Biochem Biophys
1997;347(1):9–18.
11. Cherrington NJ, Falls JG, Rose RL, Clements KM, Philpot
RM, Levi PE, Hodgson E. Molecular cloning, sequence,
and expression of mouse flavin-containing monooxyge-
nases 1 and 5 (FMO1 and FMO5). J Biochem Mol Toxicol
1998;12(4):205–212.
of para-hydroxybenzoate to the enzyme [26,27]. FMO5
was reported to catalyze N-oxygenation of benzy-
lamine [28] and methimazole [20] with very low ef-
ficiency. N-Oxygenation of n-octylamine has been re-
ported to reach an activity of 22.8, 38, and 26.1 nmol
NADPH oxidized min−1 nmol−1 for rabbit FMO5,
guinea pig FMO5 and hFMO5, respectively [20], illus-
trating the ability of FMO5 to catalyze primary amine
N-oxygenation. Experimental support obtained from
spectral characterization of the enzyme with stopped
flow analysis under anaerobic and aerobic conditions
will be essential for further understanding of this point.
This is the first report that aliphatic tertiary
amines are substrates for FMO5. Future exploration
of substrate–structure relationship for these enzymes
will also enlighten the overall understanding of the
functional significance of this family of enzymes. The
recombinant enzyme system for mFMO enzymes may
provide essential tools to follow these observations and
investigate enzyme catalytic mechanisms in conjunc-
tion with mutagenesis studies.
12. Karoly ED, Rose RL. Sequencing, expression, and
characterization of cDNA expressed flavin-containing
monooxygenase 2 from mouse. J Biochem Mol Toxicol
2001;15(6): 300–308.
ACKNOWLEDGMENTS
13. Noguchi T, Onodera A, Tomisawa K, Yokomori S. A prac-
tical procedure for the synthesis of esonarimod, (R,S)-
2-acetylthiomethyl-4-(4-methylphenyl)-4-oxobutanoic
acid, an antirheumatic agent (Part 1). Chem Pharm Bull
(Tokyo) 2002;50:1407–1412.
14. Lomri N, Yang Z, Cashman JR. Expression in Escherichia
coli of the flavin-containing monooxygenase D (form
II) from adult human liver: determination of a distinct
tertiary amine substrate specificity. Chem Res Toxicol
1993;6(4):425–429.
We thank Nayaz Ahmed and Synthia Chang for
technical assistance. We also thank Dr. Senait Ghirmai
for determination of molecule 3D coordinates using
SymApps software and Dr. Randy Rose (North Car-
olina State University) for providing mouse FMO
clones in PJL-2 vector.
15. Nagata T, Williams DE, Ziegler DM. Substrate specifici-
ties of rabbit lung and porcine liver flavin-containing
monooxygenases: differences due to substrate size. Chem
Res Toxicol 1990;3(4):372–376.
16. Brunelle A, Bi YA, Lin J, Russell B, Luy L, Berkman
C, Cashman J. Characterization of two human flavin-
containing monooxygenase (form 3) enzymes expressed
in Escherichia coli as maltose binding protein fusions.
Drug Metab Dispos 1997;25(8):1001–1007.
17. Lattard V, Zhang J, Tran Q, Furnes B, Schlenk D, Cashman
JR. Two new polymorphisms of the FMO3 gene in
Caucasian and African-American populations: compar-
ative genetic and functional studies. Drug Metab Dispos
2003;31(7):854–860.
REFERENCES
1. Krueger SK, Williams DE. Mammalian flavin-containing
monooxygenases: structure/function, genetic polymor-
phisms and role in drug metabolism. Pharmacol Ther
2005;106(3):357–387.
2. Cashman JR, Zhang J. Human flavin-containing
monooxygenases. Annu Rev Pharmacol Toxicol
2006;46:65–100.
3. Hines RN, Hopp KA, Franco J, Saeian K, Begun
FP. Alternative processing of the human FMO6 gene
renders transcripts incapable of encoding
a func-
18. Ohmi N, Yoshida H, Endo H, Hasegawa M, Akimoto
M, Higuchi S. S-oxidation of S-methyl-esonarimod by
flavin-containing monooxygenases in human liver mi-
crosomes. Xenobiotica 2003;33(12):1221–1231.
19. Lomri N, Yang Z, Cashman JR, Regio- and stereoselec-
tive oxygenations by adult human liver flavin-containing
monooxygenase 3. Comparison with forms 1 and 2.
Chem Res Toxicol 1993;6(6):800–807.
tional flavin-containing monooxygenase. Mol Pharmacol
2002;62(2):320–325.
4. Hernandez D, Janmohamed A, Chandan P, Phillips IR,
Shephard EA. Organization and evolution of the flavin-
containing monooxygenase genes of human and mouse:
identification of novel gene and pseudogene clusters.
Pharmacogenetics 2004;14(2):117–130.
5. Janmohamed A, Hernandez D, Phillips IR, Shephard
EA. Cell-, tissue-, sex- and developmental stage-specific
expression of mouse flavin-containing monooxygenases
(FMOs). Biochem Pharmacol 2004;68(1):73–83.
20. Overby LH, Buckpitt AR, Lawton MP, Atta-Asafo-
Adjei E, Schulze J, Philpot RM. Characterization of
flavin-containing monooxygenase 5 (FMO5) cloned from
J Biochem Molecular Toxicology DOI 10:1002/jbt

Volume 21, Number 4, 2007
FUNCTION OF MOUSE FMO1, FMO3, FMO5
215
human and guinea pig: evidence that the unique cat-
alytic properties of FMO5 are not confined to the rab-
bit ortholog. Arch Biochem Biophys 1995;317(1):275–
284.
25. Schreuder HA, Prick PA, Wierenga RK, Vriend G,
Wilson KS, Hol WG, Drenth J. Crystal structure of
the p-hydroxybenzoate hydroxylase-substrate complex
˚
refined at 1.9 A resolution. Analysis of the enzyme-
21. Overby LH, Carver GC, Philpot RM. Quantitation and ki-
netic properties of hepatic microsomal and recombinant
flavin-containing monooxygenases 3 and 5 from humans.
Chem Biol Interact 1997;106(1):29–45.
22. Kim YM, Ziegler DM. Size limits of thiocarbamides
accepted as substrates by human flavin-containing
monooxygenase 1. Drug Metab Dispos 2000;28(8):1003–
1006.
23. Ziegler DM. Flavin-containing monooxygenases: cat-
alytic mechanism and substrate specificities. Drug Metab
Rev 1988;19(1):1–32.
24. Ziegler DM. An overview of the mechanism, substrate
specificities, and structure of FMOs. Drug Metab Rev
2002;34(3):503–511.
substrate and enzyme-product complexes. J Mol Biol
1989;208(4):679–696.
26. Frederick KK, Palfey BA. Kinetics of proton-linked flavin
conformational changes in p-hydroxybenzoate hydroxy-
lase. Biochemistry 2005;44(40):13304–13314.
27. Ballou DP, Entsch B, Cole LJ. Dynamics involved in catal-
ysis by single-component and two-component flavin-
dependent aromatic hydroxylases. Biochem Bio Phys Res
Commun 2005;338(1):590–598.
28. Lang DH, Rettie AE. In vitro evaluation of potential
in vivo probes for human flavin-containing monooxy-
genase (FMO): metabolism of benzydamine and caf-
feine by FMO and P450 isoforms. Br J Clin Pharmacol
2000;50(4):311–314.
J Biochem Molecular Toxicology DOI 10:1002/jbt